inviroranaffiat Protection  Laboratory
         Cincira-ati OH 45268
cr'A-ouO/8-79-006
March 1979
Research arc) Davelopment
      PA Manual for
      nics Analysis
Chromatography-
Mass Spectrometry

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology.  Elimination of traditional  grouping was consciously
planned to foster technology transfer and a maximum interface in related fields:
The nine series are:

      1.  Environmental Health Effects Research
      2.  Environmental Protection Technology
      3.  Ecological Research
      4.  Environmental Monitoring
      5.  Socioeconomic Environmental Studies
      6.  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

This report has been assigned to the "SPECIAL" REPORTS series. This series is
reserved for reports targeted to meet the technical information needs of specific
user groups. The series includes problem-oriented reports, research application
reports, and executive summary documents. Examples include state-of-the-art
analyses, technology assessments, design manuals, user manuals, and reports
on the results of major research and development efforts.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia  22161.

-------
TECHNICAL REPORT DATA
(Please read Inunctions on the reverse before completing) •
1. REPORT NO. 2.
EPA-600/8-79-006
4. TITLE AND SUBTITLE
An EPA Manual for Organics Analysis- Using Gas
Chroma tography - Mass Spectrometry
7. AUTHOR(S)
William L. Budde and James W. Eichelberger
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Monitoring and Support Laboratory
Office of Research and Development
U. S. Environmental Protection Agency
Cincinnati , Ohio 45268
12. SPONSORING AGENCY NAME AND ADDRESS ""'"'
Same as above
3. RECIPIENT'S ACCESSION-NO. _ '.
DPOQ71&4
5. RE*ORi^6Xl£ / / J. W F
March 1979 issuing date
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NC
10. PROGRAM ELEMENT NO.
1BD884
11. CONTRACT/GRANT NO.
In-House Report
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/06
15. SUPPLEMENTARY NOTES " "'•' 	 •"
16. ABSTRACT
  This  procedural  manual  defines the areas of applicability of gas chromatography-
  mass  spectrometry in environmental analysis.  The manual includes sample prepara-
  tion  methods  specifically adapted to this measurement technique, data processing anc
  interpretation methods,  quality control  procedures, quantitative analytical tech-
  niques,  utility  software requirement's, preventive maintenance recommendations,
  systematic  trouble shooting methods, and a selected bibliography.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Environmental Surveys Organic
Environments Compounds
Spectroscopic Analysis
Organic Chemistry
Pollution
Ecology
Mass Spectroscopy
13. DISTRIBUTION STATEMENT • '" - '
Release to Public
b. IDENTIFIERS/OPEN ENDED TERMS
.. 'i<
19. SECURITY CLASS (This Report}
Unclassified
20. SECURITY CLASS { This page I
Unclassified
c. COSATI Field/Group
1402
0703
21
22. PRICE .-'•-' '-
EPA Form 2220-1 (9-73)

-------
                                                   EPA-600/ 8-79-006
                                                   March 1979
      AN EPA MANUAL FOR ORGANICS ANALYSIS
USING GAS CHROMATOGRAPHY-MASS SPECTROMETRY
                              by


                         William L. Budde


                              and


                       James W. Eichelberger
             Environmental Monitoring and Support Laboratory
                  Office of Research and Development
                  U.S. Environmental Protection Agency
                       Cincinnati, Ohio 45268

-------
                                DISCLAIMER

  This manual has been reviewed-by the Environmental Monitoring and Support Labpratpry-
Cincinnati, U.S. Environmental Protection Agency, and apprpved for publiGatjpn. M,er»tipn pf
trade names pr cpmmercial products does not constitute endorsement or recommendation for
use.

-------
                                  FOREWORD

  Environmental measurements are required to determine the quality of ambient waters and
the character of waste effluents. The Environmental Monitoring and Support  Laboratory-
Cincinnati conducts research to:

  •    Develop and evaluate techniques to measure  the  presence and concentration of
       physical, chemical, and radiological pollutants in water, wastewater, bottom sediments,
       and solid waste.

  •    Investigate methods for  the concentration, recovery,  and identification of viruses,
       bacteria and other microbiological organisms in water; and to determine the responses
       of aquatic organisms to water quality.

  •    Develop  and  operate an  Agency-wide  quality  assurance  program to  assure
       standardization and quality control of systems for monitoring water and wastewater.

  This publication of the Environmental Monitoring  and Support Laboratory-Cincinnati
entitled:  An EPA  Manual  For  Organics  Analysis  Using  Gas  Chromatography-Mass
Spectrometry, is  the result of an Agency-wide effort to document broad  spectrum methods
needed for the analysis of organic compounds in  environmental samples. Broad spectrum
methods  are required for situations where the nature of the  organic compounds that may be
present is unknown. Federal agencies, states, municipalities,  universities, private  laboratories,
and industry should find this manual of assistance in  monitoring and controlling organic
pollution in the environment.
                                                 Dwight  G. Ballinger
                                                 Director
                                                 Environmental Monitoring and
                                                   Support Laboratory-Cincinnati

-------
                                   ABSTRACT

  This procedural  manual defines the areas of applicability of gas chromatography-mass
spectrometry in environmental analysis. The manual includes sample preparation methods
specifically adapted  to  this measurement technique,  data processing  and interpretation
methods,  quality  control  procedures, quantitative analytical techniques,  utility software
requirements, preventive maintenance recommendations, systematic trouble shooting methods,
and a selected bibliography.
                                         IV

-------
                           TABLE OF CONTENTS


Chapter  1          Introduction	...;,...	..-.	  1

Chapter  2          GC/MS Operations and Quality Assurance	 4

         2.1         Mass Spectrometer Start-up	„..,	 4

         2.2         Resolution and Sensitivity Adjustments	,	 5
                    DC  Zero  and Balance...........	 5
                    Tune»up With Perfluorotributylamine	'.	 7

         2.3         Data System  Start-up (PPP-S)	 9
                    Loading the Disk Operating System Bootstrap
                      Program	 9

         2.4         System Zero  Adjustment	12

         2.5         Mass Scale Calibration	12
                    Calibration Diagnostic Program	13

         2.6         Quality Control	'.....'....„....	:...'	14

         2.7         Suggested Operating  Parameters for Sample Analysis	18
                    Gas Chromatographic Considerations	18
                    Data Acquisition	19
                    Integration Time as  a Function of Signal Strength
                      (IFSS)	21

         2.8         Mass Spectrometer Shut-Down	23
                    Short  Term	23
                    Long Term	.;.	23

Chapter  3          GC/MS Sample Preparations	24

         3.1         Water Samples	25
                    Direct Aqueous  Injection  Including Concentration
                      Techniques	25
                    Inert Gas  Purging and Trapping	26
                      Equipment	26
                      Discussion	;	30
                      Sample Collection  and Preservation	33
                      Procedure	v	35

-------
                 TABLE OF  CONTENTS (CONTINUED)


                    Qualitative Headspace Analysis	36
                    Extraction with -a Low Boiling Solvent	 37
                      Reagents  and Equipment	K.	40
                      Procedure	40
                    Extraction with a High  Boiling  Solvent	42
                    Adsorption  with Porous Polymers	43
                      Reagents :and Equipment	47
                      Procedure	47

         3.2         Air Samples	48
                    Direct Air  Injection	48
                    Adsorption  with Porous Polymers	49
                    Filtration with Glass Fiber Filters	50

         3.3         Sediment Samples	51

         3.4         Fatty Tissue Samples	52

         3.5         Chemical Derivatization Options	53

Chapter  4          Data output	55

         4.1         The OUTP/CRT Software	,	56
                    The Total Ion Current  Profile	56
                      Cathode Ray Tube Terminal	56
                      Plotter	57
                    Mass Spectra Histograms	57
                      Cathode  Ray Tube Terminal	•.	57
                      Plotter	58
                    Printed Mass Spectral Data	58
                    The Extracted Ion Current Profile	61
                    Queued  Output	..62

         4.2         The MSSOUT Output Software	62
                    The Total Ion Current  Profile	64
                    Mass Spectra Histograms	67
                    Printed Mass Spectral Data	70
                    The Extracted Ion Current Profile	'.	70
                    Queued  Output	72

         4.3         Interactive  Graphics Oriented Output Software
                      (IGOOS)	72
                    Total and Extracted Ion Current  Profiles	73
                    Mass Spectra Histograms	74
                    Extended Memory Output Software	75
                                      VI

-------
                TABLE  OF CONTENTS  (CONTINUED)


Chapter  5          Compound Identification	  77

         5.1         Interpretation from Theory	  77

         5.2         Empirical Spectrum Matching	  77
                    Searching Printed Data Bases	  77
                    Remote Computerized Search Systems	  78
                    Minicomputer  Search Systems	  80
                      Procedure (PDP-8)	,;	  81

         5.3         Quality Control in Compound Identification	  83
                    Evaluation of Blanks and Reagent Blanks	  83
                    Supporting Experiments	  84
                    Requirements for Pure Standards	  88
                    Similarity Index  Interpretations	  88
                    The Quality Index	:	  88

Chapter  6          Advanced Analytical Techniques	  91

         6.1         Selected  Ion Monitoring (SIM)	  91
                    Definitions of Terms	  92
                    Selection of Masses	  94
                    SIM Using the Control Program	  97
                    Specialized SIM  Programs	.	  99

         6.2     .    Quantitative Measurements	 100
                    Measurements with QNTSET/QNTATE	 103
                    Measurements with QUAN3	 104
                    Measurements with MSSOUT	 105

         6.3         Open  Tubular Columns .„.	,-.	 106
                    Operating Variables...	 106

         6.4         Chemical lonization (CI)	 107
                    Methane Chemical lonization Reference Compound	 108
                    Principles of Chemical lonization	 108

         6.5         Accurate Mass Measurements	110

Chapter  7          Auxiliary Software	114

         7.1         Real Time Operating  System	114
                    Tape  Initialization Programs	114
                    File Copy, Duplication,  and Save  Programs	115
                    File Delete and  System  Tape Dump Programs	116
                    Program  to Change Machine Language Instructions	116
                    Program  to Start the OS-8 System	,	117
                                      vu

-------
                  TABLE OF  CONTENTS (CONTINUED)
         7.2         The OS-8 Operating System	117
                    Tape Format Programs	 117
                    Program to List Files on the OS-8 System:	118
                    Programs to Build Real Time Systems from  Tape Dumps       118
                    Loading New Programs  on OS-8 or the Real Time System	118
                      Inserting a Paper Tape Handler in OS-8	119
                      Loading a Paper Tape File on OS-8	119
                      Loading a Dectape File  on OS-8	119
                    Loading a Binary Program on the Real Time System Disk	119
                    Program to Duplicate OS-8	120
                    Program to Start the Real Time System	120

Chapter  8          Preventive Maintenance	121

         8.1         The Mass  Spectrometer  Vacuum System	121
                    Mechanical Pumps	121
                    Diffusion Pumps	122
                    Ion Gauge, Manifold,  and Inlet Systems	122

         8.2         The Mass  Spectrometer  Electronics	122
                    Cooling  Fans and Dust  Filters	122
                    Vacuum Tubes	123

         8.3         The Gas Chromatograph	;	.....;.. 123

         8.4         The Data System	123
                    Software	123
                    Cooling  Fans and Dust  Filters	123
                    Lubrication of Mechanical Parts	123
                    Magnetic Storage Devices	124
                    Diagnostic Computer Programs	124

Chapter  9          Trouble  Shooting	125

         9.1         Mass Spectrometer Problems	125
                    Vacuum System	125
                    Ion Gauge	126
                    Heaters	126

         9.2         Data System Problems	126
                    Computer Diagnostic Programs.	126
                    Dectape Diagnostic Programs	129
                    Disk Diagnostic Program	130

         9.3         System  Zero Adjustment Problems	'.	131
                                       vui

-------
                  TABLE OF  CONTENTS (CONTINUED)


         9.4         Mass Scale Calibration Problems	,	131
                    Interface ~ Spectrometer Gain Alignment	132
                    Loss of Ions	132
                    Interface Diagnostic  Programs	133
                      ADZERO  Program	134
                    Peak Profile  Diagnostic Program	134

         9.5         Quality Control Problems	136
                    Poor Overall Sensitivity	136
                    Poor High Mass Sensitivity	136
                    Incorrect  Mass Assignments or Poor Resolution
                      of Adjacent Ions	136
                    High Background	137

Chapter  10         Selected Bibliography	,	138

         10.1        General Reference Books	138

         10.2        Printed Mass Spectra Collections	139

         10.3        Selected Review and Primary Journal Articles	139
                    Mass Spectrometry	139
                    GC/MS	139
                    Direct  Aqueous Injection	139
                    Inert Gas Purging and Trapping	139
                    Qualitative Headspace  Analysis.	140
                    Adsorption with  Porous Polymers	140
                    Air Samples	'.	140
                    Fatty Tissue Samples	140
                    Chemical  Derivatization	140
                    Selected Ion  Monitoring	141
                    Open Tubular Columns.::.,	141
                    Chemical  lonization	141
                                       IX

-------
                                 LIST OF FIGURES


Figure   1.1        Flow of trace organics analysis in a laboratory with
                    GC/MS capability	 3

Figure   2.1        Rough  functional schematic showing components  used  in
                    the tune-up of a Finnigan  model 1015 in the  high  mass range.... 6

Figure   2.2        Peaks at masses 219 and 220	10

Figure   2.3        Peaks at masses 69 and  70	10

Figure   2.4        Peaks at masses 502 and 503	10

Figure   2.5        Calibration diagnostic program plot	15

Figure   2.6        The IFSS algorithm flow chart	22

Figure   3.1        A purging device with a 5 ml sample capacity	:	27

Figure   3.2        The trap assembly for a purging device	29

Figure   3.3        Recoveries of selected  compounds as  a function of  purge
                    gas flow rate	'.	32

Figure   3.4        Chromatogram of organohalides	34

Figure   4.1        The mass spectrum of a GC/MS reference compound,
                    decafluorotriphenylphosphine	59

Figure   4.2        The mass spectrum of a GC/MS reference compound,
                    decafluorotriphenylphosphine, with plot options	60

Figure   4.3        The extracted ion current  profile for  mass 149 and the
                    corresponding total ion current profile	63

Figure   4.4        A total ion current profile generated  with
                    MSSOUT	65

Figure   4.5        A mass spectrum  plot generated  with MSSOUT	66

Figure   4.6        A mass spectrum  plot generated  with MSSOUT	68

-------
                      LIST OF FIGURES (CONTINUED)
Figure   4.7        An extracted ioircurrent profile generated with
                    MSSOUT	  71

Figure   5.1        The extracted ion current profiles  for mass  171
                    from a sample and the corresponding reagent  blank	  85

Figure   5.2        The extracted ion current profiles  for mass  149
                    from a sample and the corresponding reagent  blank	  86

Figure   5.3        The total ion current profiles from a sample and the
                    corresponding reagent .blank..,..	'.	  87

Figure   5.4        The mass spectrum of an unidentified compound and the
                    mass spectrum of chloropicrin	  89

Figure   6.1        A schematic diagram of continuous  repetitive  measurement
                    of spectra,  a TICP,  and an EICP	  93

Figure   6.2        A schematic diagram of continuous  repetitive  measurement
                    of spectra,  SIM, and two SICP's	  95

Figure   6.3        A TICP, SICP, and EICP from the chromatography of seven
                    chlorobiphenyl isomers...,	  96

Figure   6.4        Some calculated .chlorine-bromine  isotope distribution
                    patterns	'.	  98

Figure   6.5        Chemical ionization (methane) mass spectra of DFTPP as a
                    function  of ion source temperature on the  Finnigan  model 4000
                    GC/MS  system	109

Figure   6.6        Comparison of the El and  CI mass spectra of
                    octadecane	:	112
                                         XI

-------
                              LIST OF TABLES
Table     2.1         Functions and corresponding  potentiometer designations
                     for Finnigan mass spectrometers	  8

Table     2.2         Octal-binary equivalents of the  real time system disk
                     bootstrap program	11

Table     2.3         Sample report from the calibration  diagnostic
                     program	,	14

Table     2.4         Suggested GC columns and conditions	17

Table     2.5         Decafluorotriphenylphosphine  key ions and ion
                     abundance criteria	17

Table     2.6         Representative GC columns for  sample  extracts	18

Table     2.7         Representative GC columns for  direct aqueous
                     injection	19

Table     3.1         Recoveries of spiked organics from water after  a simple
                     distillation	25

Table     3.2         Recoveries of organics by gas purging and
                     trapping	31

Table     3.3         Effect of residual chlorine  on concentrations of
                     chlorinated methanes  in drinking water at 4C	33

Table     3.4         Recoveries of organics by solvent extraction with
                     methylene chloride	38
                                        xu

-------
                    LIST OF TABLES  (CONTINUED)
Table    3.5
Table    5.1
Table    6.1
Table    9.1
Recoveries of organics by adsorption on porous
polymers	!".	 44

User options available with the mass spectral search
system	'.	 78

The effect of the accuracy of mass  measurements  and
atom constraints on the  number of acceptable  compositions	 113
Computer diagnostic programs for the PDP-8/E
or M	'.;	
                                                                                127
Table    9.2        Switch register selection of Dectape unit  in the
                    random exerciser test	129
                                      xin

-------
               LIST OF PDP-8 DATA SYSTEM PROGRAMS

no file          bootstrap loader	  2.3
EXEC          system monitor; control and IFSS  prompts	  2.3
INITSS        .file with system configuration and  constants	  2.3
ZERO          analog baseline adjustment	  2.4
CALIBR       calibrates mass scale  only	  2.5
FC1000         calibration  driver file	  2.5
M1000         calibration  driver file	  2.5
CDIAGN      calibration  diagnostics	  2.5
MASDEF      shifts calibration masses...	  2.5
CONTRL      data aquisition including IFSS	  2.7
RTCON        real time display on  a CRT	  2.7
CONDIS       real time display on  a CRT	  2.7
CONDI1       real time display on  a CRT	  2.7
TWIN          data acquisition on a second disk  drive	  2.7
EXMOD2      data acquisition on a second disk  drive	  2.7
PERCAL      retrieves and loads operating parameters	  2.7
RSTORE       detailed sample information	  2.7
LIST           lists files on disk 0,  TD8E Dectape, or
                 overlays  MAGLST	  4
LSTDSK       lists files on TC08  Dectape, and disks
                 0,1,2, or 3	  4
MAGLST      list for 9 track tape	  4
EXMOD1      plotter/printer dialogue	.'...:.....:..:...	  4.1
PLOUT        plotter driver	:„.....	.....'...„....	  4.1
PLOT          plotter driver	  4.1
PLTP2         plotter driver	  4.1
CUEXEC      queued plotter output	  4.1
RGC           generates TICP data	  4.1
TYPE          prints mass spectral  data	  4.1
BACKGD      background subtract  overlay	  4.1
CRT           main CRT program	  4.1
CRTRGC      overlay for CRT TICP	  4.1
CRTSP         overlay for CRT histogram	  4.1
CBCKGD      overlay for CRT background subtract	  4.1
DSPLAY       overlay for CRT	  4.1
MSSOUT      main program	  4.2
MOUTP1      overlay for MSSOUT	  4.2
MOUTP2      overlay for MSSOUT	  4.2
MOUTP4      overlay for MSSOUT	  4.2
MOUTP5      overlay for MSSOUT	  4.2
MOUTP6      overlay for MSSOUT	  4.2
INCCHR      overlay for MSSOUT	  4.2
SS1S           scratch file for MSSOUT	  4.2
BCLRGC      main TICP/EICP program	  4.3
RGCOV1       overlay for main TICP/EICP program	  4.3
RGCOV2 ,     overlay for main TICP/EICP program	  4.3
                                     xiv

-------
       LIST OF PDP-8 DATA SYSTEM PROGRAMS  (CONTINUED)


CRTPLT      main histogram plot program.;..,	.-::	  4.3
PLOVL1       overlay for CRTPLT	  4.3
PLOVL2       overlay for CRTPLT	  4.3
SPRRGC      maximized mass version of BCLRGC in  16K	;•	 r.4.3
SPRPLT       maximized mass histograms in 16K	  4.3
SPRVL1       overlay for above	  4.3
BRVCCP      spectra abbreviation  for MSSS	  5.2
MDIREK     transmission to MSSS-Tektronix interface	  5.2
BMATCH     spectra abbreviation  for PDP-8 search	  5.2
BMNBVT     overlay for BMATCH	.'...':	  5.2
BMN8K       PDP-8 search in 8K	  5.2
BMN16K      PDP-8 search in 16K	  5.2
PRNTSP      print spectrum in  PDP-8 library	  5.2
CRTSIM      special SIM  data acquisition	:	  6.1
PLTSIM       special SIM  data acquisition	  6.1
CRSMPT      special SIM  output	  6.1
PTSMPT      special SIM  output	„	  6.1
CONVRT     converts SIM datafiles to standard format	  6.1
SIMEXC      expanded CRTSIM in 12K	  6.1
CRTSM2      overlay for above	  6.2
QNTSET      quantitative analysis	:	  6.2
QNTATE     quantitative analysis	  6.2
QUAN3       quantitative analysis	„.'	  6.2
NEWTAP     Dectape initialization (two versions)	  7.1
NEWMAG    9 track tape initialization	  7.1
COPY        copies file between disk 0 or  1  and TD8E  Dectape
                 and overlays MAGCPY	  7.1
COPY        copies files between  disk 0  or 1 and TC08
                 Dectape	  7.1
MAGCPY     9 track tape file copy	  7.1
CPYDSK     system disk  duplicate	  7.1
SAVE        save spectra (two  versions)	  7.1
DELE        deletes file on  disk 0 or TD8E  Dectape	  7.1
DELE        deletes file on  disk 0 or TC08 Dectape	  7.1
BCDELE     deletes groups  of files on disk 0	  7.1
SYSDMP     dump disk to Dectape (two versions)	  7.1
MAGDMP    dump disk to 9 track tape	;	  7.1
PATCH       change machine language instructions	  7.1
OSDGO       starts OS-8 system	'.	  7.1
DTFRMT     TC08 Dectape format	  7.2
TDFRMT     TD8E Dectape format	  7.2
PIP           OS-8  file transfer  and  file list	  7.2
TC8RES      restore real  time system disk from TCQ8 Dectape	  7.2
TD8RES      restore real  time system disk from TD8E Dectape	.-.	 -1.2
MAGRES     restore real  time system disk from 9  track  tape	  7.2
                                     XV

-------
       LIST OF PDP-8 DATA SYSTEM PROGRAMS (CONTINUED)


BUILD        OS-8 system  build	,	  7.2
ABSLDR      OS-8 loader	..'.	  7.2
FOTP         'OS-8 file  transfer	  7.2
LOD150       real time  system loader.	  7.2
DSKCPY      OS-8 disk duplicate	  7.2
MSDGO       starts real time system from  OS-8	  7.2
INST1         PDP-8  instruction test	  9.2
INST2         PDP-8  instruction test	,	  9.2
XADDR       extended  memory address test	  9.2
XCHKBD     extended  core memory test	  9.2
ADDER       ?PDP-8  adder test	,	  9.2
EAEST1       EAE instruction test	  9.2
EAEST2       EAE instruction test	  9.2
EAEXME     EAE-memory test	  9.2
TC08RX       TC08 Dectape exerciser	  9.2
TD8DIA       TD8E Dectape exerciser	  9.2
DDIAGN     disk format and diagnostics	  9.2
MSTEST       old MS 'interface test	,	  9.4
NEW150       new RIB interface test	  9.4
ADZERO     interface  adjustment	  9.4
MSSCAN      peak profile diagnostic	  9.4
                                    XVI

-------
DATA SYSTEM RULES  AND CONVENTIONS  USED  IN  THIS  MANUAL

1.  Each line of response to a computer program entered from a keyboard by a user must be terminated
by pressing the return.key. This is not indicated in sample dialogue.

2.  User responses in sample computer dialogue are underlined.

3.  A user response to a computer program of YES may be abbreviated Y; a user response to a computer
program of NO may be abbreviated N; in most programs an N may be replaced by merely pressing the
return key. In a few programs a NO response will not be accepted and a negative input must be made by
pressing the return key. These exceptions are indicated in the sample dialogue.

4.  Most user entered commands may be abbreviated with the first four characters.

5.  A user may interrupt computer prompts at the keyboard \f he understands the correct response
before the output is complete.                    :

6.  A user may usually return to the SELECT MODE: prompt by holding down the CTRL key while
pressing the L key. This is abbreviated CTRL/L frequently.

7.  The compound perfluorotributylamine will be abbreviated PFTBA. This is also sometimes known by
the 3M company trade designation FC-43.
                                         xvu

-------
                              CONTRIBUTORS

  Special recognition is required for several individuals who made significant contributions to
this manual. These individuals are:

       Ann Alford
       Environmental Research Laboratory, EPA
       Athens, GA   30601

       Thomas A. Bellar
       Environmental Monitoring and  Support Laboratory
       Cincinnati, OH   45268

       Harvey W. Boyle
       National  Enforcement Investigation  Center, EPA
       Denver,  CO   80225

       Robert D. Kleopfer
       Region VII,  EPA
       Kansas City,  KS   66115

       D. Craig Shew
       Robert S. Kerr Environmental  Research Laboratory,  EPA
       Ada,  OK   74820
                                     xvni

-------
                           ACKNOWLEDGEMENT

  The authors wish to acknowledge the many others who participated in the development or
review of the manual.

       Herbert J. Brass
       Division of Technical Support
       Office of Water Supply,  EPA
       Cincinnati, OH   45268

       Mike  Carter
       Environmental  Research  Laboratory, :£PA
       Athens, GA    30601

       Emile Coleman
       Health Effects  Research  Laboratory, EPA
       Cincinnati, OH   45268

       B. F.  Dudenbostel
       Region II, EPA
       Edison, NJ    08817

       Jack D. Henion
       New York State College of Veterinary  Medicine
       Ithaca,  NY    14853

       Doug Kuehl
       Environmental  Research  Laboratory, EPA
       Duluth, MN   55804

       Robert Lingg
       Health Effects  Research  Laboratory, EPA
       Cincinnati,  OH   45268

       William Loy
       Region IV,  EPA
       Athens, GA   30601

       Robert Melton
       Health Effects  Research  Laboratory, EPA
       Cincinnati,  OH  45268
                                      xix

-------
           ACKNOWLEDGEMENT (CONTINUED)


William  Middleton
Environmental Monitoring and Support Laboratory,  EPA
Cincinnati,  OH    45268

David J. Munch
Division of Technical Support
Office of Water  Supply.,  EPA
Cincinnati,  OH    45268

Curt Norwood
Environmental Research Laboratory, EPA
Narragansett,  RI   02882

Frank Onuska
Canada Center for Inland waters,
Burlington, Ontario,  Canada

James F. Ryan
Health Effects Research Laboratory, EPA
Research Triangle Park,  NC  27711

Jack M. Teuschler
Environmental Monitoring and Support Laboratory,  EPA
Cincinnati,  OH    45268

Richard J. Thompson
Environmental Monitoring and Support Laboratory,  EPA
Research Triangle Park,  NC  27711

Oilman Veith
Environmental Research Laboratory, EPA
Duluth, MN    55804

Jacqueline  Waitts
Bowne Time Sharing
Washington, DC   20036

Virgil L. Warren
Region VI, EPA
Houston, TX   77036
                              xx

-------
          ACKNOWLEDGEMENT (CONTINUED)


Ron Webb
Environmental Research Laboratory, EPA
Athens, GA   30601

Karen Zieverink
Southwestern  Ohio Regional Computer Center
Cincinnati,  OH   45221
                            XXI

-------
                                       CHAPTER 1
                                   INTRODUCTION
  The  purpose of this manual is to organize in one
place a variety of information that is needed to identify
and measure organic compounds in environmental
samples using computerized gas chromatography-mass
spectrometry (GC/MS). The  application of a  mass
spectrometer as a universal, yet extremely selective and
sensitive   detector  in  gas  chromatography   has
revolutionized the identification  and measurement of
organic compounds. The efficient use of GC/MS was
made possible by the development of the relatively low
cost  digital  minicomputer  during the  late 1960's.
During the subsequent years a large number of books,
review  articles,  and journal articles appeared which
describe   the   basic  concepts   of  the   GC/MS
instrumentation, and applications to a wide variety of
organic analytical problems. A- number of these books
and articles are cited in the Bibliography in Chapter 10
of this manual. However in the environmental field the
specific detailed information that is required to utilize
this powerful tool is  scattered  among  a variety of
sources, and  some vital information  is not written down
at all.  The  emphasis of  this manual is on the
experimental details of the GC/MS methodology.
  There  are three  distinctly different  applications
where  environmental measurements by  GC/MS are •
being or will be used widely. First, there is the broad
spectrum (BS) organics analysis which has the goal of
seeking a broad spectrum picture of whatever is present
in a sample as a major or minor component. This kind
of analysis is not guided by a predetermined list of
compounds  to be measured.  The  BS approach was
made  possible by the development of tools such as
computerized  GC/MS,   and  this  approach   is
emphasized  in this manual. The BS approach is  most
appropriate  for samples that are  likely to contain
unexpected components.
  Secondly,  the application  of GC/MS appears  more
cost  effective  than conventional  GC  methods for
routine monitoring of relatively large numbers of target
compounds,  e.g., 40-60 or more. Some possible reasons
for this are presented in the next paragraph. Thirdly,
the application of GC/MS in  real time selected  ion
monitoring is widely used in the biomedical and other
fields, and is  recognized as one of the most accurate,
precise,  and   selective  tools  known  to  analytical
chemistry.    This   application   requires   detailed
knowledge of the compound being measured, and its
mass spectrum. This application is not emphasized in
this manual  except for one  section in Chapter 6.
However it is expected  that selected ion monitoring
methods  will  become  more  important   in   the
environmental field in the future.
   There are   a  number of contrasts  between  the
GC/MS methods in this  manual and  conventional
methods, which are often  based on chromatographic
'•techniques, of determining organic compounds. Both
types  of methods  are  needed,  and  each  has  an
appropriate area of application. Conventional methods
for  organic   compounds  are  detailed  chemical-
instrumental  procedures  designed  to  separate  the
compound of interest (target compound, TC) from all
known interferences, and to measure its concentration
with   relatively  inexpensive,  simple,  easy-to-use
instrumentation.  The  separation   and   isolation
procedure is often called "clean-up" and is, in effect, a
large  part  of the  qualitative  analysis.  The  gas
chromatography-electron capture detector procedures
for chlorinated hydrocarbon pesticides are examples of
this approach. The clean-up for conventional detectors
is often quite rigorous and complex to assure that all
potential  interferences  are  eliminated.  Similarly,
because conventional methods often  rely heavily on
retention indices for the identification of compounds,
rigorous control of operational conditions such as flow
rates and temperature  is required. Quality control
measurements, such as retention index standards, must
be made at relatively frequent intervals to assure that
conditions  are  well  controlled.  By contrast,  the
GC/MS sample  preparation methods  described in
Chapter 3 were designed to be  as simple as possible.
                                                 1

-------
Rigorous clean-up is often not required because the
mass   spectrometer  output   provides   adequate
information, in most cases,  to  reliably identify the
compound. Rigorous clean-up is, of course, counter
productive in the BS approach. The GC/MS methods
do not depend on rigorous attention to operational
conditions and frequent measurements of standards for
determination of retention indices. The application of
GC/MS for routine monitoring of target compounds is
most appropriate for samples that are likely to contain
unexpected components.
  Another contrast is  that conventional methods
usually  require  a,  variety  of  detectors  to  permit
measurements  of a  broad range of compound types.
This is because conventional detectors are designed for
some  selectivity to  further minimize  the  effects of
interfering  substances. The mass  spectrometer is  a
universal detector which precludes interferences by the
very nature of its output. For monitoring a wide range
of compound types,  it appears that the additional cost
of computerized GC/MS instrumentation may  be
offset by the need to acquire and maintain only one
detector, and the use of simpler, less  costly  sample
preparation and quality control procedures.
  Several chapters of the manual (Chapters 2,4,7,8,
and 9) describe in detail the operation and maintenance
of the Finnigan Corporation's models 1015 and 3000
series GC/MS systems with datasystems based on a
Digital   Equipment Corporation   model   PDP-8
minicomputer.  This information was  included and
emphasized  because the Environmental  Protection
Agency (EPA) owned more than thirty Finnigan 1015
and 3000 series GC/MS systems and a similar number
of PDP-8  GC/MS  datasystems.  These models  of
spectrometers and datasystems are in widespread use,
and the general  user community  should  find these
specialized  chapters of interest and  value. Users of
other types of GC/MS systems will find information of
general  interest  in section  2.6,  Quality Control;
Chapter 3,  GC/MS Sample Preparation; Chapter 5,
Compound   Identification;  Chapter  6,  Advanced
Analytical   Techniques;  and,  Chapter   10,   the
Bibilography.
  Figure 1.1 shows a flow chart for the processing of a
sample with the BS approach. Several blocks in Figure
1.1 are identified with specific chapters in this manual.
In the BS approach, the conventional broad band GC
detector, e.g., flame ionization, may be an important
screening  tool  for the busy  GC/MS  laboratory.
Alternatively, if the current sample load is not high, the
GC/MS may itself perform the initial screening.  The
GC/MS Operations and Quality Assurance Chapter
(Chapter 2) is identified with the block in  Figure 1.1
labeled integer accuracy GC/MS. The importance of
reagent blanks and spectra interpretation is discussed
in  Chapter 5.  Chapter 6,  Advanced  Analytical
Techniques, includes information on methods that may
be  used to obtain  more  information  when an
identification cannot be made with integer accuracy
GC/MS.  The   Advanced   Analytical   Techniques
chapter also includes information  about quantitative
measurements with GC/MS.
  Contributions to and reviews of the draft chapters of
this manual were made by the participants in the EPA
mass spectrometer users group between January,  1975
and October, 1978. This users group was  organized by
the authors in the spring of 1972  to  promote the
exchange  of  technical  information among  EPA
laboratories  using computerized GC/MS to identify
and measure organic environmental pollutants.

-------
STANDARD3-RT
  CI/QC/MS
 EXACT
   GC-IR
                  YES
                                  SAMPLE
                                  PREPARED
                                  FOR8C
                                  INTEGER
                                 ACCURACY
                                   OC/MS
SPECTRA
INTERPRETATION
\
f
YES ,

                                   MS OR
                                   SIMPLE
                                 DITECTOR
                               IDENTIFICATION
                                     &
                                MEASUREMENT
                                                    YES

                                                     NO
                                                                      3C
                                                                     SIMPLE
                                                                   DETECTOR
                                                                       NO
                               •*<&
Figure 1.1   Flow of trace organics analysis in a laboratory with GC/MS capability.

-------
                                                         2
      GC/MS  OPERATIONS  AND  QUALITY ASSURANCE
  The purpose of this chapter is to provide the basic
information required for the successful operation of the
computerized gas chromatograph/mass spectrometer
(GC/MS). The chapter is  oriented  to  the Finnigan
Corporation's models 1015 and 3000 series quadrupole
mass spectrometers with  data  systems that  utilize
Digital   Equipment  Corporation   model  PDP-8
computers. However, certain portions of this chapter,
such as the quality control procedure in Section 2.6, are
readily applied to other mass spectrometer systems.
The  quality  control  procedure  is  of  particular
importance because one can be reasonably assured  of
obtaining  good  quality  mass  spectra  when  the
performance  criteria  outlined  in  that  section are
attained.
  For initial start-up or whenever the vacuum system
is at atmospheric pressure (after  changing a separator,
cleaning rods, etc.) one should follow the procedures
beginning with  Section  2.1  and continue through
resolution and sensitivity adjustments, data system
start-up,   system  zero  .adjustments,   mass  scale
calibration,  quality control, and  sample  analysis.
However for normal day-to-day operation one would
begin  with  data  system start-up followed by system
zero adjustment, mass scale calibration, and the quality
control  tests.  Resolution  adjustments  are  usually
required only occasionally as indicated by the quality
control   evaluation.   Some  suggested   operating
parameters are given in Section 2.7. These parameters
should be  considered as  guidelines  and may not
represent optimum  operating conditions  for  every
sample which may be encountered in an environmental
analysis laboratory.
  Recommended  shut-down procedures  for both
short-term  (overnight,   weekends)   and  long-term
situations are given. Proper shut-down procedures are
important to reduce or prevent deterioration or damage
to  certain  expensive  components   of   the  mass
spectrometer.
2.1   MASS SPECTROMETER
START-UP

  When the vacuum system is at atmospheric pressure
(after changing a separator, cleaning rods, etc.) and the
system  is  completely  shut down,  the  following
procedure is recommended for restarting the system.

  1.    Apply power  with  the vacuum  controller
       circuit breaker. This supplies line power to the
       mass  spectrometer,  the gas chromatograph,
       and the vacuum system.
  2.    Make sure that the system  to be evacuated is
       closed to the atmosphere (the  GC end of the
       separator must be capped and  the direct inlet
       system closed).
  3.    Turn  on the  forepump (analyzer and  batch
       inlet if present) switch and separator pump
       switch. These switches activate the mechanical
       vacuum pumps  used to  obtain a "rough"
     .  vacuum.
  4.    Check pressures on the forepump pressure
       meter. This meter monitors the pressure down
       to 0.01 torr at each forepump. If no leaks are
       present,  readings  below   0.1  torr  will  be
       obtained within one minute.
  5.    After  pressures  of 0.1  torr  or below  are
       obtained, turn the vacuum controller function
       switch to the override position. This turns  on
       the diffusion pump  heaters. Since the override
       position bypasses the protection circuits, the
       mass   spectrometer  should   not   be   left
       unattended during this time.
  6.    After waiting  approximately 45 minutes for
       the diffusion pumps to heat, attempt to light
       the   Bayard-Alpert  pressure   gauge   by
       depressing  the start button on the  vacuum
       controller.  Depressing  this button  for  1-2

-------
       seconds  will turn  on the ion gauge filament
       when the system pressure is below 10"J torr. If
       the filament fails to light, wait five minutes and
       repeat. The life of the ion gauge filament is
       shortened by high pressure operation. If the
       filament stays on at  10"1 torr or greater, turn
       the  function switch  to off, then back to the
       override position and wait for lower pressure
       to be obtained.
         The zero control on the vacuum controller
       module permits electrical zeroing of the high
       vacuum meter. This can be done while the ion
       gauge filament is off and the meter switch is in
       the override position. Adjust the zero control
       until the meter reads 1 X ICT'torr.
  7.    After a stable vacuum operation  of 10"' torr or
       lower is achieved, move the vacuum controller
       function  switch from  override to  protect.
       Normally several hours are required to obtain
       pressures of 10"7  torr or  lower. The  mass
       spectrometer can be safely  utilized, however,
       with an analyzer vacuum of 10"' torr or less.
  8.    Adjust temperatures of the analyzer manifold,
       transfer  line,  batch  inlet,  and interface as
       desired with the heater controls on the vacuum
       controller.  These  temperatures  should  be
       monitored  occasionally with   an  external
       pyrometer of known accuracy. In general, the
     .  analyzer manifold should  not be heated;- and'
       the  separator (interface) temperature should
       be a few degrees higher than the  maximum GC
       column   temperature  to   be   utilized.  A
       temperature of about 250C  is adequate for the
       transfer  line.  The  batch  inlet temperature
       should   be  sufficiently   high to   permit
       appreciable vaporization of any liquid to be
       introduced via  the batch inlet  (generally
       100-200C).
  9.    Turn on the  line power switch on  RF/DC
       power supply (1015 only).
  10.   Turn  power on for high  voltage  supply to
       electron multiplier (1015 only).
2.2   RESOLUTION AND SENSITIVITY
ADJUSTMENTS

  Resolution and sensitivity adjustments are necessary
with a quadrupole mass spectrometer after the system
has been shut down  for cleaning  or replacing the
ionizer, mass filter, or the electron multiplier. They are
also necessary for other reasons such as after changing
 a component on a printed circuit (PC) board in a power
 supply or in the  DC/RF generator.  Before the mass
 spectrometer resolution and sensitivity can be adjusted,
 .the electronic components must be balanced. The DC
 zero and balance procedure that follows was written for
 the Finnigan model 1015. For the Finnigan 3000 series,
 the corresponding procedure is described  as a ROD
 DC BALANCE ADJUSTMENT on page 1-16 of the
 3000 series systems maintenance manual. Figure 2.1 is
 a functional schematic of the components used in  the
 model 1015 tune-up.

 DC ZERO AND BALANCE
    1.    Set the mass spectrometer as follows:
        RF/DC Power Supply (bottom panel):
          Line Power -  On
          Standby/Operate - Operate
        RF/DC Generator:
          Ionizer:  Off
          Electron Multiplier Voltage:  Off
          Scan Time:  0.1 sec
          Scan Time Vernier:   Fully clockwise
          Scan Mode:  Repetitive
          First Mass Selector:  000
          Last Mass Selector:   000
          Scope Sensitivity:  0.5V./CM
    2.    Set the normal/test switch on the rear of the
        oscilloscope to "test" position.
, • '. 3.  i  Zero .the oscilloscope by adjusting the trace to
        the center of the  screen. Keep it zeroed during
        the  following   procedure  by  momentarily
        disconnecting the oscilloscope probe from the
        test point being  observed and readjusting the
        trace.
    4.    Set the mass range selector on the model 1015
        RF/DC  generator to a detent position. This
        removes  the  tuning capacitor from the  RF
        generator thereby curtailing the output of RF
        and places PC-5 circuitry under the mid and
        high mass mode of control. The mass selector
        is a five position switch having  one dentent
        position  between the high and medium range
        and another between  the low and medium
        range.
    5.    Turn R-81 on  PC-5  (behind the removable
        panel  on   the   RF/DC  generator)  fully
        counterclockwise then turn clockwise 10  full
        turns. This places R-81 approximately in the
        center of its adjustment range when amplifier
        A4 is balanced.
    6.    Connect  the  oscilloscope test probe (xlO) to
        TP-3 on PC-5.

-------
-15
    R81 LOW MASS COMP
                          R77 FINE RESOLUTION
                          R74 COARSE RESOLUTION
                                                                                 PC 5
._ 	 _ _ _ l»r f*
(
i R179 MASS SELECT
-B-^V?A Tf (
^- R180 HIGH MASS
COMP.
-15



1.R185
~DC Z

"?-'


POS ROD
RF DRIVER
-^_ R202 BALANCE
	
NEG ROD
RF DR9VER
(A5)
TP4
T ,

POWER
DRIVER
946
i r j
POS ROD T
TP6
NEG ROD f

| R189
i DC ZERO
1
•-(
1
r
1 	 .TP20
AX j, f!
R73
FEEDBACK
AJUST
RAMP
r*"- nmevEK
(A3)


1
' \


*%*»
RAWJP
FEEDBAC
(SALS)
?
RAMP
POWER
DRIVER
(Q44)

iW
~ ~ " "" PC 6
K,K C2
<*1
' ~~\
1
RF
I*" GENERATOR
807's
I-


~₯ C21=BALANCI
J^ C22=TUNE
*
' POS ROD RF
E
NEG ROD RF
L J
  Figure 2.1   Rough functional schematic showing components used in the tune-up of a Finnigan model
             1015inthehigh mass range.

-------
   7.    Turn R-185 fully clockwise until  a clicking
        sound  is  observed; then  turn it counter-
        clockwise  until  the oscilloscope trace shifts
        downward  about  1.5  cm (15  volts).  This
        balances amplifier A4 and allows R-81 to have
        the  maximum range of adjustment. Connect
        the oscilloscope test probe to TP-4.
   8.    Turn   R-189   clockwise;  then   turn  it
        counterclockwise until the oscilloscope display
        shifts upward about 1.5  cm (15 volts).  This
        balances amplifier A5 (see schematic of PC-5).
        The 15 volt jumps observed on the oscilloscope
        are  the saturation  voltages of amplifiers A4
        and A5. R-185 and R-189 influence the range
        of R-81.  If one  lacks  sufficient range while
        adjusting  the low  mass resolution,  R-185
        and/or R-189 may not be adjusted properly.
   9.    Switch the mass range selector to 50-750.
   10.   Set  first mass selector to 000 and last  mass
        selector to 750.
   11.   Connect one end of the rod balance  network to
        TP-5 and the other end to TP-6 on PC-5 (rod
        balance network = two 100K,  1% resistors
        connected in parallel to a  test point).
   12.   Connect test probe to the junction between the
        two resistors.
   13.   Set      oscilloscope     sensitivity      to
        20 millivolts/cm.
   14.   Adjust  R-202 until the trace on the scope is
        horizontal; then adjust  R-189 until the trace is
        at 0 volts. If R-189 needs to be changed more
        than 2 turns, go back to R-202 and repeat
        adjustment.
   15.   Set  the  normal/test  switch  on rear  of
        oscilloscope to the normal position.
  After the electronic components have been balanced,
one should proceed with the resolution adjustment for
50-750 mass range.
detailed  procedure  is  somewhat different and  not
described  in   this  manual.   Potentiometer  (pot)
designations (e.g., R-81) refer to'the model 1015 only.
These  pot  reference numbers  are marked on PC-5
which  is behind the removable panel of the RF/DC
generator. In parentheses after the 1015 pot number is
the pot function name.  In  the  Finnigan  3000 series
spectrometers these names are marked on the RF/DC
controller board in  the RF/DC generator. However,
the 3000 series  pot numbers, which are referenced in
the manuals and on schematics, do not correspond to
the pot numbers used in the model 1015. Table 2.1
shows   function  names  and   corresponding   pot
designations for the model 1015, 3000 series, and model
4000 spectrometers. Similar considerations apply to
test points.
  The EPA tune-up procedure  is essentially the same
as that given in Finnigan manuals except that certain
modifications are made to facilitate generating spectra
similar  to  that  obtained  with  magnetic  deflection
instruments. These modifications are as follows:

   1.    There is no optimization of signal intensity or
        resolution  at  masses  16-18.  This  spectral
        region is rarely,  if ever, important  in organic
        analysis.

  2.    The 10% valley definition of  resolution is
 '   -.'•  employed for tune-up at masses 502-503. This
       improves relative signal intensity in this region
       of the spectrum.
   3.    The magnet position, ion energy, and extractor
        voltages are adjusted  for maximum signal
        intensity and peak shape at masses 502-503.

  4.    Guidelines for relative abundance criteria are
        included to assist  in achieving performance
        consistent    with    magnetic    deflection
        spectrometers.

  5.    Electron energy is adjusted for optimum signal
        intensity across the 50-750 amu range.
TUNE-UP WITH
PERFLUOROTRIBUTYLAMINE
  The tune-up procedure that follows was written
specifically for the  Finnigan  model  1015  and 3000
series spectrometers. The same basic  philosophy also
applies  to the model 4000  spectrometer, but  the
  Resolution adjustment requires the introduction of
perfluorotributylamine  (PFTBA)  into  the  mass
spectrometer. PFTBA  has known mass  peaks and
adjacent  13C  peaks  up to approximately  mass  600.
Significant   ions  are   at '  the   following,  mass
numbers:  69-70,   131-132,    219-220,   264-265,
414-415, 502-503, and 614-615.

-------
  Table 2.1.   Functions and Corresponding Potentiometer Designations for Finnigan Mass Spectrometers
    Function
Model
 1015
   Models
  3000-3300
 Model
 4000
    Low mass compensator
    Coarse  resolution
    High mass threshold
    High mass resolution
    Fine resolution control
    Balance for amplifier A4
    Balance for amplifier A5
    DC Balance
 R-81
 R-74
 R-179
 R-180
 R-77
 R-185
 R-189
 R-202
R-35
R-30
R-29
R-28
Resolution  (front panel)
not applicable
not applicable
R-42
R-45
R-65

R-73
(front panel)
  Using the 50-750 setting on the 1015, one is capable
of calibrating  and scanning from  -10 to 750  amu.
Therefore, for  normal  operations only the high  mass
range is adjusted.
  1.   Set the ion source, preamp, oscilloscope, and
       DC/RF generator controls as follows:
              3.
system pressure at least a factor of five. If no
spectrum is observed, one may continue with
step 3, but serious problems are apparent and
trouble shooting may be required.
Turn   R-179  (high  mass  threshold)  fully
counterclockwise  and  R-180  (high   mass
              Control
   Ionizer. (Filament)
   lonization (Emission) Current
   Electron Energy
   Extractor
   Ion Energy
   Lens
   Preamp Sensitivity
   Preamp Filter
   Scope  Sensitivity
   Electron Multiplier
   Mass  Range  Selector
   Scan Time
   Scan Time Vernier
   Record Switch
   Scan (control) Mode
   First Mass
   Last Mass
       Model  1015
  On
  0.5 milliamp
  70 ev
  8v
  5v
  -  lOOv
  10"' amps/v
  > 3000 amu/sec
  0.5 v/cm
  Cu/Be, 3000v
  High
  0.1 sec
  Fully  Clockwise
  Off
  Repetitive
  0
  750
              3000 Series
             On
             0.5 milliamp
             70  ev
             4v
             2v
             40v
             10"* amps/v
              > 3000 amu/sec
             0.5 v/cm
             Continuous  dynode,
             Not Applicable
             GC
             Not Applicable
             Not Applicable
             On
             0
             750
       1600v
   2.    Inject 0.5 ul of PFTBA into either the batch
        inlet, if you have one, or the variable leak inlet
        system and admit a small amount into the
        mass spectrometer. Look at scope to observe a
        mass spectrum. If no  spectrum is observed,
        double check all settings given  in step 1 and
        admit additional  PFTBA to  increase  the
                   resolution) fully clockwise.
                   Adjust the first mass and last mass controls so
                   that the peaks at 219 and 220 amu appear on
                   the oscilloscope as in Figure 2.2.
                   Adjust R-77 (resolution-front  panel) for the
                   resolution between  219 and 220 as in Figure
                   2.2.  If  R-77,  the fine  adjustment,  lacks

-------
     sufficient range, set  it to the middle of its
     range, and  adjust R-74 (coarse resolution) to
     obtain roughly what is illustrated in Figure
     2.2.  Then  go  back to  R-77  for  the  fine
     adjustment. The resolution observed on the
     scope is approximate and not necessarily the
     actual resolution. In  general, if resolution on
     the scope  is not quite  baseline,  the  actual
     resolution will be adequate.
6.    Set the first mass and last mass controls so that
     the peaks at 69 and 70 amu appear on the
     oscilloscope as in Figure 2.3.
7.    Adjust R-81 (low mass compensator) for the
     resolution  as  illustrated  in  Figure  2.3.
     Trimpots   R-77 and  R-81  are interactive.
     Therefore  steps 4 thru  7 must be  repeated
     several times before resolution is obtained at
     both 69-70 and 219-220 amu.
8.    Decrease the scope sensitivity until mass 69 is
     observed completely on the scope. Adjust first
     mass and last mass controls until mass 69 is on
     the left side of scope and mass 219 is on right
     side   of   scope.  Mass   219  should   be
     approximately 35 percent as abundant as mass
     69.
9.    Set the first mass control so that the peak at
     mass  264 is on the extreme left side  of the
     scope. Set the last mass control to 750.
10.  Turn R-180  (high  mass, resolution)  fully
     counterclockwise.
11.  Slowly  turn R-179  (high mass  threshold)
     clockwise until a large peak  appears on the
     right side of the scope. Continue turning R-179
     until peak 264 on left side of scope just starts to
     increase.  This  allows  R-180  to influence
     masses down to mass 264.
12.  Turn R-180 clockwise until  the  large peak
     diminishes and changes  into  spectral  lines.
     This completes a rough tune of R-180.
13.  Set first mass and last mass controls to display
     the masses 502 and 503. Adjust R-180 so that
     resolution is as in Figure 2.4. This completes a
     fine tune of R-180,
 14.     At this  point,  with peaks  502  and  503
      displayed  on  the scope, adjust  the  source
      magnet  position,  extractor voltage, and ion
      energy for maximum peak height  and best
      peak'shape. In general the  collector current
      should be 70-90% of the emission current.
 15.   Adjust first mass and last mass  controls to
      display  the entire PFTBA spectrum. Then
      adjust the  electron energy for maximum signal
       intensity. The voltage observed on the meter
       should then be above 50 and below 80 volts.
  16.   The mass spectrometer should now be ready to
       calibrate. There  is no advantage .to checking
       resolution with the strip chart recorder.
2.3   DATA SYSTEM
START-UP (PDP-8)

  This section describes in detail the start-up of a data
system based on  a Digital Equipment Corporation
model  PDP-8 computer.  Power  is applied to  the
computer by turning the console key to POWER. The
POWER LOCK position also activates the system,  but
the front panel is disabled and the user cannot see the
blinking lights. Peripheral  devices (plotter, keyboard
terminal, etc.) may require separate activation of power
switches at some installations.
  The software operating system programs are stored
on a real time system disk or on a real time system
Dectape on older systems. The disk should be inserted
into disk drive 0 (or the tape mounted  on unit 1) and
the LOAD/RUN switch  set to the  RUN position.
About two minutes  are  required  for  the yellow
READY light to illuminate, indicating the disk is up to
speed arid ready to transfer data.
 ' the HALT and SING STEP switches must be up to
run a program.  The operating software  system is
started by running the bootstrap loader program. For
systems with a hardware bootstrap loader,  the loader
program is initiated by  pressing the  SW switch
(adjacent to the POWER key)  down, then up. If a
hardware  bootstrap  loader is  present,  the system
prompt SELECT MODE:  will be displayed on  the
console printer or cathode ray tube  (CRT).  If no
hardware  bootstrap  loader is  included  in the  data
system, the loader program must be loaded manually.
The  instructions  for loading and running the disk
operating system bootstrap loader are in the following
section.

LOADING THE DISK OPERATING
SYSTEM  BOOTSTRAP PROGRAM
   The disk  bootstrap program is  loaded  via  the
computer switch  register. The switch  register has 12
switches, which are either up or  down. A switch in the
up position represents a "1" and a switch in the down
position represents a"0". The switch register is divided
into four fields of three switches each.

-------
 Figure 2.2 Peaks at masses 219 and 220.
 Figure 2.3  Peaks at masses 69 and 70.
Figure 2.4  Peaks at masses 502 and 503.
                 10

-------
   Instructions and address locations used to load the
 disk bootstrap are given in four digit octal numbers to
 represent the position  of the 12 switches.  Actual
 loading  procedure,  however,  requires   that  the
 instructions and  address locations be keyed in with
 binary numbers.
   The binary equivalents of the octal numbers used to
 load the disk bootstrap are given in Table 2.2. Load the
 disk bootstrap in the computer as follows:

   1.    Set the starting address, octal value 7737, in
        the switch register.
   2.    Depress the LOAD ADDR key.
   3.    Set the first instruction (7200) in the switch
        register.
   4.    Depress the DEP key.
   5.    Set the next instruction (6502) in the switch
        register.
   6.    Depress the DEP key.
   7.    Repeat steps 5 and 6 until all instructions have
        been deposited.
   8.    Set the computer switch register to 7737.
   9.    Depress the computer LOAD ADDR key.
   10.   Place the rotary switch in the MD position.
   11.   Depress the computer EXAM key.
  12.   Check that  the  first  instruction,  7200,  is
       displayed in the computer's lower display.
  13.   Examine  each instruction in  sequence to
       ensure that the bootstrap  has  been loaded
       correctly.  If a location is found to be in error,
       the address of the location must first be entered
       using the load address key, before  the correct
       instruction is entered (using the DEP key).
  14.   Set the switch register to 7737.
  15.   Depress the computer LOAD ADDR key.
  16.   Depress the computer CLEAR key, then the
       CONTINUE.
  The first message displayed by  the disk operating
system is  a  designation of the  revision level of the
software,  for  example,  D7  or  EO.  The  revision
described in this manual is EO. The next line displayed
is called the system prompt. It is the return point for all
programs  and the user can  always return to it by
holding down the CONTROL (CTRL)  key  while
pressing the L key on the console keyboard. A program
called EXEC and a file named INITSS are used here.
EXEC contains the system prompt, the CONTROL
and IFSS prompts, and some utility functions. INITSS
contains  the system  configuation  information and
constants (see section 7.1).
Octal
7737
7740
7741
7742
7743
7744
7745
7746
7747
7750
7751
7752
7753
7754
7755
7756
7757
7760
           Table 2.2. Octal-Binary Equivalents of the Real Time System Disk Bootstrap Program.

                  Address                                                Instruction
    Octal
    7200
    6502
    6504
    6512
    1356
    6517
    7200
    6514
    6501
    5347
    6511
    5760
    6505
    7402
    5337
    7757
    1400
    0400
Binary
1 1 1
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
111
Oil
100
100
100
100
100
100
100
100
101
101
101
101
101
101
101
101
110
111
000
001
010
on
100
101
110
111
000
001
010
on
100
101
no
111
000
111
110
110
110
001
110
111
110
110
101
no
101
no
in
101
in -
001
000
Binary
.010 000
101
101
101
on
101
010
101
101
on
101
111
101
100
on
in
100
100
000
000
001
101
001
000
001
000
100
001
no
000
000
on
101
000
000
000
010
100
010
no
in
000
100
001
in
001
000
101
010
111
111
000
000
                                                 11

-------
2.4 SYSTEM ZERO  ADJUSTMENT
  This adjustment is extremely important and should
be checked frequently, at least once each work day. The
zero adjustment defines an electronic  threshold or
baseline for the acquisition of mass spectrometric data.
Any signal  from the mass spectrometer  that is below
the  threshold will not be observed by the data system.
All   signals  that are  above  the threshold will be
measured. If the threshold is too high, small but real
signals will not be  observed. This reduced sensitivity
will be accompanied by inaccuracies in ion abundance
ratios.  If the threshold is too low, there  will be a
decrease in  the signal/noise ratio and the mass spectra
will  contain many  2-5% relative abundance signals
that are meaningless.
  The  zero  adjustment   is   accomplished  under
computer control by running  the zero program. The
program is called by responding ZERO  to the system
prompt:

   SELECT  MODE:   ZERO
   MANUAL?:  N_
   AUTOMATIC?:  N_
   ZERO  ADJUST?:  Y_

The MANUAL prompt is used to maintain or release
computer control  of the  mass spectrometer.  A  Y_
response causes the computer  to release  control of the
spectrometer, that  is, put it in the manual mode, and
the  data system will  return  to the system prompt.
Therefore to use  the  zero program the user must
respond to MANUAL with an N. The AUTOMATIC
prompt was intended for a future implementation of an
automatic zero adjust. This has not been  accomplished
yet, and the user should respond N. After a Yjesponse
to the ZERO ADJUST?, the data system monitors the
output of the mass spectrometer at  a particular mass.
In earlier versions of the software, for example D2, the
zero program monitored a mass at the very low end of
the  mass  scale in  the  10 * 30 amu range. In  later
versions, for example D5, this  was changed to monitor
a mass in the 450  - 500 amu  range. The user should
make certain that the later version of the zero program
is used for the zero adjustment. In the lower mass range
there  is  often significant  background  from  air
components and other common gases. If the  zero
program is run with  the  ion source  and electron
multiplier   turned   on,  as   is recommended,  this
background will cause the threshold to be set too high.
In  the  range of 450 - 500 amu there is  very  little
natural background and a better zero adjustment may
be accomplished.
  The actual threshold adjustment is made, while the
zero program is running, by turning the preamplifier
zero adjust control' until just one light (octal 0001) is
observed  flashing  on  and  off  in  the  computer
accumulator. Zeroing should be performed  with the
preamplifier switch  in the position to be used during
data acquisition. The rotary switch on the PDP-8 front
panel must be set to the AC position  to display the
contents of the accumulator in the front panel lights. If
no  accumulator lights are displayed, the threshold is
too  high; if more  than  one accumulator light is
displayed, the threshold is too low.
  As  previously indicated,  the  zeroing  operation
should  normally  be performed  with the ionizer and
electron multiplier power applied. This should correct
for background noise  which exists while the mass
spectrometer  is  acquiring  data.  Many   Finnigan
GC/MS operators prefer to zero with no power applied
to the ionizer and electron multiplier. In theory, if the
ionizer is reasonably clean and the electron multiplier is
not noisy, the zero setting, with 0001 showing in the
accumulator, should be approximately the same with
ionizer and power supply on or off. Background noise,
however, often affects the preamplifier zero value.
   After  the  attainment  of  a satisfactory zero, the
CTRL/L key is used to return to the system prompt. If
a  stable zero  adjust cannot be achieved, there is a
serious hardware problem and  the trouble shooting
chapter should be consulted.
2.5 MASS SCALE  CALIBRATION

  The calibration of the mass scale should be checked
at the  beginning  of each day. See  section 2.6.  If
calibration is required the program CALIBR is used
for  this purpose.  Mass  scale  calibration,  requiring
approximately 40  seconds, is accomplished  over the
mass range or over a superset of the mass range selected
for  spectrum acquisition. The mass scale calibration
compound perfluorotributylamine (PFTBA) is injected
into the   mass  spectrometer,   and  the  system
automatically calibrates the mass scale using ions from
this compound. A file with the calibration data (mass
numbers relative to mass set voltages) is automatically
created. It should be noted that this procedure does not
calibrate ion abundances.  ••*..
                                                12

-------
  To  calibrate, the operator  calls the calibration
program:

   SELECT MODE:   CONTROL OR IFSS
   CALIBRATE?:   Y_   .
   CALIBRATION  FILE  NAME:  UP TO  SIX
     CHARACTERS
   MASS  RANGE:  20-700
   MS  RANGE  SETTING?:  H_
   NON  STND  CAL?:   N^
   DO CALIB  DIAGNOSTIC?:  Y_ OR N_

  The calibration program uses the IFSS mode which
is explained in Section 2.7. However, it may be called '
by responding to the system prompt with CONT or
IFSS
  The calibration file name defines the file which will
contain the current  calibration data. The file name is
used in each data acquisition run to define the mass set
voltages to be applied by the data acquisition routine. If
the calibration file name entered already exists in the
directory, the calibration  routine is aborted, an error
message is  displayed,  and the system returns to the
system prompt.
  The MASS RANGE prompt specifies the range of
mass units  covered  by the calibration file. Only one
mass range may be selected for calibration, it must
include masses 69 and 100, and the first and last.masses
must be separated by a hyphen.
  The MS  RANGE SETTING? prompt refers to a
mass range switch  on the model  1015 and earlier
spectrometers. The  user responds to this inquiry with
an H, M. L, or P  corresponding  to the mass range
selected on the mass spectrometer electronics console.
H, M, L, and P refer, respectively, to the high-mass
range (50-750),  mid-mass range (10-250),  low-mass
range (1-100), and peak identifier mass range (1-500).
The 3000 series systems have a single mass range and
the user responds with an H. It is  recommended that
1015 and earlier model users calibrate the high mass
range for most general purpose work.
  The  NON  STND  CAL?  prompt  is  used  for
calibration  with  PFTBA on  instruments with an
extended mass range (> 750 amu) or  for calibrations
with any alternative mass calibration compound. For
calibrations up to  750 amu  of extended mass range
instruments with   PFTBA,  the  user  responds Y.
Another prompt requests the name of the overlay file
and the correct response is FC1QOO. A separate overlay
file is, required for  each  alternative mass calibration
compound.     For    example,   with  tris (per-
fluoroheptyl) triazine as the mass scale calibrant, the
overlay filename is M1000. The user should be aware
that calibration  to 1000 amu requires a change in a
system status word as described in section 7.1 For
calibrations of  instruments  with  a mass  range  to
750 amu, the response is N_
  For the first calibration the user must respond N_to
the DO CALIB DIAGNOSTIC?  prompt.  Once the
system  has been  calibrated,  and  assuming  the
calibration  file  was  not deleted,  the  calibration
diagnostic feature may be used. After the response to
this prompt is entered, the operator must be sure the
ionizer is on, and the pressure of PFTBA is adequate
for calibration.  When calibration  is completed, the
operating system returns to the system prompt.
  A program called MASDEF may be used to shift
calibration masses by a constant amount per 100 amu
to correct for fractional masses. The program operates
on existing  calibration files, and the user has several
choices of mass shifts.  For example, if a shift of +0.06
amu is selected (one of several  possibilities),  the
program sets up a continuous shift  of 0.06 amu at 100
amu, 0.09 at 150 amu, 0.13 at 220 amu, etc. Thus one
may use a standard PFTBA calibration file and correct
for large  fractional  masses  in  saturated  aliphatic
hydrocarbons while  using  the higher  scan  speed
availabe with one sample/amu. Negative shifts are also
available and have several potential applications.
CALIBRATION  DIAGNOSTIC  PROGRAM
  An  affirmative  response  to  the  DO  CALIB
DIAGNOSTIC? prompt calls the program CDIAGN
and leads to another prompt:
   SELECT  MODE:  CONTROL  OR  IFSS
   CALIBRATE?:  Y_
   CALIBRATION  FILE NAME:  UP  TO SIX
      CHARACTERS
   MASS  RANGE:   33-700
   MS RANGE SETTING?:  H_
   NON STND CAL?:  N[
   DO CALIB DIAGNOSTIC?:  Y_
   PREV  CALIB  FILE  NAME:   UP  TO  SIX
      CHARACTERS
  After input of the name of the previous calibration
filename, which still exists on the disk, the  program
produces a report on the console printer or  CRT.  A
sample report is shown in Table 2.3.
  The first row of the report gives the  amplitudes
(AMP) of  the peaks that were used in the current
calibration. The mass numbers (AMU) that correspond
to these amplitudes are shown in the bottom  row. The
                                               13

-------
second row,  labeled DEL  N  (delta N), shows the
respective  changes  in  digital-to-analog  converter
(DAC) values from the previous calibration file. This is
perhaps the most significant set of data in the report.
Generally when calibration files are compared,  and
there has been  very  Irttle drift since the previous
calibration, the DEL N values will be no greater than
5-8. If several days have elapsed between calibrations,
DEL N values can be expected to approach 10-20 with
the Finnigan  1015. DEL N's significantly greater than
these values indicate a drifting mass spectrometer or,
less likely, a  data  system problem. The next row of
values labeled T, MS is the time, in milliseconds, over
which each ion signal was integrated. The amplitude
value must be divided by the integration time at each
mass to calculate comparable  ion abundance values.
The row of values labeled N is the DAC value, in octal,
that  was supplied to generate the mass set voltage for
each mass in the current calibration file. To reiterate,
the key value is the drift in DAC value of a given mass
between calibrations (DEL N).
require valid mass spectra of the compounds detected.
This  is  independent  of  the  actual  method  of
interpretation of the spectra, i.e., -an empirical search
for a match within a collection of authentic spectra or
an  analysis  from  the principles  of organic  ion
fragmentation. A properly operating and well tuned
GC/MS is required to obtain valid mass spectra.
  The purpose of the quality control test is to make a
quick check - about 15 minutes - of the performance
of the  total operating system  of a  computerized
GC/MS. Thus with a minimum expenditure of time, an
operator can be reasonably sure that the GC column,
the  enrichment  device,  the  ion  source, the  ion
separating device, the ion detection device, the signal
amplifying circuits, the analog to digital converter, the
data reduction system, and the data output system are
all functioning properly.
  An  unsuccessful  test  requires,  of course,  the
examination  of  the  individual   subsystems  and
correction  of the faulty  component.  Environmental
data acquired after a successful systems check are, in a
                   Table 2.3.   Sample Report from the Calibration Diagnostic Program.
MONITOR REPORT:
AMP
DEL N
T, MS
N
AMU
0
0
0
1465
20
234
0
1
2207
28
93
0
16
4053
50
726
-2
1
5522
69
70
-1 .
1
10173
100
244
-1
1
12646
131
71
-1
4
15774
169
233
0
1
22130
219
89
0
• 8
42363
414
107
1
. 64
63163
614
PLOTTER READY?

  An affirmative response to the PLOTTER READY
question in the Monitor Report will produce the plot
shown in Figure  2.5. The diagnostic program uses the
table  of mass  set voltage values that the calibration
program generated, scans over the ions that were used
in the current calibration, and plots the peaks. This plot
provides the user with the peak  shapes to determine
whether resolution was adequate. From these shapes,
the user can also determine the condition of the ionizer.
Very heavily skewed peaks often indicate that cleaning
is required. The plot also gives the operator a record of
calibration data on a day-to-day basis. The plot shows
relative abundances, the masses of the ions used in
calibration, the DAC value settings for each mass, and
the integration time spent on each ion. It also allows the
user to label the conditions used in calibration that day.

2.6 QUALITY CONTROL

  Correct identifications of organic pollutants from gas
chromatography  mass spectrometry  (GC/MS)  data
real  sense, validated  and  of far  more  value  than
unvalidated data. Environmental data acquired after an
unsuccessful  test  may be worthless and may cause
erroneous identifications.
  It is recommended  that the test be applied at the
beginning of a work day on which  the system will be
used and  also anytime there is a suspicion of a non-
obvious malfunction.
  The test was written specifically for the Finnigan
quadrupole mass spectrometer equipped with a Digital
Equipment  Corporation Model  PDP-8  computer.
However,  the test is clearly and readily adapted to any
GC/MS system by suitable modification of the detailed
procedure. Indeed data from other GC/MS systems
was used to establish the abundance criteria of the test
(Analytical Chemistry, £7,995, 1975).
  There  is a special  need  to  closely  monitor the
performance  of  the quadrupole mass  spectrometer.
Unlike the magnetic deflection spectrometer, the active
ion separating device of the quadrupole spectrometer,
the rods, is directly contaminated during operation and
                                                  14

-------
B 1 I u II I i/ft>»v J
, A. L
^ yi\, ,
*, i/lW i/JVji iAj iAjL i 1 •
Mi 20 I 28 I SO 1 69 1 108 1 131 I 169 1 219 1 261 I 111 I 332 I 611 I 650 I
N 1312 17S2 3102 1661 7031 11201 137S3 17111 22S67 3S3H 13577 S3S11 5S2B1
T 0 1 16 1 1 1 1 1 1 1 1 61 0

-------
after  prolonged  operation  is  subject  to  severely
degraded performance. Since degraded  performance
usually affects  the  high  mass region first,  the test
includes the high mass end criteria. High quality high
mass data is important  since many  environmentally
significant compounds 'have  molecular and fragment
ions in the 300-500 amu range.
  A quadrupole spectrometer which meets the criteria
of this test will, in  general, generate mass spectra of
organic compounds which are very similar, if not
identical, to spectra generated  by  other types  of
spectrometers. Thus quadrupole mass spectra will be
directly comparable to  spectra in collections which
have been developed over the years with other types of
spectrometers.
  The reference compound used in this test is available
from PCR, Inc., P.  0. Box 1778, Gainesville, Florida
32602.
Procedure:
  3.

  4.
Make    up   a    stock    solution    of
decafiuorotriphenylphosphine (DFTPP) at 1
milligram/milliliter (1000 ppm) concentration
in  acetone (or a hydrocarbon solvent). This
stock solution was shown to be 97 + % stable
after 6 months and indications are that it will
remain usable for several years. Dilute an
aliquot   of  the  stock  solution    to   10
microgram/milliliter (10 ppm) concentration
in acetone. The very small quantity of material
present in very dilute solutions  is subject to
depreciation due to adsorption on the walls of
the  glass  container,  reaction  with  trace
impurities in acetone,  etc.  Therefore this
solution may be usable only in the short term,
perhaps one week.
Adjust the  flow  from the GC column to  30
cc/min at the exit, attach the column to the
spectrometer,   and    set   desired    oven
temperature. Some suggested GC columns and
conditions are listed in Table 2.4. Parameters
should be adjusted to permit at least four mass
scans during elution of the DFTPP.  This will
permit selection  of a spectrum   that  is
reasonably free of abundance distortions due
to rapidly changing sample concentration.
Set sensitivity to  10'7 and zero instrument
using the automatic ZERO program.
Set up the real time system for control mode
with the following variables:  (see section 2.7
for additional information).
SELECT MODE:  CONT
CALIBRATE?:  N_
TITLE:  DFTPP +  THE DATE
CALIBRATION  FILE  NAME:  CAL
FILE  NAME:  UP TO  SIX CHARACTERS
MASS  RANGE:  40-450
INTEGRATION  TIME:  8_
SAMPLES/AMU:  j_
THRESHOLD:  press  return
RT GC ATTEN:  5_
FAST SCAN OPT?:  press return
MS  RANGE  SETTING?:   H_
MAX  RUN TIME:   20
DELAY BETWEEN  SCANS (SECS.)?:
   press return

5.    Inject 20 ng (2 ul) of the dilute standard into
     the GC column.
6.    After the acetone elutes from the column and
     is pumped or diverted from the system, turn on
     the ionizer and start scanning.
7.    Notethe exact retention time of the DFTPP as
     it elutes from the column as indicated on the
     realtime GC. This retention time can be used
     as a daily check of the condition of your GC
     column and  separator by  comparing  the
     values. The retention  times should not vary
     significantly from  day to day with identical
     operating conditions.
8.    Terminate the run by pressing CTRL/L, turn
     the ionizer and multiplier off, and plot the total
     ion current profile (see chapter 4.)
9.    Select a spectrum number on the front side of
     the GC peak as near the apex as possible and
     select  a   background  spectrum   number
     immediately preceding the peak.
10.  The mass spectrum can be output in various
     ways including a plot  of the full spectrum on
     the plotter or cathode ray tube or a print of the
     full spectrum on a printer or cathode ray tube
     (see chapter 4). A print of a partial spectrum
     using  the  following  responses  is usually
     convenient:

 SELECT  MODE:   OUTP
 TOTAL ION CURRENT  PROFILE:   N^
 EXTRACTED  ION  CURRENT  PROFILE?:    N^
 PLOT SPECTRUM?:   N_
 PRINT SPECTRUM?:   Y_
 FILE  NAME:  YOUR  FILENAME
 SPECTRUM  NUMBER.:   APPROPRIATE
   SPECTRUM NUMBER
                                               16

-------
   PARTITIONED OUTPUT?:  Y_
   MASS RANGE:   51;68-70;127;197-199;
     275:365:441-443
   MINIMUM  VALUE  %:  press return
   SUBTRACT BACKGROUND:  Y_
   SPECTRUM NUMBER:  APPROPRIATE
     SPECTRUM NUMBER
   BACKGROUND AMPLIFICATION:
     press  return
   SAVE SUBTRACTED  FILE?:  N_
   NORMALIZE  ON:  press return
                                                ions in Table 2.5. Figure 4.1 is a plot of an acceptable
                                                DFTPP spectrum.
                                                  If the relative abundances are not within the limits
                                                specified, the appropriate adjustments should be made
                                                using the proper trimpots on PC-5 of the Finnigan
                                                1015:  R-81 (low mass compensator) for the low mass
                                                range, R-77 (resolution, front panel) for the medium
                                                mass  range, and R-180 (high mass resolution) for the
                                                high mass range. The magnet position and ion energy
                                                may need to be adjusted very slightly to cancel  out
                                                front  end  liftoff. When an acceptable spectrum is
                       Table 2.4.  Suggested GC Columns and Conditions1
Dimension (Type)

 6' x 2 mm ID
     (Glass)
 6' x 2 mm ID
     (Glass)

 6' x 2 mm ID
     (Glass)

 6'  x' 2  mm ID
     (Glass)
                    1.95%  QF-1 plus
                    1.5% OV-17 on
                    80/100 mesh Gas-Chrom Q

                    3% OV-1  on  80/100
                    mesh Chromosorb W

                    5% OV-17 on  80/100
                    mesh Chromosorb W

                    1% SP2250 on  100/120
                    mesh Supelcoport
                                                     Flow Rate
30 ml/min
30 ml/min
30 ml/min
30 ml/min
220
220
170
                                R. Time
             4 min.
5 min.
5 min.
5 min.
'Use any column of your choice which gives at least four mass scans during elution of the DFTPP.
                                                obtained, it should be recorded for permanent storage.


            Table 2.5.   Decafluorotriphenylphosphine Key Ions and Ion Abundance Criteria.
  The spectrum obtained on the test system should
contain ion abundances within limits given for the key
                        Mass      Ion Abundance Criteria

                         51       30-60% of Mass 198
                         68       Less than 2% of Mass 69
                         70       Less than 2% of Mass 69
                        127       40-60% of Mass 198
                        197       Less than 1% of Mass 198
                        198       Base Peak, 100%  Relative Abundance
                        199       5-9%  of Mass 198
                        275       10-30% of Mass 198
                        365       At least  1%  of Mass  198
                        441       Less than- Mass 443
                        442       Greater than  40%  of  Mass  198
                        443       17-23% of Mass 442
                                             17

-------
2.7   SUGGESTED OPERATING PARA-
METERS FOR  SAMPLE  ANALYSIS
  This section deals with the instrumental portion of
the analysis after the sample or sample extract has been
properly   prepared.   Details  concerning  sample
preparation are given in Chapter 3. The first part of this
section deals with aspects of the gas chromatography
itself  and  includes a list  of some recommended
columns. The second part describes the operating
parameters required for data acquisition.
GAS CHROMATOGRAPHIC
CONSIDERATIONS
  In  order  to achieve optimum  separations,  the
selection of the column and  gas  chromatography
operating  conditions  are  of  utmost  importance.
Certainly the degree of separation of components in a
complex mixture has a direct bearing on the ability to
make  unambiquous mass  spectral  identifications.
Publications dealing with selection of the liquid phase
and  the solid  support, along with other  operating
parameters are referenced in the bibliography (Chapter
10).
  Generally, 6 ft. X 1/4 in. O.D.-(.08 in. I.D.) packed
glass columns with helium carrier gas flowing at 20-30
ml/mm  are   utilized  for  environmental  samples.
Comments on open tubular columns are in section 6.3.
Silanized glass wool plugs and silanized solid supports
of 80/100 mesh are recommended. While  using the
GC/MS system, one must be alert for the occurrence of
leaks  or a plugged  separator. Normally  with  helium
flowing through the column at 20-30 ml/min,  the
separator forepump pressure will be 0.1 torr or higher.
A pressure lower  than this indicates either, a leak
upstream from the  separator or a plugged separator.
The normal high vacuum pressure is 10"5 to  10"* torr
with helium flowing through the separator. A pressure
of 10"* torr or lower indicates a plugged separator.
  Injections  of methylene  chloride or  chloroform
solutions should be avoided on  systems without  a
diverter valve  because  injections of more  than  two
microliters may cause an excessive increase in pressure
in the mass spectrometer which may automatically
shut  down  the system.  Acetone  and hexane  are
acceptable solvents  to use. Injections should be made
with the ionizer off. The elution of the solvent from the
column is indicated by a deflection of the high vacuum
pressure gauge. After the pressure has returned to
normal the  ionizer  may  be  turned on   and data
acquisition can be initiated. However, for systems with
a diverter valve, appropriate measures should be taken
to divert the solvent  from the mass  spectrometer.
Under these conditions  .large  injections  of  even
troublesome  solvents may be  quite possible.  With
hexadecane or tetralin  extracts (extraction with high-
boiling  solvents   for  determination   of  volatile
components) data acquisition should be started prior to
injection and stopped just before elution of the solvent.
  Some representative columns  for  separation  of
environmental extracts are listed in Table 2.6. Only
well  conditioned columns should  be used and the
maximum operating temperatures  should never  be
exceeded.
  Table 2.6.  Representative GC Columns for
             Sample Extracts.
     Column
Maximum Operating
    Temperature
1.5% OV-17 +  1.95%
QF-1 on Supelcoport

5%  OV-17 on
Gas-Chrom  Q

3%  OV-101 on
Gas-Chrom  Q

4%  FFAP on
Gas-Chrom  Q

3%  Dexsil 300 on
Chromosorb G 60/80

1%  SP2250 on 100/120
Mesh Supelcoport
          250C
           300C
           300C
           200C
          400C
           375C
  Certain compounds can be determined by  direct
aqueous  injection into the gas chromatograph. The
ionizer is turned on  and data collection is  initiated
immediately  after  elution  of  the   water.  Some
representative columns for aqueous injections are listed
in Table  2.7.
                                                18

-------
 Table 2.7.  Representative GC Columns for Direct
            Aqueous Injection.
 Column Packine
Chromosorb 101
80/100 mesh
4% FFAP on
Chromosorb W
60/80 mesh
    Application

General for  Low  Molecular
Weight Compounds up to
MW 200.   Water elutes
before organics

Phenols, Acids
0.4% Carbowax 1500   Low Molecular Weight
on Carbopack A
Tenax  GC 60/80
mesh

Carbowax 400 on
Porasil C 100/200
mesh
Acids

Amines
General for  Low
Molecular Weight
Compounds
DATA ACQUISITION

  There are two data  acquisition modes available to
real time system users. The CONTROL mode utilizes a
fixed integration time for each mass that is monitored
while in the IFSS mode the integration time is varied as
a function of the signal strength. The principal file that
must be present on the disk is named CONTRL. This
handles data acquisition for the CONTROL and IFSS
modes.  For real time  display on  the CRT the files
RTCON and CONDIS must be  present.  For data
acquisition and  storage of accurate time data with the
remote in base (RIB) interface, the file CONDI 1 must
be present. For data  acquisition on a second disk drive
the files TWIN  and EXMOD2 must be present. None
of these files are  directly called by the user
  As shown in the examples in Section 2.6 and Chapter
3,  data  acquisition   is  initiated   by   answering
CONTROL to  the system prompt followed by a NO
response to the  calibrate question-.  The operator then
enters  the  required information  for each  of  the
remaining prompts which are described as follows:

TITLE:  The title, which is later printed or plotted on
the data outputs, can include up to 64 characters which
identify and  describe the samples.  Such  things  as
sample number,  date,  and  gas  chromatographic
 conditions may be included here.
 CALIBRATION FILE NAME:   The calibration file
. name is the file which contains the current calibration
 data (see Section 2.5 for details on calibration).
 FILE  NAME:  This input,  which may be up to six
 characters in length, identifies the data acquisition file.
 This file name is used later to output data.
 MASS RANGE:   This input specifies the atomic mass
 units which are to be scanned during data acquisition.
 One can have a single mass range, a set of up to 8 mass
 ranges, or, up to 8 individual atomic mass units. The
 mass  range  inputs  must be  entered in ascending
 nonoverlapping order separated  by  semicolons.  The
 mass range inputs need not be contiguous.
   Prior  to  initiating data  acquisition,  the  control
 routine checks the operator inputs against the inputs
 specified for the calibration file. This check ensures that
 the mass range is a subset of the values specified in the
 calibration file. If there is a  discrepancy in the mass
 range setting, (see below) or if the user attempts to run
 outside of the calibration file range, the run will be
 aborted, and an error message will be  printed on the
 console or CRT.
 INTEGRATION  TIME:  This  input specifies  the
 time in milliseconds over which the mass spectrometer
 monitors each individual mass. A numeric value of 1 to
 4095 is entered for each mass or mass range entry.
 These  are separated by semicolons.
 SAMPLES/AMU:  This  parameter can be  used to
 compensate for noninteger masses (for example, 243.2)
 when necessary. The number of samples to be taken per
 atomic mass unit  must be in the range of 1 to 10.
 Normally  an input  of one  is used.  This  provides
 sampling at integer mass units only. However, a larger
 number can  be utilized in which  the data acquisition
 program proceeds from the  integer amu by 0.1  amu
 steps until the number of samples per amu has been
 exhausted. The highest intensity point is then recorded
 as the  peak center. A single numeric value defining the
 number of samples/amu must be entered for each mass
 range  selected. For example,  an entry of 1;4;1  coupled
 with a mass range entry of 40-100; 101;  102-190 would
 result  in sampling at integer  mass units for each mass
 except 101. Mass 101  would be sampled at 101.0, 101.1,
 101.2, and 101.3.
 THRESHOLD:   This input is a single numeric value
 which applies to the entire mass range being monitored.
 The number, which represents an integer multiple of
 .025% of full scale, defines the background  noise
 threshold of  the system. The data acquisition software
 uses this value "as the lower'limit for the retention of
 data.  A sample point whose ion intensity  does not
                                                19

-------
exceed this limit is discarded and  reported as zero
intensity. The input range is from 0 to 4095 and the
default value is 0. This value modifies the zero setting
described in Section 2.4, and should  be used with
caution.
RT ON CRT?: A negative response to this causes the
program to put the  real time gas chromatogram on the
plotter, and even if the remote in base (RIB) interface is
present, no real time clock data is saved. An entry of Y
or YG initializes the real time clock capability if the
system has a RIB interface. A Y causes display on the
CRT of every third mass spectrum if the integration
time is 3 msec or longer; if the integration time is less
than 3 msec, the program bombs! A YG causes display
of the total ion current profile (TICP) on the CRT. Far
more important, when the DATA prompt is printed
and the user presses the space bar, the real time clock is
zeroed and started. The CRT or TTY bell  will ring
every two seconds; when the user presses RETURN,
data acquisition begins. Thus the clock may be  started
at  injection  time and   data   acquisition   started
subsequently.  Alternatively   pressing   RETURN
immediately after the DATA prompt zeros and starts
the clock and data acquisition. The real time clock data
is saved with the data file, and the MSSOUT software
only  (Chapter  4)  may   be  used   to  output  a
chromatogram with the X axis in spectrum numbers or
time in minutes.
RT GC ATTEN:   This is a value from 1 to 8 which
attenuates the real time output of the total ion current
profile  recorded  on the digital  plotter.  The  default
value is 1. For displays on the CRT, the attenuation of
the spectrum is accomplished with switch registers 0-5;
the attenuation of the TICP is accomplished with
switch  registers 6-11. The switch  registers may be
changed during the course of the run, and on both the
larger the octal number the more the attenuation of the
display.
FAST SCAN  OPTION?:   This  option modifies the
overall scan rate of the system, by reducing the settling
time (ordinarily 2.3 milliseconds) between successive
mass peaks. A response of yes should be made only if
this hardware option has been installed.
MS RANGE SETTING:   This indicates the range to
be used  (High,   Medium,  or  Low)  during data
acquisition with a model 1015. The operator responds
with  an H, M, or L, depending on the mass range
selected on the mass spectrometer electronics console.
Npte that  the calibration file  utilized  must  also
correspond 'to  the mass  range  setting. This was
described more completely  in Section  2.5. The 3000
series users should respond with an H
MAX RUN TIME:  This specifies, in  minutes, the
duration  of data acquisition. A response of zero or
pressing the return key. results in only one scan being
made. The upper limit for this entry is 2047 minutes.
DELAY  BETWEEN SCANS (SECS.):  An input
here allows one to delay between scans from one second
to 2047  seconds.  No response results  in  no delay
between scans.
SECOND DISK?: When data acquisition on a second
disk (Dl) is carried1  out, the current calibration file
must  be  present on the second disk. The  file name
created on Dl is exactly the same as the designated file
name on DO, and the file on Dl starts over at spectrum
number 1. The response to this prompt is YES or NO,
depending on whether  data  acquisition should  be
continued on the second disk if the first  disk becomes
filled to capacity.
  After all of the  prompts  have  been answered, the
computer responds with DATA and acquisition begins
when the return key is depressed. Data acquisition will
end when the previously stipulated maximum run time
is  reached   or  it  can be   halted  manually  by
simultaneously depressing the CONTROL and L keys.
  The time required to scan a complete mass spectrum
can be determined from the mass range, the integration
time,  the number of sample/amu, and the settling time
which is normally 2.3 milliseconds. For example, data
acquisition in the control mode with a mass range of
40-400, one sample/amu, and  an  integration time of
8 ms  would  result   in  scan   times   of   3.7
seconds:   (400-40) X (8 + 2.3)       =  3708 ms.
Generally scan times of 3-5 seconds are adequate to
produce good quality mass spectra. However, shorter
scan times may be necessary for early eluting narrow
GC peaks. The scan  time must be multiplied by the
number of sample/amu, and this could  increase scan
time significantly.
  The same operating parameters may be used for data
acquisition at a later time because these  parameter
values are stored with the data  file. Reuse of the same
parameters requires that the data file be present on the
disk. The program used to implement this capability is
called PERCAL. The name of an existing data file is
entered and PERCAL  retrieves the stored  operating
parameters, e.g., mass range, samples/AMU, etc. and
prints them on the CRT or TTY; also PERCAL places
them in their proper places in core memory for use in
the next data acquisition.

   SELECT  MODE:   PERCAL
   FILE  NAME?:   TEST
   TYPE OF  RUN:    CONTROL
                                                20

-------
   RUN  TITLE:   DEMO

   CAL  FILE NAME:   CAL
   NB.  OF  RANGES:   0001
   MASS RANGES:   0035-0500
   INTEG.TIME(S):   0008
   SAMPLE/AMU:  0001
   THRESHOLD:   0000
   MAX.  RUN TIME:   0025M
   DL  BET. SCANS:   OOOOS
   RT GC ATTEN:   0003
   NB.  OF  SPECTRA:   0485
   NB.  OF  POINTS:   0000
   FILE  NAME?:

  Entry of another file name will cause the program to
access another datafile and repeat the process. Entry of
CTRL/L causes a return to SELECT MODE, but with
the data from the latest PERCAL saved in core. If the
response to SELECT MODE is an asterisk, a new data
acquisition may begin svith minimum dialogue using
the same operating parameters.

   SELECT  MODE:
   FILE  NAME:        TURKEY
   MAX RUN  TIME:  20

   DATA

Another program .called RSTORE is'similar, but. has •
more  extensive capabilities  for storing  information
about the sample, etc. and works with new and old data
files.

INTEGRATION TIME AS A FUNCTION
OF SIGNAL STRENGTH (IFSS)
  The IFSS mode is a signal optimization mode which
is used when the operator wants the system to adjust
the Integration time  as  a  Function  of the Signal
Strength. The use  of this option results in a nearly
constant signal-to-noise ratio and reduces the chances
of saturation  of the detector. Figure 2.6 is a flow chart
of the IFSS algorithm. The dialogue begins when the
operator responds  to the computer prompt SELECT.
MODE by entering IFSS. The other prompts have the
same significance as in the control mode except for the
following:

MAXIMUM REPEAT COUNT:  This input (from 1
to 4096)  specifies the  maximum number of times to
repeat the specified base integration  time if the signal
level does not reach the specified upper threshold. In
effect, this input coupled with the base integration time
 allows the operator to specify a maximum integration
 time (maximum  repeat count x base  integration
 time =  maximum integration time).
 BASE   INTEGRATION  TIME:  The   operator
 responds by entering a number from 1 to 4096, which is
 the time in milliseconds desired  for base integration
 :ime. Generallv a low base  integration-time (1-4)  is
 desirable in order to avoid saturation of the detector.
 REPEAT COUNT BEFORE CHECKING LOWER
 THRESHOLD:  This input (from 1 to 8) specifies the
 lumber of times to repeat the base integration time
 aefore testing the lower threshold value.
 [f the signal strength for a given ion has not reached the
 lower threshold value, then the integration moves onto
 the next higher mass ion. This, in effect determines the
 minimum integration time (repeat count before check-
 ing    lower    .threshold x base      integration
 time =  minimum integration time).
 LOWER THRESHOLD: This value determines the
 point at which the program ceases the sampling  of a
: given ion when a weak  signal occurs. The  program
 collects the signal for the minimum integration time,
 then if the lower  threshold has  not  been  attained,
 integration continues at the next higher mass. A value
 from 1 to 8 must be entered. These values correspond to
 the strength  of  the signal  required to  continue
 integration  at a  given  mass (1  =  0% full  scale,
 2  = '0.025%,'     3 =0.05%,  '    4 = 0.10%,
 •5  = 0.20%,   .    6 =  0.39%,       7 = 0.78%,
 8  = 1.56%). An  entry of 1   would mean   that
 integration   would   continue  for  the  maximum
 integration  time  or until  the  upper threshold  is
 attained.
 UPPER THRESHOLD: This value determines the
 point at which the sampling of a given ion ceases when
 a strong signal occurs. The program collects the signal
 until the upper threshold  limit  is attained or the
 maximum integration  time is  reached  (whichever
 occurs first), then continues to the next higher mass.
 The values (1 to 8)  correspond to the signal strength
 required for integration to stop (1  = 50% full scale,
 2  = 25%,  3 = 12.5%,  4 =  6.25%,5 = 3.12%,
 6  = 1.56%, 7 =  0.7%, 8  = 0.39%). An entry of 1
 would mean that, for those signals where the lower
 threshold  has  been attained,   integration would
 continue until the signal  was 50% of full scale or until
 maximum integration time is reached.
    Numerous specific applications  of these  operating
 parameters are  contained  in  Chapter 3  with the
 appropriate   sample  preparation • methods.   The
 PERCAL program  described above also applies  to
 IFSS parameters.
                                              21

-------
    4,193,280
   SATURATION
                                                                      SAVE  ACTUAL  REPEAT
                                                                           COUNT  FOR
                                                                     NORMALIZATION  OF MASS
                                                                     SPECTRUM,CLEAR INTEGRATOR
! 6.25'% OF SATURATIONS)
                                    INTEGRATE FOR A SHORT
                                           -2 M  SEC )
                                     BASE    INTEGRATION
                                             TIME
                                             HAS
                                        SIGNAL   REACHED
                                            UPPER
                                           THRESHOLD
  UPPER
THRESHOLD
         LOWER
       THRESHOLD
                                         UNT BEFORECHECKIN
                                             HAS
                                            MAXIMUM
                                        COUNT    BEEN
                                           REACHED
(O.I %  OF SATURATION-4
0.025%  OF SATURATION
                                                              LOWER THRESHOLD
                                                              UPPER THRESHOLD
                                                              BASE INTEGRATION  TIME
                                                              MAXIMUM  REPEAT COUNT
                                                              REPEAT COUNT BEFORE
                                                              CHECKING  LOWER THRESHOLD
                        .  Figure 2.6   The IFSS algorithm flow chart
                                             22

-------
2.8 MASS SPECTROMETER
SHUT-DOWN
SHORTTERM
  When the system is to be shut down for short periods
of time (overnight,  weekends,  etc.) the following
procedure is recommended:

  1.    Turn ionizer  off.  Turn  electron  multiplier
       voltage power supply to off or standby. Turn
       RF/DC power supply to standby (1015).
  2.    Depress HALT switch on computer. Set the
       LOAD-RUN switch to LOAD and  wait for
       the white LOAD light to signal that the disk
       has  stopped spinning. Shut down computer
       terminal and other data output devices. Turn
       computer power key to OFF.
  3.    Remove beam from oscilloscope with vertical
       control or on/off switch.
  4.    If a diverter valve is present, direct the carrier
       gas from the mass spectrometer. If no diverter
       valve is present, one option is to disconnect the
       column and cap the GC inlet. This allows the
       spectrometer vacuum system to  pump  out
       residual background under a high vacuum. A
       fritted disk  filter  in the bulkhead  union
       provides adequate protection from separator
       clogging    during' •  column    disconnect.,
       Alternatively,  one may  reduce the  flow
       through   the   GC   column   to   about
        lOml/minute,  but not disconnect it from the
       separator.  This   provides  some  protection
       against oil backstreaming into the analyzer in
       the event of a power failure. Leave the column
        oven temperature  at least  50C  below the
        maximum operating temperature to minimize
        column bleed.
LONGTERM
  When the system is to be shut down for long periods
of  time  or  for  maintenance  purposes  (changing
separator, cleaning rods, etc.) the following procedure
is recommended:
  1.    Proceed as for short term shutdown.
  2.    Turn RF/DC line power switch to off (1015).
  3.    Remove GC column and  cap the  separator
       port. Cool column and turn off helium flow
       through the GC.
  4.    Turn power off/protect/protection  override
       switch on the vacuum controller to off. This
       turns  off  the  diffusion  pumps,  but  the
       mechanical pumps will remain on.
  5.    Wait at  least 45  minutes for the  diffusion
       pumps to cool. Check to make sure that the
       diffusion pump base is cool to the touch.
  6.    To  vent the system when a  solid inlet is
       present, insert a hose from the regulator on a
       cylinder  of dry helium or nitrogen  into the
       solid inlet assembly. Set no more than 5 psi
       pressure from  the  regulator. Tighten  the
       retaining knob on valve assembly. Open both
       the ^ high  vacuum  and   analyzer  valves
       simultaneously (slowly). When the valves are
       partially open (you  will be able to hear the
      • mechanical pumps pumping the venting gas),
       throw the main circuit breaker on the vacuum
       controller to shut off the mechanical pumps.
       Open the valves 'all the way, allow the system
       to fill with  helium, and then remove the gas
       line and close the solid inlet valve.

  Exposure of the copper-beryllium electron multiplier
to air for extended  periods of time (15  minutes or
longer) will result in rapid loss of multiplier gain. The
multiplier may be damaged permanently. Continuous
dynode  electron  multipliers  are not  affected  by
exposure to air. It is recommended that Cu-Be electron
multipliers be stored in an inert gas atmosphere or a
vacuum desiccator.

-------
  The  sample preparation  methods in this chapter
emphasize the broad spectrum approach explained in
chapter 1. There are no methods that are optimized for
a specific compound (parameter) or a  closely related
group of compounds.  Each method is a  generalized
module with a scope and limitations. Several different
modules must be applied to cover the broad range of
several  hundred  thousand  compounds  that  could
contribute to the pollutants in any specific sample. In
some cases there is significant overlap in the scope of
the modular methods. The  appropriate choice of a
specific module will depend on the objectives of the
survey and the scope and limitations of the method.
  Clearly there are deficiencies in some of the methods
and some environmentally significant compounds may
never be amenable to  analysis by any method in this
chapter.  For  these  classes  of compounds, -major
innovations are required such as the development of
liquid chromatography methods. For the present these
problems are beyond the scope of this manual.
  Many of the methods in this chapter are flexible in
that  they  can be combined  or  shortened, altered at
various stages to fit specific needs, or expanded to cover
unanticipated contingencies. Innovation in using these
procedures is encouraged insofar that it will improve or
otherwise  make  a particular  analysis  more efficient.
Hopefully,  these procedures, in  one form  or another,
can be used for the majority of analyses that confront
the GC/MS laboratory.
  All of the methods were designed for compatibility
with existing  GC/MS systems.  This  is  particularly
noticeable in the choice of solvents, since some solvents
cause excessive pressures in some mass spectrometers
and  can cause an automatic  shut  down  of the high
vacuum system. Generally all methods are designed to
be as  simple as possible.  The reliable  qualitative
information inherent in the GC/MS method  precludes
the need for a great deal of clean-up in most samples.
Simplicity  in methods  minimizes the  chances  of
contamination from reagents and glassware. However
where  appropriate, as in  fatty tissue  and  sediment
analyses, fractionation procedures are included in the
methods.
  Recovery data on spiked samples is presemed with
many of the  methods. It  should be  recognized that
these values  reflect the best  effort of a skilled and
experienced  laboratory  scientist  in one  or  two
laboratories. These data do not guarantee that similar
values  will  be  obtained  in  other  laboratories.
Recoveries of spikes  must  be determined  in  each
laboratory as part of the quality assurance protocol of
the method if the  data are  used  for concentration
measurements.
  All of the methods in this chapter were contributed
by EPA laboratories. Many were developed or evolved
because of the need  to have  methods oriented  to
GC/MS that can be used for a particular type of sample
or group of compounds.
  The majority of the methods are  for organics in
water, but that is not to imply that analyses of organics
in air, sediment, and fatty tissues are less important. In
reality, there are fundamentally, no differences in the
GC/MS  aspects  of the analyses  of environmental
samples from different media. The principal differences
are  in  the  sampling techniques  and  preliminary
separation procedures.  Therefore  much  of  the
information found in the section on water samples is
applicable to samples from other media. References to
these sections are included  in the sections on air,
sediment, and fatty tissue samples.
  In  the interest of  consistency, each ..method is
described in  a similar format.  The first part of the
method contains  a brief  definition  of the  method,
advantages and disadvantages, scope  and  limitations,
the rationale for the choices of materials and reagents,
detection limits, recovery efficiencies,  and any special
problems  with the  method.  All detection limits are
expressed in terms of a 33-450 amu mass spectrum with
average noise <1%  of the  base  peak. A  reagents
section  is included if reagents are required  with
emphasis  on the  purification of  materials.  An
equipment section lists and describes  unusual or new
                                                 24

-------
equipment  only.  There is no  listing  of standard,
conventional  glassware  and  other  equipment  that
should be found in any well equipped laboratory. The
procedure section  is a  step-by-step account of the
method itself including unusual sampling and storage
considerations,  cleaning  of glassware  and  sample
containers,  requirements for blanks and spikes, GC
column selection, temperature programming and other
GC conditions, and sample computer dialogue for the
GC/MS data system.
3.1 WATER SAMPLES

  Water samples encompass a wide variety of types
including processed water for human consumption,
lake water,  river  water, ground water, leachate,
municipal wastes, industrial wastes, and treated wastes.
Each of the sample preparation methods is applicable
to one or more of the water sample types and many
references to these applications appear in the following
sections.

DIRECT AQUEOUS
INJECTION  INCLUDING
CONCENTRATION TECHNIQUES
  Direct aqueous injection is the introduction of a few
microliters of water sample into the GC/MS system.
This procedure allows the identification of compounds
that are not amenable  to other methods because they
are too water soluble.
         The minimum detectable concentration using direct
       aqueous injection is 1 to 5 mg/1 (ppm) depending on
       the compound observed.  The method is suitable for
       analysis of many types of samples, expecially industrial
       and municipal wastes.
         The following types of compounds can be detected
       by  GC/MS  using  the  direct  aqueous  injection
       procedure:     aliphatic   hydrocarbons   Cl  -  C7;
       aromatic  hydrocarbons-up to  short  chain alkylated
       benzenes;  the  more   volatile   alcohols,  glycols,
       aldehydes, ketones, ethers, fatty acids, esters, amines,
       amides, and halogenated compounds.
         Simple distillation can be used to concentrate water
       samples  for direct aqueous injection. This technique
       has been used to concentrate alcohols, ketones, amines,
       nitiles, dioxolanes, and pyridines. Chlorinated solvents
       such as methylene chloride and chloroform also have
       been  identified  in distillates of industrial effluents.
       Table  3.1 shows examples  of recoveries  of  known
       amounts of several compounds from a waste sample
       after  simple distillation.  Approximately 10% of the
       spiked solution was distilled.
         The minimum detectable quantity is lowered by
       distillation to about 1-10 ug/1 in the original sample. It
       may be necessary to concentrate several liters of water
       for situations  such as the analysis of drinking water
       where taste and odor problems have been observed.
       .This can be done by using a large distillation flask or by
       distilling  several  portions  of • the  sample  'and
       compositing the distillates for a final distillation in a
       semi-micro  apparatus.  This allows  a detection limit
       lower than  1  ug/1 for some compounds.
            Table 3.1.  Recoveries of Spiked Organics from Water After a Simple Distillation.1
                      Compound Added

                      methanol
                      acetonitrile
                      propionaldehyde
                      methyl ethyl  ketone
                      1,4-dioxane
                      n-butanol
                      aniline
                      isopropanol
                      nitrobenzene2
                      o-toluidinej
                      di-n-propylaminej
                      di-n-butylamine1
Number of
Replicates

  4
  4
  4
  4
  4
  6
  5
  4
  1
  1
  1
  1
Average
% Recovery

     31
     44
     70
     69
     23
     56
     12
     54
     57
     20
     60
     68
   'Five hundred ml of boiled distilled waste was spiked at 0.2 mg/1 with the indicated organic chemicals. 'The spiked water was made
    strongly basic.
                                                 25

-------
  A  simple  evaporation  technique is useful  for
concentrating organic compounds whose boiling points
are  too  high for steam distillation.  This  has  been
applied to the identification of ethylene glycol in water.
No special equipment  is required for direct aqueous
injection. The  distillation  concentration  technique
requires a conventional, all-glass laboratory still with a
1-3 liter capacity.
  Install   a   conditioned   column   in   the   gas
chromatograph. Special care should be taken to flush
the  column sufficiently with  helium  to remove any
residual  air from the packing before heating. Be sure
the column was conditioned to minimize bleed during
the sample run. Columns which are applicable to direct
aqueous  analysis are listed in Table 2.7. Temperature
programming from  70C to 220C at 8  degrees per
minute may be utilized. Inject 2-8 microliters of the
water sample with the ionizer off. With the pressure
about 10"5 torr,  turn  on the  ionizer, start  data
acquisition, and begin temperature programming. Halt
data acquisition after about 35 minutes.
  Either the control mode  or the IFSS mode may be
used for data acquisition. The sample dialogue below
contains some suggested parameters.
   SELECT  MODE:   CONTROL
   CALIBRATE?:   N_
   TITLE:  UP  TO 64 CHARACTERS
   CALIBRATION  FILE NAME:  CURRENT  FILE
   FILE  NAME:  UP  TO  SIX  CHARACTERS
   MASS RANGE:  20-260
   INTEGRATION  TIME:   V7
   SAMPLES/AMU:  j_
   THRESHOLD:   press return
   RT  ON CRT?:   N_
   RT  GC ATTEN:   1  - 8
   FAST  SCAN OPT?:  N_
   MS  RANGE SETTING?:  H_
   MAX RUN  TIME:  30
   DELAY  BETWEEN SCANS  (SECS.)?:
      press  return
   SELECT  MODE:   IFSS
   CALIBRATE?:   N_
   TITLE:  UP TO  64 CHARACTERS
   MASS RANGE:  20-260
   SAMPLES/AMU:  1_
   MAX RPT COUNT:  32
   BASE  INTEGRATION  TIME:   1
   RPT COUNT  BEFORE  CHECKING  LOWER
      THRESHOLD:  8
   LOWER  THRESHOLD:   4_
   UPPER  THRESHOLD:   4_
   RT  ON  CRT?:  N^
   RT  GC  ATTEN:   1  - 8
   FAST  SCAN  OPT?:  N_
   MS  RANGE  SETTING?:  H_
   MAX RUN  TIME:   30
   DELAY   BETWEEN  SCANS  (SECS.)?:
      press return

  If no peaks are observed, a concentration procedure
may be applied. For distillation of industrial effluents.
transfer 500 ml of sample to a one liter distillation flask,
add boiling stones, and collect 25-30  ml of prime
distillate. This prime distillate may be redistilled in a
semi-micro distillation apparatus with the collection of
one ml of secondary distillate. The prime or secondary
distillate is analyzed by direct aqueous injection as
above.
  For distillation of samples  that are  suspected to
contain less than one ug/1 of an organic compound,
collect  only  10-15  ml  during  the first distillation.
Discard the pot residue and add an additional two liters
of sample.  Distill as above and composite the prime
distillates .by collecting the second 10-15 ml in the same
collection flask.
  A  simple  evaporation may  be  accomplished in a
beaker or evaporating dish on a hot plate. The residue
in this case is  analyzed as above.

INERT GAS PURGING  AND TRAPPING
  Gas  purging and trapping  is  a method  for the
isolation, concentration,  and  determination  of  low
boiling organics in water. The method uses finely
divided gas bubbles passing through the  water sample
to transfer organic compounds from the aqueous to the
gas phase. The compounds concentrate by adsorption
on a  porous polymer trap at room temperature as the
purge gas is vented. The compounds are subsequently
desorbed at elevated temperature by backflushing with
a carrier gas into the gas chromatographic system. The
method may  provide both qualitative and quantitative
information.  Purging may be accomplished at ambient
or elevated temperatures with helium or another inert
gas.

Equipment.   The equipment required for this method
consists of a purging device, a trap, and a trap heater or
desorber. Figure 3.1 shows construction  details for an
all glass purging device with a 5 ml sample capacity.
                                             26

-------
        OPTIONAL
         FOAM TRAP
  EXIT 1/4
  IN. O.O.

*- 14MM. O.D.

INLET 1/4
IN. O.D.
1  4 IN.
O.D. EXIT
    10MM. GLASS FRIT
    MEDIUM POROSITY
      SAMPLE INLET

        2-WAY SYRINGE VALVE
       17CM. 20 GAUGE SYRINGE NEEDLE
      6MM.O.D. RUBBER SEPTUM
                                      10MM. O.D.
                                        INLET
                                        1/4 IN. O.D.
                           1/16 IN. O.D.
                           STAINLESS STEEL
                                                      13X MOLECULAR
                                                      SIEVE PURGE
                                                      GAS FILTER
                                                         PURGE GAS
                                                         FLOW CONTROL
               Figure 3.1  A Purging Device with a 5 ml Sample Capacity.
                                   27

-------
The glass frit at the base of the sample volume allows
the  finely divided gas bubbles to pass through the
sample  while the sample is restrained above the frit.
Gaseous volumes  above the sample  are  kept  to a
minimum  to  eliminate dead  volume effects, yet
sufficient space is allowed  to permit most foams to
disperse. The inlet and exit ports are constructed of
heavy walled quarter inch glass tubing to permit leak
free  removable   connections   with  finger   tight
compression fittings containing Teflon ferrules. The
removeable foam trap is optional and recommended for
samples that foam. A 25 ml capacity purging device is
recommended for use  with a mass  spectrometer GC
detector.
   Figure 3.2  shows  a trap which is  a short gas
chromatographic column that retards the flow of the
compounds of interest at ambient temperature  while
venting the purge gas and, depending on the adsorbent
used, much of the water vapor. The trap is constructed
with a  low thermal mass to  allow  rapid heating for
efficient desorption, and  rapid  cooling  to  ambient
temperature for recycling. The trap length, diameter,
and wall thickness indicated in Figure 3.2  are critical
and variations in these will  affect  the trapping and
desorption  efficiencies of the compounds discussed in
this section.
   The trapping and desorption efficiencies are also a
function of the adsorbents, adsorbent mass, and the
adsorbent packing  order shown  in Figure  3.2. The
single adsorbent Tenax GC (60/80 mesh) is effective
for  compounds  that boil  above approximately 30C.
However compounds that  boil below approximately
30C are not strongly adsorbed by Tenax and may be
vented  under the purging  conditions. If compounds
that boil below about  30C are to be measured, a dual
adsorbent trap should  be  used. Grade-15 silica gel
effectively retards the flow of most organics at ambient
temperature and  should be packed behind the Tenax to
trap the lower boiling  components.  Silica gel is  not a
useful   single  adsorbent  because  higher  boiling
compounds do not efficiently desorb from it at  180C.
  The Tenax-silica gel combination trap  utilizes the
adsorptive properties of two materials to provide a trap
that effectively adsorbs and desorbs  a wide variety of
organic  compounds. The small amount of OV-1  on
glass wool at the trap inlet (Figure 3.2) is to insure that
all the Tenax adsorbent is within the heated zone and is
efficiently heated to the desorption  temperature. A
metal fitting at the trap inlet could act as a heat sink
and create a cool spot on the Tenax if this spacer is not
used.
  Details of the trap heater  are also shown in Figure
3.2.   The  adsorption-desorption   cycle   may   be
accomplished conveniently with the use of a six port
valve and plumbing system constructed of materials
that  neither adsorb volatile organics nor outgas them.
Several commercially available purge and trap systems
use this approach. With the six port vaJ,ve in the adsorb
position,  the effluent from the purging device passes
through  the trap where the flow of the organics is
retarded  and the purge gas is vented.  During  this
period the gas chromatograph is supplied with carrier
gas and may be used for other analyses. With the valve
in the desorb position, the trap is placed in series with
the gas  chromatographic column  which allows the
carrier gas to back flush the trapped materials onto the
chromatographic column.
   It  is strongly recommended  that the power for the
desorber   heater  be   supplied  by  an  electronic
temperature controller that is  set  to begin supplying
power as  the valve is placed in the desorb position. This
allows rapid heating of the trap to  180C with minimal
overshoot and maintenance of the desorb temperature
until desorption is complete (a four minute backflush at
20-60 ml/min is recommended). Using this procedure,
the trapped compounds are released as a narrow plug
into  the  gas chromatograph, which should  be at the
initial operating temperature.  Packed columns with
theoretical  efficiencies near  500  plates/foot  under
programmed' temperature  conditions can  usually
accept such desorb injections  without altering peak
geometry.
   Substitution of a non-controlled power supply, such
as a manually operated variable transformer, will cause
non-reproducible  retention times and may lead to
unreliable  concentration measurements. If it  is not
possible  to heat the trap  in a rapid  and controlled
manner to the desorption temperature, the contents of
the trap may be transferred onto the analytical column
at 30C or lower and once again  trapped. The analytical
column is then  rapidly heated  to the initial  operating
temperature for the analysis.
   Several gas  chromatographic columns have been
employed for the separation of the volatile components
prior to  measurement. A  recommended  column  for
general purpose work is a 8-ft x 0.1-in. id stainless steel
or glass  tube packed  with 0.2% Carbowax  1500 on
Carbopack-C (80/100  mesh). With a helium flow of 40
ml/min., the initial temperature of 60C is held for three
min., then programmed at 8C/min. to 160C.
Newly packed traps should be conditioned overnight at
230C with an inert gas flow of at least 20 ml/min. The
trap   is   also  conditioned  prior  to  daily  use  by
backflushing at 180C for 10 min.
                                                 28

-------
PACKING PROCEDURE
                                            CONSTRUCTION
DUAL A
GLASS WOOL 5MM
OSAOE IS
SIUCA GEL
8CM


TENAX ISCM




17. OV-I ICM
GLASS WOOL
SMM









ia;
oso
//
k

M
1
1
>$
If
^
1
•••-•: -T.
1
if ir
»iNT SINGLE
GLASS WOOL SMM




TENAX J3CM




37. OV-I ICM
GLASS WOOL
SMM
(LET
A









TIJ
9SO
%
\
1
|
1
1
1
^
;'«•
1

\
kf i
HUNT









411
                                                            COMPIISSION nnrno NUT
                                                            AND FEMUIEI
                                                                    MFT. PVFOOT  BISISTANCE
                                                                    WIKf WIAPKO SOUO
                                                                 TNIIMOCOUKE/CONTIOILIR
                                                                 SINSOI
                                                             TUMNO 3SCM 0.10) IN. 1.0.
                                                            • 0.1JJ IN. O.D. STAINLESS STia
       Figure 3.2   The Trap Assembly for a Purging Device
                                  29   '

-------
Discussion.  Table 3.2 contains a list of some of the
compounds that have been submitted to this type of
analysis. The recovery data is intended to be illustrative
only  since recoveries  depend  strongly  on several
important method variables. Recoveries are expressed
as a percentage of the'amount added to organic free
water. The purge time was 11-15 minutes with helium
or  nitrogen,  the  purge rate  was  20  ml/minute  at
ambient temperature, and the trap was Tenax followed
by Silica Gel. Data from the 5  ml sample was obtained
with a custom made purging  device and either flame
ionization,   microcoulometric,   or    electrolytic
conductivity  GC detectors.  Data from  the  25 ml
sample was obtained with a Tekmar commercial liquid
sample concentrator and a  mass spectrometer GC
detector using continuous repetitive measurement of
spectra.
  A  number  of  other  compounds   have   been
concentrated  and measured using  the purge and trap
method, but no recovery  data for these is available.
These    compounds     include     chloromethane,
bromomethane,  chloroethane,  1,2-dichloropropane,
trans-1,3-dichloropropene-1,  cis-1,3-dichloropropene-
1,    1,1,2-trichloroethane,    2-chloroethylvinylether,
1,1,2,2-tetrachloroethane, and  ethylbenzene. In general
the method is applicable to compounds that have a low
solubility  in water and a vapor pressure greater than
water at ambient temperature.
  All of the experiments summarized in Table 3.2 were
conducted with   the   water  sample  at  ambient
temperature,   about   22C.  Purging   at  elevated
temperatures has been investigated and clearly affects
recoveries of some compounds. However, this  is not
recommended as a  general  procedure because the
elevated temperature may promote chemical reactions
that could significantly alter  the  composition of the
trace organics.  This is  especially important  with
chlorinated drinking water or waste effluents.
  A significant method variable is the purge gas flow
rate. The total purge time is not very flexible since it
will usually be desirable to keep this as short as possible
to minimize the analysis time.  In all of the experiments
summarized in Table 3.2, a flow rate of 20 ml/min.. was
employed for 11-15 minutes, and for many compounds
an acceptable recovery was obtained. Figure 3.3 shows
the  percentage  recovery of several  representative
compounds as a function of purge gas flow rate. The
general curve shape displayed for 1,2-dichloroethane
and bromoform is typical of most of the compounds in
Table 3.2. The  low boiling compound vinyl chloride
was  trapped on Tenax  only  and  its flow rate  curve
illustrates  the sharp reduction in trapping efficiency
observed with this  type of compound  and trap at
elevated     flow     rates.   .  The     compound
dichlorodifluoromethane displays a similar  flow rate
curve even with the combination Tenax-silica gel trap.
  Because of the differences in  the  construction of
various purge and trap devices, actual recoveries may-
vary significantly from those shown in Figure 3.3 and
Table 3.2.  Therefore, it is required  that  individual
investigators determine recoveries of compounds to be
measured as a function  of  flow  rate with their
apparatus. Operation in  the optimum flow rate range
will assure maximum  sensitivity and precision for the
compounds measured.
  The recoveries of aliphatic hydrocarbons were found
somewhat more variable than the recoveries of the
other compounds investigated with this method. In all
of these experiments, known quantities were added to
organic free water,  and  the slightly soluble aliphatic
hydrocarbons probably formed a thin surface layer on
the water.  Under these inhomogeneous conditions,
special care is needed to achieve consistent results.
  In  order  to  generate quantitative  measurements
within a reasonable  purge time,  e.g.   10-15 min.,
calibration  of the method  with known  standards is
required. The recommended approach is to estimate
the concentration of the  unknown by a comparison of
its peak size with the size  of the corresponding peak in a
quality  control  check   standard  made   at  some
appropriate  concentration  level  and  measured  at
regular  intervals  during the work  day. From this
estimate  a  concentration  calibration   standard  is
prepared with the concentrations of the compounds to
be measured within a factor of two or less of the
probable concentration in the unknown.  Standards are
prepared by taking aliquots of solutions in  methanol
and injecting them into organic free water. This insures
maximum dispersal  of the  organic compound in the
aqueous system. Organic free water is  prepared  by
passing  distilled water through an activated  carbon
column.
  The standard is purged and measured immediately
after the unknown and under the identical conditions
used with the unknown. A sample of organic free water
should be purged between  each  set of  samples and
especially after all high level  standards or samples. This
will insure that the apparatus is purged of contaminats
and prevent cross contamination.
  Sample matrices significantly different than  surface
water have not been investigated extensively. Extremes
of pH, high ionic strength, or the presence of miscible
organic  solvents will likely  affect  recoveries of some
compounds. Therefore measurements of compounds in
                                                 30

-------
                 Table 3.2.  Recoveries of Organics by Gas Purging and Trapping.

Amount
Added
ug/1
% Re-
covery
5ml
Sample

Amount
Added
ug/1
% Re-
covery
25 ml
Sample
Chlorinated  Hydrocarbons

 methylene chloride                                                       12                97
 chloroform                        160                101                 12                95
 chlorodibromomethane              8,2              86,69                 12                72
 bromodichloromethane             40,2              94,82                 12                72
 carbontetrachloride                 4,2              108,82
 1,2-dichloroethane                  2                 70                 12                95
 1,1,2-trichloroethylene              2,2              118,89
 tetrachloroethylene                 2,2              81,80
 chlorobenzene                       4                 80
 p_-dichlorobenzene                  4                 71                 12                58
 dichlorofluoromethane               2                  7
 trichlorofluoromethane               2                 99
 vinyl chloride                       2                 95
 1,1-dichlofoethylene                 2                114
 1,1-dichloroethane                  2                 90
 trans- 1,2-dichloroethylene           2                100
 1,1,1-trichloroethane                2                 88
 dichloroiodomethane                2                 72
 2,3-dichloropropene-l               2                 85-

Hydrocarbons

 n-pentane                          81                100
 n^nonane                          93                 99
 n-pentadecane                     100                 56
 benzene                                                                 12                105
 toluene                                                                 12                91

Brominated  Hydrocarbons

 bromoform                        4,2               67,48                 12                73
 1,2-dibromoethane                 4,2               80,47

Nitrogen  Compounds

 nitromethane                                                            24                  2
 nitrotrichloromethane                                                    24                 22
 N-nitrosodimethylamine                                                  24                  0
 N-nitrosodiethylamine                                                    12                  0
 N-nitrosodi-n-butylamine                                                12                  0
                                              31

-------
                                                 BROMOFORA
                                           1,2-DJCHLOROETHANE
                                  DieHLORODiFLUOROMETHANE
                                              VINYL CHLORIDE   o
                           FLOW RATE ml/min.
Figure 3.3  Recoveries of Selected Compounds as a Function of Purge Gas Flow Rate.
                                32

-------
these   matrices  must   include  determinations  of
recoveries of spikes in the sample matrix and perhaps
use of a calibration based on the method of standard
additions.
  The detection limit of the method is also dependent
on a number of operational variables. For  a sample
volume of 5 ml a concentration factor of about  1000
over a direct aqueous injection is usually possible. This
places the limit of detection in the 0.1 to 1 microgram
per liter range for a GC/MS system operating in the
selected ion monitoring mode. With a GC/MS system
operating in the continuous repetitive measurement of
spectra   mode,  a  sample  volume  of  25  ml  is
recommended to achieve  this  detection limit.  The
method can be applied over a concentration range of
approximately 0.1 to 1500 micrograms per liter. Figure
3.4 shows a  chromatogram  of a mixture of  29
compounds from a purge and trap analysis.

Sample   Collection   and   Preservation.   Previous
reports emphasized the importance of sample handling,
and indeed because of the  very  volatile nature of the
compounds measured in this type of analysis, sample
collection deserves special consideration. In general,
narrow mouth glass vials with a total volume in excess
of 50 ml are acceptable. The bottles need not be rinsed
or cleaned with  organic solvents, but simply cleaned
with detergent and water,  rinsed with distilled water,
air dried, and dried in a 105C oven for one hour. The
vials are carefully filled with sample to overflowing
(zero head space) and a Teflon faced silicone rubber
septum is placed Teflon face down on the water sample
surface. The septa may be cleaned in  the'same manner
as the vials, but should not be heated more than one
hour because  the  silicone  layer slowly degrades at
105C.
                       faced septa are acceptable if the seal is properly made
                       and  maintained  during shipment. However, several
                       years of experience indicates a success rate significantly
                       less'thant 100% in making proper seals of this type in
                       the field. Therefore simple screw cap vials used with the
                       Teflon faced  septa were evaluated and found to  give
                       equivalent results and a very high rate of acceptable
                       samples. Narrow mouth screw cap bottles with Teflon-
                       faced silicone rubber septa cap  liners are  strongly
                       recommended for sample collection.
                         One special problem in sample preservation has been
                       recognized as a result of the widespread application of
                       this method to drinking waters which contain residual
                       quantities of  disinfectants, e.g., chlorine. The levels of
                       certain  chlorinated  compounds,  e.g., chloroform,
                       found in such waters will vary depending on the time of
                       analysis unless the residual chlorine is consumed by a
                       reducing agent such  as sodium thiosulfate. Table 3.3
                       shows the  concentrations   of four  compounds  in
                       Cincinnati tap water  as a function of the sample age in
                       days. The samples were  taken from the distribution
                       system at the EPA  Environmental Research Center
                       and  maintained  at 4C until  analyzed.  No  reducing
                       agent was added to the samples.
                         The data  in  Table 3.3  shows  that under  the
                       experimental   conditions   used   the  chloroform
                       concentration increased by 114% during the eight  day
                     ,. -storage period. Smaller increases in the concentrations •
                      ' of the brominated compounds were observed, but the
                       carbontetrachloride  concentration did  not  change
                       within the precision' of the  method  which averages
                       approximately 6% in  the 1-1000 ug/1 range. A similar
                       set of samples was  stored at 22C, and the same trends
                       were observed except  the concentrations of each of the
                       halomethanes after seven and eight days were the same
                       within the precision of the method. This indicates that
       Table  3.3   Effect of Residual  Chlorine on  Concentrations of Chlorinated  Methanes in
                   Drinking  Water  at  4C
                 Time,
                 days

                    0
                    1
                    2
                    7
 CHC1,
  ug/1

.  17.0
  17.9
  23.7
  32.6 .
  36.3
CHBrCU
  ug/1

  12.4
  12.9
  16.1
  19.9
  21.6
  Two types of seals for the vials have been employed
and  both give  satisfactory results.  Aluminum, one-
piece, crimp-on seals used with serum vials and Teflon
CHBr;Cl
  ug/1

 11.9
 11.7
 14.4
 17.2
 18.7
ecu
 ug/1

3.1
3.1
3.4
3.4
3.5
                       the rate of hafogenation was reduced substantially after
                       one week of storage  at  22C. The reduced rate of
                       reaction was presumed due  to a  greatly diminished
                                                  33

-------
COLUMN: 0.27. CABBOWAX 1500 ON CARBOPACK-C


PROGRAM: 60°C-3 MINUTES e'/MINUT! TO 160'C



                               S

                               at

                               O
                               Lk

                               O


                               O
                                 RETENTION TIME MINUTES
                        Figure 3.4   Chromatogram of Organohaiides.
                                              34

-------
concentration of active halogenating agents, or the
organic substrates,  or both, but the details were not
investigated further. In a third set  of samples the
reducing agent sodium thiosulfate was added at the
time of sampling to quench the halogenation reaction.
The  concentration  of the  halomethanes in   these
remained constant within the precision of the. method
over the  eight day period.  In these samples no
significant   losses  were   observed  indicating  the
effectiveness  of  the  sample  sealing   procedures
described in this paper.
  The designer of an analytical survey that will include
samples containing  residual 'chlorine or other similar
agents must be aware of these effects. The addition of a
reducing agent at the time and place of sampling will
give a measure of the instantaneous concentration of
halogenated compounds, and this may or may not be
the desired value. A measure of the maximum possible
concentration of halogenated species in the particular
sample may be obtained by storing the sample at 22C or
higher  until  the   concentration  of  halogenated
compounds is constant.


Procedure.  The following  procedure will apply to
both commercially available purging instruments and
home-made units. However, home-made units usually
demand trap manipulation which will not be discussed.
  1.

  i
Adjust the helium flow through the purging
vessel to 40 ml/min.
Newly packed traps must be conditioned at
approximately 230C with  a helium flow of 40
ml/min for 16 to 24 hours. Each day  before
use, the trap should be conditioned at 180C for
10 minutes during a backflush with helium at
30 ml/min.
Inject 25 ml of sample into a 25 ml purging
device with a syringe. A sample volume of 25
ml is recommended when GC/MS is used for
identification  and  measurement  of  the
purgeable compounds. For more concentrated
samples, a 5 ml volume can be used.
Purge the sample for 12 minutes at 40  ml per
minute at room temperature.
After the 12 minute purge time, desorb the
trapped  compounds onto the GC column
which has been cooled to room temperature
for 10 minutes. Desorption is accomplished by
rapidly heating the trap  to   180C  while
backflushing with helium  at 30 ml per minute
for 4 minutes.
  6.    When the transfer is complete, heat the GC
       column  rapidly  to  the  initial  operating
       temperature  for  analysis.   Temperature
       programming at a suitable rate (4-8 degrees
       per minute) should be used to obtain sufficient
       resolution between  GC peaks and accurate
       mass spectra.
  7.    Either the control mode or the IFSS mode may
       be used for data acquisition.
  The sample dialogue below contains some suggested
operating parameters.

   SELECT  MODE:  CONTROL
   CALIBRATE?:   N_
   TITLE:   UP TO 64 CHARACTERS
   CALIBRATION FILE  NAME:   CURRENT
       FILE
   FILE NAME:   UP  TO SIX  CHARACTERS
   MASS RANGE:   20-260
   INTEGRATION TIME:    \1_
   SAMPLES/AMU:  J_
   THRESHOLD:   press return
   RT  ON  CRT?:   N_
   RT  GC  ATTEN:  1  -  8
   FAST SCAN OPT?:   N_
   MS  RANGE SETTING?:   H_
   MAX RUN TIME:   30
.   DELAY  BETWEEN  SCANS  (SECS.)?:
       press return .

   SELECT  MODE:,, IFSS
   CALIBRATE?:   N_
   TITLE:   UP TO 64 CHARACTERS
   CALIBRATION FILE  NAME:
       CURRENT FILE
   FILE NAME:   UP  TO SIX  CHARACTERS
   MASS RANGE:   20-260
   SAMPLES/AMU:  J_
   MAX RPT  COUNT:   32
   BASE INTEGRATION  TIME:   j_
   RPT COUNT BEFORE CHECKING
       LOWER THRESHOLD:  8_
   LOWER THRESHOLD:   4_
   UPPER  THRESHOLD:   4_
   RT  ON  CRT?:   N_
   RT  GC  ATTEN:  1  -  8
   FAST SCAN OPT?:   N_
   MS  RANGE SETTING?:   H_
   MAX RUN TIME:   30
   DELAY  BETWEEN  SCANS-'(SECS.)?:
       press return
                                             35

-------
QUALITATIVE HEADSPACE ANALYSIS
  Headspace analysis is a method of measurement for
compounds that have a relatively high vapor pressure
over a water sample. The  primary purpose of the
method  is  qualitative,  analysis  because  accurate
concentration  measurements   require   extensive
information about the equilibria of many compounds
between the vapor and dissolved phases. The principal
advantage of the  method is  that it permits  the rapid
identification of very volatile compounds, e.g., vinyl
chloride, fiuorocarbons, etc., that may be missed with
some other methods.
  Samples should be collected in bottles as described in
the inert gas purge and. trap procedure. However, in
place of the standard screw-on caps, the open top caps
should be used. This permits sampling of the headspace
without removing the  screw-on cap. Glassware for
blanks and samples should be prepared as described in
the purge and trap procedure. In the laboratory, prior
to sample collection, prepare a blank by completely
filling a sample bottle with  low-organic water. Low-
organic water is  prepared by passing distilled water
through an activated carbon column and" purging the
carbon-treated  water  with an  inert gas. The filled
bottles are sealed and shipped to the sampling site along
with the empty sample bottles. While samples are being
taken, the blank bottles are opened and about 2 ml of
water is decanted. The blank bottles are resealed and
returned  to  the laboratory.  These  blanks  will
compensate for residual organics in the atmosphere at
the sampling site as  well  as  contamination  during
shipment and storage. Samples are taken by completely
filling the bottles, then  decanting about 2 ml. The
bottles are capped, shipped at 0-5C, and the samples
analyzed within 48 hours.
  For the analysis, place the sample bottles in a water
bath at 25C and  allow temperature equilibration to
standardize the headspace measurement. During this
thermal equilibration, occasionally agitate the contents
of the bottle to promote equilibrium between the gas
and  liquid phases. The GC columns  suggested for
direct  aqueous   injection  in   Table 2.7  are  also
recommended for  the qualitative headspace analysis.
  Flush a valve-controlled gas tight syringe twice with
helium. To sample the headspace, pull 2  ml of helium
into  the  syringe,  close the valve, pierce the  sample
bottle septum, and inject the helium into the headspace,
of the sample bottle. Without removing the needle from
the  septum, pull  the syringe barrel out to  the 5 ml
mark, close the valve, and withdraw the syringe from
the   septum.  To  inject into  the  chromatograph,
compress the gases in  the syringe against the closed
valve to about 2 ml, insert the'needle into the injection
block, and open the syringe valve. After completing the
injection of a compressed plug of gas, close the valve
and withdraw the syringe. Initial column temperature
should be as near ambient as possible. A holding time
of several minutes is suggested followed by temperature
programming at 4-8 C/min. A sample run  time of
about 30 min is usually 'sufficient. Suggested sample
data acquisition parameters are as follows:

   SELECT MODE:  .CONT
   CALIBRATE?:   N_
   TITLE:   UP TO  64 .CHARACTERS
   CALIBRATION  FILE  NAME:
      CURRENT  FILE
   FILE  NAME:   .UP TO  6 CHARACTERS
   MASS  RAN.GE:   14-16; 19-27:29-31;
      33-260
   INTEGRATION TIME   T7;17;17;17
   SAMPLES/AMU:    1;'1;1;1
   THRESHOLD:   press  return
   RT ON  CRT?:   N^
   RT GC  ATTEN:   1  -  8
   FAST SCAN OPT?:    N_
   MS   RANGE SETTING?:  H_
   MAX RUN  TIME:  30
   DELAY  BETWEEN  SCANS  (SECS.)?:
      press return
   SELECT MODE:   IFSS
   CALIBRATE?:   N_
   TITLE:   UP TO  64 -CHARACTERS
   CALIBRATION  FILE  NAME:
      CURRENT  FILE
   FILE  NAME:   UP TO 6  CHARACTERS
   MASS  RANGE:    14-16;19-27;29-31;
      33-260
   SAMPLES/AMU:   1;1;1;1
   MAX RPT  COUNT:   32
   BASE INTEGRATION TIME:   J_
   RPT  BEFORE  CHECKING  LOWER
      THRESHOLD:  8_
   LOWER  THRESHOLD:  4_
   UPPER  THRESHOLD:  J_
   RT ON  CRT?:   N_
   RT GC  ATTEN:   1  - 8
   FAST SCAN  OPT?:   N_
   MS RANGE SETTING?:   H_
   MAX RUN  TIME:  30
   DELAY  BETWEEN SCANS  (SECS.)?:
      press return
                                             36

-------
EXTRACTION  WITH  A LOW BOILING
SOLVENT         '    ••
  Liquid-liquid extraction with a low boiling solvent is
a time honored technique  in organic  analysis.  The
principal advantages of this method are its applicability
to a  broad  variety of  compounds and the  large
concentration   factors  that  may  be   realized.   A
disadvantage is  that very  volatile compounds, e.g.,
chloroform, vinyl chloride, etc., will not be observed as
they are either lost during extract concentration or
masked during solvent elution from the GC.
  The scope, limitations, and detection limits of the
method are a function of the solvent selected. Over the
years  a great variety of solvents have been used. The
most  popular low boiling  solvents for liquid-liquid
extraction are diethyl ether,  hexane, benzene,  ethyl
acetate,  carbon  tetrachloride,   chloroform,   and
methylene chloride. Each of these  solvents may be the
optimum for a given situation, but methylene chloride
is the only  solvent  recommended for general purpose,
broad spectrum extractions in this section. Among all
the solvents available it has the  maximum number of
favorable properties.
  Diethyl ether is a poor choice because it invariably
contains peroxides and these must be removed before
its'  use.  However, air  oxidation of  diethyl  ether
produces more peroxides and, even immediately after
purification, sufficient concentrations are present-.to,
react  with  trace organics and  solvent molecules  to,
create an  unacceptable  reagent  blank. On  balance
diethyl  ether, a hazardous flammable substance,  is
simply not of sufficient value to justify its continued
use.
   Hexane,  benzene, and  analogous   hydrocarbon
solvents are insufficiently polar  to extract efficiently a
wide variety of polar organics that are amenable to gas
•chromatography. This makes them unsuitable for the
broad spectrum approach. Ethyl acetate is used in the
biomedical field for extractions of drugs from body
fluids, but it  has  not been evaluated sufficiently for
broad spectrum applications in  water analysis. All of
the   above  solvents  have  the  additional  minor
disadvantage of being lighter than water which is an
inconvenience during separation of immiscible phases.
   Carbontetrachloride is a carcinogenic material and
its human inhalation toxic dose is 20 ppm  (central
nervous sytem effects).  In  addition its relatively  low
polarity, relatively high  boiling point,  and relatively
high  molecular weight make it clearly unsuitable for
GC/MS work.
   Chloroform is reported  to be carcinogenic and its
human  inhalation toxic dose (systemic effects)  is
 10 ppm.  By contrast,  methylene  chloride  is  not
carcinogenic  and  the  corresponding  toxic  dose. for
methylene chloride is 500 ppm (central nervous system
effects). Of particular concern to GC/MS users  is the
tendency of both of these chlorinated solvents to pass
through  the  sample/carrier  gas  enrichment device.
This causes the development  of excessive pressures in
the mass spectrometer which  may  bring  about an
automatic shut-down.  One solution  is  the  use of a
solvent venting valve. However if this is  not available,
the replacement of the chlorinated solvent during the
concentration of the solvent extract is recommended.
Methylene chloride (bp  41C) is readily replaced by
acetone (bp  56C) but chloroform (bp 62C) is not. Up
to 8 ul of acetone may be injected into the GC/MS
without recourse to a venting valve. The lower boiling
point of methylene chloride gives it a major advantage
over chloroform in reducing losses of volatile organics
during extract concentration. Finally chloroform is not
suitable for methylations with diazomethane whereas
methylene  chloride  is  a   preferred   solvent  (see
section 3.5).
  On balance, methylene chloride is the solvent of
choice for  low  boiling solvent  extractions.  It  is
commercially available in very pure form, contains no
peroxides, has a very low boiling point, a low toxicity,
and a  polarity sufficient  to extract a wide variety of
polar organics!  It is very insoluble in  water,  easily
replaced by acetone, and is generally recognized by
practicing chemists as the equivalent of chloroform for
solvent  extractions.   Some  representative   single
laboratory recovery data of spiked organics in distilled
water by solvent extraction with methylene chloride
are  shown   in Table 3.4.  Additional classes of
compounds  not  included  in Table 3.4 have   been
observed by  liquid-liquid extraction  with methylene
chloride.  These  include  aliphatic  hydrocarbons,
benzene  derivatives, organic  phosphorous pesticides,
some carbamates,  alcohols, aldehydes, ketones,  some
carboxylic acids, esters, nitriles, sulfur compounds, and
many others.
  The  detection  limits  for  .methylene   chloride
extractions depend on the size of the water sample, the
extent  of concentration of the extract, the extraction
efficiency  of specific substances,  and  the  GC/MS
behavior of the specific substances. With a three liter
water  sample and concentration  of the  extract to
100 ul, the  usual  detection limit is in  the  range of
10-50 ng/1   (10-50 ppt)  for  compounds  with  a
favorable extraction efficiency (> 80%) and favorable
GC/MS behavior.
                                                  37

-------
Table 3.4.  Recoveries of Organics by Solvent Extraction with Methylene Chloride







           Compound Added.,  Spike Con.,  ug/1  pH' % Recovery
Chlorinated Hydrocarbons
Lindane
Heptachlor
DDD-p,p'
DDT-p,p'
Endrin
Toxaphene
Aroclor 1016
Aroclor 1254
hexachloroethane
p_-dichlorobenzene
bis(2-chloroisopropyl)ether
bis(2-chloroethyl)ether
Chlordane (technical)
Aldrin
Heptachlor epoxide
DDE-p.p'
Dieldrin
Aroclor 1242
hexachlorobenzene
m-dichlorobenzene
hexachlorocyclopentadiene
hexachloro- 1 , 3-butadiene
1 -chloro— 4— nitrobenzene
Phenols
phenol
dimethylphenol
2,4,6,-trichlorophenol
pentachlorophenol
2-nitrophenol
o-chlorophenol

2
2
2
2
2
2
2
2
50
2
25
25
2
2
2
2
2
2
<^
z
2
2
25
2

5,25
2
2
.5
25
25

N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
. N
N
N
N
N

A.A + N
N
N
A
N
N

100
90
97
98
105
106
75
80
45
103
84
88
88
94
95
104
84
84
84
70
55
69
80

52,59
75
53
65
52
30
Polycyclic Aromatic Hydrocarbons
naphthalene
acenaphthylene
acenaphthene
pyrene
fluoranthene
chrysene
benzofa]pyrehe
0.5
25
0.5,25
0.5
2
2
2
N
N
N,N
N
N
N
N
86
92
101,92
104
60
73
76
                                    38

-------
Table 3.4.   Recoveries of Organics by Solvent Extraction with Methylene Chloride (continued).



                     Compound Added   Spike Con., ug/1 pH' % Recovery

                   Polycyclic Aromatic Hydrocarbons (continued)

                   indeno[l,2,3-cd]pyrene          25        N     59
                   phenanthrene                   2        N     72
                   anthracene                      2        N     88
                   dibenzo[a,h]anthracene          25        N    104
                   benzo[g,h,i]perylene             25        N    100
                   benzo[a]anthracene          .25        N    100
                   fluorene                        2        N     87

                   Nitrogen Compounds

                   pyridine                       100         B     92
                   nitrobenzene                   100         B     98
                   o-nitrotoluene                 25        N     88
                   aniline                        100         B     81
                   o^methylaniline                100         B     97
                   di-n-butylamine               100         B     20
                   benzidine                      20        N     57
                   N-nitrosodimethylamine        20    A,B,N   0,0,0
                   N-nitrosodiethylamine          20        N     49
                 •  N-nitrosodibutylamine         • '20        N     73
                   2-benzothiazole                25        N     95
                   hydrazobenzene                25        N     84

                   Oxygen  Compounds

                   n-hexanol                     25        N     64
                   camphor                      25        N     89
                   alpha terpineol                 25    '    N     82
                   isobutyric acid                 25        A       5
                   isovaleric acid                .25        A     17
                   n-octanoic  acid                30      N,A   86,15
                   methyl palmitate               25      N,A   72,31
                   n-tributyl phosphate           25        N    105
                   di(2-ethylhexyl)phthalate         2        N     95
                   dimethyl phthalate .             2        N     75
                   diethyl phthalate                2        N     87
                   di-n-butyl  phthalate            2        N     92
                   isophorone                     2        N    100

                   'N = pH 6-8;  A = pH  2^t; B=pH 10-12.
                                           39

-------
  For accurate quantitative analysis  as  well  as to
obtain  additional  qualitative  information, known
quantities of compounds should be dissolved in acetone
and aliquots mixed  with sample water. This spiked
sample is  extracted according to  the  procedure to
determine the amount recovered and to verify that the
compound was extracted in the correct pH fraction.
  For each, set of samples a reagent blank is required. A
reagent blank is defined as an experiment that uses all
procedures, quantities of materials, and glassware used
in sample preparation except that no water or aqueous
solution  is  used.   It   is   required   even   when
contamination  from glassware and reagents is  well
controlled.   The   reagent   blank   result  is  the
documentation  that proves that  good  control  was
exercised,   and it  defines  precisely   the  level of
background that was beyond control.


Reagents  and  Equipment.  Redistill the methylene
chloride and acetone in an all-glass laboratory  still
before use.  Mesityl  oxide (bp 129C)  and diacetone
alcohol (bp 168C)  are   common  contaminants  in
acetone. The entire volume of the still pot should not be
distilled to prevent contamination of the distillate with
these higher boiling  impurities. Also the pot residue
should be discarded frequently.
  Distillation of solvents is particularly important for
analyses of relatively  clean  water,  for  example,
drinking water  or water from a clean lake.  Distillation
of methylene chloride and acetone (if used) should be
carried out immediately  before  use  since  these
compounds begin to  develop low level  impurities
within hours even when stored  in the dark under
nitrogen.  For  waste  effluents,  the best  available
commercial solvents  may be adequate for use without
redistillation. This should  be verified by GC/MS
analysis of a reagent blank prior to the extraction of the
samples.
  Purify the sodium sulfate by  extraction with 50
volume percent methylene  chloride - acetone  in  a
soxhlet extractor for 8 hours. Air dry the sodium
sulfate and heat in a 120C oven for 4  hours.
  Standard  laboratory glassware  is required for  this
procedure. Extractions  of 3 liter  samples require  a
6 liter  separatory funnel  with a  Teflon stopcock.
Kuderna-Danish  (KD)  glassware  is  used for  the
concentration  of extracts.  A  standard  set  of KD
glassware consists of a 3  ball Snyder column, a 500 ml
evaporating  flask,  and   a   10 ml  receiver  ampul
graduated   in  0.1 ml  increments.  An  all  glass
laboratory still and a  Soxhlet extractor are required for
purification of reagents.
Procedure. The following  procedure is  scaled to a
3 liter sample. This is most appropriate for relatively
clean water, for example drinking or clean lake water.
For  effluents  and similar  water  containing  higher
concentrations, the quantities of sample and reagents
should be pealed accordingly. This  procedure was
designed to  isolate  the  neutral,  acidic, and  basic
compounds in three separate extracts, and the data in
Table 3.4 was obtained in this way. An alternative used
in the EPA sampling  and  analysis procedure  for
screening of industrial effluents for priority pollutants
is to combine the basic and neutral  compounds in a
single fraction by initial adjustment of the pH to 11 or
greater. This reduces the workload by one-third in that
two rather than  three extracts must be processed,  but
other complications that may result from this short-cut
are not well established.
  Glassware for blanks, spikes, and samples should be
washed with detergent, rinsed with tap water, rinsed
with distilled water,  air dried and heated in a muffle
furnace at a minimum temperature  of 300C for  one
hour. One gallon jugs for samples should be cleaned
using the same  procedure. However, these jugs  are
usually made of a soft glass and should not be heated as
rigorously as Pyrex  glassware. Heating  to  300C for
fifteen minutes is recommended. The sample jugs must
be supplied with  Teflon cap liners.
  1.    Measure the pH of  the  gallon  sample and
       transfer  3 liters to a  6 liter separatory funnel. If
       the pH is less than  6 or greater than 8, adjust
       the  pH  to  6-8 with concentrated HC1 or
       NaOH.  Extract the sample  with 125 ml  of
       methylene chloride  with vigorous shaking for
       3 minutes. Allow the layers to separate, then
       drain  the  methylene  chloride layer into a
       suitable  collection  flask.  Repeat twice with
       75  ml portions of methylene chloride and
       combine the three extracts.
       If  an   emulsion   develops,   a   number   of
       techniques may be applied to resolve the phase
       boundary. Additions of reagents or heating the
       mixture  is not recommended as these actions
       may cause more problems than they solve. One
       effective emulsion  breaking  technique is  to
       drain off the emulsion, swirl several times in
       the drain flask, and pour the mixture gently
       through  the water in the separatory funnel. If
       this produces  a  partial methylene chloride
       layer,  separate the  clear methylene  chloride,
       and  pour the remaining emulsion  through  the
       water until a satisfactory separation of layers is
       accomplished.
                                                  40

-------
Pour  the  combined extracts  through  two
inches  of  anhydrous sodium  sulfate  in a
19 mm I.D.  glass column. Collect the dried
extract in  a  500-ml  Kuderna Danish (KD)
flask fitted with a 10-ml ampul. A reagent
blank, treated exactly like the sample extracts,
should be run along  with a sample or set of
samples.
After  the  combined  extract   has  filtered
through the sodium sulfate, rinse the sodium
sulfate with 50  ml of acetone. This is done to
rinse any residual sample from the sodium
sulfate, and  to introduce a nonchlorinated
solvent into the sample for GC/MS injection.
If the  GC/MS system  is equipped with a
solvent venting valve and it is known  that
methylene  chloride  injections  are   well
tolerated, the acetone addition  may not be
necessary.
Adjust the pH of. the water sample to 2 using
concentrated HC1 and repeat steps 1, 2, and 3
this time  using 75 ml of methylene chloride
for  each   extraction.   However,  if  a
derivatization procedure is to be employed, for
example the methylation of the acid fraction to
convert carboxylic acids to methyl esters, see
section 3.5 first.
Adjust the pH of the  water sample to 12 using
saturated NaOH solution and repeat steps 1,2,
and 3 as in step 4. Three sample extracts are
now  contained in   three  KD flasks:   the
neutral compounds extracted from a solution
of pH 6-8, the acid compounds extracted from
a solution of pH = 2, and the basic compounds
extracted from a solution  of  pH=12.  The
reagent blank is in a fourth KD flask.
Fit a 3 ball Snyder column to each Kuderna-
Danish  (KD)   flask and  concentrate  the
extracts on a  steam bath to approximately
5  ml.  After  concentration the  methylene
chloride (bp=41) will be completely removed
and the sample will  be contained in acetone
(bp=56).   The  extract  can  be  further
concentrated in the graduated ampul  in a
warm water bath under a stream of clean, dry
air or nitrogen.  Concentration of the extract to
100 ul or less is entirely possible with very low
losses. However, successful completion of this
final concentration requires tender loving care!
The temperature of the water bath must be no
higher than  50C, the stream of gas must be
truly gentle, and the inside walls of the ampuls
     must be repeatedly rinsed with small quantities
     of pure solvent.
7.    Install  a conditioned  column  in  the gas
     chromatograph. The column packing can be
     any one  of a number of solid supports coated
     with a suitable silicone oil. Table 2.6 contains
     a number of suggested GC columns. Section
     6.3  contains  information  on  open  tubular
     columns. Temperature programming should
     be  utilized with a  relatively  low  initial
     temperature to enhance resolution of more
     volatile components, and a final temperature
     that does not  exceed the maximum operating
     temperature of the column.
     Inject an appropriate volume of extract into
     the system with the ionizer off and either vent
     the solvent or  allow the solvent to be pumped
     from the manifold. After about two minutes,
     or when the pressure reaches 10"' torr, turn on
     the ionizer, start data acquisition, and begin
     temperature   programming.   Halt    data
     acquisition  after  about  40 minutes or  until
     peaks cease to appear. Either the control mode
     or the IFSS mode may be utilized. The sample
     dialogue below  contains  some  suggested
     operating parameters:

 DELECT MODE:  , CONTROL
 CALIBRATE?:  N_
 TITLE:   ANY  TITLE UP  TO 64
    CHARACTERS
 CALIBRATION  FILE NAME:   CURRENT
    CALIBRATION FILE NAME
 FILE  NAME:   UP  TO  SJX  CHARACTERS
 MASS  RANGE:   40-450
 INTEGRATION  TIME:  8_
 SAMPLES/AMU:   J_
 THRESHOLD:  press  return
 RT  ON  CRT?:   N_
 RT  GC  ATTN:  1-8
 FAST  SCAN  OPT?:   N_
 MS  RANGE SETTING?:   H_
 MAX RUN  TIME:    45
 DELAY BETWEEN  SCANS  (SECS.)?:
     press  return
 SELECT MODE:   IFSS
 CALIBRATE?:   N_
 TITLE:   ANY  TITLE  UP  TO  64
     CHARACTERS
                                        41

-------
    CALIBRATION  FILE NAME:   CURRENT
       FILE
    FILE NAME:   UP TO  SIX  CHARACTERS
    MASS  RANGE:   40-450
    SAMPLES/AMU:   J_
    MAX RPT  COUNT:  32
    BASE  INTEGRATION TIME:   J_
    RPT COUNT BEFORE  CHECKING  LOWER
       THRESHOLD:   8_
    LOWER THRESHOLD:   4_
    UPPER THRESHOLD:    4_
    RT  ON  CRT?:    N^
    RT  GC  ATTEN:   1-8
    FAST  SCAN OPT?:   N_
    MS RANGE  SETTING?:   H_
    MAX  RUN  TIME:   45
    DELAY  BETWEEN  SCANS  (SECS.)?:
        press  return

EXTRACTION WITH A  HIGH
BOILING  SOLVENT
  Extraction  with a  high  boiling  solvent  is the
application of a relatively  high molecular weight late
eluting  solvent  to  the  conventional  liquid-liquid
extraction technique. This method has the advantage of
permitting the analysis of volatile compounds that are
usually lost during the concentration of a low boiling
solvent  extract or masked by  the solvent during gas
chromatography. The extraction is accomplished with
a relatively small volume of high boiling solvent and a
relatively large  volume  of water. The extract  is
analyzed without further concentration, and no volatile
compounds are lost. The late eluting character of the
solvent permits data acquisition to begin immediately
after injection, and no volatile compounds are masked
by the solvent.
  The scope of  the high boiling solvent extraction
procedure is similar to inert gas purging and  trapping.
However, high  boiling solvent extraction  has been
much  less  thoroughly  evaluated.  Very   limited
information is available about the types of compounds
extracted with various high boiling solvents,  detection
limits, or the precision and accuracy of concentration
measurements.
  High  boiling solvent  extraction has the advantage
that no special  equipment is  required.  Under these
circumstances  the method should  be  of interest
primarily for rapid qualitative analysis.
  A number of  solvents  appear  suitable  for this
application   but  hexadecane  has  been  used most
frequently. Hexadecane is available commercially  in
rather pure form, it is very insoluble in water, and it
elutes very late on many GC columns. Hexadecane
does have the disadvantage of low. polarity and is best
suited for the extraction of non-polar compounds and
mixtures  such as aliphatic hydrocarbons,  benzene,
toluene,  the  xylenes,  gasoline,  aviation  jet fuel,
kerosene, carbontetrachloride, chloroform, and related
materials. The  extraction  efficiency will vary with
different classes of compounds and any concentration
measurements must  be considered as minimum until
recoveries of spikes clearly establish the capabilities of
the method. The minimum detectable level for some
compounds is about 1 ug/1 when  analyzing a 10 ml
extract  of a  one liter water  sample.  The extraction
efficiency has been found to be 85-95% for halogenated
methanes and aromatic hydrocarbons such as benzene,
toluene, and xylenes.
  As always, reagent blanks  must be  employed  to
establish that contamination is well controlled and the
precise level of contamination that is beyond control. It
should  be verified by GC/MS that  each  batch  of
hexadecane or another solvent is free of significant
quantities of interferring substances.  Purify sodium
chloride and sodium sulfate  by  extraction with  50
volume per cent methylene chloride -  acetone  in  a
soxhlet  extractor for 8 hours.  Air dry the solids and
heat in a 120C oven for 4  hours.
   Samples are  collected in one-liter, glass jars with
 Teflon lined caps. Extracts are collected in 15 ml vials
 with Teflon  faced septa and  aluminum crimp-on or
 screw cap seals. This is required to prevent loss of
 volatile compounds. Vials are available from the Pierce
 Chemical  Company, Box 117.  Rockford,  Illinois
 61105. No other special equipment is required
   Glassware for blanks, spikes, and samples should be
 washed with detergent, rinsed with tap water, rinsed
 with distilled water, thoroughly dried, and heated in a
 muffle  furnace for  at least one hour at a minimum
 temperature of  200C.
  Add 10 ml of hexadecane to a one liter sample bottle
and collect about  950 ml of water sample. Cap  the
bottle and shake vigorously. If the water temperature is
18C or below, the hexadecane will solidify. In this case,
wait  until the hexadecane melts before shaking. Ship
the sample to the laboratory without refrigeration.
  If phenols or other low molecular weight acids are of
interest,  adjust  the pH of  the  sample to 2 with
concentrated  sulfuric acid. Shake the sample well and
transfer the contents to a 2  liter  separatory  funnel.
Shake vigorously for 2 minutes and allow the layers to
separate. If an emulsion occurs, add 5  grams of sodium
chloride, reshake, and allow the layers to separate.
                                                42

-------
  Drain off the water layer and measure its volume in a
1000 ml graduated cylinder. The added NaCl will not
increase the volume by a significant amount. Drain the
hexadecant- through a small filter funnel packed with
2 grams of anhydrous sodium sulfate over glass wool
into a 15 ml vial. Seal the vial by crimping on a Teflon
faced septum and an aluminum seal or use a screw-on
cap. The extract is now ready for GC/MS analysis.
  Suggested gas chromatographic columns  for high-
boiling solvent extract analysis are listed in Table 2.6.
Temperature programming from  ambient  to about
200C at 8 degrees per minute may be utilized. Begin
with GC  oven door  open  to  achieve  an initial
temperature of about  50C. Turn on the ionizer, start
data acquisition, and then inject 2-8  microliters of
extract. Close the oven door  and begin temperature
programming. Halt data acquisition before elution of
the hexadecane peak (generally  about 10  minutes).
Either the control mode or the IFSS mode may be used
for data  acquisition. The sample  dialogue below
contains some suggested operating parameters:

  SELECT MODE:   CONTROL
  CALIBRATE?:   N_
  TITLE:   UP  TO  64  CHARACTERS
  CALIBRATION  FILE NAME:  CURRENT  FILE
  FILE NAME:  UP  TO SIX  CHARACTERS
  MASS  RANGE:   20-260                   .
  INTEGRATION  TIME:   8_
  SAMPLES/AMU:  _1_
  THRESHOLD:   press  return
  RT ON CRT?:    N_
  RT GC ATTEN:    1-8
  FAST  SCAN  OPT?:   N_
  MS  RANGE SETTING?:    H_
  MAX RUN  TIME:   20
  DELAY BETWEEN  SCANS  (SECS.)?:
    press  return
  SELECT MODE:   IFSS
  CALIBRATE?:   N_
  TITLE:   UP  TO  64  CHARACTERS
  CALIBRATION  FILE NAME:   CURRENT FILE
  FILE  NAME:   UP  TO  SIX  CHARACTERS
  MASS  RANGE:   20-260
  SAMPLES/AMU:   j_
  MAX RPT COUNT:   32
  BASE INTEGRATION  TIME:  J_
  RPT  COUNT  BEFORE  CHECKING  LOWER
    THRESHOLD:   8_
  LOWER THRESHOLD:   4
   UPPER THRESHOLD:   4_
   RT  ON  CRT?:   N_
   RT  GC  ATTEN:   1-8
   FAST SCAN OPT?:   N_
   MS RANGE  SETTING?:   H_
   MAX  RUN  TIME:   20
   DELAY  BETWEEN  SCANS  (SECS.)?:
       press  return
ADSORPTION WITH POROUS
POLYMERS
  Adsorption with porous polymers is the application
of a solid phase, organic polymeric material to the
isolation and concentration of organic compounds in
water. The principal advantages of this technique are
the acquisition of time integrated (composite) samples
and the very low detection  limits made  possible by
sampling relatively large volumes of water..
  In  the  past,  activated  carbon   has  been used
extensively as an adsorbent for organic compounds in
water.  However  the carbon   adsorption  method,
although still widely used, has a number of limitations.
High quality activated carbon  is not always  readily
available, significant  effort  is  required  to  reduce
background, lengthy work-up procedures are required,
and many organic compounds are adsorbed so strongly
thatdesorption is accomplished in very low yield.
  The  porous polymer adsorption  method  has the
same advantages as the carbon adsorption method, but
fewer disadvantages. A variety of commercial polymers
are available, background contamination is more easily
removed, desorption does not require lengthy Soxhlet
extraction,   and  the recoveries in  Table 3.5  are
acceptable  for measurements of a wide variety of
compounds. These recoveries are based on elution with
diethyl ether and were selected from the literature cited
in section 10.3.
  The porous  polymers employed in this  method are
called macroreticular resins.  Macroreticular refers to
the relatively large, controlled  pore size of the resin
beads.  Each grain of resin is formed from many
microbeads  that are  cemented  together  during the
polymerization process.  The  pores are the spaces left
between the cemented microspheres.
  The most-frequently used resins are the Amberlite
XAD  series made by the Rohm and Haas Company.
They are hard insoluble beads of 20-50 mesh, varying
from white to light brown in color. Four materials are
available, XAD-2, 4, 7 and ». XAD-2 and XAD-4 are
styrene-divinylbenzene copolymers that are chemically
identical but differ in active surface area and average
                                             43

-------
Table 3.5.  Recoveries of Organics by Adsorption on Porous Polymers.




     Compound Added      XAD Resin Type   %Recovery
Chlorinated Hydrocarbons
sym-tetrachloroethane
bis-chloroisopropyl ether
bromodichloromethane
dibromochloromethane
chlorobenzene
o-dichlorobenzene
m-dichlorobenzene
1,2,4,5-tetrachlorobenzene
1 ,2,4-trichlorobenzene
benzyl chloride
o-chlorobenzyl chloride
m-chlorotoluene
2,4-dichlorotoluene
p-chloronitrobenzene
Lindane
Aldrin
Dieldrin
Chlordane
DDT
DDE
Phenols

phenol
o-cresol
p_-cresol
3,5-xylenol
o-chlorophenol
g-chlorophenol
2,4,6-trichlorophenol
pentachlorophenol
1-naphthol

2, 4, 7, 8
2, 4, 7, 8
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2 •


2, 4, 7, 8
4
2, 4, 7, 8
4
4
4
4
2, 4, 7, 8
4

61, 72,
76, 80,
87
99
95
88
93
74
99
88
96
80
71
100
95
47
93
82
96
.81-

9
14, 40,
73
44, 69,
79
96
95
99
84, 84,
91

35, 59
76, 77




















19, 29

33, 47




83, 77

Polycyclic Aromatic Hydrocarbons
naphthalene
1 -methylnaphthalene
2-methylnaphthalene
2-methoxynaphthalene
acenaphthene
biphenyl
fluorene
anthracene
tetrahydronaphthalene
dibenzofuran
diphenyl ether
2, 4, 7, 8
2, 4, 7, 8
2, 4, 7, 8
2
2, 4, 7, 8
2
2
2
2
2, 4, 7, 8
2
79, 80,
87, 77,
95, 77,
97
99, 81,
101
84
83
62
93, 82,
91
64, 78
64, 80
63, 77

72, 20




73, 95

                             44

-------
Table  3.5.  Recoveries of Organics  by Adsorption  on Porous  Polymers, (continued)

              Compound Added      XAD Resin Type     % Recovery


              Nitrogen  Compounds

              o^nitrotoluene              2, 4, 7, 8       82, 83, 53, 77
              nitrobenzene               2               91
              o-nitroaniline              2               100
              hexadecylamine             2               94
              indole                     2               89
              N-methylaniline            2               84
              quinoline                   2               84
              isoquinoline                2               83
              benzonitrile                2               88
              Atrazine                   2, 2             83, 100
              benzothiazole              2, 4, 7, 8       100, 82, 40, 53
              benzoxazole                2               92


              Carboxylic Acids

              octanoic acid              4               108
              decanoic acid              4               90
              palmitic, acid              4               101
              oleic  acid                  4               100
              benzoic acid               4               107

              Alcohols

              1-hexanol                  2               93
              2-octanol                  2               100
              1-decanol                  2               91
              benzyl alcohol              2               91
              cinnamyl  alcohol           2               85
              2-phenoxyethanol          2               102
              alpha terpineol             2, 4, 7, 8       81, 80, 36, 62
              2-ethylhexanol             2, 4, 7, 8       99, 91, 74, 79
              1-dodecanol               2               93


              Aldehydes and Ketones

              2,6-dimethyl-4-heptanone   2               93
              2-undecanone              2               88
              acetophenone              2               92
              benzophenone              2               93
              benzil                     2               97
              benzaldehyde     *-'        2               101
              salicylaldenyde             2               100
              isophorone                 2, 4, 7, 8       76, 86, 46, 47
                                        45

-------
        Table  3.5.  Recoveries of Organics by Adsorption on  Porous  Polymers, (continued)

                        Compound Added   XAD Resin Type    % Recovery
                        Esters

                        benzyl  acetate
                        di(methoxyethyl) phthalate
                        dimethyl  phthalate
                        diethyl phthalate
                        dibutyl phthalate
                        di(2-ethylhexyl) phthalate
                        diethyl fumarate
                        dibutyl fumarate
                        di(2-ethylhexyl) fumarate
                        diethyl malonate
                        methyl benzoate
                        methyl decanoate
                        methyl octanoate
                        methyl palmitate
                        methyl salicylate
                        methyl methacrylate

                        Miscellaneous

                        n-hexadecane
                        ethylbenzene
                        cumene
                        g-cymene
                        dihexylether
                        dibenzylether
                        anisole
                        bromoform
                        iodobenzene

pore diameter.  XAD-2 has  300 mVg  active  surface
area and 90 A average pore diameter compared to
784 mVg and 50A for XAD-4. XAD-7 and XAD-8
are both acrylate polymers having active surface areas
of 7 50  mVgand!40  mVg and average pore diameters
of 80 A  and 250  A, respectively.  The resins derive
their adsorptive properties from their macroreticular
porosity,  pore size  distribution, and high surface area.
XAD's 2, 4,  and 7 are available from Mallinckrodt
distributors such as Curtin. All four are available from
Chemical  Dynamics  Corp.,  Hadley Road, P.O.  Box
395, South Plainfield,  New Jersey 07080.
  In practice, water  is passed through a column of
XAD resin and the organic materials are adsorbed. The
organics are desorbed from the resin by elution  with an
organic solvent and the extract is submitted to GC/MS
2,  4, 7,
2
2
2
2
2
2
2
i
                100
                94
                91
                92
                99
                88
                86
                92
                84
                103
                101
                95
                98
                70
                96
                35
3,  11,  3, 18
81
93
92
75
99
87
101
81
  analysis. A number  of  elution solvents  have been
  evaluated and methylene chloride and acetone are very
  effective.  Diethyl  ether  and  chloroform  are not
  acceptable for the reasons discussed in the low boiling
  solvent extraction method.
     The method has  been  applied to a  wide variety  of
  organic compounds and  is applicable to non-ionized
  materials having boiling points above about 100C. The
  detection limit is of the order of 1-10 ng/1 in an
  aqueous matrix.  In  general,  overall recoveries are
  equivalent to those  obtained with low boiling solvent
  extraction. Table 3.5 contains a number  of values that
  have been reported using  diethyl ether for elution. For
  accurate concentration measurements,  these recoveries
  must be established in the laboratory conducting the
  procedure.
                                                  46

-------
  As in other methods, a reagent blank is required to
assure that compounds detected are not contaminants
of the glassware or materials. The reagent blank may
consist of one liter of low organic water.
Reagents  and Equipment.  Extract  glass wool  in  a
Soxhlet apparatus with methylene chloride before use.
Methartol, methylene chloride, and acetone should be
of high quality and redistilled before use in all-glass
apparatus.  (See  the low boiling solvent  extraction
method for precautions in the use of these solvents).
Sodium sulfate should be extracted  and dried before
use.
  Low organic water is prepared by passing tap water
through a filtration system consisting of cation and
anion-exchange  columns  and  activated   carbon
columns. This water is then  distilled  in an all-glass
apparatus. The resulting  deionized/distilled  water  is
then passed through an XAD-2 column as a final clean-
up step.
  Some special equipment is required for this method.
The resin  columns  are  custom made  from  liquid
chromatography columns, 25  mm O.D., 22  mm I.D.,
equipped with a Teflon stopcock. The columns are
shortened to about 180 mm (packing length) and fitted
with 24/40 female joints at the top of the column. The
column also should have glass "ears" near the top for
securing the glass stoppers or the glass adapters with
rubber bands  or  wire springs.  Other arrangements are
also acceptable; the main requirement is that the ratio
of length to inside diameter should be about 7. The
stoppers are  24/40 ground glass stoppers  equipped
with glass "ears".
  For automatic continuous sampling, a variable speed
peristaltic  pump capable of pumping at  a rate of
120  ml/min  is used.  The  pump is attached to the
stopcock end of the column with Tygon tubing. The
water sample is  drawn through the  column from the
other end using an adapter and Teflon tubing. The
recommended flow rate is four bed volumes per minute.
  Adapters are  required to connect the column with
1/4 inch O.D.Teflon tubing that is placed in the sample
source. The adapter consists of a 24/40 male joint
joined  to a short section of 1/4 O.D. glass tubing. The
glass tubing is connected to the Teflon tubing by a brass
Swagelok adapter fitted with Teflon ferrules. The glass
adapter is equipped with glass "ears"  for securing in
place rubber bands or wire springs.
   Kuderna-Danish equipment is required for the final
concentration of the eluate. (See the low boiling solvent
extraction procedure for a list of this equipment).
Procedure.  The  commercial   XAD   resins  are
contaminated.  Preextraction  to  remove  interferring
impurities is required.

  1.   Slurry the desired amount of resin in methanol
       and decant to remove the fine particles. Place
       the resin in a Soxhlet apparatus of appropriate
       size using glass wool plugs to hold the resin in
       place.
  2.   Extract the resin  sequentially with acetone,
       methanol, and methylene chloride utilizing a
       minimum often cycles per solvent.
  3.   With the resin still in the Soxhlet apparatus,
       rinse  three  times  with  methanol. Store the
       resin   in  a  closed glass  container  under
       methanol. Do not allow to dry.
          An  alternative  cleanup  procedure  uses
       extraction  with   acetone  and  acetonitrile,
       followed by  vacuum degassing at 200C and
        10"* to 10"' torr.  If a large Soxhlet is not
       available, 100 g batches can be cleaned up by
       placing them in a  5 cm I.D. column and
       washing  successively  with  two  liters  of
       acetone, methanol, and  methylene chloride.
       Concentrate  a  portion  of  the   methylene
       chloride effluent, and check it for background.
       Repeat  the   methylene  chloride  wash   if
       necessary. Remove the resin from  the column
       and store under methanol.:
  4.   Plug  the  bottom  of the resin  column  with
       1-2 cm of pre-extracted  glass wool. Add the
       XAD resiri to the column  as a methanol slurry.
       Adjust the resin height in the column to about
       8  cm. Stir with a glass  rod  to remove any
       trapped air. Plug the top of the column with
       1-2 cm of glass wool. Do not allow the column
       to dry.
  5.   Add 30 ml of low organic water, let drain until
       water reaches the top of the glass wool. Repeat
       three times.
  6.   The column is now ready for a  sample. If
       stored or shipped,  the column should be filled
       with  water  and stoppered such  that all air
       pockets are excluded.
  7.   Grab samples generally consist of 1 to 5 liters
       of water. It may be desirable to adjust the pH
       of the sample prior to passing it through the
       column. For example, the recovery efficiencies
       for phenols and carboxylic acids are greater
       when performed at lower pH. Pass the water
       through columns by gravity at a  flow rate of
       approximately four bed  volumes  per minute
                                                 47

-------
     (ca. 120 ml/min.)- Any sample except drinking
     water may contain suspended matter that is
     trapped on the glass wool and reduces the flow.
     In this case a peristaltic pump at the bottom of
     the column for samples greater than 1 liter is
     suggested. Allow sample to drain freely and
     close the stopcock. If the column is to be stored
     or shipped prior to solvent extraction, fill with
     organic free water,  exclude air bubbles, and
     stopper. Frequently air  bubbles  form in  the
     column during  the adsorption  step; ignore
     them.
8.    'For composite samples,  the columns can  be
     .used for manual compositing. For example, in
     the field,  one liter  of effluent water can  be
     passed through the column every hour over the
     desired sampling period (8-24 hours) as a time
     composite. Composites can,also  be taken  by
     attaching  the column to a peristaltic pump
     having the appropriate pumping speed (ca.  120
     ml/min or less). The water to be sampled is
     drawn to  the column through Teflon tubing
     (1/4  inch O.D.).  The  effluent  end of  the
     column is attached to  the peristaltic pump
     with Tygon tubing. The flow rate should be
     checked periodically to determine the volume
     sampled.
9.    For.desorption,  allow the  column to  drain
     freely. Close the stopcock and place a 300 ml
     Erlenmeyer flask under the column.
10.   Add 30 ml of acetone and force the solvent
     through the  resin  at a  slow rate (dropwise)
     using  nitrogen  pressure (ca.   2 p.s.i.) if
     necessary.
11.   Allow the acetone to drain until it is just to the
     top of the upper glass wool plug.
12.   Repeat steps 10 and 11.
13.   Continue desorption with eight bed volumes of
     methylene chloride.
14.   Remove the water layer from the eluate using a
     disposable glass pipette.
15.   Dry the eluate  with organic-free anhydrous
     sodium sulfate, and concentrate by Kuderna-
     Danish evaporation to a volume of about 5  ml.
     The volume  can be further reduced over a
     warm water bath by directing a stream of clean
     air or nitrogen onto the extract.  (See the  low
     boiling  solvent  extraction  procedure  for
     cautions during concentration). The extract is
     now ready for GC/MS analysis.
16.   Often the resin columns can be reused after a
     minimum amount of cleanup. However, if the
       sample  was extremely dirty and  contained
       sediment, it-is recommended that the resin be
       discarded. If the r'esin is discolored, replace the
       upper glass wool plug. Add 30 ml of acetone
       and allow to drain freely. Repeat the acetone
       wash twice. Add 30 ml of organic free water
       and .allow to drain freely. Repeat  the water
       wash twice. The column is now ready to reuse.
         The GC/MS  conditions are  the same as
       given under the low boiling solvent extraction
       procedure.   Suggested   columns   are  in
       Table 2.6.
3.2 All  SAMPLES

  Air sampling and analysis methods have been under
development for many years, and numerous methods
are available for the isolation,  concentration,  and
measurement of specific  organic compounds in air
samples. However, as emphasized in chapter one, the
purpose of this manual is to describe broad spectrum
methods for general classes of organics. In this  area
there   is   relatively   little  information  available.
Therefore, the methods given in this section probably
represent  the  first  generation  broad  spectrum air
procedures,   and  significant   improvements  will
probably occur in future years.
  The problem  of  quantitative  sampling of air is
significantly  more  complex   than  is  quantitative
sampling of water. Numerous studies of the sampling
problem have been conducted and the results reported.
It is beyond the scope of this manual to reproduce the
detailed   methodology    required   for   accurate
quantitative sampling  of  organics in air. The broad
spectrum methods described in this chapter are most
appropriate  for  qualitative or  rough  quantitative
analyses of air samples.
  After sampling and the initial  sample preparation,
the GC/MS aspects of air analyses are not unlike the
analyses of samples from other media. Therefore, a
great deal of information from the other sections of this
chapter and other chapters is applicable to air analyses.
Numerous references to these sections are made under
Air Samples.

DIRECT  AIR INJECTION
  Direct air injection is the introducion of a  few ml,
e.g., 5, of air into the GC/MS system. This procedure is
clearly similar to the qualitative headspace  analysis
described in section 3.1. The advantage of this method
                                              48

-------
is the speed of the analysis and relative simplicity of
sample handling. The disadvantage is that the method
is limited to relatively high concentrations of volatile
compounds such as vinyl chloride or chlorotrifluroro-
methane. The detection limit is estimated to be about
two milligrams .per cubic meter (kiloliter) of air. This
procedure will probably be useful only for an industrial
work site or a spill site. Generally ambient air levels of
organics are in range of tens of picograms to nanograms
per cubic  meter. Industrial work  rooms or  source
effluents usually do not exceed tens of micrograms per
cubic meter, but many exceptions are possible and have
been reported.  The  application  of selected  ion .
monitoring as described in section 6.1 may lower the
detection  limit of the method to the microgram  per
cubic meter level.
  For this procedure glass collecting tubes of 100-500
ml capacity fitted with gas tight stopcocks and rubber
septa are evacuated in the laboratory. The stopcock is
opened at the sampling site, and sampling is continued
until the pressure in the collecting tube is equivalent to
the external pressure. The stopcock is closed and the
container is returned to the laboratory. An aliquot of
the sample is  removed with a gas  tight  syringe  and
injected directly into the GC/MC system Sampling
containers are  commonly  available  from  general
laboratory supply companies.
  The  GC columns suggested  for  direct, aqueous
injection in Table 2.7 are also recommended for direct
air injection. Initial column temperature should be as
near ambient as possible. A holding time of several
minutes  is  suggested  followed   by   temperature
programming  at 4-8 C/min. A sample  run time of
about 30  min. is usually sufficient.  Suggested  sample
data acquisition parameters are as follows:
  SELECT MODE:   CONT
  CALIBRATE?:  N_
  TITLE:   UP  TO  64  CHARACTERS
  CALIBRATION  FILE NAME:    CURRENT  FILE
  FILE NAME:   UP TO  6 CHARACTERS
  MASS  RANGE:   14-16;19-27;29-31;
    33-260
  INTEGRATION TIME:   17;17;17;17
  SAMPLES/AMU:    1;1;1;1
  THRESHOLD:  press  return
  RT ON CRT?:   N_
  RT GC ATTEN:   1 - 8
  FAST  SCAN  OPT?:   N_
  MS  RANGE SETTING?:   H_
  MAX  RUN TIME:   30
  DELAY BETWEEN  SCANS  (SECS.)?:
    press return
SELECT  MODE:   IFSS
CALIBRATE?:   N_
TITLE:   UP  TO 64 . CHARACTERS
CALIBRATION  FILE NAME:   CURRENT
   FILE                       	
               UP  TO  6  CHARACTERS
                  14-1=3; 19-27:29-31;
   FILE NAME:
   MASS  RANGE:   	
      33-260
   SAMPLES?AMU:   1; 1; 1; 1
   MAX RPT COUNT:  32
   BASE  INTEGRATION TIME:   J_
   RPT BEFORE CHECKING  LOWER
      THRESHOLD:   8_
   LOWER  THRESHOLD:   4_
   UPPER THRESHOLD:   J_
   RT ON  CRT?:   N_
   RT GC  ATTEN     1 - 8
   FAST  SCAN  OPT?:   N_
   MS-RANGE SETTING?:   H_
   MAX  RUN TIME:   30
   DELAY  BETWEEN  SCANS  (SECS.)?:
      press return


ADSORPTION WITH POROUS POLYMERS
  This method involves passing a measured amount of
air through a solid adsorbent,  followed by thermal
desorption  of organics into the GC/MS system. The
advantage  of this method compared  to  cryogenic
trapping is the significantly  simpler field  equipment
handling.  The  advantage,   compared  to solvent
desorption,  is  higher sensitivity  because  the total
sample is measured and there is no background from
the solvent. The principal disadvantage is that because
the total sample is measured, there is no room for error
in the analytical procedure. Multiple samples should be
collected to provide the required back-up in the event
of a laboratory accident or equipment malfunction, and
to obtain information on the precision of the method.
The porous polymer adsorption method is applicable to
organics having compositions  in the C6 to CIS range.
  Although  several  different adsorbents  have been
used by various researchers, Tenax GC appears to have
several advantages over the others.  This  material is
thermally stable to about 350C, has good adsorption
characteristics, but does not  efficiently retain  water
vapor.  However caution must be  excercised because
some batches  of Tenax  have  been  reported  to
decompose at lower temperatures. Detection limits are
not well established for these methods, but probably are
about 1-10 ug per cubic meter"(kiloliter) of air. Good
quality  control  requires  that unused  adsorbent  be
desorbed at frequent  intervals with  the identical
                                              49

-------
conditions used with the samples. This will assure that
background contamination from the adsorbent itself or
from the equipment is under control,  or, at least, is
known to the  investigator. There is limited  recovery
data available for this method.
  Tenax GC (2,6-dlphenyl-p-phenyleneoxide polymer)
is available from several suppliers including. Applied
Science, State  College, Pennsylvania.  This  must  be
purified before use by extracting with methanol for
several  or more  hours in  a  Soxhlet  apparatus.
Purification may not be necessary if purified material is
available  from the supplier,  but  the material  is
susceptible to decay and contamination during storage.
  The Tenax GC, 60/80 mesh adsorbent is packed into
Pyrex glass tubes of 2.5 mm I.D. x 200 mm dimensions.
The adsorbent  is supported by 0.5 cm plugs of silanized
glass wool. Prior to use the tubes and glass wool may be
heated to  500 C for several hours to remove traces of
organic matter.  The  packed sampling tubes  are
conditioned at 270 - 300 C for 0.5 - 1 hour under a
helium  or nitrogen flow  of 30 - 50 ml/min. After
cooling to room temperature, cap both ends or store the
sampling tube in a clean screwcapped test tube.
  Sampling  pumps  are  available  from  several
laboratory supply companies. A flowmeter should be
precalibrated with  a sample tube connected using a
soap bubble^ flowmcter. The sample is taken by pulling
air  through the tube at a rate of about 500  ml/min.
After a suitable sampling period of 2 - 4 hours in urban
locations, the pump is turned off, and the tube is capped
or stored until the analysis can be performed.
  The sample  is analyzed by direct heat desorption at
250 - 270 C for three minutes under a 10 ml helium
flow. A desorption technique similar to that described
under inert gas purging and trapping is recommended.
The GC  column oven  is   reduced  to  ambient
temperature   before   desorption   begins,   then
programmed to an elevated temperature at 4 - 8/min.
The columns described in Table 2.6 for solvent extracts
are recommended for porous polymer trapped samples.
Data  collection  should begin  just   prior  to  the
desorption step. The following GC/MS operating
conditions are recommended for the control and IFSS
modes.

  SYSTEM   1500  IS  ON  SELECT MODE:
     CONT
  CALIBRATE?:  . N_
  TITLE:   UP  TO  64  CHARACTERS
  CALIBRATION FILE NAME:  CURRENT FILE
  FILE NAME:   UP  TO  6_4  CHARACTERS
  MASS  RANGE:   14-16;19-27;29-31;
     33-260
   INTEGRATION TIM'E:   17;1 7;1 7; 17
   SAMPLES/AMU:   1;1;1;1
   THRESHOLD:   press  return
   RT ON  CRT?:  N_
   RT GC  ATTEN:   1 - 8   .
   FAST  SCAN  OPT?:   N_
   MS RANGE SETTING?:   H_
   MAX  RUN TIME:   30
   DELAY  BETWEEN SCANS  (SECS.)?:
     press return
  SELECT MODE:   IFSS
  CALIBRATE?:  N_
  TITLE:   UP  TO  64  CHARACTERS
  CALIBRATION FILE NAME:  CURRENT
      FILE
  FILE  NAME:   UP  TO  6 CHARACTERS
  MASS  RANGE:   14-16;19-27;29-31;
      33-260
  SAMPLES/AMU:   1;1;1;1
  MAX RPT COUNT:   32    BASE
      INTEGRATION TIME:   J_
  RPT  BEFORE CHECKING  LOWER
      THRESHOLD:   8_
   LOWER THRESHOLD:  4_
   UPPER THRESHOLD:  j_
   RT  ON CRT?:   N_
   RT  GC ATTEN:    1  - 8
   FAST  SCAN OPT?:   N_
   MS  RANGE  SETTING?:   H_
   MAX RUN  TIME:   30
   DELAY  BETWEEN SCANS  (SECS.)?:
      press return
FILTRATION WITH GLASS FIBER
FILTERS
  This  method is  for the  analysis  of organic
compounds contained in paniculate matter (dirt, dust,
etc.)  collected  from  air with  a  glass fiber  filter.
Paniculate matter itself is a criteria pollutant that is
known to be rich in organic material. The advantage of
this method is that it specifically addresses the problem
of organics in particulate matter. The disadvantage is
that gas phase, low molecular weight organics are not
adsorbed on glass fibers. In this method air is drawn
into a covered housing and through a glass fiber filter
by means of high flow rate blower (HI VOL sampler),
the  mass  of  the collected  particulate  matter is
determined, the organics are  removed by Soxhlet
                                             50

-------
extraction, and analyzed by GC/MS. The extract may
be separated with silica gel into aliphatic, aromatic, and
oxygenated  fractions  prior  to  GC/MS  analysis.
Generally, for urban particulates, alkanes ranging from
C,,H36 to Cj,H6J will be found in the aliphatic fraction,
compounds   such  as   anthracene,   pyrene,   and
benzopyrene will be found in the aromatic fraction, and
phenols,  phthalate esters and other polar compounds
will be found in the oxygenated fraction.
  Methylene chloride  is  the  solvent of choice for
general  purpose  work. The  reasons  for  choosing
methylene chloride are exactly the same as given under
the low boiling solvent extraction method in section 3.1
of this chapter. The specifications for glass fiber filters
and  the  detailed  procedure for use of a  HI VOL
sampler are  beyond the scope of this manual,  but are
well documented in other publications. The user should
see  the Federal Registered, 8186 (1971). Additional
information  is  available  from the director  of the
Environmental Monitoring and Support Laboratory,
Research Triangle Park, North Carolina.
  The  detection limit for this method is  not well
established,  but is probably about one ng per kiloliter
(cubic meter) of air. Typically a 24 hour sample of 2500
kl in an urban location will  yield 125  - 150 mg of
paniculate matter  (60 - 200  ug/kl)  of which about
10% is extractable. It is absolutely essential  that an
unused glass filter from the same batch of filters be
extracted  at the   same time  and under the same
conditions as the loaded filters. The analysis of this
unused   filter  will  provide   information   about
background  and solvent contamination.  There are no
recovery data available for this method.
  The  methylene  chloride should  be distilled as
described under   low  boiling solvent  extraction
immediately before use.  The weighed filter is cut  into
pieces  small enough  to fit into  a 250 ml  Soxhlet
extractor. Usually only one section of the whole filter is
used for  the organics analysis. No  thimble should be
used, but about one inch of pre-extracted glass wool is
placed in the bottom of the extraction chamber. This
will  act as a filter and preclude the accumulation of
paniculate  matter in the distillation  pot. The filter
pieces are inserted, and 200 ml of solvent is placed in
the pot. Extraction should continue for 25 cycles, and
the cooled extract transferred to a Kuderna-Danish
(K-D) apparatus. If the optional silica gel separation is
to be used, concentrate the extract to 1 ml; if no silica
gel separation is planned, concentrate to 0.1 - 1  ml.
  For the silica gel separation, pack activated silica gel
into a glass  chromatography column (2 cm  I.D.) to  a
height of 10 cm.  To the  1 ml of extract in  a small
beaker add sufficient silica gel to adsorb the sample,
and then evaporate the  residual solvent with gentle
stirring in a hood. Wet the column with about 20 m 1 of
hexane and add the sample when the last of the 20 m 1
reaches the surface of the adsorbent. Rinse the beaker
with hexane and add the rinsings to the column. The
beaker should be rinsed several times with hexane and
each succeeding eluent when the eluent is added to the
column.  Collect the fractions in K-D ampuls and elute
the  aliphatic  fraction with  85  ml  of hexane,  the
aromatic fraction with 85 ml of.benzene or the safer
toluene,  and the polar fraction'with 1:1  methanol-
methylene chloride. Concentrate  each fraction to 0.1 -
1  ml in the K-D  apparatus, and  proceed with the
GC/MS analysis of each.
   The GC columns and conditions  recommended are
the  same  as those  used for  low  boiling  solvent
extraction. This procedure should be consulted for this
information. Also  see the previous section for sample
GC/MS system dialogue.
3.3  SEDIMENT SAMPLES

  Sediment  is defined  as  wet solids taken from the
bottom  of a stream, river or lake.  This  method  is
applicable to these, and with the elimination  of the
drying step, to the broad spectrum analyses  of organics
in dry  soils  and related  matter.  In the  method a
sediment sample is partially dried and extracted in an-
ultrasonic homogenizer (e.g., a Polytron)  with a  1:1
mixture of acetone-hexane. The residue is filtered and
the extract concentrated in a Kuderna-Danish  (K-D)
apparatus. The concentrated extract may be analyzed
directly  by   GC/MS  or  optionally  separated into
fractions with activated silica gel. The purpose of the
acetone solvent is to provide a water-soluble wetting
agent to stimulate desorption of organics from the solid
material in the sample. The advantage of this method is
that  it   uses  a reasonably  well  tested  extraction
procedure and  a familiar, well established  optional
preliminary  fractionation scheme. The detection limit
is not known, but is estimated at about 1-10 ug per kg
of dry solids. Quality  control must  include a  single
reagent blank for each group of samples extracted on a
given day. A reagent blank is the result of processing
through the entire method all the  same quantities of
materials and labware, except that the sediment  itself is
not included. The result of the reagent blank analysis
documents the background and reagent contamination.
   It is not possible to  determine recoveries with this
                                                 51

-------
method since it is not possible to spike a sediment and
simulate natural adsorption conditions. One method of
estimating the  degree of desorption is to re-extract the
processed sediment with fresh solvent or  a  second
solvent pair such as methylene chloride-acetone, and
determine if additional amounts of the same compound
are obtained.
  Purify the reagents and solvents as described under
low boiling solvent  extractions in section 3.1. Standard
laboratory glassware and equipment is required for this
method.  See the equipment section under low boiling
solvent  extractions in  section 3.1. The ultrasonic
homogenizer recommended is a Polytron model PT-
20ST,  Brinkmann Instruments, or the equivalent.
There  also  are  a   number of   ultrasonic  tissue
homogenizers  available and these are reported to give
good  results.  A  stainless  steel   container  is
recommended  for  blending to preclude breakage by
stones or other hard objects.
  Decant  and  discard the  water  layer  over the
sediment,  and  mix  the  residue to  obtain  as
homogeneous a sample as possible. Transfer the sample
to a shallow pan to partially air dry for about three days
at ambient temperature. It is a good idea to select and
remove from the sample large debris such as stones,
broken  glass,  and bottle  caps.  Drying  time varies
considerably  depending on  soil  type  and  drying
conditions. Sandy soil will be sufficiently dry when the
surface starts to split, but there should be no completely
dry  spots.  Moisture  content  will be 50-80% at this
point.
  Weigh 50 g  of the partially dried  sample into  a 600
ml  stainless steel  beaker.  Add 50 g  of anhydrous
sodium sulfate and  mix well with  a  large  spatula.
Immediately  after  weighing  the  sample,   weigh
approximately 5 g of the partially dried sediment  into a
tared crucible. Determine the percent solids by drying
overnight at HOC. Allow  the crucible to  cool in a
desiccator  before  weighing.  Optionally one   may
determine  the percent volatile solids by placing the
oven dried sample into a muffle furnace and heating at
550C for one hour. Again the crucible should be cooled
in a desiccator before weighing. The concentrations of
organic  compounds  are normally  expressed as the
percent of the dry solid after the 1 IOC treatment.
  To the 50  g  sample add 250 ml of  1:1  acetone-
hexane, lower  the homogenizer into the beaker until it
is approximately one cm  from the bottom. Tilt the
probe slightly about 15 degrees' from vertical to prevent
the solution from  swirling out of the chamber  while
blending. Blend at maximum speed for 30 seconds.
Raise the probe out of the solution and rinse it with
acetone, allowing the rinsings to drain into the beaker.
Filter the homogenized mixture through a 7 cm or
larger medium porosity fritted glass funnel into a K-D
flask fitted with a 10 ml ampul. Rinse the beaker with
acetone and wash the filter cake with the rinse acetone.
Clean the probe between samples with successive 15
second runs in tap water, distilled water and acetone.
Wipe off the  residual solid material between cleaning
runs.
  If the  optional silica gel separation is to be used,
concentrate the extract to  1  ml; if no  silica  gel
separation is planned,  concentrate to 0.1-1 ml as
described under low  boiling  solvent  extraction in
section 3.1.
  For the silica gel separation, pack activated silica gel
into a glass chromatography column (2 cm  I.D.) to a
height of 10  cm. To the 1 ml of extract in  a small
beaker add sufficient silica gel to adsorb the  sample,
and then evaporate the residual solvent with gentle
stirring in a hood. Wet the column with about 20 m 1 of
hexane and add the sample when the last of the 20 m 1
reaches the surface  of the adsorbent. Rinse the beaker
with hexane and add the rinsings to the column. The
beaker should be rinsed several times with hexane and
each succeeding eluent when the eluent is added to the
column. Collect the fractions in K-D ampuls and elute
the aliphatic  fraction  with 85 ml of  hexane,  the
aromatic fraction with 85 ml  of benzene or the safer
toluene,  and the polar fraction with 1:1 methanol-
methylene chloride. Concentrate each fraction to 0.1-1
ml in the  K-D apparatus, and  proceed with  the
GC/MS analysis of each.
  The GC columns and conditions recommended are
the  same  as those used  for low boiling  solvent
extraction. This procedure should be consulted for this
information.  Also see section 3.1 for sample GC/MS
system dialogue.


3.4 FATTY TISSUE SAMPLES

  This method is intended for the extraction of a broad
spectrum of environmentally significant organics from
fatty tissue, and the separation of these organics from
the natural fats and oils prior to GC/MS analysis.  The
method has been tested mainly with fish tissue, but
should be applicable to other types of tissues.  The
analysis of fish tissue is used to describe the method,
with optional sample handling for the analysis of the
whole fish, or just the potentially edible portions.  The
advantage of the method is that  it  is  based on a
significant   amount   of   experience   of   several
                                                  52

-------
 investigators over a number of years. The detection
 limit for the method is not known, but is estimated at
 1-10 ug per kg of fish tissue.
   The analysis of a reagent blank is required for each
 group of fish specimens extracted  on a given day. The
 reagent  blank  analysis provides  information  about
 background and solvent contamination. Recovery data
 are not available for the total procedure since it is not
 possible to spike fatty tissue with known concentrations
 of organics and simulate natural conditions for  the
 incorporation of these materials. Perhaps  the best
 method of measuring the effectiveness of the extraction
 is to  re-extract with a fresh portion of solvent, or with
 an alternative   solvent,  such as  acetone-methylene
 chloride.
   Solvents  and  other reagents should be  purified as
 described  under low  boiling solvent extraction  in
 section 3.1. For  the analysis of large whole fish, a meat
. grinder is required. In all cases, a laboratory blender of
 about one  quart capacity is also required.  A  gel
 permeation chromatograph equipped  with a 2.5 x 50
 cm column packed with BIO-RAD SX-2 beads is used
 to separate the natural fats and oils from the organics of
 interest.
   Fish at the  sampling site  should  be wrapped in
 aluminum foil, shipped in an ice chest packed with dry
 ice (preferred) or ice, and preserved in a freezer until
 analyzed. Small  fish must be combined by sampling site
 and species to obtain the weight required for analysis.
 For the analysis of whole fish, the  entire fish (or fishes)
 are  ground directly, or,  if necessary, chopped into
 pieces small enough to fit into the  meat grinder. Grind
 the fish several  times and thoroughly  mix  the ground
 material.  Clean out  any  material  remaining in  the
 grinder and add this to the sample. For an analysis of
 the edible portions only, fillet the fish or fishes, and cut
 these into small pieces no  larger than about two cubic
 centimeters each.
   Add enough  dry  ice to the blender to completely
 cover the blades. Homogenize the dry ice for about 30
 seconds,  and then add about 25 g (weighed to the
 nearest 0. Ig) of fish fillet chunks or ground whole fish
 along with more dry ice to the blender: Wait several
 minutes for the fish to freeze, homogenize for at least
 two  minutes, or until the mixture is free of lumps, and
 then add about 75 g of purified anhydrous  sodium
 sulfate to the blender. Homogenize for another  two
 minutes, and pour the contents of the blender onto an
 aluminum  foil  sheet (one square foot). Homogenize
 additional dry ice and 25 g of anhydrous sodium sulfate
 and  transfer this to the  same aluminum foil. This
 operation serves to rinse the blender of residual fish
 residue. Mix the fish, sodium sulfate, and dry ice on the
 foil with a spatula. Carefully fold the aluminum foil
 into the form  of an  envelope, label, and place in a
 freezer at -15C  for 8-12  hours. The dry ice will
 sublime and the residue should be in the form of a
 granular lump-free material.
   Extract the mixture in a 250 ml Soxhlet extractor
 with 200 ml of a 1:1  mixture of acetone-hexane for 8
 hours. Cool the extract, transfer it to a Kuderna-
 Danish (K-D) apparatus, and concentrate it to about 3
 ml. Remove the last traces  of solvent with a gentle
 stream of dry  nitrogen, and weigh the resulting oil.
 Dilute  the  oil  with  methylene   chloride   to   a
 concentration of 100 mg/m 1.
  . Inject the total amount of diluted extract, in 5 m 1
 aliquots, into  the  gel  permeation chromatographic
 system.  Elute  each 5  mlx aliquot with  225  ml  of
 methylene chloride, at a flow rate of 3.5 m 1 per minute,
 and discard the first  160 ml  of each eluate  (this
 contains the natural oils and  fats). Collect the balance
 (about 65 ml)  of each eluate from each aliquot in the
 same 500 ml  K-D flask.  Concentrate the combined
 eluate containing the  organics  of interst to  5 ml.
 Equilibrate the  gel  permeation  system  with a  1:1
 mixture of cyclohexane-methylene chloride. Inject the
 5 ml of concentrated combined eluate and elute with
 300 ml of 1:1 cyclohexanemethylene chloride at 3.5 ml
 per minute: Again discard  the  first  160 ml  which
 contains the residual lipid material.  The remaining
. eluate may be divided into 3-10 fractions  or collected
 as a single fraction. Concentrate each fraction to 0.1  - 1
 ml in a K-D apparatus as described under low boiling
 solvent extraction in section 3.1. The GC columns and
 GC/MS conditions used for the low boiling  solvent
 extraction  procedure  are  recommended  for this
 method.

 3.5 CHEMICAL DERIVATIZATION
 OPTIONS

   Chemical   derivatization   methods  have  been
 developed extensively since 1955. Much of this effort
 was directed toward functional goup selective reagents
 for  derivatization of specific target  compounds  or
 narrow classes  of compounds; e. g., steroidal alcohols.
 In terms of the broad spectrum  approach defined in
 chapter  1  of this manual, a  few reasonably general
 reagents, such  as diazomethane, have been employed
 for  derivatization . of  several   broad  classes  of
 compounds, e.  g.,  phenols,  fatty  acids, etc., that
 generally appear in the acid  fraction  of a low  boiling
                                                  53

-------
 solvent extract. Historically, derivatization procedures
 were  needed  to  allow  the  gas  chromatographic
 separation  of compounds  that  were  difficult  or
 impossible to handle with existing liquid phases and
 solid  supports.  However, there has been  a steady
 improvement in  packing'materials and open tubular
 column technology, and several classes of compounds,
 such as the steroids, phenols, and fatty acids, may now
 be    conveniently    chromatographed    without
 derivatization. Under  these circumstances,  serious
 consideration must be given to the derivatization
 option, and  the  need for this additional step in any
 broad spectrum organic analysis. The advantages and
 disadvantages of derivatization are summarized in this
 section.

   A clear advantage of chemical derivatization is that
 several or many different classes of compounds may be
• readily chromatographed with a few general purpose
 columns. Without derivatization, it may  be  necessary
 to change columns frequently to accomodate certain
 classes of compounds,  e.  g., phenols and fatty acids.
 Another advantage of derivatization is that atoms may
 be introduced into various compounds that give the
 compounds  special  properties  that  aid  in  their
 identification.   Examples   of   this   include   the
 incorporation of a high fraction of heavy atoms such as
 deuterium  or carbon-13,  and naturally   occurring
 isotope distributions, such as bromine-79 and bromine-
 81.

   The disadvantages  of derivatization  include the
 additional cost, mainly in time and reagents, of these
 procedures. On the surface,  the additional costs may
 appear slight, but additional quality controls will be
 required, .and these may double or triple the number of
 GC injections which are time consuming and produce
 significantly more  data that  needs to be stored and
 interpreted. The real question is whether the additional
 cost is worth the additional information that may be
 acquired.  Another  disadvantage  is the  uncertainty
 about yield  in any derivatization procedure. Unless
 adequate    control    samples    are    derivatized
 simultaneously   under  identical   conditions  with
 identical reagents,  the  measurement accuracy of the
 method may be seriously affected. Independent of the
 use of quality control samples, derivatization for broad
 spectrum  organics  analysis  creates   considerable
uncertainty with regard to what the reagents are doing
to many other components of the mixture. By-products
and side reactions are not uncommon, and commercial
reagents  are frequently impure and introduce  trace
impurities  that  could  be  mistaken  for  sample
components. The problems of reagent shelf life and the
frequent  poisonous and explosive  nature  of many
reagents, e.g.,  diazomethane, are disadvantages  that
need  to be  carefully  weighed  in  the  cost-benefit
equation.
  In conclusion, it is recommended that for most broad
spectrum  analyses  derivatization  is  not worth the
effort, and that  column  switching  is preferable.
However, if certain  target compounds, e.  g., 2,4,6-
trichlorophenoxyacetic acid, must be included in the
analysis, derivatization is the only  practical choice  at
this time. In the future, liquid chromatography - mass
spectrometry interfaces may be developed to an extent
to permit direct analysis of some classes of compounds
not readily chromatographed without derivatization.
  If it is necessary to form a derivative of a specific
target compound, the commercially available reagents
and procedures are strongly recommended. The Pierce
Chemical Company of Rockford,  Illinois, has  been
particularly active in developing these reagents, and a
detailed discussion of the merits of various materials
and procedures is beyond the scope of this manual. The
choice of a reagent and procedure depends on the target
compound or compounds, and usually several reagents
and procedures  will be available.  Some  important
considerations in selecting procedures are the speed of
the reaction, the convenience of the procedure, side
reactions, and reproducibility. Often very pure and dry
reagents are required, including generally anhydrous
conditions. The commercially available reaction vials
with   Teflon   lined   septum  tops  are  strongly
recommended.
  Finally, because of its particularly hazardous nature,
a word  about  diazomethane  is appropriate. It is a
highly  poisonous and  very  explosive  yellow gas.
Extreme caution must be exercised to avoid inhaling it
or allowing it too near an open flame. The diazotization
reaction must be conducted in an effective fume  hood
and safety glasses and shields  are  mandatory. No
chipped,  etched, or cracked  glassware can  be  used
because    rough   glass   surfaces   enhance   the
decomposition of diazomethane.
                                                  54

-------
                                      CHAPTER  4
                                   DATA  OUTPUT
  The  purpose  of this chapter is  to  provide  the
information that is  needed  to  output data with a
computerized gas chromatography-mass spectrometer
system. The chapter describes in detail the operation of
programs that run on a Digital Equipment Corporation
Model PDP-8 Computer.
  One  of  the  most  fundamental  and  important
programs that the user of a system requires is one that
types out a list of programs and data files. All computer
acquired data and all computer programs are stored as
named  files on either the disk or tape storage device.
The names of the data files currently  existing on the
disk are required before output can begin. Only the files
on the disk can  be directly output when a  disk
              operating system is in use. Output directly from tape is
              possible only with a tape operating system. If a disk
              operating system is in use, and output is required from
              magnetic tape, the file must first be copied from tape to
              disk before output can begin. (See the copy program in
              Chapter 7).
                There are a number of programs available to list the
              programs and data files on a disk or tape. Each or these
              programs  is  somewhat   different  in  hardware
              requirements, and user input. Only  three of these
              programs are described in this section, and only three
              are supported under the revision EO operating system.
              The program names and hardware requirements are as
              follows:
                   Program
                    Name
                   LIST


                   LIST


                   MAGLST


                   BCLIST


                   LSTDSK


                   LSTBLK

                   LSTDAT
  Status

Unsupported


Supported


Supported


Unsupported


Supported


Unsupported

Unsupported
                      Sample dialogue for the supported
                    as follows:
         Hardware

   Disk unit 0 and TC08
   Dectape  unit !

   Disk unit 0 and TD8E
   Dectape  unit 0

   9-track industry  standard
   magnetic tape

   Disk units 0 or  1
   and TC08  Dectape units 0-7

   Disk units 0-3  and
   TC08 Dectape units 0-7

   Disk unit 0 only

   Disk unit 0 only

LIST program is
                                               55

-------
   SELECT  MODE:   LIST
   TAPE?:   N_

   NAME   BLOCK
   INIT$$
   EXEC
   EXMODI
   CUEXEC
   ETC.
15
17
47
76
In the dialogue a Y_to the TAPE prompt causes the
files on TD8E Dectape unit 0 to be listed. An entry of
LIST  to  the system prompt and M to  the TAPE
prompt calls the program MAGLST which causes the
files on the industry standard tape to be listed.
  For users with TC08 Dectapes or users who wish to
list the files on disk units 1-3, the program LSTDSK
replaces  the TC08  LIST  program  and BCLIST.
LSTDSK requires  8 K  of core memory and  an
extended arithmetic unit. The only input required is the
disk number DO, Dl, D2, or D3 or the TC08 Dectape
number TO-T7.
4.1  THE OUTP/CRT SOFTWARE

  The plotter and printer programs that are called by
entering OUTP in response to the system prompt are
the original output programs for the PDP-8 GC/MS
datasystem.  As such,  they  are  relatively  simple
programs that  require only 4K of memory and no
special PDP-8 hardware. Nevertheless, they have some
capabilities not available in  the  output software
described under sections 4.2  and 4.3. Also, they are
programs  that  are  easy  to  use  by  the  relatively
inexperienced user. For these reasons they have  been
retained on the system, and are used  extensively  in
many laboratories.  Eight  files are required on the
system  disk to  run  these   programs:  EXMODI,
PLOUT, PLOT, PLTP2, CUEXEC, RGC, TYPE, and
BACKGD. These programs handle output from disk
unit 0 only. Another program, PERCAL, may be used
to retrieve stored operating parameters from any stored
data file. This program is explained in more detail  in
section 2.7.
  The CRT program requires 8K of memory and the
files CRT,  CRTRGC,  CRTSP,  CBCKGD,  and
DSPLAY must be present. ,
THE TOTAL ION CURRENT PROFILE
  The total ion current .profile (TI'CP) is defined as a
normalized  plot of the sum of the ion abundance
measurements in each  member  of a series of mass
spectra as a function of the serially indexed spectrum
number. Data for this plot are acquired by continuous,
repetitive acquisition   of mass  spectra  as  sample
components emerge from a  gas chromatograph, leak
from a batch inlet system, or volatilize  from a direct
inlet system.  Each point  on  the  ordinate  is the
normalized  sum of all  the ion abundance data  in a
single mass spectrum; and, each point on the abscissa
represents  the spectrum number or a corresponding
unit of time.
  This same plot is referred to as a reconstructed gas
chromatogram  (RGC)  but this nomenclature is not
preferred as it does not accurately define a TICP. An
RGC could just as well be the output of a  flame
ionization detector, as redrawn by a draftsman.

Cathode Ray Tube Terminal. A TICP may be output
on a cathode ray tube (CRT) in a few seconds'or on a
pen and ink plotter in several minutes. Sample dialogue
for the CRT output is as follows:

   SELECT MODE:    CRT
   FILE NAME:  SAMP1
   DOUBLE  DISPLAY?:   N_
   TOTAL ION CURRENT PROFILE?:  Y_
   EXTRACTED  I C   PROFILE?:   N^

  If the user responds Y_to  the DOUBLE DISPLAY
prompt before the carriage return, a TICP and a mass
spectrum may be displayed together.  Mass spectra
outputs are discussed later. A response of Y_ to the
EXTRACTED  I  C PROFILE?: prompt leads to
dialogue which  is used to  display  an  extracted ion
current profile (EICP). This option is described in a
subsequent section.
  Unless a TICP consists of only a few spectra, e.g., 25,
it  cannot be compressed on a small  CRT and still
permit  the  viewer  to accurately  select spectrum
numbers for other operations. Therefore, commands
exist to replot a TICP with  a better view of a smaller
area. These commands, which are entered from the
keyboard after display of the TICP, are as follows:

  E_   further dialogue requests the extreme left hand
       spectrum number
  C_   entered  repetitively after E for subsequent
       plots of narrower ranges of spectrum numbers
       beginning with  the designated left hand value
                                               56

-------
  R_   restore last plot before expansion
  G_   the complete TICP is replotted
  F_   further  dialogue requests  filename  for  a
       different TICP

  CTRL/L returns to the  system  prompt.  Several
other commands ore available from the CRT keyboard
during viewing of these plots. These produce EICP and
mass spectra plots and  are discussed  in  subsequent
sections.

Plotter. The program for output of a TICP on the pen
& ink plotter is more flexible than the CRT program.
The output dialogue is as follows:

   SELECT  MODE:   OUTP
   TOTAL ION CURRENT  PROFILE:   Y_
   FILE NAME:   SAMP1
   EXPAND  BY:   press  return

  However,  in   this  case  the  spectrum   number
resolution  is fixed and quite  adequate for  selecting
spectrum numbers for other operations. A response of a
number between 1-100 to the EXPAND BY prompt,
causes the total ion current data to be multiplied by the
expansion  factor before plotting. This will  cause some
peaks to  be  plotted  with  flat  tops  indicating the
numerical  data  exceeds  the  range of the plot axis.
Baseline as well as peak values are multiplied by the
expansion  factor. In general factors greater than about
five are not very useful.
  Long TICP plots from GC  runs are best displayed
immediately  on the pen and ink  plotter. The time
requirement of several minutes is not that long, the
quality plot will be needed ultimately for the record,
and considerable time will be expended replotting for
better viewing with the CRT.  On the other  hand the
CRT is excellent for fast preview of TICP plots and the
rapid examination of data  from a  single or several
closely spaced peaks.
MASS  SPECTRA HISTOGRAMS
  Once a TICP is available, mass spectra are selected
from individual peaks and these are used to identify the
compounds causing the peaks. As with the TICP, mass
spectra may be displayed on the CRT or the plotter.
Since a TICP often  consists of hundreds of mass
spectra,  and since it is often necessary to plot several
mass spectra per peak, the fast CRT is of enormous
value in previewing mass spectra. Ultimately, selected
spectra may be plotted on the pen and ink plotter to
make quality copy for reproduction in reports or for
preservation in the permanent record.

Cathode Ray Tube Terminal. The dialogue for CRT
output of mass spectra is as follows:

   SELECT  MODE:   CRT
   FILE NAME:   SAMPLE
   DOUBLE DISPLAY?:   N_
   TOTAL  ION  CURRENT  PROFILE?:   N_
   SPECTRUM NUMBER:  3_
   SUBTRACT BACKGROUND?:   Y_
   SPECTRUM NUMBER:  2_

   In the CRT mass spectrum plot dialogue, a negative
response to a TICP causes the program to assume that
a mass spectrum is to be plotted. Normally the best
spectrum number is found  on the leading edge of a
peak, about two-thirds of the way to the peak top. For a
single component peak, the spectra will not differ much
across the peak except for abundances at the very high
and  low amu  values.  For  a  multicomponent peak,
several spectra can be used to determine which ions in a
spectrum belong to a single component. Background
subtraction  is  extremely   important  to  remove
extraneous ion abundance data from the spectrum of a
compound.  In  addition to unresolved  components,
these extraneous ions are generated from continuous
column bleed  and background  materials   in  the
spectrometer  that  are not  pumped  out  rapidly.
Spectrum  selection and  background subtraction is
often an iterative process, and several trials may be
required to produce the optimum  spectrum.
   An affirmative response to the CRT double  display
prompt allows simultaneous display of two plots. The
following dialogue illustrates the use of this feature to
display a spectrum with background subtracted on the
upper plot and  the background spectrum on the lower
plot.

   SELECT  MODE:   CRT
   FILE NAME:   SAMPLE
   DOUBLE DISPLAY?:   Y_

   UPPER  GRAPH PARAMETERS:
   TOTAL  ION CURRENT  PROFILE?:   N_
   SPECTRUM NUMBER:  3_
   SUBTRACT BACKGROUND?:  Y_
   SPECTRUM NUMBER:   2_

   LOWER  GRAPH  PARAMETERS:
   SPECTRUM NUMBER:   2_
   SUBTRACT BACKGROUND?:  N
                                               57

-------
  The CRT commands E^ C R, G_, F,_ and CTRL/L
are also available after mass spectral plotting. In this
case E sets the extreme left hand amu value. Another
command, S,_may be entered after viewing a TICP or a
mass spectrum. It  requests a spectrum number for
plotting  as well  as information for background
subtraction.

Plotter. The program for plotting a mass spectrum on
the pen and ink plotter is  more flexible than the CRT
program and also permits one  to print digital mass
spectral data on the console printer or CRT as well as
plot on the plotter. The output dialogue used to plot a
mass spectrum is:

   SELECT MODE:  OUTP
   TOTAL  ION CURRENT  PROFILE:   N^
   EXTRACTED  ION CURRENT  PROFILE:   N_
   PLOT SPECTRUM?    Y_
   FILE  NAME:   REF2
   SPECTRUM NUMBER:   49
   AMPLITUDE EXPANSION?:   press  return
   MINIMUM  VALUE  %:  J_
   SUBTRACT BACKGROUND?:   Y_
   SPECTRUM NUMBER:   44
   BACKGROUND AMPLIFICATION:
      press return
   SAVE  SUBTRACTED  FILE?:   N_
   NORMALIZE ON:   press  return
A plot of the reference compound mass spectrum using
the  dialogue is shown in  Figure 4.1.  To illustrate
several of the plot options the following dialogue was
used to generate the spectrum in Figure 4.2.

   SELECT MODE:  OUTP
   TOTAL ION CURRENT  PROFILE:   N_
   EXTRACTED  ION CURRENT  PROFILE:   N_
   PLOT  SPECTRUM?:   Y_
   FILE NAME:   REF2
   SPECTRUM NUMBER:  49
   AMPLITUDE  EXPANSION?:   Y_
   THRESHOLD  %:   5_
   EXPAND   BY:   5_
   MINIMUM  VALUE %:  2_
   SUBTRACT BACKGROUND?:   Y_
   SPECTRUM NUMBER:  44
   BACKGROUND AMPLIFICATION:   W
   SAVE SUBTRACTED  FILE?   Y_
   FILE NAME:   REF2BS
   NORMALIZE  ON:   press  return
   TITLE:   DFTPP 5  1  75
   SPECTRUM  NUMBER:  49-44
   BASE PEAK:   152320
   TOTAL ION CURRENT:  J_223424

  An affirmative response  to the  AMPLITUDE
EXPANSION? prompt leads to dialogue in which a
per cent relative abundance threshold and an expansion
factor are entered by the user. The  product  of the
threshold and expansion factor may not exceed 30. The
principal application of this option is to emphasize very
weak  ion abundances  in the  spectrum  plot. The
minimum value option is used to eliminate plotting of
very weak abundances below the minimum. Normally
the minimum value and amplitude expansion options
are  not  used   in  the   same  plot.  Background
amplification  may be used  to  enhance the ion
abundance data in the background spectrum before it is
subtracted from  the sample spectrum.  If the user
chooses to save the background subtracted spectrum, a
file name must be supplied for the data. All spectra are
automatically plotted with the most abundant ion set to
100% relative abundance. However, a user may enter
an alternative  mass number  in  response to the
NORMALIZE ON response, and  the abundance of
that mass will be set to 100%  with a proportionate
adjustment for all other abundances. This has the effect
of amplifying weak ion abundances without the scale
expansion shown in Figure  4.2

PRINTED MASS SPECTRAL DATA
  Printed mass  spectral data may be produced on
either the CRT or the console printer, but the OUTP
mode is required in both cases. Output is directed to
one or the other via the switch on the back of the CRT.
There are  two  possible  responses to the PRINT
SPECTRUM?: prompt.  A Y_response will cause mass
numbers, per cent relative abundances, and per cent of
total  abundances to be  printed. An A_ response will
cause mass numbers and the corresponding absolute
abundance data to be printed. Absolute abundances are
in analog-to-digital converter units. Standard printout
dialogue allows the user some of the same options as
the plotter routine:

   SELECT MODE:   OUTP
   TOTAL ION  CURRENT PROFILE:    N_
   EXTRACTED  ION  CURRENT PROFILE?:   N_
   PLOT SPECTRUM?:    N_
   PRINT  SPECTRUM?:   Y_
   FILE  NAME:   SAMP2
   SPECTRUM  NUMBER:  42
                                            58

-------

8
e.
8-
P-
bS.
8.
o

SPEOfUl
DFIPP S



V

30 10
nx e
rueat t
) IS




j j, ,
X SO
9 -





A
TO
tl





/„!
eo SB 100







10 120






I 	 ..„!'.
tag IIB isa 10 no tao iao






,| 11(1 (
2nO Z10 ZE» 238 MO ZSB






. 1 , ' . .
a» ^ 280 SO » 310 320 3M 310 36D 3SO 370 380 330 «10 '«0 CO GO t







IB «0 1
Figure 4.1  The mass spectrum of a GC/MS reference compound, decafluorotriphenylphosphine.

-------
    9>EEmjM Nraoi


    HTTP s i TS
 R.
r
WB-
                  »  SB"  i
                                                IBB in ae
B SB V
                        iee us' ISB tag i« tan iw ITS
                                                         «a
za MS ZSB as ro a» ao sss sie fc 333 sra 3SB sso rw jaa 333  -as  «a' «B 'CT «> iso
                 Figure 4.2  The mass spectrum of a GC/MS reference compound, decafluorotriphenylphosphine, with

                              plot options.

-------
   PARTITIONED  OUTPUT?:  N_
   MINIMUM VALUE  %:   5_
   SUBTRACT BACKGROUND?:   Y_
   SPECTRUM NUMBER:   40
   BACKGROUND  AMPLIFICATION:
     press  return
   SAVE  SUBTRACTED FILE?:   N^
   NORMALIZE  ON:   press return

   TITLE:   SAMPLE  2 2UL  of  1ML
   SPECTRUM NUMBER:   42 -  40
   BASE  PEAK:   327680
   TOTAL ION  CURRENT:   681283


PERCENT OF
AMU INTENSITY TOTAL INTENSITY
51.0
63.0
64.0
102.0
102.0
126.0
127.0
128.0
129.0
9.27
7.44
8.12
9.21
9.21
7.39
13.90
100.00
11.45
4.45
3.58
3.90
4.43
4.43
3.55
6.68
48.09
5.51
However,  in  place  of the plotter AMPLITUDE
EXPANSION?   prompt,   the   PARTITIONED
OUTPUT? prompt is printed. A Y_ response permits
the  output of selected  ion  abundance  data.  The
following application  of partitioned output shows data
from  a  spectrum   of perfluorotri-n-butyl  amine
(PFTBA). This spectrum was measured immediately
after calibration while PFTBA was in the spectrometer.

The partitioned output gives  a rough  idea of the
condition of the spectrometer.
   SELECT MODE:   OUTP
   TOTAL ION CURRENT PROFILE:   N_
   EXTRACTED ION CURRENT PROFILE?:
   PLOT SPECTRUM?:   N_
   PRINT  SPECTRUM?:   Y_
   FILE  NAME:   J_
   SPECTRUM  NUMBER:   j_
   PARTITIONED  OUTPUT?:   Y_
   MASS  RANGE:   67-72;  219-222;
      500-505; 612-616
N
     :  TITLE:  PFTBA
       SPECTRUM  NUMBER:   1
       BASE PEAK: .7215104
       TOTAL ION  CURRENT:    15664128
                              PERCENT  OF
     AMU
     ,  67.0
       68.0
       69.0
       70.0
       71.0
       72.0
     219.0
     220.0
     221.0
     500.0
     501.0
                                                502.0
                                                503.0
                                                613.0
                                                614.0
                                                615.0
                                                616.0
       INTENSITY
                 .04
                 .49
              100.00
                1.38
                 -02
                 .00
               25.23
                1.09
                 .02
                 .00
                 .00
                 .84
                 .09
                 .00
                 .15
                 .02
                 .00
TOTAL INTENSITY
           .02
           .22
         46.06
           .63
           .00
           .00
         11.62
           .50
           .00
           .00
           .00
           .38
           .04
           .00
           .07
           .00
           .00
   MINIMUM  VALUE %:   press  return
   SUBTRACT  BACKGROUND?:   N_
   NORMALIZE  ON:  press  return
                                              Partioned output of digital data is analogous to the
                                              extracted ion current profile that is discussed in the
                                              next section.
THE EXTRACTED ION CURRENT
PROFILE
  The extracted ion current profile (EICP) is defined as
a plot of the change in relative abundance of one or
several ions as a function  of the serially  indexed
spectrum number. It is called an extracted ion current
profile because the data are literally extracted  from the
total ion current profile (TICP) data in a non-real time
process.  The  EICP   is  therefore  analogous to
partitioned output of digital data. However, it should
not be confused with real-time selected ion monitoring
(SIM) and the corresponding selected ion current
profile (SICP) that are discussed in Chapter 6. The SIM
technique is a real time process in which ion abundance
data are  measured  at  only selected  masses as
components emerge from the gas  chromatograph or
other inlet systems. The SIM technique produces a real
increase in signal/noise by time  averaging random
noise. The EICP produces an apparent increase in
sensitiyjty  by  removing  from  the  TICP  the  ion
abundance  data   from   background,  unresolved
components, and other irrelevant ions.  The  terms
limited mass output or limited mass search are often
                                            61

-------
used, but are less meaningful than extracted ion current
profile to describe this plot.
  On the CRT, the EICP is generated by an affirmative
response to the TOTAL ION CURRENT PROFILE?:
prompt  and  an  affirmative  response  to  the
EXTRACTED I C  PROFILE?: prompt. The user is
then requested to enter a mass or mass range. The CRT
commands E, C, R, G, and F are available as described
above, and an L command may be used to change the
mass range in the EICP plot. The dialogue for a CRT
EICP is as follows.

   SELECT  MODE:   CRT
   FILE NAME:    SAMPLE
   DOUBLE DISPLAY?:   N_
   TOTAL  ION CURRENT PROFILE?:
   EXTRACTED I  C  PROFILE?:  Y_
   MASS RANGE:   149
Y
  In response to the MASS RANGE? prompt, the user
may insert up to eight individual masses or mass ranges
separated by semicolons. The EICP produced is a sum
of the abundance data for each of the masses or mass
ranges specified. Figure 4.3 shows a TICP and the
EICP for mass 149 from the same data  file. The
compounds that produce  mass 149 ions are clearly
highlighted in the EICP. The plotter output dialogue is
similar to the CRT dialogue.

   SELECT MODE:   OUTP
   TOTAL ION CURRENT  PROFILE:   N^
   EXTRACTED  ION  CURRENT PROFILE?:  Y_
   FILE  NAME:   SAMP1
   MASS  RANGE:   149
   EXPAND  BY:   press return

QUEUED OUTPUT
  The user may defer actual output of various pen and
ink  plots  and  printouts  until a series of output
commands  have been stored for  later execution.  In
response to the system  prompt, the user enters the
command QUEUE. The system immediately returns to
the  system  prompt. The user  enters the information
required  to generate an  output.  However, at the
completion of the dialogue the plot is not executed, but
the  system returns  to the system prompt. A series of
output dialogues may be completed by a repitition of
this procedure.  The queue is ended with the command
ENDQ in response to  the system prompt.  All
commands  are  then sequentially executed. The user
must delete the $$ QUE file before another queued
output can begin. The following dialogue is an example
of a queue for two spectrum plots.
SELECT MODE:   QUEUE
SELECT MODE:   OUTP
TOTAL  ION CURRENT PROFILE:   N_
EXTRACTED ION CURRENT PROFILE:   N_
PLOT SPECTRUM?:   Y_
FILE NAME:   805D
SPECTRUM  NUMBER:  34
AMPLITUDE EXPANSION?:  press return
MINIMUM  VALUE %:  j_
SUBTRACT  BACKGROUND?:   Y_
SPECTRUM  NUMBER:  3^
BACKGROUND AMPLIFICATION:
    press  return
SAVE SUBTRACTED  FILE?:   N_
NORMALIZE ON:  press  return

SELECT MODE:   OUTP
TOTAL  ION CURRENT PROFILE:   N^
EXTRACTED ION CURRENT PROFILE:   N_
FILE NAME:   81-6D
SPECTRUM  NUMBER:  67
AMPLITUDE EXPANSION?:  press return
MINIMUM  VALUE %:  \_
SUBTRACT  BACKGROUND?:   Y_
SPECTRUM  NUMBER:  60
BACKGROUND AMPLIFICATION:
    press  return
SAVE  SUBTRACTED  FILE?:   N^
NORMALIZE ON:  .press  return

SELECT  MODE:  ENDQ
         4.2 THE MSSOUT OUTPUT SOFTWARE

           This output system has many of the same 'basic
         capabilities as the output and CRT software described
         in section 4.1. However,  the MSSOUT software has
         flexibility and capabilities not found in the standard
         output programs. This progam requires the presence of
         an extended arithmetic element (EAE) in the PDP-8
         computer. The EAE unit consists of two printed circuit
         boards labeled M8340 and  M8341. These provide
         enhanced  processing  speed  for certain types  of
         computer instructions.
           The MSSOUT software also requires the presence of
         seven files on the system  disk: MSSOUT, MOUTP1,
         MOUTP2,  MOUTP4, MOUTP5,  MOUTP6,  and
         INCCHR. Another file,  SS1S, is generated by the
         program.  As with  many other choices, there are
         advantages  and  disadvantages  to  adopting  the
                                           62

-------
                      i oo no  i:e 130 i«  IEO ieo no  too iso zoo  310 220 zio zte  zso
                                                                               EICP-MASS 149
                                                                              zre  ZBB zse SOB  aio 320
                                                                                                      aw jse JIB  aro JBO iaj  «B
Figure 4.3  The extracted ion current profile for mass 149 and the corresponding total ion current profile.

-------
MSSOUT  software.  The principal disadvantage, in
addi.tion to the hardware capital investment, is that a
set of user commands need to be mastered in order to
use the capability.  The MSSOUT program is non-
prompting (non-interactive). The user must know what
commands to enter and their proper sequence. Another
disadvantage is that there is a bug in the MOUTP4 file
that has proven difficult to locate and repair. This bug
causes flat topped peaks that are really not  saturated.
Another version of MOUTP4 exists in which  the flat
topped  peak  bug  was  repaired,  but  spectra are
sometimes missed. A temporary fix is to retain both
versions on the system disk under other  names. This
allows  the user to delete the current MOUTP4 and
rename with the COPY program one of the versions.
The choice depends on which bug the user wants to live
with. The  file MTP4/1  is a version of MOUTP4 that
gives flat topped peaks; the file MTP4/2 gives missing
spectra.
  The  advantages of the program appear to be very
significant and include the following:

   1.    The user has the capability to designate within
       limits the size of various plotter displays and
       alphanumeric labels.
  2.    The user  has  the  ability to switch  rapidly
       between CRT and plotter displays within the
       same program.
  3.    Because of  the EAE  hardware, processing
       speed is improved somewhat compared to the
       standard software.
  4.    Mass spectra histograms may be plotted over a
       smaller mass range than was actually observed
       during the original .data acquisition. Therefore
       if the 33-450 amu range was observed, but no
       ions  were  observed  above  mass  300, the
       histogram plot may be limited to 33-300 amu.
  5.    Similiar capabilities exist for Total Ion Current
       Profile (TICP)  and Extracted Ion Current
       Profile  (EICP)  plots,   i.e.,  portions   of
       chromatograms that contain the peaks may be
       plotted, and the leading or trailing sections
       that contain no peaks may be deleted from the
       graphics display.

  The MSSOUT outputs all begin with the same basic
dialogue:

         SELECT  MODE:  MSSOUT

         RUN NAME:   STD
         NBS STD SONG  EACH
In this dialogue the run name means the same as the
filename in the standard output software. After the run
name is  entered, the program immediately prints the
title attached to that file. If an informative title was
entered,  this will confirm the identity of the data file.
After this an asterisk is printed. This is a signal that the
program is ready to receive commands from the user. If
the user  wishes to change the data file (run name) this
may  be  accomplished  anytime after an asterisk is
printed. Simply enter the NA (name) command and the
program returns to the RUN NAME: prompt.
   •NA

   RUN
   NBS
NAME:    213
SAMPLE  213 5UL
  Anytime after an asterisk is printed the user may
enter  the EX  (exit)  command  which causes  the
program  to return to  the system prompt. The user
commands for various outputs are described  in the
following  sections. These  are organized similarly to
section 4.1 with sections on the TICP, Mass Spectra
Histograms, and  Printed  Mass Spectral Data, the
EICP, and Queued Output.

THETOTAL ION CURRENT PROFILE
  The sequence of commands in the sample dialogue
will cause a TICP to be plotted on the plotter. The
TICP will begin at spectrum number 20 and  end at
spectrum number 220. The plot will be 3 inches high by
5  inches  long. Figure 4.4 is the result  of these
commands:

          *CM
          ENTER MASSES  TO  COLLECT.
          MAX= 24

          *EM
          ENTER MASSES  TO  PLOT
          J_

          •ST=20

          *EN = 220

          •HE = 3
          •LE =
          .pp
          SCAN
                                               64

-------
                                                      TJJ
      50           100           150           200
Figure 4.4 A total ion current profile generated with MSSOUT

-------
s
         STD
    lOO-i
     80-
     60-
     40-
     20-
    SCAN  216
               i
r25
Vf*
                                   •) "I-f r | » ; v yr( i [ » |"» ; » (
         50    100     150    200     250    300
                  Figure 4.5 A mass spectrum plot generated with MSSOUT

-------
  This sample dialogue illustrates the use of some of
the commands that may be used to generate a TICP.
The CM (collect masses) command is mandatory and
causes the program to gather and save in core memory
the data necessary to plot a TICP and, optionally, a
number of  EICP's.  If the  user  responds with  a
RETURN to the ENTER MASSES TO COLLECT
prompt, as was done in the example, only  data for a
TICP will be collected, and it will not be possible to
display  an EICP. However the user may reenter the
CM anytime an asterisk  is  printed,  and  cause  the
program to collect a new  set of data.  The  MAX=X
gives  the maximum number  of EICP's that may  be
saved in core memory, and this is explained further in
the section on the EICP. If MAX=0, the  data for a
TICP is saved after entering RETURN.  If specific
masses are entered, the data  necessary for  a TICP  is
always gathered in addition to the EICP data.
  The EM command  is optional in the most recent
version of MSSOUT, and this revision should be placed
on  all disks. The functions of the EM command are
explained in the section on the EICP.
  The ST (start) = X (where X = an integer spectrum
number) and EN (end)  =  X commands are optional,
and allow the user  to define starting  and ending
spectrum numbers for the TICP. If these commands
are omitted,  the default values are the first  and last
spectrum numbers in the file.
  The HE (height) = X (where X=  integer inches)
and LE (length) = X commands are also optional, and
apply to the plotter only, not the CRT. They allow the
user to define the height and length of the plot up to the
maximum size of the paper.  If these  commands are
omitted, the  default size  is 5x8  inches.  The  final
command that must be entered is either a PP for a plot .
on the plotter or a PM for a display on the CRT.
  An extremely important feature of MSSOUT is that
the values entered for all  commands are saved until
they are changed or until the program is restarted. This
feature results in a number of convenient possibilities
for the user.  For  example one  may  issue a  PP
immediately  after a PM and produce the same plots on
the- plotter and CRT.  Alternatively, the  user  may
change just one or two commands then enter the PP or
PM, and immediately produce a modified plot without
reentering all the  commands.  The  command  OP
(options) will print out a list  of all current command
options, and this should be used in the event the user
needs to recall the current option selections. The values
of the options shown in the following list specify the
default values, and these will be used for all plots until
they are changed by the user. In some cases, e.g., HE,
the actual default value is given; however in most cases,
e.g., CM, a zero specifies the default situation.
    *OP
CO
ST
EN
BA
MA
LI
HE
LE
TH
CS
LA
FR
CM
SM
MC
su
SF
0
0
0
0
0
0
5
8
1
0
50
Y
0
0
S
0
100
These option values are saved during a terminal session
even though the user changes the data file with the NA
command.
  An IN (integrate) command also exists to integrate
peak areas. However, this is discussed in section 6.2
along with other quantitative analysis methods. If the
data was acquired with the RIB interface and time data
was stored with the datafile, MSSOUT may be  used to
give plots with time in minutes on the x axis. Setting the
parameter MC=T gives the time axis. The  default
value is MC=S or spectrum number.

MASS SPECTRA HISTOGRAMS
  The sequence of commands  in the sample dialogue
will cause a mass spectrum to be plotted on the  plotter.
A background spectrum will be subtracted, and the
plot will  begin at mass 50 and end at mass 310. The
filename (run name), spectrum number (scan number),
and  axis  labels will be  in  one-quarter  inch  high
characters.  The size of the plot will be 6 x 8  inches.
Figure 4.5 is the result of these commands:
     *SC

     SCAN
     216

     *SU
     ENTER
NUMBERS:
 BACKGROUND:    213
                                                67

-------
300-
 80-
 60-
 40-
 20-
    STD        SCAN 236
                     *t»,HJrr^
               r25
X30
      50        100       350       20O       25O       3OO
                  Figure 4.6 A mass spectrum plot generated with MSSOUT

-------
   'ST=50
   •EN = 310
   •CS =
   *LE =
   *PL
The SC command is mandatory for a mass spectrum
plot and  the spectrum number is requested. Several
spectra may be  specified by  a series of spectrum
numbers  separated  by  commas.  The only  other
mandatory command is either  a PL for a plot on the
plotter or a  DI for  a display  on the  CRT.  After a
spectrum is displayed on the CRT, entry of a return
will cause the CRT to be paged for the next spectrum, if
more than one was specified, or the next command. If
more  than  one spectrum is specified  with  the SC
command, and output is to  the plotter, the plots are
queued automatically one after the other. There are a
large number of optional commands that may be used
to refine  these graphics outputs. Some of these apply
only to displays on the CRT, some apply only to plots
on the plotter, and some apply .to both outputs.
  All of the following options apply to both the CRT
and plotter. The  SU  command causes a background
subtract  prompt. If  more than  one background
spectrum is specified,  the  background subtraction
process appears to be aborted, but no error message is
output by the program. The ST (start) = X (where X
= an integer mass) command defines the starting mass
of the plot, and the EN (end) = X command defines
the ending mass of the plot. The default values for ST
and EN are the first and last  masses observed. The BA
= X  (where X =  an integer  mass) may be used to
specify the mass used to normalize the mass spectrum
plot. The default value is the base peak in the spectrum.
The LA = X (where X =.an  integer mass) may be used
to specify the mass interval between labeled tick marks
on the mass axis of a spectrum plot. The default value is
50 amu. The SF = X (where X is a integer factor) may
be used to specify the amount of amplification of the
background spectrum before background subtraction.
The  default  value  is  100 which corresponds to an
amplification factor  of one.  Therefore entry of SF =
150 will amplify the background file by 50%.  Entry of
SF = 50 will reduce the background  spectrum by  a
factor of  two. The MA = X, Y (where X = an integer
mass number and Y = a magnification factor) may be
used to specify  that beginning at mass  X  all ion
abundances are to be multiplied by the factor Y before
plotting.  The mass  where magnification  begins  is
indicated on the plot as in Figure 4.6. The capability to
average several spectra,  e.g., across a TICP peak,  is
another optional procedure which is accomplished
with the dialogue:
   •SC

   SCAN  NUMBERS:
   A_

   •AV
   ENTER SCANS  TO AVERAGE:  214-220
The average spectrum plot is completed with the other
standard commands as above.
  A summary of optional commands that apply to the
plotter and the CRT is as follows:

  SU  Subtract background
  ST  Starting mass
  EN  Ending mass
  BA  Normalize on
  LA  Interval of mass  axis labels
  SF  Background amplification  factor
  MA Ion  abundance multiplier
  A/AV Average spectra

As  previously  described under the total ion current
profile, values  entered for most of these options are
saved until changed or until the program is restarted.
The current values or default values of the optional
commands may be printed with the OP command. A
command RE  (repeat)  is  also available  for  mass
spectrum plots. This causes a repeat of the plot that was
most recently requested.
  There are three commands that  apply just to CRT
displays. The  DI command is the command  which
causes the actual CRT display and this was described
above. The LI =  X (where X = a low integer value,
e.g., 1,2, or 3)  is an option that defines the number of
horizontal regions the CRT is divided  into for an
expanded  presentation  of -a  mass spectrum.  The
. practical limitation of LI is probably 3 or 4, and the
default value is selected by the program depending on
                                                69

-------
the mass range scanned. Finally the CO = X (where X
is an integer) command may be used to automatically
make X number of hard copies of the CRT display.
  There are several commands that apply just to the
plotter. The PL command causes the plot to begin, and
this was described above. The FR command causes a
line (frame) to be drawn across the top of the mass
spectral plot. The DF comand deletes this frame and is
the default choice. The HE = X (where X = height in
inches) and  LE = X  commands  have  the same
meaning  as  described  under the total ion  current
profile. These options may be used to specify the size of
the plot within the limits of the paper size. The default
size is 5 x 8 inches. Finally the CS = X (where X = an
integer value) may be used to define  the size of the
alphanumeric characters used in the plot labels. A size
of one inch is expressed by CS  = 100, one half inch by
CS = 50, and one-quarter inch by CS = 25, etc.
   In summary, the commands that apply either to the
CRT or the plotter, but not both are as follows:

   DI  Begin CRT plot
   LI   Division of CRT into regions
   CO Automatic CRT copies

   PL  Begin plotter plot
   FR  Draw line  on top of histogram
   DF Delete line
   HE Set height  of plot
   LE  Set length  of plot
   CS  Define character size

PRINTED MASS SPECTRAL DATA
   Lists of digital data may be printed on the CRT or
printer using the commands  TA and  RD. The TA
command gives a  normalized  list  and the   RD
command generates raw data in terms of analog to
digital converter  counts. A spectrum number  or
numbers must be specified first with the SC command.

THE EXTRACTED  ION CURRENT
PROFILE
   The EICP plot is a variation on the total ion current
plot, and all the commands that apply to the TICP also
apply to the  EICP. However the method of using the
CM  and  EM commands is crucially important. The
CM  (collect  masses) command is required as in the
TICP, but the individual masses of interest  must be
specified in response to this command. Mass ranges are
not accepted as  a response to  the CM command. The
printed  message  that  follows  the CM  command
informs the user of the maximum number of individual
masses that may be specified. This maximum will vary
as a function of the size of the data file.
   'CM
 ENTER MASSES COLLECT.
   143,167,193,208
MAX=    24
  After  the  collection is  completed  the  user  may
display the plots in a number of sequences, and create
other types of plots. These very flexible options are
described in this section. The most basic option  is to
display the EICP's specified in exactly the order and
quantity given  in the CM command.  This  may  be
accomplished on the CRT  or plotter, but the TICP is
always placed on the bottom of the display. The EICP's
are then placed in order, from the bottom to the top, as
they were listed with  the CM command. This list is
saved  until another CM command is issued, or the
program is restarted.
  The functions of the EM command are to redefine
the order in which various plots are displayed  on a
single CRT or plotter  display, and to generate subsets
of the plots specified in  the CM  command.   For
example, if the EM command below were entered, the
sequence  of plots" from the bottom to the  top of the
display would be a TICP, a mass 169 EICP, a sum plot
(defined later), a TICP, a mass 197 EICP, a mass 213
EICP, and another TICP.

  *EM
  ENTER  MASSES  TO  PLOT
  T,169,S,T,197,213,T
Figure 4.7 is an example of this type of display using a
different response to the EM command. The sequence
of plots defined in the EM command  is saved until
another EM command is issued or the program is
restarted, but the current value of EM does not appear
in the options list. Mass ranges are not accepted by the
EM command.
  The EM command may also be used to define subsets
of masses prior to the IN (integrate) command that is
described in section 6.2.

  The sum  plot referenced above displays the sum of
the abundances of two or more masses. The S response
to EM is not functional until a sum is defined by the SM
command as follows.
   "CM
  ENTER  MASSES  TO  COLLECT. MAX =
   143,167,193,200
             24
                                               70

-------
       Ld
f-    <
C/O    CQ
250
                               Figure 4.7 An extracted ion current profile generated with MSSOUT

-------
   •EM
   ENTER  MASSES  TO PLOT
   S_

   *SM
   ENTER  M/E'S  TO  SUM
   143,  167,193

   *PP
   SCAN
This dialogue produces a single plot that is an EICP for
the sum of the abundances  of the  three  masses
specified. There is one other command, the BA  = X
(where X is an integer mass), that  may be used to
specify the.  abundance used to normalize  any EICP
plot.
   In summary the commands that may be used for the
EICP are as follows:
   PP   Causes actual  plot on the plotter
   PM   Causes actual  plot on the CRT
   ST   Defines starting spectrum number
   EN   Defines ending spectrum  number
   HE   Defines height of plot on the plotter
   LE   Defines length of the plot on the  plotter
   SM   Requests abundances to  be  summed
   BA   Normalize EICP on  mass

   The user should note that Figure 4.7 has a line drawn
across the top of each EICP and that this may not be
deleted with the DF command because this command
does not apply to TICP and EICPs.
QUEUED OUTPUT
  Queued output of a series of separate displays with
MSSOUT is only possible with mass spectra plots. The
user should enter, in response to the message printed
after the SC command, a series of spectrum numbers
separated by commas. The SU command may be used
to subtract the same background spectrum from each.
After each spectrum is plotted on the CRT, pressing
the return causes the program to immediately produce
the next plot. However if a CO command is used, an
automatic hard copy of each will be generated, and the
next spectrum produced without the return. A similar
series of plots may be queued to the plotter.
   4.3  INTERACTIVE  GRAPHICS
   ORIENTED OUTPUT
   SOFTWARE (IGOOS)

     This  output  system has many of the  same  Basic
—capabilities as the plotter and CRT software described
   in  sections 4.1  and 4.2. However,  the IGOOS has
   significant additional capabilities not available with the
   previously  described  output systems. This  software
   requires the presence of an extended arithmetic element
   (EAE) in the PDP-8 computer. The EAE unit consists
   of two printed  circuit  boards labeled M8340 and
   M8341. These provide enhanced processing speed for
   certain types of computer instructions.
     The standard TICP and EICP capabilities require
   the presence  of three  files  on the  system  disk:
   BCLRGC, RGCOV1, and RGCOV2. Mass spectrum
   plotting requires the files CRTPLT, PLOVL1, and
   PLOVL2.  As with many other choices, there are
   advantages and disadvantages to adopting the IGOOS.
   The  principal  disadvantage,   in  addition  to the
   hardware  capital investment,  is  that a set  of user
   commands need to be mastered in.order to use the
   capability.  The  IGOOS  has its own  set of commands
   and the user must know what commands to enter and
   their proper  sequence.  Another disadvantage is the
   presence  of  several  program  bugs.   These are
   documented along with the user instructions. The
   advantages of the program appear to be very significant
   and include the following:

     1.    The  user  has  the capability of  selecting
          spectrum  numbers  from   CRT  graphics
          displays by placing the cross hair cursors over
          the appropriate place  on the display, and
          entering a single character  command at the
          keyboard.
     2.    The convenient spectrum number  selecting
          capability  described above  may  be used to
          build Q files  consisting of spectrum numbers
          and  corresponding  background  spectrum
          numbers. These Q  files may be used to drive
          other programs  that plot spectra on the CRT
          or plotter, that abbreviate spectra for matching
          against a  local or  remote  database with a
          spectrum matching program, or that perform
          quantitative analyses.
     3.    The  same basic software may be used to
          generate CRT and plotter displays of various
          graphics, with the CRT used as a convenient
          preview of what  will appear on the plotter.
                                               72

-------
  4.   Because  of the  EAE  hardware,  processing
       speed is improved somewhat compared to the
       software described in section 4.1.
  5.   Finally there is the capability to display TICP
       and EICP plots exactly above one another on
       the CRT or plotter.  Up to six  plots can be
       overlayed in this way.

TOTAL AND EXTRACTED ION
CURRENT PROFILES
  The user  entry  into this program  is  with  the
following dialogue which specifies the first TICP or
EICP to be displayed on the CRT:

    SELECT  MODE:    BCLRGC
    FILE?     CINTAP
   Q-FILE NAME?
   RANGE?
   EXPAND  BY?
QCINTA
 In response to the FILE? prompt the user enters the
'data filename. A Q-filename must be  entered even
 though the user has no intention of setting up a Q file.
 The program will perform if a RETURN is entered in
 response to the  Q-FILE  NAME? prompt, but an
 unnamed Q file will result if a Q file is initialized, and
 this will cause problems. If a Q file is not initialized and
 closed as explained below, the Q-filename will not be
 saved  in  the disk directory.  In  response to the
 RANGE? prompt the user may enter a RETURN for a
 TICP or a single mass or mass range for an ELCP. In
 response to the EXPAND BY? prompt the user enters
 a RETURN for no expansion, or an integer number
 which is a multiplication factor for profiles containing
 only  weak  peaks. A  CRTL/L  during  this or any
 subsequent  dialogue  causes  a return to the system
 prompt.
  At this  point  the  designated TICP  or  EICP  is
 displayed on the CRT screen, and the cross hair cursors
 are activated. If the data file consists of 260 or fewer
 mass spectra, the  entire TICP or  EICP  will be
 displayed. If the file contains more than 260 spectra,
 only the first 260 points are displayed. A block of 260
 spectra is referred to as a screen  load. The TICP's and
 EICP's are displayed  in 260 spectra screen loads in
 order to allow sufficient accuracy in positioning of the
 cross hair cursor on a given spectrum number.
  With the first screen load displayed the user has only
 two options. Entry of a series of RETURNS will cause
 all  sections of the TICP or EICP id be displayed in
 screen loads of 260 spectra with a 20 spectra overlap
 between each. Alternatively with any given screen load
the user may initialize a Q-file by entry of the character
Q. This will activate the S and B keys until Q is entered
again which closes the Q-file.
  With the S and B keys activated the user may save
spectrum  numbers in the Q-file by positioning the
horizontal and vertical cross hairs over the desired
point on the profile, and striking S or B. The S key is
normally   used   to  save   spectrum   numbers
corresponding to the apexes of peaks, and the B key is
used to save the  corresponding background spectra.
The B  key cannot be used except after an S is entered.
However,  a succession of S's may be entered without
saving  any background spectra. The spectrum numbers
saved are  not required to  correspond  to apexes and
background. For example,  they could be used to save
spectrum  numbers that correspond  to the  limits  of
intergration   for  a quantitative analysis   program.
Entering a RETURN will cause the next screen load to
be  displayed while the Q file is still open. At the
conclusion of the saving of all spectrum numbers, the Q
key must be pressed to close the Q file and cause it to  be
saved in the disk directory.
  After the last screen load  of a given data file  is
displayed, certain single  character  commands are
activated  to  permit  other displays  or plots. These
commands are the O, RETURN, F, R, and H. Several
of these commands return to the RANGE? prompt for
: specifications for displays, and several others cause the
program to remember or forget previous specifications
for CRT plots.
  Entry of an O causes the program to return to the
RANGE? prompt for specifications for another TICP
or EICP that will be overlayed on the first plot on the
CRT.  Up to  six plots may be overlayed on the CRT,
and the program  remembers which was specified first
and most  recently. There is a reported bug in the  O
option that may be present in other options. The bug
appears only when a Q file is initialized during the
viewing of a TICP. If the same TICP is overlaid with  an
EICP  using  the  O  option, the TICP  is  displayed
incorrectly.                     x
   Entry of a RETURN causes the program to forget
all  specifications for previous displays and return to the
RANGE? prompt.
   Entry of an F causes  the program to forget  all
specifications for previous displays except the first, and
to display the first TICP or EICP specified on the CRT.
This command has a slight bug in some versions of the
program.  It assumes the  first display specified was a
TICP; but, if it was an EICP,- the data will be displayed
correctly, but the EICP will be labeled a full mass range
plot.
                                                73

-------
  Entry of an R causes the program to  forget all
specifications for previous displays except the first, and
to return to the RANGE? prompt for specifications for
a TICP or EICP. This is then displayed on the CRT
overlayed with the first plot specified.
  Entry of an H causes the last TICP or EICP specified
to be plotted  on the plotter. Before this begins  the
OVERLAY? prompt is printed. A negative response
causes the plotter to draw the spectrum number axis
first. An entry of Y_ causes the plot  without the axis.
The  program  then returns  to the  COMMAND?
prompt. At this  point only the O and R as defined
above have any  meaning.  A  RETURN causes  the
program to return to the RANGE? prompt and forget
about all previous specifications.
MASS SPECTRA HISTOGRAMS
  For mass spectra displayed on a CRT or plotter the
dialogue is as follows:
   SELECT MODE:
   FILE?   CINTAP
   SPECTRUM?
CRTPLT
A summary of all commands is as follows:

                   Character    Function
                          Q    Initialize or close a Q
                               file of spectrum  numbers

                          S    Save a spectrum number
                               designated by the cross
                               hair cursors in  a Q file

                          B    Save a spectrum number
                               designated by the cross
                               hair cursors in  a Q file     ;

                  RETURN    Display next screen load
                               on  a  CRT

                          O    Overlay the next TICP  or
                               EICP specified  on the CRT

                  RETURN    Forget all information
                               about previously specified
                               TICP's or EICP's and return
                             .  to the RANGE? prompt

                          F    Forget all information
                               about specified  TICP's or
                               EICP's except the first,
                               and display it on the CRT

                          R    Forget all information about
                               previously specified TICP's
                               or  EICP's  except the first,
                               and overlay it with the  next
                               one specified on the CRT

                          H    Plot the last specified
                               TICP  or EICP on the plotter
              When Operable

              Anytime
              After a Q file
              is initialized
              After an S command
              Before final  screen
              is displayed

              After final screen
              is displayed

              After final screen
              load is displayed
              After final  screen
              load is displayed
              After final screen
              load is displayed
              After final screen
              load is displayed
                                                74

-------
The response to the FILE? prompt is the usual data
filename. There are three possible responses to the
SPECTRUM? prompt. Entry of a spectrum number
leads to the additional prompts as follows:

   SPECTRUM?   2_
   BACKGROUND SPECTRUM?    ]_
   AMPLIFY  BY?

If no background subtraction is desired, the user should
press  RETURN. Entry of a background spectrum
number  leads to the AMPLIFY BY? prompt. An
amplification factor of 1-4 may be optionally entered.
This causes multiplication of the background spectrum
by the factor  before it is  subtracted. Generally this
feature  is not  useful, and the user should  press
RETURN. At this point the spectrum is displayed on
the CRT, and the  cross hair cursors are activated.
Interactive examination may begin as described later in
this section. To  return  to the SPECTRUM? prompt,
press RETURN.
  In place of the entry  of a single spectum number as
described above, the user may enter the commands QC
or QP.  Both  commands tell the program  that the
spectrum numbers for the displays are to be retrieved
from  a  Q  file  generated  by BCLRGC. The  QC
command causes plots on the  CRT, and  the QP
command causes plots on the  plotter. Each command
leads to the Q-FILE? prompt:

   SELECT  MODE:  CRTPLT "   •'
   FILE?   CINTAP
   SPECTRUM?   QC
   Q-FILE?   QCINTA

It is extremely important that the user enter a true Q
filename generated by  BCLRGC. Simply  pressing
RETURN causes the program to hang, and there is no
apparent recovery procedure.  Entry of a -non-existent
filename causes the program to generate spectrum plots
from  apparently  random  numbers  that have  no
apparent meaning.
  With the QP command a series of spectra denned by
the Q file are plotted on the  plotter without further
operator action.  The plotter program keeps a record of
the length of each  plot. Therefore, if the pen is set
initially  one inch to the right of a paper fold, each new
spectrum will be started one inch to the right of a fold.
When the Q file is exhausted, the program returns to
the system prompt.
  With the QC command the first spectrum defined by
the Q-file is displayed on the CRT, and the cross hair
 cursors are activated.  Interactive examination of the
 spectrum  may begin,  or  the user may press  the
 RETURN to observe a CRT plot of the next spectrum
 in the Q-file. When the Q-file is exhausted, the program
 returns to the system prompt.
   Regardless of how CRT plots are generated, i.e., by
 entering individual spectrum numbers or with a Q-file,
 interactive processing  is possible when the cross hair
 cursors are  displayed. Four  commands have special
 meaning, namely B, E,  R, and H.
   The B  command is used  to  blow-up  (expand) a
 section of a spectrum defined by the cross-hairs. To use
 this command, first place the cross  hairs over  the
 smaller of the two masses that define the section to be
 expanded, and press B. Then place the cross hairs at the
 larger mass and press B again. The section defined  will
 be expanded to the maximum extent  possible. If the
 larger mass is denned first, the program blows up.
   The E command is used to multiply all abundances
 by a single  integer  expansion factor.  The masses to
 which the expansion factor will apply are defined by the
 position of the cross hairs. To use this command, place
 the cross  hairs over the mass that defines the starting
 point  for abundance expansion, e.g.,  mass 300,  and
 press E. This causes the EXPAND BY? prompt to be.
 printed. Enter an integer  expansion factor, e.g.,  10,
 press return, and the display will be redrawn with all
 abundances above mass 300 multiplied by 10.
   The R  command restores  the original spectrum to
 the screen. The B and E commands are cumulative and
 act on the current screen display, but R always restores
 the original  spectrum. The  H command  causes the
 current CRT display to be plotted on the plotter. If this
 is the last spectrum of a Q-file, the program returns to
 the system prompt. Pressing a RETURN will either
 cause a CRT display of the next spectrum in the Q-file,
 or a return to the SPECTRUM? prompt if not in the Q
'mode. Any time a prompt is  printed on the screen the
 user may enter CTRL/L  and return to the system
 prompt.
 EXTENDED MEMORY  OUTPUT
 SOFTWARE
   The program SPRRGC  performs very much like
 BCLRGC except that it requires 16K of core memory,
 and the TICP or EICP  is generated using only the
 abundances that maximize at each spectrum number.
 For example, if the mass  149 abundances at spectrum
. nuinbers  100,  101, and  102 are 40, 90, and  40
 respectively, the abundance at spectrum number 101
 would be  retained and the abundances at spectrum
                                               75

-------
numbers 100 and 102 would be discarded. The program     The program SPRPLT, and its overlay  SPRVL1,
uses the overlays RGCOV1 and RGCOV2. A minor   display spectra analogous to CRTPLT, but produce
difference from BCLRGC is that if the full mass range   only mass maximized spectra as described above. One
measured is selected for display, the first mass scanned   difference is that the RESOLUTION?: prompt, is given
is ignored. This was intended to compensate for some   and the only legal responses are 1, 2, or 3. The exact
noise  caused   by  some  interfaces  that  gave false   meaning  of this resolution is not clear. There is no
abundance  data  at  the  first  mass measured.  The   provision for background subtraction, but the QC and
program  SPRRGC  will  not  function  correctly  if   QP commands are enabled  as  are the  interactive
saturated peaks are present in the data file.              options described under CRTPLT.
                                               76

-------
                                      CHAPTER 5
                      COMPOUND  IDENTIFICATION
  The compound identification chapter is not intended
to give complete, step-by-step instructions on how to
identify organic  compounds  by  mass spectrometry.
That would be virtually impossible. The purpose of this
chapter is to set into perspective the techniques that are
available for compound identification. Major emphasis
is on quality assurance which is of utmost importance
in the environmental field, and guidelines for accurate
identifications are  presented  to  assist the  users of.
GC/MS systems.
5.1 INTERPRETATION  FROM
THEORY

  Interpretations  of  mass  spectra  for  compound
identification may be attempted based on the theory of
mass spectra and the rules of fragmentation of organic
ions in the gas phase.  Unfortunately, this approach is
frequently slow and  tedious and generally requires
considerable experience. It is often difficult to sustain
the necessary accurate deductive reasoning process for
the long periods of time required to make a large
number of correct identifications. Finally, not enough
is known about the detailed processes that occur in the
fragmentations of organic ions, and this severely limits
the approach.  However, when all  else  fails,  this
approach is clearly justified in  important situations,
such as an enforcement action or a fish kill.
  This method  will not be considered further  in this
manual. However, a number of references to excellent
books are included in the bibliography, and many of
these discuss interpretation of mass spectra.

5.2 EMPIRICAL SPECTRUM
MATCHING

  The purely  empirical method of searching a file of
reference mass spectra to find a similar or exact  match
of an experimental mass spectrum  has been  under
development for a number of years.  Any empirical
search  and match system  has  two  fundamental
components:  a data base, which is nothing more than
an  organized collection  of reference  spectra, and a
system to search the database.  Databases may be in
printed  or machine  readable  forms  and countless
search systems  are possible.  In this chapter  several
printed  spectral collections  are  recommended and
computerized spectral search systems are discussed.

SEARCHING PRINTED DATABASES
  Manual  searching  of  printed  databases  was
developed long before computerized systems, and some
elaborate indexing schemes were invented to facilitate
the  user's search.  Nevertheless,  all  manual  search
systems  are rather slow, subject to human error, and
intellectually fatiguing. It is also difficult and expensive
to update the database since the index usually requires
complete revision. In spite of these limitations, this
method  may be  rewarding  for  small  numbers of
unknowns if a printed database is available.
  For manual  searches, the two most  useful mass
spectral  catalogs are the "Eight Peak  Index of Mass
Spectra", published by the Mass  Spectrometry Data
Centre  (MSDC),  Aldermaston,  England, and the
"Compilation of Mass Spectral Data" (10 peak index)
by Cornu and Massot. (See the Bibliography for further
information.)
  The "Eight Peak Index" consists of three tables in
three volumes (31, 101 spectra):

  Volume  1.
        Table 1.   Spectra are arranged  in ascending
        molecular  weight  order,  subordered  on
        number of carbon atoms, hydrogen atoms, etc.
  Volume 2.
        Table 2.   Spectra are arranged in ascending
        molecular weight order, subordered  on m/e
        values in order of decreasing abundance.
  Volume 3.
        Table 3.   Spectra are arranged in ascending
                                                77

-------
       rn/e value order, each given where present as
       first,  second,  and  third  most  abundant;
       subordered on other m/e values in  order of
       decreasing abundance.

  An enlarged  edition of the  Compilation  of Mass
Spectral Data (10 peak index) was published in 1975 in
two volumes with three sections. This edition contains
mass spectral information on 10,000 comounds.

  Part I.   Spectra are ordered by molecular weight,
       giving the compound name, reference number
       in the original collection, molecular weight,
       molecular formula, and a listing of the 10 most
       intense peaks and their abundances.
  Part II.   Spectra are ordered by molecular formula
       (in order  of C,H,D,Br,Cl,F,I,O,P,S,Si, and
       others)  and includes the same  information as
       Part I.
  Part III.   Spectra are ordered by  frangment ion
        values in order of increasing m/e values, and
        includes the 10 most abundant peaks.

  The "Eight Peak Index" is usually  examined first,
because it contains three times as many spectra.as the
"Compilation of Mass Spectra". In both catalogs, the
fragment  ion index  is  the  most  useful.  Detailed
instructions for use  of these catalogs  is beyond the
scope  of  this  manual,  but  these  instructions are
included in the respective catalogs.
  Perhaps the most difficult part of manual searching is
the judgment of how well a reference and experimental
spectrum  match.  Most spectra in the catalogs were
measured     with    magnetic    deflection    mass
spectrometers. There  is a widespread belief that spectra
measured   with  quadrupole  and  other   types  of
spectrometers do not  agree with spectra from magnetic
deflection spectrometers. However, this problem is not
at all serious when the spectrometer is tuned according
to procedures in chapter 2. With proper ion abundance
calibration, quadrupole measured spectra often  agree
within a  few percent with  magnetically  measured
spectra. Other  factors,  such as  differences in inlet
systems, should cause greater spectral  differences, but
these are difficult to recognize in spectral collections.
  Comparisons  of  printed   lists  of masses  and
abundances  are  especially   tedious  and   it  is
recommended that final decisions concerning matches
be made with the histogram plots in the "Registry of
Mass Spectral Data", Volumes I-IV, by Stenhagen, et.
al.  There  is also available  the "EPA/NIH  Mass
Spectral Data Base" published by the National Bureau
of Standard's Office of Standard Reference Data. This
presents the 25,566 spectra of the NIH-EPA collection
in  bar  graph  form.  The spectra  are ordered  by
molecular  weight  and  each includes  the  molecular
formula, molecular structure,  Chemical  Abstracts
Service  registry  number, and compound name. The
publication is designated NSRDS-NBS  63 and the first
supplement will   contain  about  8000  additional
spectra.

REMOTE COMPUTERIZED
SEARCH  SYSTEMS
  The application of computers overcomes some of the
problems of  manual  searching,  but   computerized
search systems are constrained by the size and validity
of the  database, the thoroughness of the  searching
algorithms, and the cost of using the systems.
  An excellent mass spectral search system (MSSS) is
available on a commercial time sharing system at a
moderate cost. This MSSS was developed with the joint
and cooperative support of the Mass  Spectral Data
Centre - an agency of the British  Government, the
National Institutes of Health, the U.S. Environmental
Protection  Agency, and several other agencies. The
principal feature of the MSSS is that it  provides
different  search options and a single database in one
convenient  system. Table 5.1 lists  the  many different
search options currently available.

Table 5.1. User Options Available with the Mass
           Spectral Search System
Command        Function

PEAK     Finds all spectra that contain masses and
           abundances that  are  designated  by the
           user                      '
MW       Finds all spectra of compounds  with a
           molecular weight that is designated by the
           user.
MF        Finds all spectra of compounds  with a
           complete  molecular   formula   that  is
           designated by the user.
KB        Computer  to  computer transmission  of
           spectral data for matching; best for large
           amounts of data; Biemann algorithm
PMW      Combines  peak and  molecular  weight
           searches;  powerful  if you  know the
           molecular weight.
PMF      Combines  peak and molecular  formula
           searches;  powerful  if you  know the
           molecular formula.
                                                 78

-------
 Command       Function
MWMF    Combines   molecular   weight   and
            molecular formula searches; fastest way
            to  find  the  spectrum  of  a  known
            compound.
SIM        Computes a similarity index between two
            sets of data; automatic on KB search.
SPEC       Prints all  masses and abundances from
            spectra in the  file; user must supply an
            identification number.
FICHE     In manual mode, prints  fiche  numbers
            and matrix positions of spectra in the file;
            user must  supply  an   identification
            number;  also  in computer mode may
            drive one model automatic fiche viewer.
PLOT      Plots mass spectra from  the file  on a
            graphics  device;  user must  supply an
            identification number.
COM       User may enter comments, complaints,
            and suggestions.
NEWS      Prints news of the system
EXIT,
END,OUT  Causes  the mass spectral search system
            to fade away.
AUTH      Searches  for  literature  references  by
         ,   specific author.
COMP      Searches  for  literature  references  by
            specific compound.
ELE        Searches   for  liteature  references  by
            specific elements.
HELPxxxx  Prints explanation of the option xxxx.
INDEX     Searches literature references  by specific
            index terms.
LAB        Calculates the enrichment  of isotopically
            labeled compounds.
META      Calculates the m/e  values of various
            metastable ions that correspond to a given
            parent/daughter pair.
MOLFOR   Calculates   all   possible   molecular
            compositions that have approximately the
            same accurate mass.
PBM       A  reverse  mass  spectral   matching
            program..
PF         Finds all spectra of compounds with a
            partial   molecular   formula   that   is
            designated by the user.
PA         Searches for proton affinities
SUBJECT   Searches   for  literature  references  by
            specific subject.
OPT.HELP Prints this list of options
   For all of the spectral search options in Table 5.1
 except KB and PBM, the user enters mass,  abundance,
 or   other   data  from   a   keyboard/printer  or
. keyboard/cathode  ray  tube (CRT)  terminal that is
 interfaced to a conventional voice grade telephone line.
 This terminal may be completely independent of any
 GC/MS data system. Data entered by the user are
 transmitted to the time sharing computer that searches
 the database and prints interim search results at the
 user's terminal  in  a  few seconds. The user employs
 these interim  results  to  make judgments about
 additional data to  be entered to improve the search
 results.  This  interactive,  iterative  procedure   may
 continue until the user is satisfied that the database was
 thoroughly searched. The usual result is that a  small
 number  of spectra  are found  and  reported by
 identification numbers  and names. The  final choice
 .among these is the responsibility of the user. Complete
 spectra may be printed at the user's terminal or plotted
 on a graphics device.
   With the KB option,  a completely different mode of
 operation  is  employed. This is not an  interactive,
 iterative option, but a number crunching system to find
 the  best overall match using  the  Biemann search
 algorithm. Most importantly, with the KB  option data
 may be  transmitted automatically  from  the  mini-
 computer of the GC/MS data  system to the  time
 sharing computer  or entered  from a keyboard. The
 remote computer conducts a search for a match  based
 on the transmitted mass and abundance data, and sends
 the results  back to the mini-computer in  a matter of
 seconds.  Direct, computer controlled data  transfer
 requires special communications hardware and several
 programs for the GC/MS data system. The hardware
 and programs  are available for a number of  mini-
 computers commonly used in GC/MS data systems.
   The choice between automatic transmission or one of
 the  keyboard   entry   options  depends  on   the
 circumstances. For a few spectra or for special feature
 searches, e.g., by molecular weight, the keyboard entry
 option is best. For a  large number of spectra, the KB
 option with automatic data transfer is preferred. The
 user makes the choice  after calling  the time sharing
 system and may change options repeatedly during  a
 time-sharing session. Another option which attempts
 to find the best overall match is PBM, or probability
 based matching. Presently this requires keyboard entry
 of the entire spectrum, and its  use is limited by this
 requirement.
    If the automatic transmission KB option is' selected,
 the user must run a spectrum abbreviation  program on
 the  mini-computer before  calling  the MSSS.  This
                                                 79

-------
program, which selects the two most abundant ions in
each 14 amu interval, is used as follows:

   SELECT  MODE:   BRVCCP
   STORAGE  FILE:   SELECT A NAME  FOR
      YOUR  ABBREVIATED  SPECTRUM FILE
   SPECTRUM  FILE:   THE  NAME  OF  THE
      FILE  CONTAINING THE  SPECTRA YOU
      ARE  ABBREVIATING
   SPECTRUM  NUMBER:   46
   BACKGROUND  SPECTRUM:    43
   MINIMUM VALUE:    press  return
   ANOTHER SPECTRUM:   YES  or  NC)
      depending  on  how many  spectra  you
      are  abbreviating

  if YES  —cycle will repeat enabling you to put a
         number of abbreviated spectra in one file
  if NO   —the prompt will be ANOTHER FILE?
         which  enables  you to abbreviate spectra
         from another data file stored on the  same
         disk.
  If NO again — system returns to SELECT MODE:

  After  generation of the abbreviated spectrum  file
another  program is used  to transmit  the data. This
program, named MDIREK, uses the Tektronix 4010
or 4012 or 4014 communications interface and will not
work with the Digital Equipment Corporation KL8E
communications interface. The BAUD rate may be
110-9600 and the program will work with any rate
your modem  handles and  the time sharing system
supports. The program is used as follows:

  SELECT MODE:   MDIREK
  SPECTRAL DATA COMMUNICATION
      PROGRAM
  FILE?:  NAME OF STORAGE FILE
      SELECTED  ABOVE.
Place  the  communications interface  three position
switch in the middle position, select a BAUD rate, full
duplex, and call the time sharing system.
The standard login procedure is used and a KB is given
in answer to OPTION?:
  OPTION?:   KB
  COMPLETE SPECTRUM SEARCH
  MAIN FILE(Y OR N)?  Y_
  INPUT TWO TITLE LINES FIRST*"
Place the switch in  bottom  position and  enter two
contol P's followed by a control A. The first control P
causes   the  file  name,  spectrum  number,  and
background spectrum number to print  on  the CRT.
The second control P causes  the sample title to print
out. Control A halts the PDP-8. Place the switch in the
top position and press continue. At the end of the
spectrum transmission, DATA OK? will be  printed. It
is important at this point to place the  switch in the
middle  position  so  that the next  spectrum to be
transmitted is not disturbed.  If the data is  acceptable
enter a Yj if not enter a N_and the data may be edited.
Upon  completion of the search,  the  system  will
transmit all hits with their similarity indexes. It will
.then prompt with CONTINUE? If the response is NO,
you will exit from the KB search. If the response is
YES, the prompt will be SAME SPECTRUM (Y_or
N).  Continue  as  above  and  transmit  the  next
abbreviated spectrum, i.e., if you abbreviated a second
spectrum in the same file you  may send it by repeating
the sequence of steps beginning with placement of the
switch in the bottom position.

MINICOMPUTER SEARCH SYSTEMS
  An alternative  to the  large, remote,  time  sharing
system is the use of a search and  match system on the
mini-computer of a GC/MS data system. Several data
system manufacturers offer this kind of system and its
use may offer important time and cost savings.  The
most serious limitations of local database searching are
the size of the database, the quality of the database, the
difficulty of updating the database, and the very limited
searching capability usually offered with these systems.
  The EPA minicomputer search system operates as
an integral part of the PDP-8 GC/MS system software.
The hardware required to use this system is a PDP-8
with  8K of core memory,  an  extended  arithmetic
element (EAE), two disk drives, and a printer such as a
teletypewriter,  Decwriter,   or   line   printer.  An
alternative  search program  operates in  16K  and
reduces search time by about  5-10 seconds per search.
This is recommended for searching very large numbers
of spectra, e.g, to match 100 spectra in 8K  requires
about 17 minutes longer than  matching 100 spectra  in
16K.
  The files required to use this system are not all listed
in the disk directory even though one or more of them
may  be present.  In addition,  some files  are write
protected so they are less likely to be clobbered by a
user error. Finally, it is required that certain files be on
certain disks in specific drive units, or all will fail. The
files  required  and  their directory  status, write
                                              80

-------
protection status, and  disk  requirements are listed
below.
BMATCH  and  BMNBVT.  These  files  should  be
copied on all disks that may be used to store data files.
They are conventional directory listed files that are not
write protected. The program BMATCH is used for
spectrum abbreviation and BMNBVT is an  overlay.
Spectrum  abbreviation  is  the  same  function  as
accomplished with the program BRVCCP when using
the KB option and the remote search system. However,
its  capabilities  are significantly  enhanced  by  the
IGOOS software described in section 4.3.
BMN8K and BMN16K.  These are the actual search
programs  in  8K  and  16K  versions.  They  are
conventional directory  listed files  that are not write
protected.  They should be on only one disk, plus a
backup, which is a special skeleton system disk stripped
of the  normal  data  acquisition  and  data  output
software. This skeleton system is initially placed in disk
unit  1,  then  moved  to disk  unit 0  after spectrum
abbreviation is finished.  The file of abbreviated spectra
is also placed on this disk.
NAMES.  This file is very special. It  is contained on
the special skeleton system disk, but it will not appear
in the disk  directory,  and it  is  write-protected.  It
occupies blocks 10000-31277. Therefore, these blocks
are forever  inaccessible and  any attempt to write
another file into them will result in failure. The number
of.files on the special skeleton system disk must be kept
small and not exceed block  7777. The  NAMES  file
contains Chemical  Abstracts  nomenclature  for  all
compounds  in  the  database, a  unique Chemical
Abstracts  Registry  number  for  each compound, a
molecular formula, molecular  weight, and a purely
arbitrary identification number.
PRNTSP.  This is a conventional directory listed file
that is not write-protected. It is on the skeleton system
disk and may be used to print the masses and relative
abundances of abbreviated spectra as they appear in the
database. Please note that the database contains only
abbreviated mass spectra.
PRESEARCH.  This is.a special file that is contained
on a disk that has no operating system and cannot be
started with the bootstrap loader.  This disk is always
placed in disk unit 1 after the abbreviated spectrum file
is generated on the skeleton  system and the skeleton
system is moved to disk drive unit 0. This file contains
the mass of the  base peak for each spectrum and other
search algorithm related information. This file is write-
protected.
ABBREVIATED  SPECTRUM   FILE.  This  is  a
 special  file  that  is  on  the  same  disk  as  the
PRESEARCH file, and is write-protected. It contains
mass and abundance data for the two most abundant
ions in each  14 amu window for 25,555 mass spectra.
No two spectra  have  the same Chemical  Abstracts
Registry number.
DATA FILES.  When using  the search system it is
strongly recommended that the user have descriptive
(up to 64 characters) titles associated with each data
file.   Abbreviated   spectrum   files  may  contain
information  from  one  or  several  data  files, and
descriptive titles  will improve the readability  of the
printed report form BMN8K or BMN16K.

Procedure (PDP-8).
   1.    Place  the  disk  containing  the data  file(s),
  BMATCH, and BMNBVT in disk unit 0. Place the
  skeleton operating system in the disk unit 1. Run
  BMATCH when  both disks  are  up to speed  as
  described below using either Q file or keyboard entry
  of spectrum numbers. The Q file spectrum number
  entry mode is strongly  recommended  for large
  numbers of spectra and is described first:

   SELECT  MODE:    BMATCH
   STORAGE FILE?    TURKEY
   FILE?   CINTAP
   SPECTRUM?   (3
   Q-FILE     QCIN
   FILE?   CINTAP
   SPECTRUM?   (3
   Q-FILE?    Q2CIN
   FILE?   6SPK
   SPECTRUM?   Q_
   Q-FILE?    Q26SO
   FILE?   press return

The storage file is the file that will contain abbreviated
spectra from the data files. Each abbreviated spectrum
requires three disk  blocks.  The  storage file will  be
generated on the skeleton system in  disk unit  1. The
FILE? prompt refers to the data  file on disk  0 that
contains spectra to be abbreviated. A response of Q to
the  SPECTRUM?  prompt  leads  to  the  Q-FILE?
prompt. The user must enter the name of a Q  file
created by BCLRGC  using the IGOOS described in
section 4.3.  This Q file contains  a  list of spectrum
numbers, and BMATCH will automatically retrieve
each, subtract background if called for, abbreviate  the
spectrum, and place the latter in the storage file on disk
unit 1.  When the Q  file is' exhausted, the program
returns to the SPECTRUM?  prompt. The user may
repeat the same process with another data file and  the
                                                 81

-------
corresponding Q file, or with the same data file and a
different Q file that also refers to the data file. Both
these options are illustrated in the sample dialogue.
Please note that all abbreviated spectra from  several
data files \vill be placed in the same storage file on disk
unit 1.
  After all spectra from all files have been abbreviated,
the user must press RETURN in response to the FILE?
prompt. If CTRL/L is entered, the program returns to
the system prompt, and all abbreviated spectra will be
lost.
  For keyboard entry of spectrum numbers the user
simply enters a spectrum number in place of the Q
command as in the following sample dialogue. If no
background subtraction is required, press RETURN.

    SELECT  MODE:   BMATCH
    STORAGE  FILE?   VOLS
    FILE?   STD1
    SPECTRUM?   56
    BACKGROUND?   £3
    SPECTRUM?   92-
    BACKGROUND?   88
    SPECTRUM?      1 17
    BACKGROUND?    114
    SPECTRUM?   127
    BACKGROUND    124
    SPECTRUM?   192
    BACKGROUND?    189
    SPECTRUM?   216
    BACKGROUND?   212
    SPECTRUM?   246
    BACKGROUND?   241
    SPECTRUM?   press return
    FILE?   press return

    SELECT  MODE:

  It is important to recognize that Q file and keyboard
entry of spectrum numbers may be mixed randomly by
the  user and all abbreviated spectra will appear in the
same storage  file on disk unit  1. If spectrum numbers
are  being entered by the keyboard, the user must press
RETURN twice to return to the system prompt as in
the example above.
  2. . Remove  both disk cartridges and move  the
skeleton system from  unit 1 to unit 0. Place the disk
containing the PRESEARCH and 25,555 spectra in
unit 1. Bring both cartridges up to speed.
  3.  Run BMN8K or BMN16K  as  shown  in  the
sample dialogue. The FILE? prompt  refers  to  the
storage file for the abbreviated spectra on the skeleton
system. The  user should  switch from the CRT to
console printer to allow the printed report to appear on
the printer.

   SELECT  MODE:   BMN16K
   FILE?     TURKEY

The printed report consists of the title of the file from
which the spectrum  was taken, the spectrum number
and background subtracted spectrum (if any), the
number of presearch hits (base peak only in presearch
file), and a list of best hits. For each hit the following is
printed:
  a.   Similarity index (SI) -  see discussion of
similarity index interpretations in this chapter.
  b.   The   Chemical  Abstracts   name   -  this
nomenclature system will often  generate very  long
names for commonly known substances, e.g., dieldrin.
Future versions of the NAMES  file  will include
common names.
  c.   The molecular composition - this is printed after
the name.
  d.  The molecular weight - this is printed after the
composition.
  e.   The Chemical Abstracts Registry number - this
is a unique  number that  permanently identifies the
compound. This should be included on all reports  of
the occurrence of this  compound to allow  users  of
reports to readily  look  up the properties  of the
compound in other printed or computerized data bases.
  f.   An arbitrary  spectrum  identification  number
that has no meaning beyond this search system.

  A sample  report  is as  follows:

  STD #1
  SPECTRUM 0056 - 052
    47   HITS
   SI=0.589
   ETHANE, 1,1-DICHLORO-C2H4CL2
   75343   #     301
98
   SI=0.250
   ETHANE,  1-BROMO-2-CHLORO-C2H4BRCL
   142    107040  #     1121
   81=0.195
   CARBONIC DICHLORIDE CCL20
   75445  #     307
                                             82

-------
   SI=0.153
   ETHENE,  (METHYLSULFONYL)-C3H6O2S
   106   3680022    #      7360

   SI=0.135
   ETHANE, 1-CHLORO-2-NITRO- C2H4CLNO2
   109   625478  #     3569

   SI=0.115
   ETHANEDIOYL  DICHLORIDE C2CL2O2
   126   79378   #     460

   SI=0.111
   PROPANE,  1,1,2-TRICHLORO- C3H5CL3
   146   598776   #      3247
When the abbreviated spectrum file is exhausted the
program returns to the system prompt. Clearly this
search system is oriented to processing large numbers
of spectra in an overnight or overlunch batch mode.
One optional program on the skeleton system allows
the user to print mass and abundance data from the
25,555  abbreviated  spectra  file.  The NUMBER?
prompt refers to the arbitrary spectrum identification
number. Pressing RETURN at this prompt causes the
program to return to the system prompt.
                              measurements were  not documented  with  sufficient
                              data to support their reliability.  This caused doubt
                              about the validity of the measurements and concern for
                              the correctness of correlations and proposed standards.
                                Several aspects of quality control were emphasized in
                              previous  chapters  of this  manual.  Reagent  and
                              glassware  control  is  required  to  minimize  the
                              introduction of contamination from the materials used
                              in the sample preparation procedures and  this  was
                              emphasized in the sample preparation  methods in
                              chapter 3.
                                Instrumentation control is required to assure  that
                              the total operating GC/MS system is calibrated and is
                              in proper working order. If a computerized GC/MS
                              system is used to collect data, the computer data system
                              must be included in the performance evaluation.  The
                              recommended instrumentation control procedure was
                              described in section 2.6. This employs a  standard
                              reference compound and a set of reference criteria to
                              evaluate the performance of the overall system. This
                              evaluation should be performed on  each  day the
                              GC/MS system is used to acquire data from samples or
                              reagent blanks.  The records  from  the  performance
                              evaluations should be maintained with the sample and
                              reagent blank records as permanent documentation
                              supporting the validity of the data. The major emphasis
                              of this section is the quality control that applies to
                              compound identification.
SELECT  MODE:
NUMBER?   301
PRNTSP
ETHANE,  1 ,  1-DICHLORO-  C2H4CL2
                        98     75343
MASS    INT MASS
    14      1     15
    61      6    63
   105      1
NUMBER?
     INT MASS
       1     26
    100    65
INT MASS
  6    27
 31    83
INT MASS
 40    35
 14    85
INT MASS
  1     47
  9     98
INT MASS
  1     60
 10    100
INT
  1
  7
5.3 QUALITY  CONTROL IN
COMPOUND IDENTIFICATION

  The  importance of analytical quality control  in
organic pollutant analysis  cannot be overestimated.
Data generated  in  surveys are being  used to set
standards for drinking water, air quality, surface water
quality, and effluents. Possible correlations between the
presence of organic contaminants in drinking water
and air and human health effects are under widespread
study.  In  the   past  many   carefully  conducted
                              EVALUATION  OF BLANKS AND
                              REAGENT BLANKS
                                The blank is an  experiment that is required for all
                              samples. A reagent blank is defined as an experiment
                              that employs all procedures, quantities of materials,
                              glassware, etc. used in the sample preparation except
                              that no water, air, tissue, or sediment sample is used. A
                              low organic water blank is recommended in place of a
                              reagent blank in several procedures,  e.g., inert gas
                              purging and trapping. One of these blanks is required
                              even when contamination from glassware and reagents
                                              83

-------
is well controlled.  The data from  the  blank is the
documentation  that proves that good  control  was
exercised, and it defines the level of background  that
was beyond control. The evaluation  may be a straight
forward comparison of corresponding peaks and mass
spectra in the blank and sample.
  An effective but  rapid technique for comparison of
blanks and samples employs the extracted ion current
profile (EICP)  of  one or several ions.  An EICP  is
defined as a plot of the change in relative abundance of
one or several ions  as a function of time. The data for
this  plot is  extracted from all  the ion  abundance
measurements  made over  the  mass range  observed
during the elution  of the separated  components from
the GC. The EICP produces an apparent increase in
sensitivity by subtracting from the total ion current
profile all the ion abundance data that is contributed
from background,  unresolved components, and other
irrelevant ions. The EICP generator is a standard data
reduction program  on all modern  computerized
GC/MS systems and is described  in chapter 4 and
section  6.1.  A fast graphics display device (CRT) is
required to facilitate reviewing a large number of EICP
plots.
  It is emphasized  that it is not necessary to have even
a tentative identification of a compound to apply this
technique to blank evaluation.  To  conduct an EICP
comparison, mass spectra of all peaks in the sample are
examined. One  or several ions that are prominent in a
spectrum from each peak are selected, and sample and
reagent blank EICP's are generated on the CRT. In
most  cases  comparison  of these EICP's  permits
straightforward judgments concerning the presence of
compounds  in the  sample and  the  blank. Figure 5.1
shows EICP's for mass 171 from a sample and from the
corresponding  reagent blank.  Clearly  a  compound
having a mass 171 ion is present in the sample, but no
corresponding peak is observed above the noise level in
the reagent  blank.  Figure 5.2  shows a sample,and
reagent blank that  contain  the  same compounds by
EICP analysis.  For a valid comparison this technique
does require exactly the same GC and data collection
conditions for the sample and blank. It is possible that
the compounds in  the blank and sample that elute at
the same time are different, but contain the same ion or
ions. If this is suspected, the entire spectrum from  each
must be examined to confirm or deny this possibility.
  If the concentration of a blank component equals or
exceeds the  concentration of sample component, the
decision  is   clear and the compound  must not be
reported. A far  more difficult judgment must be made
when the concentration of a sample component exceeds
its concentration in the blank. The material could, of
course, be a true sample component. Alternatively it
has been observed empirically that compounds in the
blank  sometimes merely  appear  to  be  at  lower
concentrations  than  the same  compounds  in  the
corresponding sample. Figure 5.3 shows the total ion
current profiles from a sample and  a  corresponding
reagent blank.  Careful  comparison of  the  profiles
reveals a very similar pattern of peaks and valleys in
certain  areas, e.g., spectrum numbers 170-190 and
235-245,   yet   a  significantly   lower  apparent
concentration in the reagent  blank. There are several
possible reasons for this. One rationalization used, in
the case of solvent extracts,  is that impurities in the
solvent of the  reagent blank  are adsorbed  more
efficiently onto the drying agent and other surfaces.
With extracts containing some water, the wetting effect
of the water precludes efficient adsorption on surfaces,
and impurities are  carried  on  in  the  solvent.
Alternatively, certain solvent impurities  may be  lost
more readily from the blank than from  the sample
extract during concentration. Perhaps  the  general
organic  background  matrix  in  the  sample  extract
retains the solvent impurities. Both explanations are
reasonable but unproved. In view of the uncertainties,
any compound that is observed in the sample should
not be reported if it is part of an overall pattern of peaks
that is repeated in  the blank, although  at a  lower
apparent concentration. This same overall pattern will
usually persist in the acid, neutral, and basic fractions
of a solvent extract.

SUPPORTING EXPERIMENTS
   Chemical  ionization, field  ionization,  and  high
accuracy mass measurements are GC/MS techniques
that are capable of generating very strong evidence in
support of identifications. Some of these are discussed
in Chapter 6. However, the production of this evidence
is restricted because only a relatively few laboratories
have developed capabilities with these techniques. High
accuracy mass  measurements are further limited by
sample size, since  some  sacrifice  in  sensitivity is
required to achieve the high accuracy.
   After a tentative identification is made, several other
types  of supporting  experiments  become possible.
Retention  time data  from  the  GC/MS of a  pure
compound  (standard)  may  be   compared   with
analogous data from the sample component. Similarly.
the mass spectrum of the standard, obtained under the
same conditions that were used for the sample, may be
compared with the sample component spectrum.  The
standard may be dissolved in water at  an  appropriate
                                                 84

-------
oo
en
        s?.
        8.
        p.
        8.
         L

        8-
        8.
MASS171 IN REAGENT BLANK
                             MASS 171 IN SAMPLE
             10  20  30  10  SO  GO  70  80  SO  100 110  130 130  1« 138 ISO  170 ISO  138 TOO ZIO  ZZO 230 21O 250 2EO  270 380 Z30 380 310  3ZO 330
                        Figure 5.1  The extracted ion current profiles for mass 171 from a sample and the corresponding
                                   reagent  blank.

-------
                                MASS 149 IN  REAGENT  BLANI
                                                I
ID  US  30  10  SB
      HfffR
                    70  80  30
MASS 149 IN SAMPLE
                               	"l""'""l""""'l	I	I	l""l'"1	1	1	1	|""I"VI	I	~.|i..y...|....l...1	[	I	•••!•'	I'";	I
                               JOB 110  120 130  110 ISO  160 170  180 130 200  210 2SO  230 2K3  ZS3 260  Z10 2BO 330 300 310 32O  330 3W 3SO 3EO  370 380 330
                    Figure 5.2   The extracted ion current profiles for mass 149 from a sample and the corresponding
                                  reagent  blank.

-------
 8
 8-
 8.

 B.
*.
fc.
 8.
 »4
                  REAGENT BLANK
B  IB 20  38  «  S9 tO  78  to  SB  108  1H) ia 138 lie  ISO ISO 1TO 1»  ISO 2BS 218 228 230 2« 2SB 3SD 2TO  2BO 238 308 3JO 32D 338 310 368 3E8 37B
                   Figure 5.3  The total ion current profiles from a sample and the corresponding reagent blank.

-------
 concentration, isolated, and measured. The recovery of
 this spike in the same fraction in which the suspected
 component appeared, and the observation of equivalent
 mass spectra for the spike and the sample component,
 is strong evidence for confirmation of the identification.
REQUIREMENTS FOR PURE
STANDARDS
  Clearly  the  most  convincing  evidence  for  an
identification is obtained by the examination of pure
standards   corresponding  to   suspected   sample
components. However, the existence of this evidence is
constrained by the availability of the pure standard and
the additional cost and time required to examine it.
Because it  is not usually possible to predict which
compounds will be found, some standards will not be
available immediately. There are many very practical
limitations   imposed   on  the   development  and
maintenance  of a large  library  of pure authentic
standards. Many compounds are obtainable from fine
chemical supply  houses, but  procurement  time  is
variable and  may extend  to weeks or months. Some
compounds  are not  available  from  any  supplier,
frequently because they are by-products of industrial
processes rather than manufactured products. This
same limitation of standard availability also precludes
calibrated concentration measurements in many cases.
SIMILARITY INDEX INTERPRETATIONS
  Because of the problem of standard availability, it is
worthwhile    to   determine   whether    reliable
identifications   can  exist  without  standards.  One
criterion for a reliable identification that might be used
is a quantitative measure of the exactness of match
between  an  experimental  mass  spectrum  and  a
spectrum from  the printed literature or a  computer
readable database. A similarity index (SI), calculated
on a scale from zero to one, has been described and
used in several computerized mass spectrum matching
systems including the  KB option and PDP-8 search
systems described in this chapter. Experience with this
SI indicates that, in general, a value greater than about
four tenths corresponds to a reasonable match between
two mass spectra. A reasonable match often does imply
an identification, but sometimes it does  not. Position
isomers and  members  of  homologous  series  of
compounds  often  give  very  similar mass  spectra.
Another undetermined  number of compounds simply
are not  uniquely characterized by their mass  spectra.
Figure 5.4 shows the mass spectrum of an unidentified
compound  and  the  spectrum  of the  compound
chloropicrin, Cl.CNO,. the match is clearly good by
inspection  and  a1  SI  of  0.453   was  calculated.
Nevertheless, the unknown whose spectrum  is  in
Figure 5.4 is not chloropicrin as determined by the gas
chromatographic behavior  of the unknown and pure
chloropicrin.
  Another  problem  with  identifications based  on
empirical  spectrum  matching  is   that  significant
differences  in  ion  abundance  measurements  are
sometimes observed when the mass spectrum of a
compound is measured on two different spectrometers.
Some of these differences are probably caused by non-
uniform calibration procedures or by a failure to use an
ion abundance calibration procedure. Different types of
inlet  systems also may have  significant effects on
relative abundance measurements. A GC or batch inlet
system that is operated in  the 100-250C temperature
range   may   promote   temperature   dependent
fragmentations and reduce abundances  of molecular
and other high mass ions. With a well designed direct
inlet system, these temperature effects may be largely
precluded. As a result  of these factors, it is  quite
common for two spectra of the same compound, that
were  measured  with   different inlet   systems   or
spectrometers, to give a rather low SI. A low SI may
also   be  caused  by unresolved  or; partly  resolved
components  which generate  mass spectra containing
extraneous ion abundance measurements.
   It is concluded that the SI must be used with caution.
A relatively high SI may be regarded as an indication of
a  reasonable match, but  only  as  suggestive  of the
probability of an identification. A  relatively low  SI
cannot be regarded as complete rejection of a possible
identification.


THE QUALITY INDEX
  Another criterion for reliable identifications  when
standards are not available is based on an assessment of
the quality of the ions in the experimental and reference
mass  spectra. In the SI calculation, molecular ions
(M+),  molecular ions  having  naturally occuring
isotopes (typically M  + 1,  Cl, S,  etc.) and all key
fragment ions are weighted  the same as many  very
common fragment ions. However,  the M* ion, for
example, is unique in  every mass spectrum and has
significance to an identification that far outweighs most
other ions. Mass spectra may be categorized according
to the quality of the ions observed and a quality index
(QI) calculated that is a weighted similarity index.
                 QI  = SI * W

-------
  8
  8
 JS-
       SPECTRJi NLf€ER 161   -  1S3



       C1NC!N*TT! NElfTRFL
 SJ


 o
                                                                        Juu
 8
    20  30   «   SO   60   70   90   30   100  110  120   130  110  ISO  ISO  170  190
        M/ E


                  STf^CPPD
 S_
*8_
                                                                                        rf
                                                                                        5
    ...,.,..|...,,....|....|....|....1,...|....,....|,...,....J	,....|..n,r,..|....,....|T..^nnj ....... .|....,T. ..|. ........j. ...,„..j.. ..,,.„!„..,	|. ......


   20   30   10   SO  60   70   80   30   100   110   133  130  11O  ISO 160   170  180  130
Figure  5.4  The mass spectrum of an unidentified compound and the mass spectrum of chloropicrin.
                                            89

-------
:veral categories of quality are as follows:
 (a)   If the  molecular  ion is observed  and  the
      observed distribution of abundances for it and
      its isotope containing species are within 10%
      of the expected distribution, the W is 1.0. The
      significance of the SI value is considerably
      enhanced.
 (b)   If a molecular ion is observed but the isotope
      data are not within 10% of the expected value,
      lower confidence is assigned by a W of 0.75.
 (c)   Failure to observe a molecular ion but  the
      observation of key fragments that account for
      all the atoms of the molecular ion suggests a W
      of 0.5.  This factor may be raised or lowered
      between 0.4-0.6 depending on the observation
      of consistent isotope data in the key fragment
      ions.
 (d)   The lowest confidence  is placed on spectra
      which do not contain adequate fragment ions
        to account for all the atoms of the molecular
        ion. A W of 0.1 is assigned to these spectra.
  It  is recognized that  position isomers may not be
distinguishable under any circumstances, but this  is
often true even when pure standards are available.
  The quality index is amenable to additional positive
adjustments by 0.1-0.2 QI units if all major fragment
ions  are scrutinized and found to be reasonable and
compatible    with    the    assigned    structure.
Reasonableness should be based on compatibility with
the accepted principles of fragmentation of organic ions-
in   the  gas  phase.  With  magnetic   deflection
spectrometers additional fractional quality points may-
be added  if  fragmentations are  supported  by the
observation  of ions  from  the  decomposition  of
metastable ions.
  Good spectrum matches that have a QI of 0.75-1.0
are  considered reliable  identifications  when  pure
standards are not available.
                                                 90

-------
                                       CHAPTER  6
               ADVANCED  ANALYTICAL TECHNIQUES
  The purpose of this chapter is to describe several
GC/MS techniques  that are of a  more specialized
nature than the broad spectrum methods described in
the first  five chapters.  Each  of  these specialized
techniques  has   achieved  prominence   in   mass
spectrometry,  and each is expected to become  more
important in future applications of GC/MS.
  Selected ion monitoring  is a real  time  technique
whose operational philosophy differs sharply from the
broad spectrum approach emphasized in the previous
chapters.   Unlike the  broad  spectrum  approach,
selected  ion monitoring  requires knowledge of the
compounds that are to be measured. In a very real sense
it is the application of a mass spectrometer as a super
selective and sensitive GC detector. Its selectivity far
exceeds any known GC detector, and its sensitivity is at
least the equal of the most sensitive detectors known.
The technique is discussed in section 6.1.
  During the last fifteen years the emphasis placed on
qualitative  analysis  by  mass  spectrometry  has
somewhat   obscured  the   quantitative   analytical
strengths of the technique. Today quantitative analysis
by mass spectrometry is more significant than ever,
especially  when used with selected ion monitoring.
Quantitative analytical methods  are  discussed  in
section 6.2.
  The  development  of  open  tubular  column  GC
technology is expected to be very significant during the
next five years,  and reliable, long  lasting  precoated
columns will likely be available at relatively low cost
from a variety of suppliers. This technique is changing
so rapidly  that  section 6.3  is devoted largely to  a
presentation of  a  few basic  principles, and  to  a
discussion of several potential areas of application.
  Chemical ionization GC/MS is discussed in section
6.4,  and  because of the  very broad scope  of this
relatively new technique, the presentation is limited to
a few basic principles, quality control considerations,
and environmental applications.
  Precise mass measurements are currently beyond the
capabilities   of  most   environmental   GC/MS
laboratories. The brief discussion of this technique is
included in section 6.5 primarily for the information of
readers  who should be aware of the capabilities that
exist, and the advantages and  disadvantages of the
method.
6.1   SELECTED ION MONITORING
(SIM)

  The computer controlled mass spectrometer has two
general modes of operation as a continuous detector in
chromatographic systems. One mode is  to acquire
conventional mass spectra of  components as they
emerge from the chromatographic system. These mass
spectra are used to identify the individual components.
This broad spectrum approach was emphasized in the
first five chapters of this manual.
  The alternative mode  is  to   apply  the  mass
spectrometer as a substance selective detector. This
mode  is called selected ion monitoring (SIM) which is
defined as the dedication of a mass spectrometer to the
acquisition  of  ion abundance  data  at only selected
masses in real  time as components emerge from the
chromatographic system.
  Selected ion monitoring is  not a  new  principle in
mass spectrometry. A technique called peak stepping
or peak switching has been used for decades  in the
precise measurement of isotope ratios. In this classical
application,  the  ionic  abundances measured  were
usually limited to those separated by just a few atomic
mass units, for example "0 and "0. In recent years there
has been a very significant increase in the applications
of  SIM.  This was brought about  largely  by the
development of the computer controlled quadrupole
mass  spectrometer as vapor  phase chromatography
detector. The presence of the gas chromatography inlet
system permitted tn'e  introduction  into  the  mass
spectometer of very small samples of complex mixtures
in  an easily   handled liquid  form. The  computer
                                               91

-------
controlled quadrupole mass spectrometer provided a
method for high speed and very accurate and precise
ion monitoring over a very wide mass range. Finally the
dedicated  minicomputer  and its related peripherals
gave Uie. experimentalist, access to a  wide range of
control functions  and   real  time  ion  monitoring
techniques.

DEFINITIONS OF TERMS
  There  is  usually  a  period  of  confusion  in
terminology, concepts, etc. any time that technology is
rapidly advanced by  a  number of individuals and
organizations in a relatively  short time period. The
application of a mass spectrometer  as a substance
selective detector in chromatography is no exception.
The  terms  accelerating  voltage  alternation,   mass
fragmentography, single  ion  detection, and multiple
single ion detection are among a number of terms that
have been used to describe this technique. An analysis
of this  terminology led to the recommendation of a
standard term, selected ion monitoring  (SIM), because
it best conveys to the reader the significant information
about  the  technique  that sets  it  apart  from  other
techniques. The term SIM is general in  that it does not
imply a particular type of spectrometer, the number of
ions measured, or the type of ions measured. It must be
recognized  however  that   SIM  is   a  real   time
measurement technique and that a designation is also
required for the output obtained from SIM. The name
selected   ion   current   profile  (SICP)  has   been
recommended as the most meaningful and appropriate.
Consistent with the definition  of SIM, a SICP is then a
plot of the change in ion abundance as a function of
time using abundances measured by SIM.
   Clear precise terminology is particularly important
in computerized GC/MS  work  because there  are
several other  widely  used  techniques that  may be
confused with SIM  and the SICP. It is  important to
understand these techniques to appreciate the overall
advantages and  disadvantages of SIM. Perhaps  the
most widely used real  time data acquisition technique
in GC/MS is the continuous repetitive measurement of
spectra (CRMS). Figure 6.1 contains  a schematic
diagram of CRMS and two types of data reduction that
are in common use. The sawtooth diagram in the top of
Figure  6.1 is  a  representation of CRMS during  the
elution of components  from a gas chromatograph.
Each   solid line  represents  a  sweep  of the  mass
spectrometer Trom an  arbitrary starting mass, e.g., 40
amu, to an arbitrary ending mass, e.g.,  400 amu.  Each
dotted  line  represents  the  resetting  of the   mass
spectrometer sweep control to the starting mass. In a
typical  GC/MS  run,  several  hundred  to  over  a
thousand mass spectra may be acquired in this  way.
Each sweep of the mass range usually requires a time in
the range of 2-5 seconds, but faster or slower scans may-
be used in some cases. The second diagram from the top
in Figure 6.1  merely shows that each  solid line of the
sawtooth represents a mass spectrum as displayed in
the standard  histogram format. The  most important
idea is that CRMS produces a set of mass spectra that
are more or less complete depending on the selection of
the mass range. Each integer mass between the starting
and ending masses is measured and recorded. This is in
sharp contrast to SIM where measurements are made
at only a few masses in real time.
  The third  diagram from  the top in  Figure 6.1
illustrates a 'widely  used data reduction process that
uses  data acquired  by CRMS.  Each  point  on the
ordinate is the normalized sum of all the ion abundance
data in a single mass spectrum and, each point on the
abscissa  represents  the  spectrum  number  or   a
corresponding unit of time. This plot is referred to as a
total ion current profile (TICP) which is defined as  a
normalized plot  of the sum of the  ion  abundance
measurements in each member of a series of  mass
spectra as a function of the serially indexed spectrum
number.  This same plot  is often referred to  as  a
reconstructed gas  chromatogram  (RGC),  but this
nomenclature is not  preferred as it does not accurately
define a TICP.  An RGC could just as well be the
output of a flame ionization detector,  as redrawn by a
draftsman. An important point to recognize is that the
TICP in Figure  6.1 is the result of a computer data
reduction and not real time display. With magnetic
deflection spectrometers it  is common to continuously
monitor the unresolved ion beam and produce a total
ion current plot in real time. This plot should be similar
to the TICP in Figure 6.1, but clearly it will differ in
that it will contain contributions from ions below mass
40 and above mass 400.
   There is one  more  very  valuable data reduction
technique whose output is  most often confused with a
SICP. In this technique data acquired by CRMS, and
perhaps displayed in a TICP, is further  reduced by
plotting the change in relative abundance of one or
several ions'as a function of time. This plot is illustrated
at the bottom of Figure 6.1. The plot appears very
similar to a SICP but the data used are quite different.
The significance of  this difference is explained below,
but first a clear precise name for this output is required.
The original name suggested was mass chromatogram.
Unfortunately this is not a very descriptive name and it
may be confused with the output from a gas density gas
                                                 92

-------
        400 AMU (2-
   40 AMU (0 SEC)
                                            \
       100
  RELATIVE
    ION
ABUNDANCE
         o
                  ,
                  L
                      MASS, AMU
   REL.
 TOTAL
   ION
 ABUN
           ', .»
               V    TICP
                *',„'"'.
        SPECTRUM NUMBER (TIME)
            REL.
          PART.
            ION
                        * X
                            EICP
                      TIME (SN)
Figure 6.1 A schematic diagram of continuous repetitive measurement of spectra, a TICP, and an EICP.
                         93

-------
chromatographic  detector.  The name extracted ion
current profile (EICP) is more meaningful because the
data for the few ions used in the plot are extracted from
the larger set used to generate a TICP. The terms
limited mass  output or limited mass search are  often
used to describe this same plot. However they are less
meaningful than EICP since the nature of a  limited
mass is not clear.
  The significant difference between  a  SICP and a
EICP  is  that SIM  produces a  real  increase  in
signal/noise by  time  averaging random noise. The
EICP produces an apparent increase in sensitivity by
removing from the TICP the ion abundance data from
background,   unresolved  components,  and   other
irrelevant   ions.  The  contrast  between continuous
repetitive   measurement  of  spectra  and SIM  is
illustrated in  Figure 6.2. If the sweep of the complete
spectrum  is made  in 3.6 sec (3600 millisec) then the
data system  may  integrate the ion current at each
integer mass for 10 msec (3600 msec/360 amu).  If the
same total time is allowed for all the measurements at
each of the selected masses 99, 157, 203,  and 250 amu
in real time, then integration of signal intensity at each
may proceed for  900  msec (3600 msec/4 amu). The
longer  integration   time   during   SIM  permits
enhancement of the signal to noise ratio by averaging of
random  noise.  Therefore  there- is  a substantial
improvement in the detection limit by  SIM.'This.is in
contrast to the EICP which still uses the 10 msec data
with its inherently lower signal/noise.
  The SICP illustrated in the third diagram from the
top of Figure 6.2 was generated by summing the
abundances of all four ions  measured  during  SIM.
Clearly one could also plot the change in  abundance of
each  ion  separately  and  we make  no  distinction
between various  types of SICP plots.  However, as
illustrated in the bottom of Figure 6.2, a SICP for mass
125  would yield  no  peak  since mass  125 was not
measured during SIM.
  Figure  6.3 is a  display  of  a  TICP, SICP,  and an
EICP. The TICP was  generated from CRMS over the
40-400  amu  mass range during chromatography of
seven chlorobiphenyl isomers. Five nanograms of each
isomer was injected and an 11 msec integration time
was applied at each mass. The total time for a 40-400
amu sweep was about 5 sec. The EICP  was obtained
from  the  TICP using seven masses characteristic of
chlorobiphenyls. The SICP  is the result  of SIM  using
the same seven masses, but  an integration time of 540
msec on each. The SICP is the sum of the  data from the
seven masses. The signal to  noise contrast between the
SICP  and  EICP  is clearly  demonstrated.  This
illustration is not meant  to  imply that EICP is not.
valuable technique, but to show .the differences in the
methodoloy.
  A routine application of an EICP is shown in Figure
4.3. In this example the. TICP data has an adequate
signal/noise and the EICP was used, to effectively to
highlight  those areas of the chromatogram  having
abundant mass 149 measurements.

SELECTION OF  MASSES
  The principle potential advantages of selected ion
monitoring may be.summarized as follows:

  1.  High Selectivity
  2.  Tunable Selectivity
  3.  Qualitative Reliability
  4.  High Sensitivity
  5.  Reduced Need for Sample Purification
  6.  Quantitative Accuracy

  In practice it may not be possible to. achieve all of the
advantages simultaneously. There is a general tendency
to select the most abundant ion or ions in a spectrum in
order to optimize sensitivity.  In certain cases this may
have a significant effect on, the selectivity, and therefore
the reliability of the measurement. For, example, the
most abundant ion in. the,  electron, ionization spectrum
of 2-methylbenzothiazole is  the molecular  ion,  mass
149. However selection, of this ion for SIM could result
in measurement errors due.to the ubiquitous phthalate
esters which dis,play intense  mass 14.9, fragment ions.
The selection of a. less abundant but more selective ion
may provide adequate detection limits and preserve the
reliability of the measurement.
  One method of selecting, masses is to  employ a
computerized mass spectral search system (chapter 5)
to evaluate the selectivity of candidate masses in, the
spectrum  of the compound of interest. In this method,
the  frequency  of occurrence of several  candidate
masses  in a  large,  data base  of  mass  spectra  is
determined, and the results are used to select a mass or
masses  having an   acceptable  trade-off  between
selectivity and sensitivity.
  A method of maintaining qualitative reliability is to
monitor several ions from the same compound that
have  an  established  abundance  relationship.  The
molecular ion and its corresponding isotope containing
species  have abundance  relationships  that are  well
known. If the compound  of interest contains chlorine,
bromine,  or other  elements with several  abundant
isotopes, an excellent approach to qualitative reliability
is to monitor several ions and compare the observed
                                                 94

-------
              400AMU (2-5SEC)
CRMS
         40AMU(OSEC)
                             SIM
                                        /        / 250
                                       '        /  203
                                             '  157
                                            /   99AMU
REL
PART.
ION
ABUN

V
X
REL
PART.
ION
ABUN
\ SICP
X
x*
v TIME(SN)
SICP
                        TIME (SN)
Figure 6.2 A schematic diagram of continuous repetitive measurement of spectra, SIM, and two SICP's
                              95

-------
                   CHLOROBJPHENYLS
0    10   ZO   30
     9aECTRLM NU"BER
         SO  'GO  70   80   90  100  110  120  130  HO  ISO 160  170  183 130  200  210 220
Figure 6.3 A TICP, SICP, and EICP from the chromatography of seven chlorobiphenyl isomers.

-------
 and expected  abundance ratios. Figure 6.4 shows  a
 number  of  calculated  chlorine/bromine   isotope
 distribution patterns normalized to the most abundant
 ion of the group. Within these patterns are numerous
 possibilities for comparisons of ratios. For compounds
 that do not contain readily measurable isotopic species,
 it ,is recommended that several  ions of experimentally
 determined relative abundance be monitored and their
• abundance ratios compared.
   If several ions  from  the  same  compound are
monitored, selectivity may be improved significantly by
a  data  reduction program that plots  a sum  of the
abundances of the selected ions at a spectrum number if
and only if all the selected ions are above a defined
threshold. The advantages of this  technique will be
more generally  recognized  as  programs. with this
capability are developed.
   The reduction  in sample preparation as a result of
SIM will depend on the nature  of the sample. In the
environmental  field,  air and relatively clean  water
samples offer the best possibility for elimination of all
extract  purification. For fatty  tissue,  sediment, and
sewage samples some reduction  in extract purification
is usually possible.
   One  difficulty  with  SIM   is  the  simultaneous
measurement  of two  or more  components  with
significantly different concentrations.  Selection of a
long integration time will enhance the signal/noise of
less abundant ions, but abundant ions will saturate the
detection  system. Alternatively, a short integration
time may avoid saturation of the detector but preclude
clear observation of the less abundant ions. A partial
solution to this problem is a method for the dynamic
selection of integration time as a  function of signal
strength (IFSS), as described in chapter 2.

SIM USING THE CONTROL PROGRAM
   In  section 2.7, the control program options were
described and the possibility of using this program for
selected ion monitoring  was  implied. However all
previous examples of the application of this program in
this  manual   demonstrated   the  program   with
conventional wide mass  range  data acquisition. In
place of entering a single mass range in response to the
MASS RANGE? prompt, the user may enter up to  8
individual masses separated by semicolons or a mixture
of up to 8 individual masses and mass ranges separated
by semicolons. The masses and/or mass ranges must be
entered in ascending, non-overlapping order and they
 need not be contiguous.
   If more  than one mass or mass  range is entered,  a
 corresponding  number  of  integration  times and
samples/amu must be entered separated by.semicolons.
These have the same meaning as described in section
2.7. The following sample dialogue illustrates selected
ion monitoring with the control mode using eight ions
characteristic of chlorinated biphenyl compounds;

   SELECT MODE:    CONT
   CALIBRATE?:  N_
   TITLE:   SIM FOR  PCB'S USING  8  IONS
   CALIBRATION FILE  NAME:    CAL
   FILE  NAME:   SIM
   MASS  RANGE:    190;224;260;294;330:
      362;394;426
   INTEGRATION TIME:   400;400;400;400:
      400;400;400;400
   SAMPLES/AMU:
   THRESHOLD:
   RT  ON  CRT?:
   RT  GC  ATTEN:    5
                        1; 1; 1; 1; 1; 1; 1; 1
                              H
    MS RANGE  SETTING?:   _
    MAX  RUN  TIME:   30
    DELAY  BETWEEN SCANS  (SECS.)?


   The ions used in this example  are  selective for
 chlorobiphenyl  compounds  with different  levels  of
 chlorination and do not give optimum sensitivity. The
.choice of 400 milliseconds as an integration time is an
 example of the  use of SIM to signal average random
 noise for a significant improvement in the signal to
 noise ratio. It also embodies the idea of choosing an
 integration time such that a complete cycle through the
 selected masses requires about the same overall time as
 a conventional data acquisition from 33-450 amu using
 CRMS. Under these conditions the spectrum number
 index displayed on the SICP will be approximately the
 same as the spectrum  number index displayed on the
 TICP.
   The advantages of  using  the  control  program for
 SIM include the ability to output the results using the
 standard output programs described in section 4.1 and
 MSSOUT described in section 4.2. Output as a TICP
 will produce a  SICP  which displays the sum of the
 abundances of the selected ions as a function of time.
 Individual  masses may be displayed using the EICP
 capability.  Another significant advantage is the IFSS
 capability as defined in section 2.7. Sample dialogue for
 SIM with IFSS is as follows:

    SELECT  MODE:    IFSS
    CALIBRATE?:  N_
    TITLE:    SIM FOR  PCB'S USING 8  IONS
                                                97

-------
           8.
vo
CO
B.

o



O-
                CIBr3   CI2Br3  CI3Br3   CI4Br3   CIBr4   CI2Br4
                                                                       Br,
Br3
Br,
                                                                                                              _s
                                                                                                              .W
                                                                                                              .8

                                                                                                              IP
                                                                                                              _B
                                  Figure 6.4 Some calculated chlorine-bromine isotope distribution patterns.

-------
   CALIBRATION  FILE NAME:  CAL
   FILE  NAME:   SIM
   MASS  RANGE:   190;224;260;294;330;
      362;394;426
   SAMPLES/AMU:   1;1;1;1;1;1;1;1
   SCALE  FACTOR:   K)
   TITLE:   SIM  TEST
   MAX RPT  COUNT:   j_6
  , BASE INTEGRATION  TIME:   j_
   RPT  COUNT  BEFORE CHECKING  LOWER
      THRESHOLD:    4_
   LOWER  THRESHOLD:  2_
   UPPER THRESHOLD:   2_
   RT ON  CRT?:
   RT GC  ATTEN:    5_
   MS  RANGE SETTING?:   H_
   MAX RUN TIME:   30
   DELAY  BETWEEN SCANS   (SECS.)?

For each mass or mass range selected a sample/amu
value must be entered. A single entry is acceptable for
the  base integration  time  and   the  other  IFSS
parameters. However one may enter a string of values
separated by semicolons for IFSS parameters and these
will be applied to the masses or mass ranges in the usual
way.

SPECIALIZED SIM  PROGRAMS
  There are several data acquisition and data output
programs  that  are  specifically  for  selected  ion
monitoring, and cannot be used for more generalized
types of data acquisition or data output. However these
programs have several advantages that may favor their
use  in certain  applications. The  data acquisition
programs have a digital integration feature which may
be used to improve signal/noise with very low level
signals. The  data  display  programs  have  a digital
smoothing feature to further improve signal/noise, and
the plotter display program allows the user to exactly
overlay selected ion current profiles of two or more
ions.
  The data acquisition programs are named CRTSIM
and PLTSIM. The programs are the same except that
CRTSIM creates the real time display  on  the CRT
screen, and PLTSIM creates the real time display on
the plotter. The dialogue for the two programs is the
same:

   SELECT MODE:   CRTSIM  OR  PLTSIM
   CALIBRATION FILE:    CAL7
   MASS(ES):   198,442,443
  -'INTEGRA. TIME:   50
   NO.  POINTS:   J_0
   RUN TIME:   30
   DATA  FILE NAME:
   DATA
     DFTPP
  In response to the MASS(ES): prompt the user may
enter from one to eight masses separated by commas. It
is possible to enter fractional masses such as 440.3. An
integration time from 1 to 4096 msec may be specified
for each mass entered, or a single time may be entered
which applies to  all the masses.  If the number of
integration times entered is less than the number of
masses, the last time entered will be used for the masses
which had no integration time specified. The number of
points is the number of times each mass is integrated
before a combined value is stored in the data file. The
run time is entered in minutes, and the scale factor is an
integer amplification factor for the real time  data
display. During data acquisition three commands are
acceptable with the following results:

CTRL/L terminates data acquisition, saves the data
  file, and returns to the monitor

£ terminates data acquisition, deletes the data file, and
  returns to the monitor

Fn (where n=an integer) changes the scale factor in
  real time, but does not affect the data already stored
  in the data file.

  The data files created  by the two data acquisition
programs are compatible with the two data  output
programs named CRSMPT and PTSMPT.  The data
files are  not compatible with  the output  programs
described in chapter 4 (however see CONVRT below).
Furthermore the CRSMPT and PTSMPT  programs
cannot output data generated by the control program
in either the control or IFSS modes. Sample dialogue
for CRSMPT is as follows;
   SELECT  MODE:
   FILE?   PFTBA
   FIRST MASS?
 CRSMPT
219
   SECOND  MASS?   502
   INDIVIDUAL NORMALIZATION?   Y_

  This program plots digitally smoothed selected ion
current profiles for two ions on the CRT. Only the first
200 data points are displayed for each ion. By pressing
any key,  the next 200 data points are displayed and this
procedure may be continued  until all data points have
                                             99

-------
been plotted. The program then returns to the FIRST
MASS? prompt. A Y_ response to the INDIVIDUAL
NORMALIZATION? prompt will cause each profile
to be normalized on the largest peak in each individual
profile. Any response  except Y_ will  cause all  the
profiles to be normalized on the largest peak in  the
entire data file.
  Digital smoothing causes actual plotting to begin at
the seventh data  point  and terminate six data points
before the last  point in the file. The smoothing may
result in the most intense peak for any ion being a  few
percent  too  large or small.  A  CTRL/L  command
returns-the user to the system monitor.
  The dialogue for the plotter program PTSMPT is as
follows:

   SELECT  MODE:   PTSMPT
   FILE?   PFTBA
   MASS?   219
   SMOOTH?   Y_
   INDIVIDUAL NORMALIZATION:   Y_
   OVERLAY?   Y_
   MASS?   502
   SMOOTH?   Y_
   INDIVIDUAL NORMALIZATION?  Y_
   OVERLAY?   N_
   MASS?   503
   SMOOTH?   Y_
   INDIVIDUAL NORMALIZATION?  N_
   OVERLAY?

  This program is somewhat different in that only  one
profile is  plotted  and then the user has an option of
either overlaying one or more profiles (perhaps using a
different color ink) or plotting the additional profiles
separately. The actual plotting  begins immediately
after  the  INDIVIDUAL   NORMALIZATION?
prompt which has the same meaning as in CRSMPT.
Also, the user has the option of omitting the digital
smoothing routine.  Again the  CTRL/L  command
returns the user to the monitor.
  There  is a program  that converts  CRTSIM  or
PLTSIM data files into files compatible with the output
programs described in chapter 4. The program is called
CONVRT and is run as follows:


  SELECT MODE:  CONVRT
  FROM UNIT:  O_
  INPUT FILE: TEST
  TO UNIT:  O_
  OUTPUT FILE:  CTEST
SIMEXC (12K and EAE required) and CRTSM2 (an
overlay)  expand  the capabilities of  CRTSIM by
allowing  an  unlimited number of mass sets to be
monitored one at a time, with each set containing up to
8  masses. The mass  sets entered, ^together  with
integration times are withdrawn sequentially, so that
users should enter these on the basis of retention time.
The user only  calls SIMEXC, never CRTSM2, and
SIMEXC controls all data flow and program control.
After  calling SIMEXC, the dialogue is exactly like
CRTSIM, except after all information is input for a set
of masses the question MORE INPUT? is  asked, to
which the user should respond Y if additional sets of
masses are to be entered. Any other response is treated
as a negative one, resulting in the printing of DATA on
the screen. One pitfall must be pointed out so that it can
be avoided. A different file name is required for data
storage for  each  set of masses.  SIMEXC  tests for
duplicate file names among catalogued  files, but does
not test for duplicate entries within its  own  dialogue.
This can result in a halt in the  program overlay
CRTSM2 if duplicate names are entered in SIMEXC.
SIMEXC-CRTSM2 detects the presence of a saturated
peak during data  acquisition.  At the  time a file is
catalogued,  it  prints SATURATED PEAK on the
screen  and halts, forcing action by the operator. To
continue  the  run, the CONTINUE  switch on the
computer must be  depressed. Mass  sets  may be
changed in two ways. First, after the run time  for a
given  set has expired, the next set is automatically
withdrawn  and  data  acquisition  begins  for  that
following cataloging of the previous file. Secondly,
striking CTRL-L during data acquisition automatically
halts acquisition on the current set, cataloging of the
file, and initialization of the next set.
6.2   QUANTITATIVE MEASURE-
MENTS

  There are several important advantages gained by
making  concentration  measurements  by  GC/MS.
Quantitation  with  selected ion  monitoring  is  widely
used because it offers the ultimate in compound tunable
selectivity, high sensitivity, precision and accuracy.
Indeed the ultimate in quantitative accuracy is possible
with  SIM and a stable  isotope  labeled internal
concentration  calibration standard that is the same
compound as  the  measured analyte. Because of the
high selectivity  of SIM, some reduction  in  sample
purification is  usually possible and  this adds cost-
                                              100

-------
effectiveness to the approach. The use of data obtained
by continuous repetitive measurement of spectra for
quantitative  analysis  will  clearly  result   in  less
sensitivity and pernaps lower accuracy and precision..
However, the same tunable selectivity  is  available
through  the  extracted  ion  current  profile,  and
significant cost effectiveness  is possible by use of the
same data  file  for both qualitative and  quantitative
analyses.
  There  are a very large number of procedures for
quantitative measurements with GC/MS. These range
from standard manual measurements of peak heights
or areas in the plotter display of the real time total ion
chromatogram,  to off line computations using digital
data printed with the Adoption of the standard output
program. Measurement of peak heights is particularly
effective  with  data  acquired by  the  SIM programs
CRTSIM or PLTSIM and displayed with the program
PTSMPT.
  Manual methods  are not discussed further in this.
section  since  these are  generally understood  by
individuals experienced in chromatography work, and
many authors have written on the subject. The purpose
of this section is to define a number of terms used in
quantitative   analysis,   to   present  some   general
guidelines for use of concentration standards, and to
document  the  application  of three  programs  to
quantitative  measurements. The programs are named
QNTSET/QNTATE. QUANX (where x = a revision
number,)  and  the  IN option  of MSSOUT.  The
MSSOUT software is described fully in section 4.2, and
the user  should consult this first to  understand the
basic operations of MSSOUT.
  The following terms are defined to assure that all
readers understand the same  meanings  for several
important terms.
  External    Concentration   Calibration   Standard
(ECCS):  A known amount of a pure compound that is
measured separately (externally)  from identically  the
same compound in  a sample or sample extract. The
measured detector response from the external standard
is used to calibrate the concentration measurement of
the same compound in the sample.
  Internal   Concentration   Calibration    Standard
(ICCS): A known amount of a  pure compound that is
added to a sample or sample extract, but is not one of
the compounds found in the sample. This compound
may be labeled with an isotope or  isotopes, but usually
is not because it is not found  in the sample. The
measured detector response'from the internal standard
is used to calibrate the concentration measurements for
other compounds in the sample or sample  extract that
 are   chemically   different.   This  calibration   is
 accomplished   through   independently   measured
 response factors (RF) for each compound using the
 equation below. The independent measurements of
 RFs are made with an  external standard that also
 contains the internal standard.
                          Area (X)
            RF =
Amount (X)

  Area (S)
                        Amount (S)
   where:    Area (X) =
            Amount (X) =
            Area (S) =
            Amount (S) =
    the  peak area  of the
    compound in consistent
    units.

    the   quantity  of  the
    compound  injected  in
    consistent units.

    the  peak area  of the
    internal  standard  in
    consistent units.
    the quantity of internal
    standard   injected  in
    consistent units.
  Isotope Labeled Internal Concentration Calibration
Standard (ILICCS):  A  known amount  of  a pure
compound that contains a known  proportion of an
isotope or isotopes in an amount different from the
naturally occurring amount. This compound is added
to a sample or sample extract that contains the same
compound with a natural distribution of isotopes. This
is a  special  case  of  the internal  concentration
calibration standard which does not require the use of
response factors. The presence of isotopes does not
generally change a compound's response in a detector,
or its extraction efficiency but the labeled and natural
material   may   be  distinguished  with  a  mass
spectrometer because of the differences in the masses of
labeled ions. However, if the labeled internal standard
were used to quantify chemically different compounds
in the same sample, response factors would be required
as described above.
  Spike{s): A known amount of a pure compound that
is added to the  original environmental  sample, and is
the  same  compound  as  found  in  the   original
                                                 101

-------
environmental sample. The measured percent recovery
of the spike is taken as a measure of the accuracy of the
total analytical method with the sample matrix, and
when there is no change in volume due to the spike, it is
calculated with the equation;

            P=  100(0-X)/T

where:  P = the percent recovery of the spike.
        O = the measured value of the concentration
            of the analyte in the sample after the spike
            is added.
        X =the measured value of the concentration
            of the analyte in the sample before the the
            spike is added.
        T = the amount of the spike added expressed
            in terms  of its  concentration  in the
            sample.

  Surrogate Spike (SS): A known amount of a pure
compound that is added to the original environmental
sample, but  is not one of the compounds found in the
sample.  The  measured percent  recovery  of the
surrogate spike  is   taken  to  be indicative  of the
recoveries of the compounds in the sample when actual
spikes of each are not used.
  Laboratory  Control  Standard or Check Standard
(LCS): a sample that is prepared in the laboratory by
dissolving a known amount of a pure  compound in a
known amount of water or an organic solvent. The final
concentration  calculated from the known quantities is
the true value of  the standard. The measured percent
recovery of the laboratory control  standard is taken as
a measure of  the accuracy of the analytical method
independent of various sample matrices.
  Duplicate (D):  Two aliquots made in the laboratory
of the same environmental  sample.  Each  aliquot is
treated  exactly the same throughout the  laboratory
analytical method. The difference in the values of the
duplicates is taken as a measure of the precision of the
method.
  Regardless   of  whether  internal  or  external
concentration  calibration standards are employed, the
standards and unknowns should be run under identical
conditions in so far as this is possible. Therefore, GC
temperature programming, integration times, and mass
range in control runs or integration times, repeat count,
and number of masses monitored  in SIM runs should
be identical to cancel as much error  as possible. No
delay between scans should be utilized since it has been
found that better results are obtained on broad peaks as
opposed  to rather sharp ones. Also good  results are
often  far  more  dependent  on the  stability of the
instrument sensitivity, baseline zero, and  calibration
than on the exactness of the quantitation program.
  In general the use of an 'isotope  labeled internal
concentration calibration standard for each compound
measured   will   be  precluded   because  of  the
unavailability and/or high cost of such standards. Also
the use of isotope labeled standards in this way requires
fairly   complex   computations   to   account  for
background  ions. These standards are best applied  in
specific situations for measurements  of one  or a few
compounds.
  For  either  internal  or  external  concentration
calibration standards  it is  necessary to measure the
standards  and, for internal  standards, revalidate the
response factors on every day that measurements are
made of unknowns in samples. This is because some
GC/MS system drift is inevitable and good  results.
depend on system stability as much as any other factor.
The daily measurement of an external standard and/or
revalidation of  response  factors does  not  increase
workload  since this may be accomplished  concurrent
with a required  daily  measurement  of an  external
standard to assure proper chromatographic resolution
and system performance.
  It should also be established that concentrations are
a linear function of peak areas f6r  each compound
measured, and the assumption of linearity should not
extend beyond the established working  range. For
sample extracts it is strongly recommended that the
internal   concentration  calibration  standard   or
standards be placed in the concentrated sample extract
no more than a few minutes before the measurement.
The purpose of this  is to protect the concentration
standard from losses due to evaporation, adsorption, or
chemical reaction. An exception to this is the inert gas
purge  and  trap  procedure,  where  the   internal
concentration calibration standard must be  added  to
the original  sample.  With  this procedure  the total
method,  including the  purge  efficiency, is calibrated
with internal or external standards. However the time
constraint is still present, and the purging should begin
immediately after adding the internal standard. With
isotope labeled  internal   concentration  calibration
standards, the standard is often added to the original
sample to calibrate  the total method, including the
extraction  efficiency. However  this  is a dangerous
procedure since the relatively long contact time with
the raw sample may result in losses of the calibrant, and
cause serious errors in the measurement. This type of
calibration should be attempted only if it is established
that the  internal  standard is stable in  the sample
                                                 102

-------
 matrix.   For  internal   concentration  calibration
 standards  that  are  different  compounds  than  the
 measured analyte, additional errors may be produced if
 the standard is added to  the original sample. These
 would be caused if the extraction efficiencies of the
 standard and measured analytes were different, but the
 response factors were determined without taking into
 account this difference.
   Individual   recovery    efficiencies   should   be
 determined with spikes into representative  samples.
 Measured values of unknowns should not be corrected
 with the recovery data from a single spike, but the spike
 recovery data should be  reported to the end user.
 Laboratory control standards are used to  measure
 recoveries independent of  sample matrices. Surrogate
 spikes may be used as a rough method check.
   The final concentration  of the unknown in ug/1  is
 computed with the equation below when an internal
 concentration calibration standard is employed. Areas
 and RF are as defined above.
     C(X) =
                Area (X) *   Amount(S)
                Area(S)   *     RF *  V
   where:  C(X) = the concentration of the unknown
                 in   the   original   sample
                 micrograms per liter
in
          Amount(S).= the   quantity   of  internal
                 standard added to the concentrated
                 extract in micrograms
          V =   the volume of the original sample in
                 liters.
  One  of  the  disadvantages   of   an   internal
concentration calibration standard is that it must  be
added to a sample before  the general concentration
levels of the unknowns are known. Therefore  the
standard may be present in  much greater or smaller
quantity  than  the unknown. This may  give rise  to
serious  errors.  External  standardization  has  the
advantage of allowing the concentrations of standards
to  be  defined after making  an estimate of  the
concentration levels of the unknowns. Therefore the
standard/unknown .  concentration  ratios  can   be
adjusted  to  a  low value which may  lead to  more
accurate measurements.
  With these quidelines, it is  feasible to use chemically
reactive  compounds,  such  as  anthracene-d,0,  for
internal concentration  calibration. However,  better
choices are available  including many brominated and
fluorinated compounds that are not  likely to  be as
chemically reactive  and  variable  in  composition.
 Several    possibilities    exist     including    1,4-
 dibromobenzene,        a         tribromobenzene,
 pentafluorobromobenzene, or perfluorobenzene.
   Although  there  is  very'  little  published  data,
 quantitative  analysis  with  continuous  repetitive
 measurement   of spectra   and  a  single  external
 concentration calibration standard in  one laboratory
 gave an average bias of -30%  for aqueous laboratory
 control standards in the 0.1 - 10ug/l  range.
   All of the programs described are compatible with
 the use of internal concentration calibration standards.
 This is accomplished by entering the same data file
 name for the standard and the unknown. More specific
 instructions for internal standards are included in the
 description of each program. The QUANX program is
 particularly well adapted to internal standards since it
 contains an option to enter response factors.
   The programs QNTATE and QUANX quantify an
 unknown by a simple linear extrapolation from the
 peak area measurement of a single standard. Therefore
 it is desirable for the concentration of the standard to be
 reasonably close to the concentration of the unknown.
 A factor of 3 or 4 is considered reasonable but linearity
 should be established as described above. The program
 MSSOUT does not do  this  computation, but merely
 prints out the integrated peak  areas for various peaks
 from  various files.  The user must make the final
• calculation and  may include  a response factor  if
 desired.
   In the sections that describe the individual programs,
 the examples were generated from the same two data
 files and the results obtained give a direct comparison
 of the three programs.  The data files were obtained
 from 50ng and lOOng injections of DFTPP and are
 representative  of clean, well  resolved,  well shaped
 peaks. Therefore the results obtained are probably the
 best that can be expected from the programs.

 MEASUREMENTS WITH
 QNTSET/QNTATE
   This pair of programs has several advantages. They
 may be used with data files acquired under the standard
 control mode,  IFSS,  or one of  the specialized SIM
 programs CRTSIM or PLTSIM. QNTSET  is used to
 create a file of peak area data using information from a
 data  file generated by a concentration standard; and
 QNTATE is used to relate standard and unknown peak
 areas and  print out  concentrations  in nanograms. A
 disadvantage of this pair of programs  is  that they
 cannot be used in the extracted ion current mode, i.e., if
 the standards or unknowns were measured with the
 control or IFSS mode using a range of mass units, all
                                                 103

-------
the abundances measured at all the masses are used in
the  quantitation. There is  no  provision to extract
abundances at specific masses and use only these for
quantitation.  However,  if data was acquired  with
control, IFSS, CRTSIM, or PLTSIM at a specific mass
or masses, clearly only the specific mass or masses will
be used in  the quantitation.  Peak  areas  in  both
QNTSET and  QNTATE are  computed  with  the
baseline  subtracted.  The user  specifies  the spectral
limits of the peak and the program then assumes that
the  baseline connects the first spectrum before the
starting  value  and  the first  spectrum after  the
terminating value.
  Sample dialogue for QNTSET is as follows:

   SELECT  MODE:   QNTSET
   STANDARDS FILE NAME?   STAN
   SPECTRUM FILE?   REF100
   FIRST  SPECTRUM?   £8
   SECOND SPECTRUM?   75
   SPECTRUM NUMBER  OF  MAXIMUM?   1_\.
   NAME  OF  COMPOUND?  DFTPP
   NANOGRAMS  OF COMPOUND?   100
   SAME  FILE?    N_
   NEW FILE?  N_

The  STANDARDS  FILE NAME is not a data file
from the GC/MS of the standards, but a new file to be
created by QNTSET that will contain peak area data
from the standards. Therefore a new file must be named
by the user. The SPECTRUM FILE is the name of the
data file from the GC/MS of the standards. The FIRST
SPECTRUM is the  first spectrum number of a GC
peak, the SECOND SPECTRUM is the final spectrum
number of the same  GC peak,  and the SPECTRUM
NUMBER OF MAXIMUM is the spectrum number
from the same GC peak where the total ion abundance
reaches  a  maximum.  The  NAME  OF   THE
COMPOUND may  be  entered by the user and the
quantity of material expressed in nanograms injected.
This must be expressed as an integer number, e.g., if
one microgram was injected, the user should.enter
1000.
  The  SAME  FILE? prompt should be answered
positively if the user wishes to enter peak area data for
another peak from the same GC/MS data file into the
same peak area  file. The NEW  FILE? prompt should
be answered positively if the user wishes to enter peak
area data for another peak from a different GC/MS
data file into the same peak  area file. A negative
response to this last prompt causes the peak area file to
be closed and saved and the user is  returned to the
system prompt. A response of CTRL/L to any prompt
causes the program to abort, and all peak area data will
be lost.
  The QNTATE program requires the presence of the
peak area file generated by QNTSET. Sample dialogue
is as follows:

   SELECT  MODE:   QNTATE
   SPECTRUM  FILE?   REF50
   FIRST  SPECTRUM?   68
   SECOND  SPECTRUM?   IB
   SPECTRUM  NUMBER OF  MAXIMUM?  7_1
   STANDARDS  FILE NAME?  STAN
   PEAK  NUMBER:   j_
   DFTPP  54  NANOGRAMS
   SAME  FILE?   N_
   NEW FILE?    N_

The  SPECTRUM  FILE  is the GC/MS  data file
containing a peak or peaks to be quantified. The FIRST
SPECTRUM,   SECOND   SPECTRUM,   and
SPECTRUM NUMBER OF MAXUMUM prompts
have the same meaning  as described under QNTSET.
The STANDARDS FILE NAME is the peak area file
generated by QNTSET. The response to the PEAK
NUMBER prompt is the number of the entry in the
peak area  file which is to be  used  to quantify the
unknown peak. For example, if the user has created a
peak area  file containing three  entries such as the
following,  and wants to quantify a DDD peak, the
reply to PEAK NUMBER would be 2.
   1.  DDT
   2.  DDD
   3.  Malathion
 50 nanograms
100 nanograms
 80 nanograms
 The program then prints the name of the compound
 stored in the peak area file, and the computed number
 of nanograms in the  unknown.  In the example 50
 nanograms of DFTPP were taken and 54 nanograms
 found using a 100 nanogram standard.
  The  SAME  FILE  prompt should  be answered
 positively if the user wishes to quantify another peak
 from the same GC/MS data file. The NEW FILE
 prompt should be answered positively if the user wishes
 to quantify a peak from a different GC/MS data file.

MEASUREMENTS WITH QUAN3

  This program works with data files acquired under
the control or IFSS modes or with the specialized SIM
programs CRTSIM and  PLTSIM.  The program is
                                            104

-------
somewhat  simpler  to  use  than  QNTSET  and
QNTATE. Furthermore, it has the major advantage of
operating in the extracted ion current mode, i.e., if the
standards or unknowns were measured with the control
or IFSS mode using a range of mass units, QUAN3
may be used to quantify with a single mass  range
shorter than the measured mass range, or a single mass
unit, or even a single fractional mass unit. The program
works by summing for all spectra between given limits
(inclusive) the abundances of all masses within the
given  range (inclusive.) The baseline is  subtracted
during the integration and  the  user may  enter a
response  factor which is a weighting  factor for the
measurement. Sample dialogue for the program is as
follows:

   SELECT MODE:   QUAN3
   STANDARD  FILE NAME:  REF100
   SPECTRUM  NUMBERS:   68-75
   MASS  RANGE:    442
   SUBTRACT BASELINE:  Y_
   AMOUNT  =  100

   ABSOLUTE INTENSITY  =  +  124928.000

   UNKNOWN  FILE  NAME:  REF50
   SPECTRUM  NUMBERS:   68-75
   MASS  RANGE:    442
   SUBTRACT  BASELINE:  Yj_
   RESPONSE FACTOR:   j_

   ABSOLUTE INTENITY  =  + 67328.000

   AMOUNT  =  +  53.893

   UNKNOWN  FILE  NAME:

The data Hies used to illustrate this dialogue are the
same as  those used to illustrate QNTSET/QNTATE
and MSSOUT. The STANDARD FILE NAME is the
data file generated by GC/MS of the standards,  in this
case  100 nanograms  of DFTPP. Quantitation was
accomplished using only molecular ion data although
the original data acquisition was over the 33-450 amu
mass range. Using a response factor of 1, the  program
found 53.893  nanograms  of DFTPP in  the 50
nanogram injection. The unknown amount is given in
the same units as the standard amount.

MEASUREMENTS WITH MSSOUT
   The integrate (IN) command of MSSOUT may be
used to  integrate peak areas after the CM  and EM
commands are issued according to the standard rules.
The CM and EM commands are described in section
4.2. The. IN command has the same advantage  of
QUAN3 in that  quantitation may  be accomplished
using individual masses although abundance data from
the entire  mass  range  was acquired  originally.
However, MSSOUT has  no  apparent capability  to
relate standard and unknown peak areas. Therefore
this final calculation, including the response factor,
must be accomplished by the user with a calculator or
some other manual procedure. Sample dialogue for the
IN option of MSSOUT using the same two DFTPP
data files is as follows:

   SELECT MODE:  MSSOUT

   RUN NAME:  REF100
   DFTPP  100NG
   •CM
   ENTER  MASSES  TO COLLECT.  MAX =
   47  442

   "EM
   ENTER MASSES  TO PLOT
   442

   •IN
   START  =  68
     END  =  7j5

.   MASS  SUM
   442   126148

   •NA

   RUN NAME:   REF50
   DFTPP SONG
   •CM
   ENTER MASSES  TO  COLLECT.  MAX =
   47   442

   •EM
   ENTER MASSES  TO  PLOT
   442

   •IN
   START  =  68
     END  =  75
   MASS SUM
   442   68360
                                           105

-------
 As in plotting an EICP with MSSOUT, the specific
 mass to be used in the IN command must be collected
 with  the CM  command.  The START  and  END
 prompts produced by the IN  command request  the
 starting and ending spectrum numbers of the peak. In
 place of  entering the spectrum numbers from  the
 keyboard, the user  may  touch the space bar which
 displays the cross hair cursor. Moving the cursor to the
 left and right sides of the peak, and pressing the space
 bar after each, causes input of the  spectrum numbers.
 The NA command changes the name of the data file
 from the standard  file to the unknown  file.  Final
 calculation using the formula described above gave
 54.2 nanograms for the 50 nanogram injection.
 Peak areas for more than one ion across a single peak
 are  obtained as shown below; in this  case EM is
 required to define the sequence of events for the IN
 command.

    •EM
    ENTER  MASSES TO  PLOT
    198,442

    'IN
    START  =   67
      END  =79

    MASS SUM
    198   100992
    442    79752
6.3   OPEN TUBULAR COLUMN'S

  The purpose of this section is to review the current
advantages, disadvantages, and applications of open
tubular  columns in environmental monitoring with
GC/MS. It is clearly beyond the scope of this manual
to review in  depth  the  entire field of open tubular
columns which is a very active area of current research.
  The primary application for open tubular columns in
environmental monitoring  is in  the analysis of very
complex  samples such as industrial and  other waste
effluents. Another  application  is  in  the analysis of
samples  that contain  difficult  to  separate  but
environmentally significant components such as both
pqjychlorinated    biphenyls     and    chlorinated
berizodioxanes. However the user must recognize that
the use of high  performance,  high  resolution open
tubular  columns introduces trade-offs in  analytical
operations:-Greatly improved  resolution  invariably
requires  significantly more control of operational
variables. Therefore  there  is an important need  to
understand the additional costs required to produce the
improved  performance,  and  to  apply  the  high
performance  columns only in situations that truly
require  them.  For the   majority  of  problems
conventional  packed  columns currently offer the best
combination  of performance and operational conven-
ience. It  is also of importance to realize that there is a
far greater need for high resolution performance with
conventional   GC  detectors  than  with  a  mass
spectrometric detector. With conventional detectors,
the  retention index  is  the only output useful for
qualitative analyses.  Therefore it is vital to acquire
accurate and precise retention data,  and it  is  not
surprising to  find that the leading proponents of open
tubular columns are users  of conventional detectors.
With the mass spectrometric detector and the ability to
acquire several mass spectra across many GC peaks,
unresolved   components   may  often  be   clearly
recognized, identified, and even quantitated using
characteristic ions from each  component. Similarly
with selected  ion monitoring high resolution often may
be of no  real value. Nevetheless with complex mixtures
of compounds with  complex spectra,  open tubular
columns offer  outstanding  capabilities  for GC/MS
application and may be the method of choice.

OPERATING VARIABLES
  As previously indicated, it is not the intention of this
section   to  review  comprehensively   all  possible
operational variables. However  the ideas presented
seem to  represent  a consensus of the important
viewpoints of the most successful and experienced users
of open tubular columns.
  By all  means do not attempt to coat open tubular
columns  unless you have a great deal of experience or
the time  to acquire it. Commercial columns of high
quality are available from several sources, and the time
required to master and control the coating technique is
much  greater than  the  cost of the  commercially
available   columns.  Glass   columns  are  strongly
recommended because they  are chemically inert with
respect to most compounds.
  Splitless  injection  is strongly recommended.  The
method  of stream  splitting should  be abandoned
because of significantly reduced sensitivity due to the
discarding of a large  fraction of the injected sample,
and  the uncertainty of reproducing the split ratio.  A
particularly important variable with splitless injection
is the solvent.  In general a solvent should be selected so
the initial column temperature may be set at least 20C
                                                106

-------
below the boiling point of the solvent. This will permit
the operation of the so called solvent effect which acts
to concentrate  the  sample components into  very
narrow bands giving maximum peak  resolution. The
choice of solvent will usually  be constrained by the
capabilities  of the  GC to achieve ambient or sub
ambient  column  oven  temperatures.  If  the  GC
equipment in use does  not  have low temperature
capability, the potential for achieving optimum peak
resolution with  splitless injection  and  the  most
desirable solvent will be sharply reduced. For example,
methylene  chloride  is  strongly   recommended in
chapter 3 for low boiling solvent extraction. Methylene
chloride is also an excellent solvent for use with open
tubular columns. However the solvent effect would
cause an undesirable  peak broadening effect  with
methylene chloride  unless a satisfactory sub ambient
initial column temperature could be obtained. If sub
ambient temperature  capability is not available, and
high performance is needed, a possible alternative is
replacement of the low boiling solvent with a higher
boiling one. This of course may not be practical because
of solubility  or other considerations,  adds to the
complexity of the procedure, and the higher boiling
solvent could contribute undesirable background to the
sample.
  Two of the most critical factors which operate to
reduce column  lifetime  to  unacceptable levels are
repeated violations of the high .temperature limit of the
column, and the injection of samples containing water.
These conditions should  be avoided. With  splitless
injection samples of the order of 1-2 microliters may be
injected  into  the  system.  Columns  prepared  by
commercial suppliers using the newer  methods of
manufacture can withstand routine direct injections of
several  microliters  of pentane, hexane, methylene
chloride, or  carbon  disulfide  for months  without
damage.  Some  solvents,  such as  methanol, are not
recommended,  but the  manufacturer's  instructions
should be carefully noted when choosing solvents.
  The application of open tubular columns has several
implications for all data systems. Very sharp GC  peaks
require scan speeds that  are relatively fast, and if all
mass spectra  are  saved  on a computerized storage
device, sufficient space must be available to store the
potentially large number of spectra generated during a
long elution period. The PDP-8 data system has several
features that facilitate the use of open tubular columns,
but  reasonable  caution is required for this and any
other data system that was not specifically designed for
this application.
   In section 2.7 under FAST SCAN OPTION and the
following  paragraphs,   the  basic   sweep   speed
considerations of the PDP-8 data system are explained.
Insertion of the fast scan option hardware causes the
settling time for each mass set voltage to be reduced
from 2.3  msec to 1 msec, with some improvement in
scan speed.  Another method of  improving  sweep
speeds is to apply the IFSS method of data acquistion.
The following dialogue shows operating parameters
that give a 2.5 sec sweep with a 2.3 msec settling time.
Use of a 1 msec base integration time reduces the sweep
time to about 2 sec.

          SELECT MODE:  IFSS
          CALIBRATE?:   N_
          TITLE:   OPEN TUBULAR TEST
          CALIBRATION  FILE  NAME:    CAL
          FILE  NAME:   SAMPLE
          MASS  RANGE:   33-450
          SAMPLES/AMU:   J_
          MAX  RPT  COUNT:    5_
          BASE  INTEGRATION  TIME:   2_
          RPT COUNT  BEFORE  CHECKING
              LOWER THRESHOLD:    2_
          LOWER THRESHOLD:   3_
          UPPER THRESHOLD:   J_
          RT ON CRT?:   N_
          RT GC ATTEN:   5_
          MS  RANGE SETTING?:   H_
          MAX  RUN  TIME:  30
          DELAY BETWEEN  SCANS  (SECS.)?

          DATA

  Software exists to cause a data file to  begin on the
first disk drive (unit 0) of a two disk  system,  and
conclude on the second disk drive (unit 1) if insufficient
space is  available on unit  0.  The  files  required are
TWIN and  EXMOD2, and bit  11  of the  hardware
status word at location 2060 in INITSS must be a one
(see section 7.1).


6.4  CHEMICAL lONIZATION(CI)

  The purpose of this section is to review the current
advantages,  disadvantages,  and   applications   of
chemical  ionization (CI) in environmental monitoring.
It is not a purpose to review the entire field of CI, which
is very broad in scope and a very active area for current
research.                   .         -*
  The primary application  of chemical ionization in
environmental monitoring is to obtain the additional
                                               107

-------
information   necessary   to   accurately   identify
compounds that are not reliably  identified through
their  electron ionization  (El) spectra.  Most  often
additional information  will be required in  situations
where  a molecular ion is not observed in the El
spectrum, and  where ail' authentic  standard  is not
available  for  acquisition   of   other  supporting
information such as a GC retention time. The problem
of quality assurance in  compound  identification  is
discussed further in section 5.3.
  Another" important application of  CI is in selected
ion monitoring where ions formed in CI processes are
sometimes more selective and abundant  than ions  in
the El spectrum of the same compound. An  advantage
of CI is that  significant diagnostic fragment ions are
sometimes present that, are not observed  in the El ;
spectrum. This  advantage would be of value to those
engaged in the  interpretation  of spectra of unknown
compounds whose El spectra do not appear in mass
spectral libraries.  A potential future application  is
negative ion CI. Negative ions have been shown to be
generated by some compounds  in  much greater
abundance under  CI   conditions  than  under  El
conditions. However, most current instruments cannot
observe negative ions without major modifications, and
the significance of this  technique  is not  yet fully
demonstrated.
  At  the  present time  most  types  of  chemical
ionization cannot stand  alone as the primary method of
ionization in broad spectrum  organic analyses  as
defined in this manual. There are several reasons for
this situation. First many .compounds give, with the
popular reagent gases, CI spectra consisting mainly of
molecular ions and molecular ions containing a proton
or  some  hydrocarbon  fragment.  Therefore  these
spectra  are  less  useful for identifications  than the
corresponding El spectra which often contain  many
structurally   significant  fragment  ions.  Secondly,
although many  compounds show fragmentation  in
their CI spectra, CI spectral libraries have not yet been
developed.  Therefore,  interpretations  of spectra are
required, and this is a relatively slow process compared
to empirical spectrum matching.
  In  addition  to  the  problem  of the  sometimes
confusing M + 29, M+41, and other extraneous ions in
some types of CI spectra, there are problems associated
with the large number  of operational variables of CI.
These   variables   sometimes   lead  to  significantly
different  CI  spectra  "produced  from   the  same
compound by different laboratories  and instruments.
This makes interlaboratory comparisons difficult and
discourages   spectral   library^  development.   The
principal variables appear to be reagent gas, ion source
design,  ion source  temperature,  and  ion  source
pressure. Other variables^ such as certain ion source
potentials, electron energy, etc. may also be important.
One approach  to generating comparable spectra from
different instruments in different laboratories is the use
of a standard reference compound  to calibrate  the
experimental variables. Of course this should be used in
connection with temperature and. pressure gauges and
other operating  parameter  controls.  Unfortunately
some equipment  sold commercially does not include
essential controls. Also  temperature, pressure,  etc.
gauges are not always reliable indicators because these
devices  depend on. transducers  that are subject to
variable-output after aging or contamination.

METHANE  CHEMICAL IONIZATION
REFERENCE  COMPOUND
  The El GC/MS reference compound decafluorotri-
phenylphosphine (DFTPP) is particularly well suited
as a methane CI reference compound. The methane CI
spectrum  of  DFTPP displays,  in  addition  to  an
abundant  M+l   ion  at mass  443,  several major
fragment ions that are of very low abundance in the El
spectrum.  Therefore the CI spectrum is an  excellent
indicator of the  presence of correct CI conditions.
Figure 6.5 shows the dependence  of the fragmentation
pattern  of  DFTPP (methane CI) on the ion source
temperature. At temperatures below  150C the ions at
masses 423 (loss of F) and 365 (loss of a phenyl group)
are the most abundant in  the spectrum. In  the
corresponding El spectrum both of these ions are below
5% relative abundance. As the temperature of the ion
source in the CI  experiment  was  increased, thermally
induced fragmentation increases until, at temperatures
above 200C, the base peak is mass 169 and very little
M + l ion was observed.  The methane CI spectrum of
DFTPP at an ion source  temperature of 90C as
displayed  in  Figure  6.5  is proposed as a  reference
spectrum for methane CI measurements.
  The methane CI of DFTPP is much less sensitive to
ion source pressure in the range commonly used for CI
measurements.  However  overall  sensitivity  is  a
significant function of ion source pressure.


PRINCIPLES  OF CHEMICAL
IONIZATION
  In chemical ionization electrons are used to ionize a
reagent  gas   which   subsequently  ionizes  sample
molecules by chemical reactions, i.e. proton transfer,
hydride abstraction,  ion attachment,  etc.  Chemical
                                                108

-------
         Ion source 90C
                                                                   -w--'-
        ioc •
         Ion source 150C
                                                                                L JL
        IOC       150
          Ion source l80C
        iOC
.r>c
,L ,.,, Ul ...1 	 , ...

id ,lj
,„,,,„.] „
.|.,H.iti.
• J,,
1 I- M • L '
                                      4'jc
          Ion  source  23
                                ...  L  1
        100       150      20
2£0       -ICC      3^0
Figure 6.5 Chemical ionization (methane) mass spectra of DFTPP as a function of ion source temperature on the
         Finnigan model 4000 GC/MS System.
                                           109

-------
ionization is usually a much less energetic process than
electron  ionization.  Electron  ionization  generates
molecular ions with an internal energy of 500 to 1000
kcal/mole,  and in order to release such energy, these
ions may^undergo rearrangements and fragmentations.
Often the most important piece of information in  a
mass spectrum, the mass of the molecular ion, is lost,
and only'fragment ions will be apparent. In contrast,
initial ions formed by a chemical ionization process
usually  have  an  internal energy of only 50  to  100
kcal/mole, which means that CI spectra usually exhibit
intense ions in the molecular weight region. Chemical
ionization has been shown to offer sensitivity at least as
good as electron ionization, and offers the advantage of
allowing -.characterization  of the sample's  chemical
reactivity through the choice of reagent gases.  While
methane has been the most widely used reagent gas, a
number of others, singly and in combination, have been
successfully used to produce CI mass spectra. Popular
reagent gases include hydrogen, helium, argon-water,
ammonia, and nitric oxide.
  When methane  is bombarded with high  energy
electrons, primary reaction 1 takes place.
                   Figure 6.6 shows a comparison of the El and CI mass
                   spectra of octadecane. The El spectrum shows a base
                   peak  at  mass 57.  The  CI  spectrum  shows  the
                   characteristic ion at mass 253 (M-l) formed according
                   to equations 7 and 11.
                     There  are  usually  two options  for  handling  the
                   carrier gas in CI.  The carrier gas may be used  as the
                   chemical ionization reagent  gas, and no  enrichment
                   device is required. This has the advantage of improved
                   sensitivity since all of the chromatographic effluent
                   enters the  spectrometer. It has  the  disadvantage of
                   requiring the use of unusual carrier gases and perhaps
                   changing chromatographic behavior. Alternatively the
                   usual GC carrier gases may be used with an enrichment
                   device, and then the CI reagent gas introduced into the
                   GC inlet line just prior to the effluent entering the ion
                   source. With this method some sample will be lost, but
                   the customary GC carrier gases may be retained with
                   the considerable  flexibility  of  adding a  variety  of
                   reagent gases. The latter approach is necessary with
                   open tubular columns since the low flow rates used are
                   inadequate to give appropriate reagent gas ion source
                   pressures. A modern, well designed chemical ionization
CH4
CH,+
                                CH:"
                                         CH'  +  H/
If the gas is present in a confined chamber at a pressure
of approximately one mm of Hg, i.e. if it is in a CI mass
spectrometer source, the following secondary reactions
can take place (2 through 5):
                               IT
                                                                       2e
(i)
CH4+
         CH4 -^
CH,* + CH4

CH,+ + CH, -

C5H3+  +  CH<
CH3

 H2
 Hr
                                             (2)

                                             (3)

                                             (4)

                                             (5)
In a normal CI source, at a pressure of one mm, CH,*
and  CH;+ will comprise about  90%  of  the  total
ionization, while CjH^ will be about 5%. If a sample
molecule, MH, is present at a pressure  much less than
the reagent gas, it can react with the above ions in the
following manner (Reactions 6 through  11):
                       mass spectrometer should be equip'ped for all of the
                       above options, and a system for rapid switching among
                       them.

                      6.5   ACCURATE  MASS
                      MEASUREMENTS

                        The purpose of this section is to briefly review the
                      concept  of  accurate mass measurements and  their
                      application  to  compound  identification.  As  with
                      several other sections in this chapter, a comprehensive
                      review of the subject is beyond the scope of this manual,
                      and the user is directed for more information to several
                      excellent  books   referenced  in  the  Bibliography.
                      Although  only  a  small  fraction  of  the   mass
                      spectrometers   manufactured  have  accurate   mass
                      measuring capability, it is important that the user of a
                      conventional GC/MS system recognize the existence of
      CHJ + MH — CH4  +  (MH:)*   (proton transfer)        (6)

      CHS+ + MH -^CH<  +  H, + (M)T   (hydride abstraction)       (?)

      C2Hr  + MH -» (MH + C2H,r  (ion  capture)         (8)
                                                110

-------
      C2H,+ + MH -~ C2H4  +  (MH,)*   (proton  transfer)         (9)

      CjH," + MH —»(MH  +  CjH,)*  (ion capture)         (10)

      C,H5* + MH —»CjH6  +  (M)+  (hydride abstraction)      (ii)
this  capability.  For certain samples  of particular
significance  the  most important decision may be to
make arrangements for an accurate mass measurement
of one or more ions.
    Accurate  mass  measurements are  defined  as
  measurements made to 0.01 amu or better and are
  important because this degree of accuracy may allow
  an accurate determination of the atomic composition
  of the ion. This information is a powerful aid to the
  identification of unknowns, particularly in situations
  where a  matching  or similar  spectrum  is  not
  available in a mass spectral library.  For example,
  both acetone and njsutane exhibit a molecular ion at
  mass 58. However, because of the mass defect of the
  constituent atoms of each ion, i.e., the deviation of
  the actual  atomic  mass from  the nominal  integer
  number, the acetone ion has an accurate mass of
  58.0417 while the isobutane  ion  has a mass of
  58.0780. Therefore, these two ions differ in mass by
  36.3 millimass units (mmu).  Both have a positive
  mass defect, that is, their mass defects are in excess of
  the nominal integer value. Mass defects can also be
  negative, in which case the accurate mass is less than
  the integer mass  number. In  order to distinguish
  between these two adjacent mass spectral peaks, they
  must be  separated  or resolved.  One common
  definition of resolution states that  two peaks are
  resolved when the valley between them is ten percent
  of   the   overall   peak  height.  Resolution   is
  mathematically defined as M/deltaM,  where M  is
  the nominal mass of the two ions to be separated, and
  deltaM  is the difference between them. Using the
  example of mass 58 ions of acetone and njsutane, a
  resolution of 1600 would be needed to resolve them.
  The calculation is shown below.
      Resolution  =
  M
delta M
                                   58
                                 .0363
=  1600
  This is considered moderate resolution. Resolution
in excess of 8000 is considered high since that amount is
usually  necessary to resolve most mass doublets. In
order to achieve such resolution, double focusing of the
ion beam is usually necessary. A magnet focuses only
on the basis of mass, but an electrostatic analyzer, i.e. a
 set of curved plates with a voltage impressed on them,
 will focus  ions on the basis of their kinetic  energy.
 Adding  such  extra  focusing  reduces  the   overall
 quantity of ions traversing the mass spectrometer, and
 thus reduces overall sensitivity. To overcome  such a
 reduction,  the mass range is usually scanned at a slow
 rate. To minimize the effects from slow scanning and
 decreased sensitivity,  one should only use as much
 resolution  as  necessary .to perform  the required
 analysis, always keeping in mind that the accuracy of a
 mass measurement is independent of resolution as long
 as any mass doublets are separated.
  Accurate masses of ions  can be determined  by
 several methods, but the  two most common are peak
 matching and computer  acquisition. Peak  matching
 involves  a comparison of the unknown mass spectral
 peak wih a known ion peak, usually from a standard
 material  such  as  perfluorokerosene  (PFK).   By
 switching the accelerating voltage rapidly such that the
 unknown and known  ions are alternately focused on
 the collector, one can overlay or match the two ions,
 measure  the  ratio of their masses, and  therefore
 determine the mass of the unknown ion. The computer
 acquisition  method of determining accurate  masses
 depends  upon a computer finding two relatively close
 reference masses, and then interpolating between them
 to identify the exact mass of an unknown ion.
  Once  the  mass  of  an ion  has been accurately
measured, it is a relatively  easy matter to determine
what combination of atoms will add up to that number.
Several tables have been published which list accurate
masses and the corresponding composition  formulas.
 However the  most effective approach  is  to  use a
computer program  to find  all possible compositions
 whose  masses are within the probable error  of the
 accurately  measured   mass.  Listed  below   is   an
 interactive program written in the Basic language that
 finds  all possible  carbon,  hydrogen, nitrogen,  and
 oxygen compositions that fall within the range of a
 measured mass plus or minus the probable error. The
 program is constrained  further by a  user imposed
 limitation on the number of oxygens or  nitrogens it
 may use in any given formula.  This constraint  was
 inserted  to rule  out  a  large  number of formulas
 containing unreasonable numbers- of these-elements. It
                                                 111

-------
 E

 8.

 8.
     OCTfPEDTC El

   219  3D  «  SO   GO   70   88  SO  108 U8 128  130  J«  1SB  160  170  188 ISO 8BB 218 220  230  Vffi  2SB
       MX E
 B


 8_

 8.
     N-OCTHECFIC Cl
5R-
   TO  BO  so   ice  iie  120  130 i« iso ira ITS  in  ise  aao  210  220 230 z«
          E
2SB Z9B 273 2B9 29  369
                Figure 6.6  Comparison of the El and C! mass spectra of Octadecane.
                                                112

-------
is usually possible to guess at the maximum number of
oxygens and nitrogens, run the program, and if too few
or too many formulas are found, change the guess and
run it again. Also  no formulas will be found  which
contain  an extremely small  number of hydrogens
relative to the number of carbons. Atoms other than N
or O could be included in the program by modifying it,
or by substituting  their  masses for N or O  in the
program.

0002 PRINT "WHAT  IS  YOUR PRECISE  MASS,
     INTEGER AND DECIMAL PARTS?"
0004 INPUT M
0006 PRINT "WHAT  IS  YOUR ERROR,  E.G.,
      .002,?"
0008 INPUT E
0010 PRINT "WHAT IS THE MAXIMUM
     NUMBER  OF  OXYGENS TO BE
     CONSIDERED?"
0012 INPUT O
0014 PRINT "WHAT  IS  THE MAXIMUM
     NUMBER OF  NITROGENS TO BE
     CONSIDERED?"
0016 INPUT N
0018 LET A=M/12
0020 LET A=INT (A)
0022 LET L=INT (M)
0024 FOR J=0'TO  0
0026 FOR K=0 TO N
0028 LET C = A
0030 LET B=L-C*12-J*16-K»14
0032 IF B
-------
                                      CHAPTER  7
                            AUXILIARY
  A number of computer programs are used directly in
the acquisition   and   output  of  data  from  the
computerized  GC/MS  system. The use of  these
programs is described, with specific examples, in the
chapters on GC/MS operations and Quality Assurance
(Chapter 2), Data Output  (Chapter 4),  Compound
Identification (Chapter 5), and Advanced Analytical
Techniques (Chapter 6).  There is another group of
computer programs which  are used to evaluate the
performance of certain hardware components of the
data acquisition and control system. The use of these
programs is described, with specific examples, in the
chapter  on Preventive Maintenance (Chapter 8) and
Trouble Shooting (Chapter 9). Finally  there is a
miscellaneous  group of computer programs that are
useful in working with the GC/MS system, but do not
fit into any of the above categories. The purpose of this
chapter  is to  describe the application of these with
specific  examples. Some  of these  programs  operate
under the GC/MS real time  operating system, and
some  of them are best used only with an auxiliary
operating system. The choice of an auxiliary operating
system is of some importance as several systems  are
available.
  The OS-8 operating system requires  8K  of core
memory and has the advantage of being supported and
maintained by the Digital Equipment Corporation. It is
very widely used by thousands of installations across
the country, and a great deal of utility  software is
available for  use with  it. However, it would be
impossible  to  switch  all  the mass spectrometer
programs over to OS-8 because it is not  a real time
operating system.
  The X-System is an older utility system that requires
only 4K of core memory.  It was developed at Stanford
University and System Industries acquired it from the
Digital   Equipment  Corporation   Users   Society
(DECUS). As far as is known it is not supported or
maintained by anyone, a large number of bugs exist in
some of the programs on some versions, and the kind of
software usable with it is extremely limited.
  For the reasons given above the  OS-8 system  is
strongly  recommended  as  the  standard  PDP-8
auxiliary operating system. The X-system will not be
discussed in  this manual, and  any user who wishes
information   about    it  may   acquire   limited
documentation from DECUS (DECUS No. 8-64a) or
System Industries.


7.1 REAL TIME  OPERATING SYSTEM

  All the programs discussed  in this section are found
on the standard  system disk and are  used under that
operating system. Additional programs may be loaded
onto and run from the real  time operating-system.
Instructions for loading these are given in section 7.2.

TAPE INITIALIZATION PROGRAMS
  These programs are required to initialize a formatted
Dectape or industry standard 9 track magnetic tape for
use with  programs  or  data  files. The  program
NEWTAP is used to initialize  Dectapes with either
type controller (see section 9.2):
   SELECT  MODE:   NEWTAP
   MOUNT  NEW DECTAPE  ON UNIT  1
   (TC08  controller)
   MOUNT  NEW DECTAPE  ON UNIT  0
   (TD8E  controller)

   SELECT  MODE:   LIST
   TAPE?:   Y_

   NAME  BLOCK
   ENTRIES  LEFT  =174
   BLOCKS  LEFT  =  5525
   NEXT AVAILABLE  BLOCK  =
15
                                              114

-------
If you have available  an  industry standard 9-track
magnetic  tape  with  the  Datum  controller,  the
corresponding program is  NEWMAG. The Dectape
unit must be write enabled and an industry standard
magnetic tape must have a write ring inserted.
FILE COPY,  DUPLICATION,  AND
SAVE PROGRAMS
  Program and  data files may be copied from one
storage device to another (Dectape, disk, or industry
standard  magnetic tape) with  the COPY program.
Sample dialogue for disk to disk transfer is as follows:

   SELECT  MODE:    COPY
   FILE  NAME:  FILE TO BE COPIED
   RENAME?   FILE  MAY  BE  RENAMED
   FROM DISK?   Y_
   UNIT NO.:   0_
   ONTO DISK?   Y_
   UNIT NO.:  1
If a transfer to or from Dectape is desired, a Njesponse
is made to the FROM DISK? prompt or the ONTO
DISK? prompt. This causes the program to enter the
Dectape dialogue. The user must be aware that there
are two different programs named COPY. One version
copies files among disk units 0-3  and TC08 Dectape
units 0-7. The other version copies files among disk
units 0-3 and TD8E Dectape unit 0, and calls the
program MAGCPY which copies files to and from 9
track industry standard magnetic tape. If a transfer to
or from industry standard magnetic tape is desired, a M
response is made to these prompts.  The industry
standard magnetic tape must be unit 0. To copy files to
an industry standard magnetic tape, a write enable ring
must be installed on the tape reel. This function is
accomplished  with a Dectape by pressing the write
enable switch.
   A system disk or Dectape may be duplicated (copied
in its entirety) with the program CPYDSK if two disk
drives or two Dectape drives are present, which implies
a TC08 Dectape. The dialogue for this program is as
follows:
   SELECT MODE:   CPYDSK
   FROM  DISK?   Y_
   UNIT  NO.:  0_    ~
   ONTO  DISK?   Y_
   UNIT  NO.:  1
   SELECT MODE:   CPYDSK
   FROM  DISK?   N_
   TAPE  UNIT NO.:   _1_
   ONTO  DISK?   N^
   ONTO  TAPE UNIT  NO.:   0_

 Industry standard magnetic tape to industry standard
 magnetic tape duplications are not supported.
   A very versatile and useful program called SAVE is
 available to transfer parts of data files from one system
 disk or Dectape to another, or to segment data files on a
 single disk. The original intent of this program appears
 to have been to give the operator the capability of
 saving  a  collection  of  mass  spectra  of various
 compounds  using  different spectra  selected  from
 different data files. However, the program SAVE has
 broader applicability and may be used to reduce the
 size  of data files by saving, under a new file  name,
 contiguous groups of spectra that correspond  to the
 parts of a  chromatogram that contain the peaks. The
 original data file is then deleted  and many useless
 spectra of leading, interspersed, or trailing baseline are
 purged from the storage device. It is very important to
 recognize  that the SAVE program does not maintain
 the original spectrum index numbers, but renumbers
 the spectra  in  the saved file. Therefore all sense of
 retention time is lost, and if this information is desired,
 a TICP of the original data file should be plotted before
 executing  the SAVE program. In the sample dialogue
 shown below, the SAVE program is used to segment a
 data file containing  372  spectra but contains all the
 relevant information except retention time. All of the
 conventional output programs described in chapter 4
 may be used with the reduced and saved data file to plot
 TICP's, EICP's and mass spectra.

   SELECT  MODE:   SAVE
   FILE  NAME:   RW147B
   RENAME:  Y_
   NEW  FILE  NAME:   SRW147
   FROM  DISK?   Y_
   UNIT NO.: J_
   ONTO  DISK?   Y_
   UNIT NO.:   0_
   SPECTRUM  NUMBERS:
   170-200;230-290;300

  In this example the original data file was on disk unit
1 and the new data file is saved on disk unit 0. It is not
necessary to transfer a file to another  unit with the
SAVE program. If the original and saved data files
have different names, they may be located on the same
                                              115

-------
device.  A  negative response  to the ONTO DISK?
prompt causes the  program  to  enter the Dectape
dialogue,  and the new 'data  file  may be • saved on
Dectape. The SAVE program does not  support an
industry standard magnetic tape. The user should be
aware that there'are two version of SAVE;  one handles
TC08 Dectapes 0-7 and disks 0-3; the other handles
TD8E unit 0 only and disks 0-3.
  In the same dialogue a single spectrum No. 300, was
saved illustrating  this  capability.  It should be  noted
that there is no background subtract capability  in the
SAVE program and it  is the background subtracted
spectrum  that should  be saved in any collection of
individual • mass spectra. Therefore the feature  of the
plotter  output routine  (Chapter  4) that allows the
operator to save a subtracted file should be used first
and the background subtracted spectrum saved with
the SAVE program.

FILE  DELETE AND SYSTEM
TAPE DUMP  PROGRAMS
  The program DELE is a  housekeeping program
which removes programs or data files from the disk or
Dectape directory. It is essential for  a GC/MS operator
to maintain an orderly  storage system and 'delete all
unneeded files. The dialogue for the program is:

   SELECT  MODE:   DELE
   FILE NAME:   FILE TO  BE  DELETED
   TAPE?:   Y_(if  the file  is  on  tape)

            N_ (if  the file  is  on  disk)

This program operates most efficiently if files at the end
of the list are deleted first, followed by older files. The
user  should be aware that there are two versions of
DELE; one handles TC08  Dectape unit 1 and disk 0;
the other handles TD8E unit 0 and disk 0.
  An alternative program, BCDELE, is more efficient
in deleting files,  but  requires the presence  of an
extended arithmetic element (EAE) and 8K core in the
PDP-8. More information  on the EAE is included in
section  4.2 on the MSSOUT software.  BCDELE will
remove from disk unit 0 only contiguous groups of files
as in the following example:

   MOUTP4 7431
   SAMP1  7457
   SAMP2  7465
   SAMPS  7473
   SAMP4  7501
   SAMP5  7507
   SAMP6  7515
   SAMP7  7'523
   SAMPS  7531
   SAMP9  7537
   SAMP10 7545

   SELECT  MODE:   BCDELE
   •SAMP1:SAMP2
   "SAMP5:SAMP6
   "SAMP8:SAMP9
   " press  return

   MOUTP4 7431
   SAMPS  7457
   SAMP4  7465
   SAMP7  7473
   SAMP10 7501

  System dump programs cause the entire contents of
the system disk to be copied to Dectape or industry
standard magnetic tape in a form  that is suitable for
their recopying by another program onto a formatted,
but otherwise empty disk cartridge. The value of this
program is in creating a number of new disk cartridges
containing identical  system software from a  single
dump  tape.  Sample dialogue for  the  SYSDMP
program, which dumps the system to Dectape, is:

   SELECT  MODE:   SYSDMP

   7  TAPES WILL BE  NEEDED
   MOUNT  REEL  1  ON UNIT 1

The corresponding program for  industry standard
magnetic tape is called MAGDMP. The  method of
restoring dump tapes to system cartridges is discussed
under  the OS-8 operating  system. Again the user
should be aware  that there are two  versions  of
SYSDMP, one for TC08 Dectape units 0-7 and one for
the TD8E Dectape unit 0.
PROGRAM TO CHANGE MACHINE
LANGUAGE INSTRUCTIONS
  The program PATCH is an extremely dangerous one
because it allows the user to change machine language
instructions in programs as they are stored in disk files.
The program is important because certain changes are
sometimes essential. An example is the change of status
bit 10 in location 2060 of INITSS to permit use of
Ml000 to calibrate with  tris(perfluoroheptyl)triazine.
The procedure for this change is as follows:
                                             116

-------
  SELECT MODE:
  FILE:   INITSS
  (computer halts)
PATCH
Examine location 2060 and modify to assure that bit
10=1; for example,  load address  2060, examine
contents (this process bumps the address to 2061); load
address 2060,  set all switch registers  to 0 or 1, press
deposit key; set switch  register to 0200,  press load
address, clear,  and continue. The system will boot.
Bits in location 2060 have the following significance:

  bit   0=no    l=yes

  0    Second TC08 Dectape for data acquisition
  1    TC08  Dectape system
  2    Dual  drive TC08  Dectape
  3    Disk  system
  4    Dual  disk system
  5    IFSS  interface
  6    Plotter
  7    TD8E Dectape  or 9  track tape or cassette
  8    CRT
  9    Disk  monitor in core
  10    1000  amu file handling
  11   One millisecond timing board
       (fast scan option)
PROGRAM TO START THE
OS-8 SYSTEM
  The OS-8 system may be started by inserting an OS-8
disk  cartridge, and  then  entering into  the  switch
register a bootstrap loader. An alternative method is to
start  a system program, then insert an OS-8  disk
cartridge. The system program is called OSDGO and
sample dialogue is as follows:

   SELECT  MODE:   OSDGO

   PLACE  OS/8  CARTRIDGE  IN  DRIVE;
   WAIT  TIL  READY; THEN
   PRESS  CONTINUE
7.2 THE OS-8  OPERATING SYSTEM

  If a real time system disk is not up and running, it is
simpler and quicker to start OS-8 by entering tfie OS-8
bootstrap program into the switch register and running
it in the usual manner. See section 2.3 for instructions
on loading a program  into the switch register. If a
system disk is up and running, it is simpler to start OS-8
with the program described in the previous section. The
OS-8 bootstrap is as follows:
       Address                 Instruction
                                      0000
                                      0001
                                      0002
                                      0003
                                      0004
                                      0005
                                 6502
                                 0000
                                 6517
                                 6512
                                 6514
                                 5005
                             If this program runs successfully, a period will  be
                             printed on the console output terminal. If the program
                             fails to run properly the first time, it must be reloaded
                             and restarted. This short bootstrap is of the type that is
                             destroyed  as the OS-8  system loads,  but  a new
                             bootstrap is created at address 7600. Therefore after a
                             successful  start if it is necessary to halt the computer,
                             OS-8 may be restarted at address 7600.
                                An extensive discussion of the OS-8 operating system
                             and many of the programs found on the OS-8 system is
                             well beyond  the scope of this chapter or this manual.
                             Users who desire additional information should consult
                             the OS-8 handbook (Digital Equipment Corporation
                             part number DEC-S8-OSHBA-A-D, about $5.00). The
                             only  programs discussed in this  chapter  are those
                             relevant   to  GC/MS   system   operations  and
                             maintenance.

                             TAPE FORMAT PROGRAMS
                               Dectape must   be  formatted using a  program
                             analogous to the disk cartridge  format program. Two
                             programs  are available   named  DTFRMT  and
                             TDFRMT and these apply to Dectapes using the TC08
                             and TD8E controllers respectively. Mount the Dectape
                             to be formatted on the drive and select a unit number
                             with the thumbwheel switch. Turn each  tape twelve
                             turns  onto  the  take-up  reel, place  the WRITE
                             LOCK/WRITE ENABLE switch in the write enable
                             position, and the  LOCAL/REMOTE  switch in the
                             remote position. The NORMAL/WRTM switch is  on
                             the front of the TC08 tape controller behind the facing
                             panel and this must be placed in the WRTM position.
                             With the TD8E controller the corresponding switch is
                             located on the upper right corner of the M868 board on
                             the PDP-8 bus. It is necessary to slide the PDP-8
                             forward and remove the lid to see the switch.  Its
                             normal position is off and  the format position is labeled
                             WTM. Call the  format  programs  under OS-8  as
                             follows:
                                              117

-------
    .R  DTFRMT (TC08 controller)

    DTA?  j_
    DIRECT? MARK
    0201  WORDS,  2702  BLOCKS.  OK?
    (YES OR  NO)
    YES
    SET  SWITCH TO NORMAL

    .R  TDFRMT (TD8E controller)
                     2702  BLOCKS.  OK?
    UNIT?  0_
    FORMAT?  MARK
    0201  WORDS,
    (YES OR  NO)
    YES'
    SET  SWITCH TO WTM
    (after  one  pass  down the tape)
    SET  SWITCH TO NORMAL

The response to the first prompt is the Dectape unit
number selected. On a dual  drive system two unit
numbers may be entered to format two dectapes. After
one pass of the tape the message SET SWITCH TO
NORMAL is printed. Set the switch to NORMAL (or
OFF) and press the return. At the conclusion of the
formatting, the program returns to the second prompt
and additional Dectapes may be formatted. After all
the Dectapes are formatted,  halt the  program and
restart OSS at address 7600.
PROGRAM TO LIST FILES ON THE
OS-8 SYSTEM
  There are several methods of listing the names of the
files on the OS-8 system. These are described in detail
in the OS-8  handbook,  and only  one method  is
described briefly in this section. One point to remember
is that the OS-8 system disk is divided into two parts, a
system area and a default storage area. Files in both
areas may be listed with the program PIP as follows.
   .R  PIP
   •SYS:/E
   '/E
Entry of the command /E  without designating the
SYS: area causes the program to list the files in the
default storage area.
PROGRAMS TO  BUILD REAL TIME
SYSTEMS FROM TAPE DUMPS
  A real time system tape dump contains all the files
necessary to regenerate a new disk. These dumps are
prepared using the programs SYSDMP or MAGDMP
as described in section 7.1. A single dump tape may be
used to generate a number  of system  disks. Three
programs are available which  are named  TC8RES
(TC08 Dectape controller), TD8RES (TD8E Dectape
controller), and MAGRES industry standard magnetic
tape).
  The two Dectape programs operate very much the
same. In each case the first reel of the dump tape (there
may be one or more reels depending on the number of
data files included in the dump) should be mounted
before  calling the program. If  a  TC08  Dectape
controller is present, mount  the tape on unit 1;  if a
TD8E controller is present, select unit 0. The Dectape
drive should be in the write lock position  to protect the
dump tape in case of an error. The sample dialogue is as
follows:
                                                    .R  TC8RES
                                                    .R  TD8RES
                      (TC08  controller)
                      (TD8E  controller)
   MOUNT REEL  1  ON UNIT  1
   (TC08  controller)
   MOUNT REEL  1  ON UNIT  0
   (TD8E  controller)

 At this point remove the OS-8 cartridge and install a
 formatted disk cartridge. This cartridge may contain
 files and programs, but they will all be replaced by the
 real time system files. Press the return when the disk is
 up to speed. After the restoration  is  complete, the
 operator may start the new system by responding Y_to
 the START SYSTEM prompt. A negative  response
 causes the computer to halt. Another cartridge may be
 restored without  reinserting the OS-8  system. Place
 another formatted cartridge in the drive and restart the
 restore program at address 0200.
   The MAGRES program is used very much the same.
 The write ring should be removed from  the dump tape
 and the disk controller format switch on the rear of the
 disk controller should be placed in the format position.


LOADING NEW PROGRAMS ON  OS-8 OR
THE  REAL  TIME SYSTEM
  Programs may be transferred from  the OS-8 system
to the real time system and run under this system.
Similarly  programs  may be loaded  onto  the  OS-8
                                             118

-------
system from OS-8 Dectapes or paper tape. Finally new
programs may be written in PDP-8 assembly language
with  the  use  of  the  OS-8  operating  system,  and
transferred as above. All of these transfers are possible
using a variety of OS-8 options, and is well beyond the
scope of this manual to describe all of the possibilities.
The user  should  consult the  OS-8  handbook  for
detailed documentation.  This  section does present
instructions  for  a few  simple  transfers  that  are
especially valuable.

Inserting a Paper Tape Handler in OS-8. Some OS-8
system disks are not set up to handle paper tape input
from  a teletypewriter. This insertion is  required just
once  and  is accomplished by entering the  following
commands to the OS-8 monitor:

   .RUN  SYS  BUILD
   $ INSERT KS33,PTR
   $  BOOT
   SYS  BUILT
   .SAVE  SYS BUILD

Loading a Paper Tape File on OS-8.   Paper tape files
come in two basic varieties: binary and ASCII formats.
Binary tapes are usually labeled FILENAME.BN and
ASCII tapes are labeled FILENAME.PA.  One  may
load a program from a binary file onto the OS-8 system
and  run it.with the OS-8 R_ command  by  using the
following procedure:

   .R ABSLDR
   •PTR:
   (insert  paper tape  into the  reader in the
   region where only the parity bit is punched;
   start the reader)

The program will print an asterisk after reading the
tape. Stop the paper tape reader and enter a control/C.
After  the  period  is  printed,  enter  the  following
command:

   .SAVE  SYS:   FILENAME

  Either ASCII or binary paper  tape files may be
transferred to the OS-8 system area with the program
PIP. If the file is in ASCII format, use the /A option; if
binary use the /B option.

   .R PIP
   •SYS:   FILENAME.PA  <   PTR:/A
   (insert  paper  tape  into  the  reader
   In  the region  where only  the  parity
   bit is punched;  start the  reader)

 The program will print an asterisk after reading the
 tape. Stop the paper tape reader and enter a Control/C.

 Loading a Dectape File on OS-8.    An OS-8 Dectape
 file  may  be  transferred to the OS-8 disk by several
 programs. The program FOTP is convenient since it
 automatically generates the same filename on the disk
 as was present on the Dectape, and several files may be
 transferred with one command:

   .R  FOTP
   •SYS:  <  DTA1:Filename.BN

 This command would cause a binary file on Dectape
 unit 1 to be transferred to the OS-8 disk. Under the OS-
 8 system Dectape units 0 and 1 are usually reserved for
 TD8E controllers  and Dectape  units 2 and 3  are
 usually  reserved  for  TC08  controllers.  More
 information  on this program  is  found in  the OS-8
 handbook.
LOADING A  BINARY PROGRAM ON
THE REAL TIME SYSTEM  DISK
  A  binary file program on the OS-8 system may be
transferred  to the real  time  system disk with the
following procedure:

   .R ABSLDR
   •SYS:    FILENAME, LOD150/G

This  causes the file to be loaded on the real time system
to be placed in core memory; a second binary file called
LOD150 is also loaded and started by the /G option.
LOD150 is a program that has several other versions
named DLOADO and LOADIT. LOD150 produces
the dialogue:
                  FILENAME
FILE NAME:   	
P.A.:   enter page  addresses
P.A.:   e.g.,  0200-7577
P. A.:
                       (1 if  Dectape)
          0020-0157
   ONTO UNIT  NO.:   0_
   ONTO TAPE?:   N
                       (Y  if Dectape)
Before pressing the return remove the OS-8 disk and
insert the  real  time system, disk. Press return  when
ready and  the message FILE HAS BEEN LOADED
wilfbe printed at the completion of the transfer.
                                              119

-------
PROGRAM TO DUPLICATE OS-8
  The  'program CPYDSK, described in  7.1, will
duplicate the OS-8 system. On OS-8 disks it is listed as
DSKCPY, but the dialogue is the same.

PROGRAM TO START THE
REAL 'TIME SYSTEM
  The  OS-8 program MSDGO will start a GC/MS
system disk as follows:

        .R  MSDGO
        PUT S/150  CARTRIDGE  IN D'RIVE;
        WAIT  TIL READY;  THEN
        PRESS  CONTINUE
                                       120

-------
                        PREVENTIVE  MAINTENANCE
  The purpose of this chapter is to gather together in
one place preventive maintenance information for the
complete computerized GC/MS system. As in several
other   chapters,  some   detailed  information   is
specifically oriented to the Finnigan quadrupole mass
spectrometer with a data system that  uses a Digital
Equipment  Corporation   model  PDP-8  computer.
However a significant amount of information is of a
general nature and is- applicable to any computerized
GC/MS system. The sections in this chapter do not
repeat detailed instructions that are adequately covered
in manufacturers' manuals. The user is referred to these
instructions in the appropriate manual.
  Preventive maintenance is perhaps more important
with a mass spectrometer than any other instrument
system. This is because the mass spectrometer vacuum
system is turned on 24 hours per day, 7  days per week.
The failure of a relatively simple component at an
inappropriate time can result in the total destruction of
an expensive and vital part of the system and therefore
make the total system useless.
8.1 MASS  SPECTROMETER
VACUUM  SYSTEM

MECHANICAL PUMPS
  The oil in mechanical pumps must  be changed
occasionally.  The   frequency  of  oil  changes  is
determined by the number of samples introduced that
contain high concentrations of materials. Heavy use of
a  direct  inlet  system  invariably  leads  to  rapid
contamination of the pump oil. In general for a heavily
used  system, the mechanical pump  oil should  be
changed at least once a year. If the system is  used
mainly for GC/MS work with relatively clean samples,
a longer period may be acceptable. Similarly, heavy use
of a direct inlet system may require an oil change every
six  months.  The direct drive  pumps may require
attention every  six months regardless of the sample
load. Edwards No. 16 oil and Welch DUO-SEAL oil
are acceptable as pump oils. Direct  drive mechanical
vacuum pump  fluid  should be used in the Alcatel
pumps.
  Mechanical pump oils should be purged occasionally
for a  few  seconds   to  vent  impurities.  This  is
accomplished by  opening the oil reservoir cap and
allowing the pump to gulp air for a few seconds. A
mechanical pump  pressure of 0.1  Torr or higher after
purging is an indication that an oil change is required.
A clean  mechanical  pump operates at 0.01 - 0.05
Torr.  Unreliable  pressure  measurements can  result
from dirty contacts on the pressure gauge switch. Spray
contact cleaner  should be used as required  to insure
accurate pressure measurements.
  The mechanical pump oil  level  gauges should be
inspected periodically and cleaned  if necessary. Soak
the tubes  in acetone  for 30  minutes,  rinse with
methanol, and dry with a heat gun. Some pumps have
small windows instead of tubes for observing oil levels.
These cannot be cleaned. Tygon vacuum lines may be
cleaned with methanol but not with acetone which
dissolves Tygon. This should be done whenever the oil
is changed and  the lines should be inspected for soft
'spots  or wear,  particularly at the connection points.
The metal tube  connecting  the pump to the  separator
should be cleaned  at the same time since oil can collect
in the horizontal part of the fitting and contribute to a
high background.  Metal fittings may be washed with
acetone and rinsed with methanol.
   Mechanical pump  belts  should be  checked  for
correct tension  and wear at least every six months or
any time the system is shut down. A correctly adjusted
belt can be twisted 90 degrees. The rough pump belts
often give a warning signal when they start  to fray or
develop cracks prior to breaking; they usually develop a
noise and  the operator should be aware  of any change
in the sound of the pumps. If the  pump guards are
removed and left off, it is much more conducive to
                                               121

-------
regular checks on the condition of the belts.
  Special attention should be given to oil leaks if they
develop. On direct drive pumps the motor to pump
seals may be replaced if necessary.

DIFFUSION PUMPS
  Oil changes are required for diffusion pumps and the
considerations for  the frequency of changes are the
same as discussed under mechanical pumps. In general,
diffusion  pump  oils  should  remain  cleaner  than
mechanical pump oils because  the mechanical  pumps
are at the end of the vacuum system.  Diffusion pump
oil  can be used for up to three years if the pressure
remains good and the background low.  Use Santovac
No. 5 oil (a polyphenyl ether made by  Monsanto) in
the main' analyzer diffusion pumps. This oil has a very
low vapor pressure and its background contribution is
negligible. A silicone oil such as Dow-Corning 704 or
705 is used in the smaller diffusion pump attached to
the batch and direct inlet systems. This oil has a greater
resistance to oxidation and is better suited to the inlet
systems  where  oxygen pressures  may  be  high.
Whenever a diffusion pump oil is changed,  the old O-
rings should be  replaced.  Tygon tubes  should be
inspected and cleaned as described under mechanical
pumps.
  The cooling water exit!, of the diffusion  pumps
should be checked daily for proper flow. Occasionally
solid particles from corrosion plug the exit fittings from
a pump and cause it to overheat. This is  more likely to
happen in an older system. If flow is bypassing  one
pump, the cooling coils will be hot to the touch. The
cooling water lines should  be  cleaned  whenever the
pump oil is  changed. A complexing agent such as
EDTA (Versene) or other commercial cleaning agents
work well for this. If cooling water is 50 ppm calcium
hardness or higher, lines will plug within two years if
not cleaned. If the cooling water has a high sediment
content, a reliable sediment filter is recommended for
the cooling water line. This filter should be maintained
regularly to prevent reduced flow rates.

ION GAUGE,  MANIFOLD,  AND
INLET  SYSTEMS
  The Bayard-Alpert ion gauge should be outgassed
daily for  approximately  five minutes. The  ion gauge
may be wrapped in aluminum foil to  increase  the
temperature of the tube. This  helps keep  the gauge
clean and expedites outgassing.  Some Finnigan models
are fitted with an aluminum protective cover for the ion
gauge. This also serves as a reflector to increase the tube
temperature. If this shield is" not provided, it is  a good
idea to build  a plexiglass cover for the ion gauge. A
dropped tool and a suddenly shattered ion gauge tube is
a catastrophic event.
  The manifold  may be baked  out  occasionally at
200-250C  if background  is a problem. The magnet
should be  removed  and  the  ionizer  and electron
multiplier potentials should be turned off. In the event
of extreme contamination, the manifold can be cleaned
with acetone or methanol.
  If the direct inlet system is used frequently, pump oil
can backstream and settle in the solid probe area  and
contribute to background. The direct  inlet manifold
may be cleaned with a cotton swab dipped in acetone or
methanol. However care must be exercised in baking
the system since the direct inlet valves contain Viton O-
rings. These O-ring seals should be replaced at regular
intervals if the direct inlet system is used frequently.
  The metal disk filter between the GC column and the
separator must be cleaned if the gas chromatograph is
operated at high temperatures. The frit is cleaned with
methylene chloride in an ultrasonic bath.
  The  manifold,  batch  inlet,  transfer  line,   and
separator oven temperature monitoring systems must
be verified  occasionally with a reliable thermocouple
and a good external potentiometer. The contacts on the
temperature switch should be cleaned with an abrasive
paper and  spray  contact  cleaner to  insure accurate
temperature readings.
  The batch  inlet  system  septum should be  changed
each  time the system is brought  to atmospheric
pressure.
8.2  THE  MASS SPECTROMETER
ELECTRONICS

COOLING FANS AND DUST  FILTERS
   With modern solid state electronics the single most
important preventive maintenance factor is the flow of
cool air across and around  electronic components.
Clean dust filters and working fans provide these flows
and it is absolutely essential to maintain them properly.
The proper operation of cooling fans is extremely easy
to check and this should be done once a week. The time
interval for  cleaning cooling fan filters varies widely
depending on location and the air quality in the mass
spectrometer laboratory. Frequent checks of air flows
and the cleaning of filters quarterly is essential.
   The Finnigan  model 1015  has  cooling fans with
filters  on the oscilloscope,  RF/DC generator, and
power supply modules.  A  fan without a  filter  is
                                                 122

-------
contained in the oscillographic recorder. Other models
have a similar placement of fans and filters.

VACUUM TUBES
   The older Finnigan model 1015  systems  have  in
excess of 25   vacuum  tubes  and  the  3000  series
instruments have four vacuum tubes in the RF/DC
generator (one 6AL5,one 6EB8, and two 8458).There is
wide disagreement about changing vacuum tubes on a
regular basis in any preventive maintenance program.
One extreme philosophy is to never change a tube until
unstable  operations or failures  occur. At that  time
tubes are prime trouble shooting targets. It  is more
prudent  to regularly change tubes  and the schedule
suggested in this manual is intended as a guideline for
the 1015  system.

RF/DC  Generator.    Replace 5726 and 807  tubes
annually. The 5726 is a 6AL5 replacement.
Oscilloscope.   Replace the 68Q6 tube annually.
Electron   Multiplier  Power  Supply (Keithley).
Replace the 8068 and OA2 tubes annually.
Low  Voltage  Power Supply   (Dressen-Barnes).
Replace  the  6AV6,  OA2,   12AX7, 655OA,  and
5R4GYB tubes annually. Check 655OA and 5R4GYB
tubes semi-annually.
Low Voltage Power Supply (Finnigan).   Replace the
6LQ6, OA2, and 12AX7 tubes annually.
Oscillographic  Recorder  (Century).     Clean  the
optical surfaces annually. These are cleaned in place
with a cotton swab and  alcohol; wipe clean with a
tissue. Check the condition of the drive belt  and the
galvos annually.

8.3 THE GAS CHROMATOGRAPH

  The GC septum should be changed weekly. The oven
and injection block  temperature  monitoring systems
must   be  verified  occasionally  with   a   reliable
thermocouple and a good external potentiometer. The
dirt filter  should be cleaned regularly. The utilization of
drying agents,  filter traps, and an oxygen  scrubber  is
recommended for the carrier gas line, and these should
be regenerated or replaced as necessary.

8.4 THE DATA SYSTEM

SOFTWARE
   Computer programs are called software because they
can  be changed relatively easily in contrast to the
resistors,  transistors, and integrated  circuits of the
 hardware. This capability of relative ease of change of
 software can  provide  enormous  flexibility  and  the
.possibility   of continuous  system  updating  and
 improvement.  It also can create a chaos of magnetic
 tape and disk cartridges whose contents are all different
 and unknown to the user.
   The very first consideration  of software preventive
 maintenance is an organized,  systematic method of
 dating and labeling all paper tapes, magnetic tapes, and
 disk cartridges. Similarly all documentation, which
 may be received as loose update sheets, should be dated
 and stored in a single file for ease of reference. All
 operating system software should be duplicated before
 any attempt is made to run the programs. Instructions
 for  duplicating system  disk  cartridges, OS-8  disk
 cartridges, and Dectapes are given in chapter 7.  It is a
 sound policy to make several copies of important disks
 and tapes  and place  one copy of each in  a secure
 cabinet. These copies should be reserved for ultimate
 back-up that is never used for anything except making
 copies.
COOLING FANS AND  DUST FILTERS
   The considerations discussed  in section 8.2  are
equally applicable to the data system. The PDP-8 data
system contains numerous fans and filters.
   All floor standing cabinets contain a double fan and a
large filter at the rear cabinet base. The disk controller
and disk controller power supply  modules also have
fans and filters. The  PDP-8 has an internal fan and
filter. The  Dectape drive and its power  supply have
cooling fans. The Houston plotter has a cooling fan and
the Tektronix 4610 hard copy unit has a dust filter. The
System Industries digital interface  has a cooling  fan
and dust filter.  The RIB interface has no fan or filter.
LUBRICATION OF MECHANICAL
PARTS
   Because  of  the  mechanical complexity of  the
teletype, an annual service call or service contract is
recommended.  If service  is not readily available, the
motor should be oiled, the motor gears greased, the belt
checked, and  dashpot  and  dashpot cylinder cleaned
and reoiled at least annually. It is  recommended that
 the teletype be turned off at all times when not in actual
 use.
   The plotter's sliding  metal surfaces  should  be
cleaned quarterly with isopropyl alcohol. Do  not oil or
 grease any moving parts or sliding surfaces; all moving
 parts  are  permanently  lubricated  or made of low
 friction combinations.    .
   The slide mechanisms of electronic modules may be
 lubricated with machine oil or graphite as needed.
                                                123

-------
MAGNETIC STORAGE DEVICES
  The read/write heads of the magnetic tape and disk
drive units must be kept clean and free of dirt and iron
oxide.  The  tape  heads  are readily  cleaned  with
isopropyl alcohol and a soft tissue monthly. The disk
drive  heads  should  be  cleaned  semi-annually  with
isopropyl  alcohol and'  a lint free  wiper.  Further
information is contained in the disk drive manual.

DIAGNOSTIC COMPUTER PROGRAMS
  Certain    diagnostic   tests   are   utilized   for
troubleshooting, and several of these are recommended
for checking  the integrity of the system on a routine
basis. These tests are explained in detail in chapter 9.
                                              124

-------
                                       CHAPTER  9
                              TROUBLE SHOOTING
  Trouble shooting  is often  the  most difficult and
frustrating part of operating a computerized GC/MS
system. The failure of an extremely simple component
or electrical connection .may be the cause of a complete
failure  of the  system.  However,  because  of  the
complexity of the total system, it may be extremely
difficult to isolate the simple problem.
  The purpose of this chapter is to develop a systematic
method of trouble shooting a computerized GC/MS
system.  As  in  other chapters,  some  sections  are
oriented  to  Finnigan model 1015  and  3000  series
quadrupole mass spectrometers with data systems that
employ Digital Equipment Corporation model PDP-8
computers. However, the overall systematic approach
is applicable to any GC/MS system.

  The chapter is divided into five sections. The order of
the sections is  the  same as the probable order of
discovery of problems by an operator following the
recommended operational  procedures  in  chapter 2.
Within each section the problem  solving approach is
the same. The first things to look for are simple, easy to
fix  problems such as loose or  broken cables, blown
fuses, dirty PC board connections, inoperable cooling
fans, dirty dust filters, water or oil leaks, air leaks, and
uncontrolled  heaters. If  this  fails  to  locate  the
difficulty, it is necessary to isolate the problem as much
as possible to a specific module or modules. Once this is
accomplished, manufacture's service manuals should
be  consulted to  determine specific remedies.  It is
beyond the scope of the manual to catalog a large
number of specific failures and remedies. There are
simply too many variations in hardware even within
one manufacturer's product line. If the manufacture's
service manual does not provide a solution to the
problem, it is then appropriate to call a service engineer
for the specific module that is suspect.
9.1   MASS SPECTROMETER
PROBLEMS

VACUUM SYSTEM
  There are several warning signals that  should  be
checked each day on entering the mass spectrometer
laboratory. A diffusion pump over heat light will come
on if the cooling  water flow was too low or the water
insufficiently chilled to maintain the pumps at a proper
temperature. A Fenwall switch senses the temperature
at each diffusion pump and automatically shuts off the
high vacuum system  in the event of overheat.  If the
overheat  is limited to just one diffusion pump,  the
entire system  is shut down. The  remedy for this
problem is clearly to adjust the cooling water flow to a
correct level and to assure that the outlet stream is cool.
The Fenwall switch may require resetting  before the
high vacuum system is reactivated. See chapter 2 for
the pump down procedure.
  Another  warning signal is the illumination of the
pressure failure light or an indication of an abnormally
high pressure on the vacuum gauge. With the GC
column disconnected or the carrier gas diverted from
the mass spectrometer,  a pressure of 5x10"' to 10"'
Torr should be maintained. Under these conditions air
leaks  are insignificant and  the  presence  of ions  of
masses 18,28,32,  and 40 are of no consequence. If the
pressure becomes sufficiently high, about 10"1 Torr, the
high vacuum system will be shut down.  The most
common causes of vacuum system failures are a broken
belt on a mechanical pump, a break in a Tygon or glass
vacuum line, or a leaking O-ring or Teflon ferrule.
Teflon ferrules flow when hot and the metal  fittings
need to be tightened occasionally. Electrical problems
may  also  cause  high  vacuum  system failures.  A
momentary  power failure will shut down the high
                                                125

-------
 vacuum system but the mechanical pumps will restart
 when the power returns.  A  blown fuse or a  faulty
 component on the ion gauge controller board may also
 cause'a pressure failure.
   For abnormally high pressures that  do not shut
 down the vacuum system, the first remedy is to tighten
 metal fittings containing Teflon ferrules about one-half
 turn. If this fails to eliminate the leak, inert gas such as
 argon  or helium should be sprayed around fittings,
 gaskets,   tube  connections,  and   O-rings.   The
 observation of a sudden  increase in pressure is  an
 indication of a leak. If this procedure fails to detect the
 leak, turn the ionizer and electron multiplier on and set
 the first and last mass controls to display the prominent
 ions from the inert gas. Continue spraying the gas and
 look for a sudden increase in ion abundance  at the
 characteristic masses. It may be necessary to change a
 gasket or leaking O-ring.
ION GAUGE
  The glow of the Bayard-Alpert tube is an indication
of the state of the vacuum system. If the tube glows
extremely brightly, flickers intermittently, or flickers
rapidly when the solvent from a GC injection reaches
the manifold, trouble is indicated. Similarly, if the high
vacuum gauge seems insensitive to changes in pressure
as during solvent elution or the admitting of PFTBA,
the ion gauge should be serviced.
  Intermittent   flickering   may   be   caused   by
contamination  from solvents or pump oil.  Degassing
overnight may help as will the application of external
heat with a heat gun. See also chapter 8 for preventive
maintenance measures. For  other problems, the ion
gauge controller board should be examined for bad
components. If this fails to solve the problem,  replace
the vacuum tube.
 HEATERS
   Manifold,  separator,  transfer line, or batch inlet
 heaters may lose control and overheat. If this occurs,
 the respective triac controllers are suspect and should
 be tested  and  replaced.  If  higher  than  normal
 potentiometer  settings   are  required  to  maintain
 temperatures, one-half of a  clamshell heater may be
 burned out. Also, bad thermocouples and dirty meter
 switch contacts cause inaccurate temperature readings.
9.2   DATA SYSTEM PROBLEMS

  Many data system problems are caused by simple
failures that are  readily corrected.  Power cables and
signal transmission cables may work loose and all such
connections should be checked at the first sign  of
trouble. Similarly fuses, cooling fans, and filters should
be checked at once. Some modules have normal/test
switches and if a  switch was inadvertantly left in a test
position, the module may not funtion normally.
  Magnetic read/write  heads should  be cleaned  as
described  in  the preventive maintenance chapter. If
software  on  a  disk or tape  was altered  by some
malfunction or accident, programs may suddenly fail.
An alternative disk or tape of known good  quality
should be tried. If these measures fail, it is advisable to
check  all   PC boards  for  secure, tight,  and clean
connections. It is worth lifting each PC board in a unit
from its connector, removing dirt corrosion with  an
ordinary pencil eraser, and replacing each board.
  Electronic component and some mechanical failures
are best found with the use of a diagnostic computer
program. The programs in  the following sections are
for the PDP-8 data system, but comparable programs
should  be  made  available  by  each data  system
manufacturer. After isolation of the particular problem
with the diagnostic program, it is appropriate to call a
service engineer for the faulty module.
COMPUTER DIAGNOSTIC PROGRAMS
  A large number of computer diagnostic programs
are  available  for  the  PDP-8   from  the  Digital
Equipment Corporation. Each of these programs tests
the operation of one or several computer functions.
Digital  Equipment  Corporation  maintenance  and
diagnostic   programs   are   recognized   by   the
characteristic name, MAINDEC, followed by a series
of digits and letters.  Each progam is supported  by a
document that describes the use of the program and the
error messages that are printed. These documents are
required  to  gain  full  advantage of  the  diagnostic
programs.
  The  most important decisions that  must be made
when trouble  shooting  the  data  system are which
diagnostic program to run and when to call the service
engineer. In  general it is possible to save considerable-
time and expense by running diagnostics to isolate the
problem before calling a service  engineer.  Reporting
the results of the diagnostics at the time the service call
is  made sometimes allows a  quick solution  to the
                                                 126

-------
problem  by replacement of a specific printed circuit
board. In a few cases the choice of a diagnostic program
is quite clear.  For example, if a problem  repeatedly
occurs when using a magnetic tape drive and only when
using the tape, find a diagnostic for the magnetic tape
system and run it. Dectape diagnostics are discussed in
the next section of this manual. If all the more readily
recognized  device  problems   are  eliminated,  the
magnetic disk  system is the next peripheral to check
since it contains  mechanical funtions that  are more
susceptible  to  failure  than  the  purely  electronic
functions of the computer itself. The disk diagnostic
program is described in a subsequent section and  it is
not  a  standard  Digital  Equipment  Corporation
program. The  PDP-8 disk system  employs a Diablo
Corporation series 30 moving head disk drive and a
controller manufactured  by System  Industries.  The
diagnostic program was written by System Industries
and it has its own detailed documentation.
  If it is determined that the disk system is functioning
properly, it is appropriate to run computer diagnostics.
Table 9.1 contains a list of OS-8 program names for a
selection  of computer diagnostic programs, brief
operation  instructions,  and  the  satisfactory  test
indications.  This list is a small subset of the total
number  of available computer diagnostics,  but it
includes  some of the  most general  tests  that  are
recommended  for reasonably comprehensive testing of
most  basic  computer  functions.  It  is  strongly
recommended  that the GC/MS operator run several or
                             more of these programs at regular intervals in order to
                             become familiar with  their correct operation.  With
                             systems that are not used often, these programs should
                             be run regularly to exercise the system and keep it in
                             good operating condition. With  these  diagnostic
                             programs stored on the  OS-8 operating system, it is
                             easily possible to test the computer's function in  a few
                             hours. If the complete  program documentation is not
                             available,  it  is  possible  to   gain  some  valuable
                             information with the general directions that are  given
                             in Table. 9.1.  To load the programs  from the  OS-8
                             system, enter R, then  a space,  and the name of the
                             diagnostic program. The programs should be run in the
                             sequence shown in Table 9.1 since some of the  later
                             tests assume correct functioning of earlier tests. All
                             programs have the standard .starting address 0200 if it
                             is necessary to restart them. The programs are stopped
                             by   pressing  the  HALT  switch.   The  XADDR,
                             XCHKBD, and EAEXME programs  require at  least
                             8K of memory, but there are corresponding programs
                             for 4K systems. Error messages are printed in the event
                             errors  are  found,   and  the  complete program
                             documentation is  required to  interpret  the  error
                             numbers. The  error number explanations may  have
                             little meaning to  most  operators,  but  should be
                             available before calling a service engineer. After halting
                             the program, OS-8 may be restarted at address  7600.
                             However .the two memory diagnostics write over all
                             contents of memory, and the OS-8 bootstrap must be
                             reloaded to restart OS-8 in these cases.
                    Table 9.1   Computer Diagnostic Programs for the PDP-8/ E or M
    OS-8
    Program Name

    INST1
    INST2
    XADDR
    Function

CPU  instruction
tests
CPU  instruction
tests

extended memory
address tests
         Operation

halts on  loading; enter  7777
in switch register  and press
clear and continue

starts-on  loading
halts on loading; enter
starting address 0200 in
switch register, press load
address,  clear,  and continue;
halts and prints message;
enter  in  the switch  register
0001 for 8K,  0002 for 12K,
etc. and press  continue
   Satisfactory
 Test Indication

bell every few
seconds
bell every few
seconds

prints  the numeral
5 every five minutes
                                                 127

-------
              Table 9.1   Computer Diagnostic Programs for the PDP-8/E or M (continued)
 OS-8
 Program Name

 XCHKBD
    Function                  Operation

extended core        halts on  loading;  enter in
memory tests         the switch  register 0200,
                     press load  address, clear
                     and continue;  halts and prints
                     message;  enter in the switch
                     register 0001 for  8K, 0002
                     for 12K, etc. and press
                     continue
                                      Satisfactory
                                    Test  Indication

                                 prints the numeral
                                 5 every  five minutes
  ADDER
CPU  adder
circuit tests
  EAEST1
  EAEST2
  EAEXME
extended arithmetic
element instruction
tests
extended arithmetic
element instruction
tests
extended arithmetic
element - memory
tests
starts on loading; before
calling program  enter in the
switch register 0210 for 8K,
0220  for 12K, etc and then
call program

halts  on loading; enter 0200
in switch register, press load
address;  enter 1000 in switch
register;  press clear, and
continue

halts  on loading; enter 0200
in switch register, press load
address,  clear and continue;
halts; enter  0400 in switch
register and press continue

halts  on loading; enter 0200
in switch register and press
load address; enter 4000 in
switch register for 4K, 4001
for 8K,  4002 for 12K, etc;
press clear  and  continue;
after  message enter 0 for
4K,  1 for 8K, 2 for  12K, on
the terminal keyboard and
press return
prints the characters
SIMAD, SIMROT, FCT,
and RANDOM  after 35
minutes
prints the
characters KLE81
every 6 minutes
rings bell every
minute
prints the
characters KE8  EME
every 22 seconds
  For   problems  that  are  intermittent,   it   is
recommended  that one of the  memory or  general
instruction  test   programs  be   permitted  to  run
overnight. Raising the temperature of the room during
testing is a method of causing marginal components to
fail completely.
  A  complete  list  of the  MAINDEC  diagnostic
programs is found under acceptance tests in the Digital
                              Equipment  Corporation  PDP8/E,  PDP8/F  and
                              PDP8/M Processor Maintenance Manual - Volume 1.
                              Additional tests for modules not included in Volume 1
                              are found in Volume 2, Internal Bus Options or in
                              Volume 3, External Bus Options, in Volumes 2 and 3
                              the available MAINDEC programs are usually listed
                              under  Software, Acceptance Tests, Maintenance,  or
                              Trouble Shooting. Lists of maintenance and diagnostic
                                                128

-------
programs are also found in other Digital Equipment
Corporation publications. For example a list is found in
appendix A of the 1971 edition of the PDP8/E Small
Computer Handbook.

DECTAPE DIAGNOSTIC PROGRAMS
  The  Dectape diagnostics are  described separately
because problems may be readily identified with this
unit  when  failures  occur  only  during  Dectape
operations.  Also included is a relatively  simple tape
drive brake ajustment  that is a solution to several
problems.
  There  is  one  aspect  of  Dectapes  that causes
confusion. All Dectapes are not alike although they
may appear physically identical and data files  and
programs may be transferable between them. The most
common Dectape drive is the TU56 in either a single or
double tape drive configuration.  The confusion arises
because this drive is used with the relatively slow TD8E
interface board  and the relatively fast TC08 Dectape
controller. The TD8E and the TC08 are not compatible
and Dectape access programs written for one will not
work with the other. Each  controller has a different
format program and different diagnostic programs but
data  files  and  non-Dectape  programs  are  fully
compatible with each system. To  add to the confusion,
older documentation often refers to the  older TU55
tape drive and TC01 Controller which are very similar
totheTU56andtheTC08.
  The diagnostic programs for  Dectape units are listed
in the appropriate maintenance manual, e.g., the TC08
Dectape  Controller   maintenance  manual  under
Software.  If a  Dectape  is  suspected  of causing
transmission errors or  exhibits sluggish performance,
an  exerciser  program  should  be  run.  A  random
exerciser program called TC08RX is available for the
TC08 Dectape controller. A program called TD8DIA
is available for TD8E Dectapes. The use of the TD8E
test  program   requires  the  complete   program
documentation and this program  is not discussed here.
The  following information  refers  to the TC08  test
program, but TD8E users should take note of the TU56
brake drive adjustments and other comments in the last
two paragraphs of this section.
  After the program   is loaded into  memory, the
computer will halt. Mount a formatted Dectape on the
drive or drives to be tested. More than one drive may be
tested at the same time if desired. The tape or  tapes
must not contain any data or programs that are needed,
as the diagnostic destroys all information  on the tape.
Set the desired  drive number  using the thumbwheel
switch  on the Dectape drive and place the unit in write
enable and remote. Load the starting address of the
program,  0200, into the switch register and  press
LOAD ADDRESS. Set the switch register to select the
drive or drives to be tested as shown in Table 9.2.
Table 9.2.   Switch Register Selection of Dectape
             Unit in the Random Exerciser Test.
   Dectape  Unit  on
  Thumbwheel Switch

          0
           1
          2
          3
          4
          5
          6
          7
Octal Value of
Switch Register

     4000
     2000
     1000
     0400
     0200
     0100
     0040
     0020
Several drives may be tested at one time by combining
the switch register settings  for the corresponding
Dectape  drives.  Press CLEAR and  CONTINUE
switches. The program will halt again.  Place 0000 in
the switch register and press CONTINUE. To halt the
test, press the HALT switch. The complete program
documentation  is  required  to  interpret  the  error
messages and use all program options.
  Errors will be found if a read/write head is dirty, the
tape is badly worn, or the unit is overheating due to an
inoperable cooling fan. A worn tape  may be reclaimed
by reformatting. It  is recommended that the first five
feet of a worn tape be removed before reformatting.
This may eliminate the most worn section. Other errors
may  be due to  an  incorrect  tape drive motor brake
adjustment. This  undesirable situation can also be
observed by sluggish performance of the tapes on stops
and turn arounds. This problem  can wipe out existing
files on a tape or make it  impossible to store sample
data.
  To  correct this problem  on a dual transport TU 56
tape unit, there are two adjustments on  printed circuit
board M302 which is the dual delay motor voltage. The
bottom trimpot  adjusts the right transport and the top
trimpot adjusts the left transport. Whether the pots
should  be turned  clockwise  or  counterclockwise
depends on which motor is most out of adjustment. The
adjustment is most easily made if one person operates
the transport manually while the second  adjusts the
trimpot. Adjust voltages until stops and starts are crisp.
                                                129

-------
DISK  DIAGNOSTIC PROGRAM
  This  extremely  important  program  has  several
functions and it is described separately for the reasons
given in the section on computer diagnostics. The
program's  functions include  formatting new disk
cartridges,  testing the surface of the magnetic disk,
testing the functions of the disk drive, and testing the
disk  controller.  If any  errors are found, an  error
number is printed. The error numbers and explanations
of the malfunctions are listed in the System Industries
disk maintenance manual. These error  numbers and
explanations have little or no meaning to the average
operator,  but  it is important to run the diagnostic
program and  have  the  error message  ready  before
calling a service engineer.
  The disk diagnostic program  is named DDIAGN
and it may be loaded into the computer from an OS-8
disk as indicated below. In the following discussion, the
SWR refers to the switch register on the PDP-8 front
panel. The first prompt printed by the disk diagnostic
program is:
ENTER DISK TEST INFORMATION
After this and all other  prompts, the computer halts
(the run light goes out). The procedure followed is the
same for  all prompts: Enter a twelve  digit  binary
number into the SWR  and  press the  CONTINUE
switch on the  PDP-8 front panel. In the instructions
that follow the twelve digit binary number is expressed
as a four  digit octal  number. The sample dialogue is
recommended  for most  purposes, but some optional
responses  to several prompts  are discussed  below.
Insert and start the OS-8 disk.

   .R DDIAGN

   ENTER  DISK TEST INFORMATION

  Before responding to this prompt, remove the OS-8
disk and insert a new or test cartridge. Enter 300x into
the SWR where x = the disk unit number ( 0 for a
single disk drive, 1 for the second unit of a two drive
system, etc.) and press continue. If more than one drive
is present, more than one test can be started at the same
time as follows:

   ENTER  DISK TEST INFORMATION
     (enter  3000  and press continue)
   ENTER  DISK TEST INFORMATION
     (enter  3001. and press continue)
   ENTER  DISK TEST INFORMATION
      (enter  0000  and  press continue)
  ENTER UNIT ADDRESS
     (enter 0050  and  press continuej
  ENTER FORMAT LOOP CONTROL
     (enter 0000  and  press continue)
  ENABLE  FORMAT  SWITCH,  ERROR  AND
  TYPE OUT  CONTROLS
     (enter 0002  and  press continue)

  UNIT 0
  UNIT 1

  DISABLE  FORMAT  SWITCH

  UNIT 0
  UNIT 1
  UNIT 0
  UNIT 1
  UNIT 0
  UNIT 1
  UNIT 0

  Each time the unit number is printed a new cycle of
tests begins for that drive. Note that  any and  all
programs or data stored  on  the test  disk will  be
destroyed. One of the prompts printed is:
ENTER FORMAT LOOP CONTROL
If a 0000 is entered, the disk will be formatted once at
the beginning  of  the  first pass of the  tests. On
subsequent passes, the  disk will not be reformatted.
Any non-zero entry will cause the program to reformat
the disk before every  pass of the tests. A  single
formatting is adequate and 0000 should be entered. The
last prompt printed is:
ENABLE FORMAT SWITCH, ERROR AND
TYPEOUT CONTROLS
Several  responses  to  this  prompt  are possible  to
perform a variety of functions. Place the format enable
switch on the rear  panel of the  disk  controller in the
format  position, enter one of the following codes into
the switch register, and press CONTINUE. Code 0002
is recommended for most situations  and is especially
suited to over-night testing for intermittent errors.

0002.   Causes  the disk to be formatted  and runs
       diagnostics. If an  error occurs,  an  error
       message is printed and the program continues
       with the next test. At the conclusion of all tests
       the sequence of tests is repeated. The program
       runs about one hour with the format switch
       enabled and completes  one  cycle of tests in
       another hour with the format switch disabled.
                                              130

-------
 0000.   Causes  the  disk  to  be formatted and  runs
        diagnostics.  If an  error occurs, an  error
        message is  printed  and  the  program  halts.
        Running time is the same as 0002.
 0003.   This entry is used following a halt with 0000.
        Enter 0003  and press  CONTINUE and the
        diagnostic will continuously  repeat the test
        that caused the error.
. 0040.   This  is  a fast format option. The  disk is
        formatted in  about two minutes and  the
        program halts without  requesting the user to
        disable  the  format  switch.  This option is
        acceptable  in  emergency situations when a
        formatted disk is required immediately. It is
        not recommended in general as all disks should
        be thoroughly tested after formatting.
   To stop the test, press the HALT switch. The OS-8
 disk may be reinserted and the OS-8 system restarted at
 address 7600. Alternatively a  new  test disk may  be
 inserted and DDIAGN restarted at 0200.


9.3   SYSTEM ZERO  ADJUSTMENT
PROBLEMS

   If the system  zero adjustment described in section
2.4 cannot be achieved, or if the signal is  excessively
noisy, there are  several likely causes. Instabilities are
sometimes caused by overheating of electronics due to a
cooling fan failure or  a dirty  filter  that  blocks the
passage of an adequate amount of cool air. Vacuum
tubes in the RF  generator (1015 and 3000 series) and
power supplies should be checked or  replaced. The
preamplifier  may be faulty  and  zeroing  should be
attempted on all  of the preamp ranges. If possible swap
the  preamp for  a spare  and attempt to zero.  Some
preamplifiers  have   a   coarse  zero   screwdriver
adjustment. Set  the  fine zero pot to  mid-range and
adjust the coarse zero using the ZERO program.  Some
preamplifiers  have  a  screwdriver  hum  balance
adjustment. Consult the  manufacturers  manual  for
details of this adjustment.
   The other potential major problem area  is the mass
spectrometer interface. The proper functioning of the
interface is  best tested using the  interface diagnostic
program described  in  section 9.4. If the diagnostic
program indicates the interface is functioning properly,
yet it is not possible  to zero, the fault  may be a noisy
electron multiplier. Before replacing or reconditioning
the multiplier, refer the subsequent sectionsjrtcluding
LOSS of IONS.  As with many other problems,  failure
to zero properly also suggests a dirty ion source.
 9.4   MASS SCALE CALIBRATION
 PROBLEMS
  The  automatic  mass  scale  calibration program
should complete a successful calibration of the mass
range 20-650 amu at a PFTBA pressure of about 1 X
10"* Torr in about 40 seconds or less. If a longer time is
required, or the system will not calibrate, moderate to
serious  trouble is  indicated. There  are a  very large
number of potential causes for mass scale  calibration
problems and it is simply not possible to describe each
in detail. Therefore a general approach to the diagnosis
of the problem is given here and the manufacturer's
maintenance manual should be consulted for  specific
details.
  After a calibration failure, the system tune-up should
be checked following the procedure in the section 2.2.
During  this tune-up check it may not  be  possible to
observe any signals from PFTBA ions. This is classified
as a major  problem  and one should check the next
section titled LOSS OF IONS. If ions are observed the
calibration problem should be attacked in the following
manner.
  If the calibration diagnostic program described in
section 2.5 was routinely applied,  past records should
be examined for signs of day to day drift. The vacuum
.tubes in the RF generator, particularly the 6AL5 or its
5726 replacement option, are prime suspects in cases of
calibration drift. Problems with these tubes may not be
observed with a vacuum tube checker. A new tube
should  be  inserted  and  its  stability  determined
experimentally. If drift continues,  try several new tubes
and  select the most  stable one  of  the group. The
Finnigan 3000 series  instruments have four vacuum
tubes including a 6AL5 in the RF generator.
  A  drift  in  an  RF tuning capacitor   can cause
calibration problems. This  capacitor compresses or
expands the mass range observed during a single sweep
of the rod voltages. There is one tuning capacitor for
each range on the 1015 and a single tuning capacitor for
the 3000 series mass  range. The proper  adjustment of
the tuning capacitors is described in the  first mass, last
mass  calibration procedures in   the Finnigan  1015
operator's manual. A similar procedure exists for the
3000 series spectrometers.
  A problem  in the  mass spectrometer interface can
cause calibration failures.  The proper functioning of
the interface  diagnostic  program is described in  a
subsequent section of this chapter. Another valuable
program for diagnosing instability problems is the peak
profile diagnostic program  which is  also described in
                                                  131

-------
this chapter. Again a dirty ion source may be the cause
of a calibration failure.

INTERFACE-SPECTROMETER
GAIN ALIGNMENT
  Calibration problems with the Finnigan 1015, 3000,
and 4000 mass  spectrometers  could  be caused by
incorrect alignment with the' PDP-8  interface. The
System Industries (SI) interface outputs 0-10V which
corresponds  to 1-750 amu or 1-1000 amu. The amu
range is  handled by  the  driver  file.  The  standard
calibration driver file  that is built  into the calibration
program  is set up for 43 digital to analog converter
(DAC) units per  amu (43 x 750 = 32,250). The 15 bit
DAC has a range of 0-32,768 DAC units. The FC1000
and Ml000 driver files are set up for 32 DAC units per
amu (32x1000 = 32,000).
  The problem that develops is that the 0-10V range of
the SI interfaces  have to be matched with the 0-10V
range  of the Finnigan  mass  spectrometers.  The
calibration files work as follows. The driver files consist
of a list  of DAC ranges that  correspond  to the
calibration masses. The DAC ranges cover +3 amu
around ,the  calibration  mass.  For example in the
standard driver file mass  100 has a DAC range of 43 x
100 =  4300 DAC units + (3 x 43 =) 129 DACunits;
in otherwords mass  100 + 3amu corresponds to the
range of  4171-4429  DAC units. The SI  interface
outputs i the 4171-4429  DAC  unit  range  which
corresponds   to  1.2728-1.3516V   sent   to   the
spectrometer. The most important idea is that the mass
100 peak must be in this range and all other calibration
masses must be in their ranges. On the Finnigan 1015
R-73 on PCS is an offset adjustment to make mass 100
fall in the correct  range.
  If the Finnigan ramp voltage (PC-5 on 1015.) is set up
so that mass 100 is passed through the quadrupole at
1.262 volts, calibration will fail because the SI interface
and calibration program look for mass 100 between
1.2728-1.3516V.  R-73 (on the 1015) shifts the ramp up
or  down without changing  the  slope.  Thus the
spectrometer may be set up so mass 100 will be passed
through the quadrupole in the correct voltage range.
  A problem was created when Finnigan eliminated
the R-73 adjustment in the 3000 and 4000 series mass
spectrometers. R-73 may be thought of as a very coarse
resolution, coarser than R-74. To compensate there is a
"MASSOUT GAIN" pot  in the  new RIB  interface.
This allows  matching of the interface ramp  with the
instrument ramp. To make this  adjustment run the
peak profile diagnostic program MSSCAN,  which  is
described in section 9.4 and introduce PFTBA. Center
 mass 100, or 219, or etc. on the screen with about a 500
 DAC unit range. Then calculate where the peak should
 be in DAC  units as follows: 1-1000 amu instruments:
 mass  x  32  DAC  units  =  	;  1-750  amu
 instruments : mass x 43 DAC units =	
   Using MSSCAN determine where your peak is, and
 adjust R-73 on the 1015 or the MASSOUT GAIN on
 th RIB until the mass is centered about- where it should
 be. Remember there is a +3 amu (+  96 or 129 DAC
 unit) range  in the calibration file.. The user should be
 aware that the MASSOUT GAIN pot may be mounted
 over the  letters SSOUT; therefore it looks like MA
 GAIN on the label in the RIB interface.
 LOSS OF IONS
   If no ions are observed during an attempted tune-up,
 the first check is the 0-200V ramps. On the model 1015
 these are observed at TP5 and TP6 on PCS (see Figure
 2.1).  On the 3000 series instruments the ramps are
 observed at TP4 and TP5 on RF/DC  control card.  If
 the ramp is present the loss of ions is probably due to a
 failure in the ion source, electron multiplier, or one  of
 their  power supplies.  A  non-existent  ion source
 collector (trap) current suggests a broken filament: a
 lower than normal trap current suggests a sagging
 filament that must be replaced. If source potentials are
 missing or  low in value, the problem may be on the ion
 source  control PC board.  The electron multiplier
 power supply output should  be checked. Some power
 supplies have vacuum tubes that should be replaced
 regularly.  See  chapter  8  for   a   recommended
 replacement schedule.
   Another good trouble shooting approach is to  shut
 down all power to the mass spectrometer and check all
 electrical   leads  to the ionizer  and  multiplier for
 continuity  or no continuity as appropriate. Consult the
 manufacturers  maintenance  manual  for  specific
 directions.  If shorts are indicated it will be necessary to
 remove the ionizer and multiplier assemblies from the
 manifold and check for shorts and correct assembly.
   If the -200V ramp was not present, check TP20 on
 PCS of the 1015 for the ramp. Check TP1 and TP6 on
 the   RF/DC  control  board  of  the 3000 series
 instruments. If this is present the problem is probably a
 component on the PCS board or the RF/DC control
-board.. If the ramp is not present  at TP20 (TP1/6)
 check the  output of the +1200V and +200V power
 supply. Vacuum tubes should be replaced if necessary.
 Also there  are vacuum tubes in the RF generator of the
 1015 and  3000 series  instruments.  See chapter 8 for
 more details on these.
                                               132

-------
INTERFACE DIAGNOSTIC PROGRAMS
  Problems in the mass spectrometer interface are
often revealed during an attempt to calibrate or during
operations with an apparently acceptable calibration.
The  interface diagnostic  program  is an  excellent
approach to finding interface problems. It is a good idea
to run the interface  diagnostic program at regular
intervals as a preventive maintenance routine when the
GC/MS system is not in use.  This will  exercise the
electronic components of the interface and familiarize
the operator with the method of running the program.
  There  are  two versions of the  program  that
correspond to two models of the interface. Both models
perform identical functions, but the later  model takes
advantage of advances in solid state technology and is a
smaller overall package that may be somewhat remote
from the data system. This interface is often called the
remote in base (RIB) interface. The interface diagnostic
programs are identified with hardware as follows:
    OK
    OK
    OK
    OK

    (after completion of  test,  place manual
    test switch  in  off  position)
    .R NEW150
    DO YOU
    DO YOU
       NO
    DO YOU
    DO YOU
       YES
    DO YOU
       TEST?
    DO YOU
WANT  STATIC TEST?   NO
WANT  TRANSMISSION  TEST?

WANT  ANALOG  ADJUST?  NO
WANT  MASS SPEC  TEST?

WANT  REMOTE  SEQUENCING
 YES
WANT  IFSS  TEST?   YES
             Interface Model Number     Interface Diagnostic Program
   1152-7016 (separate analog
   and digital drawers)

   5000-7073 (RIB or remote interface)
MSTEST

NEW 150
The interface diagnostic programs are stored on the
OS-8 system and may be loaded into the computer as
described  in  the  previous  section  on  computer
diagnostic  programs.  Both programs have a starting
address of 0200. For general trouble shooting surveys,
operator familiarization, and interface exercising, the
followiing  responses  to  the  prompts are strongly
recommended:
  R MSTEST

   DO YOU  WANT STATIC  TEST?  NO
   DO YOU  WANT MASS  SPEC TEST?
      YES
   DO YOU  WANT OFFSET ADJUST
      PROMPT?   NO
   DO YOU  WANT TRACKING ADJUST
      PROMPT?    NO
   DO YOU  WANT  IFSS TEST?   YES
   DO YOU  WANT  PLOTTER  TEST?  NO
   PUT LOGIC  DRAWER  IN TEST  MODE
      (place manual test switch on back  of
      logic drawer in ON  position)
      (press continue)
    DO  YOU WANT  PLOTTER TEST?   NO

    OK
    OK
    OK

  The prompts that are answered in the negative refer
to specialized tests and adjustment aids that are best
used after trouble is indicated in the Mass Spec and/or
IFSS  tests. It is extremely important to  run the IFSS
test  since this mode  is  used  during  mass scale
calibration.  A  major  advantage of  running  the
suggested tests only is that after completion of each
test,  the  program prints  OK,   then  repeats  the
diagnostic.  Therefore it  is  possible, and  strongly
recommended, to run the test continuously over a long
period of time, including overnight. This will enhance
considerably  the  chances  of finding crucial  but
intermittent  errors. Be certain to  enter 0000  in the
switch register to insure the program will continue even
if an  error is detected. To stop the program press the
HALT'Switch.
   If an error is detected,  the program  will prinr an
error message. As usual most of the error messages will
                                             133

-------
probably have little meaning to the operator even after
consulting the appropriate System  Industries mass
spectrometer interface maintenance manual. However
having the error  numbers available when a  service
engineer is called could save considerable time and
expense. However, there are several errors that  are
related  to  relatively  simple   potentiometer  (pot)
adjustments that should be attempted by the operator.
Errors 3 and 4 occur if the analog to digital converter
(ADC)  zero is not in agreement  with the digital to
analog  converter (DAC) zero.  The  zero may  be
adjusted by a trial and error procedure until errors 3
and 4 are corrected. Trial and error consists of turning
the zero adjust potentiometer on the right side of the
ADC (Phoenix analog drawer) one-half turn in one or
the other direction while the diagnostic is running and
observing the errors messages. A YES response to the
MSTEST  prompt  DO  YOU   WANT  OFFSET
ADJUST  PROMPT? was intended  to facilitate this
adjustment by continuously rerunning the diagnostic
that leads to errors 3 and 4. However there is little point
in continuously rerunning the one specific diagnostic,
since the entire MASS Spec and IFSS  tests only require
several minutes. A much quicker method of making
this adjustment in  the two drawer interface is to run the
program ADZERO.
ADZERO  Program.   This  program is a modfied
version  of the preamplifier zero program,  ZERO,
described in section 2.4. It is loaded as follows:

   SELECT  MODE:   ADZERO
   MANUAL?    N_
   AUTOMATIC?   N_
   ZERO  ADJUST?  Y_

  This program causes the DAC to output zero volts.
This output is then connected (through software) to the
input of the ADC  with  the ADC's output  being
displayed in the accumulator of the PDP-8. If no lights
are visible,  turn the zero adjust pot CCW until a light
appears in  the accumulator. Rotate the pot CW until
the light goes out completely (stops flickering). At this
point rotate the pot another 1/2 turn  CW. This will
insure operation in the safe region  of the operating
range. If a  light or lights are initially displayed in the
accumulator,  eliminate   them  using  the  above
procedure.
  The small  remote interface may require the same
adjustment after an occurrence of the same errors. The
correct pot is labeled integrator offset on the top board
of the interface.  The  ADZERO  program  is fully
compatible with the remote interface and may be used
to facilitate the adjustment.
  Another error that may be corrected by a simple
adjustment on either interface is error 6. This is caused
by an improper gain pot adjustment. This pot on  the
remote interface is labelled the unity gain pot. A trial
and error process should be used during the Mass Spec
and IFSS tests to make this adjustment.
  With the excepton of the Houston plotter test,  the
other  prompts of  MSTEST  and NEW 150 were
intended   to  facilitate  certain   adjustments  and
measurements. These are not as generally useful to the
system operator,  but  may  be used if desired. The
System Industries interface  manuals describe these
tests. The plotter  test may be the most useful  of the
remaining tests,  but a failure does not  necessarily
indicate  trouble  in the  plotter itself. The  plotter
interface  is contained in  the digital bay of the two
drawer interface,  and is on a separate  PDP-8 board
when the remote interface is present.
PEAK PROFILE DIAGNOSTIC PROGRAM
  This  program  is a  valuable  tool  to  investigate
problems  of system  zero adjustment,  calibration,
quality  assurance, or  intermittent  instability.  For
example, if the mass scale calibration is failing because
of a weak 6AL5 (5726) detector diode tube in either a
model  1015 or 3000 series, a clear observation of this
instability can be made with the program.
  The  program gives  the  operator the capability  of
observing repeatedly an entire mass spectrum, a small
portion of a mass spectrum, or a single peak  in a
spectrum at  the data system CRT. These same basic
capabilities are available at  the  mass spectrometer
console oscilloscope, but the CRT is easier to use for
diagnostics because of its storage capability.  With the
CRT and the program, the operator has a very much
better opportunity to  examine clearly the  details  of
peak shape and its stability.  Also since this program
uses the data system, if a spectrum is clearly observed
on the oscilloscope, but not  with the program, the
problem is immediately isolated to the data system
itself or the data system-mass spectrometer interface.
  The  name of the diagnostic program is MSSCAN
and the program is available on the OS-8 and the real
time operating  systems. Before the program is run,
perfluorotributylamine  (PFTBA) must be introduced
into the mass spectrometer. Turn on the ionizer and the
electron multiplier. Under OS-8 the program is called
                                                134

-------
with the command R MSSCAN. Under the real time
system the dialogue is as follows:

   SELECT MODE:   MSSCAN
   CONST,  SCAN OR?

A user entry of a question mark causes the program to
print a list of MSSCAN commands that is reproduced
and explained in detail in this section. A user entry of
the command C (not CONST) causes the program to
enter a non-scanning mode. In this mode additional
commands  are available as  shown below.  One may
cause the program, by the various keyboard entries to
set and hold various rod voltages. These voltages  are
expressed in terms of the corresponding  digital to
analog converter (DAC) units. Also one may print  the
corresponding analog to digital converter (ADC) data
after integration of the electron multiplier signal  for
various time periods. This constant mode of operation
has some utility for trouble shooting, but is is generally
less useful than the other mode of operation.
  The most useful application of the program is  the
scanning mode that is entered by the command S (not
SCAN). Subsequent prompts are shown in the sample
dialogue, which should cause mass 219 (from PFTBA
for example) to be displayed near the center of the CRT
screen with a 1-750 amu mass range instrument.

   CONST,  SCAN OR  ?S_
   STARTING DAC  9150
   RANGE   500
   INCREMENT   3_

If the peak is not prominently displayed, any of  the
single character commands under the scan mode may
be entered to modify the program scanning conditions.
The screen  will be  paged automatically,  and  the
command printed. Enter the new value and scanning
will be resumed. The first command to try is the M
command which multiplies the peak intensity by  the
value entered. One may also change the starting DAC,
range and increment slightly  since the  exact values
required to  display mass 219 in the center of the CRT
will vary somewhat from instrument to instrument. In
general one amu corresponds to about 43 DAC units on
a 750 amu mass range intrument and 32 DAC units on
a 1000 amu mass range instrument. A starting DAC of
about 2000, a range of 20000, and an increment of 10
will display the spectrum from about 40-700 amu. To
exit from the program or to recover from an error after
paging the  screen manually, enter the E command.
This will return the program to the first  prompt. Under
OS-8 enter a CONTROL/C and under the real time
system a CONTROL/L  to return  to the system
monitor.

MSSCAN Variables

  C    —  constant .yoltage (DAC)

  S    —  set DAC and read  ADC
  U    —  increment DAC and read ADC
  D    —  decrement DAC and read ADC
  P    —  print  current DAC
  T    —  set time register (max. 4095)
  R    —  repeat integration at current register
           settings
  E    —  back  to start

  S    —  scan on 4010 display

  S    —  change  starting DAC
  R    —  change  range
  I    —  change  increment
  T    —  change  time  register
  P    —  print  parameters
  O    —  one step each time  space  bar is pressed
  C    —  clear  one step - full speed
  M   —  multiply intensities
  D    —  divide intensities
  E    —  back  to start

Integration time will be equal to the value in the time
register, in milliseconds. This register is set to 1  at the
start of the program.
  The most important observations to make from the
mass  219 ion of PFTBA  are peak shape, resolution
from  the mass  220  ion,  position  stability,  and
abundance stability. The peak should be symmetrical
without excessive noise or a bra shaped top. Its stability
in position and abundance should be very constant with
little or no movement. It may be necessary to examine a
broader mass range, e.g.,  50 amu, to adequately judge
position stability. A noisy or unstable condition is a
clear sign of trouble and a frequent cause of mass scale
calibration failures.
  Some of the conditions observed with MSSCAN are
listed below together with  a few possible causes  of the
problem.
Position Instability:  A bad 6AL5 (5726)  tube in the
model 1015 or 3000 series RF generator; also three 807
RF tubes are contained in  the 1015 generator and two
8458 tubes and a 6EB8 are contained in the generator
of the 3000 series  instruments. The power supply for
                                                135

-------
the model 1015 has vacuum tubes that can cause severe
instability.
Abundance Instability: A bad preamp; bad tubes in the
electron multiplier power  supply or a  bad electron
multiplier; a sagging filament  or unstable ion source
potentials.
Noisy Peaks: Improper grounding which produces 60
Hz noise in a ground loop; bad power supply tubes or
unstable source voltages. The mass set voltage at TP18
on PC-4 of theFinnigan 1015 should be stable to better
than a millivolt regardless of the source, i.e., internal or
datasystem.
9.5   QUALITY CONTROL PROBLEMS

  Quality control problems are defined as problems
that are recognized only when examining  the overall
peak size caused by a known amount of DFTPP or by
examining the mass spectrum of DFTPP. At this stage
the  system  should be  tuned and  the  mass  scale
calibrated. The quality control  performance evaluation
is conducted as described in section 2.6. Examine the
output from the evaluation and consult the appropriate
section that follows.


POOR OVERALL SENSITIVITY
  Poor overall sensitivity may be caused by a variety of
factors arid is roughly judged  by the size of the peak
obtained from 20ng of DFTPP. The operator must be
familiar with the capabilities of the instrument when it
is in good working order. The following list of possible
causes of poor overall sensitivity is ordered from the
first things to check to some actions of  nearly last
resort:

  1.   For  packed  columns connected  to a  jet
       separator,  the mass  spectrometer  pressure
       should be  about !0~5 mm  of Hg. A lower
       pressure  suggests f. plugged separator; check
       the temperatures of the separator and transfer
       line oven for a possible heater failure. A cool
       oven will reduce sensitivity by condensation of
       the test compound.
  2.   The dilute standard solution of DFTPP should
       be remade if it .is noj fresh. Old solutions of all
       dilute  standards  suffer concentration losses
       due to adsorption of  trace quantities on the
       walls of containers or oxidation by  trace
       quantities of oxygen or other oxidizing agents.
  3.    If the shape of the peak is broader than normal
       or exhibits tailing, the column may be poorly
       conditioned or suffering from old age. Also the
       metal fritted filter or separator may be partly
       plugged.
  4.    The tune-up should  be checked for over
       resolved peaks in any or all  mass ranges. Use
       only  as  much  resolution as  needed  as
       illustrated in section 2.2. Note  carefully the
       proper    functioning   of   all   tune-up
       potentiometers. Recalibrate and  repeat the
       DFTPP performance evaluatibn.
  5.    The peak profile diagnostic program should be
       run with particular attention to noisy peaks
       and possible ground loop problems and other
       instabilities as discussed in section 9.4.
  6.    Clean  the ion  source, check for a sagging
       filament, and clean the rods according to the
       manufacturers's procedures.
  7.    If all else fails, recondition electron multiplier
       using  procedures  from  the  manufacturer's
       manual.

POOR HIGH MASS  SENSITIVITY
  This condition  is  revealed by   unacceptable ion
abundance  at  masses 365 and  442 in  the  DFTPP
spectrum. Poor high mass sensitivity  is most frequently
caused by dirty rods or over resolution of masses 219
and 220 and 502 and 503 as in manufacturer's tune-up
procedure.  If  a  tune-up according to the  EPA
procedure in section 2.2 is not successful, clean the ion
source and rods.

INCORRECT MASS
ASSIGNMENTS  OR POOR
RESOLUTION OF  ADJACENT IONS
  This condition is revealed by an examination of ion
abundances in  the DFTPP spectrum at masses  68-70,
197-199 and  441-443. If the mass scale calibrated
correctly, but incorrect mass assignments appear in the
DFTPP spectrum, a slow drift of the  mass scale
calibration is indicated.  This is most  frequently caused
by a weak 6AL5 tube in the RF generator (1015 and
3000 series instruments). Refer to section 9.4 and locate
the cause of the drift.
  If the mass assignments are correct but the ions a:
masses 68, 197, or 441 are too abundant, front end lift
off is  indicated. This condition is ofter caused by an
incorrect alignment of the source magnet or incorrect
adjustment of the ion source potentials (lens, extractor,
ion volume, etc.).
                                                136

-------
  Insufficient  resolution  will  cause  excessive  ion
abundance at masses 68, 197 or 441 as well as at masses
70,  199, or 443. Acceptable ion abundance at masses
68,  197,  or 441,  but inaccurate 69/70,  198/199, or
442/443  ratios may be  caused  by  a  system  zero
adjustment that sets the threshold too high and clips off
real ion abundance data.
HIGH BACKGROUND
  High background is  defined as  the  presence  of
extraneous ions in an otherwise acceptable spectrum of
20ng of DFTPP.  A more or less continuous band of
ions across a broad mass range suggests a poor system
zero adjustment, in which the threshold is set too low.
Repeat the zero adjustment  in section 2.4 and make
sure one accumulator light is either just flashing or just
off. If a steady zero cannot be achieved, see section 9.3
for possible causes.
  Chemical  background can result  from a  serious
accumulation of pump oil in the manifold, solid probe,
or transfer line flange area. This could be caused by a
lengthy power failure and the resultant diffusion of oil
vapor or by a failure in the overheat protect circuit that
allowed a diffusion pump to overheat.  Removal of
pump  oil is accomplished  by  a thorough  internal
cleaning   of the  system with solvents  and  the
manufacturer should be consulted for assistance.
  Chemical background can also result from liquid
phase bleeding  from the GC column, contaminated
carrier gas, a contaminated injection system, or a small
leak in the  GC column.  As appropriate the column
should   be   operated   at   a   lower   temperature,
reconditioned or changed. If this fails to eliminate the
background change or regenerate the  carrier gas filter
drying trap or change the carrier gas cylinder. Injector
contamination  can be  eliminated  by  a  thorough
cleaning of the injection port with a series of solvents.
However, if on-column injection is employed, cleaning
the injection port will not help. If the head of the
column  is  highly  contaminated,  discard the  old
packing, thoroughly clean the column, and repack.
                                                 137

-------
                                    CHAPTER  10
                        SELECTED  BIBLIOGRAPHY
  It is not the purpose of this chapter to present a
comprehensive listing of all the known literature of
GC/MS 'and  related   articles  on  enviriomental
applications. Rather a limited number of books, review
articles, and original journal articles are cited which are
considered important contributions and keys to further
information about these subjects.
 10.1   GENERAL REFERENCE BOOKS

  S. Safe and O. Hutzinger, "Mass Spectrometry of
  Pesticides and Pollutants", CRC Press, Cleveland,
  OH, 1973.

  R.E. Gould, "Fate of Organic Pesticides in the
  Aquatic Environment", Advances  in  Chemistry
  Series  No.  Ill,  American   Chemical  Society,
  Washington, D.C., 1972.

  F.W. McLafferty, "Interpretation of Mass Spectra",
  2nd ed., W.A. Benjamin, Inc., New York, N.Y.,
  1973.

  G.W.A. Milne,  "Mass Spectrometry, Techniques
  and Applications",  Interscience Publishers,  John
  Wiley and Sons, Inc., New York, N.Y., 1971.

  H. Budzikiewicz, C. Djerassi, and D.H. Williams,
  "Mass  Spectrometry  of Organic  Compounds",
  Holden-Day Inc., San Francisco, CA, 1967.

  J.  Roboz,  "Introduction to Mass  Sriectrometry,
  Instrumentation  and   Techniques",  Interscience
  Publishers, John Wiley  and Sons, Inc., New York,
  N.Y. 1968.
W. McFadden, "Techniques of Combined  Gas
Chromatography/Mass Spectrometry-Applications
in Organic Analysis", Interscience Publishers, John
Wiley and Sons, Inc., New York, N.Y.,1973.

G.R. Waller,  "Biochemical Applications  of Mass
Spectrometry", Interscience Publishers, John Wiley
and Sons, Inc., New York, N.Y., 1971.

Q.N. Porter and J. Baldas, "Mass Spectrometry of
Heterocyclic Compounds", Interscience Publishers,
John Wiley and Sons, Inc., New York, N.Y., 1971.

F.W. McLafferty, "Mass Spectrometry of Organic
Ions", Academic Press, Inc., New York, N.Y., 1963.

J.H. Beynon,  R.A.  SauhderSi and A.E. Williams,
"The Mass Spectra of Organic Molecules", Elsevier
Publishing Co., New York, N.Y., 1968.

K.   Biemann,  "Mass   Spectrometry-Organic
Chemical Applications", McGraw-Hill Book Co.,
Inc., New York, N.Y., 1962.

M.C. Hamming and N.G. Foster, "Interpretation of
Mass Spectra  of Organic Compounds", Academic
Press, New York; N.Y., 1972.

H.M.  McNair  and E.J.   Bonelli,  "Basic  Gas
Chromatography",  Varian  Aerograph,  Walnut
Creek, CA, 1969.

W.R;  Supina, "The  Packed  Column   in  Gas
Chromatography", Supelco, Inc., Bellefonte, PA,
1974.

T.R. Lynn, "Guide to Stationary Phases  for Gas
Chromatography", Analabs, Inc., North Haven, CT,
1975!
                                             138

-------
10.2   PRINTED MASS SPECTRA
COLLECTIONS

  A. Cornu and R. Massot, "Compilation of Mass
  Spectral Data",  2nd ed.,  Heyden and Son, Ltd.,
  London, England, 1975.

  "Eight Peak Index of Mass Spectra", Vol. 1-3, Mass
  Spectrometry  Data  Centre,  Atomic  Weapons
  Research Establishment,  AJdermaston, England,
  1975

  E.  Stenhagen,   S.  Abrahamssen,  and   F.W.
  McLafferty, "Registry  of Mass Spectral Data",
  Interscience Publishers, John Wiley and Sons, Inc,
  New York, NY, 1974.

  "EPA/NIH Mass Spectral Data Base", National
  Bureau of Standards, National Standard Reference
  Data Service publication No. 63 (1978).
10.3   SELECTED REVIEW AND
PRIMARY JOURNAL ARTICLES

MASS SPECTROMETRY

  Burlingame, A.L.,  Shackleton, C.H.L.,  Howe, I.,
  Chizhov, O.S., Anal. Chem. Annual Reviews, 50,
  346R(1978).

  Burlingame, A.L., Kimble, B.J., and Derrick, P.J.,
  Anal. Chem. Annual Reviews, 48,368R (1976).

  Burlingame, A.L.,  Cox,  R.E., and Derrick, P.J.,
  Anal. Chem. Annual Reviews, 46,248R (1974).

  Burlingame, A.L. and Johanson, G.A., Anal. Chem.
  Annual Reviews, 44, 337R (1972).

  DeJongh, D.C., Anal. Chem. Annual Reviews, 42,
  169R(1970).

  Large, R. and Knof, H.,  Org. Mass Spectrom., 11,
  582(1975).

  Fales, H.M., Milne, G.W.A., Winkler, H.V., Beckey,
  H.D., Damico, J.N., and  Barron, R., Anal. Chem.,
  47(2), 207 (1975).
  Alford. A., Biomed. Mass Spectrom., 2, 229 (1975).

  Heller, S.R., Milne, G.W.A., and Feldmann, R.J.,
  Science, 195,253(1977).

GC/MS

  Fenselau, C, Appl. Spectrosc.. 28 (4), 305 (1974).

  Brooks, C.J. and Middleditch, B.S., Mass Spectrom.,
  2,302(1973).

  Junk, G.A., Int. J; Mass Spectrom. Ion Phys., 8(1), 1
  (1972).

  Oswald, E.O., Albro, P.W., and McKmnery, J.P., L_
  Chromatogr., 98(2), 363 (1974).

  Heller, S.R.,  McGuire, J.M.,  and  Budde, W.L.,
  Environ. Sci. Technol., 9,210 (1975).

  Eichelberger, J.W., Harris, L.E., and Budde, W.L.,
  Anal. Chem.. 47,995 (1975).

DIRECT AQUEOUS INJECTION

  Harris, L.E., Budde, W.L., and Eichelberger, J.W.,
  Anal. Chem., 46, 1912(1974).

  "Standard Methods for the Examination of Water
  and Wastewater", 13th Edition, 1971.

  Baker, R.A., J. Amer. Water Works Ass., 58, 751
  (1966).

  Fujii, T., J. Chromatogr., 139,297 (1977).

INERT GAS  PURGING  AND TRAPPING

  Bellar, T.A. and Lichtenberg, J.J., J. Amer. Water
  Works Ass., 66, 739 (1974).

  F.C. Kopfler, R.G. Melton, R.D. Lingg, and W.E.
  Coleman in "Identification and Analysis of Organic
  Pollutants in Water",  1st ed., L.H. Kieth, Ed., Ann
  Arbor Science Publishers Inc., Ann Arbor, MI,
  1976. Chapter 6.
                                             139

-------
  •Bellar, T.A., Lichtenberg, J.J., and Kroner, R.C., J._
   Amer. Water Works Ass.. 66,703 (1974).

   "The  Analysis of Trihalomethanes  in Finished
   Waters  by  the   Purge   and  Trap  Method",
   Environmental Monitoring and Support Laboratory,
   U.S.'JEnvironmental Protection Agency, Cincinnati,
   OH, 1977.

   Bellar,  T.A.  and  Lichtenberg,  J.J.,   "Semi-
   Automated Headspace Analysis of Drinking Waters
   and  Industrial  Waters  for  Purgeable  Volatile
   Organic Compounds" submitted  for publication to
   ASTM, July 1978.

   Bellar, T.A.,  Lichtenberg, J.J. and Eichelberger,
   J.W., Environ. Sci. Technol., 10, 9"26 (1976).
ADSORPTION WITH POROUS
POLYMERS

  Junk, G.A., Richard, J.J., Grieser, M.D., Witiak, D.,
  WTtiak, J.L., Arguello, M.D., Vick,  R., Svec, H.J.,
  Fritz, J.S., and Calder, G.V., J. Chromatogr., 99, 745
  (1974).

  R.G.Webb,  EPA  Report  EPA-660/4-75-003,
  Athens, GA, June 1975.

  G.R.  Harvey,  EPA  Report  EPA-R2-73-177,
  March, 1973.

  Van Rossum, P., and Webb, R.G., J. Chromatogr.,
  150,381(1978).
   Coleman, W.E., Lingg,  R.D.,  Melton, R.G.  and
   Kopfler, F.C., in "Identification  and Analysis of
   Organic Pollutants  in Water", L.W. Keith, Ed.,
   305-327, Ann Arbor Science, Ann  Arbor, Mich.,
   1976.

   Symons,  J.M.,   Bellar,  T.A.,  Carswell,  J:K.,
   DeMarco, J., Kropp, K.L.,  Robeck, G.G., Seeger,
   D.R., Slocum, C.J., Smith, B.L. and Stevens, A.A., J._
   Amer. Water Works Assoc., 67, 634, (1975).

   Brass, H.J., Feige, M.A., Halloran, T., Mello, J.W.,
   Munch, D. and Thomas, R.F., in "Drinking Water
   Quality Enhancement Through Source Protection",
   R.B. Pojasek, Ed.,  393-416, Ann Arbor Science,
   Ann Arbor, Mich., 1977.

   Bellar, T.A. and Lichtenberg, J.J., presented at the
   ASTM  Symposium on  "The  Measurement of
   Organic Pollutants in Water  and  Wastewater",
   Denver, CO., June, 1978.
AIR SAMPLES

  Grob, K., J. Chromatogr., 62, 1 (1971).

  Lao, R.C., Thomas, R.S., Oja, H., and Dubois, L.,
  Anal. Chem., 45(5), 908 (1973).

  Pellizzari, E.D., Bunch, J.E., Berkley, R.E., and
  McRae, J., Anal. Chem., 48(6), 803 (1976).

  Bunn,  W.W.,  Deane, E.R.,  Klein,  D.W.,  and
  Kleopfer, R.D.,  Water, Air, and Soil Pollution,. 4,
  367(1975).
FATTY  TISSUE SAMPLES

  Kuehl,  D.W. and Leonard, E.N., Anal. Chem. 50,
  182(1978).
   Bellar, T.A., Budde, W.L. and Eichelberger, J.W. in
   "Monitoring Techniques for Toxic Substances" D.  CHEMICAL DERIVATIZATION
   Schuetzle,, Ed., American Chemical Society,  1978.
QUALITATIVE HEADSPACE ANALYSIS

  McAullife, C, Chem. Tech., 46 (1971).
  A.E.  Pierce, "Silylation of Organic Compounds",
  Pierce Chemical Co., Rockford, IL, 1968.

  K. Blau  and   G.  King,  "The  Handbook  of
  Derivatives for Chromatography", Heyden and Son,
  Ltd., New York, NY, 1977.
                                              140

-------
SELECTED ION MONITORING

  •Bu'dde.   W.L.. .  and   Eichelberger.   J.W.,   I_
  C):rornaiOgr., 1J4. 147(1977).

  Hichs!b.er»er, J.W., Harris, L.E., and Budde, W.L.,
  Anal. Chem., 46, 227 (1974).
OPEN TUBULAS COLUMNS

 Grab, K., Anal. Chem.. 45,1783(1973).

 Grab, K. and Grob K. Jr., J. Chrotnatogr., 94, 53
 (1974).

 Novotny, M., Anal. Chem.. 50,16A (1978).


CHEMICAL  IONIZATION

  Jelus,  B.L.  and  Munson,  B., Biomed.  Mass
  Spectrom.. 1,96(1974).

  Hatch, F. and Munson, B., Anal. Chem. 49(1), 169
  (1977).
 Field, F.H., Advnn. Mass Spectrom.. 4, 645 (1953).

 Munson, B.. Anal. Cheni., 49, 772A (1977).

 Hunt,  D.F.,  StatTord,  G.C.,  Crow,  F.W., and
 Russell, J.W., Anal. Cherr... 48 (14), 2098 (1976).

 Price, P., Martinsen, D.P., Upham, R.A., SwotTord,
 H.S., and  Buttrill, S.E., Anal.  Chem., 47 (1), 190
 (1975).

 Murata, T., Takahashi, 7.,  and Takcda, T., Anal.
 Chem. 47 (3), 573 (1975).

 Oswald, E.O., Levy, L.. Corbett,  B.J. and Walker,
 M.P..J. Chromatogr., 93, 63 (1974).

 Fales, H.M.,  Milne, G.W.A., and Nicholson, R.S.,
 Anal. Chem., 43 (13), 1785 (1971).

 Holmstead, R.L. and Casida, J.E., J. Ass. Offlc.
Anal. Chem., 57(5), 1050(1974).

Biros, F.J., Dougherty, R.C., and Dalton, J., Org.
Mass Spectrom.. 6(11), 1161(1972).
                                              141

-------