&EPA
             United States
             Environmental Protection
             Agency
             Municipal Environmental Research EPA-600/2-80-098
             Laboratory' '••       August 1980
             Cincinnati OH 45268
             Research and Development
Design
Considerations for a
Thermal
Decomposition   .
Analytical System

-------
                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 ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution-sources to meet environmental quality standards.
 This document is available to the public through the National Technical Informa-
 tion Service, Springfield, Virginia 22161.

-------
                                         EPA-600/2-80-098
                                         August  1980
DESIGN CONSIDERATIONS FOR A THERMAL DECOMPOSITION

                ANALYTICAL SYSTEM
                        by

                  Wayne A. Rubey
     University of Dayton Research Institute
               Dayton, Ohio  45469
              Grant No.  R805117-01-0
                 Project Officer

                Richard A.  Carnes
   Solid and Hazardous  Waste Research Division
   Municipal Environmental  Research Laboratory
              Cincinnati, Ohio  45268
  MUNICIPAL ENVIRONMENTAL  RESEARCH  LABORATORY
       OFFICE OF RESEARCH  AND DEVELOPMENT
      U.S. ENVIRONMENTAL PROTECTION AGENCY
            CINCINNATI, OHIO  45268

-------
                           DISCLAIMER
     This report has been reviewed by the Municipal Environmental
Research Laboratory, U.S. Environmental Protection Agency,  and
approved for publication.  Approval does not signify that the
contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation
for use.

-------
                            FOREWORD


     The U.S. Environmental Protection Agency was created because
of increasing public and government concern about the dangers of
pollution to the health and welfare of the American people.  Nox-
ious air, foul water, and spoiled land are tragic testimonies to
the deterioration of our natural environment.  The complexity of
that environment and the interplay of its components require a
concentrated and integrated attack on the problem.

     Research and development is that necessary first step in
problem solution; it involves defining the problem, measuring its
impact, and searching for solutions.  The Municipal Environmental
Research Laboratory develops new and improved technology and sys-
tems to prevent, treat, and manage wastewater and solid and
hazardous waste pollutant discharges from municipal and community
sources, to preserve and treat public drinking water supplies, and
to minimize the adverse economic, social, health, and aesthetic
effects of pollution. .This publication is one of the products
of that research and provides a most vital communications link
between the researcher and the user community.

     The safe, permanent disposal of highly toxic organic wastes
is vitally important.  The study reported here presents a
sophisticated laboratory system that has been designed and
assembled to provide fundamental thermal decomposition data on a
wide assortment of organic waste materials.
                                Francis T. Mayo, Director
                                Municipal Environmental Research
                                   Laboratory
                               111

-------
                            ABSTRACT


     The safe, permanent disposal of highly toxic organic wastes
is vitally important, and controlled high-temperature incin-
eration is one of the most promising methods for achieving it.
For safety reasons, the thermal decomposition properties of a
toxic organic substance must be well established before large
quantities are subjected to high-temperature incineration.  How-
ever, there is a serious lack of basic experimental data concern-
ing the thermal decomposition behavior of these organic
materials.

     In response to this need, a sophisticated laboratory system
has been designed and assembled to provide fundamental thermal
decomposition data on a wide assortment of organic materials.
This thermal decomposition analytical system (TDAS) is a
specially designed, closed, continuous system consisting of a
versatile and fully instrumented thermal decomposition unit which
is connected to dedicated in-line gas chromatography, mass
spectrometry, and data reduction equipment. .With the TDAS, a
wide variety of organic samples can be subjected to analysis,
including gases, liquids, solids, and even polymeric materials.

     In the TDAS, precisely controlled thermal decompositions
are conducted in a narrow-bore quartz tube reactor.  The various
thermal decomposition products are subsequently collected using
a cryogenic adsorptive trap.  After thermal desorption, the
products are subjected to high-resolution gas chromatographic
analysis, and identification of the various decomposition pro-, ::
ducts is accomplished using in-line coupled mass spectrometry,
A dedicated minicomputer is used for reducing the analytical
data.  From the product analyses obtained through a sequence of
thermal exposures, the thermal decomposition properties for a
given organic sample can be readily obtained.

     The TDAS has been designed to generate fundamental thermal
decomposition data rapidly, economically, and safely.  These
laboratory data should be instrumental in establishing acceptable
criteria for controlled high-temperature incineration of toxic
organic wastes    ,  ,

     This report was submitted in partial fulfillment of Grant
No. R805117-01-0 by the University of Dayton Research Institute
under the sponsorship of the U.S. Environmental Protection
Agency.  The report covers the period September 1, 1977 to
April 15, 1979, and work was completed as of May 31, 1979.

-------
                          CONTENTS
Foreword	
Abstract	    v
Figures	] *   vj_
Tables		'. '. '. '. '.   ix
Abbreviations and Symbols	'.'.',    x
Acknowledgments	

   1.  Introduction	    ^
   2.  Conclusions	    4
   3.  Recommendations	    5
   4.  System Precursors	    6
   5.  Concept and Basic Design of TDAS	   14
   6 .  Thermal Decomposition Unit  . .	, .	   20
            Sample handling and insertion	   24
            Transport of sample and effluent products  ...   29
            Quartz tube reactor	   30
            Collection of effluent products	   48
            TDU instrumentation	. . .	   50
   7.  Chromatographic Separation and Analysis ..........   59
   8.  Mass Spectrometer with Dedicated Minicomputer
       and Associated Data Reduction Equipment	   68
   9.  Utilization of the TDAS in Thermal
      . Decomposition Studies	   73

References	   75
Bibliography	   89
Appendices                                    :   .

   A.  Gas flow analysis:  a detailed analysis of
       gas flow and dispersion in a" narrow-bore
       circular cross section tube of folded
       configuration	   g^
            Straight tube of circular cross section .....   93
            Folded tube of circular cross section 	  102
   B.  Error analysis of mean residence time
       determination	  109
   C.  Quartz tube reactor heat transfer analysis .......  112
   D.  TDAS transducer calibrations 	.,..„,*...««	  123
   E.  Preparation of gas chromatographic open
       tubular columns	  -j.31
   F.  Application of thermal focusing and programmed
       temperature gas chromatography in the TDAS .......  139
   G.  Approximate costs of TDU components	  142

                             v

-------
FIGURES
Number
1
2

3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

19
20

Transport between adjacent analytical components
Schematics of the first generation discontinuous
system 	
First generation thermal equipment 	
Variety of residence time distributions 	
Block diagram of TDAS . 	
Simplified schematic of TDAS 	 ......
Artist's rendering of assembled TDAS 	 	
Diagrams of the TDU 	 	
TDAS residence time distribution 	
Sample insertion assembly 	
Glass rod probe 	
Packed tube probe 	
Sample handling devices 	
Transport tubing tee connector 	
Heated transport lines 	
First generation reactor configuration 	
Proposed second generation reactor 	
Secondary flow and circular cross sectional
flow path 	
Sectional view of furnace interior 	
Quartz tube reactor components 	
Page
8

• 9
10
12
15
16
18
21
23
25
27
27
28
31
32
34
34

. . . . 36
38
39
    VI

-------
Number                                                       Page

  21    Assembled quartzware 	  40

  22    Quartzware assembly in three-zone furnace 	  42

  23    Detail of quartz tube reactor 	  43

  24    Additional details of reactor assembly	  45

  25    Square wave profile of.gas phase temperature
        versus time	  46

  26    Scanning electron microscope photographs of
        quartz surfaces 	........	  47

  27    Schematic of collection trap assembly 	  51

  28    Adsorption trap insert .	  52

  29    TDU console and instrumentation modules 	  53

  30    Interior of gas flow control module	  55

  31    Fundamental schematic of  mass flow transducer 	  55

  32    Pressure transducer schematic	'.	  58

  33    Schematic of GC column installation ,«,.„,..„...	  60

  34    Glass OTC installation in modified Pye 104	  61

  35    Glass OTC separation of normal alkanes	  63

  36    Chromatographic separation of a complex mixture  ....  64

  37    Separation of complex chlorinated hydrocarbon
        mixture	  65

  38    LKB 2091 GCMS and TDU	  69

  39    Minicomputer for data reduction 	,	  71

  40    Mass spectra searching terminal	  72

 A-l    Effect of residence time  bandwidth 	  92

 A-2    Quartz tube reactor data	 . .	105

 A-3.    Relative bandwidth as a function of mean
        residence time ..„,....		108
                              Vll

-------
Number

 C-l

 C-2


 C-3

 C-4

 D-l

 D-2


 D-3

 D-4


 E-l

 E-2

 E-3
                                                    Page

Details of quartz tube reactor assembly 	  113
Simplified version of fine bore connecting
tube 	
114
Graph of bulk temperature versus distance 	  120

Sketch of heat transfer condition 	  121

Mass flow transducer calibration 	  125

Calibration of transducer readout
instrument	  126
Calibration of T2 readout instrument
129
Calibration of temperature selector
readout instrument	 . .	  130

Capillary cleaning apparatus 	  133

Whisker-textured tubing bore	  135

Apparatus for drying glass tube interior 	  136
                               vi a, i

-------
                             TABLES

Number                                                       Page

   1    Sample Handling Modes	    26

   2    TDU Temperatures		    33

 A-l    Calculated Fluid Dynamic Values 	......,., ..'. , .   106

 D-l    Differential Pressure Transducer Calibration 	   124

 D-2    Mass Plow Transducer Calibration	   125

 D-3    Calibration of EP-3465A Digital Multimeter .......   126

 D-4    Reactor Temperature Measurements 	    127

 D-5    Omega K-l Thermocouple Chromel-Alumel	   127

 D-6    Calibration of Omega Digital Temperature 	   129

 D-7    Calibration of Omega Digital Temperature
        Indicator Model 175, Type K	   130
                               IX

-------
                   ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS

AMP
CHAZ
DDT
GC
GLC
GCMS
GCMS-COMP

ID
MS
OD
OTC
PBB
PCB
PGC
PTGC
RTD
TDAS
TDU
UV
amu
eV
kV
ms
mv
tc
w
operational amplifier
N-cyclohexyl-3-azetidinol
dichloro-diphenyl-trichloroethane
gas chromatography
gas-liquid chromatography
gas chromatography mass spectrometry
gas chromatography mass spectrometry with
dedicated computer
inside diameter
mass spectrometer
outside diameter
open tubular column
polybrominated biphenyl
polychlorinated biphenyl
pyrolysis gas chromatography
programmed temperature gas chromatography
resistance-temperature detector
thermal decomposition analytical system  ..
thermal decomposition unit
ultraviolet
atomic mass unit
electron volt
kilovolt
milliseconds
millivolts
thermocouple
watts
SYMBOLS
A    — cross-sectional area
 m
C    — concentration
C    — average radial concentration

D    — effective diffusion coefficient
 e
D    — gas diffusion coefficient

F    — flow
                                x

-------
  m
 F
 K
 L
 L
 N.
 N.
 N.
 N
 Dn
 Nu
 Re
 Sc
 R
 R,
 R
 n

-------
k.
 i
m
P
Pa
P2
q
r
ro
rs
rw
 v
 v
 ve
 vm
 x
 y
 z
 21
 n
 e
 x
 V
 IT
 P
 a
 cr.
— initial partition coefficient
— mass rate of flow
— pressure
— differential pressure      '.
— outlet pressure
— pressure at condition 1
— pressure at condition 2
— rate of heat transfer
— rate of heat transfer by convection
— rate of heat flow by conduction
— radius
— inner tube radius
— radius of inner  surface
— radius of outer  wall surface
— time
— residence time
— average  residence  time
— time  coordinate
— velocity
— average  velocity
— exit velocity
— mobile phase velocity
 — coordinate
 — coordinate
 — distance and coordinate
 — revised coordinate
 — fluid viscosity
 — cylindrical coordinate
 — coil to tube diameter ratio
 — kinematic viscosity
 — geometric constant
 — density
 — statistical standard deviation
 — time-based standard deviation

-------
0)
distance-based standard deviation
identity in Appendix A
probability
flow profile parameter
                             xi li

-------
                        ACKNOWLEDGMENTS
     The author wishes to acknowledge the support and encourage-
ment of the EPA Project Officers, R. A. Carnes of the
Municipal Environmental Research Laboratory and L. Weitzman of
the Industrial Environmental Research Laboratory.  Thanks are
also due to D. S. Duvall, the Principal Investigator for this
grant and Head of the Environmental Sciences Division at the
Research Institute.  During the course of this research effort,
many of my colleagues graciously offered suggestions and advice
pertaining to the development of this system.  Particular
appreciation is extended to R. A. Grant for his exceptional
scientific glass blowing and his many ideas concerning the
fabrication of the system quartzware.  We also wish to express
our gratitude to J. A. Mescher for her consultation with
respect to the mass spectrometry requirements, and to R. N._Ely,
who assisted with the transducer installations and calibrations.
                               xiv

-------
                            SECTION 1

                           INTRODUCTION
      Thousands of organic compounds are used daily in our
 modern industrial society.  Unfortunately, it seems that many of-
 the_more useful industrial chemicals also have associated
 toxicities of one form or another.   Even though there are     •  :
 continual searches to find chemical substitutes that are less
 toxic,for the foreseeable future,  there will be large quantities
 of toxic substances that must eventually be disposed of once they
 have fulfilled their intended industrial use.           .    ,   , :

      Many procedures have been proposed for the disposal of  toxic
 organic compounds (1,2).   Some of  these procedures are temporary•
 that is,  they consist essentially  of containment methods.  Other'
 methods are more  permanent in nature,  as they  produce chemical
 changes in the substance  with the  intention of rendering the
 material  safe.  Of the various methods involved with changing the
 structure of  the  toxic organic molecule,  chemical  and thermal
 techniques have received  the  most attention (3).   in addition to
 the  straightforward chemical  and thermal  reaction  procedures,
 special techniques such as microwave plasma detoxification  (4,5)
 and  molten salt decomposition methods  (6)  have been  employed.

      Controlled high-temperature incineration  is one of  the most
 promising methods  for  disposal of hazardous  organic  wastes.   Both
 land-based incineration and incineration  on  ocean-going  vessels
 have  been utilized for the permanent disposal  of organochlorine
 wastes  (7)  and  large quantities of other  types  of  organic
 industrial  wastes  (8).  Incineration-has  also  been employed for
 the disposal of various pesticide formulations  (9).  Although
waste disposal by  incineration at sea  is  expensive,  it is
 receiving  increased consideration  (10).

     Before methods can be developed for  the immense variety of
organic materials, precise information is needed concerning their
 thermal decomposition properties.  Much remains to be learned
about the thermal decomposition of these organic materials and
the nature of the decomposition products formed.  It is antici-
pated that this basic information can be obtained in the
laboratory using very small quantities of sample—micrograms or
J-6SS •

-------
     There are many advantages to generating fundamental thermal
decomposition data in the laboratory.  First, it is anticipated
that these thermal decomposition experiments can be conducted
safely in a properly equipped laboratory.  There would be a
much higher element of risk if the initial studies were conducted
in larger throughput units.  Second, data generated in the
laboratory can be much more precise and comprehensive than
thermal decomposition data generated in field-scale studies.
Third, laboratory thermal.decomposition data- can be obtained
economically and in a short period of time.

     Once the thermal decomposition properties of a particular
material have been characterized in the laboratory, the pre-
liminary decision can be made as to whether high-temperature
incineration is a viable disposal route for that particular
material.  If no adverse characteristics are detected during the
laboratory experiments then,with due caution, the material can
be subjected to larger-scale thermal decomposition studies.
However, if the laboratory data indicate difficulties or problem
areas, then incineration is probably not a viable disposal method
for that particular substance.

     As an example, chlorinated hydrocarbons are indeed valuable
industrial chemicals; however, they are also known for their
chronic and sometimes acute toxicity.  Many of the compounds on
the recent priority pollutant list are members of the chlorinated
hydrocarbon family.  High-temperature incineration is being con-
sidered for the disposal of many of these chlorinated hydro-
carbons.  Prom previous thermal decomposition experiments with
chlorinated hydrocarbons and chlorocarbons,  it was determined
that detailed thermal decomposition data were necessary before
conducting large-scale thermal disposal  operations.  Specifi-
cally, it was concluded that although the parent molecule may be
dealt with at certain thermal exposures, other  secondary or
intermediate reaction products may be produced  that  are more
toxic than the parent substance.  Consequently,  there  is an
urgent need to determine  the  full thermal destruction  data  for
toxic organic materials before .progressing  to the  larger-scale
disposal units.

      Laboratory-scale thermal decomposition studies  of various
organic materials  are being performed at the University of  Dayton
Research  Institute.  These experiments  are  being conducted  using
recently  developed second generation thermal decomposition
 instrumentation.   The present instrumentation is referred  to as a
 thermal decomposition  analytical system (TDAS).   This  system
 incorporates  a versatile  in-line thermal decomposition unit with
 sophisticated analytical  instrumentation capable of  analyzing the
various  decomposition products.

      The objective of  this report is to describe,  in detail,  the
 various  aspects associated with the design of  the TDAS.   This

-------
second generation system has been designed specifically to pro-
vide precise and accurate thermal decomposition data for a wide
variety of organic samples.  The TDAS has-, also been designed to
provide this information safely and in a short period of time.

-------
                           SECTION 2

                          CONCLUSIONS
     A versatile thermal decomposition analytical system (TDAS)
has been designed, assembled, and tested.  With .this laboratory
system, the thermal decomposition behavior of toxic organic   ..,
substances and a wide assortment of other organic compounds.can
be experimentally determined.  The effects of the major thermal
decomposition variables  (i.e., exposure temperature,.composition
of atmosphere, mean residence time and residence time distribu-_
tion, and reactor internal pressure) can be individually investi-
gated with the TDAS.                                       .

     A number of specific design features have been incorporated
into this comprehensive  system, with the following results:

     o  Because of the closed continuous design concept of this
        system, toxic samples can be safely tested.

     o  A sample can be  subjected to a very precise thermal ex-
        posure in the TDAS.  At selected reactor temperatures
        ranging from 200° to 1150°C, the maximum temperature
        variation is less than ±2°C.

     o  This  system is capable of subjecting a  sample to a
        precise mean residence time ranging from 0.25 to 5.0
        seconds.  In addition, this system provides a narrow
        Gaussian residence  time distribution.

     o  With  the flexibility designed  into the  TDAS, thermal  de-
        composition studies can be  conducted using pure organic
        substances or organic mixtures,  and only small quantities
        of  sample are needed  (micrograms).  Samples can be
        either gases, liquids, or solids.  In addition, these
        samples can be  subjected to thermal decomposition  studies
        in  any of a wide variety of atmospheres.

     o  Comprehensive analyses of the  thermal decomposition
        products  are performed through the use  of  gas chromato-
        graphy/mass  spectrometry  (GCMS)  analytical  instrumen-
        tation .

     o The TDAS  can provide fundamental thermal decomposition
        data economically  and on  a  quick response  basis.

                                4

-------
                           SECTION 3

                        RECOMMENDATIONS
     The thermal decomposition properties of organic substances
can be determined with the use of a specially designed thermal
decomposition analytical system  (TDAS) .  It is therefore   •••-',-
recommended that this laboratory system be utilized for estab-
lishing the thermal decomposition behavior of individual toxic
organic substances  and multicomponent industrial organic wastes.
Such fundamental thermal decomposition data would be instrumental
in determining the thermal requirements necessary for environ-
mentally acceptable disposal of hazardous materials.'       '-••  •

-------
                           SECTION 4

                       SYSTEM PRECURSORS
     The incineration of toxic organic wastes is profoundly
different from the typical incineration of, municipal refuse.  The
temperatures required for adequate thermal decomposition of
hazardous organic materials are.much higher than those employed
in other types of incineration.  These incineration temperatures
must be sufficient to decompose not only the parent organic
molecule, but also the various intermediate thermal reaction pro-
ducts  (11,12.).

     The residence time of molecules at these elevated tempera-
tures must be sufficient to produce complete thermal decomposi-
tions.  Therefore, to insure that organic compounds experience
sufficient temperature and residence time, high-temperature
afterburners are usually incorporated in the process stream.  The
design and selection of materials for these high-temperature
afterburners has received a considerable amount of study, and
work is ongoing in this area  (13,14).

     Recommended guidelines have been prepared for selecting the
exposure temperature, residence time, and excess oxygen needed
for the incineration of hazardous organic waste materials  (15).
Most of the data compiled thus far have been associated with^the
decomposition of the parent molecule.  Detailed data pertaining
to the various products formed during the thermal decomposition
are practically nonexistent.

     Instrumental chemical analysis techniques have been employed
for many years in the monitoring of effluents from thermal re-
actors and various thermal analysis instruments.  Some of these
effluent monitoring techniques  (15-22) rely upon periodic GC
separations and quantitations of the emerging products.  Other
methods of monitoring effluents employ different forms of
analytical instrumentation  (23-26).  Gas chromatography  (GC) has
been used extensively for monitoring combustion devices  (27), and
also for monitoring the effluents from microreactor devices that
have been incorporated into the GC flow path  (28).

     Thermal reactors with  in-line effluent sensing instrumen-
tation have been used in basic reaction kinetic experiments  (29,
30).  The coupling of thermal analysis instrumentation with^GC
has been used extensively in working with nonvolatile organic

-------
materials  such  as  plastics  and polymers.   The  term pyrolysis  gas
chromatography  (PGC) has been applied  to  this  coupled  instrumen-
tation  technique,  and  it is used  routinely for the analysis and
characterization of organic 'polymeric  substances  (31-35).

     A  gas chromatograph is an extremely  versatile and sensitive
analytical instrument.  If  due consideration  (36)  is given to
the transport of sample between adjacent  analytical components
 (see Figure  1), it is  then  possible  to assemble analytical
systems that by their  nature can  provide  a vast amount of
information.  There are many benefits  associated with  the
analytical system  approach.  Briefly,  precise  information can be
obtained rapidly from  analytical  systems  that  cannot be gathered
from isolated instrumental  analysis  techniques.

     Prom previous experiences with  systems  (37,38) that used GC
in conjunction  with thermal analysis instrumentation,  a system
was developed that involved a high-temperature quartz  tube re-
actor and a  GC.  Schematics of this  discontinuous  system are
shown in Figure 2, and a photograph  of the  assembled thermal  de-
composition  equipment  is presented in  Figure 3.  With  this
system, a sample was. first  gradually vaporized in  a continuously
flowing carrier gas and then passed  through a  narrow-bore, high
temperature  quartz tube.  Upon emergence  from  the  controlled
high-temperature exposure,  the various condensable effluent pro-
ducts were captured by an adsorptive trap.  This trap was then
inserted into the  modified  chromatograph, where the products  '
were thermally desorbed and subsequently  analyzed by programmed
temperature  GC  (39-42).

    ^A number of environmentally  important compounds have been
examined for their thermal  decomposition  properties with the
above relatively simple discontinuous  system:  DDT, Mirex,
Kepone, hexachlorobenzene, PCBs,  hexachlorocyclopentadiene, and
PBBs.  Other workers (43-44) have employed similar instrumen-
tation to determine the thermal decomposition  properties of some
volatile compounds of high environmental  interest such  as
benzene, vinyl chloride, and acrolein.

     The determination of thermal decomposition properties in the
laboratory can be accomplished with very  small amounts  of
sample; in addition, the various parameters' affecting the thermal
decomposition can be closely controlled.   How,.well these
laboratory data extrapolate to field scale units remains to be
seen.  However,  in one particular case (the Kepone Incineration
Test Program at Toledo, Ohio), close agreement was found between
laboratory thermal decomposition data and that obtained from a
special pilot-scale test facility (45).

     Although the relatively simple discontinuous system as
shown in Figure 2 provided some much needed information in a
relatively short period of time,  it was already apparent that

-------
             EXTENDED CHROMATOGRAPHIC SYSTEM
        UPSTREAM
        ANALYTICAL
        INSTRUMENT
DOWNSTREAM
ANALYTICAL
INSTRUMENT
                       CHROMATOGRAPH
Figure 1.  Transport between adjacent analytical components

-------
                              A.  . COMPRESSED AIR,, BREATHING .QUALITY GRADE


                              B.   TWO-STAGE PRESSURE REGULATOR


                              C,   "HYOROPURGE" FILTER " "•   •- '• •".  5-


                              D   FLOW CONTROL VALVE


                              'E   PRESSURE TRANSDUCER    " ''.',.-,-


                              F   SAMPLE ' HOLDER", PY.REX  ' ..   '  '


                              Gi •  HEATED INLET CHAMBER '     '.-'••


                              H   QUARTZ TUBE . -     '        '


                              I   HEATED OUTLET CHAMBER


                              J   EFFLUENT TRAP, TENAX-GC OR CHARCOAL


                              K   FLOW METER    -. . . „,.... . r ... ,.
                                       INSULATION
EXTERNAL POWER
• SUPPLY AND
TEMPERATURE
CONTROL


A
INJECTOR
CAP





'
\\\\\\\\
ir^1






WUi
vm,

OVEN DOOR
TRAPPING
MEDIUM

h^^%$M
^\ ^^
ALUMINUM BLOCK
WITH EMBEDDED
CARTRIDGE HEATER

t
HELIUM
CARRIER

v\\\\\V<

; HFID

^s^
x




^INSULATION

	 1» VENT


GLC COLUMN
NPYREX)
I.8M OV-I



Figure  2.
Schematics  of  the  first  generation  discontinuous
system.

-------
            TRANSDUCER
           CONTROL UNIT
DIGITAL READ-OUT
 OF TEMPERATURE
 AND PRESSURE
LINDBERG
FURNACE
                                                                 FLOW METER
 TRANSDUCER
POWER SUPPLY
 PRESSURE
 TRANSDUCER
          HYDRO-PURGE
            FILTER
       COMPRESSED
           AIR
      . HEATED
       OUTLET
                                       SATURATOR
         Figure  3.  First  generation thermal  equipment.
                                     10

-------
this thermal concept could be vastly improved to provide a highly
versatile in-line closed system capable of dealing with a wide
range of thermal decomposition problems.  An expanded system with
increased versatility is definitely needed to handle a wider
range of toxic organic samples and to accommodate the very com-
plex mixtures of thermal decomposition products.  The large
number of products formed at intermediate temperatures and the
inability to sufficiently separate and identify these products
necessitated the use of more sophisticated analytical instrumen-
tation.  Also, the advantages presented by an in-line closed
system would be such that gases, liquids, solids, mixtures
thereof, and even nonvolatile organics could be tested.  At the
same time, it was realized that a more sophisticated reactor
could be designed and fabricated.  This device could be fully
instrumented and thereby provide very accurate thermal decom-
position data along with the corresponding detailed chemical
analysis of the effluent products.

     GC is an extremely valuable technique for separating and
quantitating organic compounds that can be volatilized.  However,
it has a weakness when it comes to qualitative analysis or
identification of the various constituents of a complicated
organic mixture (46).  This shortcoming of GC has been all but
overcome by the combined instrumental analysis technique of GCMS.
This powerful instrumental combination is ideally suited for the
analysis of organic pollutants  (47-49).  Also, coupled GCMS seems
to be compatible with an upstream in-line thermal reactor.

     With respect to the thermal reactor, it is important that
the mean residence time and the residence time distribution be
accurately known.  This is one of the distinct advantages of
determining a material's thermal decomposition properties while
in the laboratory.  For example, in the much larger pilot- and
field-scale units, a wide variety of residence time distributions
might be encountered, as typified by the residence time profiles
shown in Figure 4.  For experiments conducted in the laboratory,
a narrow bore flow path can be used that produces residence
distributions as shown in Figure 4(f).  Thus if the residence
distribution is known, one can avoid the situation in which some
molecules rapidly pass through the high-temperature zone, and
other similar molecules experience a sustained exposure.

     After considering these concepts and desired improvements,
we decided that a suitable thermal decomposition system could
be designed and assembled that would have a wide range of
versatility.  Such a system would be capable of handling and
testing various toxic organic waste materials—both relatively
pure substances and very complicated organic mixtures.  Also,
broader ranges of exposure temperature and mean residence time
were anticipated for this system.  Likewise, with appropriate
instrumentation, this system should be capable of conducting
tests at different reactor pressures.  In addition, this system
                               11

-------
         (a)
                                                   MIXING CHAMBER
         (b)
 MIXING CHAMBER
 WITH CHANNELING
         id
                                                   EXPONENTIAL DECAY
               <
               CO
               o
         tdi
BI MODAL OR
MULTIMODAL
         lei
GAUSSIAN WITH
LARGE VARIANCE
         (0
                                                   GAUSSIAN WITH
                                                   SMALL VARIANCE:
                                TIME-
Figure  4.   Variety of residence  time  distributions
                                12

-------
should be capable of conducting thermal decomposition, tests in a
variety of flowing atmospheres — for example, air, nitre ?,n,
helium, and mixtures thereof.
                              13

-------
                           SECTION 5

                CONCEPT AND BASIC DESIGN OF TDAS
     The rationale behind the design concept of the TDAS has not
changed basically from that of the earlier discontinuous system
(39,40).  The sample is still inserted into the system and then
gradually vaporized in a flowing carrier gas.  The vaporized
compounds are subsequently subjected to a controlled, high-
temperature exposure.  The components that emerge from the high-
temperature environment are then collected and subjected to
instrumental chemical analysis.  This same thermal analysis
format has been employed with respect to the TDAS, but each
operation within the system is much more sophisticated, thereby
producing greatly increased experimental versatility.  Each of_
these procedural operations will be discussed at length later in
the report.

     The major design changes over the earlier system are
centered around the design of the reactor, the closed continuous
system concept, and also the vastly increased analytical cap-
ability that is now provided by an in-line gas chromatograph/
mass spectrometer/dedicated computer  (GCMS-COMP).  Numerous re-
finements  have also been designed into the TDAS, and these are
detailed in Sections 6 and 7.

     Many design objectives were associated with the develop-
ment of the TDAS.  This system should be capable of conducting
precise thermal decomposition tests.  More precisely, it should
be capable of experimentally determining the effects of the five
prominent thermal decomposition variables—exposure temperature,
gaseous atmosphere, pressure, mean residence time, and residence
time distribution.  In addition, the  TDAS  should be able_to ,
accommodate almost any type  of organic material.  Also, it
should be capable of analyzing all of the  thermal decomposition
effluent products.  This closed continuous  system  should be
capable of dealing with toxic materials.  Also,  the TDAS should
be capable of generating data  on a quick response basis.

     Figure 5 shows  a block  diagram of  the  TDAS, and Figure 6
is a simplified  schematic  of the assembled  TDAS  components.
Studies can be conducted with  the TDAS  using almost  any com-
pressed gas as the carrier and thermal  decomposition atmosphere.
Indeed, pyrolytic  studies  can  be performed using inert  gases,
and oxidative studies  can  be conducted  using air or  any other

                                14

-------
                                       Q

                                       O
                                       o

                                       VI
                                                     C/3
                                                     
-------
                                           IN-LINE
                                          GC/MS/DS
A.  HELIUM GAS
B.  COMPRESSED AIR
C.  INERT CARRIER
D.  PRESSURE REGULATOR
E. .OXYGEN SCRUBBER ;
F.  DIRECTIONAL VALVE ,
G.  FILTER
 H.  FLOW CONTROL VALVE
.).  FLOW TRANSDUCER
 J.  PRESSURE TRANSDUCER
 K.  INSERTION CHAMBER"
 L  TEMPERATURE  PROGRAMMER
 M.  REACTOR IN FURNACE
 N.  PRODUCT COLLECTION TRAP
    Figure 6.   Simplified schematic of  TDAS.
                         16

-------
 oxygen-containing carrier.   Accurate measurements of pressure and
 flow can be readily obtained with the TDAS.   The internal
 pressure in the reactor and mass flow rate (thus mean residence
 time)  can be continuously monitored by in-stream instrumentation.

      In^the earlier work with the discontinuous system,  only low-
 volatility organic samples  could be tested.   The TDAS has been
 designed for measuring the  thermal decomposition properties  of a
 wide range of organic  samples—gases,  liquids,  solids, and even
 polymers.   Complicated organic mixtures can  also be  tested with
 the  TDAS by using the  process of slow vaporization of the sample.
 After  vaporization,  the gas phase molecules  are subjected to pre-
 cise high-temperature  conditions (these range between 200° and
 1150°C and are held  with ±2°C)  in a fused  quartz tube reactor.
 The  emerging products  are then rpaidly, swept into a  cryogenic,
 adsorbent in-line trap where the condensable products are
 captured at temperatures down to minus 110°C.   The collected
 effluent products are  subsequently thermally desorbed from the
 trap and then subjected to  GC analysis using high-resolution
 glass,  open tubular  columns.   The separated  compounds are then
 subjected  to detailed  analysis  by mass spectrometry  (MS).  The
 thermal  decomposition  test  parameters  and  the chemical analysis
 data are_then retained,  and a complete analysis  of the thermal
 decomposition process  is carried out.

     Although the TDAS  is a complicated system with many  inter-
 acting  components, the  operation of  this system  and the conduct-
 ing  of  thermal decomposition tests  can be  accomplished with
 relative ease (that  is,  after suitable familiarization and
 experience).   Most of  the crucial  instrumentation  for conducting
 experiments  with  the TDAS have  been  mounted  in an  instrumentation
 console, and most of the test functions can be continuously
 monitored  on this console.

     An artist's  rendering  of the  assembled TDAS is presented  in
 Figure 7.  This equipment is  located in a  temperature and
 humidity controlled laboratory  that was specifically  designed for
 conducting sensitive chemical analyses.

     The basic procedure for  testing a particular  sample* begins
with selection of  the thermal decomposition atmosphere or carrier
 gas.   Next,  the mode of  sample  insertion must be selected.  Then
 the actual exposure temperature  is chosen, and each of the
 furnace zones are  set to this temperature.  The internal pressure
under which  the test is  to be conducted can now be established
by adjusting the gas flow restrictors  situated downstream of the
reactor.  The next parameter to be established is the mean
*A11 samples subjected to analysis by the TDAS would be screened
beforehand using other chromatographic instruments to verify that
the sample would indeed be conducive to a TDAS examination.
                               17

-------
                                            0)
                                            H
                                             tn
                                             CQ
                                             to

                                            4-1
                                             O

                                             tn
                                             G
                                            •H
                                             in
                                             (U
                                             cu
                                             tn
                                            •H
                                            -P
                                             M
                                             t~-

                                             Q)
18

-------
residence time.  Once the other variables have been stabilized
and measured, it is a simple matter of calculating the desired
carrier flow and dialing off that value with the adjustable flow
control.

     After establishing the thermal decomposition test parame-
ters, the next step in the procedure is to cool down the
effluent collection trap so that the condensable products can be
captured in the adsorptive cryogenic trap. The remaining steps
are relatively routine.  The sample must then be admitted, the
products collected, and after switching to helium carrier gas,
the captured products must be subjected to in-line analysis bv
GCMS-COMP.                              -     ;  •  '     ••     Y

     The above test procedures would yield one set of data points
in the thermal decomposition procedure.  Any one of the above
variables could now be changed for the next thermal decomposition
test, and so on, until the thermal decomposition properties of
the sample 'are adequately characterized.

     Extensive design details covering the various TDAS com-
ponent parts and assemblies are presented in the following
sections.
                               19

-------
                           SECTION 6

                   THERMAL DECOMPOSITION UNIT
     The TDAS is a continuous system that can be viewed as three
in-line instrumental stations or units.  The first unit in the
system is designated as the thermal decomposition unit (TDU).
Block diagrams of the TDU are shown in Figure 8.  Specifically,
this unit consists of the various gas supplies, an instrumen-
tation control console, the sample insertion chamber along with
the high-temperature reactor-, the effluent collection trap, and
the various associated transfer lines.  All of these components
will be discussed later in this section.

     The primary function of the TDU is to subject gas phase
molecules to well-defined thermal exposures and to collect the
various emerging chemical compounds.  The remaining stations of
the TDAS will subsequently separate and analyze these collected
compounds.

     The  schematic shown in Figure 8(a) can be redrawn as in
Figure 8(b) to depict the physical relationships within the gas
flow path of the TDU.  From this sketch, it is seen that the
pressures, temperatures, and volumes can be related according
to the ideal gas law as follows:
                          V
                           2p2
V p
 0^0
 T
                                                               (1)
where V is volume, p  is pressure, T is temperature, and T J-S
temperature averaged  over  the  length of  the reactor.  The
subscript 2 denotes the quartz tube interior, and the subscript
o represents  the  ambient outlet conditions.

     If the impedance or resistance to gas flow presented by the
downstream transfer lines  %,  the packing within the trap Rt, and
the  restrictor  Rr, is such that
                           R
      R
                                                               (2)
 where RCT represents  the resistance to gas  flow of  the  quartz  tube
 reactor assembly,  then the differential  pressure Pd can  be
                                20

-------
                            SAMPLE
                                        HEATED
                                        TRANSFER
       FILTRATION
         AND
        CONTROL
                                                     CRYOGENIC'
                                                      TRAP
                                                     i    t
                          a)  BASIC SCHEMATIC






IMI FT















F
~m








PO























REA(
INSERTION
p
'd








\ / CHAMBER






TRANSDUCERS






\J

V





;TOR




P
2







T2
EFFLUENT
TRAP





VARIABLE
RESTRICTOR



^
pl I








; VACUUM
                                                         ^>
                                                           V F P
                                                           V0'r0'r0
                                                            OUTLET
b)  SCHEMATIC REDRAWN TO  SHOW PHYSICAL RELATIONSHIPS WITHIN
    THE  GAS  FLOW PATH
                   Figure  8.   Diagrams  of  the TDU.
                                   21

-------
expressed as
                                    =   Pl  "  Po
                            (3)
Now, volume flow F can be written as
where t is time, and equation 1 can be rewritten as
                          V
                           2p2
Fotpo
                                                               (4)
                                                               (5)
or
                        irr
F tp
 o ^
  T
                                                               (6)
where r and L are the tube radius and length, respectively.
Therefore, it is seen that time, or, in this case, residence
time, can be written simply as
                                                               (7)
where tr represents the calculated mean residence time of a
substance in the quartz tube interior.

     The equipment used to measure these parameters will be dis-
cussed at length later in this section.  Also, an error analysis
associated with these measurement variables is presented as part
of the appendices.              ,

     One of the distinct advantages in conducting thermal decom-
position experiments using the laboratory scale TDAS is that each
molecule is subjected to essentially the same thermal exposure.
To illustrate, consider the transport of like molecules through
the high-temperature reactor.  First, the individual molecules
experience random entry into the gas flow over a relatively long
period of time.  However, as the pressure, temperature, and
linear gas velocity within the reactor are constant, the
individual molecules generate a residence time distribution as
shown in Figure 9.

     There are many advantages in having a Gaussian or normal
residence time distribution.  Linear combinations of Gaussian
random variables lead to new random variables that are also
Gaussian.  Such simplicity is not the case for most other
                                22

-------
                 
-------
distribution functions.  This is but one of the desirable
features of working with the narrow Gaussian function as opposed
to some of the skewed and distorted distribution profiles depict-
ed earlier in Figure 4.

     Before the sample handling in the TDU is described, it is
interesting to note from Figure 8 that both the mass flow trans-
ducer and the pressure transducer are upstream, yet in-line with
the reactor and product collection trap.  Therefore, with the
use of these in-line sensing devices, thermal decomposition
experiments can be conducted at subatmospheric, atmospheric, or
elevated pressure.
SAMPLE HANDLING AND INSERTION

     The sample insertion section of the TDAS was designed to be
as versatile as possible.  This fabricated sample insertion
assembly is rigidly mounted on a machined aluminum support locat-
ed between the TDU instrumentation console and the high-tempera-
ture reactor furnace.  A photograph of this assembly is presented
in Figure 10.  The actual sample reception chamber consists of
fused quartz tubing surrounded by an isolated aluminum block that
can be readily heated and temperature programmed.

     The exit end of the sample reception chamber empties into
1.0 mm ID quartz tubing that is connected to the inlet of the
quartz tube reactor assembly.  An outer jacket of quartz tubing
surrounds this small bore transfer tube, so that by using a
counter flow of heat transfer air, a uniform temperature
distribution is obtained for the transport of the gas sample to
the reactor assembly.

     There are numerous modes for inserting samples into the
TDAS, and because of the built-in flexibility, almost any type of
organic sample can be examined.  Some of these sample handling
modes are listed in Table 1.  A sketch of the glass or quartz
rod probe is shown in Figure 11, and the special packed tube
probe is depicted in Figure 12.  A photograph of the different
sample handling devices is shown in Figure 13.

     The one major requirement with respect to sample handling
with the TDAS is that substances leaving the sample insertion
chamber must be in the gas phase.  Precautions must be taken to
prevent particulate material from migrating into the heated
transfer lines and the high-temperature quartz tube assembly.

     TDAS  samples are either volatilized or thermally degraded
in the sample insertion chamber.  Volatilization is accomplished
by controlled heating of  the insertion chamber with a special
temperature programmer.   If  sample thermal degradation is
required,  it can be  accomplished using the probe sample holders

                                24

-------
Figure 10.  Sample insertion assembly,
                  25

-------
        TABLE 1.  SAMPLE HANDLING MODES
    Types of
organic samples
    Mode of insertion
        into TDAS
Gases
Liquids
(low and high
volatility substances)

Solids
(molecular weights
less than 800)

Non-volatiles or
polymeric organics
• Gas sampling valve
• Gas-tight syringe

• Syringe
• Ampoule
• Quartz rod probe

• Platinum ribbon probe
o Quartz rod probe
• Packed tube probe

• Platinum ribbon probe
• Packed tube probe
                         26

-------
    SAMPLE DEPOSIT REGION
    2.0 cm BY,0.4 cm DIAMETER
        (SAND BLASTED)
  GAS-TIGHT SEAL
(3-SILICONE 0-RINGS)

              PYREX ROD
            25 cm BY 0.6 cm
                                           \
       ADJUSTABLE
       SET-SCREW
       RETAINER
               Figure  11.    Glass  rod  probe
                 COILED PLATINUM HEATER — CONTROLLED BY
                   FLASH PYROLYSIS POWER SUPPLY
                                   GRANULATED POLYMER SAMPLE
                                     MIXED WITH QUARTZ WOOL
OUTGOING GAS
    PLUS    0	1
 VOLATILES
  ft  c\
                        b /o  \3  V)
             <]	1 INCOMING GAS
                        QUARTZ
                         WOOL

                THIN WALL QUARTZ TUBING
            Figure  12.    Packed  tube  probe,
                                 27

-------
Figure 13.  Sample handling devices
                 28

-------
 and a versatile pyrolysis unit.  These two programmable thermal
 controllers will be described later when discussing the TDU
 instrumentation.

      The sample insertion section of the TDAS has been designed
 so that a Plexiglass glove box can be installed that will
 completely encompass the sample entry region.  If needed, samples
 could actually be prepared in the isolated environment of the"
 glove box and then inserted into the TDAS.


 TRANSPORT OF SAMPLE AND EFFLUENT PRODUCTS

      Three basic criteria must be met for proper transport of gas
 phase samples.   First,  the transport tubing material must not
 interact with any of the chemical substances that constitute the
 flowing media.   Second, it is important that a constant flow of
 sweeping gas be maintained throughout every portion of the trans-
 port line.   Third,  the  gas transfer must be maintained at
 elevated temperatures that are compatible with the flowing media.
 Each of these criteria  require some elaboration relative to the
 transport of sample and products in the TDAS.

      Wherever possiblef fused quartz was used as the tubing
 material for sample transport in the TDAS.   However, certain
 portions of  this continuous  system required metal to metal
 connections,  and the 'decision was  made  to use  glass-lined
 stainless  steel tubing  (from Scientific Glass  Engineering)  at
 these locations.  Other possible tubing materials were  nickel 200
 and  stainless steel.  Indeed,  these  latter  materials are  physi-
 cally more rugged,  but  the glass-lined  SS tubing was selected
 primarily because of its  increased inertness.   Other types of
 materials are also  encountered in  the transport path downstream
 of the reactor.   Stainless steel,  Teflon,  and  polyimide  surfaces
 are  found in  such components  as  Swagelok  unions,  the switching
 valve  interior,  and tubing ferrules.

      The second basic requirement  for good  transport of sample
 involves the  continual  gas sweeping of  the  transfer  tubing.   It
 is desirable  that the transport  gas flow  be maintained at
 relatively high linear  velocities  (that is, greater  than  20 cm
 sec  -1-) .  Thus the inside diameter  (ID) of the  transport tube
 should be relatively  small, e.g.,  less than 1.0 mm ID.  Unswept  •
 recesses in a gas flow  path can present major  transport problems.
 Consequently, the use of various tube fittings  should be
minimized.  Several unions were necessary in this system;
 however, to minimize difficulties at these junctions, the ends
 of the tubing material were faced off with a grinder before
 assembly.

     A common problem in transporting gases arises when a gas
 stream must be split into two or more branches.  In the TDAS,
                               29

-------
this splitting was accomplished with the use of tee connectors
that had been filled with quartz wool (Figure 14).   With the
placement of quartz wool at the point of branching, considerable
gas phase mixing is induced.  This mixing essentially avoids such
difficulties as molecular weight fractionation and non-lxnear
splitting.

     The selection of transport temperature and the maintenance
of uniform temperatures along the transport path, constitutes the
third criteria for proper transfer of sample.  The selected
transport temperature must be high enough to avoid any serious
condensation and possible adsorption effects.  However, it must
be low enough that additional gas phase reactions do not occur
during transport.  Localized cold or hot regions must be avoided.

     The transfer tubing used in the TDAS was not directly heated
but was actually jacketed with much larger diameter glass or
quartz tubing.  The outer tube was first instrumented with
thermocouples located at key positions along the tube.  The
thermocouples were attached to the outer tube with the use of
glass adhesive tape  (No. 27 from 3M Co.).  After electrically
insulating the various conductors along the transport line with
woven glass tape  (Ware Apparatus Co.), the tubular assembly was
wrapped with high-temperature heating tape  (Electrothermal
flexible tape).  The heating tape was then wrapped with another
layer of woven glass tape.  A photograph of an assembled trans-
port line is shown in Figure 15.  Table 2 also shows the
various thermocouple locations and their corresponding maximum
temperatures for gas transport in the TDAS.

     During the course of conducting,experiments with the TDAS,
certain vital  components  in the  transport path are  intentionally
cooled to very low temperatures.  This  localized cooling is
necessary  (50,51) for capturing  the  products  emerging  from  the
TDAS.  Details concerning this  low temperature procedure will
be discussed  later  in this  section.
 QUARTZ TUBE REACTOR

      The design and performance of the high-temperature reactor
 is vitally important to the entire TDAS.  Consequently, 'con-
 siderable attention has been centered upon this component.   As
 stated earlier, the important thermal decomposition variables
 (i.e., gas phase environment, mean residence time and distri-
 bution profile, exposure temperature, and pressure) apply direct-
 ly to the interior of the reactor.

      In the earlier thermal decomposition studies conducted with
 the first generation unit, a narrow-bore folded quartz tube was
 used  (Figure 16).  This particular quartz tube reactor was of
 heavy wall construction and occupied most of the internal volume

                                 30

-------
                             SPLITTER
                               AND
                             VENT TRAP
TRANSPORT
   LINE
GCMS
                                  QUARTZ
                                   WOOL
       Figure  14.  Transport tubing tee  connector.
                            31

-------
Figure 15.  Heated transport lines.
                  32

-------
TABLE 2.  TDU TEMPERATURES
Temperature ' . ,,
selector . Ma™m
_ • , • temperature
positxon Thermocouple location (<>c)
1
2
3
4
5
6
7
8
9
10

11
12

13

14
15
Room temperature (tc in connector)
Insertion block
Interior of pre-reactor tube
Mid point of inlet heat transfer cavity
Reactor end region of inlet cavity
Reactor end region of outlet cavity
Mid point of outlet heat transfer cavity
Junction of glass-lined and quart tubing
Transfer tube between valve and GC
Glass-lined tubing upstream of collection
trap
Effluent collection trap
Glass-lined tubing downstream of
collection trap
Junction of glass-lined tubing and
valve inlet
Switching valve
GC oven interior
30
300
300
400
350
350
400
350
300

300
300

300

250
240
300
             33

-------
  GAS FLOW
  HEAVY 'WALL FOLDED QUARTZ
 TUBE WITH CONSTANT BORE SIZE
(FLOW PATH • 0.84 m by 0.8 mm  ID)
         \
                                                        INSULATION
                                     HEATING ELEMENT  ,
                                   OF UNDBERG FURNACE
                                     (R.T. to 1000°C)
Figure 16.   First generation  reactor  configuration,
      GAS FLOW -»






FINE BORE
ENTRANCE
QUARTZ TUBE
(0.5 mm ID)
\
/////A/////


THIN WALL QUARTZ
(FLOW PATH =1.0
/
///////////,


HELICAL


TUBING
m by 1.0 mm ID)

1
	 /
/////// X /
1 1 \ rK-rrrr?rtf7??v?v?r> 	
t


'///////////


//////////Z^.

\
/X///.
\

.1
//A

FINE BORE
                                                          INSULATION
                          HIGHER TEMPERATURE
                         FURNACE (R.T. to 1150°C)
      Figure  17.   Proposed  second generation reactor
                                 34

-------
 of_its high-temperature furnace.  Many tests were conducted with
 this reactor, and it experienced considerable use at temperatures
 up to 1000°C.  Even so, a number of design refinements were con-
 sidered for the second generation quartz tube reactor.  Better
 control of exposure temperature throughout the gas flow path was
 desired, and the minimization of end effects would be beneficial.
 Also, the ability to conduct thermal decomposition tests at some-
 what higher temperatures was considered most advantageous.
 Therefore, the initial conceptual design for the second genera-
 tion quartz tube reactor was as.shown in Figure 17.   Again, the
 major portion of the proposed new reactor was a narrow-bore flow '
 path of considerable length.  (It is this design concept that is
 of paramount importance in obtaining narrow residence time
 distributions of Gaussian profile.)

      A variety of studies have been conducted on laminar fluid
 flow through narrow-bore flow paths (52-57).  Methods for deter-
 mining gas diffusion coefficients have been developed using the
 precise control of dispersion afforded by the' harr.ow flow path
 (58,59).   Thermal decomposition testing using this narrow-bore
 flow path concept requires a high-temperature tubing material,
 and_fused quartz (also referred to as fused silica and vitreous
 silica)  seems to be the best material presently available for
 high temperature.use (60,61) .   The proposed quartz tube reactor'.
 shown in  Figure 17 uses very fine bore entrance and  exit tubes
 to transport the gas rapidly into and out of  the central portion'
 of the reactor.   Later in this section even further  refinement of
 this  transport concept is discussed.

      Unfortunately,  fabrication difficulties  are encountered  in
 preparing tightly coiled quartz  tubing.   Specifically,  when
 requests  were sent to  various  quartz  fabricators,  the  common
 reply was that they could not  provide helical  quartz  tubing with
 a  precise circular cross section  flow path.   It was  our under-
 standing  that when quartz  tubing  is tightly coiled,  the circular
 cross  section bore distorts  into  an ovalized  cross section  and
 even  takes  the shape of a  "D"  with the flat side being  on the
 inner  diameter of  the  helix.   As  a result of  the apparent
 fabrication  difficulties,  and  before  contemplating other  geo-
metric configurations  for  the  reactor,  it was  decided to  assess
 the possibility  of bending narrow bore quartz  tubing without
changing  the  circular  cross  section of the  tubing bore.

     Maintenance of  the  circular  cross section  flow path  is
essential because  it affects the  ability  to describe the  effects
of secondary  flow  on the residence time distribution.   Secondary
flow occurs whenever a  fluid is passed through  a tube that
exhibits a significant  amount of  axial curvature.  Figure 18(a)
shows^the typical  streamlines associated with  secondary flow,
and Figure 18(b) depicts the corresponding radial mixing pattern.
Secondary,flow has been observed and studied for decades  (62-67).
Dean  (62) studied the effects of  secondary flow as early as  1927."

                                35

-------
         (a)  STREAMLINES
(b)   RADIAL MIXING
                 (c)   CURVED QUARTZ TUBE OF
                      CIRCULAR CROSS SECTION
Figure 18.  Secondary flow and circular  cross  section flow path
                                36

-------
 Also,  pressure losses (68)  were associated with secondary flow
 in bent tubes as early as 1937.  To characterize accurately the
 residence time contribution in a tube that exhibits some degree
 of secondary flow,  it is imperative that the flow, path be of
 circular cross section.   The description of secondary flow in an
 elliptical,  ovalized, or noncircular flow path is practically
 an impossible task.

     Therefore,  upon giving this fabrication problem to the
 University scientific glass blower, thin-wall , 1. 0. mm' ID quartz
 tubing was eventually obtained that retained its circular
 cross  section when  bent  through a 180°  arc at, a radius of 2.1
 cm.  Figure  18 (c) shows  a photograph of a typical cross section
 of curved tubing.   Indeed,  there is no  visible ovalization
 of the tubing bore.   Circular cross section was maintained by
 locally heating  the  tubing on a mandrel and allowing the quartz
 to bend by its own weight.

     The ability to  maintain the circular cross section flow
 path in a curved tube permitted the eventual design of a compact
 folded tube  reactor.   The actual configuration and  dimensions
 of this quartz tube  reactor will be presented later,  along
 with a detailed  description of the gas  flow behavior.
          the ovalization difficulty overcome, it was then
possible to design a complete quartz tube reactor assembly that
could be fitted into a high -temperature tubular furnace.  The
furnace selected for heating this quartz assembly was a Lindberg
Model 54357, a three-zone furnace of hinged construction that
was designed for continuous operation at temperatures up to
1200°C.  This furnace has its own control console (Model 58744-A)
which separately regulates the temperatures of each of the three
zones .  Only the. middle portion of the furnace ' s central zone is
used to heat the high-temperature reactor, although during
operation the entire length of the tubular furnace is held under
isothermal conditions.  Figure 19 shows a sectional view of the
furnace . interior along with the locations of the thermocouples
used for controlling temperatures within the three separate
zones.

     The assembled quartzware, which was eventually positioned in
the high-temperature furnace, is detailed in Figure 20.  This
quartzware, and all of the other glassware employed in the TDAS ,
was fabricated in the University Glass Shop by a scientific
glassblower.  Even though the assembled quartzware is delicate,
it has .experienced considerable use without problems at
extreme temperatures.

     A better understanding of the actual behavior of the entire
quartz tube assembly can be obtained by close inspection of the
schematic shown in Figure 21.  In this figure, three separate
gas flow paths are depicted.   The main flow is the carrier
                               37

-------
                                                  28.0
                                         STEEL HOUSING
  [DIMENSIONS ARE IN CENTIMETERS!
Figure 19.  Sectional view of  furnace interior.
                          38

-------
rf-

S  E!
    _o


    E
                ©
                ©
       C£


       g

       S
       o;
               ©
                                         g
                            o
                            ce:
                        ©
                        ©
                                                                                                                  cr

                                                                                                                  S
                                                                                                                  oo
                                                              g   E
O   5
Q-   E

CL.  ^,
                                                                                                                 3=   O
                                                                                                                 O   Di
                                                                                                                 r=c   ar
                                     "'  Q
                                     >-  o
                                                                                                   oo

                                                                                                   <
                                                                                                   D;
                                                                                                   P
                                                                                                                 o
                                                                                                       a
                                                                                                       i—i
                                                                                                       oo
                                                                  CO
                                                                  -P


                                                                  0)

                                                                  O


                                                                  i1
                                                                  O
                                                                  o

                                                                  S-l
                                                                  o
                                                                  -P
                                                                  o
                                                                  fd
                                                                  Q)
                                                                                                                                Q)
                                                                                                                               -P


                                                                                                                                N-

                                                                                                                               -P
                                                                                                                                d
                                                                                                                               O!
                                                                 O
                                                                 CN


                                                                  (L)


                                                                  S
                                                                  tn
                                                                 •H
                                                                 PL,
                                                               39

-------
n  -> o
°-  UJ O
V  "- ffl
E   CM
o  o —

IIR
III
               55
        v	/
                          0)
                          Csl
                          -P
                          
-------
 stream through the transfer tubing, into the reactor, and out
 the exit transfer tubing.  This gas flow path carries the
 vaporized sample molecules into the reactor, where they are
 subjected to the controlled high-temperature conditions and
 then passed downstream where the various products are collected
 and eventually subjected to analysis.  The other two gas flow
 paths consist of adjustable flows of heat transfer air that
 cool the entrance and exit chambers of the quartz tube assembly
 Thus the main gas flow into the reactor is maintained at lower
 nondestructive temperatures right up to the point of entry into
 the high temperature reactor.   As soon as the sample or product
 molecules leave the reactor,  the gas temperature is abruptly
 decreased to the same nondestruction transport level (approxi-
 mately 275°C).  The gas then flows on to the collection trap at
 approximately the same temperature.  Figure 22 shows the
 quartzware assembly centrally  located in the three-zone furnace.

      The reactor was constructed from quartz tubing obtained
 from Heraeus-Amersil, Inc.,  and their smooth-wall tubing was
 found to possess uniform wall  thickness and bore diameter over
 the entire tubing length.  The tubing diameter measurements, as
 shown in Figure 23,  were obtained by carefully filling the
 particular tubing section with mercury and then removing the
 mercury  and obtaining its respective weight with the use of  an
 analytical balance.   This mercury displacement technique was
 very similar to that described by Barr and Anhorn (69).

      A narrow-bore,  multiple-fold quartz  tube  reactor of race
 track configuration  (Figure 23)  was selected for several reasons
 First, with the relatively short overall  length  of  the main
 portion  of  the  reactor  and with utilization of  the  multiple-fold
 tube design, any  existing longitudinal temperature  gradient
 within the  furnace would be effectively averaged out  over  seven
 cycles,  thereby assuring that  sample molecules encounter
 essentially  isothermal  conditions  during  their controlled
 traverse  through  the  reactor.   Second,  these flow path di-
 mensions  would  permit mean residence time  studies of  between
 0.25  and  5.0 seconds, which was  a  TDAS  design objective.   In
 addition, as there are  no protrusions  into  the flow path,  and
 as  the tubing diameter  transitions  are  very gradual,  eddy
 currents would  be negligible.  With this  size reactor, gas
 phase  thermal decomposition studies would be conducted in the
 laminar flow region at Reynolds  numbers that are orders of
magnitude below the onset of turbulence.  Calculated Reynolds
number extremes for this quartz  tube reactor range from  0.7  to
 140.  Furthermore, there would be a very low pressure drop
across_this reactor even at the very high linear gas velocities.
Also, in view of the narrowness of the gas flow path, back
diffusion in the reactor would be nonexistent.

     In the TDAS design, gas phase molecules are subjected to
a narrow residence distribution while maintaining, within the
                               41

-------
  CARRIER

   SAMPLE
                   QUARTZ  TUBE  ASSEMBLY

                     IN 3-ZONE FURNACE

                     (CAPABLE OF I20O C) .
  CARRIER
fe   +
  PRODUCTS
Figure 22.   Quartzware assembly in three-zone  furnace,
                               42

-------
                       RACE-TRACK
                      CONFIGURATION
                      (3.5 CYCLES)
 CONNECTING TUBING
 70 mm by 0.43 mm ID
 2.1 cm RADIUS-
                      •11.8'cm
70 mm by 0.43 mm  ID

                980 mm by 0.97 mm ID
Figure 23.   Detail of quartz tube reactor.
                       43

-------
reactor, an acceptable ratio of gas volume to surface area.  The
design of this quartz tube reactor permits a detailed descrip-
tion of gas flow behavior within the reactor.  (Such a detailed
analysis is presented in Appendix A.)

     As derived earlier in Section 6, the mean residence time
for gas phase molecules in the above reactor is
                                     1 +
                                (7)
and the standard deviation of the residence time distribution is
                                                  1
                                 ^  _      	fc> T.
         a   =
                2D (I,  + L
                     (v)3
24 v D
24 v' D
                                                               (8)
as obtained from the derivation presented in Appendix A  (see
list of symbols for the designations of the various terms).
An error analysis pertaining to the mean residence time
determination for this reactor has also been conducted  (see
Appendix B).

     Additional details concerning the high-temperature  gas
flow into and out of the quartz tube reactor are illustrated in
Figure 24.  Basically, this design permits negligible flow path
end effects and entrance length distortions.  The steady state
heat transfer effects for this reactor were studied in Appendix
C, and a very rapid temperature transition time was found for
molecules passing through the quartz tube reactor.  Specifically,
Figure 25 shows the incoming gas reaching the' elevated tempera-
ture in less than 5, milliseconds.  This essentially square-wave
profile of  temperature versus time is accomplished through the
use of the  low-temperature heat transfer principle, whereby the
incoming gas is maintained at transport temperatures right up to
the point of entry into the high-temperature region.  Then by
employing fine bore connecting tubes, the linear gas velocity is
increased by a factor of five at the entrance and exit portions
of the high-temperature reactor.

     This high-temperature quartz tube reactor  is capable of
withstanding intense temperatures for long periods of time.
However, it can be easily damaged by some corrosive gases.  For
example, the scanning electron microscope photographs shown^in
Figure 26 were obtained from various sections of quartz  tubing
that were intentionally exposed  to corrosive and destructive
gaseous environments.  From this figure, it is  observed  that the
smooth quartz surface can be drastically changed by corrosive
agents.  Consequently, care must be exercised with respect to
determining what  samples are to be subjected to analysis by the
                                44

-------
                   HIGH -
          VENT    TEMPERATURE
           i'     INSULATION
 CARRIER
  4-
 SAMPLE
CARRIER
  +
PRODUCTS
Figure  24.  Additional details of  reactor assembly,
                             45

-------
               TEMPERATURE VARIATIONS
               CYCLIC OVER SEVEN
               CYCLES
±2.0°(cOl2p00K
                                   • tr = 2.00 SEC -
                       < 5ms
Figure 25.   Square wave profile of  gas phase -temperature versus
              time.                                i   '
                                   46

-------
     CLEAN QUARTZ SURFACE
           (10,OOOX)
QUARTZ SURFACE AFTER ATTACK
BY HF AT HIGH TEMPERATURE
         (10,OOOX)
QUARTZ SURFACE AFTER EXPOSED      QUARTZ SURFACE AFTER EXPOSED
TO NF4HF~ AT 400°C FOR  4 HOURS    TO NF.HF.-, AT 400°C FOR 4 HOURS
            (3,OOOX)                           (10,OOOX)
Figure 26,  Scanning  electron microscope photographs of quartz
            surfaces.
                               47

-------
TDAS.  If indeed the surface of the quartz tube is attacked, and
increased surface area and adsorptive behavior result, then the
precise residence time control no longer applies.  A certain
amount of sample screening is necessary before an unknown sample
can be subjected to thermal decomposition analysis by the TDAS.
COLLECTION OF EFFLUENT PRODUCTS

     As the gas phase sample and its various decomposition pro-
ducts emerge from the high-temperature reactor, they are swept
downstream through the heated transfer line and into the effluent
product collection trap.  This product collection trap is an
important component in the TDAS and requires considerable dis-
cussion.

     As stated in Section 4, the product collection trap must
possess a slight resistance to gas flow, as should the smaller
bore transfer tubes.  However, the primary function of the pro-
duct collection trap is to quantitatively capture the_unreacted
sample and the various products of thermal decomposition that
emerge from the high-temperature reactor.

     The utilization of collection traps at cryogenic
temperatures has been in practice for many years in the field of
gas chromatography  (70-74).  Indeed, the benefits of  low-
temperature trapping of samples are well known.  However, there
are also numerous problems associated with low-temperature
condensing traps.  One of the major difficulties is that of the
condensing out of very large quantities of water along with the
various organic compounds of interest.

     Many procedures have been developed for  thermally desorbing
collection traps, including cryogenic  sample  collections  (75-79).
Some of these techniques can liberate  the captured  sample in a
very short period of time.  It can then be passed downstream into
either a chromatograph or  some other analytical  instrument.

     One of  the most useful and  intriguing sample trapping
methods was  devised by Kaiser  (80-83) .  He employed the
properties of an  adsorbent  trapping media along  with  cryogenic
trapping for the  collection of sample.  He then  used  thermal
desorption for  liberating  the  trapped  contents.  Basically,
Kaiser employed  the concept of a reversible  countercurrent heat
exchanger for capturing  and liberating substances while  in an
in-line  system.   This  trapping technique has  seen a considerable
amount of use  in both  in-line  and discontinuous  sampling systems.

      The selection  of  trapping media  is very important,  and  there
are many different  trapping materials  commercially  available._
Some of  the  commercially available trapping  media used primarily
 for collecting  organic compounds are  as  follows: carbon

                                48

-------
 molecular sieves, graphitized carbon black, Ambersorb adsorbents,
 XAD resins, chromosorb polymer series, Porapak polymer series,
 Tenax GC, activated charcoal, and polyurethane foam.  Studies
 have been conducted concerning the physico-chemical behavior and
 trapping properties for many of the adsorbent trapping media (84-
 89).  There are numerous applications to which these various
 trapping materials have been applied (90-100).  Also, to increase
 trapping versatility,  some of these trapping media have been
 placed in series arrangements and then used in the same trappincr
 device,   -      .           .   ..                          .   c.c  ^

      The adsorbent trapping material that was eventually selected
 for  use in the TDAS was Tenax GC.   There were many reasons for
 selecting this particular porous polymer.   Tenax GC can withstand
 high temperatures  before appreciable thermal degradation occurs.
 High molecular weight  materials  can be trapped and readily
 liberated by thermal desorption  from this  particular adsorbent.
 Also,  water in the gas flow path does not adversely affect the
 behavior of this adsorbent to any great extent (101) .   Tenax GC
 does  have some disadvantages relative to some of the other
 trapping materials,  but for the  intended uses in the TDAS,  it
 seemed to be the best  material presently available for  in-line
 cryogenic trapping of  organic products.

      Tenax GC was  introduced in  1969  and was initially  intended
 for  use as a gas chromatographic column packing.  -However,  it
 was  quickly recognized that it could  also  be used most
 beneficially as  an adsorbent trapping material.   Tenax  GC  is  a
 porous  polymer based on 2,6-diphenylparaphenylene oxide and is
 reported to have an  upper  temperature limit  in the vicinity of
 450°C  (102).   Even so,  for  sensitive  analytical  work where  a   '
 minimum of degradation or bleed  can be  tolerated,  its upper
 temperature limit  is significantly below this  value.  Many
 workers  routinely  condition Tenax GC  traps at  325°C  in  a flowing
 inert gas.

     In  the  past five  years,  the use  of  Tenax  GC  as a trapping
 medium has  seen  wide use.   It has been used  to capture  and
 analyze  organic  compounds in  environmental samples  (103,104),
 volatile polar organics found  in water  (105), organic com-
 ponents  present  in human breath  (106) , pesticides, poly-
 chlorinated biphenyls  and polynuclear aromatics  found in surface
 and drinking waters  (107,108), the gas phase of cigarette smoke
 (109), organics  in marine biota  (110), and the exhaust of net
 engines  (111) .

     If a Tenax GC trap is swept with a high velocity gas
 stream while at high temperatures, high molecular weight
 substances can be quantitatively liberated from the adsorbent
particles  (12,112).  Accordingly, this high velocity sweeping
procedure has been incorporated into the subsequent design of
the in-line trap used in the TDAS.
                               49

-------
     A schematic of the effluent collection trap used in the
TDAS is shown in Figure 27.  This design is modeled after the
trapping concept as introduced by Kaiser.  However, rather than
using a bidirectional flow of nitrogen, the gas flow through the
jacket is discontinued during the thermal desorption mode.  To
liberate the sample from the in-line trap, the temperature of
the aluminum heating block that surrounds the trap tubing is
linearly increased to an eventual desorption temperature of
280°C.  Note in Figure 27 that the trap contents are .designated
simply as in-line trapping media.  The reason for this is that
for most of the work done with the TDAS, an adsorbent porous
polymer trapping medium is used  (specifically, an 80/100 mesh
Tenax GC trap as shown in Figure 28).  The remainder•of the
sample collections are, obtained using a trap containing
phosphoric-acid-treated glass wool or chemically bonded packings
at cryogenic temperatures.

     Very recently a problem concerning irreversible adsorption
of certain types of organic molecules by various porous polymers
has received considerable attention  by analysts and environmental
scientists.  Therefore, to avoid any difficulties with the very
low levels of residual adsorptivity, the Tenax GC trap is
employed for the thermal decomposition studies in which the
concentrations that go onto the  trap are relatively large and
the interest concerning low-level trace components is relatively
minor.  When the objective is to determine the trace level
remnants of the unreacted parent sample after exposure to very
high  temperatures, a glass wool  trap or a chemically bonded
packing trap is used along with  cryogenic trapping temperatures.


TDU INSTRUMENTATION

      Most of the measurements pertaining  to  the  thermal
decomposition unit have been instrumented, and the various  con-
trol  and measurement devices* have  been  incorporated into an
instrumentation console.   A photograph'of  this instrumentation
console is presented in Figure  29,  along with  the  descriptive;
designations of the modules that measure  or  control  the  various
operations performed by the TDU.  These  modules  are  mounted in
a Cabtron Series  C modular electronic  cabinet.                ,,-.   .

      The temperature programming function for  both the  sample: •'
 insertion process and  the desorption of  the  collection  trap is
 accomplished with the  use of  a  Theall  Engineering  TP-2000 _
 temperature  programmer.   This  device is  capable  of generating'
 linear temperature programs  from minus 100°C to  plus 700°C.  •
A switch has been installed  in  this module so that power can be•
 *calibration data are presented in Appendix D,

                                 50

-------
GLASS-LINED'..as.  TUBING
 /INSULATED  S  WRAPPED\
 V WITH  HEATING TAPE '  /
IN-LINE  TRAPPING  MEDIA
3i_ [—1 KNN\\y^ , — | i L.UW
T 	 	 	 -^==3-lJ-T~
FLOW FROM: [
THERMAL i
DESTRUCTION
UNIT






N« V F M T -<- f3a
p V L. 1 M 1 ^ txjr -•
jJ ^^^'t
1 S\\ \\\ X\
' t
ALUMINUM
BLOCK ,
a :
CARTRIDGE]
HEATER 1
1
1
1
1
1
• INSULATED;
LINES !
1

LIQUID — -U
,J - - ^CHROM/
k N2GAS
INLET
(N2-!60°C)






v,
*i
r^ ^ N? GAS
SUPPLY

N2 ^^
Trapping Mode, valve V1 ls open and trap temperature approaches.- 160°C.
.Desorb.Mode: Valve V1 is closed and trap is subjected to a temperature
1 ' rate"™ °f "16°° tO +3°°°C at a Pre-selected programming
     Figure 27.   Schematic of collection trap assembly.
                             51

-------
             GLASS LINED S.S. TUBING
         0.16
                      TENAX-GC

                    (8tng, 8O/IOOMESH)
                                            .  GLASS WOOL
                                            (S1LANE TREATED)
0,31
 U-r- 1.5 —J   1-0  I*—'1.5 —*


	   18.9    	'-.	-	
                                                             7.4
         TCHMENSIONS ARE IN CENTIMETERS|
                Figure 28.   Adsorption  trap -irisert,
                                    52

-------
                     TEMPERATURE PROGRAMMING
                       •SAMPLE INSERTION
                       •.PRODUCT DESORPTION
                     PRESSURE AND FLOW READ OUT
                     PROBE TEMPERATURE/TIME CONTROL
                     TEMPERATURE MEASUREMENT
                     GAS FLOW CONTROL
 f  **
*•'•««   • •   • »-<
  m m   w_4   * -m
                     TRANSPORT TEMPERATURE CONTROL
Figure 29.  TDU Console and instrumention modules.
                 53

-------
applied to either the sample insertion chamber or the product
collection trap.

     The pressure and flow readout module consists essentially
of a Hewlett-Packard 3465A digital multimeter.  The signal
output from the mass flow transducer is in the range of volts
DC, whereas the output from the differential pressure trans-
ducer is read out in millivolts DC.

     Certain low volatility samples can be inserted into the
TDU with the use of an electrically heated probe.  A modified
Chemical Data Systems Pyroprobe 100 solids pyrolyzer_is used to
perform this sample insertion and subsequent volatilization.  A
number of sample insertion techniques have been developed using
this type of programmed heated probe.  This pyrolyzer can also
be used to study nonvolatile materials such as polymers.

     The temperature measurement module consists of 18
electrically shielded thermocouple inputs, a thermocouple
selector switch  (Omega MTG-OSW3-18), and two digital temperature
indicators (Omega 175-KCl and Omega 250-KC1-05).  The average
temperature of the quartz tube reactor is continually monitored
by the Omega 250-KC1-05 instrument.

     The gas flow control module contains the gas input ,   •'.•:.
connections, a carrier gas filter  (Applied Science Laboratory's
Hydro-Purge), a gas switching valve  (Carle Instruments'. 5511).,
needle valves that control the flow of heat transfer air in the
high-temperature reactor assembly, the variable flow controller
 (Porter Instruments' VCD 1000) that controls the actual gas
flow to the reactor, a switch for the solenoid valve that
controls the nitrogen gas flow in the cryogenic trapping unit,
and the flow and pressure transducers along with their
associated electronics.

     Because of  its direct interaction with the high-temperature
reactor, the gas flow control module is a very crucial component,
A photograph of  the interior of this isolated module is
presented in Figure 30.   (Notice the positioning of the mass
flow and differential pressure sensing elements relative to the
thermometer and  the mounted circular level.)

     The principle behind the operation of the mass flow meter
is well established  (113,114).  Although there is'a number_of
different methods for measuring in-line gas flow  (115), this_
particular flow  measurement device has performed quite well in
laboratory studies where accurate measurements of in-line gas
flow were required  (116) .  A schematic showing the principle  of
operation of the transducer and the  simplified associated
electronic circuitry is shown in Figure 31.   The mass flow
transducer is position sensitive.  Therefore, a circular level
was mounted on  the same flat base  that supports the mass flow

                               54

-------
      MASS FLOW
      TRANSDUCER
                    DIFFERENTIAL
                      PRESSURE
                     TRANSDUCER
Figure 30. Interior of gas flow control module.
                   55

-------
                                  f  TO
                                   CONSTANT
                                       1Y

                                   HEATER
            FLOW-
             TO
          EXCITATION"
          SUPPLY *l
          RTD.,:  RESISTANCE-TEMP
            '1,2
                    DETECTOR
          R3>4  :  STABLE RESISTORS
       TO
    EXCITATION
    SUPPLY # 2

      ,.(•*-) .'  ,
                                                            0-5V OUT
* \ FLOW
 v  ' INDICATION
Figure  31.   Fundamenetai schematic  of mass  flow transducer;
                                    56

-------
 sensing element, thus providing the means of establishing the
 required perpendicularity of the device.

      The differential pressure transducer used in this module is
 somewhat similar to the device used in our earlier work with
 the_ first generation system.  However., a  number of circuitry
 revisions have been made that resulted in improved signal
 stability.   A schematic drawing of the transducer used in this
 module is shown in Figure 32.

      Both of these transducers have a fixed working range.  The
 mass  flow transducer is calibrated to handle flows of air up to
 50  cm-5 min L,  and the differential pressure transducer is
 designed for use up to 2.0 absolute atmospheres.   Each of these
 sensitive components has a built-in safety factor that allows
 them  to handle somewhat higher gas flows  and pressures.   However,
 the recommended upper values are,as stated above.   If TDAS work
 is  anticipated at flows and pressure beyond these limits,
 different transducers would have to be substituted.

      Two additional measurements are routinely made  in the
 course of conducting a test with the TDAS.  ;One of these
 measurements is ambient temperature.   This  temperature
 measurement  is made with a mercury-in-glass -thermometer,  which
 is  located as  shown previously in Figure  30.:   The  other  important
 measurement  is-that of barometric, pressure,  and this  measurement
 is  taken using a Sargent-Welch S-4565  vernier rack barometer,
 which reads  the height of the mercury, column  to the nearest  0.1
 mm.                 '              ;     ....-.,.

      The transport temperature control module consists of a
 number of different variacs,  adjustable power supplies,
 temperature  controllers,  and assorted  fuses and switches,  which
 are used for heating the  transport lines, gas switching valve
 chamber,  and the GC inlet block.   These power supplies have
 adjustable outputs,  and the  temperatures  of the various heated
 components are  routinely  measured  using the temperature
 measurement  module.   The  power lines for  the  various  heaters are
 sufficiently separated  from  the  shielded  thermocouples and the
 temperature  measurement module.
     There are
tation modules
advantageous in
to have at one
test parameters
modular concept
to the TDAS can
several advantages to housing these instrumen-
in one compact console.  First, it is particularly
 conducting detailed experiments with the TDAS
location the controls and readouts of the various
   In addition, by employing the versatile
 for the TDU instrumentation, future modifications
 be readily accommodated.
                               57

-------
 SOLA 81-05-230-1
CONTROLLED
  POWER  .
  SUPPLY

  (5.0VDC)
                 500ft
STATHAM STRAIN GAGE
DIFFERENTIAL PRESSURE
TRANSDUCER
PL-295TC-I5-350
                                               DIGITAL
                                            * VOLTMETER
                                            •  (mv)
   Figure 32.   Pressure transducer schematic.
                           58

-------
                           SECTION  7

            CHROMATOGRAPHIC SEPARATION AND ANALYSIS
     The TDAS concept relies  strongly on  the  analysis of  the
chemical compounds that emerge from  the high-temperature
reactor.  This analysis establishes  the extent of  thermal
decomposition of the parent substance and at  the same time
provides information on the various  thermal reaction products.

     The seven stages of a typical instrumental chemical
analysis are as follows:                .            .

          1.  ; Collection             .      ,..       •.••>•
          2.  Storage                :
          3.  Extraction       ,         ,;.  <
          4.  Concentration           , =.
          ,5.......Isolation
          6.  Identification
         i 7.  Quantification

Now the primary function of the GC located in the  TDAS  is to
separate the constituents of  the various  complicated chemical
mixtures.  In addition, this  GC unit is also  responsible  for
stages 4 and 7 .of the above analysis scheme.

     The GC installation within the  TDAS  is shown  schematically
in Figure 33.  The glass open tubular column  (see  Appendix E) and
the bypass tubing are located in the interior of a modified Pye
104 gas chromatograph (Figure 34).   This  particular instrument
is part of the LKB-2091 GCMS-COMP system.  A  high-temperature
gas switching valve has been  attached to  the  side  of this GC,
and the interior of the heated valve chamber  is maintained at a
constant elevated temperature.

     Water and volatile acids can present difficulties  to most
glass open tubular columns.   Therefore, a sample bypass feature
has been incorporated into this installation  so that harmful low
molecular weight species bypass the  open  tubular column (OTC)
and pass on to the mass spectrometer (MS), where they can be
continually monitored.  This  bypass  mode  can  also  be used for
MS analysis of the low molecular weight products.  The columns
that have been prepared and used in  the TDAS  can accept small
quantities of water without adversely affecting the performance
                                59

-------
                                GC COLUMN OVEN
  HEATED
'JUNCTION
  AUXILIARY
   HELIUM

     TRAP.
   MASS
SPECTROMETER
§£ n
RESTRICJOR


s,.;.
DETECTOR
(TIC)
RESTRICTOR /
. = /
— • ^nnnrinri ~ •f-*-ii
HR-OTC L-H
CARBON -^$1

^-^
^f
!
'
MOLECULAR ; ^
SIEVE TRAP ' I
       '   HEATED
       J   VALVE
         CHAMBER

        CHARCOAL
         TRAPS

        CAPILLARY
       RESTRICTORS
        (VARIABLE)
                                                    VENTS
      Figure  33.   Schematic of GC  column installation.
                                 60

-------

            SOT^,
Figure 34.  Glass OTC installation in modified Pye 104
                      61

-------
of the column.  Even so, if possible, it is desirable to keep
oxygen, water, and various harmful substances from entering the
glass capillary.

     As mentioned earlier, another vital function of the TDAS
chromatographic column is to trap the condensable products that
are liberated from the effluent collection trap during the _
thermal desorption process.  A chromatographic column can indeed
be an excellent trap, and if held at room temperature, it can
quantitatively capture most substances having a molecular weight
of 100 or greater.  If the column oven is maintained at sub- ;
ambient temperatures with the use of auxiliary coolant (e.g.,
dry ice, etc.), much lower molecular weight species can be
successfully trapped  (117-119).  This thermal focusing process
(see Appendix P) is a crucial step in the sample handling with
the TDAS.  Also, for the analysis of products from the TDAS,
programmed temperature GC is essential.
          ,               '                                      i
     The TDAS depends on high-resolution separation of the
various chemical compounds.  Therefore, this necessitates  the
use of glass capillary GC columns.

     The gas-liquid chromatographic  separation of volatile
organic compounds was first accomplished in 1951  (120) , and  ,
OTC's  (121-124) were  introduced  in 1958.  The latter  continue
to receive the  attention of many researchers.  The various
mechanisms that influence chromatographic separation  have  been
placed on a firm technical foundation  (125), and  today it  is not
uncommon to find a  sample containing more than a  hundred
different compounds being adequately separated for qualitative
and quantitative analysis.  Research work with glass  OTC's is
relatively recent,  although glass capillaries were introduced
as early as 1960.   Much of the recent  interest in glass OTC's
has come about  through  increased attention  being  devoted  to  .
surface chemistry and physical compatibility with ,the various
stationary liquid phases.  A  bibliography has been prepared_
 (126)  concerning  some of the  more recent developments in  this
field of open tubular gas  chromatographic  columns.  Also,  a  text
on glass OTC's  has  recently been published  (127.).           .

      The various  columns used in the TDAS,  and  those  used in ,the
other supporting  GC's,  are fabricated  in our  laboratory.
Appendix E describes the procedures  used, for  preparing these
OTC's.   Our  OTC preparation procedures have been influenced
collectively by the works of  many researchers  (128-137)  in the
 field of capillary columns.   The columns  that are used_in the
TDAS are  first subjected to  a performance  examination in a
 separate GC  instrument.  Figures 35, 36,  and 37  show chromato-
 grams obtained by such columns.   Figure 35 indicates the
 separating efficiency for a series  of  normal alkanes; Figure 36
 indicates the tremendous complexity.of some samples  (e.g., a
 distillation fraction of crude oil), and Figure 37 shows the
                                62

-------
                                               GAS CHROMATOGRAPHlb CONDITIONS
                                              Instrument
                                              Analyst	
                                              Date
                                                          Varian 1860-1
                                     W.  Rubey
                                                        5-23-79
                                              'Column:
                                               tubing material  ". '  Pyrex-'
                                               tubing length       15 .
                                               tubing bore '        0. 25
                                                    -meters
                                              Stationary Phase   SE-30 Silicons
                                               weight	.  -    ,.    .,'•',''  '
                                              Suppojt       	
                                               mesh
                                              Carrier Gas_
                                               inlet pres
                                                           Helium
                                           , 6
                                               linear velocity_
                                               outlet flow	^
                                             Chart Advance
                                             Read Out
                                             Sample	
                                               size
                                             30
                                         •\-l.l
abs atmo *
_cm/sec •.
                                          0.25
                                                     cm/min
                                         1.. 0
                                                 my full scale
                                        normal  alkane  mixture
                                                       0.5  ul
                                             Detector_
                                               range	
                                                         HFID
                                           AFS.
                                               attenuation
                                                            64
                                             Gas flows, ml/min
                                              hydrogen	30
                                              air
                                              split ratio_^	
                                             Temperatures, °C
                                              detector      300
                                              injector     • -200
                                              column:
                                                initial       35
                                                final     '..  220-
                                                            300
                                                            40 to  1
                                               program rate_
                                             Sample Solvent	
                                              concentration
                                         n-Hexane
                                                  °C/min
                                                                  98
                                                                       solvent
                             T
                                       T
30   ...  .      ,.",,             60


         TIME,, minutes ....
                                                                             ,90
Figure  35.    Glass  GTC separation of  normal  alkanes;
                                   63

-------
                                   .  s
                                               CU
                                               X
                                              •H
                                               g
                                               (1)
                                              •H
                                               o
                                               o
                                               (0
                                               IH
                                               o
                                               o
                                               •r-\
                                               -P
                                               (0
                                               
-------
                     uJuntojLj[jUt_Jvj
                                                    GAS CHROMATOGRAPHIC  CONDITIONS
                                                  Instrument  Varian 1860-1
                                                  Analyst  W.  Rubey
                                                  Date   5-25-79
                                                  Column:
                                                   tubing material	Pyfex
                                                   tubing length   15	
                                                   tubing bore	Q . 25
                        meters
                        mm
                                                  Stationary Phase SP-210Q Silicone
                                                   weight	~                 ~
                                                  Suppo rt   •  -	.
                                                   mesh
                                                  Carrier Gas_
                                                   inlet pres
                                                              Helium
            1.6
                                                   linear velocity_
                                                   outlet flow
                25
                                                                0.25
                                                                1.0
abs atmo
cm/sec
ml/min
cm/min
Chart Advance
Read Out	'   	
Sample	Industrial Waste Mixture
 size
                                                                       mv full scale
                                                          0.5 ul
                                                            HFID
                                                  Detectoi
                                                   range	; IP"1
                                                   attenuation  32
               AFS
                                                  Gas Flows, ml/min
                                                   hyd rogen   30	
                                                   air
                                                             300
                                                   split ratio
                                                              40 to 1
                                                  Temperatures, °C
                                                   d et e cto r   3QQ
                                                   inj e cto r	260
                                                   column:
                                                   initial      35
                                                   final      2~TO~~
                                                    program rate_
                                                 Sample Solvent	
                                                  concentration
                                                                       "C/min
              Benzene
              	95	i
                                                                         solvent
             20
                              40
                                                60
                               —i	r
                                100
                                    TIME,  minutes
Figure  37.    Separation of  complex chlorinated  hydrocarbon
                    mixture.
                                           65

-------
separating capability for a very complex mixture of chlorinated
hydrocarbons and other organic compounds.

     The columns that are presently used in the TDAS are
relatively thin film OTC's in that the stationary phase film
thickness is on the order of 0.1 microns.  This produces a
relatively low stationary phase bleed level and also permits
low temperature elution of relatively heavy compounds (138).
The one disadvantage is that these thin film OTCs have low
sample capacity.  The columns that are usually used in -the TDAS
are from 10 to 20 meters long.  However, the columns that are
used in the other GC's which support the TDAS operation, are
much longer; consequently, higher efficiency separations can be
obtained with these instruments.  Presently, the separation
capability within the TDAS is adequate for the samples currently
being tested.  In the future, other types of OTC's will also be
incorporated in the TDAS.  These will contain chemically bonded
stationary phases  (139,140) and other types of high-temperature
substrates.

     The gas flow to and from a gas chromatograph must be
especially free of extraneous organic contaminants.  _Otherwise
the chromatographic separation and subsequent analysis are
adversely influenced by these contaminants  (141).  Also, the
geometry of the flow path  should be free of mixing chambers and
unswept regions  (142).  In a system as complicated as the TDAS,
there are a number of potential sources for organic contami-
nation.  However, by using thermally stable materials throughout
the,gas flow path, the level of bleed contaminants can be
minimized.

     With the TDAS, there  is an,additional  trapping procedure
that can be invoked when it is desirable to conduct an_analysis
with a very low background level.  Specifically, this is
accomplished by cooling the cryogenic adsorptive trap, immediately
after the collected products have been thermally desorbed  from
the trap.  In short, by cooling the in-line trap to  subambient
temperatures, any  residual compounds that would normally be
migrating during the  time  period of the  GC  analysis would  now be
captured by this cryogenic trap.   These contaminants are
therefore not continuously migrating through  the GC  column
during the programmed  temperature GC analysis.  Any  condensable
contaminants that  are  admitted  to the column  before  or during
the thermal desorption of  the product collection trap are  sub-
sequently captured by the  thermal focusing  feature of the  GC':
column before the  start of the  programmed  temperature GC
analysis.  And  of  course,  these condensable contaminants, would
eventually  be chromatographed'as  typical solutes.

      In  the TDAS,  the GC  effluent is passed into an  electron-^
impact  source mass spectrometer,  which  is  described  briefly in
Section  8.  The influence  of  the  GCMS coupling on  the
                                66

-------
performance of the OTC can be negligible, providing the gas
transport between the two instruments is fully understood  (143-
145).  In reality, the MS probably has far less influence on the
behavior of the chromatographic column than the column has on
the MS (e.g., column bleeding, etc.)

     The ability to separate complex volatile mixtures is at a
relatively advanced technological state.  Even so, recent
developments indicate that this already high/resolving cap-
ability will probably be advanced even further (146-149).
Therefore, if situations arise in the, future in which greater
resolution is needed in the TDAS, it would appear that such
could be readily accomplished.
                               67

-------
                         is an
                           SECTION 8

         MASS SPECTROMETER WITH DEDICATED MINICOMPUTER
            AND ASSOCIATED DATA REDUCTION EQUIPMENT


     In the TDAS, the identification of the various thermal
decomposition products is accomplished through the use of_
in-line MS.  As shown earlier in Figures 5 and 7, the MS i
integral and vitally important component of the TDAS.  The
particular MS instrument incorporated into the TDAS is an
LKB 2091 GCMS.  A photograph of this instrument along the up-
stream TDU is shown in Figure 38.

     A combined GCMS is a very powerful analytical tool  (46,
150, 151) and is especially applicable for analyzing complex
mixtures of organic compounds.  The LKB 2091 is a single focusing
MS equipped with a 90° sector magnetic analyzer.  The MS
assembly consists of one unit, housing both analyzer and con-
trols .  There are individual high-capacity vacuum, systems _ for
the analyzer tube, the source, and the inlet systems.  This
instrument employs the Ryhage two-stage molecular separator
 (152), which is capable of operation at temperatures up  to 350°C.

     The MS incorporated into the TDAS uses an electron  impact
source, and this instrument has a mass range of 0 to 2000 at
1.17 kV acceleration voltage.  In the GC mode, the resolution
of the mass spectra is better than 700 at 3.5 kV  acceleration
voltage, and at an ionization potential of 70 eV.  This  MS can
be automatically scanned at different rates over  the full
spectrum range or parts thereof.  Parabolic scanning rates range
from 1 to  240 seconds for a full-scale scan.

     There are two detectors in the LKB 2091 GCMS.  A total ion
current detector is used for generating a chromatographic
signal that represents the ionized  effluent at the exit  of the
ion source.  The output from this detector is displayed  on a
potentiometric recorder.  The other detector is  an electron
multiplier that  is positioned at  the outlet of the magnetic
analyzer tube.   The output of the electron multiplier-
preamplifier can be displayed on  a  UV-recording  oscillograph,
or data  can be fed to the data reduction  system.   This  instru-
ment has its own mass marker  (LKB 2091-110) , which has  an
 accuracy of  ±0.3 amu  at  3.5 kV acceleration voltage.  This MS
 is also  equipped with a  direct inlet  system  (LKB 2091-210)
68

-------
Figure 38.  LKB 2091 GCMS and TDU.
                69

-------
whereby a sample can be admitted to the spectrometer through the
use of a special probe.

     The data reduction equipment that supports the GCMS is an
LKB 2130 data system.  A photograph of this equipment is shown
in Figure 39.  The data system consists of a Digital Electronics
Corporation PDP 11/04 Minicomputer.  The graphics terminal is a
Tektronix 4012, and the line printer/plotter is a Versatec 800.
Included with the data system is the LKB applications software
package and the LKB 2130-220 library.

     In addition to the above data handling components, another
terminal has been incorporated into the TDAS.  This additional
terminal is a Texas Instruments 733 ASR, which uses dual
cassettes.  This terminal functions as an external data base
interface through which searches can be conducted of the EPA-
NIH Chemical Information System  (153, 154) mass spectral library,
A photograph of this mass spectra searching terminal is shown
in Figure 40.
                                70

-------
Figure 39.  Minicomputer for data reduction.
                     71

-------
Figure 40.  Mass spectra searching terminal,
                     72

-------
                            SECTION 9

                UTILIZATION OP THE TDAS IN THERMAL
                       DECOMPOSITION STUDIES
     _The TDAS has been designed with the objective of conducting
 precise thermal decomposition experiments in the laboratory.
 This system is capable of investigating a wide variety of organic
 materials and experimentally examining the effects of the major
 thermal decomposition variables.  The TDAS can handle relatively
 pure substances and also very complex samples (e.g.,  multi-
 component industrial organic wastes).

      The most important immediate use of the TDAS is  in deter-
 mining  the thermal decomposition behavior of toxic organic
 substances.   This is a very  sensitive problem area in which
 information  is urgently needed with  respect to the tremendous
 variety of toxic organic materials.

      With the flexibility built into  the TDAS,  many types  of
 thermal decomposition experiments can be conducted.   Essentially
 the  various  types of information that are obtained from TDAS
 experiments  can be grouped into four  categories:

      1.   Establishment of the  thermal decomposition profile
          of  a substance.

      2.   Determination of residence time  effects  and  associated
          kinetic  behavior.

      3.   Investigation of different gaseous atmospheres,
          pressure  effects, and  other  related  thermal             :
          decomposition variables.

      4.   Collection of  data for prediction of thermal
          decomposition  behavior.

     _The  thermal decomposition profile of a substance is
obtained by subjecting  a  known quantity of the material to a
controlled thermal exposure.   Analyses are then conducted to
determine the amount of the unreacted parent substance and also
to identify and quantitate the various thermal decomposition
products.  By subjecting a sample to a series of thermal
exposures covering a broad experimental range, the thermal
                               73

-------
decomposition behavior of that sxi±>stance can be established.
Although the thermal decomposition behavior of the parent sub-
stance is of major importance, the detection and identification
of various thermally stable decomposition products is also of
vital concern  (12).  There are many instances where an organic
substance decomposes and forms other chemical species of equal
or greater toxicity  (155).  The major objective in obtaining a
thermal decomposition profile is to determine the thermal
exposure conditions that are necessary to produce essentially
complete destruction of the parent substance and its various
products.  This profile also shows the thermal regions where
toxic products are formed  (12,156).  such thermal decomposition
data are of considerable value in providing guidance for large-
scale thermal disposal operations.  The thermal decomposition
profile of a substance can form the basis for establishing its
safe incineration conditions  (15,45).  This same type of TDAS
experiment can be conducted with very complex industrial wastes,
only here the objective is simply to determine at what thermal
conditions no toxic  substances are emerging in the effluent.

     The effect of residence time on the extent of thermal
decomposition can be readily measured with the TDAS.  This
system can also be used for a variety of reaction kinetic
studies, as the reactor incorporated into the TDU is ideally
suited for determining residence time and kinetic effects  (12,
41).  Some excellent data  have been obtained for lower molecular
weight organic substances  using a similar laboratory reactor
 (43,44).       -

     The third group of experiments  encompasses studies  in
which the thermal decomposition behavior of a substance  is
evaluated while  it is in different 'gaseous environments.  Also,
studies can be conducted where the macroscopic reaction  pressure
is varied to determine the effect of pressure on the thermal
reaction.

     The fourth  experimental  area in which the TDAS can  have
broad application is in  the prediction  of the thermal_decom-
position behavior of different organic  molecules.  _This  is  a
promising experimental area of considerable.potential.   From
laboratory data  generated  thus far  (12,41,43),  there  seems  to be
a good  possibility that  eventually  thermal decomposition be-
havior  can be  predicted  (at  least in part) by  simply  knowing
the molecular  structure  of a  substance  (157)-.   It  is  anticipated
that by conducting special TDAS  experiments  aimed  at  collecting
 thermal decomposition data on different chemical  families or
bond structures, some level  of predictability, will emerge for
 organic compounds of similar  structure.  The ability  to  predict
 the thermal  decomposition behavior  of  a waste  substance  strictly
by knowing  its molecular structure  would be  of  considerable
practical value.
                                74

-------
1.
                     REFERENCES


Paulson, E. G.  How to Get Rid of Toxic Organics.
Eng./Deskbook Issue, Oct. 1977, p. 21.
                                                       Chem.
    Bell,  B.  A.,  Welday,  J.,  and Gilley,  W.   Alternate  Methods
    for the Disposal of the Pesticide Kepone.  Paper  presented
    at the Third  Annual Conference on Treatment and Disposal of
    Industrial Wastewaters and Residues.   April 1978, Houston,
    TX.

    Kennedy,  M. V.,  Stojanovic,  B. J.,  and Shuman, F. L., Jr.
    Chemical  and  Thermal  Aspects of Pesticide Disposal.  J.
    Environ.  Qual.,  1:63, 1972.                          ~

    Bailin, L. J., Hertzler,  B.  L., and Oberacker, D. A.
    Development of Microwave  Plasma Detoxification Process for
    Hazardous Wastes-Part 1.   Environ.  Sci.  Tech., 12:673,
    1978.                      ~	

    Oberacker, D. A.,  and Lees,  S.  Microwave Plasma  Detoxifi-
    cation of Hazardous and Toxic Materials.   NERC Topic,
    Oct.  1977.                                	
    Molten Salt Decomposes  Pesticide Wastes.
    News,  Sept. 1977,  p.  44.
                                          Chem.  and Eng.
    Wastler,  T.  A.,  Offutt,  C.  K., Fitzsimmons, C. K., and
    Des  Rosiers, P.  E.   Disposal  of Organochlorine Wastes by
    .Incineration at  Sea.  EPA-430/9-75-014, U.S. Environmental
    Protection Agency, Washington, DC,  1975.

   .Paige,  S. F.,  Baboolal,  L.  B., Fisher, H. J., Scheyer,
    K. H. ,  Shaug,  A. M. , Tan., R.  L. , and Thorne, C. F .
    Environmental  Assessment:   At-Sea and Land-Based Inciner-
    ation of  Organochlorine  Wastes.  EPA-600/2-78-087, U.S.
    Environmental  Protection Agency, Research Triangle Park,
    NC,  1978.

    Shih, C.  C.,  Tobias, R.  P., Clausen, J. F., and Johnson,
    R. J.   Thermal Degradation  of Military Standard Pesticide
    Formulations,  Contract  No. DADA 17-73-C-3132, U.S. Depart-
    ment of the  Army, Washington, DC, 1975.
                             75

-------
10,


11,



12,
Waste Disposal by Burning at Sea.
1979, p. 44.
Chem. Week, April 11,
13,
14,
15.



16.


17.



18.



19.



20,


21,



22,
Benbow, A. W., and Cullis, C. F.  The Formation of Toxic
Products During the Combustion of Halogen-Containing Poly-
mers.  European Polym. J.  11:723, 1975.

Duvall, D. S., and Rubey, W. A.  Laboratory Evaluation of
High Temperature Destruction of Kepone and Related Pesti-
cides.  EPA-600/2-76-299, U.S. Environmental Protection
Agency, Cincinnati, OH, 1976.           •

Rolke, R. W., Hawthorne, R. D., Garbett, C. R., Slater,
E. R.f Phillips, T. T., and Towell, G. D.  Afterburner
System Study.  EPA-R2-72-062, U.S. Environmental Protection
Agency, Research Triangle Park, NC, 1972.       ,

Barrett, R. E.  Environmental Assessment of Afterburner
Combustion Systems.  In: Proceedings of the Second
Stationary Source Combustion Symposium, p. 107.  Vol.. 111.
EPA-600/7-77-073c, U.S. Environmental Protection Agency,
Research Triangle Park, NC, 1977.

Hazardous Waste:  Proposed Guidelines and Regulations and
Proposal on Identification and Listing.  Federal Register,
43, Dec. 1978.

Girard, L., and Harkness, A. C.  Method for Analyzing the
Effluent from a Microreactor.  J. of G.C.  5:516, 1967.

Keulemans, A. I. M., and Perry, S. G.  Identification of
Hydrocarbons by Thermal Cracking.  In:  Gas Chromatography
1962.  Van Swaay, M., ed. Butterworth, London, 1962, p. 356.

Kokes, R. J., Tobin, H. , Jr., and Emmett, P, H,  New
Microcatalytic-Chromatographic Technique for Studying
Catalytic Reactions.  J. Am^ Chem. Soc. 77: 5860, 1955.

Munday, C. W., and Primavesi, G. R.  A Modular Gas-
Chromatography System for Analyzing Exit Streams from
Reactors.  Analyst 88:551, 1963.

Chiu, J. Polymer Characterization by Coupled Thermogravi-
metry-Gas Chromatography.  Anal. Chem. 40:1516, 1968.

Chang, T., and Mead/ T. E, Tandem Thermogravimetric
Analyzer-Gas Chromatograph-High Resolution Mass Spectro-
meter System.  Anal. Chem. 43:534, 1971.

Shen, J. In-Situ Sampling System for Thermal Volatilization
Analysis of Nonvolatile Materials by Gas Chromatography-
Mass Spectrometry,  Anal. Chem. 49:886, 1977.
                               76

-------
 23.
28.
29.
30.
31.
32.
33.
34.
 Johnson,  D.  R.,  and Smith,  J.  W.   Bureau  of Mines Apparatus
 for Thermal  Analysis:   Simultaneous  DTA-TG-EGA  for Oil
 Shale and Solid  Fuels.   Report of  Investigations 7429, U.S.
 Department of  Interior,  Bureau of  Mines,  Washington, DC,
 1970.
 24.


 25.



 26.



 27.
 Lodding,  W.,  ed.   Gas  Effluent Analysis.
 Inc.,  New York, NY,  1967.
Marcel Dekker,
 Ayres,  W. M.,  and Bens, E. W.  Differential Thermal Studies
 with Simultaneous Gas Evolution Profiles.  Anal. Chem.
 33:568,  1967.                              	

 Wendlandt, W.  W. A New Apparatus for Simultaneous Differen-
 tial Thermal Analysis and Gas Evolution Analysis.  Anal.
 Chem. Acta 27:309, 1962.                           —~—

 Norem,  S. D.   A Combustion Device for Use in Conjunction
 with Chromatographic Columns.  In:  Proceedings of First
 ISA  Symposium  on Gas Chromatography.  Academic Press
 New  York, NY,  1958, p. 191.

 Steingaszner,  P.  Microreaction Gas Chromatographic
 Techniques.  In:  Ancillary Techniques of Gas Chromato-
 graphy,  Ettre, L. S., and McFadden, W. it., eds. Wiley-
 Interscience,  New York, NY, 1969, p. 11.

 Kwart, H., Sarner, S. F., and Olson, J. H.  Homogeneous
 Gas-Phase Thermolysis Kinetics.  An Improved Flow Technique
 for  Direct Study of Rate Processes in the Gas Phase.  J.
 Phys. Chem. 73:4056, 1969.                            ~

 McCarter, R. J.  Apparatus for Rate Studies of Vapor Pro-
 ducing Reactions.  In:  Proceedings of a Symposium on the
 Current  Status of a Thermal Analysis.  National Bureau of
 Standards, Washington, DC, April, 1970.     '    .   .

 Iglauer, N., and Bentley, F. F.  Pyrolysis GLC for the
 Rapid Identification of Organic Polymers.  J.  Chrbm. Sci.
 12:23, 1974.                               	:	;	

 Derby, J. V., and Freedman,  R. W.  Vapor-Phase Pyrolysis
 and GC Analysis of Fire-Retardant Materials Containing PVC.
Amer. Lab. May 1974.

 Cramers, C.   Pyrolysis Gas Chromatography of Volatile
 Hydrocarbons.  In:   Gas Chromatography 1968.   Elsevier Pub.,
Amsterdam, Netherlands,  1969,  p.  395.

 Uden, P. C.,  Henderson,  D.  E., and Lloyd, R.  J.   Compre-
 hensive Interfaced Pyrolysis Gas  Chromatographic Peak
 Identification System.   J.  Chromatog. 126:225,  1976.
                              77

-------
35.
36.
 37
. 38
 39
 40
 41,
 42,
 43
Uden, P. C., Henderson, D. E., and Lloyd, R. J.  An Inter-
faced Vapor Phase Instrumental System for Thermal Analysis
and Pyrolysis.  In:  Proceedings of the Third International
Symposium on Analytical Pyrolysis, Amsterdam, Elsevier
Scientific, Amsterdam, Netherlands, Sept. 1976.

Rubey, W. A.  The Measurement of Concentration Zone Profile
in Elution Chromatography.  Technical Report, UDRI-TR-70-45
University of Dayton Research Institute,  Dayton,  OH,
Dec. 1970.

Rubey, W. A.  Design of a System for Simultaneous Thermo-
gravimetric-Gas Chromatographic Analysis of  Stabilized
Polyacrylonitrile Fiber.  Technical Report,  UDRI-TR-71-28
University  of Dayton Research Institute, Dayton, OH,
Sept.  1971.

Rubey, W. A., and Duvall, D. S.  Description of Apparatus
for Conducting Isothermal Aging-Evolved Gas  Analysis
Tests  with  High Temperature  Polymers.  Technical Report,
UDRI-TR-75-01, University of Dayton Research Institute,
Dayton,  OH,  Jan. 1975.

Duvall,  D.  S., Rubey, W. A., and Carnes, R.  A.  High_
Temperature Destruction of Kepone  and Related  Pesticides.
Report presented at 173rd American Chemical  Society
National Meeting, New Orleans,  LA, March 1977.

Carnes,  R.  A., Duvall, D. S., and  Rubey, W.  A.  A  Laboratory
Approach to Thermal Degradation of Organic  Compounds.
In:  Proceedings of 70th  Annual Meeting  of  Air Pollution
Control Association, Toronto, Ontario, June 1977.

Duvall,  D.  S., and  Rubey, W. A.  Laboratory Evaluation  of
High-Temperature Destruction of Polychlorinated Biphenyls
and Related Compounds.  EPA-600/2-77-228, U.S. Environ-
mental Protection Agency, Cincinnati, OH, Dec. 1977.

Duvall,  D.  S., Rubey,  W,  A., and  Mescher, J. A.   Laboratory
Characterization of the Thermal Decomposition of  Hazardous
Wastes.  In:  Proceedings of SWRI-EPA Conference  on Dis-
posal of Hazardous  Wastes,   EPA-600/9-78-016,  U.S. Environ-
mental Protection Agency, Cincinnati, OH,  1978.

Lee, K., Jahnes,  H. J.,  and Macauley, D.  C.  Thermal
 Oxidation Kinetics  of Selected  Organic  Compounds.   In:  Pro-
 ceedings of 71st Annual Meeting of the  Air  Pollution Con-
 trol Association,  Houston,  TX,  June 1978.
                                 78

-------
 44,
 45.



 46.



 47.


 48.
49.



50.



51.


52.


53.



54.


55.


56.
 Lee, K., Hansen, J. L., and Macauley, D..S.  Predictive
•-,Model of the Time-Temperature Requirements for Thermal
 Destruction of Dilute Organic Vapors.  In:'  Proceedings of
 72nd Annual Meeting of Air Pollution Control Association,
 Cincinnati, OH, June 1979.

 Bell, B. A., and Whitmore, F. C.  Kepone Incineration
 Test Program.  EPA-600/2-78-108, U.S. Environmental Pro- .
 tection Agency, Cincinnati, OH, 1978.

 Leathard, D. A., and Shurlock, B. C.  Identification Tech-
 niques in Gas Chromatography.  Wiley-Interscience, New
 York, NY, 1970. ,                                       ..

 McFadden, W.H. Introduction of Gas-Chromatographic Samples
 to a Mass Spectrometer..  Sep. Sci. 1:736, 1966.

 Giger, W. , Reinhard, M., Schaffner, C., and Zurcher, F.
 Analyses of Organic Constituents in Water by High-Resolution
 Gas Chromatography in Combination with Specific Detection
 and_Computer-Assisted Mass Spectrometry.  .In:   Identifi-.
 cation and Analyses of Organic Pollutants,in Water.  Keith,
 L. H., ed. Ann Arbor Press, Ann Arbor, MI,  1977,  p. 433.

 Gudzinowicz, B. J., Gudzinowicz,  M. J.,'and Martin, /H. F.
 Fundamentals of Integrated GC-MS.  Marcel Dekker, New York,
 NY, 1976.                                          ......

 Kirkland, J. J.  Residue Analysis by Gas  Chromatography.
 Paper presented at ACS Analytical Summer  Symposium,
 Houston,  TX, June,  1960.

 Harris,  W.  E.,  and Habgood, H., W.,  Programmed  Temperature
 Gas Chromatography.  Wiley, New York,  NY, 1966.

 Westhaver,  J. W.   Theory of Open-Tube  Distillation Columns.
 Ind.  Eng. Chem. ,  Jan.  1942, p.  126.  ,               ,.-.,

 Westhaver,  J. W.   Concentration of  Potassium 39 by Counter-
 current  Electro-migration:   Some  Theoretical Aspects of the
 Operation.   J.  Res. NBS  38:169, 1947.

 Taylor,  G.   Dispersion of  Soluble Matter  in Solvent Flowing
 Slowly Through  a  Tube.   Proc.  Roy.  Soc. A 219:186,  1953.

 Aris,  R.   On the  Dispersion of  a  Solute in  a Fluid Flowing.
 Through  a Tube.   Proc. Roy.  Soc.  A  235:67,  1956.    .

 Golay, M.J.E.   Theory  of Chromatography in.Open and Coated
 Tubular Columns with Round  and  Rectangular  Cross-Sections.
 In:   Gas  Chromatography  1958.   Desty, D.  H., ed.  Butter-
 worths, London, 1958,  p. 36.
                              79

-------
57.


58.



59.




60.


61.
62.


63.


64.



65.



66.



67.


68.


69.



70.
Bode, H.  Transfer Behavior of Narrow Bore Tubes  under
Laminar Flow.  Chromatographia 3:113, 1970.

Giddings, J. C., and Seager, S. L.  Rapid Determination of
Gaseous Diffusion Coefficients by Means of Gas Chromato-
graphic Apparatus.  J. Chem. Phys. 33:1579, 1960.

Fuller, E. N., Ensley, K., and Giddings, J. C.  Diffusion
of Halogenated Hydrocarbons in Helium:  The Effect of
Structure on Collision Cross Sections.  J. Phys.  Chem.
73:3679, 1969.

Sutton, R., and Harris, W. E.  Pyrolysis Gas Chromatography
of Some Hydrocarbons.  Can. J. Chem. 45:2913, 1967.

Carter, M. K., Moore, M., and Nadel, F.  High Efficiency
Oxidation Pyrolysis Tube for Analytical Determination of
Trace Elements.  Paper presented at 1974 Pittsburg
Conference on Analytical Chemistry and Applied Spectro-
scopy.

Dean, W. R.  Note on  the Motion of Fluid in a Curved Pipe.
Phil. Mag. S7 4:208,  1927.

Dean, W. R.  The Stream-line Motion of Fluid in a Curved
Pipe.  Phil. Mag. S7  5:673, 1928.

Koutsky, J. A., and Adler,  R.  J.  Minimization of Axial
Dispersion by Use of  Secondary Flow  in Helical Tubes.
Can. J.  Chem. Eng. 42:239,  1964.

Tijssen, R.  Effect of Column-Coiling on  the Dispersion
of  Solutes  in Gas Chromatography.  Part  I:  Theory.
Chromatographia 3:525, 1970.

Tijssen, R., and Wittebrood,  R.  T.   Effect of Column-
Coiling on  the  Dispersion of  Solutes  in  Gas Chromatography.
Part II:  General Theory.   Chromatographia 5:286,  1972.
   %
Janssen, L.A.M. Axial Dispersion in Laminar  Flow  Through
Coiled  Tubes.   Chem.  Eng.  Sex. 31:215,  1976.

Keulegan,  G. H.,  and  Beij, K.  H.   Pressure Losses  for
Fluid Flow in  Curved  Pipes.  J.  Res'.  NBS  18:89,  1937.

Barr, W. E., and  Anhorn,  V. J.  Scientific-and Industrial
Glass  Blowing  and Laboratory Techniques.   Instruments
Publishing Co., Pittsburgh, PA,  1949, p.  198.
      Rasmussen,  R.  A.   A Quantitative  Cryogenic  Sampler.
      Lab. July 1972,  p. 19.
                                                     Amer.
                               80

-------
71.  Brenner, N., and Ettre, L. S.  Condensing System for Deter-
     mination of Trace Impurities in Gases by Gas Chromatography,
     Anal. Chem. 31:1815, 1959.

72.  Drozd, J., Novak, J., and Rijks, J. A.  Quantitative and
     Qualitative Head-space Gas Analysis of PPB Amounts of
     Hydrocarbons in Water.  J.. Chroma tog. 158:471, 1978.

73.  Tyson, B. J., and Carle, G. C.  Gas Chromatographic
     Separation of Water from Cryogenically Collected Air
     Samples-Enhanced Concentration of Trace Organic Volatiles
     Prior to GC-MS Analysis.  Anal. Chem. 46:610, 1974.

74.  Des Marais, D. J.  Variable-Temperature Cryogenic Trap
     for the Separation of Gas Mixtures.  Anal. Chem. 50:1405,
     1978.                                	
75.
76.
77.
78.
79.
80.
81.
82.
83.
 Aue, W. A.,,Vogt, C. R. H.,  and Younker, D.  R.  A Gas
 Chromatographic Cartridge Desorption Port.   J. Chromatog.
 114:184, 1975.              •                	"	

 Nawar, W. W.,  Sawyer, F. M., Beltran, E. G., and Pagerson,
 I. S.  An Injection System for Gas Chromatography.   Anal.
 Chem. 32:1535, 1960.                    ' .           —	

 Russell, J.  W.  Analysis of  Air Pollutants  Using Sampling
 Tubes and Gas  Chromatography.   Environ. Sci. Tech.  9:1175,
 1975.                            ;—:	 .

 Russell, J.  W., and Shadoff, L. A.  The Sampling and Deter-
 mination of  Halocarbons in Ambient Air Using Concentration
 on Porous Polymer.  J. Chromatog.  134:375,  1977.

 Hopkins, B.  J., and Pretorius, V.   Rapid Evaporation of
 Condensed Gas  Chromatographic  Fractions. J. Chromatog.
 158:465, 1978.                           	T
 '         •                      »                '      ' •
.Kaiser, R. E.   Temperature Gradient Chromatography:   Part
 1.  Reversion  Gas Chromatography.   Chromatographia  1:199,
 1968.                              	

 Kaiser, R. E.   Ultra Trace Analysis of Volatile Substances
 with GC Methods in the Sub ppm and ppb Range Down to 10~10%.
 J. Chromatog.  Sci. 9:227,  1971.

.Kaiser, R. E.   Enriching Volatile  Compounds  by a Temperature
 Gradient Tube.  Anal. Chem.  45:965, 1973.

 Kaiser, R. E.   Gas Chromatography  in Environmental Analysis.
.;J. Chromatog.  Sci. 12:36,  1974.
                              81

-------
84.
85
86
87
88.
89
90
91.
92.
93.
94.
Feairheller, W., Mara, P. J., Harris, D.  H.,  and Harris,
D. L.  Technical Manual for Process Sampling  Strategies
for Organic Materials.  EPA-600/2-76-122, U.S.  Environ-
mental Protection Agency, Research Triangle Park, NC,  1976.
                                              *

Snyder, A. D., Hodgson, F. N., Kemmer, M. A., and
McKendree, J. R.  Utility of Solid Sorbents for Sampling
Organic Emissions from Stationary Sources. EPA-600/2-76-
201, U.S. Environmental Protection Agency, Research
Triangle Park, NC, 1976.

Adams, J., Menzies, K., and Levins, P.  Selection and
Evaluation of Sorbent Resins for the Collection of Organic
Compounds.  EPA-600/7-77-044, U.S. Environmental Protection
Agency, Research Triangle Park, NC, 1977.

Jones, P. W., Wilkinson, J. E., and Strup, P. E.  Measure-
ment of Polycyclic Organic Materials and  Other Hazardous
Organic Compounds in Stack Gases:  State  of the Art.
EPA-600/2-77-202, U.S. Environmental Protection Agency,
Research Triangle Park, NC, 1977.

Gallant, R. F., King, J. W., Levins, P. L., and Piecewicz,
J. F.  Characterization of Sorbent Resins for Use in :
Environmental Sampling.  EPA-600/7-78-054, U.S. Environ-  •
mental Protection Agency, Research Triangle Park, NC,'1978.

Vidal-Madjar, C., Gonnord, M. F., Benchah, F., and Guiochon,
G.  Performances of Various Adsorbents for the Trapping and
Analysis of Organohalogenated Air Pollutants by Gas
Chromatography.  J. Chromatog. Sci. 16:190, 1978.

Bruner, F., Ciccioli, P., and Di'Nardo, F,  Use of      .
Graphitized Carbon Black in Environmental Analysis.,  J.
Chromatog. 99:661, 1974.
Mueller, F. X., and Miller, J. A.
Vapors in Industrial Atmospheres.
Determination of Organic
Amer. Lab.  May 1974.
 Zlatkis, A., Kaufman, H. R., and Durbin, D. E.  Carbon
 Molecular' Sieve Columns for Trace Analysis in Gas
 Chromatography.  J. Chromatog. Sci. 8:416, 1970.

 Grob, K., Grob, K., Jr., and Grob, G.  Organic Substances
 in Potable Water and its Precursor.  J. Chromatog. 106:299,
 1975.

 Thomas, T. C., and Jackson, J. W.  A Technique for Sampling
 of 27 4-D; 2,  4, 5-T Herbicides in Air.  Paper presented at
 1975 Pittsburg Conference on Analytical Chemistry and
 Applied Spectroscopy.
                               82

-------
96.


97.
98.
99.
  95.   Coburn,  J.  A.,  Valmanis,  I. A.,  and Chau, A. S. Y.
       Evaluation  of XAD-2  for Multiresidue Extraction of Organo-
       chlorine Pesticides  and Polychlorinated Biphenyls from
       Natural  Waters.   J.  AOAC  60:224,  1977.

       Ambersorb Carbonaceous Adsorbents:  Technical Notes.  Un-
       published data.   Rohm and Haas Co., Philadelphia, PA, 1977.

       Ciccioli, P., Bertoni, G., Brancaleoni, E., Fratarcangeli,
       R.,  and  Bruner, F.   Evaluation of Organic Pollutants in the
       Open Air and Atmospheres  in Industrial Sites Using
       Graphitized Carbon Black  Traps and Gas Chromatographic-Mass
       Spectrometric Analysis with Specific Detectors.  J. Chroma-
       tog. 126:757, 1976.                              	

       Lee, K.  Y., Nurok, D., and Zlatkis, A.  Combined Headspace
       and  Extraction Technique  for Profile Analysis by Capillary
      •Gas  Chromatography.  J. Chromatog. 158:377, 1978.

       Lewis, R. G., Brown, A. R., and Jackson, M. D.   Evaluation
       of Polyurethane Foam for  Sampling of Pesticides, Poly-
       chlorinated Biphenyls, and .Polychlorinated Naphthalenes in
       Ambient Air.  Anal.  Chem. 49:1668, 1977.

100.   Averill, W., and Purcell, J»E.  Concentration and Analysis
       of Organic Compounds from Air and Water,  Paper presented
       at 1978 Pittsburgh Conference on Analytical Chemistry and
       Applied Spectroscopy.

101.   Janak, J., Ruzickova, J., and Novak,  J.  Effect of Water
       Vapour in the Quantitation of Trace Components  Concentrated
       by Frontal Gas Chromatography on Tenax-GC.   J.  Chromatog.
       99:689, 1974.                               	~

102.  Van Wijk, R.  The Use of Poly-Para-2,  6-Diphenyl-Phenylene
       Oxide as a Porous Polymer in Gas Chromatography.   J.
      Chromatog. Sci.  8:418,  1970.                       ~

103.  Bursey, J. T.,  Smith, D,,  Bunch,  J. E., Williams,  R.  N.,
      Berkley, R.  E.,  and Pellizzari,  E. D.   Application of
      Capillary GC/MS/Computer Techniques to  Identification and
      Quantitation of  Organic"Components in Environmental
      Samples.   Amer.  Lab.   Dec. 1977.

104.  Bellar, T. A., Lichtenberg,  J. J. The  Determination  of
      Volatile Organic Compounds at the yg/1  Level in Water by
      Gas Chromatography.   EPA-670/4-74-009,  U.S.  Environmental
      Protection Agency, Cincinnati, OH, 1974,
                             83

-------
105.
106.
107.
108.
109.



110.



111.




112.




113.


114.


115.
Kuo, P. P. K., Chian, E. S.  K.,  DeWalle,  F.  B.,  and Kim,
J. H.  Gas Stripping, Sorption,  and Thermal  Desorption
Procedures for Preconcentrating  Volatile  Polar Water-
Soluble Organics from Water Samples for Analysis by Gas
Chromatography.  Anal. Chem. 49:1023, 1977.

Gearhart, H. L., Pierce, S.  K.,  and Payne-Bose,  D.   A
Sampling Technique for Organic Components in Human  Breath.
J. Chromatog. Sci. 15:480, 1977.

Leoni, V.,  Puccetti, G., and Grella, A.   Preliminary
Results on the Use of Tenax for the Extraction of
Pesticides and Polynuclear Aromatic Hydrocarbons from
Surface and Drinking Waters for Analytical Purposes,
J. Chromatog. 106:119, 1975.

Leoni, V., Puccetti, G., Colombo, R. J.,  and D'Ovidio, A.
M.   The Use of Tenax for the Extraction of Pesticides  and
Polychlorinated Biphenyls from Water:  II. Tests with
Artificially Polluted and Natural Waters.  J. Chromatog.
125:399, 1976.

Zeldes, S. G., and Horton, A. D.  Trapping and Determi-
nation of Labile Compounds in the Gas Phase of Cigarette
Smoke.  Anal. Chem.  50:779, 1978.

Chesler,  S. N., Gump, B. H., Hertz,  H. S., May, W.  E., and
Wise,  S. A.   Determination of Trace  Level Hydrocarbons in
Marine Biota.  Anal. Chem.  50:805, 1978.

Black, M. S., Rehg,  W.  R.,  Sievers,  R. E., and Brooks,
J.  J.  Gas Chromatographic  Technique for Compound Class
Analysis  of  Jet Engine  Exhaust.  J.  Chromatog. 142:809,
1977.

Peterson, H., Eiceman,  G. A., Field, L. R., and  Sievers,
R.  E.  Gas  Chromatographic  Injector  Attachment for  the
Direct Insertion  and Removal  of a  Porous Polymer Sorption
Trap.  Anal.  Chem.  50:2152, 1978.

Katys, G. P.   Continuous  Measurement of Unsteady Flow.
Pergamon Press, New York, NY, 1964,

Laub,  J.  H.   An Electric  Flow Meter.  Electrical Engineer
Dec. 1947,  p. 1216.

 Grey,  J., and Liu,  F.  F.   Methods  of Flow Measurement,  J.
Amer.  Rocket Soc.  May-June  1953, p.  133.
                                84

-------
116.  Lochmuller, C. H., and Gordon, B. M.  The Determination of
      Binary Gas Diffusion Coefficients Through Measurement of
      the Time-Based, Chromatographic Peak Variance.  J.
      Chroma tog. Sc i . 16:74, 1978.                    ~~

117.  Baumann, F., Klaver, R. F., and Johnson, J.'P.  Subambient
      Programmed Temperature Gas Chromatography. -  In: Gas
      Chromatography, 1962, Butterworths, London, 1962, p. 152.

118.  Rasmussen, R. A., Harsch, D. E., Sweany, P. H., Krasnec,
      J. P., and Cronn, D. R.  Determination of Atmospheric
      Halocarbons by a Temperature-Programmed Gas Chromatographic
      Freezeout Concentration Method.  J. Air Poll. Control
      Assoc. 27:579, 1977.             	:	

119.  DeRuwe, H. J. J., Schulting, F. L., and Van Grondelle, J.
      A Simple, Inexpensive Cryogenic Unit for Subambient Gas
      Chromatographic Analysis.  HRC and CC 1:211, 1978.

120.  James, A. T., and Martin, A. J. P,  Gas-Liquid Partition
      Chromatography:  The Separation and Micro-estimation of
      Volatile Fatty Acids from Formic Acid to Dodecanoic Acid.
    •  Biochem. J. 50:679, 1952.

121.  Golay, M. J. E.  Theory and Practice of Gas Liquid
      Partition Chromatography with Coated Capillaries.  In:
      Gas Chromatography.  Coates, V. J., ed. Academic Press, New
      York, NY, 1958, p.  1.

122.  Golay, M. J. E.  Theory of Chromatography in Open and Coat-
      ed Tubular Columns with Round and Rectangular Cross-
      Sections.  In:  Gas Chromatography 1958.  Desty, D. H., ed.
      Butterworths, London, 1958, p. 36,  p.  62. '••'.'

123.  Golay, M.  Brief Report on Gas Chromatographic Theory.
      In:  Proceedings of Third International Symposium on Gas
      Chromatography 1960.  Edinburgh, Scotland,  Butterworths,
      Washington, 1960, p. 139.

124.  Golay, M.  Height Equivalent to a Theoretical Plate of an
      Open Tubular Column Lined with a Porous Layer.  Anal. Chem.
      40:382, 1968.                                   	

125.  Giddings, J. C.  Dynamics of Chromatography:   Part I -
      Principles and Theory.  Marcel Dekker, New York, NY, 1965.

126.  Rubey, W. A.  Bibliography of Open Tubular Column Gas
      Chromatography.  Technical Report,  UDR-TR-79-11 University
      of Dayton Research Institute, Dayton,  OH,  Jan. 1979.

127.  Jennings, W.  Gas Chromatography with  Glass Capillary
      Columns.  Academic  Press, New York, NY, 1978.

                               85

-------
128.



129.


130.


131.



132.



133.



134.




135.



136.
Sandra, P., and Verzele, M.  Surface Treatment, Deacti-
vation and Coating in (GC)2 (Glass Capillary Gas Chromato-
graphy).  Chromatographia 10:419, 1977.
Sandra, P., Verstappe, M., and Verzele, .M.
28, 1978.
HRC and CC 1:
137.
138.
Grob, K.  The Glass Capillary Column in Gas Chromatography.
A Tool and a Technique.  Chromatographia 8;423, 1975.

Schomburg, G., Dielmann, R., Hussmann, H., and Weeke, F.
Gas Chromatographic Analysis w,ith Glass Capillary Columns.
J. Chromatog. 122:55, 1976.

Novotny, M., Blomberg, L., and Bartle, K.  Some Factors
Affecting the Coating of Open Tubular Columns for Gas
Chromatography.  J. Chrom. Sci. 8:390, 1970.

Ettre, L., and March, E.  Efficiency, Resolution and Speed
of Open Tubular Columns as Compared to Packed Columns.  J.
Chromatog. 91:5, 1974.

Rijks, J., van den Berg, J., Diependaal, J.  Character-
ization of Hydrocarbons in Complex Mixtures by Two-
Dimensional Precision Gas Chromatography.  J. Chromatog.
91:603, 1974.

Blomberg, L.  Uniformity of Stationary Phase Films in
Dynamically Coated Glass Capillary Columns for Gas
Chromatography.  Chromatographia 8:324,  1975.

Onuska, F., and Comba, M.  Comparison of Surface
Modification Techniques Used for the Preparation of Wall
Coated Open Tubular Glass Capillary Columns.  In:  Pro-
ceedings of Second International Symposium on Glass
Capillary Chromatography, Hindelang, Germany, Institute of
Chromatography, Bad Durkheim, Germany, 1977, p. 283.

Madani, C., Chambaz, E., Rigaud, M., Durand, J., and
Chebroux, P.  New Method for the Preparation of Highly
Stable Polysiloxane-Coated Glass Open-Tubular Capillary
Columns and Application to the Analysis  of Hormonal
Steroids.  J. Chromatog. 126:161, 1976.

Grob, K., Jr., and Grob, K.  Are We Using  the Full Range of
Film Thickness in Capillary-GLC.  Chromatographia 10:250,
1977.
                               86

-------
 139.  Madani, C., Chambaz, E. , Rigaud, M., and Bretton, J.  Use
      of Hydrolyzed Chlorosilanes for the Preparation of High
      Resolution Glass Capillary Columns.  Second International
      Symposium on Glass Capillary Chromatography, Hindelang,
      Germany, Institute of Chromatography, Bad Durkheim,
      Germany, 1977, p. 249.

 140.  Blomberg, L. , Buijten, J., Gawdzik, J., and Wannman, T.
      Preparation of Thermostable, Phenyl Silicons Coated, Glass
      Capillary Columns for Separation of Polyaromatic Hydro-
      carbons.  Chromatographia 11:521, 1978.

      Purcell, J. E., Downs, H. D., and Ettre, L. S.  Extra-
      neous Peaks in Gas Chromatography:  Their Source and
      Elimination.  Chromatographia 8:605, 1975.

      Guiochon, G.  Influence of the Apparatus on the Apparent
      Efficiency of Columns in Gas Chromatography.  J. Gas
      Chroma tog. 2:139, 1964.                       	'-	

      Sellier, N., and Guiochon, G.  Influence of GC-MS Coupling
      on the Performances of Chromatographic Columns.  J.
      Chroma tog. Sci. 8:147, 1970.                     ~~

      Eyem, J.,  The Role of Wall-Coated Capillary Columns in
      GC-MS Techniques.  Chromatographia 8:456, 1975,

      Novotny, M.  Coupling of Open Tubular Columns with a Mass
      Spectrometer through the Jet-Type Molecular Separator.
      Chromatographia 2:350, 1969.

      Ducass, A., Gonnord,  M. F., Arpino, P.,  and Guiochon, G.
      Simple Techniques for Two-Dimensional Gas Chromatography.
      J. Chromatog.  148:321, 1978.

      Anderson, E.  L.,  Moskot, M., and Bertsch, W.   High
      Resolution Columns in Gas Chromatography.  Paper presented
      at 174th ACS Meeting, Chicago,  IL,  Aug.  1977.

      Cram, S. P.  Extended Capabilities  in Gas Chromatography
      through High Resolution Techniques.  Paper presented at
      1979 Pittsburgh Conference on Analytical Chemistry and
      Applied Spectroscopy.

149.  Jennings, W,  G.   Food and Essential Oil  Analysis with Glass
      Capillary Columns.  Paper presented at 1979  Pittsburgh
      Conference on  Analytical Chemistry  and Applied Spectro-
      scopy .
                  J
150.  Fenselau, C.   Gas Chromatography Mass  Spectrometry:   A
      Report on State of the Art.   Appl.  Spectrosc.  28:305,
      1974.                        —	
141.
142
143,
144,
145,
146.
147
148.
                               87

-------
151.




152.



153.


154,


155.



156,
157
Gudzinowicz, B. J., Gudzinowicz, M.  J.,  and Martin,  H.  F.
Integrated GC-MS Analytical Systems:  Part I - Recent
Technological Advances Affecting Integrated GC-MS System
Utilization.  Chem. Instr. 8:225, 1977.

Ryhage, R.  Use of a Mass Spectrometer as a Detector and
Analyzer for Effluents Emerging from High Temperature Gas-
Liquid Chromatography Columns.  Anal. Chem. 36:759,  1964.

Heller, S. R., McGuire, J. M., and Budde, W. L.  Trace
Organics by GC/MS.  Environ. Sci. Tech.  9:210, 1975.

Milne, G. W. A., and Heller, S. R.  The MSDC/EPA/NIH Mass
Spectra Search System.  Amer. Lab. Sept. 1976.
Edwards, J. B.
Trace Species.
p. 64.
Combustion:  The Formation and Emission of
Ann Arbor Science, Ann Arbor, MI, 1974,
Carnes, R. A., Rubey, W. A., Mescher, J. A., and Duvall,
D. S.  The Thermal Decomposition Analytical System (TDAS):
Its Development and Application.  In:  Proceedings of the
72nd Annual Meeting of Air Pollution Control Association,
Cincinnati, OH, June, 1979.

Hutzinger, O., Sundstrom, G., Karasek, F. W., and Safe, S.
The Chemistry of Some Potential Polyhalogenated Water
Pollutants.  In:  Identification and Analyses of Organic
Pollutants in Water.  Keith, L. H., ed.  Ann Arbor Press,
Ann Arbor, MI, 1977, p. 15.
                                88

-------
                           BIBLIOGRAPHY


 Benedict,  R.  P.   Fundamentals  of  Temperature, Pressure, and
      Flow  Measurements.   Wiley, New York, N.Y.,  1969.

 Bird, R. B.,  Stewart, W.  E., and  Lightfoot, E. N.  Transport
      Phenomena.   Wiley, New York, N.Y., 1960.

 Dal  Nogare, S.,  and  Juvet, R.  S., Jr.  Gas-Liquid
      Chromatography.  Wiley -  Interscience, New  York, N.Y.,
      1962 .

 Danials, T.   Thermal Analysis.  Wiley, New York, N.Y., 1973.

 Deelder, R. S.,  Kroll, M.G.F., and Beeren, A.J.B.  Post-
      Column Reactor  Systems in,Liquid Chromatography.
      J. Chromatog. 149:669, 1978.

 Edwards, J. B.   Combustion:  The Formation and Emission of
      Trace Species.  Ann Arbor Science, Ann Arbor, Mich.,  1974.

 Harris, W. E., and Habgood, H. W.  Programmed Temperature  Gas
      Chromatography.  Wiley, New York, N.Y., 1966.

 Hettmann, E.  Chromatography.  Van Nostrand Reinhold,  New
      York, N.Y., 1975.

 Jones, C. E. R., and Cramers, C. A.   Analytical Pyrolysis.
      Elsevier, Amsterdam,  Netherlands, 1976.

 Kays, W. M.  Convective Heat and Mass Transfer.   McGraw-Hill
      New York, N.Y., 1966.                                  '

 Keith, L. H.   Identification and Analysis  of Organic Pollutants
      in Water.  Ann Arbor  Science, Ann Arbor,  Mich.,  1976.

 Kennedy, M. V., ed.  Disposal and Decontamination of Pesticides
     ACS Symposium Series  No.  73,  The American Chemical  Society
     Washington,  D.C.,  1978.

Kreith, F.   Principles  of  Heat Transfer.   International  Textbook
     Co., Scranton, Pa., 1958.

Lewis, B.,  and Von Elbe, G.   Combustion, Flames,  and Explosions
     of Gases.  Academic Press, New York,  N-.Y.,  1951.

                               89

-------
Littlewood, A. B.  Gas Chromatography.   Academic  Press, New
     York, N.Y., 1970.

Mickley, H. S., Sherwood, T. K., and Reed,  C.  E.   Applied
     Mathematics in Chemical Engineering.  'McGraw-Hill, New
     York, N.Y., 1957.

Miller, J. M.  Separation Methods in Chemical  Analysis.  Wiley-
     Interscience, New York, N.Y., 1975.

Purnell, H.  Gas Chromatography.  Wiley, New York, N.Y.,  1962.

Reddick, H. W., and Miller, F. H.  Advanced Mathematics
     for Engineers.  Wiley, New York, N.Y., 1955.

Reid, R. C., and Sherwood, T. K.  Properties of Gases and
     Liquids.  McGraw-Hill, New York, N.Y., 1958.

Sternberg, J.  C.  Extracolumn Contributions to Chromatographic
     Band  Broadening.  Advances in Chromatography.  Vol.  2.
     Marcel Dekker, New York, N.Y., 1966.

Wiedemann, H.  G.  Thermal Analysis.  Academic Press, New York,
     N.Y., 1969.

Wilkinson, R.  R., Kelso, G, L., and Hopkins,. F. C.  State-of-
     the-Art  Report:  Pesticide Disposal Research.  EPA-600/2-
     78-13, U.S. Environmental  Protection Agency, Cincinnati.,
     Ohio, 1978.
                                90

-------
                           APPENDIX A

      GAS FLOW ANALYSIS:  A DETAILED ANALYSIS OF GAS FLOW
           AND DISPERSION IN A NARROW-BORE CIRCULAR
          CROSS SECTION TUBE OF FOLDED CONFIGURATION


     Residence time is an important parameter in thermal decom-
position studies.  To illustrate the importance of residence
time, or more emphatically the residence time distribution,
consider the two dramatically different residence time distri-
butions that are represented in Figure A-l(a).  Curve I is
typical of the narrow and symmetric residence time distributions
obtained using laboratory instrumentation.  Curve II is typical
of the residence time distributions associated with a large-
scale unit having a high volume throughput and a minimum of
mixing.  The thermal decomposition profiles corresponding to
these two different residence time distributions would be as
depicted in Figure A-l(b).  Notice that the thermal decomposition
profile corresponding to the wide and asymmetric residence time
distribution shows  (1) the onset of thermal decomposition1 at
lower temperatures than the narrow residence time distribution,
and  (2) a requirement of higher exposure temperatures for the
same level of efficient thermal decomposition.

     Theoretically, the narrower'the residence time distribution,
the more closely the thermal decomposition profile represents the
actual thermal decomposition behavior of a given substance.
Thus in the ideal thermal decomposition experiment, each of the
molecules would be subjected to precisely the same residence
time.  Also, the value of this "line-function" 'residence time
would be experimentally adjustable and measurable to a high
degree of accuracy.

     In practice, these uncompromising ideal conditions are
unattainable, although accurately determined symmetric and
narrow residence time distributions can be generated in
laboratory-scale experiments.  For large-scale work that in-
volves large throughputs, such narrow residence time
distributions are not presently feasible, even though progress
is being.made in this design area.  Nevertheless, for purposes
of comparison, interpretation, and extrapolation, it is of vital
importance that the residence time distribution of a substance
be known along with the other thermal decomposition test
parameters.
                                91

-------

  m

  3
  o

  Q. co
    o
  LU
  UJ
  cc
                    TIME
             a)  RESIDENCE TIME DISTRIBUTIONS
cc


h-

UJ
o
o:
UJ
D_


h-

(3

UJ
            100
             10
            O.I
           0.01
                 T2 .EXPOSURE  TEMPERATURE




          b) THERMAL DECOMPOSITION  PROFILES







Figure A-l.  Effect of residence time bandwidth
                         92

-------
  «   In this appendix, a detailed transport analysis is present-
ed.  The objective of this analysis is to characterize the gas
flow behavior and the residence time variation associated with a
laboratory-scale thermal decomposition analytical system.

Straight Tube of Circular'Cross Section

     There are two simultaneous factors, i.e., dynamic processes,
which influence the dispersion of a dilute zone of like gas
molecules as they are being transported by a gaseous carrier
through a straight tube of circular cross section.  Funda-
mentally, these two factors are convection and diffusion.

     The laminar flow of gases through such a tube exhibits a
paraboloidal velocity profile; therefore, instantaneous axial
velocity depends on radial position.  Concurrently, the flowing
gas will experience three dimensional diffusion according to
Pick's second law.

     An equation that describes the transport behavior and the
concentration changes associated with these two effects can be
written as
                3C
                3t
   -  V7
      3C
         +  D
(Al)
where C is the concentration of like molecules in the transport
gas, t is time, v represents the gas axial velocity, z is axial
distance, Dg is the intermolecular diffusion coefficient, and
x and y represent the mutually perpendicular Cartesian
coordinates.

     Fluid flow through a straight tube of circular cross section
can be examined by equating the applied hydrostatic force to the
opposing viscous force.  Then, over an axial fluid increment dz
A  dp
 m  ^
                            =  (n
              >  dz)
              s     dr
(A2)
where Am is the cross-sectional area, r\ is the fluid viscosity,
PS is the surface perimeter perpendicular to the z axis, r is
radial distance, and dp/dz is the negative pressure gradient.

Accordingly, for a circular cross^section flow path
irrdp  =
                                  2trr dz)
(A3)
                                93

-------
                          civ  =   /r \  d£
                          dr     \2n/  dz
                                    (A'4)
T
•'/-.
                                           r  dr
                                    (A5)
where r  is the tube inner radius.  Then,
                        v  = —
                               ,22'
                                r  ~ r
                                          _
                                  4n   /  dz
                                                              (A6)
and the maximum velocity, V, can be written as
                    V  =  v
 L  •  -fe)
                                          (A7)
Therefore, it is seen that the axial gas velocity varies with
r according to                                  ,
                       V  =   V 1  -  f^rl                       (A8)
Next, the volumetric flow rate through the tube can be expressed
as
                                v(2frr) dr  .
                                    (A9)
and upon substituting equation A8,
               F  =  2Vrr
r
            |r - ±" 1 dr
                 r
                                                TrVr
(A10)
Thus the average linear gas velocity, v, can be expressed as

                                F     V
                          v-  =
                                   (All)
                                m
                                94

-------
and equation Al can then be rewritten  as
It  -  -^
                           r2
                        i _/ r
                           r2
                           1 o,
   ,   32C   32C .  32C
3z   I ,..2   ...2   n_2
                                                   (A12)
      3x
                                         3z'
With the axial velocity term expressed  relative to radial
distance, it is likewise advantageous to  rewrite the spatial
diffusion term in cylindrical coordinates.   Therefore,  starting
with the following relationships:     ,        . :..   •
                           x
                              =   r  cosG
                                                              (A13)
                           y   =   r sinG
                                                             (A14)
                              z = z
                                                             (A15)
                          r  =   Vx2 + y2
                             (A16)
the spatial diffusion  can  eventually be equated according to
1      9      9
'C_ ,  3 C    3 C
2  *   2  +  ,_2
                                       3C
                   r*    3
                   N^  ,  a
                                                 80'
                                                             (A17)
However, for  the  special  case involving laminar flow through a
straight tube of  circular cross  section, it is seen that C is
symmetrical about the  z axis. Therefore,
                         9C_
                         36,
    =  0
                             (A18)
thus
   32C
        32C   3 C
          2~     2~
        3y    3z
 32C
                                            3C
                                         r/  3r
ifc
3z2
                             (A19)
                                95

-------
 and finally,  the basic equation can be rewritten as
           32C

       =  lg+ 2v
                                                       IT     (A20)
Prom equation A20,  it is seen that
                     C  =   f(Dg,r,z,t,v,ro)
                                    (A21)
Therefore,  as  v represents the axial movement of the
concentration _centroid,  and so that the axial point of reference
moves with  this centroid,  a new coordinate,  z,,  is introduced
and defined as                              -
                          z 1   =   z  -  vt
                                    (A22)
Along with  this axial  coordinate change,  the time coordinate
correspondingly must be  revised.  Therefore, since

                            at-j^   =  3t                        (A23)

and as
                            3t
                                =   - v
                                                             (A24)
then by adopting this new  set of  coordinates,  z   and tn,  it is
coon -f-Viu-f-                                      J-       -L
seen that
                 3C
                 3t
3C
3C
       F
      3t
                     (A25)
Hence,
                   3C
                   3t
  3C
        - v   +
  3C
  3t,
(A26)
and equation A20 can be rewritten as
                                96

-------
                          3r'
                                         32C
  3z
                                      -   _
                                   ~ ^ V J-
                                            r    9C
                                                7T~
                                                9z
                                                   l.
3.tn
                          v  1 -
?r  \  ar
£L.  \  O\~,
~7}^
                                                              (A27)
If now the' stipulation is made (Golay's assumption)  that  C  never
differs greatly  from the average radial concentration  C~,  then
where
          C  =
                           C  =  C + AC
                       /r
                        (
       C 2-rrr dr  =  ~

                    ro
           /
           «'r\
C r dr
                                                              (A28)
(A29)
and
                             /'
                             «'<-\
               AC r dr
                             (A'30)
                             AC «  C
                                                              (A31)
As equation A28 has  been introduced and defined, equation A27
can be rewritten  as
                D
                    3r'

    2
    :1
                                3C    3AC
                                3t,
                              1  -
                                          (A32)
                                97

-------
and with the following assumed  conditions
                        32C  _
                        3r2  "
                                   (A33)
                            8 AC
                            at
                                 -»•   0
                              1  '
                                   (A34)
32C

3z?
    32AC
      2
                                                             (A35)
 3C
3z,
    3 AC
    3z,
                                                             (A36)
then, equation A32 can be rewritten  as
°g
3C
9tl
32AC , /1\ 3AC , 32C
9 IT-/ 3v 0
.%„•£ \-L/ or n„^
dr \ / 3z ,
f V
/2r2\
U2 /
3C
8Z1
                                                             (A37)
Wow, if each of  the terms  in  equation  A27  is  multiplied by the
operator
                               f
          r  dr
                             (A38)
and simplified through  the  use  of  equations A33 to A36, along
with substitution of  the boundary  condition
                            .
                           3r
                                     =  0
                                   (A39)
                               r=r
then the following relationship  is  obtained:
                      2v
                       2
f
1 -
     2r"
3AC   ,   .  3C
TT—— r dr + TT.—
(A40)
                                98

-------
Therefore,  by subtracting equation A40 from equation A37  and
taking  into account the condition set forth in equation A36 ,
then  the  following important equation is obtained:
Now if
32AC
3r2
                    +   —
                          3AC
                                  v
                                      1 -
                                   8C
                                                              (A41)
and
                            3AC
            32AC    ' _3J_

            3r2   =  Sr
                                                              (A42)
                                                              (A43)
then equation A41 can be  rewritten as
3r   r
                               v
                            ,  \D
                            1  X -g,
                       3C  / v \ 2r"
                      3z,  ID  /   2
                        1  \ g/  r
(A44)
and a solution is obtained as  follows:
                              dr
                                                             (A45)
        3C  JT_
        3z,  ID..
                                             3 AC
                                             3r
                                               (A46)
               AC  =
        3C  /v \  1   /".  2
        v^~  FT~  —^-  I (rr
        8Z1 \Dg/ 2r2 7
                                                 dr
                                               (A47)
                 AC   =   AC  +

                          o    4D
                         r    3C
                                               (A48)
                               99

-------
Upon differentiation, the ACQ term vanishes, and
                    3 AC
                    3z.
                      4D
                           1 -
                                r2 \ 32C
                      Ji      "g  .

Then by substitution  into equation A40,
  ? ;   2
2r / 3zf
  o    J-
                       (A49)
 o       _ 9
afc  _   (v)2

     "
                      9 —
                     3c
(2r4  -  5r2r2
   o      °
         2r4) r3dr
which eventually is rewritten as
Dg"
(vro)2"
48Dg
32C
3z2
                                          3C
                                                            -1
                                                             (A50)
                                                             (A51)
It  is now apparent  that  equation A51  can  be  expressed  in  the
same form as Pick's second  law, but with  an  effective  diffusion
coefficient defined as
                 D   =  D  +
                  e      g
                                    (vr )2
                                      o
                                    48D
                                                             (A52)
                                       g
 If  now De is  inserted into the one dimensional diffusion
 equation, that is

                            d C      dC
                            dz
                                    dt
                                                             (A53)
 then consistent with appropriate initial conditions,  the
 following is obtained as a solution:
                C  =  C.
                                      4Detx
                                                             (A54)
 This equation is readily recognized as a Gaussian distribution
 function.  Also, it is seen to have a distribution variance of
                                                             (A55)
                                100

-------
Thus  for the defined uniform flow path,  the behavior of the
distribution variance,  relative  to axial distance,  can be
equated as
                      3z
=  2D
                                          2D
                                           v
                                                             (A56)
and from equation A52,
                                2D
                                 v
                                       - 2
                                      24D
                                                             (A57)
Thus equation A57 represents  the  longitudinal  (or  distance-
based) variation in transport of.like gas phase molecules.

     If now the initial condition is
and from equation A57
                       -z=L


                     J r\
                                    -»•  0
                               z=0
                                 (A58)
'2D
  v
      dz +
L
z=L / - 2
    / vro
    I —
                      O               "O

then at distance z=L, it is seen that
    \24Dg,
           .dz
(A59)
                              2D L   vr L
                                v
                                     24D
                                 (A.6.0)
     Throughout the above examination, D  has been assumed
to be independent of concentration and only influenced by
pressure and temperature.   (For dilute binary mixtures, this is
indeed a valid assumption.)  Therefore, equation A57 also
applies for the situation where like molecules are randomly  •
entering and passing through the defined uniform flow path.
However, it would be more convenient if the migration variation
could be represented with a time-based independent variable.
Hence; it is necessary to convert the az value of the Gaussian
distributed population to a time-based sigma value.  This is
readily accomplished as
                               101

-------
                           atF
                          VAm ,
(at)2  (v)2
(A61)
Equation A60 can now be rewritten as

                         ~      2D_L
                     (at)2   =
 r2L
   o

24VD
(A62)
and finally it is seen that the time-based variation in
residence time is
                                                             (A63)
Folded Tube of Circular Cross Section

     What has been stated thus far applies strictly for laminar
gas flow through a nonreactive, smooth-walled, straight tube of
circular cross section.  When the axis of the flow path exhibits
significant curvature, then additional considerations must be
included in the descriptive analysis of gas flow.

    •At low gas flow velocities, both convection and diffusion
strongly contribute to axial dispersion in a narrow gas flow
path.  However, when the average linear gas velocity within a
1.0 mm ID straight tube is large  (v > 100 cm sec x),_then the
vastly dominant dispersion mechanism is associated with the
convection.  Longitudinal diffusion represents a negligible
contribution to total dispersion at these higher gas velocities.
Thus at high linear gas velocities,

                            2DgL  _^n                         (A64)
 and  for  the  case where  v is  large  in  a  straight  tube,  equation
 A63  can  be written  simply as
                                                             (A65)
                                  24vD
                              •  102

-------
However, the velocity profile in a curved tube is not para-
boloidal.  It includes an element of flow that is perpendicular
to the tube axis, and this additional element of flow is called
"secondary flow"  (see Section 6, and particularly Figure 18).
This additional flow component can be beneficial in that at the
higher gas velocities, increased radial mass transfer occurs,
which thereby produces less axial dispersion.

     Equation A65 applies for high velocity paraboloidal flow;
however, the general variance equation for the non-paraboloidal
velocity profile  (which contains secondary flow) can be expressed
as                          '
                             2    tor L
                         (at)  =	                      (A66)
                                  24vD
                                      g
where the parameter to characterizes the actual flow profile,
which in turn depends on the axial curvature, circular bore
diameter, and surface roughness of the gas flow path, along
with the mean axial velocity and the physical properties of the
flowing gas mixture.  Recent curve fitting calculations obtained
from experiments with flowing liquids have empirically expressed
to in terms of the Dean number, NDn, and the Schmidt number, NSc
(see the study by Deelder, R. S., Kroll, M.G.F., and Beeren,
A.J.B. Post-Column Reactor Systems in Liquid Chromatography.
J. Chromatog. 149 : 669 ,  1978).  Specifically, for values of
                       N_  J N_   >  12.5
                        Dn \  Sc
                          (A67)
it is observed that
                      = 5.6
                                        -0.67
                          (A68)
When
                       N.
                        Dn
< 12.5
                                                             (A69)
then to is equal to unity.

     Gas flowing through a narrow7bore tube of folded con-
figuration (such as that shown in text Figure 23) will encounter
dispersion in both the straight and curved sections of the
folded tube assembly.  As the circular cross section has been
maintained throughout the entire lengt£ of the tube, the total
                               103

-------
residence time variance is amenable to calculation.  First, the
assumption is made that for this folded tube, there are
instantaneous transitions between the laminar secondary flow and
the concentric laminar flow (i.e., pure paraboloidal velocity
profile) and vice versa.  Then by incorporating into equation
A62 the variance contributed by secondary flow as expressed in
equation A66, a general residence time variance equation can be
obtained.  Specifically, since independent Gaussian variances
are additive,
                                                             (A70)
and as
and
                            2D
                            (v)3
                            •2D
       ,2
       "o
     24vDg_,
     .2
     r
      o
                                     24vD
                                         g J
                                                             (A71)
                                                             (A72)
Then, the general variance equation for this tube will be as
follows:
 \     2
 1   tor
^+ —2-
                                    24  v D
r2 L
 o  s

24 v D
                                                             (A73)
where Lc is the total length of curved tubing, and Ls represents
the sum of the straight lengths.

     Finally, to complete this analysis of gas flow, the specific
data corresponding to the quartz tube reactor shown in Figure A-2
have been calculated.  Table A-l shows the range of various
flow-related parameters that cover  the intended mean residence
time extremes of 0.20 to 5.0 seconds.  These calculation values
correspond to dilute levels of naphthalene in a flowing air
atmosphere at a reactor internal pressure of 1.0 absolute
atmospheres.  With the narrow bore reactor and the relatively
small curvature of the flow'path, secondary flow is only a
                               104

-------
                             RACE-TRACK
                           CONFIGURATION
                           (3.5 CYCLES)
2.1 cm  RADIUS
       L = 100.75 cm
       l_s = 80.95 cm
       Lc = 19.80 cm
Vq = 0.7445 cm
r = 0.0485 cm
X = 43.3
                                              3
       *Includes effective volume of connecting tubing
        Figure A-2,   Quartz  tube  reactor data.
                         105

-------
     TABLE A-l.  CALCULATED FLUID DYNAMIC VALUES
Dimensionless
    term
_Lower extreme
v = 20 cm sec
T2 = 1423°K
 Upper extreme
v = 500 cm sec
T2 = 473°K
    N.
    N
    N
Re

Dn

Sc

 N
       Sc
 N
  Dn
   0.71

   0.108

   4.28


   0.050

   ,0.223
   134

    20.3

     2.97


  1224

    35.0
                            106

-------
factor at the very short residence times while at the lower
reactor temperatures.

     The effects of temperature, molecular weight, and mean
residence time on the residence time variance were investigated
in a previous study (EPA-600/2-77-288,  U.S. Environmental
Protection Agency, Cincinnati, Ohio, 1977, p. 59).  Therefore,
in concluding this analysis, Figure A-3 represents the calcu-
lated relative residence time bandwidth as a function of mean
residence time for the quartz tube reactor described in
Figure A-2.
                              107

-------
40
       0.3
  t    0.2
        0.1
                   1423°K
                             NAPHTHALENE IN AIR AT
                             1.0 ABS. ATMOSPHERES
                       MEAN RESIDENCE TIME,  seconds
Figure A-3.  Relative bandwidth as a function of mean residence
            time.
                             108

-------
                           APPENDIX  B

      ERROR ANALYSIS OF MEAN RESIDENCE  TIME  DETERMINATION


     The TDAS uses in-line instrumentation for  determining  the
mean residence of a sample in  the high-temperature  reactor.
Therefore, as presented in Section 6, equation  7  describes  the
mean residence time of a  substance that has  passed  through  the
quartz tube reactor.  This residence time equation
                                        1 +
                                            Jd
                                                              (Bl)
contains many variables, some of which have a pronounced  effect
on tr and are. difficult to measure.
     Upon investigation, it is found that fused quartz has a
very, _small coefficient of expansion  (approximately  0.59 x 10~6)
Therefore, for a given tube, the radius r and the length L can
be assumed temperature invariant.  Thus the effective volume
of a specific quartz tube, i.e.
                            Vg =
                             (B2)
remains essentially constant,
     With the refined instrumentation incorporated into the
T]DAS, there are now five variables that can have an effect on
tr.  Therefore, the maximum change in tr, which can be
attributed to an addition of errors in F ,
can be expressed as                     °
              T.
and
                3F~
                      AF
                        o
                      Ap
                        d
at
~3T~
                                        AT
                                                    3T.
                                                          AT,
                            (B3)
                               109

-------
From equations Bl and B2, we find
                   3t
 r
p
' O
           V T
                                9	

                               F~T9
                                O 2
I 1 +
                                                              (B4)
                    9t
           V
                              F^To
                               O 2
                                       1 +
                                          (B5)
                 3t.
                 3T,
          V T.
          F T^
           O 2
                    1 +
                                                              (B6)
                                   V T
                                    q
                                                              (B7)
                        3t

                                                              (B8)
Now the maximum errors  typically associated with the selected

instrumentation are



                     AF0 = 1 x 10~3 cm3 sec"1


                     AT   = 0.2°K
                     Apd = 2 x 10   atm


                   •  Ap   = 1 x 1O   atm
                      o


 Next,  an example case is presented where
                      t   =1.00 sec


                      F   = 0.3287 cm  sec
                      o

                      T   = 296°K
                      o
                                110

-------
                         T2 = 773° K

                         Pd = 0.15 atm

                         PQ = 0.98 atm

                         V  = 0.7445 cm3

     By substituting into  equation B3 the data for this case
and the typical measurement errors, the worst condition is
found to be
 At. =
At
             (F>
At
At
                                    At
                                    At
                                                              (B9)
      At  = 0.00304 + 0.00068 + 0.00129 + 0.00177 + 0.00014
Thus
                        At
                                  0.00692
                                                   (BIO)
or, the maximum relative error in t  is approximately 0.7%.

     With this in-line instrumentation and the use of cali-
bration correction curves, accurate determinations of mean
residence time can be made between the extremes of 0.25 and 5.0
seconds.  Even so, most TDAS experiments will be conducted with
the selected mean residence time falling between 0.5 and 2.0   '
seconds.  Thus in this region, and with the present instru-
mentation, the measurement of gas flow represents the dominant
source of error.
                               Ill

-------
                           APPENDIX C

           QUARTZ TUBE REACTOR HEAT TRANSFER ANALYSIS
     In designing the fused quartz tube reactor for the TDAS,
one of the specific objectives was to minimize the end effects
(disturbances) produced by the entrance and exit regions of the
reactor.  This end effect minimization was accomplished to a
considerable extent by employing the counterflow cooling
principle as shown in Figure C-l.  Using this design feature,
an abrupt change in the temperature of the flowing gas _ is
experienced as,it moves from the transport region and into the
high-temperature reactor.

     In view of the small bore of the gas flow path, it would be
very difficult to determine experimentally the exact temperature
rise profile as a function of either time or distance.  However,
this temperature rise profile is amenable to heat transfer
calculations.

     The approach used for obtaining this information was to
determine initially the bulk gas temperature within the fine
bore connecting tube as a function of longitudinal distance.
Then the effects of the longitudinal conduction within the
quartz junction assembly were examined.

     First, consider the steady state heat transfer condition
v;here 550°K gas is flowing into a quartz tube that is contained
in a 900°K isothermal chamber  (see Figure C-l).  Of primary
importance is  the time required for the incoming flowing gas to
attain effectively the 900°K equilibrium.

     A simplified version of the fine bore connecting tube is
depicted in Figure C-2, and the rate of heat flow by radial
conduction, qk, into the bore of this tube can be expressed as
   _
=  kA
       dr
                                     _
                                     k27rrL  —
                                            dr
(Cl)
where k is the average thermal conductivity of fused quartz, A
is radial area, T is temperature, r is radial distance,  z is
                               112

-------
                                 CONNECTING TUBING
                                 70 mm by 0.43 mm ID
                                 2.1 cm RADIUS
                                                             70 mm by 0.43 mm ID
                                            980 mm by 0.97 mm ID
                     HIGH
           vp..T    TEMPERATURE
            ^       INSULATION
VENT
 CARRIER
  +
 SAMPLE
  L
        CARRIER •
      " •  -K
      • PRODUCTS
Figure c-1.    Details of quartz  tube  reactor assembly,
                                 113

-------
Figure C-2.  'Simplified version
of fine bore connecting tube,
                              114

-------
 axial distance, and L is length  of  the  tube.   If conduction

 the longitudinal direction is neglected,  then
                                   in
                           dr  _  k2irL
                                       dT
                                    (C2)
 The temperatures corresponding to the boundary conditions
                                 are
                             at r  =  r
                           w           w
 and
                                    (C3)
                          T  at r  =  r
                           3           s
where  the  subscript s denotes the tube inner surface, and

subscript  w represents the outer wall.  Thus
                                    (C4)
                       "w
                                         ,T
dr _ 2-rrLk


r     q,.
                                          w
                                             dT
                                       T
                                                               (C5)
or
                               27TLk(T -T
                                     w  s
        r .


        r~
                                                              (C6)
and therefore for the dz axial element  depicted in Figure C-2,
  2irk(T -T ]
 	w  s

       r
       _j

       r
                                        dz
                                                              (C7)
                               115

-------
     Next, to obtain a comparison of the heat transfer mecha-
nisms, the following steady state energy balance was con-
structed :
                  dq  =
= dqc =
                                                              (C8)
where h is the average convective heat  transfer  coefficient  of
the laminar flowing gas, Tfc  is  the bulk temperature  of  the gas
in the bore of the dz increment, and  qc.is  the rate  of  heat
transfer by convection.   (To simplify the analysis,  heat
transfer by thermal radiation is neglected.)  Upon inserting
equation C7 and  incorporating the appropriate ATs  to maintain
the balance,  the equation  may be rewritten  as
                   2frk(AT1J
                            dz = h 2irr (AT )dz
                                      s   *-"
                               (C9)
 If now
                                     1.3 Wm
                               (C10)
                 -"quartz at 900°K
 and
                        h  =  500 Wm 2 °K
                               (Cll)
 then by evaluating the respective AT's required for maintaining
 the balance for the dz element, it is found that
                                                              (C12)
 Therefore, for the quartz connecting  tube  of Figure  C-2,  the
 valt?y dominant mechanism of heat  transfer's  radial conduction
 through the  tubing wall.  Then  for the entire  connecting  tubing
 length L, the assumption can be made  that
                              T  = T
                               s    w
                                                              (CIS)
                                 116

-------
      An energy balance can now be written for the heat transfer

 case as illustrated in Figure C-2.  Specifically,
      _
 q   =  h
                           T  -
                            w
                                 T-+T,
                              =  m  c
                                                  (ci4)
 where To is the bulk temperature of the gas at z=0,  m is the

 mass_rate of flow, and cp is the specific heat.  If  the flowing

 gas is air, then the properties of the flowing media at 900°K
 are as follows:                                              ,
                     density,  p  =  0.393  kg m




              specific heat, c   =  1.12 k J kg"1 °K~1
dynamic viscosity, y = 3.9.0 x 10~5 kg m"1 s
                                                  "1  '1
and as
        thermal conductivity, k = 0.0628 Wm"1 "^
             Prandt number, N   =0.696
                                =  900°K
                           z=0
                                =  550°K  =
                                =  0.000215m
                                              o
                             L  =  0.070m
and the average linear gas velocity,  v, is
                        v  =  2.56m s
                                     -1
                                                 (C17)
                               117

-------
then the Reynolds number  is
                             pv2r
                                 £. = 11.09
                                    (CIS)
     With this  low NRe,  along with a"constant Tw, the Nusselt

number  (from  Graetz's  classical solution) is
                                       h2re

                       .   =  3.66  = • —^r
                       Nu              .. k
                                     (C19)
and  therefore
                   h  =
3.66k

 2r_
                                   •535 Wnf 2°K
(C20)
 Also
                 m = pirr2v =1.461 x 10~^ kg  s~
                        s
                                     (C21)
 Equation C14 can now be rewritten  as
                     T  -
                      w
                           T +T,
                            o  b
                        T,-T
                         b  o
               •me ,
                 P


             h2irr z
                 s
(C22)
 which for this particular  case reduces to
                     z  =
  (T,-550)  0.000453
    D   	

       1250-Tb
                                                               (C23)
                                 118

-------
 Using equation C23,  a graph of Tfc, versus :z was prepared and is-.
 shown in Figure C-3,.  This graph reveals that the flowing bulk
 air will experience  a 350 °K increase in the very short axial
 distance of 0.5 mm.   This distance is considerably smaller than
 the axial length of  the quartz barrier between the two heated
 zones (see the enlarged detail drawing of Figure C-l) .
 Consequently,  as a result of the above calculations,  it is
 apparent that  the major factors that affect this gas  phase
 temperature increase are:  1)  the axial length of the barrier
 between  the heated zones and 2)  the volume flow rate  through
 this transition region. '..A .simplified sketch of this  particular
 heat transfer  condition is presented in Figure C-4 .

     _As  was seen from Figure C-3, the axial distance  required to
 attain thermal equilibrium is approximately 0.5 mm.   Therefore,
 if  this  length is added to the barrier length, a worst case
 situation is presented.  Now if it is assumed that the average
 temperature of the air in this transition region is 725 °K, and
 as  the interior volume, Vt,  of this region is       ,
                                         zl + A2Z2
                    —  1.130 mm                              ( C 2 4 )


and the average air  flow rate, F, corresponding  to  725°K  is
                  F  =
       725 *   =   300  mm s"1
                                                             (C25)
where Vr is reactor volume and tr is'mean residence time.  Then
it is seen that the time required- to displace the Vt transition
volume is               •             .
      vt
t  =  -
                                t±   0.0038s
                                                            (C26)
or 3.8 milliseconds.

     From previous measurements and calibrations, the
temperature of the central portion of the quartz tube reactor
has been well established.  And now, with the calculation of
                               119

-------
      T,°K
             1000
              750
              500
              250
                0  .  .
                 0
 550
 600
 700
 775
 850
 890
                                           (mm)
 0
0.0348
0.124
0.215
0.340
0.428
  1.0       1.5
z , mm
           2.0
Figure C-3.  Graph  of bulk temperature versus distance,
                               120

-------
     550°K
900° K
                                            0.43mm
                TEMPERATURE_
                 TRANSITION.
                  REGION
Figure C-4 .   Sketch of heat  transfer condition,
                       121

-------
the temperature rise time, the gas phase temperature can be
represented relative to mean time as shown in Figure 25, which
is presented in Section 6.

     An even more abrupt temperature change occurs at the exit
of the reactor.  Specifically, the temperature of the gas emerg-
ing from the reactor changes almost instantaneously to the
axial transport temperature.  This very rapid change is
attributed to the combination of induced mixing, greatly reduced
axial velocity, and thermal quenching.
                               122

-------
                           APPENDIX D

                  TDAS TRANSDUCER CALIBRATIONS
     To obtain accurate measurements of the various thermal
decomposition parameters, it is necessary to have precise
measurement'devices that have been accurately calibrated.
Therefore, the individual TDAS measurement transducers have
been calibrated using measuring equipment possessing trace-
ability to the National Bureau of Standards.  In this appendix,
the callibration data and associated correction curves are pre-
sented for those transducers that were described earlier in
Section 6 under the heading "TDU Instrumentation."

     The differential pressure transducer calibration data are
presented in Table D-l.  This device was calibrated using a
Merian Instruments 30-PAlO mercury manometer, which was cali-
brated for 22°C and 45° north latitude.  The corresponding
voltage readings were made with a Fluke 8300A digital voltmeter.

     The calibration data for the mass flow transducer are pre-
sented in Table D-2 and Figure D-l.  This particular calibration
was conducted by the manufacturer.  Both the pressure and flow
transducer outputs were alternately registered with a Hewlett-
Packard 3465A digital multimeter.  Calibration data and
corrective curves for this instrument are shown in Table D-3
and Figure D-2.

     Many temperature measurement calibrations were performed.
First, it was necessary to determine the longitudinal dis-
tribution of temperature throughout the reactor region.  This
was accomplished using a longitudinally movable unsheathed #24
AWG chromel-alumel thermocouple, and the actual temperature
measurements throughout the reactor region are shown in
Table D-4.  Using this same unsheathed thermocouple, the
temperature variations as a function of time at ,a given location
were determined to be less than ±0.5°C over a 10-minute interval,
For the actual measurement of the average temperature of the
reactor, that is T2, a special calibrated sheathed thermocouple
was purchased  (Omega Engineering Inc.).  This thermocouple was
calibrated by the manufacturer at temperatures commonly used;in
the TDAS-work.  These temperature calibrations are shown in
Table D-5.  The readout device for the T2 thermocouple was an
                               123

-------
TABLE D-l.  DIFFERENTIAL PRESSURE.TRANSDUCER  CALIBRATION*
Transducer output
(millivolts DC)
0.527
1.131
1.743
2.379
2.941
3.601
4.148
5.357
5.80
8.90
12.2
15.1
18.1
21.2
24.4
27.5
30.7
Differential
(in Hg) or
0.20
0.40
0.60
0.80
1.00 .
1.20
1.40
1.80
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
10.00
pressure
(mm Hg)
in "t A
J-U . J4
20.68
31.02
41.37
51.71
62.05
72.39 r
QT no ... .
y j . u o
103 4
155.1
206.8
258.5
310.2
361.9
413.6
465.3
517 14



1
correlation =
slope =
intercept =
. t

correlation =
slope =
intercept =
t


0.99991
0.0584
-0.0667
0.99994
0.0599
-0.3764
 *Manufacturer:   Statham Instruments, Inc
  Model #PL-295TC-15-350 Serial #316.
                            124

-------
  TABLE  D-2.  MASS FLOW TRANSDUCER CALIBRATION*
Transducer output
    (volts DC)
                   Actual air flow
                   (STP,  cm  min   )
0.005
1.250
2.500
3.750
5.000
2.500
0.006
0
12.479
24.878
37.295
49.909
24.977
0
*Manufacturer:  Brooks Instrument Division
                 Emerson Electric  Company
 Model  #5810-1C214A   Serial #7804 H52092
         AIR FLOW
         CORRECTION
         (cm3 min"')
 03

 0.2

 0.1

  0

-0.1

-0.2

-0.3
                         12    3   4
                        INDICATED AIR FLOW
                        (cm3 min"1 = VOLTS DCS-IO)
                        5 VOLTS DC
    Figure  D-l.  Mass  flow transducer calibration
                        125

-------
TABLE D-3.
CALIBRATION OF  HP-3465A DIGITAL MULTIMETER
      SERIAL  1521AO0452
Standard
(mv)
0.000
5.000
10.000
15.000
20.000 ,
25.000
30.000
35.000
40.000
45.000
50.000
55.000
60.000
65.000
70.000
,75.000
0.06 •
0.04 •
0.02 '
VOLTAGE ^^^^*.- 	 _.
CORRECTION ° >^ ® ®^< ^ xv
(mv) -0.02 - " ^>~^-^-
-0.04 -
-0.06 -....,
O 25
Meter
reading .
(mv)
0.00
5.00
10.00
15.00
20.00
25.01
30.01
35.01
40.01
45.01
50.01
55.01
60.01
65.02 •
70.02
, 75.02 . . .






• i .... j
5O 75
                         INDICATED VOLTAGE, mv
 Figure D-2.  Calibration  of transducer readout  instrument.
                             126

-------
TABLE D-4. REACTOR' TEMPERATURE MEASUREMENTS
  Distance from
reactor centerline
       (cm)
 Measured temperature (°K)
@ 599*      @.903*   @ 1198*
-6
-4
-2
0
+ 2
+4
+ 6
597
597
598
599
599
599
599
901
902
903
903
.,,903
903
' 902
1195
1196
1197
1198
1198
1198
1197
*Each of 3  zones of  the  furnace  had  the  same dial  setting
TABLE D-5.  OMEGA K-l THERMOCOUPLE  CHROMEL-ALUMEL
           Actual
         temperature
 Temperature
 deviation*
400
600
800
950
-0.75
-1.79
-2.24 :.; .
-0.90
*When deviation is plus, add to  actual  reading,
 When negative, subtract from  actual  reading.
                          127

-------
Omega Model 250 digital temperature indicator.  The calibration
of this instrument is presented in Table D-6 and Figure D-3.

     Again, there are many temperature measurements made in the
course of a single experiment with the TDAS.  With the exception
of the T2 measurement, these temperature measurements are
selectively readout on a single Omega Model 175 digital
temperature indicator.  The calibration of this particular
instrument is presented in Table D-7 and Figure D-4.
                               128

-------
TABLE  D-6.
CALIBRATION OF OMEGA DIGITAL TEMPERATURE
INDICATOR-MODEL  250, TYPE K
Standard
(°C)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
(mv)
4.095
8.127
12.207
16.395
20.640
24.902
29.128
33.277
37.325
41.269
45.108
48.828
52.398
54.125
Meter
reading
(°C)
1
101
200
302
399
501
602
702
801
901
. 1002
1103
1202
1300
    TEMPERATURE
    CORRECTION
  0
              -3
              -6 -
                 O    3OO         6OO     9OO     I2OO   °C

                           INDICATED  TEMPERATURE


   Figure D-3.   Calibration of T2 readout instrument,
                             129

-------
 TABLE D-7.
CALIBRATION OF OMEGA DIGITAL  TEMPERATURE
INDICATOR MODEL 175, TYPE K
(°c)
-200
-100
0
100
200
300
400
500
600
700
800
900
1000
Standard
(mv)
-6.891
-4.553
0.000
4.095
8.137
12.207
16.395
20.640
24.902
29.128
33.277
37.325
41.269
Meter
reading
(°C)
-197
-101
0
100
200
298
398
498
597
696
797
898
998
            6
            4

            2
  TEMPERATURE _
  CORRECTION  °

           -2

           -4

           -6
                 -200   0    200    400    600   800  1000
                         INDICATED  TEMPERATURE
Figure  D-4.  Calibration of  temperature  selector readout
              instrument.
                            130

-------
                            APPENDIX E

     PREPARATION OF GAS CHROMATOGRAPHIC OPEN TUBULAR COLUMNS
 ^h(a 4-^       °f .chemical compounds can be formed upon
 the thermal decomposition of an organic material.  Therefore
 to analyze these complicated mixtures, high-resolution gas  '
 chromatographic columns are generally needed.

      Before a sample is subjected to tests by the TDAS , it is
 ?Shn?J    3 Vari°US SC'S that Contain high-resolution glass open
 tubular columns.  And of course, the separation columns that are
 used in- the TDAS must be capable of separating complex mix?urS?

      There are a number of procedures for preparing high-
 resolution OTG's.  The preparation procedures that follow are
 those currently used in our laboratory for, fabricating OTC ' s .
 These preparation procedures have evolved over a long period of
 time _ and _ are Continually being updated.   Our experience with
             anS USe OTC'S dates back to  the early 1960 's when
recent     '^^^^E sSf a^cleaT
advantage -for 'borosilicate  glass  (Pyrex)  over soda-lime (soft)
glass.  Specifically,  it  has  been  found  that Pyrex is  a better
tubing material  for high-temperature work,  as there is less
stationary --phase bleed.   The  increased phase bleed with soft
glass has been attributed to  the composition of  that material
having a eatalytic effect on  the thin film  of stationary phase.
Consequently, at the present  time, Pyrex  is used almost
exclusively  as the glass  tubing stock for, our OTC's.

     OnceUithe1 anticipated dimensions of the finished OTC have

     SS   2 UP°n' ^ S±Ze °f the glass  twbin9  stock  ^ be
selected.  For example, if the finished bore, size  of the OTC
is to be .approximately 0.25 mm ID, then a; 'stock  having  a 2.5 mm
T«I Ca* b6Viused with a draw ratio of 100 to  one.  Th6i Ictual
wall thickness of the drawn tubing, is of  little  importance, as
  SS                                       a -i§e variety
                              131

-------
     After selecting the glass stock, it is necessary to clean
the glass tubing before drawing.  Some workers resort to cleaning
with acids; however, it has been our experience that a simple
cleaning with distilled water followed by acetone and drying
with nitrogen gas is sufficient.  The clean glass stock is next
imSrted in?o" ?he glass drawing machine  (Shimadzu Model GDM-1)
anfSSseguently drawn to capillary bore diameters and tubing
lengths of up to 150 meters.

     After the capillary tubing has been drawn, it is necessary
to determine She actua/bore size of the finished tubing   This
is accomplished by placing small sections of the tubing in a
Nikon optical microscope, Model 43622.  With the use of an
optiSa? filar micrometer, Nikon Model 66728, the bore diameters
are measured to within ±0.01 mm ID.  The length of the coiled
capillary  tubing is determined by merely counting the coils and
multiplying by the circumference of  an  average coil of the drawn
tubing .

     At this point  in the column fabrication process, it  is
necessary  to select the  stationary phase with which the tubing
will eventually be  coated.   The  selection  of the  stationary
ohase will govern much of -the  surface modification and de-
activating procedures that will  be applied to  the tubing  surface.
For example, if a dimethylsilicone stationary  phase  is  selected,
then surf ace roughness  is  a  relatively  minor  aspect,  and  surface
deactivation is of  major concern.  These concerns are Precisely
reversed if a  relatively polar stationary phase  is  to be  coated
on the tubing  wall.

      In our laboratory,  the  non-polar stationary phases are
 applied to tubing materials  that have been treated  with HC1 gas
 at elevated temperatures.   This is to etch or leach the surface
 of the glass.   Specifically, this procedure involves f^l^g
 the tube with HC1 gas and then sealing the ends with a torch
 Thl tubing is then heated in an oven at 380°C for approximately
 15 hours.  After removal from the oven and cooling to room
 temperature, the tube is vented and immediately flushed with
 nitrogen gas.   Next, the tubing is mounted in the solvent
 cleaning apparatus shown in Figure E-l.  It is then flushed with
 l-mrgSIntities of methanol and methylene chloride, respectively.
 After cleaning with the solvents, the  tubing is sealed and
 stored until ready for the deactivation process, which will be
 described later.

      For medium polarity and polar  stationary phases, the glass
 tubing bore is modified by a technique that produces whiskers
 of SS Interior surface.  This is accomplished by deconjpoaxng in
 the tubing a specified amount  (approximately 0.3_mm  cm  .) of
 2-chloro-l,l,2-trifluoroethyl methyl ether, or simply, etching
 ether.  After drawing this quantity of liquid into the central
 part of the tubing length,  a gentle vacuum is applied to the
                                 132

-------
Figure E-l.  Capillary cleaning apparatus'.
                    133

-------
ends of the tube to remove excess air, and the tubing is quickly
sealed with a torch.  The tubing is next subjected'to 375°C for
a period of 15 hours.  It is then vented in a hood while still
warm, and a vacuum is quickly applied to remove volatiles.  Next,
the tube is flushed with oxygen at 460°C for 15 hours in a
special oven.  After this, a whisker-textured surface is observed
on the tubing interior wall.  Figure E-2 is a scanning electron
micrograph of these whiskers.  Chemically, these whiskers are
considered to be relatively pure SiC>2.  After the flushing With
oxygen at high temperature, the ends are sealed until the tube  -
is ready for the deactivation step.

     At present, there are just two deactivating procedures used
in our laboratory.  The first involves the chemical bonding of
polyethylene glycol  (Carbowax 20M) to the glass surface.  This
is accomplished by passing through.the tube a 5%  (w/v) solution  ,
of Carbowax 20M dissolved in methylene chloride.  With a flow_of
dry nitrogen gas, the solvent is evaporated and the tubing dried.
The tube is then sealed while containing a nitrogen gas atmo-
sphere.  Next, the tubing is heating for 15 hours at 280°C.
After this thermal treatment, the tubing is removed front the oven
and flushed sequentially with nitrogen gas and 5 ml quantities
each of methanol and methylene chloride.

     The second deactivation procedure involves the chemical
N-cyclohexyl-3-azetidinol (CHAZ).  This material is dissolved
in methanol on a 5%  (w/v) basis and is passed through the glass
tubing.  After removing the solvent and drying with a flow of
nitrogen gas, the ends of the tube are sealed.  It is then
heated for 15 hours at 125°C and followed by the same thorough
flushing and clean-up procedures as mentioned earlier with the
Carbowax 20M.  These deactivation procedures are applied to
both the HCl-treated Pyrex tubing and whisker wall tubes.

     The next step in the OTC preparation procedure is_very
important.  It involves mounting the deactivated tube in an
oven with a special  "X" tubing connection that permits the
heating of the glass tube interior while in a stream of dry
nitrogen gas  (see Figure E-3).  This step is performed on each
tube before it is coated with stationary pha,se.  Specifically,
the tube is mounted  in the oven, and a flow of approximately 1.0
ml min"1 of nitrogen gas is established at the exit of the tub-
ing.  Next, the tubing is heated for approximately 3 hours at
150°C.  This is to remove excess water from the tubing interior.
After the tube has returned to room temperature,  it is ready for
filling with the stationary phase coating solution.  It is
important that this  filling process be accomplished quickly so
that moist atmospheric air will not back diffuse  into the dried
tubing.

     A considerable  amount of attention should be  given to the
selection of the stationary phase.  Many workers  have  found that

                               134

-------
                              lOym
Figure E-2.   Whisker-textured tubing bore
                   135

-------
                        RESTRICTOR
                        3 m by 0.25 mm
                        -w—•
             METAL CAP
           GLASS WOOL
                            VESPEL FERRULES
                                GLASS CAPILLARY
MOLECULAR SIEVE
  (LINDE 5A)
                      FILTERED
                     CARRIER GAS
Figure E-3.  Apparatus  for drying  glass tube interior,
                           136

-------
 for a thermally stable column of high efficiency,  it is  pre-
 ferable to have a gum-type stationary phase,  as  opposed  to the
 free flowing liquid phases.   This presents  some  difficulties
 with respect to the actual coating of the stationary phase.     '
 However,  it has been observed that not only is there an  improve-
 ment in efficiency at the elevated temperatures  with gum phases,
 but also,  these columns seem to have  somewhat higher thermal
 stability.   In addition,  there seems  to be  less  adsorptivity  with
 the gum-type stationary phases.   The  free-flowing  liquid phases
 do  have some advantages when one is trying  to separate volatile
 substances  at the lower chromatographic temperature  regions,  such
 as  room temperature or less.   The free-flowing coating solutions
 can^be applied by the dynamic coating technique, or  various
 modifications therefore (e.g.,  the mercury  piston  coating pro-
 cedure). However,  the gum-type stationary phases are normally
 best applied using the static coating procedures.

      In any event,  the preparation of the coating  solution is an
 important  step.   Clean and filtered solvents  are necessary.
 Also,  the mercury used for the piston should "be  free of  con-
 taminants.   The coating solution should be  freshly prepared and
 placed in a Teflon screw-capped  vial.   When hot  in use,  the
 coating solution should be stored in  the dark.

      Immediately after drying and returning to room  temperature,
 the  tubing  is  ready for application of  the  coating solution..
 For  the dynamic  coating procedure,  approximately 10  percent of•
 the  tubing  length  is  filled with the  coating  solution and
 immediately followed  with  about  a 4 cm  length of mercury,  which
 acts  as a piston.   The tubing is filled by  applying  a partial  '
 vacuum at the  exit  and with the  use of  a rubber suction  bulb
 connected to  the capillary by silicone  rubber tubing.  The inlet
 end  of the  tubing  is  immersed in the  coating  solution, which
 readily fills  the  first 10 percent  of the tubing length.   Then a
 short  length of  clean mercury is  pulled into  the tube in the  same
 manner.  The  tubing  is  next placed  in the horizontal  position.
 Silicone rubber  tubing is  attached  to the tubing entrance, and
 a nitrogen  gas pressure is applied  that will  push  the liquid
 coating solution through the  tube at  an average velocity:of
 approximately  0.5 cm  sec-1.  At  the exit end  of the  tubing, a
 short  length of  the  same size capillary glass .tubing  is applied
 so that the solution  coating  plug will  not be accelerated  upon
 emerging from the main  capillary.  After this process, the
 solvent is  evaporated by flowing nitrogen gas through the-tube
 for approximately 1 hour.  The .tube is  then ready  for condition-
 ing.   This  conditioning step will be described later.

     For the static coating technique,  the entire tube is  filled
with a relatively dilute degassed solution consisting of
 stationary phase in a volatile solvent  (e.g., pentane or
methylene chloride).  It is then mounted on a rack, and a  seal is
made at one end of the tubing by pulling into the capillary bore
                               137

-------
Apiezon "N".  After the seal has been secured, the tubing
assembly is mounted in a chamber located within a double-door
cupboard in the temperature-controlled laboratory.  Flexible
tubing from a mechanical vacuum pump is attached to the open end
of the filled tube, and the solvent is gently removed from the
tubing bore, thus depositing a known and even film thickness of
stationary phase along the tubing length.  This static coating
technique is a time-consuming process and can involve many days,
depending on the length and bore diameter of the capillary
tubing.                                        ''

     After the coating of the glass tube, it is mounted in the
same oven arrangement as shown in Figure E-3 that was employed
for the drying of the glass capillary tubing.  The capillary is
then purged with nitrogen gas as it is .gradually heated to its
elevated conditioning temperature.  The conditioning time period
for a newly prepared OTC is usually 15 hours.

     After the conditioning of a column, it is mounted in a GC
and evaluated with respect to its chromatographic performance.
If the column meets established criteria of column efficiency,
shape of elution profiles, and thermal stability,.it is removed
from the GC, and immediately the ends are sealed .with a..micro
torch.  The column is then stored on a special rack contained
in a darkened cabinet.  The column will remain stored in this
manner until ready for use on a routine basis in one- of the
GC's, or in the GC that is an integral part of the ,TDAS .v
                                138

-------
                            APPENDIX P  -    .

          APPLICATION OF THERMAL FOCUSING AND PROGRAMMED
            TEMPERATURE GAS CHROMATOGRAPHY IN THE TDAS


     Appropriate  application of temperature  programming
 techniques  is  essential to the operation of  the  TDAS.  Programmed
 increases of temperature with respect to time are applied
 sequentially to the  sample insertion chamber,  the effluent
 collection  trap,  and eventually the gas  chromatographic column.

     The  transfer of thermal decomposition products  from the
 collection  trap and  into the gas chromatographic  column is a very
 important step and one that is  crucial to the  success of experi-
 ments conducted with the TDAS.  This vitally  important  transfer
 is the subject of this appendix.

    . Chromatography  is basically a  dynamic separation process  -•
 that_uses the  concept  of differential migration.  However, to
 obtain chromatographic separations,  it is  required that the un-
 separated mixture be deposited  in a  small  locale  at  the start
 of the chromatographic process.  Therefore, in the case of a
 mixture of widely different  chemical compounds, which are emerg-
 ing _in somewhat random order  from an effluent collection trap,
 it is necessary that the condensable compounds be concentrated
 in the inlet region of  the gas  chromatographic column.   If the
 multicomponent sample  can be  placed within a low  temperature
 region of a gas chromatographic column, it can subsequently be
 efficiently eluted,  provided  it is initially located in  the
 column as a narrow longitudinal zone.

     This thermal focusing process can be explained in  funda-
mental migration  terms.  Specifically, the axial location of the
centroid of like molecules migrating through a gas chromato-
graphic column can be expressed as
                            /•fc     ,

                        =   /   Vm  CT
                           Jn
dt
                      (Fl)
                              139

-------
where z is the distance along the column axis, t is time, vm is
mobile phase velocity, and k is the partition ratio.  The
localized velocity of the migrating solute zone can then be
written as
                      vs  =
        dz
        dt
                                 =  v
m Vl+k
                                                              (F2)
Even if k remains constant through the column length, it has
been determined that the solute zone velocity experiences a
very gradual increase as the zone migrates through the column.
This small increase is due to the decompression of the carrier
gas since
                                   v
                 v
                  m
=  ftp)   =
                                                              (F3)
where p  is pressure, p0  is outlet pressure, ve is_mobile phase
outlet velocity, K  is column permeability, and n is the carrier
gas viscosity.  However,  in usual GC practice, vm does not change
by more  than  a  factor of two or  three over the length of the
column.           t

     Now if k is initially very  large and then diminishes with
time, as in PTGC where  it can  eventually approach zero, it is
seen from equation  F2 that vs  experiences a many-fold increase
with the passage of time. Therefore, if the  given  solute can
be deposited  in the column inlet while. having a  corresponding
initial  k or  k. such that
                            k.   »
                                         (F4)
 where ke is the average partition ratio upon elution,  it is   _
 evident that the initial vs is very small indeed.   Thus  the  time
 period over which the solute zone is deposited is  of small con-
 sequence.  If ki is sufficiently large, the sample depositing
 time can be tens of minutes.

      To capture a solute zone as a narrow distribution in the
 inlet portion of the chromatographic column, it is necessary
 that there be a pronounced longitudinal temperature gradient
 between the heated inlet block and the effective inlet portion
 of the chromatographic column.  For this in-line thermal
 focusing of solute molecules to be efficient, it is important
 that the thermal gradient between the heated inlet block_and
 the effective region of the column inlet be an even gradient.
 That is, there can be no intermediate cold spots or changes  in
 the sign of the gradient.  If this occurs, the chromatographic
 performance will definitely be affected, both qualitatively  and
 quantitatively.
                                140

-------
     The selection of the liquid phase used for thermal focusing
and subsequent PTGC is also important.  A good liquid phase
should have a broad liquid range.  It should have low viscosity
at the initial temperature of the program, and it should also
have good thermal stability at the elevated final temperature
At no time should the stationary phase be allowed to solidify
because the solute will then be trapped by adsorption and not
liquid partitioning.   As a result of such capturing of sample,
the resulting chromatographic peaks will be distorted.  Also
the retention characteristics will be unknown, since the com-
ponents experience undetermined extents of both adsorption and
liquid partitioning during their, subsequent migration.
                             141

-------
r
                                     APPENDIX G

                        APPROXIMATE COSTS OF TDU COMPONENTS
               A number of companies have expressed an interest in the
          TDAS and its capabilities.  Representatives from the following
          companies have visited the Research Institute facilities at the
          University of Dayton and observed,the TDAS in various stages of
          its development:

                    Acurex Corporation
                    Babcock & Wilcox Company
                    Dow Chemical Company                      . .
                    Goodyear Tire & Rubber  Company
                    Tennessee Eastman Company
                    TRW Incorporated
                    Shell Development Company
                    Union Carbide Corporation

               There have been many inquiries about the cost  of assembling
          a TDAS, particularly the costs of the various components that
          constitute the TDU.  For this reason, this appendix was compiled
          showing a breakdown of the approximate  costs  (estimated 1978
          prices) of the TDU components.
           Gas purification filters 	 	
           Pressure regulators,  flow regulators,  and flow
             control valves
           Tubing and tube fittings 	
           Temperature programmed power supply 	
           Digital voltmeter and mounting accessories
           Pyrolysis probe power supply, probes,  and
             accessories
           Console,  instrumentation panels,  name plates,  etc,
           Shielded thermocouple wire, thermocouple selector
             switch, and terminals
           Digital temperature readout devices	
           Power switch boxes 	•
           Gas switching valves . . ..	•
           Variacs, powerstats, and accessories	
           Pressure transducer and associated electronics ,
           Mass flow transducer and associated electronics
           Electrical terminals, switches, and wiring
           Materials and machining of support brackets
             for sample inlet device
 395

 640
 480
 970
 650

1300
 680

 390
 510
 130
 360
 115
 610
 645
  75

 400
                                         142

-------
Fabricated heating blocks and cartridge heaters            s
Heating tapes  	           	
Quartz stock  	!.."."!!.!..!!..	*	
Quartzware fabrication costs  . ] .........'..'.'.. '.	
High-temperature, 3-zone furnace and'control "console ".'.'.'.
Rigid table for furnace and reactor  	
Glass insulating tape and adhesive glass"tape""	
Vespel ferrules 	    " "	
High temperature insulation ...........     "	
Trapping media	...]...	
Liquid nitrogen reservoir	
Glass-lined stainless steel tubing 	'.....	
Valve oven	^ ^ _  	
Support clamps and associated hardware .'.......'..	
Calibrated sheathed thermocouple ,	    	
Total estimated costs*	 •
*Minus assembly costs.
                              143

-------
1. REPORT NO.
  EPA-600/2-80-098
                             TECHNICAL REPORT DATA    .
                       (Please read Instructions on the reverse before completing)
                         2.
                                                  3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
 DESIGN CONSIDERATIONS FOR A THERMAL
 DECOMPOSITION ANALYTICAL SYSTEM
                                                   RE
          August 1980  (Issuing Date]
          6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
  Wayne A. Rubey
                                                  8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 University of Dayton Research Institute
 300 College Park
 Dayton, OH  45469
           SOS#4  Taskll   A73D1C
          11. CONTRACT/GRANT NO.

            R805117-01-0
12. SPONSORING AGENCY NAME AND ADDRESS              Gin. , OH
  Municipal Environmental Research Laboratory—
  Office of Research and Development
  U.S. Environmental Protection Agency
  Cincinnati,  OH  45268	
          13. TYPE 01
          14. SPONSORING AGENCY CODE

            EPA/600/14
15. SUPPLEMENTARY NOTES

  Richard A.  Carnes (Project Officer)   513/684-4303
16'ABSTRCoTntrolled high-temperature incineration  is  one of the most pro-  _
 raising methods for obtaining  safe and permanent disposal of highly  toxic
 organic wastes.  A sophisticated laboratory  system has been designed and
 assembled  to provide fundamental thermal decomposition data on a wide
 assortment of organic materials.  This thermal decomposition analytical
 system  (TDAS)  is a specially  designed, closed, continuous system_con-
 sisting of a versatile and  fully instrumented  thermal decomposition unit
 Which is connected to dedicated in-line gas  chromatography, mass spec-
 trometry,  and data reduction  equipment.  The TDAS has been designed to
 generate fundamental thermal  decomposition data rapidly, economically,
 and safely.  These laboratory data should be instrumental in
 establishing acceptable  criteria for controlled high-temperature
 incineration of toxic organic wastes.
 17.
                           KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
                                       b.lDENTIFIERS/OPEN ENDED TERMS
  Organic Compounds
  Degradation
  Temperature
  Detoxification
  Recombination Reactions
  Research
 Retention Time
 Excess Air
13B
 18. DISTRIBUTION STATEMENT
     Release to Public
19. SECURITY CLASS (ThisR

 UNCLAS SIFIED
  158
                                        20. SECURITY CLASS (Thispage)

                                         UNCLASSIFIED	
                                                              22. PRICE
 EPA Form 2220-1 (9-73)
                                      144
                                                  •h U.S. GOVERNMENT PRINTING OFFICE: 1980--657-165/0054

-------

-------
r

-------

-------
                                                                                       •T) O
                                                                                       o> -•.
                                                                                       3 =!:
                                                                                       OJ O
                                                                                        r-i ^

                                                                                        CD
m
-o
CTl
O
O

ro

c»
o
 i
o
IQ
CO
C O. =5


ill


£?*
HI
m ° ~*



  fi) O
                 ssl
                      i
li
3-3
o |

5 3
           sf
           3 01
           3 |
           C S
           s I
           5 S.
                                                                                        03
                                                                                        O
                                                                                               > me
                                                                                                       3  =!' CD
                                                                                                       O  O Q.


                                                                                                       "*  3 w
                                                                                                          D >-«
                                                                                                          CD 0)

                                                                                                          5 *
                                                                                                          0) CO
                                                                                                          CD
                                                                                                          O
                                                                                                       O T  O
                                                                                                       =• 5, CD
                                                                                                       2 O
                                                                                                       fj -I  *
                                                                                                       cn
                                                                                                       10
                                                                                                       cr>
                                                                                                       00
                                                                                                             D
                                                                                                             <
                                                                                                    3
                                                                                                    CD
                                                                                                    D
                                                                                                             5"
                                                                                                             en
                                                                                                             CD
                                                                                                             0)

                                                                                                             i-i
                                                                                                m > TI m -n TO!
                                                                                                TJco  n ^  5 pi
                                                                                                >
-------