United States Atmospheric Research and Exposure
Environmental Protection Assessment Laboratory
Agency Research Triangle Park, NC 27711
Research and Development OCTOBER, 1991
jl —•*• TECHNICAL ASSISTANCE DOCUMENT
TECHNICAL ASSISTANCE DOCUMENT FOR
SAMPLING AND ANALYSIS OF OZONE PRECURSORS
-------
Technical Assistance Document for
Sampling and Analysis of Ozone Precursors
Mr. Larry J. Purdue
Atmospheric Research and Exposure Assessment Laboratory
Human Exposure and Field Research Division
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27709
Mr. Dave-Paul Dayton
Ms. Joann Rice
Ms. Joan Bursey
Radian Corporation
P.O. Box 13000
Research Triangle Park, N.C. 27709
October 31, 1991
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approver for publication. Mention of trade names
or commercial products does not constitute endorsement
or recommendation for use.
-------
ABSTRACT
This document contains guidance and discussion on methods applicable to the
proposed revisions to Title 40 Part 58 of the Code of Federal Regulations. The
proposed revisions pertain to the enhanced monitoring of ozone precursors and
meteorological monitoring. The precursors addressed include volatile organic
compounds, carbonyl compounds, oxides of nitrogen, and total reactive oxides of
nitrogen. The meteorological parameters include surface meteorology and upper air
meteorology. The document is structured in a document control format to
accommodate expected revisions due to the emerging nature of the technologies
discussed. The primary users of this document are expected to be Regional, State,
and local Environmental Protection Agency personnel addressing the new enhanced
ozone ambient air monitoring provisions.
The document consists of the following sections and appendices:
(1) Introduction
(2) Methodology for Measuring Volatile Organic Ozone Precursors In
Ambient Air
(3) Methodology for the Determination of Total Nonmethane Organic
Compounds In Ambient Air
(4) Methodology for Measuring Oxides of Nitrogen and Total Reactive
Oxides of Nitrogen In Ambient Air
(5) Methodology for the Determination of Carbonyl Compounds In Ambient
Air
(§) Meteorological Monitoring
(7) References
(A) Discussion, Issues, and Selected Procedures Related to Canister
Sampling
(B) List of Materials, Equipment, and Vendors
in
-------
PREFACE
The United States Environmental Protection Agency proposes to revise the air
quality surveillance regulations in Title 40, Part 58 of the Code of Federal Regulations
to include provisions for enhanced monitoring of ozone and oxides of nitrogen, and for
additional monitoring of volatile organic compound and meteorological parameters.
The proposed revisions are in accordance with congressional mandates set forth in
the Clean Air Act Amendments of 1990.
This technical assistance document has been prepared to provide direction on
sampling and analysis methodology to Regional, State, and local Environmental
Protection Agency personnel during enhanced monitoring planning, implementation
and operation activities.
This document does not circumvent the need for highly skilled technical
personnel who understand and can perform the procedures described. It should not
be used as a standard .operating procedure.
IV
-------
Revision 0
OCTOBER 1991
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1-1
1.1 Purpose 1-2
1.2 Organization 1-2
1.3 Summary of the Monitoring Regulations 1-3
2.0 METHODOLOGY FOR MEASURING VOLATILE ORGANIC OZONE
PRECURSORS IN AMBIENT AIR 2-1
2.1 Identification of Target Volatile Organic Ozone
Precursors 2-2
2.2 Chromatography Discussion and Issues 2-4
2.2.1 Gas Chromatography with Flame lonization Detection 2-4
2.2.2 Identification Issues 2-6
2.2.3 Moisture Issues 2-7
2.2.4 Detector Types 2-8
2.2.5 Calibration Standards 2-9
2.2.5.1 Primary Calibration Standard 2-9
2.2.5.2 Retention Time Calibration Standard 2-9
2.2.6 Analytical System Calibration 2-11
2.2.7 Column Selection and Configuration 2-13
2.2.7.1 Column Recommendations 2-14
2.2.8 Data Quality Objectives 2-19
2.3 Manual Method for Collecting and Analyzing Volatile
Organic Ozone Precursor Samples 2-20
2.3.1 Sampling 2-20
2.3.1.1 Canister Sample Collection 2-21
2.3.2 Analysis 2-28
2.3.2.1 Analytical Equipment and Configuration 2-28
2.3.2.2 Analtyical System Operation 2-31
-------
Revision 0
OCTOBER 1991
TABLE OF CONTENTS
Continued
Page
2.4 Automated Method for Collecting and Analyzing Volatile
Organic Ozone Precursor Samples 2-33
2.4.1 Method Description • • • 2-33
2.4.1.1 Analytical Equipment and Configuration 2-34
2.4.1.2 Operation 2-39
2.5 Gas Chromatography/Mass Spectrometry 2-45
2.5.1 Identification Confirmation 2-46
2.5.1.1 Use of Selected Ion Monitoring Techniques 2-47
2.5.2 Equipment 2-48
2.5.3 Interferences 2-48
2.5.4 Standards 2-48
2.5.4.1 Instrument Performance Check Standard 2-49
2.5.4.2 Calibration Standards 2-50
2.5.4.3 Internal Standard Spiking Mixture 2-50
2.5.5 Instrument Operating Conditions • 2-50
2.5.5.1 Sample Concentration Conditions 2-50
2.5.5.2 Desorption Conditions 2-51
2.5.5.3 Gas Chromatographic Conditions 2-51
2.5.5.4 Mass Spectrometer 2-51
2.5.5.5 Calibration 2-52
2.5.6 Analysis Procedures 2-54
2.5.6.1 Vaporization of Volatile Organic Compounds .... 2-55
2.5.6.2 Initiation of Analysis 2-55
2.5.6.3 Initial Review of Data 2-55
2.5.6.4 Sample Dilution 2-56
VI
-------
Revision 0
OCTOBER 1991
TABLE OF CONTENTS
Continued
Page
2.5.7 Qualitative Analysis 2-56
2.5.8 Quantative Analysis 2-58
2.5.9 Additional Uses 2-58
3.0 METHODOLOGY FOR DETERMINING TOTAL NONMETHANE ORGANIC
COMPOUNDS IN AMBIENT AIR 3-1
3.1 Method Description 3-1
3.2 Summary of Method TO-12 3-3
3.3 Significance 3-4
3.4 Interferences 3-4
3.5 Equipment 3-4
3.5.1 Direct Air Sampling 3-4
3.5.2 Remote Sample Collection in Pressurized Canisters 3-4
3.5.3 Sample Canister Cleaning 3-6
3.5.4 Analytical System 3-6
3.6 Reagents and Materials 3-9
3.7 Direct Sampling 3-9
3.8 Sample Analysis 3-10
3.8.1 Analytical System Leak Check 3-10
3.8.2 Sample Volume Determination 3-10
3.8.3 Analytical System Dynamic Calibration 3-11
3.8.4 Analysis Procedure 3-13
3.9 Performance Criteria and Quality Assurance 3-17
3.9.1 Standard Operating Procedures 3-17
3.9.2 Method Sensitivity, Accuracy, Precision, and
Linearity 3-18
3.10 Method Modifications 3-19
3.10.1 Sample Metering System 3-19
3.10.2 Flame lonization Detection System 3-19
3.10.3 Range 3-19
VII
-------
Revision 0
OCTOBER 1991
TABLE OF CONTENTS
Continued
Page
4.0 METHODOLOGY FOR MEASURING OXIDES OF NITROGEN AND TOTAL
REACTIVE OXIDES OF NITROGEN IN AMBIENT AIR 4-1
4.1 Oxides of Nitrogen 4-1
4.1.1 Measurement of Oxides of Nitrogen - 4-1
4.1.1.1 Method Description 4-1
4.1.1.2 Methods and Equipment 4-2
4.2 Total Reactive Oxides of Nitrogen . 4-3
4.2.1 Measurement of total Reactive Oxides of Nitrogen 4-3
5.0 METHODOLOGY FOR DETERMINING OF CARBONYL COMPOUNDS
IN AMBIENT AIR 5-1
5.1 Method Description 5-1
5.2 Summary of Method .5-3
5.3 Significance 5-4
5.4 Interferences 5-5
5.5 Equipment 5-5
5.6 Reagents and Materials '5-10
5.7 Preparation of Reagents and Cartridges 5-11
5.7.1 Purification of DNPH 5-11
5.7.2 Preparation of DNPH-Formaldehyde Derivative . 5-15
5.7.3 Preparation of DNPH-Formaldehyde Standards 5-16
5.7.3.1 DNPH Coating Solution 5-16
5.7.3.2 Coating the Cartridges 5-17
5.8 Sampling 5_ig
5.9 Sample Analysis 5.23
5.9.1 Sample Desorption 5.24
5.9.2 HPLC Analysis 5.24
5.9.3 HPLC Calibration 5.27
VIII
-------
Revision 0
OCTOBER 1991
TABLE OF CONTENTS
Continued
Page
5.10 Calculations 5-30
5.11 Performance Criteria and Quality Assurance 5-32
5.11.1 Standard Operating Procedures 5-32
5.11.2 HPLC System Performance 5-33
5.11.3 Process Blanks 5-33
5.11.4 Method Precision and Accuracy 5-34
5.12 Detection-of Other Carbonyl Compounds 5-34
5.12.1 Sampling Procedures 5-34
5.12.2 HPLC Analysis 5-34
5.13 Ozone Interferent 5-36
5.13.1 Equipment 5-36
5.13.2 Procedure 5-38
5.14 Alternative Substrate 5-38
5.14.1 Advantages of C18 Substrate 5-38
5.14.2 Disadvantages of C18 Substrate 5-38
6.0 METEOROLOGICAL MONITORING 6-1
6.1 Measurement - Surface Meteorology 6-1
6.1.1 Equipment 6-2
6.1.2 Procedure 6-3
6.2 Upper Air Meteorology - Mixing Heights 6-3
6.2.1 Measurements 6-6
6.2.1.1 Direct Measurements 6-6
6.2.1.2 Vertical Temperature Profile Measurements 6-7
IX
-------
Revision 0
OCTOBER 1991
TABLE OF CONTENTS
Continued
Page
6.2.2 Mixing Height Measurements 6-7
6.2.2.1 Doppler SODAR Measurement Approach ........ 6-7
6.2.2.2 Vertical Temperature Profile Measurement
Approach 6-11
7.0 REFERENCES 7-1
-------
Revision 0
OCTOBER 1991
TABLE OF CONTENTS
Continued
APPENDIX A - DISCUSSION, ISSUES, AND SELECTED PROCEDURES
RELATED TO CANISTER SAMPLING
Page
A1.0 Canister Sampling Issues A-1
A2.0 Precautions in the Use of Canister A-1
A2.1 Contamination A-1
A2.2 Sample Stability A-3
A2.3 Canister Leakage A-4
A2.4 Historical Canister Log A-6
A3.0 Canister Cleaning A-6
A3.1 Equipment A-7
A3.2 Cleaning Procedure A-11
A3.3 Blanking Procedure A-13
A3.4 Final Evacuation Procedure A-14
A4.0 Canister Sampling System Certification A-14
A4.1 Equipment A-18
A4.2 Certification Procedure A-19
A5.0 Sample Stability A-21
A5.1 Positive Pressure Samples A-22
A5.2 Diluted Samples A-22
A5.3 Repeated Analyses A-22
APPENDIX B LIST OF MATERIALS, EQUIPMENT, AND VENDORS
XI
-------
Revision 0
OCTOBER 1991
LIST OF FIGURES
Page
1-1 Isolated Area Network Design "|-6
2-1 Target Volatile Ozone Precursors 2-3
2-2 Example Chromatogram for J&W® DB-1 Capillary
Analytical Column . 2-16
2-3 Example Chromatogram for Chrompack® AI2O3/KCI PLOT
Capillary Analytical Column 2-17
2-4 Automated Multi-event Canister Sampling System 2-22
2-5 Single-event Canister Sampling System 2-24
2-6 Sampling Manifold Assembly 2-36
3-1 Schematic of Analytical System for NMOC -
Two Sampling Modes 3-5
3-2 Cryogenic Sampling Trap Dimensions 3-7
3-3 Construction of Operational Baseline and Corresponding
Correction of Peak Area 3-16
5-1 Formation of a Stable Derivative 5-2
5-2 Typical HPLC System 5-7
5-3 Typical Sampling System Configurations . 5-8
5-4 Special Glass Apparatus for Rinsing, Storing, and
Dispensing Saturated DNPH Stock Solution 5.9
5-5 Impurity Level of DNPH After Recrystallization 5.13
5-6 Sampling Data Sheet 5_2^
5-7 Chromatogram of DNPH Formaldehyde Derivative 5-26
xii
-------
Revision 0
OCTOBER 1991
LIST OF FIGURES
Continued
Page
5-8 HPLC Chromatogram of Varying Concentrations of
DNPH-Formaldehyde Derivative 5-28
5-9 Calibration Curve for Formaldehyde 5-29
5-10 Crossectional View of the O3 Scrubber Assembly 5-37
A3-1 Canister Cleaning System A-8
A4-1 Canister Sampling System Certification Schematic -
Zero Gas A-16
A4-2 Canister Sampling System Certification Schematic
Challenge Gas A-17
XIII
-------
Revision 0
OCTOBER 1991
LIST OF TABLES
Page
1-1 Enhanced Ozone Monitoring Network Sampling Schedule 1-5
2-1 Primary Quantitation Ions for Compounds of Interest 2-57
5-1 Sensitivity (ppb, V/V) of Sampling/Analysis for Aldehydes and
Ketones in Ambient Air Using Adsorbent Cartridge Followed
by Gradient High Performance Liquid Chromatography 5-14
XIV
-------
Revision 0
OCTOBER 1991
LIST OF ABBREVIATIONS
BFB p-bromofluorobenzene
C Centigrade
CAAA Clean Air Act Amendments
CBD central business district
CFR Code of Federal Regulations
cm centimeter
CMSA Consolidated Metropolitan Statistical Area
CRM Certified Reference Material
C18 octadecylsilane-bonded silica gel
DNPH 2,4-dinitrohpenylhydrazine
DQOs Data Quality Objectives
BCD electron capture detector
EPA Environmental Protection Agency
EKMA Empirical Kinetic Modeling Approach
F Fahrenheit
FID flame ionization detector
GC gas chromatography
GC/MS Gas Chromatography/Mass Spectrometer
GMT Greenwich Mean Time
HCI Hydrochloric Acid
Hg Mercury
HN03 Nitric Acid
HPLC High Performance Liquid Chromatography
I.D. inside diameter
L
LST
m
mg
mL
mm
MSA
MSD
liter
Local Standard Time
meter
milligram
milliliter
millimeter
Metropolitan Statistical Area
mass selective detector
xv
-------
Revision 0
OCTOBER 1991
LIST OF ABBREVIATIONS
Continued
N2
ng
NAAQS
NAMS
NIST
nm
NMOC
NOAA
NO
NO2
NOX
NO
NWS
02
O3D.
PAMS
PAN
PDFID
PID
PLOT
ppbC
ppbv
ppmC
ppmv
psig
QA
QC
RF
RRF
Nitrogen
nanogram
National Ambient Air Quality Standard
National Air Monitoring Stations
National Institute of Standards and Testing
nanometer
Nonmethane Organic Compounds
National Oceanographic and Aeronautic Administration
Nitric Oxide
Nitrogen Dioxide
Oxides of Nitrogen
Total Reactive Oxides of Nitrogen
National Weather Service
Oxygen
Ozone
outside diameter
Photochemical Assessment Monitoring Stations
Peroxyacetyl Nitrate
Preconcentration Direct Flame lonization Detection
photo-ionization detector
Porous Layer Open Tubular
parts per billion Carbon
parts per billion volume
parts per million Carbon
parts per million volume
pounds per square-inch gauge
quality assurance
quality control
response factor
relative response factor
XVI
-------
Revision 0
OCTOBER 1991
LIST OF ABBREVIATIONS
Continued
SIM selective ion monitoring
SIP State Implementation Plan
SLAMS State and Local Air Monitoring Stations
SODAR Doppler sound detection and ranging
SOPs standard operating procedures
SRM Standard Reference Material
UAM Urban Airshed Model
UHP Ultra High Purity
UV ultraviolet
VOCs volatile organic compounds
WCOT Wall Coated Open Tubular
/L/g microgram
fjL microliter
%RSD percent relative standard deviation
XVII
-------
Revision 0
OCTOBER 1991
SECTION 1.0
INTRODUCTION
The Environmental Protection Agency (EPA) is proposing to revise the ambient
air quality surveillance regulations in Title 40 Part 58 of the Code of Federal
Regulations (40 CFR Part 58)1 to include provisions for enhanced (1) monitoring of
ozone (0.3) and oxides of nitrogen (NOX), (2) monitoring of volatile organic compounds
(VOCs), (3) monitoring selected carbonyl compounds, and (4) monitoring of
meteorological parameters. The revisions require States to establish additional air
monitoring stations as part of their existing State Implementation Plan (SIP) monitoring
network. The EPA's authority for proposing the enhanced monitoring regulations are
provided for in Title I, Section 182 of the Clean Air Act Amendments of 1990 (CAAA).
The principal reasons for requiring the collection of additional ambient air
pollutant and meteorological data are: (1) the lack of successful attainment of the
National Ambient Air Quality Standard (NAAQS) for O3, and (2) the need to obtain a
more comprehensive air quality database for O3 and its precursors. Data acquired
from enhanced ambient air monitoring networks will have a variety of uses, including:
Developing and evaluating new O3 control strategies;
• Determining NAAQS attainment or non-attainment for O3;
• Tracking VOCs and NOX emissions inventory reductions;
• Providing photochemical prediction model input;
• Evaluating photochemical prediction model performance;
• Analyzing ambient air quality trends; and
• Characterizing population exposure to VOCs and O3.
1-1
-------
Revision 0
OCTOBER 1991
1.1 PURPOSE
The purpose of this document is to support the proposed revisions to the
40 CFR Part 58. The document provides technical information and guidance to
Regional, State, and local Environmental Protection Agencies responsible for
measuring O3 precursor compounds in ambient air. Sampling and analytical
methodology for speciated VOCs, total nonmethane organic compounds (NMOC) and
selected carbonyl compounds (i.e., formaldehyde, acetaldehyde and acetone) are
specifically addressed. The document also addresses methodology for measuring
NOX, discusses issues associated, with total reactive oxides of nitrogen (NOy), and
provides technical direction for measuring the meteorological parameters prescribed
by the revised regulations.
The technical guidance provided for measuring volatile organic O3 precursors is
based on emerging and developing technology. Guidance for automated applications,
in particular, is based to a significant extent on experience obtained from the
application of this technology during an O3 precursor study conducted by the EPA in
Atlanta, Georgia, during the summer of 19902. Because these methods are based on
emerging technology and reflect current state of the art, they will be subject to
continuing evaluation, and improvements or clarifications are anticipated in the future.
Users should consider this guidance a basic reference to assist in developing
and implementing their O3 precursor monitoring program. The method descriptions
are generic in approach and should not be considered a set of final standard
operating procedures (SOPs). The document is prepared in a document control
format to accommodate revisions that are anticipated as the emerging technologies
develop.
1.2 ORGANIZATION
The guidance provided in Section 2 of this document addresses the
measurement of volatile organic O3 precursors and includes method descriptions for
manual and automated sample collection and analysis. Detailed discussions on
selected topics such as which volatile organic O3 precursors to measure, critical gas
1-2
-------
Revision 0
OCTOBER 1991
chromatography issues, and how canister sampling should be approached are
presented.
Section 3 discusses the measurement of NMOC using Method TO-12 from the
Compendium of Methods for Sampling and Analysis of Toxic Organic Compounds in
Ambient Air3. Although measurement of total NMOC has limited application to the
implementation of the proposed 40 CFR Part 58 requirements, it is included because
of its applications to canister cleanliness verification and potential application to
alternative monitoring strategies. The alternative monitoring strategies would involve
the use of an automated version of Method TO-12 complemented with an extensive
canister sampling and manual VOCs speciation analysis program.
Section 4 addresses the measurement of NOy and issues associated with NOV.
x y
Section 5 addresses the measurement of selected carbonyl compounds using
Compendium Method TO-11 from the Compendium of Methods for Sampling and
Analysis of Toxic Organic Compounds in Ambient Air3 and includes new information
on the interference of O3.
Section 6 provides direction on meteorological monitoring, including parameters
to be measured and approaches to measurement. A detailed discussion on
techniques for characterizing upper atmospheric conditions is also provided.
Section 7 lists references cited in this document. These references should be
consulted when more definitive information is required.
Appendix A contains detailed procedures for canister cleaning and canister
sampling system certification. Precautions in the use of canisters and on sample
stability in canisters are also presented.
Appendix B lists materials and equipment referred to in this document, along
with vendors names.
1.3 SUMMARY OF THE MONITORING REGULATIONS
The 1990 CAAA require EPA to promulgate regulations to enhance existing
ambient air monitoring networks. Existing SIP stations are identified as State and
Local Agency Monitoring Stations (SLAMS) and National Air Monitoring Stations
1-3
-------
Revision 0
OCTOBER 1991
(NAMS). The proposed new monitoring stations will be identified as Photochemical
Assessment Monitoring Stations (PAMS). The U.S. EPA is also preparing a guidance
document on enhanced O3 monitoring network design entitled Enhanced Ozone
Monitoring Network Design and Siting Criteria4 which provides assistance regarding
the number of PAMS required, and station location and probe siting criteria.
The monitoring revisions proposed by EPA include changes to 15 separate
Sections, Subparts, or Appendices of 40 CFR Part 58, and vary in complexity and
impact on State and local agencies.
The areas of the revised 40 CFR Part 58 regulations most relevant to enhanced
ambient air monitoring are operating schedules, PAMS methodology, and quality
assurance. Section 58.13 of 40 CFR Part 58 contains the operating schedule for
SLAMS and NAMS. This section would be revised to require sampling for VOCs and
carbonyl compounds on a schedule specified in Section 4.4 of Appendix D of the
revised regulations.1
Five PAMS site types are described in the regulations. The number of PAMS
required is dependent on the population of the Metropolitan Statistical Area (MSA) or
Consolidated Metropolitan Statistical Area (CMSA). The specified minimum sampling
schedules for VOCs and carbonyl compounds for each site type are presented in
Table 1-1. The sampling schedule applicable to a specific area is dependent on •
population and PAMS site types. The example of an isolated area network design
shown in Figure 1-1 identifies the location of the five PAMS site types referred to in
Table 1-1. The PAMS site types are described below.
Type 1 PAMS characterize upwind background and transported O3 precursor
concentrations entering the MSA or CMSA. Type 2 PAMS characterize the type and
magnitude of precursor emissions in the MSA or CMSA at the point where O3
precursor concentrations are expected to be highest. Type 3 PAMS characterize O,
o
precursor concentrations and ratios downwind of the sources of emissions. Type 4
and Type 5 PAMS are maximum 03 concentration locations occurring downwind of
Type 2 PAMS sites.
1-4
-------
Revision 0
OCTOBER 1991
TABLE 1-1
ENHANCED OZONE MONITORING NETWORK SAMPLING SCHEDULE
Population
;: :of1vlSA/CMSA
Less than 500,000
500,000 to
1,000,000
1,000,000 to
2,000,000
More than
2,000,000
Required
Site Type
(1)
(2)
0)
(2)
(4)
(1)
(2)
(3)
(4)
(1)
(2)
(3)
(4)
(5)
Minimum VOCs
Sampling
Frequency1
A
A
A
B
A
A
B
C
A
A
B
C
A
A
Minimum Carbonyl
Compounds Sampling
Frequency1
-
D
-
E
-
-
E
E
-
-
E
E
-
-
frequency Requirements Are As Follows:
A
B
D
E
Eight 3-hour samples every third day and one 24-hour sample
every sixth day during the monitoring period.
Eight 3-hour samples, every day during the monitoring period and
one 24-hour sample every sixth day year-round
Eight 3-hour samples, every day and one 24-hour sample every
sixth day during the monitoring period.
Four 6-hour samples, every third day during the monitoring period.
Four 6-hour samples, every day during the monitoring period.
1-5
-------
Revision 0
OCTOBER 1991
U2
Central Business District
Urbanized Fringe
U1
U1 and U2 represent the first and second most predominant
wind direction during the ozone season
Figure 1-1. Isolated area network design.
1-6
-------
Revision 0
OCTOBER 1991
Appendix A of 40 CFR Part 58 would be revised to reference this guidance
document for general quality assurance recommendations for VOCs and
meteorological measurements. Specific quality assurance criteria for VOCs were not
included in Appendix A because the emerging nature of VOCs measurement
technology has not allowed development of quality assurance criteria equivalent to
those for criteria pollutants.
Appendix C of 40 CFR Part 58 would require that methods used for O3 and NOX
be reference or equivalent methods. Because there are no reference or equivalent
methods promulgated for VOCs and meteorological measurements, Appendix C of the
revisions would refer agencies to this guideline document for direction. Appendix C of
the revisions would also allow the use of approved alternative VOC measurement
methodology. This provision would require States that choose to pursue alternatives
to the methodology described herein to provide details depicting rationale and benefits
of their alternative approach in their network description as required in 40 CFR Part 58,
Section 58.4 - PAMS Network Establishment. This provision would also require that
the proposed alternative be published in the Federal Register, subjected to public
comment and subsequently approved by the Administrator of EPA.
1-7
-------
Revision 0
OCTOBER 1991
SECTION 2.0
METHODOLOGY FOR MEASURING VOLATILE ORGANIC
OZONE PRECURSORS IN AMBIENT AIR
The term VOCs generally refers to gaseous nonmethane organic compounds
that have a vapor pressure greater than 10~2 kilopascals and, for the most part, have a
carbon number in the range of C2 through C12. Many of these compounds play a
critical role in the photochemical formation of O3 in the atmosphere. Volatile organic
compounds are emitted from a variety of sources. In urban areas the dominant
source is automobiles.. This section provides two generic method descriptions for
measuring VOCs in ambient air and information and guidance to assist in the
development, implementation, and use of these methods. A number of issues are
addressed, including the target volatile organic O3 precursors to be measured; gas
chromatographic issues associated with identification and quantification of these
precursors; and a description of the methodology used for identification confirmation
by gas chromatography/mass spectrometry (GC/MS).
Measuring volatile organic O3 precursors is a complex process involving the
application of gas chromatographic techniques for qualitative and quantitative
determination of individual hydrocarbon compounds and total NMOC. Two methods
are presented for collecting and analyzing volatile organic O3 precursor samples; (1) a
manual method (see Section 2.3) involving the collection of integrated, whole air
samples for subsequent analysis at a central laboratory, and (2) an automated method
(see Section 2.4) where sample collection and analysis and data collection is
performed automatically on-site. Ideally, agencies responsible for designing,
implementing, and operating PAMS will satisfy their monitoring requirements by initially
using some combination of the manual and automated gas chromatographic
approaches, detailed in these two method descriptions. Eventually, most agencies
should primarily use the automated gas chromatograph methodology; however,
manual sampling and analysis capability will always be needed to verify the proper
2-1
-------
Revision 0
OCTOBER 1991
operation of the automated systems, characterize the quality of the collected data, and
address the identification of unknown compounds.
Because of the complexity of the measurement process and the numerous
choices of instrumentation (e.g., sampling equipment, gas chromatographs, data
acquisition hardware and software, etc.) the method descriptions in this document are
presented generically. Background information on the potential benefits and limitations
of the methods are provided. Users are ultimately responsible for equipment selection
and set-up, method development, and preparation of SOPs for their specific systems.
2.1 IDENTIFICATION OF TARGET VOLATILE ORGANIC OZONE PRECURSORS
Volatile organic O3 precursors are nonmethane organic compounds typically in
the C2 through C12 carbon number range. Figure 2-1 presents a list of typical volatile
organic O3 precursors that should be measured and reported to satisfy the
requirements of the proposed revisions to 40 CFR Part 58. Users should initially
consider these target compounds for developing their measurement systems and
monitoring approach. As experience is gained regarding the occurrence of specific
VOCs at each specific site, target compounds may be either deleted from, or added
to, the list depending on their frequency of occurrence.
The compounds listed in Figure 2-1 are presented in the order of their expected
chromatographic elution from a J&W® DB-1 dimethylsiloxane capillary analytical
column. Compounds with lower boiling points elute first on this particular analytical
column, followed by the heavier, higher molecular weight components with higher
boiling points. Concentrations of target volatile organic O3 precursors and unknown
components (unidentified peaks) are calculated in units of parts per billion Carbon
(ppbC). An estimate of the total NMOC in ppbC is made by summing all identified and
unidentified gas chromatographic peaks. The concentration in ppbC of a compound
can be divided by the number of carbon atoms in that compound to estimate the
concentration of the compound in parts per billion volume (ppbv).
2-2
-------
Revision 0
OCTOBER 1991
Acetylene
Ethylene
Ethane
Propylene
Propane
Isobutane
1-Butene
n-Butane
trans-2-Butene
cis-2-Butene
3-Methyl-1rButene
Isopentane
1-Pentene
n-Pentane
Isoprene
trans-2-Pentene
cis-2-Pentene
2-Methyl-2-Butene
2,2-Dimethylbutane
Cylcopentene
4-Methyl-1-Pentene
Cyclopentane
2,3-Dimethylbutane
2-Methylpentane
3-Methylpentane
2-Methyl-1-Pentene
n-Hexane
trans-2-Hexene
cis-2-Hexene
Methylcyclopentane
2,4-Dimethylpentane
Benzene
Cyclohexane
2-Methylhexane
2,3-Dimethylpentane
3-Methylhexane
2,2,4-Trimethylpentane
n-Heptane
Methylcyclohexane
2,3,4-Trimethylpentane
Toluene
2-Methylheptane
3-Methylheptane
n-Octane
Ethylbenzene
p-Xylene
Styrene
o-Xylene
n-Nonane
Isopropylbenzene
n-Propylbenzene
a-Pinene
1,3,5-Trimethylbenzene
1,2,4-Trimethylbenzene
/3-Pinene
Total NMOC
Figure 2-1. Target Volatile Ozone Precursors
2-3
-------
Revision 0
OCTOBER 1991
2.2 CHROMATOGRAPHY DISCUSSION AND ISSUES
2.2.1 Gas Chromatographv with Flame lonization Detection
Gas chromatography with flame ionization detection (GC/FID) is the
recommended technique for monitoring volatile organic O3 precursors in ambient air.
The sensitivity, stability, dynamic range, and versatility of GC/FID systems make
themextremely useful for measuring very low concentrations of VOCs in ambient air.
The basic components of GC/FID systems applicable to these measurements are:
• The carrier gas supply and regulation system;
• The sample concentration and injection system;
• The analytical or chromatographic separation column;
• The analytical column oven;
• The detection device; and
• The recording or integration device.
In the GC/FID technique, an air sample is taken from a canister or directly from
the ambient air, and passed through the sample concentration system. The
concentrated sample is then desorbed and injected onto the analytical column of the
gas chromatograph. The VOCs are separated by taking advantage of each
compound's distribution between the mobile phase (i.e., the carrier gas) and the
stationary phase (i.e., the solid or liquid phase coating on the analytical column). The
sample is introduced into the carrier gas stream just before it encounters the
stationary phase of the analytical column. The mixture is separated into individual
compounds based on the distribution equilibrium between the mobile and stationary
phases. The compounds then elute from the column and enter the detector. The
time of elution, or retention time, aids in identification because it is a characteristic of
each particular compound.
2-4
-------
Revision 0
OCTOBER 1991
Typically, a sample taken from an urban environment contains 100 to 150
detectable compounds in the C2 through C12 carbon range, that may be reasonably
separated into quantifiable peaks. These compounds are generally present at
concentrations varying from less than 0.1 ppbC to greater than 1000 ppbC with the
typical concentration being 0.1 to 50 ppbC. Detection of typical urban concentration
levels generally requires that samples be concentrated cryogenically in order to
selectively concentrate the compounds of interest and not the components of the
sample that are not of interest (i.e., air, moisture, and carbon dioxide). Modern GC
technology, coupled with sophisticated peak identification software, estimates the
identity and quantity of each compound to the extent that the analytical column has
been characterized and validated. The retention characteristics of the analytical
column must be determined for each target compound using pure compounds or
mixtures of pure compounds diluted with inert gas. The Flame lonization Detector
(FID) responds nearly uniformly to all nonmethane hydrocarbon compounds on a per
carbon atom basis. This uniformity of response simplifies calibration in that a single
hydrocarbon compound can be used to calibrate the detector response for all gas
chromatographic peaks. The GC/FID response is calibrated using a hydrocarbon
standard (e.g., propane) of known concentration. Compound identifications obtained
from GC/FID may be periodically verified using more definitive techniques such as'
GC/MS. (See Section 2.5 for a discussion regarding the use of GC/MS for verification
of compound identifications.)
An automated GC system developed and manufactured in Bilthoven,
Netherlands, and marketed in the United States by Chrompack® Inc., Raritan, New
Jersey, was used during the Atlanta O3 precursor study2 to obtain hourly VOC
measurements. This system was equipped with a preconcentration adsorption trap, a
cryofocusing secondary trap, and a single analytical column with a thick film liquid
phase loading. In order to adequately concentrate and speciate the light hydrocarbon
compounds with this configuration, large amounts of cryogen were required for
cooling the preconcentration trap and operating the oven of the gas chromatograph at
2-5
-------
Revision 0
OCTOBER 1991
sub-ambient temperatures. Much of the discussion that follows focuses on reducing
the consumption of cryogen and improving the performance of the automated GC
system by using alternative column(s) and column configurations.
2.2.2 Identification Issues
A GC/FID system relies primarily on retention times to make compound
identifications. Gas chromatographic retention times are subject to shifting and
interferences from non-target co-eluting compounds. Misidentification typically occurs
when the chromatogram is complex, making peak identification difficult. Under these
conditions the FID may not adequately deconvolute the peaks of interest, which
increases the probability of incorrect identification and inaccurate quantitation.
The potential for identification errors can be reduced or eliminated by:
• Using relative retention times to designate reference peaks;
• Using dual column configurations to provide improved resolution (see
Section 2.2.7 regarding column selection);
• Using specific detectors such as a mass spectrometer or a
mass-selective detector (MSD);
• Having an experienced chromatographer conduct frequent visual
inspection of the chromatograms to verify proper system operation; and
• Re-analyzing the samples on a better characterized GC system.
The effort devoted to peak identification and quantification confirmation is critical
to the quality of the collected data. Users must determine the appropriate level of
effort to be devoted to this activity, based on their specific needs and capabilities.
Expert chromatography software packages, such as MetaChrom®6, can be used to
explore, and subsequently process, large or complex chromatographic data sets to
improve qualitative and quantitative information.
In order to accommodate the large number of sample analyses required by the
proposed revisions to 40 CFR Part 58, total analytical run time must be considered.
2-6
-------
Revision 0
OCTOBER 1991
Parameters affecting analytical run time are column selection and configuration,
temperature program rate, carrier gas flow, and instrument configuration. Analytical
run time is most critical when automated methods of sampling and analysis are used
because the analytical instrumentation must be capable of completing sample
collection and analysis within a specified time period. Typical and practical analytical
run times for automated applications usually do not exceed 45 minutes. For manual
applications, the ultimate analytical run time is less critical and may be modified to
meet the user's analytical needs.
2.2.3 Moisture Issues
The effects of moisture should be considered in any monitoring program where
ambient sample preconcentration is required to increase detection sensitivity.
Cryogenic techniques are commonly used for sample preconcentration of C2 through
C12 hydrocarbons. The collection of moisture in the cryogenic trap during sample
preconcentration can cause several problems. These problems include retention time
shifting of the earlier-eluting compounds7, column deterioration, column plugging due
to ice formation, FID flame extinction, and adverse effects on adsorbent concentration
traps and some analytical detectors. If "cold spots" exist in the sample concentration
or transfer system, water can collect and cause sample carryover or "ghost" peaks in
subsequent sample analyses. This carryover may affect the data by causing
chromatographic interferences which affect the resolution, identification, and
quantitation of components of interest.
Moisture removal from the sample stream prior to sample concentration
minimizes these problems and allows larger sample volumes to be concentrated, thus
providing greater detection sensitivity. However, any device used to remove moisture
from the sample can result in the loss of certain VOCs of interest and potentially
introduce contaminants into the system.
Moisture can be removed from the air sample stream using a Perma-Pure®
permeable membrane or equivalent drying device. The permeable membrane drying
device generally consists of a copolymer of tetrafluoroethylene and fluorosil monomer
2-7
-------
Revision 0
OCTOBER 1991
that is coaxially mounted within a larger polymer or stainless steel tube. The sample
stream is passed through the permeable membrane tube, allowing water to permeate
through the walls into a dry nitrogen (N2) or air purge stream flowing through the
annular space between the membrane and the outer tube. To improve drying
efficiency and prevent memory effects, the dryer can periodically be cleaned using a
procedure that involves heating (typically at 100 degrees centigrade for 20 minutes)
and purging with dry N2 or air.
For the manual analysis method, the incorporation of a moisture removal device
is left to the discretion of the user. This decision requires a trade-off between the
advantage (i.e., improved chromatographic sensitivity and reduced retention times
shifting) and the disadvantages (i.e., potential contamination problems, loss of certain
compounds of interest, and increased retention times shifting). For automated
applications, moisture removal is critical and will be required.
2.2.4 Detector Types
Several non-specific but selective GC detectors are available for determining
hydrocarbon, aliphatic, aromatic, and halogenated compounds. The FID is the most
widely used and universal GC detector. The FID provides good sensitivity and uniform
response based on the number of carbon atoms in the compound. The FID is well
suited for ambient air analysis because a majority of VOCs in ambient air are
hydrocarbons. The uniformity of response allows reasonable estimates of
hydrocarbon compound concentration to be determined. This estimate of
concentration is achieved by calibrating the FID response with a single representative
compound (e.g., propane). The FID also has a broad linear dynamic range of
response, allowing analysis of samples with concentrations ranging from nanogram
(ng) to milligram (mg) quantities of hydrocarbons.
An electron capture detector (ECD) is appropriate for the analysis of electron
deficient materials, particularly the poly-halogenated and nitro-substituted compounds.
Many of the compounds in this select group are considered toxic VOCs. The ECD is
10 to 100 times more sensitive than the FID for specific halogenated compounds.
2-8
-------
Revision 0
OCTOBER 1991
Using an ECD in combination with an FID can be useful in both the qualitative
identification and quantitative determination of specific halogenated compounds.
The photoionization detector (PID) is useful in ambient air analysis because of
its sensitivity to, and selectivity for, aromatic compounds. Although the PID has
inherent stability and drift problems, its ability to detect aromatic compounds can be
10 to 20 times greater than that of the FID. Using a PID in conjunction with an FID
can aid in qualifying and quantifying most aromatic compounds.
2.2.5 Calibration Standards
Calibrating a GC/FID system to measure VOCs requires two distinctly different
types of calibration mixtures: a primary standard to calibrate detector response for gas
chromatographic peak quantitation and a qualitative mixture of known hydrocarbon
compounds to determine gas chromatographic peak retention times.
2.2.5.1 Primary Calibration Standard --
The GC/FID response is calibrated in ppbC using a propane primary calibration
standard traceable to the National Institute of Standards and Technology (MIST).
Standard Reference Materials (SRM) from NIST and Certified Reference Materials
(CRM) from specialty gas suppliers are available for this purpose. Less expensive
working standards needed for calibration over the range of expected concentrations
can be manually prepared by the user or purchased from a gas supplier, provided
they are periodically traced to a primary SRM or CRM. Based on the uniform carbon
response of the FID to VOCs, the response factor determined from the propane SRM
is used to convert area counts into concentration units for every peak in the
chromatogram.
2.2.5.2 Retention Time Calibration Standard -
The retention time calibration standard is a multiple-component mixture
containing all target VOCs at concentration levels from 10 to 30 ppbC for each
compound. It is a working standard and is used during the initial setup of the GC/FID
system to optimize critical peak separation parameters and determine individual
2-9
-------
Revision 0
OCTOBER 1991
retention times for each of the target compounds. The retention time calibration
standard is also used during the routine operation of the GC/FID system as a quality
control (QC) standard for verifying these retention times. The response of GC/FID to
selected hydrocarbons in this standard can be used to monitor FID performance and
determine when recalibration of the FID using the primary calibration standard is
necessary. The concentration of each compound in the retention time standard need
not be directly traced to an SRM or CRM (as is the case for the primary calibration
standard); rather it can be determined with reasonable accuracy using the FID
propane carbon response factor from the calibrated GC system. The
multiple-component mixture is not readily available and must either be prepared by the
user or obtained as a special order from a specialty gas vendor. A general procedure
for preparing the multiple-component mixture is given in the following discussion.
2.2.5.2.1 Retention Time Calibration Standard Preparation - A stock
retention time calibration standard containing the compounds of interest should be
prepared at a concentration level approximately 100 times that of the anticipated
working standard concentration. The stock standard can be prepared by blending
gravimetrically weighed aliquots of neat liquids or by adding aliquots of gaseous
standards with a inert diluent gas. The aliquot of each compound is introduced
through a heated injector assembly into an evacuated SUM MA® passivated stainless
steel canister. For the neat liquid aliquots, the pre-injection and post-injection syringe
weights are recorded, and the difference used to determine the amount of liquid
actually transferred to the canister. Following injection of all neat liquid and gaseous
components, the canister is pressurized to at least 2 atmospheres above ambient
pressure with clean, dry N2. Concentrations are calculated based on the amount of
compounds and diluent injected and the final canister pressure, using ideal gas law
relationships.
The stock retention time calibration standard is used to prepare humidified
working retention time calibration standards at the ppbC level. As indicated earlier, it
is not necessary to determine exact component concentrations in the multi-component
2-10
-------
Revision 0
OCTOBER 1991
mixture because it is not being used to determine compound specific response
factors. However, the approximate concentration of the stock standard must be
known in order to prepare the working retention time standards. Preparation of the
working standards is accomplished by syringe injecting a gaseous aliquot of the stock
standard into a SUM MA® passivated stainless steel canister, and subsequently diluting
with humidified zero air. The working standard should be prepared to provide enough
volume to last throughout a monitoring period (3-4 months). A 33-liter SUMMA®
passivated canister pressurized to at least 3 atmospheres above ambient pressure
should provide sufficient supply of the working standard.
2.2.6 Analytical System Calibration
The detector response of the analytical system should initially be calibrated with
multiple level propane standards over the expected sample concentration range. The
primary calibration standard is used to generate a per Carbon response factor based
on propane for determining the concentration of each target VOC, as well as the
summation of all detected VOCs (total NMOC). It is impractical and unnecessary to
determine compound specific response factors for each of the volatile organic O3
precursors shown in Figure 2-1, because the per Carbon response of the FID to these
compounds is approximately equal. It is appropriate to measure each compound
concentration in terms of ppbC using the relative response factor determined from the
propane standard.
For a known, fixed sample volume, the concentration is proportional to the area
under the chromatographic peak. The area is converted to ppbC using the following
equation:
CA = RF (AC)
Where:
RF = Response Factor
AC = Area Counts
CA = Concentration (ppbC)
2-11
-------
Revision 0
OCTOBER 1991
The response factor (RF) is an experimentally determined calibration constant
(ppbC/area count), and is used for all compound concentration determinations. The
response factor is determined by the analysis of the ppbC level NIST standard and is
determined using the following equation:
RF =
MAC
Where:
3 = Carbon Atoms - Propane
CB • = Concentration of the NIST Propane Standard (ppbv)
MAC = Median area count, determined from three replicate
analyses of the primary standard or analysis of the
multi-level calibration standards
Retention time identification of target compounds is performed by analyzing the
retention time calibration standard prepared as described in Section 2.2.5.2. This
standard is analyzed at least in duplicate to establish the correct retention times and
retention time windows for the peaks of interest.
A calibration check and retention times check of the analytical system should be
performed every 2 days in order to determine system variability and overall
performance. Acceptance criteria for analytical system performance (e.g., ±10%)
should be established during the method development stage of the program.
The calibration and retention times check are performed concurrently using the
retention time calibration standard. The retention time calibration standard is analyzed
and the resulting compound concentrations and retention times are compared to the
values obtained during the initial or most recent calibration of the analytical system.
The compound concentrations and retention times should compare within the limits of
the acceptance criteria. If they do not re-calibration of the analytical system should be
performed.
2-12
-------
Revision 0
OCTOBER 1991
2.2.7 Column Selection and Configuration
Column selection for analysis of the target volatile organic O3 precursors is
dependent to a large extent on whether the application involves manual or automated
methods. However, column selection for both methods is primarily dictated by total
sample analysis time and target compound resolution requirements. Other column
selection factors to be considered include practical and cost considerations, such as
the need to minimize cryogen consumption. Selecting columns that will provide the
desired separation of the C2 through C4 hydrocarbons without cooling the column
oven to sub-ambient temperatures will decrease cryogen consumption significantly.
Column configurations for use in enhanced monitoring programs are generally
limited to single-column, single-detector, or dual-column, dual-detector applications.
The simplest analytical column configuration involves the use of a single column with a
single FID. However, this configuration imposes limitations on the separation of
selected target volatile organic O3 precursors. Analyzing the full range of C2 through
C12 hydrocarbons using a single analytical column may result in less than optimum
separation characteristics for either the light or heavy hydrocarbons, depending on the
analytical column chosen. For example, to improve resolution of the C2 through C4
hydrocarbons, a thick liquid phase column and sub-ambient column oven
temperatures is desirable. However, the use of a thick liquid-phase column results i'n
less than optimum resolution of the C5 through C12 hydrocarbons, and sub-ambient
column oven temperatures result in increased cryogen consumption. In order to
improve the separation characteristics for the light hydrocarbons (C2 through C4) as
well as the heavier hydrocarbons (C5 through C12), a more complex dual-column,
dual-detector configuration may be considered. In this case, the two columns can be
judiciously selected to provide optimum separation of both light and heavy
hydrocarbons without sub-ambient column oven temperatures. Because both
columns must be contained in one gas chromatographic oven for automated
applications, columns must be selected that will provide the desired separation with a
single oven temperature program.
2-13
-------
Revision 0
OCTOBER 1991
Despite the limitations, users are strongly encouraged to pursue a
single-column, single-detector approach during the initial operation of their enhanced
monitoring program. A single-column configuration may result in less than optimum
separation characteristics of either the light or heavy hydrocarbons, but is simpler to
use than a dual-column configuration. As experience is gained with the GC/FID
system(s), a more complex dual-column configuration should be pursued.
2.2.7.1 Column Recommendations -
Several columns applicable to either single- or dual-column applications are
discussed below. The column conditions described are recommendations provided
from laboratory applications or conditions determined by the column manufacturer that
should provide acceptable separation of the VOCs of interest. However, these
conditions must be evaluated and optimized by an experienced chromatographer to
verify acceptable peak resolution prior to use. Both column configuration and column
selection are left to the discretion of the users. The columns described below have
been used in either a single- or dual-column configuration in conjunction with a single-
or dual-FID for separation of the C2 through C12 hydrocarbons.
The dual-column configuration recommended does not require the use of
sub-ambient column oven temperatures and therefore results in reduced cryogen
consumption. When using dual-column configurations, a single gas chromatographic
column oven temperature program must be optimized for both analytical columns in
order to provide the best target compound separation.
The heavy hydrocarbons (C5 - C12) maybe resolved using a 0.32 millimeter
(mm) inside diameter (I.D.), 60 meter (m) J&W® DB-1 fused silica column with a
1-micron dimethylsiloxane coating. However, this column will not provide complete
separation of the light hydrocarbons (C2 - C4) even at sub-ambient column oven
temperatures. The DB-1 column has been historically and extensively used in ambient
air applications. It can be used in conjunction with a 0.32 mm I.D., 50 m Chrompack®
Porous Layer Open Tubular (PLOT) fused silica analytical column, with a 5-micron
AI203/KCI coating. The PLOT column provides acceptable light hydrocarbon
2-14
-------
Revision 0
OCTOBER 1991
separation under the same column oven temperature program conditions used for the
DB-1 column but does not provide complete separation of C9 - C12 hydrocarbons.
Simultaneous use of these columns lends itself to dual-column automated applications.
The PLOT analytical column is susceptible to moisture, which may cause peak
retention times shifting and column deterioration; therefore, moisture must be
removed from the sample using a permeable membrane dryer or other drying device.
If manual sample analysis using these columns is performed, sequential analyses or
the use of separate GO systems may be considered to optimize and obtain complete
C2 through C12 separation. Figures 2-2 and 2-3 are example chromatograms from the
DB-1 and PLOT columns. It is recommended that users give primary consideration to
the DB-1 and PLOT columns described above during their column selection process.
The user is encouraged to perform the initial system set-up, optimization, and
check-out with a less complex single-column configuration utilizing the DB-1 for the C4
through C12 hydrocarbons, since these hydrocarbons represent the majority of the
target volatile organic O3 precursors. Once this single-column application has been
optimized, the second analytical column may be installed and optimization of
simultaneous operation of both columns for complete C2 through C12 resolution can be
performed.
There are a large number of alternate column options that can be used for C2
through C12 analysis for both single- or dual-column approaches. The column
selection process should be based on the capability of the column to separate the
target volatile organic O3 precursors listed in Figure 2-1 in conjunction with desired
overall sample analysis time. The recommended manufacturer conditions, along with
the carrier gas flow rates, must be evaluated and optimized in order to verify
acceptable peak resolution prior to use. When choosing alternate columns, the user
should consult directly with the analytical column manufacturer or developer for advice
regarding column characteristics, optimum gas chromatographic oven temperature
programs, or other considerations.
2-15
-------
ro
—L
O)
Analytical Conditions
Column: DB1, 60m x 0.32mm
Temperature: -60°C/2 min at 6°C/min to
180°C/3min
Det: FID, 3QO°C
Gamer Gas: Helium, 2 mL/min
0)
c
CD
£ 5?
J
Figure 2-2. Example Chromatogram for J&W DB1 Capillary Analytical Column.
8
a
03 3°
a I
8 5
-------
Analytical Conditions
Column: A^Q/KCI, 25m x 0.32mm
Temperature: 50°C to 75°C at
5°C/min to
125°Cat10°C/minto200°Cat
15°C/min, hold 30 minutes
Det: FID, 300°C
Carrier Gas: Helium, 5 mL/min
S « 4
CD <5 c a
2
a
c
s
Q.
c!j
^-
£ S: i? fl s a
i * ft ?. ! ii
5 °- < ~l L
I,
li 1
i H ni
JL i ii
i r rr'r i i i i i i i
"" CNJ C*J ^ lO (O f*1* CO
LJIU
i i
1
1
III
UUI
1 1 1
i
o
in
I
I
u
1
.
||
II
|
1
If
III
0
c
ID
i
m
J
1
1
1
1
ul
I I I I I
O
£
3
£
r^
y g
m 1
V —*•
f 5"
i £ £ 8
0) ^ S
1 -
1
.
T I I I I I I I I I I I I I I I I I
ttOT-cjCT^ioujr-coogSJSjSSSijgiD
Figure 2-3. Example Chromatogram for Chrompak®^^^! PLOT
Capillary Analytical Column.
o
o
CD 3°
m 9
a
-------
Revision 0
OCTOBER 1991
The following columns are alternatives for single-column, light (C2 - C4)
hydrocarbon separation and, in some cases, require sub-ambient oven temperature
conditions:
1. J&W® DB-1 with a 5-micron dimethylsiloxane phase thickness, an internal
diameter of 0.32 mm, and a length of 60 m. The recommended oven
temperature program is -60 degees centigrade (°C) for 2 minutes, to
180°C at 8°C per minute. The final oven temperature is maintained for
13 minutes for a total analytical run time of 45 minutes.
2. J&W® GS-Q fused silica capillary column with an internal diameter of
0.53 mm and a length of 30 m. The recommended oven temperature
program is 40°C to 200°C at 4°C per minute. The final oven temperature
is maintained for 5 minutes for a total analytical run time of 45 minutes.
The following columns are alternatives for single-column, heavy (C5 - C12)
hydrocarbon separation and, in some cases, require sub-ambient oven temperature
conditions:
1. Chrompack® WCOT (Wall Coated Open Tubular) capillary fused silica
column with a 5-micron CP-SIL 5CB dimethylsiloxane stationary phase
thickness, an internal diameter of 0.32 mm, and a length of 50 m. The
recommended oven temperature program is -20°C for 5 minutes, to
200°C at 7°C per minute. The final temperature is maintained for
9 minutes, which results in a total analytical run time of 40 minutes.
2. Restek® RTx-502.2 capillary fused silica column with a 3-micron phase
thickness, an internal diameter of 0.53 mm, and a length of 105 m. The
recommended GC oven temperature program is 35°C for 10 minutes, to
200°C at 4°C per minute. The final oven temperature is maintained for 7
minutes, which results in a total analytical run time of 58 minutes. This
column is capable of separating the C4 through C12 hydrocarbons
without the need for sub-ambient column oven temperatures.
A combination of these light and heavy hydrocarbon separation columns may
be used to accommodate dual-column approaches.
2-18
-------
Revision 0
OCTOBER 1991
2.2.8 Data Quality Objectives
Data quality objectives (DQOs) must be established to ensure that the quality of
the data collected is consistent with the overall method goals and intended use of the
data. Data quality objectives reflect the level of uncertainty that the data user can
tolerate from data collected. The method performance characteristics of the DQOs
should include, at a minimum, detection limits, accuracy, and precision guidelines. It
is the responsibility of the user to implement the steps necessary to determine and
ensure that the data meet the established DQOs.
Detection limits are expressed in units of concentration and reflect the smallest
volume of a compound that can be measured with a defined degree of certainty.
Instrument detection limits can be assessed using a procedure specified in 40 CFR
Appendix B, Part 136.15. This procedure is device- or instrument-dependent and
involves the replicate analysis of at least seven samples at very low ppbv
concentrations. The procedure uses the standard deviation of the concentration
measured from replicate sample analyses in conjunction with the Student's t-value at
the 99% confidence level and a standard deviation estimate with n-1 degrees of
freedom. The detection limit for hydrocarbon VOCs should be approximately 1 ppbC
for each compound.
Accuracy involves the closeness of a measurement to a reference value, and
reflects elements of both bias and precision. The SRM standards from NIST or CRM
standards from specialty gas vendors should be used to determine the reference value
and must be analyzed under conditions that duplicate the analysis method. In the
absence of specific DQOs, absolute accuracy should be within 25% of a reference
value. Accuracy for at least 25% of the target compounds should be determined.
Precision is a measure of the repeatability of the results. In the absence of
specific DQOs, absolute precision of repeated measurements should be within 20%
relative standard deviation or coefficient of variation.
2-19
-------
Revision 0
OCTOBER 1991
2.3 MANUAL METHOD FOR COLLECTING AND ANALYZING VOLATILE
ORGANIC OZONE PRECURSOR SAMPLES
The manual methodology for obtaining volatile organic O3 precursor
measurements involves collecting integrated, whole air canister samples for
subsequent analysis at a central laboratory. The proposed revisions to 40 CFR
Part 58 require States to obtain 3-hour and 24-hour integrated measurements of
volatile organic O3 precursors at specified sample collection frequencies based on
individual PAMS site type requirements. The sample collection frequencies range from
one 24-hour sample every sixth day to eight 3-hour samples every day. Specific
sample collection frequencies are presented in Figure 1 -1.
Application of the manual methodology to the enhanced O3 monitoring
regulations requires the collection and analysis of a large number of canister samples
depending on the combination of manual and automated (see Section 2.4)
approaches selected to satisfy the requirements. The extent and success of a manual
monitoring program will be dependent on the number of canisters available for use,
and the analytical capacity of the central laboratory involved. An integrated, well
planned sampling and analysis program is necessary to address the numerous
aspects of a canister-based monitoring operation, which include canister cleaning and
transport; sampling frequency and procedures; analysis procedures; and data
acquisition and reporting. These details must be addressed in accordance with the
specific needs of the user. The intent of this section is to provide general guidance on
the manual method approach. Users of manual methodology are responsible for the
selection, set-up, and optimization of their specific system(s), and for the preparation
of SOPs that delineate the details of each operation.
2.3.1 Sampling
This section describes the configuration and use of SUMMA® passivated
canisters and associated sample collection systems. These systems provide samples
for subsequent analysis at a central laboratory using a gas chromatographic analytical
system with computerized data reduction and reporting capabilities.
2-20
-------
Revision 0
OCTOBER 1991
The sample collection system should be capable of unattended sampling in
order to allow collection of samples in accordance with the specified schedule
presented in Table 1-1. Procedures for collecting canister samples are described
below. Precautions pertaining to the use of canisters, canister cleaning procedures,
and sampling system certification procedures are discussed in detail in Appendix A of
this document.
Collecting time-integrated whole ambient air samples for subsequent analysis of
target compounds is a widely accepted practice. Systems currently in use incorporate
diverse operating approaches. The primary difference among the various approaches
is the technique and associated hardware used to perform time-integration of a
canister sample collection. Time-integration techniques generally involve the use of
electronic and mechanical devices, either separately or in combination. Canister
sampling systems can be obtained commercially or can be custom-built for a specific
application.
The use of canisters to satisfy the 3-hour sampling schedules specified in
Table 1-1 will require automated multiple-event canister sampling systems. Back-to-
back collection of the individual 3-hour samples may not be practical using single-
event systems, due to the required attendance of an operator to change the sample
canisters between events. A limited number of multiple-event canister sampling
systems are commercially available. Additional systems should become available as
the need for them increases. Multiple-event systems can also be custom-built. A
conceptual automated multiple-event canister sampling system configuration is
presented in Figure 2-4.
2.3.1.1 Canister Sample Collection --
The following sections generally describe single-event canister sampling
equipment and procedures that can be incorporated into multiple-event collection
system operation.
2-21
-------
ro
ro
ro
©
® Sample Manifold Inlet
® 2 micron Filter
© Flow Control Device
© Metal Bellows Pump
© Digital Timer
© Latching Solenoid Valve
(D 6-L Sample Canister
@ By-Pass Pump
Continuous
Sample
System
3-hr Sample
System
24-hr Sample
System
3-hr Sample
System
3-hr Sample
System
3-hr Sample
System
3-hr Sample
System
5616536R
Figure 2-4. Automated multi-event canister sampling system.
o
CD 33
m 9
331
£§'
CD -3
-*• O
-------
Revision 0
OCTOBER 1991
2.3.1.1.1 Sampling Equipment - A typical single-event sampling system
configuration is presented in Figure 2-5. The single-event canister sampling system is
comprised of the following primary components:
Sample Pump - A stainless steel bellows pump, capable of 2 atmospheres
output pressure. The pump must be free of leaks and determined nonbiasing.
Sample Inlet Line - Chromatographic-grade stainless steel tubing used to
connect the sampler to the sample probe and manifold assembly.
Sample Canister A SUM MA® passivated leak-free stainless steel sample
containment vessel of desired internal volume with a bellows valve attached at the
inlet.
Stainless Steel Vacuum/Pressure Gauge - Capable of measuring vacuum
(0-30 in Hg) and pressure (0-30 pounds per square-inch gauge). Gauge should be
leak-free and shown to be nonbiasing.
Fixed Orifice. Capillary. Adjustable Micrometering Valve or Electronic Mass Flow
Controller - Capable of maintaining a constant flow rate (± 10%) over a specific
sampling period under conditions of changing temperature (20-40°C) and humidity
(0-100% relative).
Particulate Filter - 2 micron sintered stainless steel in-line filter.
Electronic Timer - To allow unattended operation (activation and deactivation) of
the collection system.
Latching Solenoid Valve - Electric-pulse-operated, stainless steel solenoid valve,
with Viton® plunger seat and o-rings.
Chromatographic Grade Stainless Steel Tubing and 316 Grade Stainless Steel
Fittings - Used for system interconnections (all tubing in contact with the sample prior
to analysis should be Chromatographic grade stainless steel and all fittings should be
316 grade stainless steel).
2-23
-------
Revision 0
OCTOBER 1991
By-Pass Pump
Sample Inlet
Timer
Latching
Solenoid Valve
V^^y Pressure
Gauge
Metal Bellows
PumpMB151
Sample inlet is an inverted glass funnel.
Lines and fittings are stainless steel.
Figure 2-5. Single-event canister sampling system.
2-24
-------
Revision 0
OCTOBER 1991
By-pass Pump - Used to continuously draw sample air through the inlet probe
and manifold assembly at a rate in excess of the sampling system total uptake.
Sample is extracted from the manifold at a lower rate, and excess air is exhausted to
the atmosphere.
Elapsed Time Indicator - Measures the duration of the sampling episode.
2.3.1.1.2 Sampling Procedure -
The sample is collected in a canister using a pump and flow control device. A
stainless steel metal bellows style pump draws in ambient air from the sampling probe
and manifold assembly at a constant flow rate to fill and pressurize the sample
canister.
A flow control device is used to maintain a constant sample flow rate into the
canister over a specific sampling period. The flow rate used is a function of the final
desired sample pressure and the specified sampling period and assumes that the
canisters start at a pressure of 5 mm Mercury (Hg) absolute. The flow rate can be
calculated by:
PxV
Tx60
Where:
F = flow rate (mL/min)
P = final canister pressure, atmospheres absolute
V = volume of the canister (ml_)
T = sample period (hours)
60 = minutes in an hour
For example, if a 6-L canister is to be filled to 2 atmospheres absolute pressure in
3 hours, the flow rate can be calculated by:
F = 2 x 6000 = 67.7 mL/min
3x60
2-25
-------
Revision 0
OCTOBER 1991
For automatic operation, the timer is programmed to activate and deactivate the
sample collection system at specified times, consistent with the beginning and end of
a sample collection period.
The use of the latching solenoid valve avoids any substantial temperature rise
that would occur with a conventional, normally closed solenoid valve that would have
to be energized during the entire sampling period. The temperature rise in the valve
could cause outgassing of organic compounds from valve components. The latching
solenoid valve requires only a brief electrical pulse to open or close at the specified
start and stop times; therefore, the valve experiences no temperature increase. The
pulses may be obtained either with an electronic timer that can be programmed for
short (5 to 60 seconds) actuation pulses, or with a conventional mechanical timer
incorporating an attached electric pulse circuit.
Canister sampling systems can collect sample from a shared sample probe and
manifold assembly as described in Section 2.4.1.1.1 or from a system specific
stainless steel sample probe and by-pass pump. If a system specific probe and by-
pass pump are used, a second electronic timer should be incorporated to start the by-
pass pump several hours prior to the sampling period to flush and condition the
components. The connecting lines between the sample inlet line and the canister
should be as short as possible to minimize internal surface area and system residence
time.
The flow rate into the canister should remain relatively constant over the entire
sampling period. If a critical orifice is used as the flow control device, a drop in the
flow rate may occur near the end of the sample period.
Prior to field use, each sampling system should be certified nonbiasing. (Refer
to Appendix A, Section A4.0 of this document for details pertaining to canister
sampling system certification.) The canisters should also be determined nonbiasing or
clean before each use. (Refer to Appendix A, Section A3.0 of this document for
details pertaining to canister cleaning.)
2-26
-------
Revision 0
OCTOBER 1991
The following provides specific details for operating a typical single-event
sampling system.
1. Verify the correct sample flow rate using a calibrated mass flow meter or
rotameter. The flow can be measured directly at the inlet of the system. The
calibrated mass flow meter or rotameter is attached to the sample inlet line, prior to
the particulate filter. The sampling system is activated, and the reading obtained is
compared to the desired collection flow rate. The values should agree within
±10 percent. If a mass flow controller is being used as the system flow control
device, allow the system to equilibrate for 2 minutes. After 2-minute equilibration, the
desired sample flow rate is attained by adjusting the system mass flow controller until
the calibrated mass flow meter or rotameter indicates the correct flow rate. If the
sampling system uses a micrometering valve instead of a mass flow controller as the
flow control device, adjust the micrometering valve until the correct flow rate is
achieved (see Figure 2-5). If the sampling system uses a fixed critical orifice assembly
as the flow control device, change the orifice to a size consistent with the desired flow
rate.
2. Deactivate the sampler and reset the elapsed time indicator to show no
elapsed time.
3. Disconnect the calibrated mass flow meter or rotameter and attach a
clean canister to the sampling system.
4. Open the canister bellows valve.
5. Record the initial vacuum in the canister, as indicated by the sampling
system vacuum gauge, onto the canister sampling field data sheet.
6. Record the time of day and elapsed time indicator reading onto the
canister sampling field data sheet.
7. Set the electronic timer to begin and stop sampling at the appropriate
times.
8. After sample collection, record the final sample pressure onto the
sampling field data sheet. Final sample pressure should be close to the desired
2-27
-------
Revision 0
OCTOBER 1991
calculated final pressure. Time of day and elapsed time indicator readings should also
be recorded.
9. Close the canister bellows valve. Disconnect and remove the canister
from the sampling system.
10. Attach an identification tag to the canister. Canister serial number,
sample number, location, and date should be recorded on the tag.
2.3.2 Analysis
Gas chromatographic analysis of whole air ambient samples collected in
SUM MA® passivated canisters for measurement of VOCs is well established and
routinely performed. The analytical system should incorporate a sample concentration
trap, a gas chromatograph, appropriate analytical column(s) and detector(s), and a
data acquisition system. A sample drying device may also be incorporated. The user
may choose either a single analytical column with a single-FID for ease of operation or
a more complex two-column, two-FID configuration to improve the resolution of the
VOCs over the C2 through C12 range. An alternative approach for improving C2
through C12 resolution is to use two gas chromatographs with two separate specific
analytical columns and oven temperature programs. Several column configuration
alternatives are described in Section 2.2.7.
2.3.2.1 Analytical Equipment and Configuration ~
The following sections identify and describe the primary components of the
analytical system.
2.3.2.1.1 Sample Drying Device - Sample moisture removal can be
performed to prevent or reduce potential adverse effects on the sample concentration
device, the analytical column(s), and the detector(s). Moisture removal allows analysis
of larger sample volumes, which provide enhanced detection sensitivity. Enhanced
detection sensitivity is critical to the measurement of less abundant volatile organic O3
precursors.
2-28
-------
Revision 0
OCTOBER 1991
Moisture should be removed from the sample stream using a permeable
membrane drying device or equivalent drying system (Section 2.2.3). Nonpolar
hydrocarbon compounds pass through the drier unaffected but certain polar VOCs
may be removed (e.g., C10 terpenes). To prevent moisture build-up and memory
effects, the drier must be initially and periodically cleaned using a procedure
recommended by the manufacturer.
2.3.2.1.2 Sample Concentration Device - Ambient air samples can be
concentrated using commercially available or custom built adsorbent and/or
cryo-focusing devices, referred to as traps. Adsorbent traps use individual or
combinations of absorbent media (e.g., Carbotrap, Carbotrap C, and Carbosieve III) to
achieve sample concentration. The adsorbent(s) is typically contained in a tubular
housing. The sample stream is passed through the adsorbent(s) and selectively
concentrated. Cryogenically cooling the adsorbent trap during the concentration
process can increase the collection efficiency of the adsorbent media. Cryogenic
concentration of samples can also be performed using a trap consisting of
chromatographic grade stainless steel tubing packed with commercially available
60/80 mesh deactivated glass beads maintained at -185°C during sample
concentration. Most sample concentration devices use the process of thermal
desorption to revolatize the concentrated VOCs for transfer to the analytical column for
separation.
2.3.2.1.3 Gas Chromatoqraph - Commercially available GC/FID systems
can be used to analyze concentrated VOC samples. Sub-ambient oven capability may
be required depending on the analysis approach selected.
2.3.2.1.4 Analytical Columns — Because of the complexity of the
relationship between instrument configuration, operating parameters, sample matrix,
and target compounds, the selection of the applicable analytical column(s) should be
based on the specific criteria and recommendations for column selection discussed in
Section 2.2.7. Consultation with GC column manufacturers for guidance in selecting
2-29
-------
Revision 0
OCTOBER 1991
alternative columns to meet the specific requirements of a program is recommended.
Column selection should be based on the capability of a column to separate the target
compounds, while considering the desired overall analysis run time. Once the column
configuration is selected, it is the responsibility of the chromatographer to determine
the optimum analytical conditions for each critical operational parameter in order to
achieve the best gas chromatographic performance.
2.3.2.1.5 Support Gases —
GC Support Cylinder Gases:
Ultra High Purity (UHP) grade helium (99.999% purity)
Hydrocarbon Free Air (< 0.1 parts per million Carbon total hydrocarbon)
UHP grade Hydrogen (99.999% purity)
Cryogen:
Liquid Nitrogen
Primary Calibration Standard:
30 ppbC propane standard diluted from an NIST stock standard
Retention Time Calibration Standard:
10 to 30 ppbC/compound calibration standard containing each of the
target compounds.
2.3.2.1.6 Data Acquisition System - The data acquisition system consists
of, but is not limited to, a PC-DOS personal computer with appropriate
chromatography data acquisition and peak integration software. Chromatography
software is typically comprised of subroutines that perform data acquisition, peak
integration and identification, hard copy output, post-run calculations, calibration, peak
re-integration, and user program interfacing. Acquired data should be automatically
and permanently stored on magnetic media (e.g., hard disk, floppy diskette).
2-30
-------
Revision 0
OCTOBER 1991
2.3.2.2 Analytical System Operation ~
This section provides general guidance for initial system set-up, analytical
system parameter optimization, calibration, and data acquisition and reporting.
Specific and detailed SOPs must be prepared by the user, with appropriate input and
assistance from the system manufacturer or developer.
2.3.2.2.1 Set-up and Parameter Optimization - A minimum period of 3 to
6 months prior to field sampling and sample analysis should be anticipated for
acquisition of canisters, acquisition and/or fabrication and certification of canister
sampling systems, initial equipment set up, user familiarization and development of
SOPs. All systems should initially be set up in accordance with the manufacturers or
developer's operational performance specifications.
During initial setup of the GC analytical system, numerous critical parameters
must be evaluated to determine optimum operating conditions for obtaining
quantitative data. These parameters include, but are not limited to, selection of
sample concentration flow rate; sample concentration trap collection and desorption
time and temperatures; the gas chromatographic oven temperature program
parameters; calibration of the detector, and gas chromatographic integration methods
to be used for peak identification. These parameters should be optimized and
characterized using; (1) pure compounds in air and (2) the retention time calibration
standard (see Section 2.2.5.2) containing a mixture of the components of interest.
Once the critical parameters are established, the retention time calibration standard is
used to determine the exact retention times of the target compounds for input into the
chromatography data reduction and peak integration software.
2.3.2.2.2 Calibration -- The analytical system should initially be calibrated
with propane standards at concentrations over the expected sample concentration
range. Analysis of the primary calibration standards should be performed every
two days in order to determine analytical system variability and overall performance.
Refer to Section 2.2.5 for details.
2-31
-------
Revision 0
OCTOBER 1991
Retention time identification of target compounds is determined by analyzing
individual species and the retention time calibration standard described in
Section 2.2.5.2.
2.3.2.2.3 Data Acquisition and Reporting - The data from the GC system
are collected and processed with a data acquisition system selected by the user. The
data acquisition system contains the algorithms to acquire, integrate, and identify the
chromatographic peaks by retention times. The system should also be capable of
producing a report file for every sample analyzed and interfacing with other data
processing equipment. Each report file should include, at a minimum, each peaks
retention time, area counts, concentration in ppbC, and the peaks name if it is an
identified component. The system should calculate an estimate of the total NMOC,
determined by the summation of the concentrations of all detected peaks in the
chromatogram. The report files should be archived on magnetic media to allow for
future processing.
To aid the chromatographer in obtaining more exact qualitative and quantitative
results, additional information such as variation in retention time during an analysis
should be considered. To obtain this additional information, it is recommended that
the data be processed using a software package such as that designed and marketed
by MetaChrom®. This software allows the user to process large and/or complex
chromatographic data sets and improve data validation. Reassignment of peak
identification can be performed using parameters such as reference compounds,
retention times and peak search windows, and appropriate algorithms. The software
program compensates for variations in sample matrix and instrument performance.
The software also provides information on the presence of unknown and unidentified
peaks.
2-32
-------
Revision 0
OCTOBER 1991
2.4 AUTOMATED METHOD FOR COLLECTING AND ANALYZING VOLATILE
ORGANIC OZONE PRECURSOR SAMPLES
The PAMS requirements prescribed in 40 CFR Part 58 recommends that States
should obtain continuous, 3-hour measurements of volatile organic O3 precursors at
at least one site within a network. This section describes an automated sampling and
analysis methodology to be used in conjunction with the manual sampling and
analysis methodology described in Section 2.3. The automated methodology may
provide a viable, less costly approach for obtaining volatile organic O3 precursor
measurements at all sites within a network. The automated approach
(e.g., configuration, columns, detectors, etc.) described in this section is a modification
of the GC approach described in Section 2.3. Exceptions and limitations have been
applied to facilitate the automation of the example GC system. An automated system
offers an additional advantage in its inherent capability of providing short-term
(e.g., 1-3 hour) measurements on a continuing basis for long time intervals.
The following description is based on a commercially available example
automated system, the Chrompack® CP-9000, and is described only in general terms.
The intent is to provide guidance on configuration and operation of an automated
system. It is not intended to serve as an SOP. Alternative approaches using other
commercially available or custom fabricated systems are acceptable. Users must '
recognize that they are response for optimizing the critical parameters for their specific
system (consistent with the manufacturers instructions if applicable) and are also
responsible for the preparation of SOPs for their specific system.
2.4.1 Method Description
This section describes the equipment, data acquisition hardware, calibration,
operation, data reporting and validation for an automated GC system. This guidance
should be used to define equipment specifications and prepare SOPs consistent with
the proposed 40 CFR Part 58 enhanced O3 monitoring requirements.
The GC system must be designed for automatic sample collection, sample
analysis and data collection on site. For automated, real-time analysis of target volatile
2-33
-------
Revision 0
OCTOBER 1991
organic O3 precursors, the GC system must be housed in a temperature-controlled
shelter at the PAMS. The analytical system should incorporate a sample concentration
system (for concentrating VOCs continually at a constant flow rate over a
predetermined sampling time) a cryo-focusing trap (for improving the resolution
capability of the gas chromatographic column); an appropriate analytical column(s); an
FID detector(s); and a data acquisition and integration system.
The VOCs are collected on the sample concentration trap, thermally desorbed
onto the sample cryo-focusing trap, and finally thermally desorbed onto the analytical
column(s) for separation, qualitative identification, and quantification. The cycle of
collection, desorption, cryo-focusing, and analysis should be completely automated.
The automated system should provide the flexibility of adjusting critical sampling and
analysis parameters (e.g., sample integration times, trap temperatures and flow rates,
column temperatures, and support gas flow rates) in order to facilitate optimization.
System flexibility will allow the evaluation and consequent use of alternate components
within the system (e.g., alternate sample collection traps, cryo-focusing approaches,
cooling systems, and columns). For typical automated GC applications, the system is
set up to collect and analyze one sample each hour. To achieve this frequency of
sampling and analysis, the cycle of sampling, refocusing, analysis, and report file
generation takes approximately 1 hour and 20 minutes, with the last 20 minutes of the
cycle occurring concurrently with the sample collection phase of the next hourly cycle.
By overlapping the cycle activities in this way, the overall cycle time is effectively
reduced to 1-hour.
2.4.1.1 Analytical Equipment and Configuration -
The automated system should consist of a sample probe and manifold, a
sample drying device, a solid adsorbent primary sample collection trap, a secondary
cryo-focusing trap, an analytical column(s), and FID(s). Although more complex, the
two-column, two-FID configuration may be more appropriate if resolution over the full
range of C2 though C12 hydrocarbons is required. Several column configurations and
2-34
-------
Revision 0
OCTOBER 1991
approaches are described in Section 2.2.7. Information identifying and describing the
primary components of an automated system are presented in this section.
2.4.1.1.1 Sample Probe and Manifold - A sample probe and manifold
assembly should be used to provide a steady representative flow of air for collection
and subsequent analysis. Figure 2-6 presents a sample probe and manifold assembly
recommended for use at PAMS.
The sample probe is constructed of glass approximately 1 inch in outside
diameter (O.D.). The inlet of the sample probe is configured as an inverted funnel
approximately 4 inches in O.D. The sample manifold is constructed of glass
approximately 1-1/2 inches in O.D. Ports for sample extraction are located on the
sample manifold. Teflon® bushings are used as section connectors. A bleed adapter
and blower are located at the exit of the sample manifold. The blower is used to pull
sample air through the sample probe and manifold. The bleed adapter is used to
control the rate at which the sample air is pulled through the manifold.
An excess of sample air should be pulled through the sample probe and
manifold to prevent back diffusion of room air into the manifold and to ensure that the
sample air is representative. Sample air flow through the sample probe and manifold
should be at least 2 times greater than the cumulative flow of sample air being
removed from the manifold for collection and analysis.
The vertical placement of the sampling probe and inlet funnel should be at a
height of 3 to 15 meters above ground level. Because enhanced O3 monitoring
requirements involve multiple-pollutant measurements, this range serves as a practical
compromise when determining suitable probe position. In addition, the probe inlet
should be positioned more than 1 meter vertically and horizontally away from any
supporting structure.
The probe inlet should be positioned away from nearby obstructions such as a
forest canopy or building. The vertical distance between the probe inlet and an
obstacle should be at least two times the height difference between the obstacle and
the probe inlet. Unrestricted air flow across the probe inlet should occur within an arc
2-35
-------
Revision 0
OCTOBER 1991
SAMPLING
CANE
TEE
OPTIONAL
CANE
COLLECTION
BOTTLE
MANIFOLD
W/PORTS
BLOWER
MOUNT,
BLEED
ADAPTER
BLOWER
Figure 2-6. Sampling probe and manifold assembly.
2-36
-------
Revision 0
OCTOBER 1991
of at least 270 degrees. The predominant and second most predominant wind
direction must be included in this arc. If the probe inlet is positioned on the side of a
building, a 180° clearance is required.
More specific details of probe positioning are presented in the "Enhanced
Ozone Monitoring Network-Design and Siting Criteria Guideline Document".4
2.4.1.1.2 Sample Drying Device - Sample moisture must be removed for
automated applications to prevent or reduce adverse effects on the adsorbent trap,
the analytical column(s), and the detector(s) as described in Section 2.2.3. Moisture
removal allows analysis of larger sample volumes, which provide enhanced detection
sensitivity. Enhanced detection sensitivity is critical to the measurement of less
abundant volatile organic O3 precursors.
Moisture should be removed from the sample stream using a permeable
membrane drying device or equivalent drying system (see Section 2.2.3). Nonpolar
hydrocarbon compounds pass through the drier unaffected but some polar VOCs may
be removed (e.g., C10 terpenes). Although the potential loss of compounds is a
disadvantage, drying the sample is still warranted based on the advantages listed
above. To prevent moisture build-up and memory effects, the dryer must be initially
and periodically cleaned using a procedure recommended by the manufacturer.
2.4.1.1.3 Sample Concentration Device -- Ambient air samples are
concentrated using commercially available or custom built adsorbent and/or
cryo-focusing devices, referred to as traps. The Chrompack® CP-9000 uses a primary
adsorbent trap that is a cryogenically cooled. The primary trap is comprised of a
1/4-inch glass cartridge containing at least 1 gram of each of the following
adsorbents:
• Carbotrap C (for trapping high molecular weight VOCs);
• Carbotrap (for trapping C5 through C8 VOCs); and
• Carbosieve III (for trapping low-molecular-weight compounds).
2-37
-------
Revision 0
OCTOBER 1991
The three adsorbents are separated within the glass cartridge by layers of deactivated
glass wool. The trap is cryogenically cooled to -20°C during the concentration
process to ensure collection of the C2 hydrocarbons. The primary trap is used in
conjunction with a secondary cryogenic trap. The function of the secondary trap is to
refocus the desorbed sample from the adsorbent trap prior to injection onto the
analytical column. The secondary cryogenic trap is comprised of a 12 cm length of
deactivated fused silica with an I.D. of 0.53 mm. It is coated with the porous polymer
Poraplot U. The secondary cryogenic trap is maintained at -110°C during the sample
concentration process.
2.4.1.1.4 Gas Chromatoqraph -- A Chrompack® CP-9000 equipped with
automated control, sub-ambient oven capability, and injector assembly systems or an
equivalent gas chromatograph may be used.
2.4.1.1.5 Analytical Columns - The user should develop specific criteria
for column selection and anticipated gas chromatographic operating conditions, based
on parameters discussed in Section 2.2.7. For the initial operation of the automatic
GC system, the less complex single-column configuration is recommended. The
column selected should be capable of a reasonable separation of a majority of the
compounds listed in Figure 2-1, preferably without requiring sub-ambient gas
chromatograph oven temperatures. Elimination of the need for sub-ambient column
oven temperatures will minimize the cryogen required to operate the GC system.
Once the column is selected, it is the responsibility of the user to determine the
optimum conditions for each critical operating parameter for the best performance of
the column (i.e., best separation characteristics for the components of interest). Once
experience is gained, a more complex dual-column, dual-FID configuration can be
implemented and optimized.
2.4.1.1.6 Support Gases -
GC Support Cylinder Gases:
Ultra High Purity (UHP) grade helium (99.999% purity)
Hydrocarbon Free Air (< 0.1 parts per million Carbon total hydrocarbon)
2-38
-------
Revision 0
OCTOBER 1991
UHP grade Hydrogen (99.999% purity)
Cryogen:
Liquid Nitrogen
Primary Calibration Standard:
30 ppbC propane standard diluted from an NIST stock standard
Retention Time Calibration Standard:
10 to 30 ppbC/compound calibration standard containing each of the
target compounds as described in Section 2.2.5.2.
2.4.1.1.7 Data Acquisition System — The data acquisition system consists
of, but is not limited to, a PC-DOS personal computer with Chrompack PCI®
chromatography data acquisition and integration software. Chrompack®
chromatography software is comprised of subroutines that perform data acquisition,
peak integration and identification, hard copy output, post-run calculations, calibration,
peak re-integration, and user program interfacing. Acquired data should be
automatically and permanently stored on magnetic media (e.g., hard disk, floppy
diskette).
2.4.1.2 Operation-
This section provides general guidance for the initial setup of an automated GC
system, preparation of operational procedures, calibration, field operation, and data
acquisition and validation. Site specific and detailed SOPs must be prepared by the
user, with appropriate input and assistance from the manufacturer or developer of the
system.
2.4.1.2.1 Set-up and Parameter Optimization -- Users should anticipate a
minimum of 3 to 6 months for initial setup, configuration, familiarization, and
development of SOPs, prior to field deployment of the automated system. The system
should be initially set up and tested for conformance with the manufacturer's or
developer's operational specifications. Under the terms of the agreement for purchase
2-39
-------
Revision 0
OCTOBER 1991
of the system, the manufacturer or developer should be required to provide a detailed
instruction manual for operation of the system. The agreement should also require the
manufacturer or developer to provide initial system setup, user training, and
demonstration of adequate performance of the system in accordance with a
predetermined protocol (developed cooperatively by the purchaser and the
manufacturer). The initial setup, testing, and training phase of the operation should be
conducted in a laboratory setting with adequate space and accessible support
equipment.
During the initial set-up of the automated GC system numerous critical
parameters must be evaluated to determine optimum system operating conditions.
Critical parameters include, but are not limited to, the sample collection flow rate and
sampling (integration) time; sample concentration and desorption times; flow rates and
temperatures for the sample concentration traps; oven temperature program
parameters; detector calibration; and the GC integration methods used for peak
identification based on retention times. These parameters are optimized by varying
the operating conditions to achieve the best resolution of the target compounds using
pure component mixtures and the retention time calibration standard (see
Section 2.2.5.2). Once all the critical parameters are optimized, the retention time
calibration standard can be used to establish the exact retention times for each of the
target compounds as input to the internal chromatography software.
The following example is typical of optimized settings for the automated hourly
analysis of C2 to C8 hydrocarbons, using a single-column configuration in a
Chrompack® CP 9000 automated GC system:
Sample collection rate and time : 10 mL/min for 30 minutes
Primary trap sample collection temperature: -20°C
Total sample volume: 300 ml_
Primary trap desorption temperature: 250 C°
2-40
-------
Revision 0
OCTOBER 1991
Secondary trap collection rate and rate:
Secondary trap collection temperature:
Secondary trap desorption temperature:
Injector temperature:
Column:
Column flow rate:
GC oven initial temperature:
GC oven final temperature:
GC oven ramp rate:
GC oven final time:
Detector temperature:
Detector hydrogen flow-rate:
Detector air flow-rate:
7 mL/min for 7 minutes
200°C
200°C
Chrompack® AI2O3/KCI,
0.32 m x 25 m
10 mL/min
50°C
200°C
5°C/min to 75°C, 10°C/min to
125°C, 15°C/min to 200°C
30 minutes
300°C
30 mL/min
300 mL/min
2.4.1.2.2 Standard Operating Procedure -- An SOP for field operation of
the system should be prepared. The SOP should be based on information obtained
during the set-up and familiarization period and the requirements of the specific
program. The SOP should address the details of the routine operation of the system
and should include:
A detailed description of the system, including the required settings for all
critical operational parameters;
A listing of ancillary equipment and materials;
2-41
-------
Revision 0
OCTOBER 1991
• A detailed operating procedure addressing initial start-up, calibration
schedule, retention time check schedule, required maintenance, and
record keeping;
• Data acquisition and reporting aspects; and
• Associated data quality issues.
2.4.1.2.3 Calibration ~ The analytical system is calibrated in units of ppbC
using an NIST traceable propane primary standard. (See Section 2.2.5 for details on
calibration standards.) Based on the carbon response of the FID to the primary
standard, a relative response factor (ppbC/area count) is determined. This factor is
used to convert area counts from every peak in a chromatogram into concentration
units. An estimate of the total NMOC is made by summing the concentrations of
every peak (identified and unknown) detected in a sample. For an identified
compound, the concentration in ppbC can be divided by the number of carbon atoms
in the compound to estimate the concentration in ppbv. A minimum of three different
concentrations of propane are required to develop a suitable calibration curve for
determining the FID response. The three concentrations should span the working
range of the analytical system. At least two replicate analyses at each concentration
are recommended.
Retention time identification of target compounds is determined by analyzing the
retention time calibration standard described in Section 2.2.5.2. This standard is
analyzed at least in duplicate to determine the correct retention times and retention
time windows for the peaks of interest.
2.4.1.2.4 Sampling Parameters - Critical automated GC sampling
parameters are closely interrelated. Determination of specific optimum sampling
conditions is dependent on field conditions (i.e., expected compound concentration
ranges, humidity, temperature, etc.), desired sensitivity (detection limit), cryogen
consumption, and sample trapping efficiency. During the setup period, these sampling
parameters should be evaluated to determine the optimum conditions for each
sampling parameter. Recommended primary sampling parameters are:
2-42
-------
Revision 0
OCTOBER 1991
A sample collection rate of 1 each hour; and
A sample collection (integration) time of 15-45 minutes.
For hourly sampling, the sample collection time must be limited to allow for the
refocusing and analyzing cycle of the system. The refocusing and analysis cycle
requires approximately 45 minutes. Shorter sampling times (no less than 15 minutes)
may be used to reduce the use of cryogen. A sample collection volume of 200 to
600 ml_ is recommended. The sample collection volume used requires a judicious
trade-off between required detection limit (the larger the sample the lower the
detection limit) and potential moisture problems (the lower the sample volume the less
the potential for moisture-associated problems). Integration times of up to 45 minutes
may be considered by using an intermediate sample collection/integration device.
This device should consist of a sample collection/integration vessel configured to
provide integrated collection of one sample while a previously collected sample is
being analyzed. Advantages to using an intermediate collection/integration device
include longer integration times and reduced cryogen use during the concentration
step of sample analysis. Intermediate sample collection/integration devices are under
development and not currently commercially available.
2.4.1.2.5 Field Operation -- The automated GC system should be installed
in a temperature-controlled shelter at the field location. The system should be
operated in accordance with an SOP that is prepared by the user, based on the
information obtained during the setup and familiarization period. The system should
be serviced by a qualified operator. The operator should perform the routine
operational and QC functions specified in the SOPs. Critical operational checks
(e.g., calibration checks for the FID response) should be performed daily or as
frequently as practical. Operational parameters should be adjusted, if necessary, so
that the quality control criteria are maintained. Retention time checks should be
performed twice each week or preferably every other day to provide retention time
reference information for validating compound identifications. The retention time
2-43
-------
Revision 0
OCTOBER 1991
calibration standard can also be used to track the FID response to determine when
generation of a new calibration curve is necessary.
2.4.1.2.6 Confirmation Samples Analysis - There are various levels of
peak identification and quantitation confirmation that should be considered when
validating data obtained from the automated GC system. These levels vary in
complexity and extent of confirmation information provided. Confirmational
approaches that use canister sampling technology include the following:
• Duplicate analyses on a selected number of samples on the same
automated GC system;
• Secondary analysis on a better characterized GC system;
• Establishment of an exchange sample analysis and comparison program
with neighboring State agencies;
• External or internal audit sample analysis; and
• Confirmation analysis of selected target compounds using a GC/MS for
peak identification.
2.4.1.2.7 Data Acquisition and Reporting - Data from the automated GC
system are collected and processed using an on-site personal computer to operate
the GC acquisition and integration software program. The GC software is developed
and supplied by the manufacturer or developer of the system, and should contain the
necessary algorithms to acquire, integrate, and identify the chromatographic peaks by
retention times. The system should be capable of producing a report file for every
sample analyzed and interfacing with other data processing equipment. This file
should contain all the necessary information needed to identify the sample, the exact
time it was collected, the chromatogram generated by the GC for the sample, and a
sequentially numbered listing for all peaks from the chromatogram. This listing should
contain the "name of the peak" (if the peak is one of the target compounds and has
been identified by retention time) and all unidentified peaks and the associated
2-44
-------
Revision 0
OCTOBER 1991
concentration, retention time, relative retention time to the selected reference peak or
peaks, peak area, and peak width. The listing should also contain an estimate of total
NMOC calculated by summing the concentrations of all of the peaks detected from the
chromatogram. The report files for each day's operation should be transferred daily to
magnetic media for subsequent validation, evaluation, and interpretation.
2.4.1.2.8 Data Validation - The analyst must develop systematic
procedures to determine that the quality of the data is consistent with DQOs. These
procedures should include examination of the calibration, QC, and data report files.
Ideally, the report file for each sample analysis should be examined by an experienced
chromatographer to verify that the sample was properly collected, analyzed, and
correct peak identifications were made. However, this examination may not be
practical due to the large number of report files that are generated when the
automated GC system is operated continuously on an hourly basis. The validation
procedure in the QC section of individual site SOPs should address this issue by
recommending a minimum number of files that must be examined to ensure that the
system operated properly during data collection.
A computer software program such as the one designed and marketed by
MetaChrom® should be considered to facilitate the data validation process. This
software is designed to aid the chromatographer in generating more exact qualitative
and quantitative results. The MetaChrom® software is designed to reprocess the gas
chromatographic result files in order to verify the retention time identifications for each
peak. In this process, retention time and peak area information are extracted from the
gas chromatographic result file to create a large database matrix. Using this matrix,
consistent peak identifications can be made for all result files, including both identified
and unknown peaks.
2.5 GAS CHROMATOGRAPHY/MASS SPECTROMETRY
A GC system utilizing an FID cannot conclusively establish the identification of
VOCs. Peak identifications are based on matching peak retention times with those
obtained from previous analysis of neat compounds and mixtures of neat compounds.
2-45
-------
Revision 0
OCTOBER 1991
The selection of neat compounds is based on prior knowledge of the compound
composition expected in the sample mixture. Knowledge of the complexity of VOCs in
ambient air suggests that identification only by retention times is not conclusive.
The combination of capillary chromatographic retention times with the specific
ions observed by a mass spectrometer or MSD provides specificity for the
identification of organic compounds. Chromatographic coelution of two compounds
can be resolved by specific mass data unless the two compounds are isomers of each
other and have identical or very similar mass spectra. Because of the cost of the
instrumentation, use of mass-specific detection is not always possible. However, it is
recommended that samples be analyzed periodically by mass spectrometric detection
both to confirm peak identification and to provide identification information for the
observed unknown peaks. Use of a mass-specific detector to confirm compound
identification provides confidence in the identifications that are made in the course of
routine chromatographic operations.
2.5.1 Identification Confirmation
In the full-scan mode, the mass spectrometer scans over a specified mass
range in a period of time, then repeats the cycle. On each scan cycle, a complete
mass spectrum is generated. In the full-scan mode, identification of unknown or
unanticipated chromatographic peaks can be made because all of the mass spectral
information is acquired and available for examination. The following steps can be
taken to identify an unknown compound:
The mass spectrum of the unknown compound can be subjected to a
computerized library search for comparison to more than 42,000 mass
spectra, and a successful match of acceptable quality may be obtained
for the sample mass spectrum and a reference mass spectrum;
If a reference compound for the specific analyte is not included in the
library, the library search algorithm will frequently allow the experienced
mass spectra interpreter to establish the class of the unknown
compound and possibly the composition of the compound, although
usually not the exact isomer;
2-46
-------
Revision 0
OCTOBER 1991
The mass spectrum of the unknown compound can be interpreted by an
experienced mass spectral data interpreter, and the compound can be
identified from first principles, or additional information such as functional
groups, composition, and other characteristics can be obtained from the
mass spectrum; and
A quantitative value can be estimated for the unknown compound by
assuming an equivalent system response to a standard, a level of
quantitation that is usually accurate to within an order of magnitude.
The following criteria determine compound identification:
The capillary retention time must be within ±0.06 relative retention time
units of the retention time of the compound in either a calibration
standard or the daily calibration check standard;
The mass spectrum of the compound of interest must correspond to the
mass spectrum of a standard of the compound generated on the same
analytical system on which the sample analysis is being performed; and
The major ions of the mass spectrum must maximize within two scans
(one second) of each other.
If the identification criteria can be met, then the identification made by GO is confirmed.
2.5.1.1 Use of Selected Ion Monitoring Techniques --
If the mass spectrometer is programmed to look only at the regions of the
chromatogram where specific peaks occur more time is spent in monitoring the
masses of specific interest and consequently sensitivity is enhanced. This mode of
operation, in which only a few masses are monitored per compound, is known as
Selected Ion Monitoring (SIM). Because analysis of volatile organic O3 precursors in
ambient air involves very low concentrations (typically 1 to 15 ppbv), the additional
sensitivity that can be gained in the SIM operating mode is frequently required.
However, the SIM mode provides data only for the compounds specified -- unknown
compounds cannot be identified in this mode.
2-47
-------
Revision 0
OCTOBER 1991
2.5.2 Equipment
Gas Chromatoaraph/Mass Spectrometer - A gas chromatograph/mass
spectrometer and data system capable of acquiring data in either the full-scan or SIM
mode and of processing the data. The mass spectrometer must be capable of
scanning over the mass range from 35 to 350 amu in 1 second or less, using 70 eV
(nominal) ionizing energy in the electron ionization mode, and producing a mass
spectrum that meets instrument acceptance criteria.
Sample Interface - A sample interface capable of taking a constant volume
sample from the sample canister and cryogenically preconcentrating the sample.
Gas Chromatograph - A gas chromatograph capable of sub-ambient oven
operation.
Chromatoaraphic Columns - Chromatographic columns appropriate to provide
compound separation in approximately 45 minutes of analysis time.
2.5.3 Interferences
Impurities in the gases used and solvent vapors present in the laboratory where
the analysis is performed can contaminate the samples. The analytical system must
be demonstrated free from contamination under the conditions of the analysis by
analyzing humid zero air blanks. Use of tubing other than Chromatographic grade
stainless steel, non-Teflon® thread sealants, and flow controllers with rubber internal
components must be avoided. High levels of VOCs can result in carryover to
subsequent samples. Whenever a high VOC concentration is encountered in a
sample, humid zero air should be analyzed until the absence of contamination can be
demonstrated. The laboratory where analysis of VOC is performed should be
completely free of volatile organic solvents.
2.5.4 Standards
Standards may be generated by dynamic dilution of certified gaseous standards
or by serial dilution of a stock prepared from liquid standards.
2-48
-------
Revision 0
OCTOBER 1991
2.5.4.1 Instrument Performance Check Standard -
A standard of p-bromofluorobenzene (BFB) in humidified zero air at a
concentration that will allow collection and analysis of 20 ng of BFB (full-scan
operating mode) or 1 ng of BFB (SIM operating mode) under the standard
preconcentration parameters is required. The following criteria will apply:
Full-scan operating mode:
Mass
50 .
75
95
96
173
174
175
176
177
Abundance Ratio
8 - 40% of mass 95
30 - 66% of mass 95
base peak, 100% relative abundance
5 - 9% of mass 95
less than 2% of mass 174
50 - 120% of mass 95
4 - 9% of mass 174
93 - 101% of mass 174
5 - 9% of mass 176
SIM operating mode:
Mass
174
175
176
177
Abundance Ratio
100
4 - 9% of mass 174
93- 101% of mass 174
5 9% of mass 176
2-49
-------
Revision 0
OCTOBER 1991
2.5.4.2 Calibration Standards -
Multiple working standards at specific concentrations must be prepared for
GC/MS calibration because response factors vary widely for compounds analyzed by
this technique. Five initial working calibration standards are required. For full-scan
mode, the standards should be at 0.5, 1.0, 5.0, 10.0, and 15.0 ppbv for each target
analyte. For SIM mode, the standards may be prepared at the same level, or a
standard at lower level may be substituted if the instrument saturates at a level of
15.0 ppbv. The calibration standards must be prepared in humidified zero air.
2.5.4.3 Internal Standard Spiking Mixture —
An internal standard spiking mixture containing bromochloromethane,
chlorobenzene-d5, and 1,4-difluorobenzene at 2 ppbv each in humidified zero air is
required for performing quantitative calculations in the SIM mode. If the internal
standard spiking mixture is being used in the full-scan mode, a concentration of
5 ppbv is required. The internal standard spiking mixture may be prepared from liquid
standards if certified gaseous standards are not available.
2.5.5 Instrument Operating Conditions
All instrument operating conditions must be optimized by the analyst for the
particular instruments and analytes being used.
2.5.5.1 Sample Concentration Conditions --
The VOCs of the canister sample are concentrated on a cryotrap cooled to
approximately -185°C. A predetermined volume of sample air from the collection
canister, consistent with the volume used during calibration, is passed through a
cryotrap. The exact trapping flow rate and time period over which the sample is
concentrated in the cryotrap will vary with the analytical system, but the total volume
trapped must be the same each time. These parameters must be optimized by the
analyst for the particular analytical system used. After concentration, the sample is
transferred from the sample interface system to the head of the capillary column.
2-50
-------
Revision 0
OCTOBER 1991
2.5.5.2 Desorption Conditions --
The following conditions may be used for desorption of cryogenically collected
VOCs:
Desorption temperature: 180°C
Desorption gas flow rate: carrier gas flow rate
Desorption time: <60 seconds
2.5.5.3 Gas Chromatographic Conditions —
Gas Chromatographic conditions must be optimized for each analytical system
in order to provide adequate compound separation and sensitivity. Baseline
separation of the three dichlorobenzene isomers and between ethylbenzene and
m-/p-xylene is indicative of adequate Chromatographic column resolution. It is
desirable that the identical GC column be used for the GC/MS analysis as for the GC
analysis. In this way, sample peak resolution will be nearly identical and confirmation
or unknown compound identification will be simpler and more meaningful.
2.5.5.4 Mass Spectrometer —
The mass spectrometer must be operated at an electron energy of 70 eV
(nominal). The mass range and monitoring time in the SIM mode will be determined
by the target compounds selected. The analytical system must meet the specifications
of the manufacturer for mass calibration in order to provide accurate mass
assignments. The acceptability of the tuning conditions of the mass spectrometer is
established by the analysis of the instrument performance check standard. The
instrument performance check standard should be analyzed at a frequency that
verifies the stability and acceptability of the tuning conditions of the mass
spectrometer.
2-51
-------
Revision 0
OCTOBER 1991
2.5.5.5 Calibration -
Prior to analyzing blanks and samples and after analyzing the instrument
performance check standard, the analytical system must be calibrated at a minimum of
five concentrations to establish instrument sensitivity and linearity of response. (See
Section 2.5.4.2 for details on calibration standards.)
2.5.5.5.1 Multipoint Calibration - Each calibration standard must be
analyzed, adding the appropriate amount of internal standard spiking mixture to each
calibration standard during the sample collection in the cryotrap. The area response
for the primary ion and the corresponding concentration for each compound and
internal standard are recorded. A relative response factor (RRF) for each compound
of interest is calculated using the following equation:
RRF = [VCJ / [Ais/Cis]
Where:
Ax = integrated area of the primary ion for each
compound to be measured
Cx = concentration of the compound to be measured,
ppbv
Ais = integrated area of the primary ion for the internal
standard
Cls = concentration of the internal standard, ppbv
The RRF for each compound is calculated using the values for area and
concentration of the closest-eluting internal standard.
2.5.5.5.2 Acceptance Criteria for Calibration Curve - The average RRF
must be calculated for each compound by averaging the values obtained at the five
concentrations used for the calibration range. The percent relative standard deviation
(% RSD) for each compound must be less than 30 percent. The % RSD of the RRF
2-52
-------
Revision 0
OCTOBER 1991
values across the working range of the calibration curve is calculated using the
following equation:
% RSD = Standard Deviation x 100
Average RRF
If some of the calibration compounds do not meet the criterion, additional
standards are analyzed until the criterion can be met. If the analysis of additional
calibration samples does not produce data that will meet the calibration acceptance
criteria, instrument maintenance and/or repreparation of calibration standards is
required.
2.5.5.5.3 Daily Calibration Check -- A check of the stability of the
calibration curve must be performed every 12 hours. A calibration standard from
approximately the middle of the calibration range (a 5 ppbv standard) should be used
as a calibration check standard. The percent difference (% D) between the RRF
calculated from the daily calibration check standard and the RRF for the original
calibration curve is calculated. The % D is calculated according to the following
equation:
% D = (MRRFX - MRRFJ x 100
MRRFX
Where:
MRRFX = average RRF from the initial calibration curve
MRRFC = RRF obtained from the daily calibration check
standard
The % D must be within ±30% for the calibration curve to be considered valid.
If % D for a given compound is not within ±30%, the analysis of the daily calibration
2-53
-------
Revision 0
OCTOBER 1991
check standard is repeated. If the criterion cannot be met for one or more
compounds after repeated analysis of the daily calibration check standard, the
calibration curve is no longer valid and a new multipoint calibration curve must be
generated.
2.5.5.5.4 System Performance Checks - Internal standard signal levels
and retention times must be monitored for each calibration sample, blank, and field
sample analyzed. If the retention time for any internal standard changes by more than
30 sec from the most recent daily calibration check standard, the analytical system
must be inspected for malfunctions and corrections must be made as indicated. The
peak area for the primary ion for each internal standard must be monitored in each
analysis performed. If the peak area for the primary ion changes by more than 40%,
the analytical system must be inspected for malfunctions and corrections must be
made as indicated. If corrections are made to the analytical system to repair a
malfunction, analyses performed during the period when the analytical system was
malfunctioning must be repeated, if possible. If it is not possible to repeat analyses,
any data obtained from samples analyzed by the malfunctioning analytical system
must be qualified and possibly rejected.
2.5.5.5.5 Frequency of Multipoint Calibration - The multipoint calibration
must be repeated when daily calibration checks no longer meet acceptance criteria. If
adjustments, corrections, or repairs are made to the analytical system, the multipoint
calibration curve must be regenerated if the daily calibration check no longer meets
acceptance criteria.
2.5.6 Analysis Procedures
When the analytical system has been optimized and calibrated, sample analysis
may be initiated. To analyze a canister sample, the canister is connected to the
interface to cryogenically concentrate the VOCs of the air sample. A sample of
accurately known volume is drawn from the canister through a concentration trap,
which is cryogenically cooled to -185°C. During the sample collection process, the
2-54
-------
Revision 0
OCTOBER 1991
gas chromatograph is cooled to the sub-ambient starting point of its temperature
program.
2.5.6.1 Vaporization of Volatile Organic Compounds -
After the volatile organic components of the sample and the internal standard
mixture are condensed on the cryotrap, the sample is injected and the cryotrap is
heated to desorption temperature and swept with helium. The contents of the
cryotrap are vaporized and recondensed on the head of the subambient capillary
column. The canister is closed and removed from the sampling manifold. The
cryotrap is heated and swept with helium until the next analysis to ensure that the
cryotrap will not exhibit any carryover of components from one sample to the next.
2.5.6.2 Initiation of Analysis -
When the sample is transferred to the head of the subambient capillary column,
the GC/MS scan cycle is initiated to allow full-scan analysis or the first SIM monitoring
window.
2.5.6.3 Initial Review of Data -
Upon completion of the analysis, data must be checked for saturation. If the
mass spectrometric peak saturated, no quantitative data can be obtained. Saturation
is unusual when analyzing ambient air samples. Quantitation by a secondary ion may
not be performed if the primary ion is saturated because the calculated values will not
be accurate; a negative bias will be introduced and, depending upon the level of
saturation, the calculated value may be biased negative by a factor of more than ten.
If an analysis is performed and detector saturation is encountered, then (1) the
analysis must be repeated with a diluted sample to obtain accurate quantitative values;
and (2) the analysis which shows saturated compounds must be followed by analyses
of blanks to demonstrate that there will be no carryover of compounds in the analytical
system.
2-55
-------
Revision 0
OCTOBER 1991
2.5.6.4 Sample Dilution -
If the concentration of any analyte exceeds the calibration range, the sample
must be diluted to an appropriate level and the analysis repeated. The need for
dilution is a function of the following:
The level of dilution chosen should be sufficient to keep the
concentration of the largest analyte peak in the upper half of the initial
calibration range of the instrument; and
If other sample components in the original sample analysis are within the
calibration range, the concentration values for these components should
be reported from the original undiluted analysis.
2.5.7 Qualitative Analysis
The compounds listed in Table 2-1 must be identified by the following factors:
The capillary retention time must be within ± 0.06 relative retention time
units of the retention time of the compound in either a calibration
standard or the daily calibration check standard;
The mass spectrum of the compound of interest must correspond to the
SIM mass spectrum of a standard of the compound generated on the
same analytical system on which the sample analysis is being performed;
and
Where multiple ions are monitored for a given compound, these ions
must maximize within 1 second of each other.
Because SIM analysis is being performed, the mass spectra will in most cases
consist of no more than three peaks. For halogenated compounds, two peaks of an
isotopic cluster will be selected so that a theoretical ratio for the presence of halogen
can be verified. A minimal requirement for the identification of a compound in the SIM
mode is that the primary ion be present within an interval of 0.06 relative retention time
units of the retention time of the authentic compound. For chlorinated compounds,
the primary ion must be present at the correct retention time and the secondary ion of
the isotope cluster must be present in the correct ratio to the primary ion.
2-56
-------
Revision 0
OCTOBER 1991
Table 2-1
Primary Quantitation Ions for Compounds of Interest
Compound Name
Propane
Propylene
Isobutane
n-Butane
Isopentane
n-Pentane
2-Methylpentane
3-Methylpentane
Benzene
Toluene
Ethylbenzene
m-/p-Xylene
o-Xylene
1 ,2,4-Trimethylbenzene
Isoprene
2,2-Dimethylbutane
Cyclopentane
2,3-Dimethylbutane
Methylcyclopentane
2,4-Dimethylpentane
2,2,4-Trimethylpentane
n-Heptane
Methylcyclohexane
2,3,4-Trimethylpentane
2-Methylheptane
3-Methylheptane
trans-2-Butene
cis-2-Butene
trans-2-Pentene
cis-2-Pentene
3-Methyl-1-butene
2-Methyl-1-butene
alpha-Pinene
beta-Pinene
1-Butene
Primary Ion
43
41
57
57
57
57
57
57
78
92
91
91
91
105
67
57
42
43
56
43
57
43
83
71
43
85
41
41
55
55
55
55
93
93
41
Confirmation
44
42
58
58
72
72
72
72
-
91
106
106
106
120
53
71
70
42
84
85
56
100
98
70
57
84
56
56
70
70
70
70
121
69
56
NOTE: This table must be expanded by the inclusion of relative retention times (not available) and theoretical ratios
between the primary and secondary ions (not calculated until a definite list of compounds is available.
2-57
-------
Revision 0
OCTOBER 1991
2.5.8 Quantitative Analysis
Identified target compounds are quantified by the RRF method using peak
areas for the primary ions. If internal standard retention times or peak areas are within
the acceptable range, quantitative analysis can be performed. The concentration of
the compound of interest in air (in ppbv) is calculated by the following equation:
Where:
DF
Ais
RRF
(AJ (Cis) (DF)
(Ais) (RRF)
integrated area of the primary quantitation ion for the
compound of interest
concentration of internal standard, ppbv
dilution factor, if applicable. If no dilution has been
performed, the dilution factor is 1
integrated area of the primary quantitation ion for the
nearest-eluting internal standard
relative response factor for the compound of interest
from the calibration curve
2.5.9 Additional Uses
The GC/MS analytical system can also be used for primary analysis in either
the full-scan mode or the SIM mode. Qualitative identification of unknown compounds
is possible using full-scan GC/MS. If positive identification is possible and these
tentatively identified compounds occur frequently, the GC/FID system used for routine
analysis may be characterized for these compounds and they may be added to the
target compound list.
2-58
-------
Revision 0
OCTOBER 1991
SECTION 3.0
METHODOLOGY FOR DETERMINING
TOTAL NONMETHANE ORGANIC COMPOUNDS IN AMBIENT AIR
Qualitative and quantitative determinations of VOCs using the GC based
methodology described in Section 2.0 requires instrumentation that is expensive,
complex, and difficult to operate and maintain. The method described in this section
provides a similar measurement of total NMOC. Although this method is not directly
applicable to the enhanced O3 monitoring, it is included here because (1) it is a viable
and effective method of post clean-up determinations of canister cleanliness; (2) it can
be used for ambient total NMOC measurements of input into O3 predictive models that
do not require VOC speciation; and (3) used in combination with the manual (canister)
methodology described in Section 2.3, it may provide a viable alternative to the
automated methodology described in Section 2.4. The method described in this
section is Method TO-12, taken from the "Compendium of Methods for the
Determination of Toxic Organic Compounds in Ambient Air".3 Method TO-12 involves
a simple, preconcentration procedure with subsequent, direct flame ionization
detection and provides accurate and sensitivity measurements of total NMOC
concentrations. The instrumentation for this method can be configured for either •
automated in-situ measurements or for analyzing integrated samples collected in
canisters.
3.1 METHOD DESCRIPTION
In recent years, the relationship between ambient concentrations of precursor
organic compounds and subsequent downwind concentrations of O3 has been
described by a variety of photochemical dispersion models. The most important
application of such models is to determine the degree of control of precursor organic
compounds that is necessary in an urban area to achieve attainment of the NAAQS for
3-1
-------
Revision 0
OCTOBER 1991
The more elaborate theoretical models generally require detailed organic
species data obtained by multicomponent GC. The Empirical Kinetic Modeling
Approach (EKMA), however, requires only the total NMOC concentration data -
specifically, the average total NMOC concentration from 6 a.m. to 9 a.m. daily at the
sampling location. The use of total NMOC concentration data in the EKMA
substantially reduces the cost and complexity of the sampling and analysis system by
not requiring qualitative and quantitative species identification.
The method presented in this section is Compendium Method TO-12, which
combines the same type of cryogenic concentration technique used in Method TO-01
for high sensitivity total NMOC measurements, without the GC columns and complex
procedures necessary for species separation.
In an FID, the sample is injected into a hydrogen-rich flame where the organic
vapors burn, producing ionized molecular fragments. The resulting ion fragments are
then collected and detected. The FID is a nearly universal detector; however, detector
response varies with the species of the organic compound in an oxygen atmosphere.
Because Method TO-12 employs a helium or argon carrier gas, the detector response
is nearly uniform for many hydrocarbon compounds. Thus, the historical shortcoming
of the FID, varying detector response to different organic functional groups, is
minimized.
Method TO-12 can be used either for direct, in-situ ambient measurements or
(more commonly) for analyzing integrated samples collected in specially treated
stainless steel canisters. EKMA models generally require 3-hour integrated NMOC
measurements over the 6 a.m. to 9 a.m. period, and are used by State or local
agencies in preparing their SIPs for O3 control to achieve compliance with the NAAQS
for O3. For direct, in-situ ambient measurements, the analyst must be present during
the 6 a.m. to 9 a.m. period, and repeat measurements (approximately six per hour)
must be taken to obtain average NMOC concentration for the period. The use of
sample canisters allows the collection of integrated air samples over the 6 a.m. to
3-2
-------
Revision 0
OCTOBER 1991
9 a.m. period by unattended, automated samplers. Method TO-12 incorporates both
sampling approaches.
3.2 SUMMARY OF METHOD TO-12
A whole air sample is either extracted directly from the ambient air and analyzed
on site by the GC system or collected into a precleaned sample canister and analyzed
off site.
The analysis requires drawing a fixed-volume portion of the sample air at a low
flow rate through a glass-bead-filled trap that is cooled to approximately -185°C with
liquid argon. The cryogenic trap simultaneously collects and concentrates the NMOC
(via either condensation or absorption), while allowing the methane, nitrogen, oxygen,
etc., to pass through the trap without retention. The system is dynamically calibrated
so that the volume of sample passing through the trap does not have to be
quantitatively measured, but the sample volume must be precisely repeatable between
the calibration and the analytical phases.
After the fixed-volume air sample has been drawn through the trap, a helium
carrier gas flow is diverted to pass through the trap in the opposite direction to the
sample flow and into an FID. When the residual air and methane have been flushed
from the trap and the FID baseline restabilizes, the cryogen is removed and the
temperature of the trap is raised to approximately 90°C.
The organic compounds previously collected in the trap revolatilize because of
the increase in temperature and are carried into the FID, resulting in a response peak
or peaks from the FID. The area of the peak or peaks is integrated, and the
integrated value is translated to concentration units via a previously-obtained
calibration curve relating integrated peak areas to known concentrations of propane.
By convention, concentrations of NMOC are reported in units of parts per
million Carbon (ppmC), which, for a specific compound, is the concentration by parts
per million volume (ppmv) multiplied by the number of carbon atoms in the compound.
The cryogenic trap simultaneously concentrates the NMOC while separating
and removing the methane from air samples. The technique is thus direct reading for
3-3
-------
Revision 0
OCTOBER 1991
NMOC and, because of the concentration step, is more sensitive than conventional
continuous NMOC analyzers.
3.3 SIGNIFICANCE
Accurate measurements of ambient concentrations of NMOC are important for
the control of photochemical smog because these organic compounds are primary
precursors of atmospheric ozone and other oxidants. Attainment of the NAAQS for
ozone is therefore dependent on control of ambient levels of NMOC and other O3
precursors.
The NMOC concentrations typically found at urban sites may reach 5-7 ppmC
or above. In order to determine transport of precursors into an area, measurement of
NMOC upwind of the area may be necessary. Upwind NMOC concentrations are
likely to be less than a few tenths of 1 ppmC.
3.4 INTERFERENCES
In field and laboratory evaluations, water in the air sample was found to cause a
positive shift in the FID baseline. The effect of this shift is minimized by carefully
selecting the integration termination point and adjusting the baseline used for
calculating the area of the NMOC peak(s).
When using helium as a carrier gas, FID response is quite uniform for most
hydrocarbon compounds, but the response can vary considerably for other types of
organic compounds.
3.5 EQUIPMENT
3.5.1 Direct Air Sampling (See Figure 3-1)
Sample manifold or sample inlet line - to bring sample air into the analytical
system.
Vacuum pump or blower - to draw sample air through a sample manifold or
long inlet line to reduce inlet residence time. Maximum residence time should be no
greater than 1 minute.
3.5.2 Remote Sample Collection in Pressurized Canisters (See Section 2.3)
3-4
-------
Vacuum
Vatve
Vacuum
Pump
Canister
Valve
Absolute Pressure
Gauge
Revision 0
OCTOBER 1991
Pressure
Regulator
Gas
Purifier
1 A
• —
Dewar
Flask
Glass
Beads
Fine
Needle
Valve
(sample flow
adjustment)
Vacuum
Reservior
6-Port
Gas
Valve
Direct Air Sampling
Cryogenic
Trap Cooler
(liquid argon)
Vent —'
Pressure
Regulator
optional fine
needle valve)
Pressure
Regulator
Integrator
Recorder
Figure 3-1. Schematic of analytical system for NMOC-two sampling modes.
3-5
-------
Revision 0
OCTOBER 1991
3.5.3 Sample Canister Cleaning (See Section A3.0)
3.5.4 Analytical System (See Figure 3-1)
FID detector system - including flow controls for the FID fuel and air,
temperature control for the FID, and signal processing electronics. The FID burner air,
hydrogen, and helium carrier flow rates should be set according to the manufacturer's
instructions in order to obtain an adequate FID response while maintaining as stable a
flame as possible throughout all phases of the analytical cycle.
Chart recorder - compatible with the FID output signal, to record FID response.
Electronic integrator - capable of integrating the area of one or more FID
response peak(s) and calculating peak area corrected for baseline drift. If a separate
integrator and chart recorder are used, care must be exercised to ensure that these
components do not interfere with each other electrically. Range selector controls on
both the integrator and the FID analyzer may not provide accurate range ratios, so
individual calibration curves should be prepared for each range to be used. The
integrator should be capable of marking the beginning and ending of peaks,
constructing the appropriate baseline between the start and end of the integration
period, and calculating the peak area.
Trap - the trap should be carefully constructed from a single piece of
chromatographic-grade stainless steel tubing (0.32 cm O.D., 0.21 cm I.D.) as shown in
Figure 3-2. The central portion of the trap (7-10 cm) is packed with 60/80 mesh glass
beads, with small glass wool (dimethyldichlorosilane-treated) plugs to retain the
beads. The trap must fit conveniently into the Dewar flask and the arms must be of an
appropriate length to allow the beaded portion of the trap to be submerged below the
level of liquid cryogen in the Dewar. The trap should connect directly to the six-port
valve, if possible, to minimize line length between the trap and the FID. The trap must
be mounted to allow the Dewar to be slipped conveniently onto and off of the trap and
also to facilitate heating the trap.
3-6
-------
Revision 0
OCTOBER 1991
Tube Length: ~30cm
O.D.: 0.32 cm
I.D.: 0.21 cm
"13 cm
Cryogenic Liquid Level
60/80 Mesh Glass Beads
'4cm-
(to fit dewar)
Figure 3-2. Cryogenic sample trap dimensions.
3-7
-------
Revision 0
OCTOBER 1991
Six-port chromatoaraphic valve - six-port valve and as much of the
interconnecting tubing as practical should be located inside an oven or otherwise
heated to 80-90°C to minimize wall losses or adsorption/desorption in the connecting
tubing. All lines should be as short as practical.
Multistage pressure regulators - standard two-stage, stainless steel diaphragm
regulators with pressure gauges, for helium, air, and hydrogen cylinders.
Pressure regulators - optional single stage, stainless steel, with pressure gauge,
if needed, to maintain constant helium carrier and hydrogen flow rates.
Fine needle valve - to adjust sample flow rate through trap.
Dewar flask - to hold liquid cryogen to cool the trap, sized to contain
submerged portion of trap.
Absolute pressure gauge - 0-450 mm Hg (with 2 mm Hg scale divisions), to
meter reproducible volumes of sample air through cryogenic trap.
Vacuum reservoir - 1-2 L capacity, typically 1 L
Gas purifiers - gas scrubbers containing Drierite® or silica gel and 5A molecular
sieve to remove moisture and organic impurities in the helium, air, and hydrogen gas
flows. Check the purity of gas purifiers prior to use. Gas purifiers are clean if they
produce less than 0.02 ppmC NMOC.
Trap heating system - chromatographic oven, hot water, or other means to heat
the trap to 80°-90°C. A simple heating source for the trap is a beaker or Dewar filled
with water maintained at 80-90°C. More repeatable types of heat sources are
recommended, including a temperature-programmed chromatograph oven, direct
electrical heating of the trap itself, or any type of heater that brings the temperature of
the trap up to 80-90°C in 1-2 minutes.
Toggle shut-off valves - leak free, for vacuum valve and sample valve.
Vacuum pump - general purpose laboratory pump capable of evacuating the
vacuum reservoir to an appropriate vacuum that allows the desired sample volume to
be drawn through the trap.
3-8
-------
Revision 0
OCTOBER 1991
Vent - to keep the trap at atmospheric pressure during trapping when using
pressurized canisters.
Rotameter to verify vent flow.
Fine needle valve (optional) - to adjust flow rate of sample from canister during
analysis.
Tubing and Fittings - Chromatographic-grade stainless steel tubing and
stainless steel fittings are used for interconnections. All such materials that contact the
sample, analyte, or support gases prior to analysis should be stainless steel or other
inert metal. Do not use plastic or Teflon® tubing or fittings.
3.6 REAGENTS AND MATERIALS
Gas cylinder of helium and hydrogen - ultrahigh purity grade.
Combustion air - cylinder containing less than 0.02 ppmC hydrocarbons, or
equivalent air source.
Propane calibration standard - cylinder containing 1-100 ppmv (3-300 ppmC)
propane in air. The cylinder assay should be traceable to an NIST SRM or to an
NIST/EPA-approved CRM.
Zero air - cylinder containing less than 0.02 ppmC hydrocarbons. Zero air may
be obtained for a cylinder of zero-grade compressed air scrubbed with Drierite® or
silica gel and 5A molecular sieve or activated charcoal, or by catalytic cleanup of
ambient air. All zero air should be passed through a liquid argon cold trap for final
cleanup, then passed through a hydrocarbon-free water bubbler for humidification.
Liquid cryogen - liquid argon or liquid oxygen may be used as the cryogen --
experiments have shown no differences in trapping efficiency between the two.
However, appropriate safety precautions must be taken if liquid oxygen, is used.
Liquid nitrogen should not be used because it causes condensation of oxygen and
methane in the trap.
3.7 DIRECT SAMPLING
For direct ambient air sampling, the cryogenic trapping system draws the air
sample directly from a pump-ventilated distribution manifold or sample line. The
3-9
-------
Revision 0
OCTOBER 1991
connecting line should be of small diameter (1/8" O.D.) stainless steel tubing and as
short as possible to minimize its dead volume.
Multiple analyses over the sampling period must be made to establish hourly or
3-hour NMOC concentration averages.
3.8 SAMPLE ANALYSIS
Preanalysis and analysis procedures are contained in this section.
3.8.1 Analytical System Leak Check
1. Before sample analysis, the analytical system is assembled and leak
checked.
2. To leak check the analytical system, place the six-port gas valve in the
trapping position. Disconnect and cap the absolute pressure gauge. Insert a
pressure gauge capable of recording up to 60 pounds per square-inch gauge (psig) at
the vacuum valve outlet.
3. Attach a valve and a zero air supply to the sample inlet port. Pressurize
the system to about 50 psig and close the valve.
4. Wait approximately 3 hours and re-check pressure. If the pressure did
not vary more than ± 2 psig, the system is considered leak tight.
5. If the system is leak free, de-pressurize and reconnect absolute pressure
gauge.
6. The analytical system leak check procedure should be performed during
the system checkout, prior to a series of analyses, or if leaks are suspected. This
procedure should be part of the user-prepared SOP manual.
3.8.2 Sample Volume Determination
1. The vacuum reservoir and absolute pressure gauge are used to meter a
precisely repeatable volume of sample air through the cryogenically cooled trap. To
do this, the sample valve is closed, the vacuum valve opened, and the reservoir is
evacuated to a predetermined pressure (e.g., 100 mm Hg) at which point the vacuum
valve is closed. The sample valve is then opened to allow sample air to be drawn
3-10
-------
Revision 0
OCTOBER 1991
through the cryogenic trap and into the evacuated reservoir until a second
predetermined reservoir pressure (e.g., 300 mm Hg) is reached. The fixed volume of
air sampled is determined by the pressure rise in the vacuum reservoir (difference
between the predetermined pressures), as measured by the absolute pressure gauge.
2. The sample volume can be calculated by:
Vs = (AP) (Vr)
(Ps)
Where:
Vs = Volume of air sampled (standard ml_)
A P = Pressure difference measured by gauge (mm Hg)
Vr = Volume of vacuum reservoir (ml), usually 1000
Ps = Standard pressure (760 mm Hg)
For example, with a vacuum reservoir of 1000 ml_ and a pressure change of 200 mm
Hg (e.g., 100 to 300 mm Hg), the volume sampled would be 263 mm. A typical
sample volume using this procedure is between 200-300 ml_.
3. The sample volume determination need only be performed once during
the system check-out and shall be part of the user prepared SOP manual.
3.8.3 Analytical System Dynamic Calibration
1. Before sample analysis, a complete dynamic calibration of the analytical
system should be carried out at five or more concentrations on each range, to define
the calibration curve. The calibration procedure should be carried out initially and
periodically thereafter, and should be part of the user-prepared SOP manual. The
calibration should be verified with two- or three-point calibration checks (including
zero) each day the analytical system is used to analyze samples.
3-11
-------
Revision 0
OCTOBER 1991
2. Concentration standards of propane are used to calibrate the analytical
system. Propane calibration standards may be obtained directly from low
concentration cylinder standards or by dilution of high concentration cylinder
standards with zero air. Dilution flow rates must be measured accurately and the
combined gas stream must be mixed thoroughly for successful calibration of the
analyzer. The calibration standard should be sampled directly from a vented manifold
or tee. The propane NMOC concentration in ppmC is three times the volumetric
concentration in parts per million volume.
3. Select one or more combinations of the following parameters to provide
the desired range or ranges (e.g., 0-1.0 ppmC or 0-5.0 ppmC): FID attenuator setting,
output voltage setting, integrator resolution (if applicable), and sample volume. Each
individual range should be calibrated separately and should have a separate
calibration curve. Modern GC integrators may provide automatic ranging so that
several decades of concentration may be covered in a single range. The user-
prepared SOP manual should address variations applicable to a specific system
design.
4. Analyze each calibration standard three times according to the procedure
in Section 3.8.4. Insure that flow rates, pressure gauge start and stop readings, initial
cryogen liquid level in the Dewar, timing, heating, integrator settings, and other
variables are the same as those that will be used during analysis of ambient samples.
Typical flow rates for the gases are: hydrogen, 30 mL/minute; helium carrier,
30 mL/minute; and burner air, 400 mL/minute.
5. Average the three analyses for each concentration standard and plot the
calibration curve(s) as average integrated peak area reading versus concentration in
ppmC. The relative standard deviation for the three analyses should be less than 3%
(except for zero concentration). Linearity should be expected; points that appear to
deviate abnormally should be repeated. Response has been shown to be linear over
a wide range (0-10,000 ppbC). If nonlinearity is observed, an effort should be made to
identify and correct the problem. If the problem cannot be corrected, additional points
3-12
-------
Revision 0
OCTOBER 1991
in the nonlinear region may be needed in order to adequately define the calibration
curve.
3.8.4 Analysis Procedure
1. Ensure that the analytical system has been assembled properly, leak
checked, and properly calibrated through a dynamic standard calibration. Ignite the
FID detector and allow to stabilize.
2. Check and adjust the helium carrier pressure to provide the correct
carrier flow rate for the system. Helium is used to purge residual air and methane
from the trap at the end of the sampling phase and to carry the revolatilized NMOC
from the trap into the FFD.
3. Close the sample valve and open the vacuum valve to evacuate the
vacuum reservoir to a specific predetermined value (e.g., 100 mm Hg).
4. With the trap at room temperature, place the six-port valve in the inject
position.
5. Open the sample valve and adjust the sample flow rate needle valve for
an appropriate trap flow of 50-100 mL/minute.
6. Connect the sample canister or direct sample inlet to the six-port valve,
as shown in Figure 3-1. For a canister, either the canister valve or an optional fine
needle valve installed between the canister and the vent is used to adjust the canister
flow rate to a value slightly higher than the trap flow rate set by the sample flow rate
needle valve. The excess flow exhausts through the vent, which ensures that the
sample air flowing through the trap is at atmospheric pressure. The vent is connected
to a flow indicator such as a rotameter as an indication of vent flow to assist in
adjusting the flow control valve. Open the canister valve and adjust the canister valve
or the sample flow needle valve to obtain a moderate vent flow, as indicated by the
rotameter. The sample flow rate will be lower and the vent flow rate will be higher
when the trap is cold.
7. Close the sample valve and open the vacuum valve to evacuate the
vacuum reservoir. With the six-port valve in the inject position and the vacuum valve
3-13
-------
Revision 0
OCTOBER 1991
open, open the sample valve for 2-3 minutes to flush and condition the inlet lines(with
both valves open, the pressure reading won't change).
8. Close the sample valve and evacuate the reservoir to the predetermined
sample starting pressure (typically 100 mm Hg) as indicated by the absolute pressure
gauge.
9. Switch the six-port valve to the sample position.
10. Submerge the trap in the cryogen. Allow a few minutes for the trap to
cool completely as indicated when the cryogen stops boiling. Add cryogen to the
initial level used during system dynamic calibration. The level of the cryogenic liquid
should remain constant with respect to the trap and should completely cover the
packed area of the trap.
11. Open the sample valve and observe the increasing pressure on the
pressure gauge. When it reaches the predetermined pressure representative of the
desired sample volume (typically 300 mm Hg), close the sample valve.
12. Add a little cryogen or elevate the Dewar to raise the liquid level to a
point slightly higher (3-15 mm) than the initial level at the beginning of the trapping.
This ensures that organics do not escape from the trap before being integrated as
part of the NMOC peak(s).
13. Switch the six-port valve to the inject position, keeping the cryogenic'
liquid on the trap until the methane and pressure peaks have diminished
(10-20 seconds). Now close the canister valve to conserve the remaining sample in
the canister.
14. Start the integrator and remove the Dewar flask containing the cryogenic
liquid from the trap.
15. Close the GC oven door and allow the GC oven (or alternate trap heating
system) to heat the trap at a predetermined rate (typically, 30°C/min) to 90°C. Heating
the trap volatilizes the concentrated NMOC. A uniform trap temperature rise rate helps
reduce variability and facilitates more accurate correction for the moisture-shifted
baseline. With a chromatograph oven to heat the trap, the following parameters have
3-14
-------
Revision 0
OCTOBER 1991
been found to be acceptable: initial temperature, 30°C; initial time, 0.20 minutes
(following start of the integrator); heat rate, 30°C/min; final temperature, 90°C.
16. Use the same heating process and temperatures for both calibration and
sample analysis. Heating the trap too quickly may cause an initial negative response
that could hamper accurate integration. Some initial experimentation may be
necessary to determine the optimal heating procedure for each system. Once
established, the procedure should be consistent for each analysis as outlined in the
user-prepared SOP manual.
17. Continue the integration only long enough to include all of the organic
compound peaks and to establish the end point FID baseline, as illustrated in
Figure 3-3. The integrator should be capable of marking the beginning and ending of
peaks, constructing the appropriate operational baseline between the start and end of
the integration period, and calculating the resulting corrected peak area. This ability is
necessary because the moisture in the sample, which is also concentrated in the trap,
will cause a slight positive baseline shift. This baseline shift starts as the trap warms
and continues until all of the moisture is swept from the trap, at which time the
baseline returns to its normal level. The shift always continues longer than the
ambient organic peak(s). The integrator should be programmed to correct for this
shifted baseline by ending the integration at a point after the last NMOC peak and '
prior to the return of the shifted baseline to normal so that the calculated operational
baseline effectively compensates for the water-shifted baseline. Electronic integrators
either do this compensation automatically or they should be programmed to make the
correction. Alternatively, analyses of humidified zero air prior to sample analyses
should be performed to determine the water envelope and the proper blank value for
correcting the ambient air concentration measurements accordingly. Heating and
flushing of the trap should continue after the integration period has ended to ensure
that all water has been removed. The six-port valve should remain in the inject
position until all moisture has purged from the trap.
3-15
-------
Revision 0
OCTOBER 1991
NMOC Peak
8.
en
-------
Revision 0
OCTOBER 1991
18. Use the dynamic calibration curve to convert the integrated peak area
reading into concentration units (ppmC). Note that the NMOC peak shape may not
be precisely reproducible because of variations in heating the trap, but the total NMOC
peak area should be reproducible.
19. Analyze each canister sample at least twice and report the average
NMOC concentration. Problems during an analysis occasionally will cause erratic or
inconsistent results. If the first two analyses do not agree within ± 5 %RSD, additional
analyses should be made to identify inaccurate measurements and produce a more
accurate average.
3.9 PERFORMANCE CRITERIA AND QUALITY ASSURANCE
This section summarizes required quality assurance measures and provides
guidance concerning performance criteria that should be achieved within each
laboratory.
3.9.1 Standard Operating Procedures
1. Users should generate SOPs that describes and documents the following
activities in their laboratory:
• Assembly, calibration, leak check, and operation of the specific sampling
system and equipment used;
• Preparation, storage, shipment, and handling of samples;
• Assembly, leak check, calibration, and operation of the analytical system,
addressing the specific equipment used;
• Canister storage and cleaning; and
• All aspects of data recording and processing, including lists of computer
hardware and software used.
3-17
-------
Revision 0
OCTOBER 1991
2. The SOP should provide specific stepwise instructions and should be
readily available to, and understood by, the laboratory personnel conducting the work.
3.9.2 Method Sensitivity. Accuracy. Precision, and Linearity
The sensitivity and precision of Method TO-12 are proportional to the sample
volume. However, ice formation in the trap may reduce or stop the sample flow
during trapping if the sample volume exceeds 500 ml_. Sample volumes below
100-150 ml may cause increased measurement variability because of dead volume in
lines and valves. For most typical ambient NMOC concentrations, sample volumes in
the range of 200-400 mL appear to be appropriate. If a response peak obtained with
a 400 mL sample is off-scale or exceeds the calibration range, a second analysis can
be carried out with a smaller volume.
The actual sample volume used need not be accurately known if it is precisely
repeatable during both calibration and analysis. Similarly, the actual volume of the
vacuum reservoir need not be accurately known, but the reservoir volume should be
matched to the pressure range and resolution of the absolute pressure gauge so that
the measurement of the pressure change in the reservoir is repeatable within
1 percent. A 1000 mL vacuum reservoir and a pressure change of 200 mm Hg,
measured with the specified pressure gauge, have provided a sampling precision of
± 1.31 mL A smaller volume reservoir may be used with a greater pressure change
to accommodate absolute pressure gauges with lower resolution, and vice versa.
Some FID systems associated with laboratory chromatographs may have
autoranging. Others may provide attenuator control and internal full-scale output
voltage selectors. An appropriate combination should be chosen so that an adequate
output level for accurate integration is obtained down to the detection limit; however,
the electrometer or integrator must not be driven into saturation at the upper end of
the calibration. Saturation of the electrometer may be indicated by flattening of the
calibration curve at high concentrations. Additional adjustments of range and
sensitivity can be provided by adjusting the sample volume used.
3-18
-------
Revision 0
OCTOBER 1991
System linearity has been documented from 0 to 10,000 ppbC.
Some organic compounds contained in ambient air are "sticky" and may require
repeated analyses before they fully appear in the FID output. Also, some adjustment
may have to be made in the integrator off-time setting to accommodate compounds
that reach the FID late in the analysis cycle. Similarly, "sticky" compounds from
ambient samples or from contaminated propane standards may temporarily
contaminate the analytical system and can affect subsequent analyses. Such
temporary contamination can usually be removed by repeated analyses of humidified
zero air.
3.10 METHOD MODIFICATIONS
Potential modifications to the method are contained in this section.
3.10.1 Sample Metering System
Although the vacuum reservoir and absolute pressure gauge technique for
metering the sample volume during analysis is efficient and convenient, other
techniques should work. A constant sample flow could be established with a vacuum
pump and a critical orifice, with the six-port valve being switched to the sample
position for a measured time period. A gas volume meter, such as a wet test meter,
could also be used to measure the total volume of sample air drawn through the trap.
These alternative techniques should be tested and evaluated as part of a user-
prepared SOP manual.
3.10.2 Flame lonization Detection System
A variety of FID systems should be adaptable to the method. The specific flow
rates and necessary modifications for the helium carrier for any alternative FID
instrument should be evaluated prior to inclusion in the SOP manual.
3.10.3 Range
It may be possible to increase the sensitivity of the method by increasing the
sample volume. However, limitations may arise such as plugging of the trap by ice.
3-19
-------
Revision 0
OCTOBER 1991
SECTION 4.0
METHODOLOGY FOR MEASURING OXIDES OF NITROGEN
AND TOTAL REACTIVE OXIDES OF NITROGEN IN AMBIENT AIR
This section addresses the monitoring provisions of 40 CFR Part 58 for
measuring NOX and provides a general discussion of NOy.
4.1 OXIDES OF NITROGEN
Nitric oxide (NO) and nitrogen dioxide (NO2) are the constituents of NOX.
Oxides of nitrogen are a principal precursor to the formation of O3.
Models such as EKMA, that are used to predict downwind concentrations of 03
require NOX and NMOC concentrations as inputs. The Urban Airshed Model (DAM),
another type of prediction model, requires NOX, total NMOC, and VOC speciation
information as inputs.
4.1.1 Measurement of Oxides of Nitrogen
4.1.1.1 Method Description -
The NOX compounds are typically measured using a chemiluminescent
instrument. The principle of operation of this instrumentation is based on the
chemiluminescent reaction of NO and O3 specifically,
NO + O3 — > NO2 + O2
The reaction causes electronically excited NO2 molecules to revert to their ground
state, resulting in an emission of light or chemiluminescence.
The NO is measured by blending the sample gas with O3 in a reaction
chamber. The chemiluminescence that results is monitored through an optical filter by
a high-sensitivity photomultiplier. The filter and photomultiplier respond to light in a
narrow-wavelength band unique to the reaction presented above. The signal
developed by the photomultiplier is proportional to the NO concentration. To
4-1
-------
Revision 0
OCTOBER 1991
determine concentration of NOX (i.e., NO + NO2), the sample gas is routed through an
N02-to-NO converter. The chemiluminescent response in the reaction chamber to the
converter effluent gas stream is proportional to the concentration of NOX entering the
converter.
There are basically two type of NO2-to-NO converters. These are:
A catalytic converter using a catalyst such as molybdenum or gold; and
• A photolytic converter using a high energy light source.
Catalytic conversion NO2 analyzers permit accurate measurement of NOX as
long as the nitroxyl compounds occurring in the sampled atmosphere are limited to
NO and NOX. Peroxyacetyl nitrate (PAN) and nitric acid (HNO3) are primary potential
interferents to the accurate measurement of NOX when a catalytic converter is used. A
catalytic converter may partially convert HNO3, and PAN to NO. This conversion
causes artificially high values for NO2 and NOX. The potential for biasing the NOX
measurement because of the catalytic conversion of PAN and HNO3 is greatest in
areas where these compounds comprise a significant percentage of the total airborne
compounds containing N2 and oxygen (O2). In this situation, the photolytic converter
can be more accurate. The photolytically activated converter is a high energy light
source at wavelengths from 300 nanometer (nm) to 430 nm that converts NO2 to
NO+O2, which is then mixed with O3 and measured by the chemiluminescent detector.
The photolytic converter is not 100% efficient and must be calibrated to determine its
efficiency.
4.1.1.2 Methods and Equipment --
Methods for measuring ambient concentrations of NO2 that have been
designated as reference or equivalent methods are presented in 40 CFR Part 538.
Subject to any limitations specified in the applicable designation, each method listed in
40 CFR Part 53 is acceptable for use at PAMS unless the applicable designation is
4-2
-------
Revision 0
OCTOBER 1991
subsequently canceled. Instruments designated as reference methods for NO2 are
also approved for measuring NO, and are acceptable for determination of NOX.
4.2 TOTAL REACTIVE OXIDES OF NITROGEN
The nitroxyl compounds in ambient air included in the group of specific
compounds referred to as NOy have not been definitively determined. This group of
compounds should contain all of the oxides of nitrogen that react in the troposphere
to any significant extent, and therefore, potentially contribute to the formation of O3.
Identified NOy constituents include:
NO;
N02;
nitrogen trioxide;
• nitrogen pentoxide;
• nitrous acid;
HNO3;
• peroxy nitrite;
PAN;
• other organic nitrates; and
• aerosol nitrates.
Although the measurement of NO is not required by the proposed revisions to 40
CFR Part 58, NOy is discussed because of its potential as a precursor to O3 formal
4.2.1 Measurement of Total Reactive Oxides of Nitrogen
There are two primary techniques to measure NO These techniques are:
4-3
-------
Revision 0
OCTOBER 1991
Analyzing separately for each constituent using an appropriate specific
method, and summing the results to determine NOy; and
Catalytically converting all the compounds containing N2 and O2 to NO
and measuring the NO using a chemiluminescence instrument.
Measured NO concentration is assumed to equal NOy.
Analyzing separately for each constituent may be impractical because it requires that
the user know all the compounds that are to be measured so that appropriate
individual methods can be applied. Comprehensive catalytic reduction to NO usually
yields NOy concentrations higher than if the concentrations of the constituent parts of
NOy are measured separately and summed.
4-4
-------
Revision 0
OCTOBER 1991
SECTION 5.0
METHODOLOGY FOR DETERMINING
CARBONYL COMPOUNDS IN AMBIENT AIR
Formaldehyde and other carbonyl compounds have been shown to be major
promoters in the formation of photochemical O3. Determination of ambient
formaldehyde, acetaldehyde, and acetone concentrations is a required specification of
the proposed ambient air monitoring revisions to 40 CFR Part 58. Details on carbonyl
sampling frequency are presented in Table 1-1. The methodology contained in
Section 5.1 presents procedures for sampling and analyzing carbonyl compounds
utilizing a solid adsorbent and high performance liquid chromatography (HPLC)
detection. This method is sensitive and provides accurate measurements of carbonyl
compounds.
5.1 METHOD DESCRIPTION
This section presents Compendium Method TO-11 which is used for
determination of formaldehyde and other carbonyl compounds in ambient air, utilizing
a solid adsorbent followed by HPLC detection.
Method TO-11 is a modification of Compendium Method TO-05, "Method For
the Determination of Aldehydes and Ketones in Ambient Air Using High Performance
Liquid Chromatography." Carbonyl compounds readily form a stable derivative with
2,4-dinitrophenylhydrazine (DNPH) reagent. The DNPH derivative is analyzed by
HPLC. In Method TO-11, the DNPH reagent is coated on a solid adsorbent and used
for sample collection. The method is currently based on the specific reaction of
organic carbonyl compounds (aldehydes and ketones) with DNPH cartridges in the
presence of an acid to form stable derivatives according to the equation shown in
Figure 5-1.
The sampling method gives a time-weighted average sample. The same
sampling methodology can be used for long-term (1-24 hr) sampling of ambient air
where the concentration of carbonyl compounds is generally in the low (1-20) ppb
5-1
-------
NO2
NO2
\
C=O + H2N-NH
\
N02
C = N - NH
NO2 + H2O
Carbonyl Group
(Aldehydes and Ketones)
2,4-dlnttrophenylhydrazine
(DNPH)
DNPH - Derivative
Wafer
01
rb
R and R1 are organic alkyl or aromatic group (ketones) or either substituent
is a hydrogen (aldehydes)
5616538R
o
o
CD
m
Figure 5-1. Formation of a stable derivative.
CD
-------
Revision 0
OCTOBER 1991
(v/v), or for short-term (5-60 min)sampling of source-impacted atmospheres where the
concentration of carbonyl compounds could reach the ppm (v/v) levels.
The sampling flow rate, as described in this document, is presently limited to
about 1.5 L/min. This limitation is principally due to the high pressure drop across the
DNPH-coated silica gel cartridges. Because of this limitation, the procedure is not
compatible with pumps used in personal sampling equipment.
Method TO-11 instructs the user to purchase Sep-PAK® chromatographic grade
silica gel cartridges and apply acidified DNPH in-situ to each cartridge as part of the
user-prepared QA program. Commercially pre-coated DNPH cartridges are also
available. Recent studies have indicated abnormally high formaldehyde background
levels in commercially prepacked cartridges. It is advised that three cartridges
randomly selected from each production lot should be analyzed for formaldehyde prior
to use to determine acceptable levels. Thermosorb/F cartridges are 1.5 cm internal
diameter x 2 cm long polyethylene tubes with Luer®-type fittings on each end. The
adsorbent is composed of 60/80-mesh Florisil (magnesium silicate) coated with DNPH.
The adsorbent is held in place with 100 mesh stainless steel screens at each end.
The precoated cartridges are used as received and are discarded after use. The
cartridges are stored in glass culture tubes with polypropylene caps and placed in
cold storage when not in use.
Method TO-11 may involve hazardous materials, operations, and equipment.
The method does not purport to address all the safety problems associated with its
use. It is the responsibility of the user to consult and establish appropriate safety and
health practices and determine the applicability of regulatory limitations prior to use.
5.2 SUMMARY OF METHOD
A known volume of ambient air is drawn through a prepacked silica gel
cartridge coated with acidified DNPH at a sampling rate of 500-1200 mL/min for an
appropriate period of time. Sampling rate and time are dependent upon carbonyl
concentration in the test atmosphere.
5-3
-------
Revision 0
OCTOBER 1991
After sampling, the sample cartridges are capped and placed in borosilicate
glass culture tubes with polypropylene caps. The capped tubes are then placed in a
friction-top can containing a pouch of charcoal and returned to the laboratory for
analysis. Alternatively, the sample vials can be placed in a styrofoam box with
appropriate padding for shipment to the laboratory. The cartridges may either be
placed in cold storage until analysis or immediately washed by gravity-feed elution of
6 ml_ of acetonitrile from a plastic syringe reservoir to a graduated test tube or a 5 ml_
volumetric flask. The eluate is then topped to a known volume and refrigerated until
analysis.
The DNPH-carbpnyl derivative is determined using isocratic reverse phase
HPLC with an ultraviolet (UV) absorption detector operated at 360 nm. A cartridge
blank is likewise desorbed and analyzed as per Section 5.9. Formaldehyde,
acetaldehyde, and acetone in the sample are identified and quantified by comparison
of their retention times and peak heights or peak areas with those of standard
solutions.
5.3 SIGNIFICANCE
Formaldehyde, acetaldehyde, and acetone emissions result from incomplete
combustion of hydrocarbons and other organic materials. The major emission
sources appear to be vehicle exhaust, waste incineration, and fuel burning (natural
gas, fuel oil, and coal). In addition, significant amounts of atmospheric carbonyl
compounds can result from photochemical reactions between reactive hydrocarbons
and NOX. Moreover, these carbonyl compounds can react photochemically to
produce other products, including O3, peroxides, and PAN. Local sources of
formaldehyde, acetaldehyde, and acetone may include manufacturing and other
industrial processes that use the chemicals. In particular, formaldehyde emissions are
associated with any industrial process that results in the pyrolysis of organic
compounds in air or oxygen. This test method provides a means to determine
concentrations of formaldehyde, acetaldehyde, and acetone in emissions sources in
various working environments and in ambient indoor and outdoor atmospheres.
5-4
-------
Revision 0
OCTOBER 1991
5.4 INTERFERENCES
This procedure has been prepared specifically to address the sampling and
analysis of formaldehyde, acetaldehyde, and acetone. Interferences in the method
arise from unresolved components in the HPLC chromatogram. Organic compounds
that have the same retention time and significant absorbance at 360 nm as the DNPH
derivatives of carbonyl compounds will interfere. Such interferences can often be
overcome by altering the separation conditions (e.g., using alternative HPLC columns
or mobile phase compositions). Other carbonyl compounds can be detected with a
modification of the basic procedure. In particular, chromatographic conditions can be
optimized to separate acrolein, acetone, and propionaldehyde (within an analysis time
of about one hour) by utilizing two Zorbax ODS columns in series under a linear
gradient program. The following gradient program was found to be adequate to
achieve a resolution of other potentially interfering carbonyl compounds. Upon sample
injection, linear gradient from 60-75% acetonitrile/40-25% water in 30 minutes, gradient
from 75-100% acetonitrile/25-0% water in 20 minutes, hold at 100% acetonitrile for 5
minutes. A reverse gradient is then necessary to reequilibrate the HPLC column. This
gradient changes solvent from 100% acetonitrile to 60% acetonitrile/40% water in 1
minute, and maintains isocratic at 60% acetonitrile/40% water for 15 minutes.
Formaldehyde or acetone contamination of the DNPH reagent is a frequently
encountered problem. The DNPH must be purified by multiple recrystallizations in UV
grade acetonitrile. Recrystallization is accomplished at 40-60°C by slow evaporation of
the solvent to maximize crystal size. The purified DNPH crystals are stored under UV
grade acetonitrile until use. Impurity levels of carbonyl compounds in the DNPH are
determined by HPLC prior to use and should be less than 0.025 micrograms (/L/g)/mL
5.5 EQUIPMENT
Isocratic HPLC system - consisting of a mobile phase reservoir; a high pressure
pump; an injection valve (automatic sampler with an optional 25-microliter (^/L) loop
injector); a Zorbax ODS (DuPont Instruments, Wilmington, DE) or equivalent C-18,
reverse phase column or equivalent (25 cm x 4.6 mm internal diameter); a variable
5-5
-------
Revision 0
OCTOBER 1991
wavelength UV detector operating at 360 nm; and a data system or strip chart
recorder (See Figure 5-2).
Sampling system - capable of accurately and precisely sampling
100-1500 mL/min of ambient air (See Figure 5-3). The dry test meter may not be
accurate at flows below 500 mL/min, and should then be replaced by recorded flow
readings at the start, finish, and hourly during the collection. The sample pump
consists of a diaphragm or metal bellows pump capable of extracting an air sample
between 500-1200 mL/min. A normal pressure drop through the sample cartridge is
approximately 14 cm Hg at a sampling rate of 1.5 L/min.
Stopwatch - for time measurement.
Friction-top metal can or a stvrofoam box with air bubble padding - to contain
the sample vials.
Thermometer - to record ambient temperature.
Barometer (optional) to measure barometric pressure.
Suction filtration apparatus - for filtering HPLC mobile phase.
Volumetric flasks - various sizes, 5-2000 mL.
Pipets - various sizes, 1-50 mL
Helium purge line (optional) - for degassing HPLC mobile phase.
Erlenmeyer flask. 1 L - for preparing HPLC mobile phase.
Graduated cylinder. 1 L - for preparing HPLC mobile phase.
Svrinae. 100-250 pL - for HPLC injection.
Sample vials - to contain the samples.
Melting point apparatus - to determine melting points.
Rotameters - for flow measurement.
Calibrated syringes - for precision injections and extractions.
Special glass apparatus - for rinsing, storing and dispensing saturated DNPH
stock reagent (See Figure 5-4).
5-6
-------
en
Injection
Valve
Column
Mobile
Phase
Reservoir
Variable
Wavelength
UV
Detector
5015918R
O
CD ID
m
CO
Figure 5-2. Typical HPLC System.
-------
Mass Flow
Controllers
Revision 0
OCTOBER 1991
Couplings to connect
DNPH-coated Sep-PAK
Adsorbent Cartndges
Oil-less Pump
Vent
(a) Mass Flow Control
Rotameter
..
Dry
Test
Meter
-
=
Pump
^ ,. — .
Needle r
Valve ,~
Vent
(Dry test meter should not be used
for flow of less than 500 ml/minute)
Couplings to connect
DNPH-coated Sep-PAK
Adsorbent Cartridges
(b) Neddie Valve/Dry Test Meter
Figure 5-3. Typical sampling system configurations.
a
§
5-8
-------
DNPH Crystals
High-Porosity Frit
Acetonitrile
Revision 0
OCTOBER 1991
DNPH-Coated SiO,
/
V
ml
mi.
— 30 —
— 25 —
— 20 —
— 15 —
— 10 —
~~ 5 "~~
\
Three-way Stopcock
Figure 5-4. Special glass apparatus for rinsing, storing, and
dispensing saturated DNPH stock solution.
5-9
-------
Revision 0
OCTOBER 1991
Mass flow meters and mass flow controllers - for metering/setting air flow rate
through sample cartridge of 500-1200 mL/min. The mass flow controllers are
necessary because cartridges have a high pressure drop and at maximum flow rate,
the cartridge behaves like a "critical orifice." Recent studies have shown that critical
flow orifices may be used for 24-hour sampling periods at a maximum rate of 1 L/min
for atmospheres not heavily loaded with particulates.
Positive displacement, repetitive dispensing pipets - (Lab-Industries, or
equivalent), 0-10 ml_ range.
Cartridge drying manifold - with multiple standard male Luer® connectors.
Liquid syringes. 10 mL - (polypropylene syringes are adequate) for preparing
DNPH-coated cartridges.
Syringe rack - to enable batch processing of 45 cartridges for cleaning, coating,
and/or sample elution.
Luer® fittings/plugs - to connect cartridges to sampling system and to cap
prepared cartridges.
Hot plates, beakers, flasks, measuring and disposable pipets. volumetric flasks.
etc. - used in the purification of DNPH.
Borosilicate glass culture tubes (20 x 125 mrm" with polypropylene screw caps -
used to transport Sep-PAK® coated cartridges for field applications.
Heated probe - necessary when ambient temperature to be sampled is below
60° Fahrenheit (F) to ensure the effective collection of formaldehyde as a hydrazone.
Cartridge sampler - silica gel cartridge, Sep-PAK® coated in-situ with DNPH
according to Section 5.7.
Polyethylene gloves - used to handle Sep-PAK® silica gel cartridges.
5.6 REAGENTS AND MATERIALS
2.4-Dinitrophenvlhvdrazine - reagent grade or equivalent. Recrystallize at least
twice with UV-grade acetonitrile before use.
Acetonitrile - UV-grade, "distilled-in-glass," or equivalent.
Deionized-distilled water - charcoal filtered.
5-10
-------
Revision 0
OCTOBER 1991
Perchloric acid - analytical grade, best source.
Hydrochloric acid - analytical grade, best source.
Formaldehyde - analytical grade, best source.
Aldehydes and ketones. analytical grade, best source - for preparation of DNPH
derivative standards.
Ethanol or methanol - analytical grade, best source.
Seo-PAK® silica gel cartridge or equivalent - for sample collection.
Nitrogen - high purity grade, best source.
Charcoal - granular, best source.
Helium - high purity grade, best source.
5.7 PREPARATION OF REAGENTS AND CARTRIDGES
This section describes procedures used to prepare reagents and cartridges.
5.7.1 Purification of DNPH
This procedure should be performed under a properly ventilated hood.
1. Prepare a supersaturated solution of DNPH by boiling excess DNPH in
200 ml of acetonitrile for approximately one hour.
2. After one hour, remove and transfer the supernatant to a covered beaker
on a hot plate and allow gradual cooling to 40-60°C.
3. Maintain the solution at 40-60°C until 95% of solvent has evaporated.
4. Decant solution to waste and rinse crystals twice with three times their
apparent volume of acetonitrile. Various health effects result from inhalation of
acetonitrile. At 500 ppm in air, brief inhalation has produced nose and throat
irritation. At 160 ppm, inhalation for 4 hours has caused flushing of the face
(2-hour delay after exposure) and bronchial tightness (5 hour delay). Heavier
exposures have produced systemic effects, with symptoms ranging from
headache, nausea, and lassitude to vomiting, chest or abdominal pain,
respiratory depression, extreme weakness, stupor, convulsions and death
(dependent upon concentration and time).
5-11
-------
Revision 0
OCTOBER 1991
5. Transfer crystals to another clean beaker, add 200 ml_ of acetonitrile,
heat to boiling, and again let crystals grow slowly at 40-60°C until 95% of the solvent
has evaporated.
6. Repeat rinsing process as described in Step 4.
7. Take an aliquot of the second rinse, dilute 10 times with acetonitrile,
acidify with 1 mL of 3.8 mole perchloric acid per 100 ml_ of DNPH solution, and
analyze by HPLC.
8. The chromatogram illustrated in Figure 5-5 represents an acceptable
impurity level of <0.025 A/g/mL of formaldehyde in recrystallized DNPH reagent. An
acceptable impurity level for an intended sampling application may be defined as the
mass of the analyte (e.g., DNPH-formaldehyde derivative) in a unit volume of the
reagent solution equivalent to less than one-tenth the mass of the corresponding
analyte from a volume of an air sample when the carbonyl (e.g., formaldehyde) is
collected as DNPH derivative in an equal unit volume of the reagent solution. An
impurity level unacceptable for a typical 10 L sample volume may be acceptable if
sample volume is increased to 100 L. The impurity level of DNPH should be below the
sensitivity (ppb, v/v) level indicated in Table 5-1 for the anticipated sample volume. If
the impurity level is not acceptable for intended sampling application, repeat
recrystallization. A special glass apparatus should be used for the final rinse and '
storage according to the following procedure:
A. Transfer the crystals to the special glass apparatus (See Figure 5-4).
B. Add about 25 mL of acetonitrile, agitate gently, and let solution
equilibrate for 10 minutes.
C. Drain the solution by properly positioning the three-way stopcock. The
purified crystals should not be allowed to contact laboratory air except for a brief
moment. Minimal contact is accomplished by placing a DNPH-coated silica cartridge
on the gas inlet of the special glass apparatus.
5-12
-------
DNPH Reagent
cn
I
CO
Sotvent Front
5015921R
O
O
CD 33
m 9
CD
10
20
30
40
50
Figure 5-5. Impurity level of DNPH after recrystallization.
-------
Table 5-1
Sensitivity (ppb,V/V) of Sampling/Analysis for Aldehydes and
Ketones in Ambient Air Using Adsorbent Cartridge Followed
by Gradient High Performance Liquid Chromatography
en
Sample Volume, L
Compound
Formaldehyde
Acetaldehyde
Acrolein
Acetone
Propionaldehyde
Crotonaldehyde
Butyraldehyde
Benzaldehyde
Isovaleraldehyde
Valeraldehyde
o-Tolualdehyde
m-Tolualdehyde
p-Tolualdehyde
Hexanaldehyde
2,5-Dimethylbenzaldehyde
10
20
Sensitivity
.45
.36
.29
.28
.28
.22
1.21
1.07
1.15
1.15
1.02
1.02
1.02
1.09
0.97
0.73
0.68
0.65
0.64
0.64
0.61
0.61
0.53
0.57
0.57
0.51
0.51
0.51
0.55
0.49
30
(ppb,
0.48
0.45
0.43
0.43
0.43
0.41
0.40
0.36
0.38
0.38
0.34
0.34
0.34
0.36
0.32
40
v/v) of
0.36
0.34
0.32
0.32
0.32
0.31
0.30
0.27
0.29
0.29
0.25
0.25
0.25
0.27
0.24
50
60
100-
200
DNPH/HPLC Method
0.29
0.27
0.26
0.26
0.26
0.24
0.24
0.21
0.23
0.23
0.20
0.20
0.20
0.22
0.19
0.24
0.23
0.22
0.21
0.21
0.20
0.20
0.18
0.19
0.19
0.17
0.17
0.17
0.18
0.16
0.15
0.14
0.13
0.13
0.13
0.12
0.12
0.11
0.11
0.11
0.10
0.10
0.10
0.11
0.10
0.07
0.07
0.06
0.06
0.06
0.06
0.06
0.05
0.06
0.06
0.05
0.05
0.05
0.05
0.05
300
Carbonyls
0.05
0.05
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.04
0.03
400
500 1000
in Ambient Air
0.04
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.02
0.03 0.01
0.03 0.01
0.03 0.01
0.03 0.01
0.03 0.01
0.02 0.01
0.02 0.01
0;02 0.01
0.02 0.01
0.02 0.01
0.02 0.01
0.02 0.01
0.02 0.01
0.02 0.01
0.02 0.01
[Note: Ppb values are measured at 1 atm and 25°C; sample cartridge is eluted with 1.5 mL acetonitrile, and 25 mL are
injected onto HPLC column.]
[Note: Maximum sampling flow through a DNPH-coated SEP-PAK® is about 1.5 L per minute.]
O
03 33
m 9
= §•
CD
-«• O
-------
Revision 0
OCTOBER 1991
D. After draining, turn stopcock so drain tube is connected to measuring
reservoir.
E. Introduce acetonitrile through measuring reservoir.
F. Rinsing should be repeated with 20 ml_ portions of acetonitrile until a
satisfactorily low impurity level in the supernatant is confirmed by HPLC analysis. An
impurity level of < 0.025 ^g/ml_ formaldehyde should be achieved, as illustrated in
Figure 5-5.
G. If a special glass apparatus is not available, transfer the purified crystals
to an all-glass reagent bottle, add 200 ml_ of acetonitrile, stopper, shake gently, and let
stand overnight. Analyze supernatant by HPLC according to Section 5.9. The
impurity level should be comparable to that shown in Figure 5-5.
H. If the impurity level is not satisfactory, pipet off the solution to waste, then
add 25 ml_ of acetonitrile to the purified crystals and repeat Step F.
I. If the impurity level is satisfactory, add another 25 ml_ of acetonitrile,
stopper and shake the reagent bottle, then set aside. The saturated solution above
the purified crystals is the stock DNPH reagent.
J. After purification, purity of the DNPH reagent can be maintained by
storing in the special glass apparatus.
K. Maintain only a minimum volume of saturated solution for day-to-day
operation. To minimize waste of purified reagent should it ever become necessary to
re-rinse the crystals to decrease the level of impurity for applications requiring more
stringent purity specifications.
L Use clean pipets when removing saturated DNPH stock solution for any
analytical applications. Do not pour the stock solution from the reagent bottle.
5.7.2 Preparation of DNPH-Formaldehyde Derivative
1. Titrate a saturated solution of DNPH in 2N hydrochloric acid (HCI) with
formaldehyde (other aldehydes or ketones may be used if their detection is desirable).
2. Filter the colored precipitate, wash with 2N HCI and water and let
precipitate air dry.
5-15
-------
Revision 0
OCTOBER 1991
3. Check the purity of the DNPH-formaldehyde derivative by melting point
determination table or HPLC analysis. If the impurity level is not acceptable,
recrystallize the derivative in ethanol. Repeat purity check and recrystallization as
necessary until acceptable level of purity (e.g., 99%) is achieved.
5.7.3 Preparation of DNPH-Formaldehvde Standards
1. Prepare a standard stock solution of the DNPH-formaldehyde derivative
by dissolving accurately weighed amounts in acetonitrile.
2. Prepare a working calibration standard mix from the standard stock
solution. The concentration of the DNPH-formaldehyde compound in the standard mix
solutions should be adjusted to reflect relative distribution in a real sample. Individual
stock solutions of approximately 100 fjg/L are prepared by dissolving 10 fig of the
solid derivative in 100 mL of acetonitrile. The individual solution is used to prepare
calibration standards containing the derivative of interest at concentrations of
0.5-20 fjg/L, which spans the concentration of interest for most ambient air work.
3. Store all standard solutions in a refrigerator. They should be stable for
several months.
4. Preparation of DNPH-Coated Sep-PAK® cartridges must be performed in
an atmosphere with a very low aldehyde background. All glassware and plastic ware
must be scrupulously cleaned and rinsed with deionized water and aldehyde-free
acetonitrile. Contact of reagents with laboratory air must be minimized. Polyethylene
gloves must be worn when handling the cartridges.
5.7.3.1 DNPH Coating Solution -
1. Pipet 30 mL of saturated DNPH stock solution to a 1000 mL volumetric
flask, then add 500 mL acetonitrile.
2. Acidify with 1.0 mL of concentrated HCI. The atmosphere above the
acidified solution should preferably be filtered through a DNPH-coated silica gel
cartridge to minimize contamination from laboratory air. Shake solution, then make up
to volume with acetonitrile. Stopper the flask, invert, and shake several times until the
5-16
-------
Revision 0
OCTOBER 1991
solution is homogeneous. Transfer the acidified solution to a reagent bottle with a
0-10 ml_ range positive displacement dispenser.
3. Prime the dispenser and slowly dispense 10-20 mL to waste.
4. Dispense an aliquot of the solution to a sample vial and check the
impurity level of the acidified solution by HPLC according to Section 5.7.1, as
illustrated in Figure 5-5.
5. The impurity level should be < 0.025 /ug/mL formaldehyde, similar to that
in the DNPH coating solution.
5.7.3.2 Coating the Cartridges -
1. Open the-Sep-PAK® package, connect the short end to a 10 mL syringe,
and place it in the syringe rack. Prepare as many cartridges and syringes as possible.
2. Using a positive displacement repetitive pipet, add 10 mL of acetonitrile
to each of the syringes.
3. Let liquid drain to waste by gravity. Remove any air bubbles that may be
trapped between the syringe and the silica cartridge by displacing them with the
acetonitrile in the syringe.
4. Set the repetitive dispenser containing the acidified DNPH coating
solution to dispense 7 mL into the cartridges.
5. Once the effluent flow at the outlet of the cartridge has stopped,
dispense 7 mL of the coating reagent into each of the syringes.
6. Let the coating reagent drain by gravity through the cartridge until flow at
the other end of the cartridge stops.
7. Wipe the excess liquid at the outlet of each of the cartridges with clean
tissue paper.
8. Assemble a drying manifold with a scrubber or "guard cartridge"
connected to each of the exit ports. These "guard cartridges" are DNPH-coated and
serve to remove any trace of formaldehyde in the nitrogen gas supply.
9. Remove the cartridges from the syringes and connect the short ends to
the exit end of the scrubber cartridge.
5-17
-------
Revision 0
OCTOBER 1991
10. Pass nitrogen through each of the cartridges at about 300-400 mL/min
for 5-10 minutes.
11. Within 10 minutes of the drying process, rinse the exterior surfaces and
outlet ends of the cartridges with acetonitrile using a Pasteur pipet.
12. Stop the flow of nitrogen after 15 minutes and insert cartridge connectors
into the long end of the scrubber cartridges.
13. Connect the short ends of a batch of the coated cartridges to the
scrubbers and pass nitrogen through the cartridges at about 300-400 mL/min.
14. Repeat Step 11.
15. After 15 minutes, stop the flow of nitrogen, remove the dried cartridges
and wipe the cartridge exterior free of rinse acetonitrile.
16. Plug both ends of the coated cartridge with standard polypropylene
Luer® male plugs, and place the plugged cartridge in a borosilicate glass culture tube
with polypropylene screw caps.
17. Put a serial number and lot number label on each of the individual
cartridge glass storage containers and store the prepared lot in the refrigerator.
18. Store cartridges in an all-glass stoppered reagent bottle in a refrigerator
until use. Plugged cartridges could also be placed in screw-capped glass culture
tubes and placed in a refrigerator. Cartridges will maintain their integrity for up to.
90 days stored in refrigerated, capped culture tubes.
19. Before transport, remove the glass-stoppered reagent bottles (or screw-
capped glass culture tubes) containing the adsorbent tubes from the refrigerator and
place the tubes individually in labeled glass culture tubes. Place culture tubes in a
friction-top metal can containing 1-2 inches of charcoal for shipment to sampling
location.
20. As an alternative to friction-top cans for transporting sample cartridges,
the coated cartridges could be shipped in their individual glass containers. A large
batch of coated cartridges in individual glass containers may be packed in a styrofoam
box for shipment to the field. The box should be padded with clean tissue paper or
5-18
-------
Revision 0
OCTOBER 1991
polyethylene-air bubble padding. Do not use polyurethane foam or newspaper as
padding material.
21. The cartridges should immediately be stored in a refrigerator or in a
chilled cooler upon arrival in the field.
5.8 SAMPLING
The sampling system is assembled and should be similar to that shown in
Figure 5-3 above. Figure 5-3 illustrates a three-tube/one-pump configuration. The
tester should ensure that the pump is capable of a constant flow rate throughout the
sampling period. The coated cartridges can be used as direct probes and traps for
sampling ambient air when the temperature is above freezing. For sampling ambient
air below freezing, a short length (30-60 cm) of heated (50-60°C) stainless steel tubing
must be added to condition the air sample prior to collection on adsorbent tubes.
Two types of sampling systems are shown in Figure 5-3. For purposes of discussion,
the following procedure assumes the use of a dry test meter. The dry test meter may
not be accurate at flows below 500 mL/min and should be backed up by recorded
flow readings at the start, finish, and hourly intervals during sample collection.
Before sample collection, the system is checked for leaks. Plug the input end
of the cartridge so no flow is indicated at the output end of the pump. The mass flow
meter should not indicate any air flow through the sampling apparatus.
The entire assembly, including a "dummy" sampling cartridge, is installed and
the flow rate checked at a value near the desired rate. In general, flow rates of
500-1200 mL/min should be employed. The total moles of carbonyl in the volume of
air sampled should not exceed that of the DNPH concentration (2 ^g/cartridge). In
general, a safe estimate of the sample size should be 75% of the DNPH loading of the
cartridge. Generally, calibration is accomplished using a soap bubble flow meter or
calibrated wet test meter connected to the flow exit, assuming the system is sealed.
Ideally, a dry gas meter is included in the system to record total flow. If a dry
gas meter is not available, the operator must measure and record the sampling flow
rate at the beginning and end of the sampling period to determine sample volume. If
5-19
-------
Revision 0
OCTOBER 1991
the sampling period exceeds 2 hours, the flow rate should be measured at
intermediate points during the sampling period. Ideally, a rotameter should be
included to allow observation of the flow rate without interruption of the sampling
process.
Before sampling, remove the glass culture tube from the friction-top metal can
or styrofoam box. Let the cartridge warm to ambient temperature in the glass tube
before connecting it to the sample train.
Using polyethylene gloves, remove the coated cartridge from the glass tube
and connect it to the sampling system with a Luer® adapter fitting. Seal the glass tube
for later use, and connect the cartridge to the sampling train so that its short end
becomes the sample inlet. Record the following parameters on the sampling data
sheet (See Figure 5-6):
• Date;
• Sampling location;
• Time;
• Ambient temperature;
« Barometric pressure (if available);
• Relative humidity (if available);
• Dry gas meter reading (if appropriate);
• Flow rate;
• Rotameter setting;
Cartridge batch number; and
5-20
-------
Revision 0
OCTOBER 1991
SAMPLING DATA SHEET
(One Sample per Data Sheet)
PROJECT:
SITE:
DATE(S) SAMPLED:
LOCATION:
TIME PERIOD SAMPLED:
OPERATOR:
INSTRUMENT MODEL NO:
PUMP SERIAL NO:
CALIBRATED BY:
ADSORBENT CARTRIDGE INFORMATION:
Tvoe:
Adsorbent:
SAMPLING DATA:
Start Time:
Time
Avg.
Dry Gas
Meter
Reading
Rotameter
Reading
Row
Rate (O)*,
Ml/min
Serial Number:
Sample
Stop T
Ambient
Temperature,
°C
j Number:
me:
Barometric
Pressure,
mm Hg
Relative
Humidity, %
Comments
Row rate from rotameter or soap bubble calibrator (specify which) Total Volume Data (Vm) (use data from dry gas
meter, if available)
Vm = (Final - Initial) Dry Gas Meter Reading, or
or
1
Uters
Liters
1000 x (Sampling Time in Minutes)
Figure 5-6. Sampling Data Sheet
5-21
-------
Revision 0
OCTOBER 1991
Dry gas meter pump identification numbers.
The sampler is turned on and the flow is adjusted to the desired rate. A typical
flow rate through one cartridge is 1 .0 L/min and 0.8 L/min for two cartridges in
tandem.
The sampler is operated for the desired period, with periodic recording of the
variables listed above.
At the end of the sampling period, the sampling parameters are recorded and
the sample flow is stopped. If a dry gas meter is not used, the flow rate must be
checked at the end of the sampling interval. If the flow rates at the beginning and end
of the sampling period differ by more than 15%, the sample should be marked as
suspect.
Immediately after sampling, remove the cartridge (using polyethylene gloves)
from the sampling systems, cap with Luer® end plugs, and place it back in the original
labeled glass culture tube. Cap the culture tube, seal it with Teflon® tape, and place it
in a friction-top can containing 1 -2 inches of granular charcoal or a styrofoam box with
appropriate padding. Refrigerate the culture tubes until analysis. The refrigeration
period prior to analysis should not exceed 30 days. If samples are to be shipped to a
central laboratory for analysis, the duration of the non-refrigerated period should be
kept to a minimum, preferably less than 2 days.
If a dry gas meter or equivalent total flow indicator is not used, the average
sample flow rate (in mL/min) must be calculated according to the following equation:
Q2 + ..... Qn
N
Where:
Qv Q2 — QN = flow rates determined at beginning, end, and
intermediate points during sampling.
N = number of points averaged.
5-22
-------
Revision 0
OCTOBER 1991
The total flow rate is then calculated using the following equation:
Where:
V..
T2-Tl
J x Q
1000
total volume (L) sampled at measured temperature
and pressure.
stop time (minutes).
start time (minutes).
total sampling time (minutes).
average flow rate (mL/min).
The total volume (in L) at standard conditions, 25°C and 760 mm Hg, is
calculated from the following equation:
Where:
V.
V,
m
Vm x PA x 298
+
760 273
total sample volume (L) at measured temperature
and pressure
average ambient pressure (mm Hg)
average ambient temperature (°C)
5.9 SAMPLE ANALYSIS
This section presents procedures for sample analysis.
5-23
-------
Revision 0
OCTOBER 1991
5.9.1 Sample Desorption
1. Remove the sample cartridge from labeled culture tube. Connect the
sample cartridge (outlet end during sampling) to a clean syringe. The liquid flow
during desorption should be in the reverse direction of air flow during sample
collection.
2. Place the cartridge/syringe in the syringe rack.
3. Backflush the cartridge (gravity feed) by passing 6 ml_ of acetonitrile from
the syringe through the cartridge to a graduated test tube or to a 5 ml_ volumetric
flask. A dry cartridge has an acetonitrile holdup volume slightly greater than 1 ml_.
The eluent flow may stop before the acetonitrile in the syringe is completely drained
into the cartridge because of the air trapped between the cartridge filter and the
syringe Luer® tip. If this happens, displace the trapped air with the acetonitrile in the
syringe using a long-tip disposable Pasteur pipet.
4. Dilute to the 5 mL mark with acetonitrile. Label the flask with sample
identification. Pipet two aliquots into sample vials with Teflon®-lined septa. Analyze
the first aliquot for the derivatized carbonyls by HPLC. Store the second aliquot in the
refrigerator until sample analysis.
5.9.2 HPLC Analysis
The HPLC system is assembled and calibrated as described in Section 5.9.3.
The operating parameters are as follows:
Column: Zorbax ODS (4.6 mm I.D. x 25 cm), or equivalent.
Mobile Phase: 60% acetonitrile/40% water, isocratic.
Detector: ultraviolet, operating at 360 nm.
Flow Rate: 1.0 mL/min.
Retention Time: 7 minutes for formaldehyde with one Zorbax ODS column.
13 minutes for formaldehyde with two Zorbax ODS
columns.
5-24
-------
Revision 0
OCTOBER 1991
Injection Volume: 25
1. Before each analysis, the detector baseline is checked to ensure stable
conditions.
2. The HPLC mobile phase is prepared by mixing 600 mL of acetonitrile and
400 mL of water. This mixture is filtered through a 0.22 pm polyester membrane filter
in an all-glass and Teflon® suction filtration apparatus. The filtered mobile phase is
degassed by purging with helium for 10-15 minutes (100 mL/min) or by heating to
60°C for 5-10 minutes in an Erlenmeyer flask covered with a watch glass. A constant
back-pressure restrictor or a short length of 0.25 mm inside diameter Teflon® tubing
should be placed after the detector to eliminate further mobile phase outgassing.
3. The mobile phase is placed in the HPLC solvent reservoir and the pump
is set at a flow rate of 1.0 mL/min and allowed to pump for 20-30 minutes before the
first analysis. The detector is switched on at least 30 minutes before the first analysis,
and the detector output is displayed on a strip chart recorder or similar output device.
4. A 100-/L/L aliquot of the sample is drawn into a clean HPLC injection
syringe. The sample injection loop is loaded and an injection is made. The data
system, if available, is activated simultaneously with the injection, and the point of
injection is marked on the strip chart recorder.
5. After approximately 1 minute, the injection valve is returned to the "inject"
position and the syringe and valve are rinsed or flushed with acetonitrile/water mixture
in preparation for the next sample analysis. The flush/rinse solvent should not pass
through the sample loop during flushing. The loop is cleaned while the valve is in the
"inject" mode.
6. After elution of the DNPH derivatives (See Figure 5-7), data acquisition is
terminated and the component concentrations are calculated as described in
Section 5.10.
5-25
-------
en
I
ro
en
Operating Parameters HPLC
Column:
Mobile Phase:
Detector:
Flow Rate:
Retention Time:
Sample Injection: 25 /A
60% Acetonitrile/40% Water
Zorbax OD3 or 18 RP
Ultraviolet, operating at 360 nm
1 mU/min
7 minutes for formaldehyde
10
20
Time, tnln
Figure 5-7. Chromatogram of DNPH-formaldehyde derivative.
o
o
ro
m
co
O
-------
Revision 0
OCTOBER 1991
7. After a stable baseline is achieved, the system can be used for further
sample analyses as described above. After several cartridge analyses, buildup on the
column may be removed by flushing with several column volumes of 100% acetonitrile.
8. If the concentration of analyte exceeds the linear range of the instrument,
the sample should be diluted with mobile phase, or a smaller volume can be injected
into the HPLC.
9. If the retention time is not duplicated (±10%), as determined by the
calibration curve, the acetonitrile/water ratio may be increased or decreased to obtain
the correct elution time. If the elution time is too long, increase the ratio; if it is too
short, decrease the ratio. The chromatographic conditions described here have been
optimized for the detection of formaldehyde. Analysts are advised to experiment with
their HPLC system to optimize chromatographic conditions for their particular analytical
needs.
5.9.3 HPLC Calibration
1. Calibration standards are prepared in acetonitrile from the carbonyl
derivatives. Individual stock solutions of 100 /jg/L are prepared by dissolving 10 mg
of solid derivative in 100 ml_ of mobile phase. These individual solutions are used to
prepare calibration standards at concentrations spanning the range of interest.
2. Each calibration standard (at least five levels) is analyzed three times and
area response is tabulated against mass injected (See Figure 5-8). All calibration runs
are performed as described for sample analyses in Section 5.9.2. Using the UV
detector, a linear response range of approximately 0.05-20 /jg/L should be achieved
for 25 /L/L injection volumes. The results may be used to prepare a calibration curve,
as illustrated in Figure 5-9. Linear response is indicated where a correlation coefficient
of at least 0.999 for a linear least-squares fit of the data (concentration versus are
response) is obtained. The retention times for each analyte should agree with
2 percent.
3. Once linear response has been documented, an intermediate
concentration standard near the anticipated levels of each component, but at least 10
5-27
-------
Operating Parameters HPLC
Column:
Mobile Phase:
Detector:
Flow Rate:
Retention Time:
Sample Injection: 25 pL
Zorbax ODS or 18 RP
60% Acetonrtrile/40% Water
Ultraviolet, operating at 360 nm
1 mlVmin
~7 minutes for formaldehyde
(a)
Time —*•
61 pg/mL
(b)
Time —*•
1.23 pg/mL
Revision 0
OCTOBER 1991
CD
"c
I
(c)
Time —•-
6.16pg/mL
Cone
.61 pg/mL
1.23pg/mL
6.16f/g/mL
12.32f/g/mL
Area Counts
226541
452166
2257271
4711408
6953812
(d)
(e)
" Time ——
13 12.32pg/mL
' Time —*•
CD
C
Figure 5-8. HPLC chromatogram of varying concentrations of
DNPH-formaldehyde derivative.
5-28
-------
800 H
Revision 0
OCTOBER 1991
700 -
600 -
500 -
8
5.
Correlation Coefficient: 0.9999
400 -
I
300 -
200 -
Operating Parameters HPLC
100 -
Column:
Mobile Phase:
Detector:
Flow Rate:
Retention Time:
Sample Injection: 25 pL
ZorbaxODSor 18 RP
60% Acetonitrile/40% Water
Ultraviolet, operating at 360 nm
1 mUrnin
~ 7 minutes for formaldehyde
i
6
I
12
I
15
T
18
DNPH-Formaldehyde Derivative (/jg/mL)
Figure 5-9. Calibration curve for Formaldehyde.
5-29
-------
Revision 0
OCTOBER 1991
times the detection limit, should be chosen for daily calibration. The day-to-day
response for the various components should be within 10% for analyte concentrations
1 ^g/ml_ or greater and within 15-20% for analyte concentrations near 0.5 ^/g/mL If
greater variability is observed, recalibration may be required or a new calibration curve
may be developed from fresh standards.
4. The response for each component in the daily calibration standard is
used to calculate the response factor according to the following equation:
RF.
Cc x V,
Where:
V,
R-
response factor (usually area counts) for the
component of interest in ng injected/response unit.
concentration (mg/L) of analyte in the daily
calibration standard.
volume (/jL) of calibration standard injected.
response (area counts) for analyte in the calibration
standard.
5.10 CALCULATIONS
The total mass of analyte is calculated for each sample using the following
equation:
Where:
Wd
RF
RFC x Rd x VE / V,
total quantity of analyte (pg) in the sample.
calculated response factor.
response (area counts or other response units) for
analyte in sample extract, blank corrected.
5-30
-------
Where:
Vc
V,
v
Revision 0
OCTOBER 1991
(Vb / Vs)]
area counts, sample.
area counts, blank.
volume (ml), blank.
volume (ml_), sample.
final volume (ml_) of sample extract.
volume of extract (/A) injected into the HLPC system.
redilution volume (if the sample was rediluted).
aliquot used for redilution (if sample was rediluted).
The concentrations of carbonyl compounds in the original sample is calculated
from the following equation:
Where:
V,
m
X1000
(or Vs)
concentration of analyte (ng/L) in the original
sample.
total quantity of analyte (ng) in sample, blank
corrected.
total sample volume (L) under ambient conditions.
total sample volume (L) at 25°C and 760 mm Hg.
5-31
-------
Revision 0
OCTOBER 1991
The analyte concentrations can be converted to ppbv using the following equation:
CA(ppbv) = CA(ng/L) x 2AA.
MWA
Where:
CA (ppbv) = Concentration of analyte ppbv.
CA (ng/L) = is calculated using Vs.
MWA = molecular weight of analyte.
5.11 PERFORMANCE CRITERIA AND QUALITY ASSURANCE
This section addresses performance criteria and QA issues.
5.11.1 Standard Operating Procedures
Users should generate SOPs describing the following activities in their
laboratory:
Assembly, calibration, and operation of the sampling system, with make
and model of equipment used;
Preparation, purification, storage, and handling of sampling reagent and
samples;
Assembly, calibration, and operation of the HPLC system, with make and
model of equipment used; and
All aspects of data recording and processing, including lists of computer
hardware and software used.
Standard operating procedures should provide specific stepwise instructions and
should be readily available to and understood by the laboratory personnel conducting
the work.
5-32
-------
Revision 0
OCTOBER 1991
5.11.2 HPLC System Performance
The general appearance of the HPLC system should be similar to that illustrated
in Figure 5-2 above. HPLC system efficiency is calculated using the following
equation:
N = 5.54 (tr / W1/2)2
Where:
N = column efficiency (theoretical plates).
t = retention time (seconds) of analyte.
W1/2 = width of component peak at half height (seconds).
A column efficiency of > 5,000 theoretical plates should be obtained.
Precision of response for replicate HPLC injections should be ± 10% or less,
day to day, for analyte calibration standards at 1 ^/g/mL or greater levels. At
0.5 pg/mL level and below, precision of replicate analyses could vary up to
25 percent. Precision of retention times should be ± 2% on a given day.
5.11.3 Process Blanks
At least one field blank or blanks to equal 10% of the field samples, whichever
is larger, should be shipped and analyzed with each group of samples. The number
of samples within a group and/or time frame should be recorded so that a specified
percentage of blanks is obtained for a given number of field samples. The field blank
is treated identically to the samples except that no air is drawn through the cartridge.
5-33
-------
Revision 0
OCTOBER 1991
5.11.4 Method Precision and Accuracy
At least one duplicate sample or duplicates to equal 10% of the field samples,
whichever is larger, should be collected during each sampling episode. Precision for
field replication should be ± 20% or better. Precision for replicate HPLC injections
should be ± 10% or better, day-to-day, for calibration standards. At least one sample
spike with analytes of interest or spiked samples to equal 10% of the field samples,
whichever is larger, should be collected. Before initial use of the method, each
laboratory should prepare and analyze triplicate spiked samples at a minimum of three
concentration levels, bracketing the range of interest for each compound. Triplicate
nonspiked samples must also be analyzed. Spike recoveries of >80 ± 10% and blank
levels as outlined in Section 5.7.1 should be achieved.
5.12 DETECTION OF OTHER CARBONYL COMPOUNDS
Ambient air contains other carbonyl compounds. Optimizing chromatographic
conditions by using two Zorbax ODS columns in series and varying the mobile phase
composition through a gradient program will enable the analysis of other aldehydes
and ketones.
5.12.1 Sampling Procedures
Same as Section 5.8.
5.12.2 HPLC Analysis
The HPLC system is assembled and calibrated as described in Section 5.12.
The operating parameters are as follows:
Column: Zorbax ODS, two columns in series
Mobile Phase: Acetonitrile/water, linear gradient
(A) 60-75% acetonitrile/40-25% water in 30 minutes.
(B) 75-100% acetonitrile/25-0% water in 20 minutes.
(C) 100% acetonitrile for 5 minutes.
(D) 60% acetonitrile/40% water reverse gradient in 1 min.
5-34
-------
Revision 0
OCTOBER 1991
(E) 60% acetonitrile/40% water, isocratic, for 15 minutes.
Detector: Ultraviolet, operating at 3607 nm
Flow Rate: 1.0 mL/min
Sample
Injection Volume: 25 /L/L
The gradient program allows for optimization of chromatographic conditions to
separate acrolein, acetone, propionaldehyde, and other higher molecular weight
aldehydes and ketones in an analysis time of about 1 hour. Table 5-1 above illustrates
the sensitivity for selected carbonyl compounds that have been identified using two
Zorbax ODS columns in series.
The chromatographic conditions described here have been optimized for a
gradient HPLC system equipped with a UV detector, an automatic sampler with a
25 /L/L/L loop injector and two Zorbax ODS columns (4.6 x 250 mm), a recorder, and
an electronic integrator. Analysts are advised to experiment with their HPLC systems
to optimize chromatographic conditions for their particular analytical needs. Highest
chromatographic resolution and sensitivity are desirable but may not be achieved.
The carbonyl compounds in the sample are identified and quantified by
comparing their retention times and area counts with those of standard DNPH
derivatives. Formaldehyde, acetaldehyde, acetone, propionaldehyde, crotonaldehyde,
benzaldehyde, and o-, m-, p-tolualdehydes can be identified with a high degree of
confidence. The identification of butyraldehyde is less certain because it coelutes with
isobutyraldehyde and methyl ethyl ketone under the stated chromatographic
conditions. Concentrations of individual carbonyl compounds are determined as
outlined in Section 5.10.
Performance criteria and Quality Assurance (QA) activities should meet the
requirements outlined in Section 5.11.
5-35
-------
Revision 0
OCTOBER 1991
5.13 OZONE INTERFERENT
Laboratory tests conducted by EPA have determined that O3 present in ambient
air interferes with the measurement of carbonyl compounds concentrations when
using Method TO-11 procedures. Ozone reacts with carbonyl compounds collected
on the DNPH-coated silica cartridges. Carbonyl compound losses have been
estimated to be as great as 48% on days when the ambient O3 concentration reaches
120 ppbv. To eliminate this interference problem, an O3 scrubber has been
developed. This section presents the ozone scrubber equipment and procedures for
use.
5.13.1 Equipment
Figure 5-10 presents the O3 scrubber schematically. The equipment required to
perform O3 removal includes:
Copper Tubing - A 36 inch length of 1/4-inch O.D- copper tubing is used as the
body of the O3 scrubber. The tubing should be coiled into a spiral approximately
2 inches in O.D.
Potassium Iodide Solution - The inside surface of the copper coil is coated with
a saturated solution of potassium iodide to provide the ozone scrubbing mechanism.
Cord Heater - A 24 inch long cord heater, rated at approximately 80 watts, is
wrapped around the outside of the copper coil to provide heat to prevent
condensation of water or organic compounds from occurring within the coil.
Thermocouple - A Chromel-Alumel (Type K) thermocouple is located between
the surface of the copper coil and the cord heater to provide accurate temperature
measurement and control.
Temperature Controller - A Type K active temperature controller is used to
maintain the O3 scrubber at 66°C as referenced by the Type K thermocouple.
Fittings - Bulkhead unions are attached to the entrance and exit of the copper
coil to allow attachment to other components of the sampling system.
5-36
-------
Top View (cutaway)
\
3' Potassium
Iodide Coated
1/4" O.D. Copper
Tubing
Aldehyde Probe
1/4" Brass
Bulkhead Union
2' Glas-Col®
Heater Cord
• Steel from
Heated Probe
Revision 0
OCTOBER 1991
110V Heater Plug
Reducer
1/4" Brass Bulkhead
~1 1/4"
Side View (cutaway) Front View
Figure 5-10. Crossectional view of the O3 scrubber assembly.
5-37
-------
Revision 0
OCTOBER 1991
Chassis Box - The scrubber assembly is housed in a conveniently sized
aluminum chassis box.
5.13.2 Procedure
Position the O3 scrubber assembly between the sample inlet probe and the
DNPH collection cartridges. Set the temperature controller to maintain the scrubber at
66°C. After placement and temperature control are complete, sampling is conducted
as described in Section 5.8.
5.14 ALTERNATIVE SUBSTRATE
A number of cartridge devices containing solid sorbents coated with DNPH
have been employed for the sampling and analysis of carbonyl compounds. These
techniques have met with varying degrees of success. Solid sorbents that have been
used include XAD-2®, silica gel, glass beads, Fluorisil, glass fiber filters, and
octadecylsilane-bonded silica gel (C18). The use of the C18 substrate coated with
DNPH is discussed below.
5.14.1 Advantages of C18 Substrate
The EPA has determined that O3 reacts with formaldehyde collected on solid
sorbents coated with DNPH. The loss from silica gel DNPH tubes has been estimated
to be as much as 48% when O3 concentrations reach 120 ppbv. Preliminary data'
show that the C18 cartridges may be less prone to O3 interferences than the silica gel.
5.14.2 Disadvantages of C18 Substrate
Several disadvantages of the C18 cartridge have been observed. One of the
major disadvantages is seen in the breakthrough volume for the higher molecular
weight aldehydes. Losses depend on the sampling flow rate, sample volume, and
compound concentration. The C18 cartridges also appear to have a lower capacity
for aldehydes in general. The usefulness of C18 depends on sample volume and
compound concentration. Problems with background levels of acetone and certain
aldehydes have also been noted. Care must be exercised during the purification of
reagents and preparation of the cartridges in order to minimize the presence of these
5-38
-------
Revision 0
OCTOBER 1991
potential interferents. Batches of cartridges that are prepared or purchased
commercially must be subjected to rigorous quality control before being used.
5-39
-------
Revision 0
OCTOBER 1991
SECTION 6.0
METEOROLOGICAL MONITORING
Surface and upper air meteorology play an important role in the formation and
transport of O3. Consequently, meteorology has an impact on population exposure to
03. In order to support monitoring objectives associated with model inputs and
performance evaluations, meteorological monitoring is required at each PAMS.
Surface meteorological measurements should begin within the first year of network
operation. Upper air meteorological data for determining mixing heights should be
collected for approximately 10 to 20 key days per year corresponding to specific
model input requirements. Upper air meteorological measurements should begin as
soon as practical after regulation promulgation or the date nonattainment designation
is made, whichever comes later.
6.1 MEASUREMENT - SURFACE METEOROLOGY
Ground level meteorological variables to be measured as part of enhanced O3
network monitoring are as follows:
• Wind direction;
• Wind speed;
• Ambient temperature;
• Barometric pressure;
• Relative humidity; and
• Solar radiation.
These variables are to be monitored at a height of 10 m above ground level.
6-1
-------
Revision 0
OCTOBER 1991
6.1.1 Equipment
The equipment examples presented in this section are routinely used to
measure the meteorological parameters listed. Any equipment meeting performance
specifications (e.g., accuracy, precision, resolution, etc.) required by EPA may be
used. The EPA meteorological equipment specifications are presented in the "Quality
Assurance Handbook for Air Pollution Measurement Systems: Volume IV -
Meteorological Measurements9".
Wind Direction - Wind direction measurements are performed using an air foil
vane wind direction sensor. This sensor provides azimuth data for use in
micrometeorological characterization.
Wind Speed - Wind speed measurements are performed using a 3-cup
anemometer. This device provides information on horizontal wind velocity.
Ambient Temperature - Representative ambient temperature measurements are
performed using a resistance temperature sensor housed in a motor aspirated
radiation shield. These devices provide continuous ambient air temperature
measurement with negligible radiation errors for temperature and differential
temperature.
Barometric Pressure - Barometric pressure measurements are performed using
an atmospheric pressure sensor. This device provides information on barometric'
pressure of the atmosphere in which it is located.
Relative Humidity - Relative humidity measurements are performed using a
relative humidity sensor. This device provides information on relative humidity of the
atmosphere in which it is located.
Solar Radiation - Solar radiation measurements are performed using a
pyranometer or solarimeter. This device provides information on total sun and sky
radiation.
A meteorological tower is used to locate the wind direction, wind speed,
ambient temperature, barometric pressure, relative humidity, and solar radiation
6-2
-------
Revision 0
OCTOBER 1991
sensors at the required monitoring height of 10 m. The sensors are attached to a
boom located at the top of the tower. There are 3 basic tower designs available:
Rxed Tower - the tower is a continuous structure from ground level to
the 10-meter height. Access to the sensors requires that the tower be
climbed;
Tilt-over Tower - the tower is a continuous structure from ground level to
the 10-meter height, but is hinged at point that allows the tower to be
tilted over. The tilting action allows ground level access to the sensors;
and
Telescopic Tower - the tower is constructed of three sections, each
approximately 4 meters in length. The top section is smallest in diameter
and fits inside the middle section which, in turn, fits inside the base
section. The top of the telescopic tower can be positioned at the
10-meter height. The top may also be positioned at a height of 4 meters,
to provide for ladder access to the sensors.
6.1.2 Procedure
Operating procedures for surface meteorological equipment will vary with
manufacturer and sensor style. Specific procedures of operation should be obtained
from the manufacturer. Generally applicable procedures are contained in the "Quality
Assurance Handbook for Air Pollution Measurements Systems: Volume IV -
Meteorological Measurements9".
6.2 UPPER AIR METEOROLOGY - MIXING HEIGHTS
The mixing height is the maximum depth of the atmosphere from the surface up
to a vertical height below which thorough mixing of pollutants can occur. The degree
of dispersion within this mixed layer is a function of the atmospheric turbulence.
Mechanical mixing, a function of wind flow and surface roughness, and thermal mixing,
a function of surface heating or cooling, are the main factors that produce atmospheric
turbulence.
Numerous studies have been performed to estimate mixing heights. Daytime
mixing heights have been determined by means of instrumented aircraft. Mixing
6-3
-------
Revision 0
OCTOBER 1991
heights were defined by utilizing three parameters: temperature, turbulence, and
aerosol concentration. Strong correlations were demonstrated during the daytime
between the temperature profile and the depth of the aerosol layer. These studies,
along with other related studies, have led to the use of radiosonde data and surface
temperature to estimate the daytime mixing height. The criterion for defining the
nocturnal mixing layer or the top of the surface based stable layer is less straight-
forward. The stable layer height has been shown to correspond to the height of the
low-level nocturnal wind-speed jet. The determinations of the mixing height are
complicated by the presence of clouds, changing air masses, and the influences of
large bodies of water or terrain.
The use of mixing heights information in dispersion model plume rise can be
limited because of the possibility of elevated inversions at the top of mixing layers.
Nearly one-half of a plume that normally would disperse upward would be reflected
toward the ground if a strong inversion occurred just above the plume height. This
situation could lead to maximum pollutant concentrations. Some of the plume will
penetrate the top of the mixing layer depending on the strength of the inversion and
the buoyancy of the plume, but this occurrence is not routinely treated in dispersion
models.
An infinite series term in the gaussian equation accounts for the effects of the
restriction on vertical plume growth at the top of the mixing layer. The method of
image sources is used to account for multiple reflections of the plume from the ground
surface and at the top of the surface mixing layer.
Mixing heights will be required as input to photochemical air quality models.
The models will be applied in determining transport and boundary conditions, and to
evaluate emission reductions in O3 nonattainment areas.
Vertical temperature profiles can be used to estimate mixing heights. These
temperature profiles are obtained from radiosonde balloon measurements,
tethersondes, or aircraft. These temperature profiles are derived from 0 and 1200
Greenwich Median Time (GMT) upper air sounding measurements, obtained from
6-4
-------
Revision 0
OCTOBER 1991
radiosonde releases at National Weather Service (NWS) stations. This balloon sensor
also measures humidity, wind speed and direction, and pressure at incremented
height levels. Tethersondes are balloon sensors tethered to a ground station that
measures atmospheric parameters during their vertical ascent. Measurements with
this system have been recorded at heights of 1500 meters. Tethered balloon
measurements should be taken during moderate to light wind conditions when the
balloon would not present an aviation hazard.
In this case, the mixing height is not measured directly but is calculated
approximately from the vertical temperature profile or the lapse rate. In the mixing
layer, the lapse rate is assumed to be roughly dry adiabatic (unsaturated conditions).
Therefore, the mixing height is estimated in terms of this adiabatic process. The urban
morning mixing height is interpreted as the height above ground at which the dry
adiabatic extension of the morning minimum surface temperature plus 5°C intersects
the 1200 GMT vertical temperature profile. The minimum temperature is determined
between 0200 and 0600 local standard time (LST). Because NWS upper air
measuring stations are located in rural areas, 5°C is added to account for the
nocturnal and early morning "heat island" effects. The afternoon mixing height is
calculated similarly except the maximum temperature determined between 1200 and
1600 LST is used.
Routine estimates of mixing height can be made using alternative National
Oceanographic and Aeronautics Administration (NOAA) methods that employ an
adiabatic temperature modification to determine the convective height and surface
friction velocity, Monin-Obukhov length, and empirical expressions are used to
estimate the nocturnal or mechanical mixed depth.
On-site meteorological measurement programs utilizing these instruments may
help provide much needed data for input to air quality models. Existing NWS upper
air data are available for large-scale atmospheric motions. Much improvement is
needed in terms of the spatial scale of meteorological measurements for special air
pollution studies, particularly for neighborhood or Central Business District (CBD)
6-5
-------
Revision 0
OCTOBER 1991
urban O, studies. Enhancement of current upper air data would provide the means of
o
achieving more accurate wind field and mixing height characterizations. Special
studies carried out by State and local agencies, academia, and private researchers
should include as an integral part of the monitoring plans, recording of upper air wind
and temperature data beyond data already available from the NWS or previous air
pollution studies.
Because balloon techniques require maintenance and operation, long-term
measurement programs using this type of sensor could result in considerable costs.
However, costs could be minimized by renting equipment or using a type of balloon
sensor requiring fewer personnel for maintenance and with lower material costs. The
sensor could be restricted for use only during critical pollution episodes or relevant
months of the year.
6.2.1 Measurements
Mixing heights can be determined from direct measurements or estimated from
vertical temperature profile measurements. The following is a description of methods
used to perform these measurements. The equipment and approaches presented are
only recommendations. Alternative measurement devices and approaches may be
used based on approval of local regulatory agencies.
6.2.1.1 Direct Measurements -
Remote sensing devices such as acoustic sounders can be used to obtain
direct measurements of mixing heights. These devices are commonly known as
Doppler Sound Detection and Ranging (SODAR) systems. These devices are effective
tools for remote measurement of meteorological variables at heights up to several
hundred meters above the surface. The SODAR systems operate on a fundamentally
simple principle; however, the systems that control the operations can be complex.
The SODAR systems transmit a strong (typically 100-300 watts) acoustic pulse into the
atmosphere and listen for the portion of the transmitted pulse that returns. Monostatic
acoustic SODAR systems measure temperature structures by using the same acoustic
driver to both transmit the pulse and receive the return signal. A bistatic system uses
6-6
-------
Revision 0
OCTOBER 1991
different antennas to transmit and receive. Monostatic systems generally have
collocated antennas, while a bistatic configuration requires an antenna separation
distance based on the desired measurement height.
6.2.1.2 Vertical Temperature Profile Measurements -
Upper air sounding measurements can be made to obtain vertical temperature
profiles. Alternatively, existing upper air sounding measurements can be obtained
from NWS reporting stations. These vertical temperature profiles can be used to
estimate mixing heights needed as input to air quality models. The upper air sounding
measurements can be made by several means, including:
Untethered balloons measuring wind speed and direction, pressure,
temperature, and humidity by means of sensors that transmit readings by
radio to a ground station. This method is commonly used by the NWS.
Tethered balloons or wiresondes can measure similar atmospheric data.
The tether carries signals from the sensor attached to the balloon. Many
atmospheric soundings have been recorded using tethered systems,
especially by research laboratories.
6.2.2 Mixing Height Measurements
The following procedures are approaches that may be used to measure or •
determine mixing heights and other upper air data.
6.2.2.1 Doppler SODAR Measurement Approach -
The SODAR approach provides another more direct method for determining
mixing height data. The SODAR is a tool for remote measurement of meteorological
variables at heights up to several hundred meters above the surface. The SODAR
systems are being used more often than in the past for developing meteorological
databases required for input to air quality models. The SODARs have been accepted
in a few cases for on-site meteorological measurement programs. The applicability of
SODAR data must be approved for use on a case-by-case basis.
6-7
-------
Revision 0
OCTOBER 1991
The SODAR system measurements of wind speed and direction have been the
subject of performance demonstrations. Some of these demonstrations have been
encouraging. However, measurements of mixing heights through interpretation of
SODAR facsimile charts have been more ambiguous and thus are not as reliable as
more standard methods for estimating mixing heights.
Successful SODAR operation requires several factors that are unrelated, such
as low background noise and optimum meteorological conditions. The level at which
90% data recovery can be achieved varies between daytime and nighttime because of
the proximity of noise sources. Theoretically, SODAR daytime range should be
superior to night time range.
Meteorological conditions can have a profound effect on SODAR operations.
Incident acoustic energy from the SODAR is scattered by a volume of air depending
on small temperature and wind speed discontinuities and the direction of propagation.
The SODAR detects the backscatter based on the strength of the return signal
expressed by velocity and thermal structure functions. Strong return signals can be
produced within a convective or stable layer; therefore, the backscattering varies
diurnally, seasonally, and synaptically.
The SODAR systems perform best for pronounced temperature differences and
perform poorly during stable to neutral conditions. Summer periods result in the •
greatest thermal structure functions and thus provide the best SODAR range statistics.
On the downside, precipitation can enhance noise or destroy the SODAR antenna
diaphragm or dish. Other drawbacks include the instrument range and interpretation
of data. A convective layer appears on the facsimile chart as a series of spikes.
Elevated stable layers are not always strong enough to produce distinct traces. Both
of these types of layers may also be out of range of the instrument. Facsimile charts
are commonly set at 500 or 1000 meters.
The SODAR data are useful provided a well trained expert interprets the
facsimile charts. Strong signals indicating the presence of atmospheric "targets" do
not by themselves define mixing heights. Close scrutiny and analysis by a trained
6-8
-------
Revision 0
OCTOBER 1991
person experienced in meteorology and instrumentation would be required to infer
mixing height information through the analysis of time-height patterns of signal
strength.
Wind speed and direction can be determined for many heights. For air quality
models, particularly dispersion models, wind data at plume height or stack height may
be determined by selecting a single measurement height representative of an average
plume height or stack height. If model input data are required for heights exceeding
the SODAR range, wind speeds can be determined based on a logarithmic profile
based on data available from at least three levels. The SODAR system produces a
vector-averaged wind speed. Wind direction substitution should come from a level
that is at least 100 m high. The cut-off level for model input should be the highest
level in which data capture is at least 80 percent. If the data are reviewed and
validated, mean wind data values can be used as in put to air quality models.
Mixing heights can be analyzed using the facsimile charts produced by SODAR.
Although these data are applicable to air quality models on a case-by-case basis, it is
not recommended for routine model use. The primary reason for this is the SODAR's
inability to detect signals from atmospheric "targets" for all possible mixing height
ranges. The translation of the SODAR facsimile information into usable data is
primarily for use in a regulatory context or for model evaluation studies.
Use of a Doppler SODAR system in an on-site program should be closely
coordinated with the reviewing agency. An overall operational plan, including QA
procedures, should be prepared prior to data collection and data use. The details of
the operational plan will differ based on the specific SODAR instrument chosen.
The SODAR equipment siting criteria should be followed closely to obtain return
signals with sharp atmospheric peak frequencies. To identify potential noise
interferences, the following steps should be taken before siting:
Determine acceptable noise levels survey to concur with manufacturer's
minimum noise requirements.
6-9
-------
Revision 0
OCTOBER 1991
Identify potential noise sources.
Perform a quantitative noise survey to concur with manufacturer's
minimum noise requirements.
During siting, orientation of SODAR antennas may depend on how independent
air parcel measurements are to be made combined with the complexity of the terrain.
The instrument should be aligned with "true north," based on techniques described in
the "Quality Assurance Handbook for Air Pollution Measurement Systems: Volume IV,
Meteorological Measurements9". The following steps should be considered part of the
overall operation of the SODAR equipment.
Siting and Installation:
Noise survey - qualitative followed by quantitative, if necessary;
Identification of potential reflection targets;
Disturbance potential; and
True north alignment.
Operation and Maintenance Quality Control:
Specifications of the manufacturer;
Initial settings of 15 minutes for averaging period and at least 300 m for
height;
Collocated tower (at a minimum height of 10 m); and
Standard Operating Procedures.
Timely and thorough data review:
Daily; Weekly; and Monthly procedures.
6-10
-------
Revision 0
OCTOBER 1991
Quality Assurance Plan:
Major elements are QC procedures, periodic audits (conducted at
6 month intervals), and data validation.
Data Validation:
Should be carried out, on a component-specific basis prior to using the
data in a model for regulatory purposes. Management plan should
incorporate timely review and archiving of data.
Data Use:
An upper bound should be established where data capture is at least
80%, for developing model inputs. Mixing height may be acceptable on
a case-by-case basis.
Data Capture Requirements:
Valid hours defined as at least three complete valid levels for 30 minutes
out of an hour (two 15-min values), must be available 90% of the time.
6.2.2.2 Vertical Temperature Profile Measurement Approach --
Radiosonde balloon techniques and operations involve tethered balloons. The
tethered balloon technique has limitations, including manual operation and general
safety problems. These systems should be sited as far from aircraft routes or airports
as possible. Caution should be taken while using during adverse weather conditions,
such as lightning or high winds.
6-11
-------
Revision 0
OCTOBER 1991
SECTION 7.0
REFERENCES
1. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40,
Part 58. Enhanced Ozone Monitoring Regulations. Washington, D.C. Office of
the Federal Register. October 23, 1991.
2. L Purdue, "The Atlanta Ozone Precursor Study," Paper No. 91-68.8, Air and
Waste Management Association - 84th Annual Meeting, Vancouver, BC,
June 16-21, 1991.
3. U.S. Environmental Protection Agency, Compendium of Methods for the
Determination of Toxic Organic Compounds in Ambient Air, EPA-600/4-89/017,
Research Triangle Park, NC, June 1988.
4. D. Grassick, R. Jongleux, Enhanced Ozone Monitoring Network - Design and
Siting Criteria Guideline Document. Contract No. 68D00125/Work Assignment
21. OAQPS MD-14. U.S. EPA, RTP, NC May 1991
5. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40,
Part 136.1, Appendix B. Washington, D.C. Office of the Federal Register,
July 1, 1987.
6. H. Bloemen, Exploratory Chromatographic Data Processing, Paper No. 91-68.4,
Air and Waste Management Association - 84th Annual Meeting, Vancouver, BC,
June 16-21, 1991.
7. J.D. Pleil, Enhanced Performance of Nafion Dryers in Recovering Water From
Air Samples Prior to Gas Chromatographic Analysis, JAPCA 37:244-248, 1987.
8. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40,
Part 53. Washington, D.C. Office of the Federal Register, July 1, 1987.
9. U.S. Environmental Protection Agency, Quality Assurance Handbook for Air
Pollution Measurement Systems: Volume IV Meteorological Measurements.
EPA 600/4-82-060. Research Triangle Park, NC 27711. February 1983.
10. D-P. Dayton, D. A. Brymer, R. F. Jongleux, Canister Based Sampling Systems -
A Performance Evaluation, Proceedings VIP-17, 1990 International Symposium
on Measurement of Toxic and Related Air Pollutants, Raleigh, NC, April 1990.
7-1
-------
Revision 0
OCTOBER 1991
11. K.D. Oliver, J.D. Pleil, W.A. McClenny, Sample Integrity of Trace Level Volatile
Organic Compounds in Ambient Air Stored in SUMMA®-Polished Canisters,
Atmospheric Environ., 20, 1403 1986.
12. M. Holdren, D. Smith, W. McClenny, Storage Stability of Volatile Organic
Compounds in SUMMA®-Polished Stainless Steel Canisters, Final Report, U.S.
Environmental Protection Agency, Research Triangle Park, NC,
January 22, 1986.
13. R. McAllister, D-P. Dayton, J. Rice, P. O'Hara, D. Wagoner, and R. Jongleux,
Stability Study - Final Report. Scientific Instrument Specialists, Moscow, ID,
March 23, 1988.
7-2
-------
Revision 0
OCTOBER 1991
APPENDIX A
DISCUSSION, ISSUES, AND SELECTED PROCEDURES
RELATED TO CANISTER SAMPLING
-------
Revision 0
OCTOBER 1991
A1.0 CANISTER SAMPLING ISSUES
The use of canister sampling for collecting and consequently determining
concentrations of VOCs in ambient air is an integral part of the sampling strategy and
recommended monitoring requirements specified in the proposed revisions to 40 CFR
Part 58. The technology utilizes stainless steel canisters with interior surfaces
conditioned by the SUMMA® process. The SUMMA® process is a proprietary
treatment that passivates the internal surfaces of the canister to minimize surface
reactivity. This process allows stable storage for many of the compounds of interest.
Passivated stainless steel canisters in a variety of volumetric sizes are commercially
available from several manufacturers.
An important advantage of the canister based methodology is that the collected
whole air sample can be divided into portions for replicate analyses (permitting
convenient assessment of analytical precision) and reanalyses using different analytical
systems for specific peak identification and confirmation. General canister sampling
procedures are described in Section 2.3 of the document.
A2.0 PRECAUTIONS IN THE USE OF CANISTERS
The canister sampling technique is not without potential problems. Primary
problem areas associated with canister sampling include contamination and sample.
stability. If not controlled, these problems can significantly reduce the quality and
usefulness of the data obtained using the canister sampling technique. The general
discussion and guidance presented below are intended to provide users with
information that should minimize these problems.
A2.1 CONTAMINATION
Contamination may cause additional, unsampled compounds to appear in the
sample or increase the concentrations of sampled compounds. Contamination may
also cause losses of sampled compounds or may introduce compounds that interfere
with gas chromatographic sample analysis. Contamination can originate from the .
sample canisters, canister cleanup systems, components in the sampling systems or
A-1
-------
Revision 0
OCTOBER 1991
analytical system, and improper canister storage practices. These problems become
more significant as analytical sensitivities are lowered.
To minimize collection system contamination, canisters should be purchased
from a reputable supplier who uses high-quality manufacturing and final cleaning
procedures. Purchase requirements should specify contamination-free valves and
criteria for maximum residual concentrations of target compounds. New canisters
should be inspected carefully for proper welding and fittings and should always be
blanked (filled with humidified zero air and analyzed) before use to check for
contamination. Canisters with excessive contamination should be returned to the
supplier or cleaned repeatedly until usable. Some contaminated canisters may appear
uncontaminated immediately after cleaning but will outgas contaminants upon storage
for several weeks. All canisters in routine use should be blanked frequently, and
particularly after extended periods of storage, to be sure that significant contamination
does not appear.
Canisters used for ambient or low-level measurements should be segregated
from those used for higher-level concentrations or for higher-molecular-weight
compounds. Higher-molecular-weight compounds are more difficult to remove from
the internal canister surface.
Canister cleanup systems should be constructed of clean, high-quality stainless
steel components, contain suitable cryogenic traps, and be operated systematically
and meticulously to avoid system contamination from vacuum pump oil, poor quality
zero air, water used in humidification systems, room air, or other sources.
Sampling and analytical systems should be constructed of clean, high quality
components, with particular attention paid to pumps, valves, flow controllers, or
components having any non-metallic surface. Before installation and at periodic
intervals, samplers should be carefully tested for contamination or compound loss by
analyzing collected samples of zero air and known concentrations of target
compounds. This procedure is termed "certification" and allows the potential
contamination characteristics of each specific sampling system to be assessed. A
A-2
-------
Revision 0
OCTOBER 1991
canister sampling system certification protocol, as presented in Section A4.0, should
be implemented to ensure that the status of each sampler is known prior to use.
Equipment found to be contaminated should be tested further to attempt to
identify the source of the contamination. Contaminated components should be
replaced or cleaned, and the system recertified. Minor contamination can often be
reduced by purging the system extensively with humidified zero air.
The entire measurement system (sampling and analytical) should be checked
regularly for additive and subtractive biases to ensure that measurements obtained are
representative. Such checking involves extensive and continual testing of the
analytical systems, sampling systems, sample canisters, and canister clean up
systems. Program checks also involve using humidified zero air and standards of
known concentration, to perform canister blanks and system audits. Collection of
samples from collocated systems and other quality assurance techniques should also
be performed. This effort represents a substantial overhead for a monitoring project,
but there is no other way to ensure that the resulting data are credible.
A2.2 SAMPLE STABILITY
While many compounds have been shown to be stable in canisters over one or
more weeks of storage, it is not known how these results extend to the variety of
conditions that may be encountered. These conditions include variable quality of the
canisters and their SUMMA® treatment, variable moisture content in the sample air,
previous history or residual contamination of the canister, sample pressure in the
canister above or below atmospheric pressure, storage temperatures, and canister
age.
Current information appears to indicate that hydrocarbon VOCs with vapor
pressures above 0.5 mm Hg at 25°C store well in canisters. Substituted
hydrocarbons, particularly the halogenated hydrocarbons with similar vapor pressure
properties, also store well in canisters. Laboratory tests indicate that many
oxygenated hydrocarbons such as aldehydes, ketones, and alcohols have less
desirable storage properties. As a general rule, organic compounds that are soluble
A-3
-------
Revision 0
OCTOBER 1991
in water do not store well in canisters. Target analytes for which there is little stability
information or for which storage stability characterization is questionable should be
specifically tested for storage stability in the canisters that will be used and under
typical conditions of use. Sample stability is addressed in more detail in Section A5.0.
A2.3 CANISTER LEAKAGE
There are three potential sources of canister leakage. These sources are:
• Faulty canister welds;
• Leakage at the connection of the valve to the canister; and
• Leakage through the valve.
A faulty weld is a manufacturing defect. Faulty welds are fairly rare and can be
detected by conducting leakage acceptance tests. Canisters may also sustain
physical damage. Damaged canisters should be repaired and retested for leaks.
Leaks at the connection of the valve to the canister are the most troublesome
type of leak. Welding the valve to the canister virtually eliminates such leaks but
makes subsequent valve replacement impractical and expensive. Usually, the valve is
connected to the canister using a standard tubing compression fitting. Properly •
installed, these fittings are very reliable. However, these fittings can loosen when an
operator improperly opens and closes the valve. If the valve rotates with respect to
the canister during opening and closing, small leaks in this fitting can occur.
Overtightening the fitting in an attempt to prevent such movement exacerbates the
problem, as does any other physical strain on the connection. Short of welding the
valve to the canister, vulnerability to leakage in this connection can be greatly reduced
by:
Using an oversize fitting (e.g., 5/16-inch or 3/8-inch rather than
1/4-inch);
A-4
-------
Revision 0
OCTOBER 1991
Equipping the canister with a valve guard to protect the valve from
physical strain; and
Mechanically clamping or fastening the valve to the canister or valve
guard to prevent rotation during opening or closing.
These measures are offered by some canister manufacturers and should be specified.
Even with these precautions, periodic retesting of canisters is necessary to ensure that
no significant leaks in the valve connection develop with extended use.
Leaks through the valve can occur if the valve seat has become damaged
through wear or overtightening. The practice of installing a cap on the valve
connection when the canister is not connected to a sampling system effectively
minimizes sample or vacuum loss during periods of storage.
A canister may quickly be tested for obvious leaks by pressurizing it with zero
air and submerging it in clean water to look for bubbles. To check for microleaks, the
canister should be evacuated and its pressure observed for several days with a
sensitive absolute pressure gauge connected. This test is performed with the canister
valve open. To check the valve for leakage, evacuate the canister, check the absolute
pressure, close the valve, and disconnect the pressure gauge, leaving the valve
connection open. Then reconnect the pressure gauge and check the pressure several
days later. The canister pressure should not increase more than a few mm Hg during
that period.
Canisters with excessive leaks must be repaired and repassivated or replaced,
but those with relatively minor microleaks can be used for many applications if
precautions are taken. Canisters determined to have microleaks can be prepared for
use just prior to sample collection and analyzed promptly after sample collection.
Reduction of the pre- and post-sampling time reduces the potential for bias. Between
evacuation and analysis, the canister connection should be tightly capped, and the
canister should be stored in a well ventilated, non-contaminated area. Avoid storing
the canisters in automobiles and laboratories where organic materials are used.
A-5
-------
Revision 0
OCTOBER 1991
Storage or shipping containers should be of an all-metal construction, well ventilated,
and should avoid the use of foam padding and other organic materials.
A2.4 HISTORICAL CANISTER LOG
A historical record should be maintained for each canister to record the type of
sampling service for which the canister is employed. This record will include the dates
samples were taken, type of samples taken (ambient or source sample), dates of
cleaning, dates of measuring cleanliness of the canister, and dates for leak checking
the canister. Canisters used for ambient sampling should not be used for source
sampling, and vice versa. Canisters used for source sampling typically contain much
higher concentrations of organic compounds. Corrosive gases may damage not only
valve seats in the inlet canister valve, but also the interior surface of the canister itself.
The SUMMA®-treatment of the interior of a stainless steel canister is designed to
passivate the surface, not only to decrease any catalytic potential of the surface for
chemical reaction of the components in the sample, but also to decrease the active
centers for adsorption of compounds, (e.g., organic pollutants), which are found in the
ambient air samples taken.
A3.0 CANISTER CLEANING
The canister cleaning procedure and equipment described in this section are.
recommended when obtaining integrated whole ambient air samples for subsequent
analysis of VOCs. The cleaning procedure involves purging the canisters with cleaned
humidified air and then subjecting them to high vacuum.
The purpose of canister cleaning is to ensure that the canister interior surfaces
are free of contaminants and that the canister meets a predetermined cleanliness
criterion (i.e., <0.20 ppmC NMOC). This level of cleanliness minimizes the potential for
carryover of organic pollutants from one sample to the next, and helps ensure that the
samples collected are representative.
A-6
-------
Revision 0
OCTOBER 1991
A3.1 EQUIPMENT
The equipment required to clean canisters includes a source of clean,
humidified air to pressurize the canisters to a pressure of 20 psig, and a vacuum
system for evacuating the canisters to 0.5 mm Hg absolute pressure. Air from a
standard oil-less air compressor will contain pollutants from the ambient air. In
addition, various VOCs will be found in the compressed air because of the lubricants
used in the air compressor. Hydrocarbon-free air may be purchased in cylinders and
humidified before being used in the cleaning process; however, this approach may be
cost-prohibitive. Figure A3-1 presents the schematic of a canister cleanup system that
is suitable for cleaning up to 16 canisters concurrently. This, and any alternative
system must include a vacuum pump capable of evacuating the canisters to an
absolute pressure of 0.5 mm Hg. The equipment is designed so that one manifold of
eight canisters is undergoing the pressurization portion of the cleaning cycle while the
other manifold of eight canisters is undergoing the vacuum portion of the cleaning
cycle.
The following equipment is incorporated in a canister cleaning system.
Air Compressor - A shop or laboratory oil-less air compressor used to provide
the air supply for the canister cleanup apparatus.
Coalescing Filter - A coalescing filter designed to remove condensed moisture
or hydrocarbon contaminants present in the air supplied from the air compressor.
Permeation Dryers - Permeation dryers used to dry the air prior to introduction
into the catalytic oxidizers. Two permeation dryers are installed in parallel.
Moisture Indicators - Visual moisture indicators installed in the transfer lines
between the permeation dryers and the catalytic oxidizers to monitor the performance
of the permeation dryers.
Catalytic Oxidizers - Catalytic oxidizers installed in the clean-air system to
oxidize any hydrocarbon contaminants that may be present in the air supplied by the
air compressor. For best results and most efficient operation of the catalytic oxidizers,
manufacturer's specifications should be strictly followed.
A-7
-------
>
00
Coalesing
Filter
Filter Assemblies
Permapure Dryers
Moisture
Indicators
Air Compressor
Air Flow
Rotameters
1
1
-O
Pressure
Regulators
Air Cryotrap
Purge Valves
? 3£
f
•3
M
„
Air
Cryotrap
|
1
Vacuum
Absolute
Pressure
Gauge
Air
Humidifier
O
A)r Bypass •*-
j t Routing
^? Valve
A][,B
Vacuum Source V 7
Selector Valve j_ £
C
-&*-
i&B*
DDDDDDDD
D D D D D
To Certification System
A. Manifold Air Pressure Valve
B. Manifold Vacuum Valve
C. Manifold Pressure Release Valve
D. Manifold Port for Connecting Canisters to be Cleaned
Catalytic
Oxidizers
Turbomolecular
Pump
Roughing
Pump
C
-£*}-
D D
8-Port
Manifold
5015837R
o
CD 3D
m 9
= §'
CD
O
-------
Revision 0
OCTOBER 1991
Filter Assemblies A 5-micron sintered stainless steel filter installed in the filter
housing assembly downstream of each catalytic oxidizer to trap any particulate
material that may be present in the air stream leaving the catalyst bed of the oxidizer.
Air Cryotrap and Purge Valves - The air cryotrap allows the cleaned air supply
lines to be subjected to cryogenic temperatures to condense (1) water formed during
the oxidation of hydrocarbons, (2) any remaining unoxidized hydrocarbons, and
(3) other condensables. Air cryotrap purge valves are used to purge these condensed
components from the air cryotrap, as described in the operating procedure described
below.
Pressure Regulators - A high purity dual stage pressure regulator installed in
each branch of the air supply line so that the maximum pressure attained during the
cleanup procedure is controlled at 20 psig.
Flow Controllers - The flow control devices shown in the canister cleanup
schematic (Figure A3-1) are metering valves. The flow rates are set not to exceed the
maximum recommended flow rate through the catalytic oxidizers.
Air Flow Rotameters - Rotameters installed in the air supply lines to allow
monitoring of the flow rates through the catalytic oxidizers.
Air Humidifier - The air humidifier shown in Figure A3-1 is a SUMMA®-
passivated, double-valve stainless steel canister with an inlet dip tube that projects1 to
the bottom of the sphere. HPLC-grade water is placed in the canister prior to use.
Two rotameters are connected to control air flow so that about 80% of the flow rate
can be directed to the humidifier (to bubble through the water to become saturated),
while the other 20% bypasses the humidifier. This procedure allows the humidification
apparatus to supply cleaned, dried air that has been humidified to a relative humidity
of -80 percent.
Manifold Air Pressure Valves - Manifold air pressure valves used to isolate the
air supply system from the manifold, or to make the pressurized air available to the
manifold.
A-9
-------
Revision 0
OCTOBER 1991
Eight-Port Manifolds - Eight-port manifolds designed to allow up to eight
canisters at a time to be connected. Fewer canisters may be connected to the
manifold if the vacant ports are sealed off with a plug fitting.
Roughing Pump - The roughing pump shown in Figure A3-1 is a high-capacity
diaphragm vacuum pump used to remove the moist cleaning air from the canisters
while evacuating the canisters to about 100 mm Hg absolute. The high moisture
content of the cleaning air contained in the canisters will not impede the function of
this diaphragm style pump, but will impede the performance of the high-vacuum pump.
High-Vacuum Pump - A high-vacuum pump capable of reducing the pressure in
the canisters to 0.5 mm Hg absolute. High moisture content will impede the
performance of the high-vacuum pump.
Vacuum Crvotrap - A U-shaped trap located in the vacuum manifold that is
sized to fit inside a Dewar flask filled with cryogen. The purpose of this trap is to
condense water vapor from the air that is pulled from the canisters during the vacuum
cycle and prevent back-diffusion of organic vapors from the high-vacuum pump into
the canisters during the vacuum cycle of the cleaning procedure.
Vacuum Source Selector Valve - The vacuum source selector valve is a
multiposition valve used to route either the roughing pump or the high vacuum pump
to the eight-port manifold assemblies or isolates both pumps from the manifold •
assemblies.
Compound Absolute Pressure Gauge - An absolute pressure gauge used to
measure the pressure attained in the canisters during the vacuum and pressurization
cycles of the cleaning procedure. The absolute pressure gauge must be able to
measure absolute pressures from 40 psig down to 0.5 mm Hg absolute.
Air Bypass Valve - The air bypass valve is used to allow for a 1.0 L/min flow of
air to be maintained through the catalytic oxidizers when the cleaning system is not in
use. This flow prevents the oxidizers from overheating when the clean up system is
not in use.
A-10
-------
Revision 0
OCTOBER 1991
Manifold Valves The manifold vacuum valve and the manifold pressure valve
are used to apply vacuum or pressure to the canisters, as required during the cleaning
procedure.
Manifold Ports - The manifold ports permit connection of the canisters to the
manifold. Fittings that mate directly with the canister valve fittings are used. These
connections will not leak during the pressurization portion or the vacuum portion of the
cleaning procedure.
A3.2 CLEANING PROCEDURE
The cleanup system is prepared for use by checking the position of all the
valves. All valves should be closed initially, with the exception of the air bypass valve.
Fill both the air source and vacuum pump vacuum flasks with cryogen and actuate the
high-vacuum pump. Ensure that these vacuum flasks remain filled with cryogen
throughout all cleanup activities. The inlet bellows valve on the humidifier is opened
and the valve on the wet air rotameter is also opened. Close the valve on the dry air
(bypass) rotameter to allow the air to become humidified. Allow the system to stabilize
for 10 minutes. After preparing the cleanup system, canister cleaning is performed
using the following procedure.
1. Connect the canisters to be cleaned to the cleaning manifolds. Record
the canister numbers and pre-cleanup concentrations, if available, as determined by
the last analysis, in the appropriate cleanup and canister history log book. Record
data pertinent to the vacuum and pressure cleanup cycles as they are completed.
2. Remove collected moisture from the air cryotraps by opening and
immediately closing the air cryotrap purge valves. Removal of the collected moisture
should be performed at the beginning of each pressure cycle, so that the cryotraps do
not plug with ice.
3. Release pressure from the canisters by opening all the canister bellows
valves and then opening the manifold pressure release valve. When venting is
complete, leave the canister bellows valves open and close the manifold pressure
release valve.
A-11
-------
Revision 0
OCTOBER 1991
4. Begin the first vacuum cycle by actuating the roughing pump, placing the
vacuum source selector valve in the roughing pump position, and opening the
manifold vacuum valve.
5. Evacuate the canisters to approximately 100 mm Hg, as indicated by the
absolute pressure gauge.
6. Position the vacuum source selector valve in the high-vacuum pump
position.
7. Evacuate the canisters to 0.5 mm Hg absolute pressure (or less) and
maintain the vacuum for 30 minutes.
8. Close the manifold vacuum valve after the 30-minute high-vacuum period
has been completed.
9. Begin the first pressure cycle by purging the air cryotraps (refer to
Step 2) and then closing the air bypass valve. Open the manifold air pressure valve.
Using the air flow control valves, adjust the air flow rate to the manufacturer's
recommended optimum flow rate for the oxidizers, as indicated by the air rotameters.
10. Check the pressure regulators to verify that they are set to deliver a final
pressure of 20 psig. Fill the canisters to 20 psig. As the final pressure is attained, the
flow rates indicated on the air rotameters will drop to zero, regardless of the setting on
the flow controllers because the pressure in the canisters and the pressure at the. exit
of the regulators reach equilibrium.
11. Close the manifold air pressure valve when filling is complete. Open the
air bypass valve and adjust the air flowmeters to 1.0 L/min.
12. Release the pressure from the canisters after they have been under a
20 psig pressure for 30 minutes by opening the manifold pressure release valve.
13. Repeat steps 4, 5, 6, 7, and 8 for Vacuum Cycle 2.
14. Repeat steps 9, 10, 11, and 12 for Pressure Cycle 2.
15. Repeat steps 4, 5, 6, 7, and 8 for Vacuum Cycle 3.
16. Repeat steps 9, 10, and 11 for Pressure Cycle 3.
17. Close all of the bellows valves on the canisters.
A-12
-------
Revision 0
OCTOBER 1991
A3.3 BLANKING PROCEDURE
Prior to initial use, the cleanliness of all canisters should be assessed. After the
initial blanking of 100% of the canisters, the blanking frequency can be reduced. One
canister on a cleaning bank of eight canisters is considered representative and should
be blanked. The selection of the canister to be blanked (from the bank of eight
canisters) is determined by selecting the canister with the highest pre-cleanup total
NMOC concentration on the manifold. This canister is selected because the potential
for compound carryover is most likely to be the largest of any of the canisters on the
manifold. The blank sample is analyzed using Compendium Method TO-12 (See
Section 3.0 in the main document). If this measurement meets the predetermined
cleanliness criterion (i.e., <0.020 ppmC), then the other canisters on the manifold are
considered clean. Blanking is a part of the overall canister cleanup procedure, and is
described below.
1. Select the canister to be blanked by referencing the cleanup history
logbook to determine the canister with the highest pre-cleanup total NMOC
concentration.
2. Verify that all the canister bellows valves are closed. Disconnect the
canister selected to be blanked.
3. Analyze the air in the canister using Compendium Method TO-12 (See
Section 3.0). If the canister analysis meet the predetermined concentration criterion
(i.e., <0.020 ppmC), then the blanked canister and all the other canisters on the bank
of eight canisters are considered clean.
4. If the canister does not meet the concentration criterion, it is reconnected
to the manifold. The entire bank of canisters is given another vacuum and pressure
cycle. After the additional cycle, the same canister is blanked again.
5. After the canister is blanked and has met the concentration acceptance
criterion, it is reconnected to the manifold.
A-13
-------
Revision 0
OCTOBER 1991
A3.4 FINAL EVACUATION PROCEDURE
After cleaning and blanking, the canisters are ready for final evacuation in
preparation for sample collection. The procedure for final evacuation is described
below.
1. Release the pressure from the canisters by opening the manifold
pressure release valve and opening all of the canister bellows valves. When venting is
complete, close the manifold pressure release valve.
2. Begin final evacuation of the canisters by actuating the roughing pump,
placing the vacuum source selector valve in the roughing pump position and opening
the manifold vacuum valve.
3. Evacuate the canisters to approximately 100 mm Hg, as indicated by the
absolute pressure gauge.
4. Activate the turbomolecular vacuum pump, checking to be sure there is
liquid cryogen in the vacuum cryotrap.
5. Switch the vacuum source selector valve to the high-vacuum pump
position. Allow the canisters to evacuate to 0.5 mm Hg, as indicated by the absolute
pressure gauge.
6. Close the canister bellows valves on all of the canisters on the manifold.
Close the manifold vacuum valve.
7. Disconnect the canisters from the manifold and remove any old
identification tags. Store the cleaned canisters in the designated storage area.
A4.0 CANISTER SAMPLING SYSTEM CERTIFICATION
Canister sampling systems should exhibit non-biasing characteristics before
being used to collect samples. These sampling systems should be subjected to
laboratory certification to quantify any additive or subtractive biases that may be
attributed directly to the sampling system. The following procedure is recommended
for certifying canister sampling systems. Alternative approaches are acceptable
provided they are properly described and documented.
A-14
-------
Revision 0
OCTOBER 1991
A challenge sample, consisting of a known concentration blend of organic
compounds in clean humidified zero air, is collected through the sampling system. A
reference sample is concurrently collected using a dedicated mass flow controller that
has been characterized prior to each use. The samples are then analyzed using a GC
system that is equivalent to or better than the GC system that will be used to analyze
field volatile organic O3 precursor samples. The percent recoveries for target
challenge compounds are calculated, based on the determined reference sample
concentrations. Recoveries of each of the challenge compounds should be in the
range of 80-120% of the concentrations determined for the reference sample. A
system-specific overall recovery should also be calculated. The overall recovery is the
average of the individual compound recoveries. Each sampling system should have
an overall recovery of 85-115 percent. The challenge sample percent recoveries are
used to gauge potential additive and/or subtractive bias characteristics for each
specific sampling system10.
In addition to characterizing the sampling system with a blend of VOCs, the
system should also be characterized using humidified zero air. A humidified zero air
blank sample is collected through the sampling system to further gauge the potential
for additive bias. The blank samples can be analyzed for specific target analytes, total
NMOC, or both, depending on individual program requirements. Two criteria apply to
the blank portion of the certification process: a determined concentration criteria of
0.2 ppbv or less for any individual target compound is required if speciation analysis of
the blank sample is performed, and a total NMOC concentration criteria of 10 ppbC or
less is also required.
Sampling is accomplished using dedicated manifolds for both the zero and
challenge phases of the certification procedure (See Figures A4-1 and A4-2). Zero air
supplied to the zero manifold should be hydrocarbon-free and humidified to
approximately 70% relative humidity. The zero air should be supplied from a canister
cleaning system similar to the one described in Section A3.1.
A-15
-------
Cleaned Humidified
Zero Dilution Gas (In)
O)
Sample
Canisters
Canister
Sampling
Systems
Flow
Controller
Canister
Sampling
Systems
Sample
Canisters
Exit
Rotameter
O
§
CD
m
33
_x
CO
CO
o'
a
-------
Certified Challenge
Stock Gas (In)
Cleaned Humidified
Zero Dilution Gas (In)
Mixing Chamber
Controlled Temperature Manifold
Characterized
Flow
Controller
Sample
Canisters
Canister
Sampling
Systems
Canister
Sampling
Systems
Reference
Sample
Sample
Canisters
O
O
CO 30
m 9
CD
CD
O
13
Figure A4-2. Canister sampling system certification schematic - challenge gas.
-------
Revision 0
OCTOBER 1991
A4.1 EQUIPMENT
The equipment required to perform canister sampling system certification is
described below. The equipment listed is consistent with the systems presented in
Figure A4-1 and A4-2.
Mass Flow Controllers - Mass flow controllers located at the inlets to the
manifolds. Mass flow controllers are used to regulate the certification pollutant,
diluent, and zero air flow rates. Also, a dedicated, characterized mass flow controller
is used to collect reference samples.
Mixing Chamber - A mixing chamber located between the outlets of the mass
flow controllers and the inlet of the challenge manifold. The mixing chamber is a
stainless steel vessel with opposed inlet and outlet ports that cause the blend of
challenge gases and the diluent gas to swirl and mix prior to entering the challenge
manifold. The mixing chamber is used to ensure that a homogeneous blend of
challenge gas is delivered to the challenge manifold. The zero manifold does not
require a mixing chamber.
Challenge Gas Manifold - A challenge gas manifold constructed of 1 /8-inch
O.D. chromatographic grade stainless steel tubing and 1/8-inch tee fittings. The
challenge manifold is used to distribute challenge gas to the individual
sampling systems being certified. The number of sample ports provided on the •
challenge gas manifold is determined by the number of sampling systems to be
certified simultaneously.
Zero Air Manifold - A zero air manifold constructed of 1/4-inch O.D.
chromatographic grade stainless steel tubing and 1 /4-inch fittings. The zero manifold
is used to distribute zero air to the individual sampling systems being certified. The
number of sample ports provided on the zero air manifold is determined by the
number of sampling systems to be certified simultaneously.
Exit Rotameter - An exit rotameter located at the outlet of both the challenge
gas and zero air manifolds. The exit rotameter is used to visually indicate that an
A-18
-------
Revision 0
OCTOBER 1991
excess of challenge gas or zero air is present in the respective manifolds during
certification sample collection.
Cord Heater - A cord heater rated at 80 watts spiraled around the outside of the
challenge manifold. The cord heater is used to heat the challenge manifold to 80°C.
Heating the challenge manifold helps to reduce the potential for loss of challenge gas
compounds to the walls of the challenge manifold. The zero manifold is not heated.
Temperature Controller - A temperature controller used in conjunction with the
cord heater to actively regulate the challenge manifold temperature at 80°C.
A4.2 CERTIFICATION PROCEDURE
The procedure to perform canister sampling system certification is presented
below.
1. Perform a positive pressure leak check of all sampling system fittings.
Attach source of pressurized air to the inlet of the system. Coat the fittings with
indicating bubble solution to locate leaks. Repair any leaks found. Perform a negative
pressure leak check. Attach an evacuated canister to the exit of the sampling system.
Open the canister bellows valve and record the initial vacuum, indicated by the sample
pressure gauge. Close the canister bellows valve and view the sample pressure
gauge and determine whether vacuum is maintained. The system is leak free if the
vacuum is maintained. If vacuum is not maintained, the system is not leak free.
Repair leaks and retest the system.
2. Connect the sampling systems and the reference sample flow controller
to the zero manifold and purge them with humidified zero air for 48 hours. The purge
air should be simultaneously routed to the challenge manifold to clean and prepare it
for challenge sample collection. Terminate the humidified zero air flow at the end of
the 48 hour period.
3. Purge the sampling systems, reference system, and manifold with dry
zero air for 1 hour to removed accumulated moisture. During the dry purge,
determine the certification flow requirements using the following equation:
A-19
-------
QT
Where:
QT
Revision 0
OCTOBER 1991
[((Q, x N,) + (QR x N2)) x FJ
Individual sampling system collection flow rate
(mL/min)
Number of sampling systems
Reference system collection flow rate (mL/min)
Number of reference systems
2.0; Excess flow factor
Total required flow rate (mL/min)
4. Determine the pollutant and diluent flows required to generate the
desired concentration of challenge gas using the following equations:
Where:
Step 2
Where:
QT
C1/C2
Desired challenge gas concentration (ppbv)
Concentration of the stock cylinder (ppbv)
Dilution factor (for use in next equation)
F2xQT
Total required flow rate
Pollutant flow rate (mL/min)
A-20
-------
Revision 0
OCTOBER 1991
Where:
QD = Diluent flow rate (mL/min)
5. Generate and deliver the challenge gas to the challenge manifold and
sampling systems. Condition the challenge manifold with the challenge gas for 10
minutes, with the sampling systems off. Condition the challenge manifold an additional
10 minutes with the sampling systems on, and in the bypass mode. Connect a clean
evacuated canister to each sampling system.
6. Collect the challenge and reference samples. Conduct challenge sample
collection according to the normal specified operation of the sampling system.
7. Connect the sampling systems to the zero manifold and purge with zero
air, humidified to 100% relative humidity, for 48 hours. Dry the manifold and samplers
with dry zero air for 1 hour. Adjust the zero air stream to 70% relative humidity.
Condition the zero manifold for 10 minutes with the sampling systems off. Condition
the zero manifold an additional 10 minutes with the sampling systems on, and in the
bypass mode. Connect a clean evacuated canister to each sampling system.
8. Collect the humidified zero air blank samples. Conduct the blank sample
collections using the same sampling system operating procedures used during the
challenge sample collection.
9. Analyze the zero and challenge samples and calculate the % recoveries.
A5.0 SAMPLE STABILITY
Sample stability refers to the representativeness of the ambient air sample
contained in a canister after sample collection and storage. For the sample to be
stable, the compound matrix and concentrations of the sample must not change
significantly with time. There are at least four ways that the concentration of target
compounds in an ambient air sample may change after sampling:
A-21
-------
Revision 0
OCTOBER 1991
Chemical reaction of a compound with another compound or with itself;
• Adsorption or desorption on the interior surfaces of the canister or on
particulate matter in the sample from the ambient air;
Dilution of the sample with another gas or liquid after sampling; and
• Stratification of the sample in the canister.
A number of studies11'12'13 have shown that a wide range of VOCs are stable in
canisters for at least 30 days. Most of the reported studies were performed in
SUMMA®-treated stainless steel canisters at pressures above atmospheric pressure.
SUM MA® passivation of the interior surfaces of the canisters is designed to passivate
the surfaces to minimize catalytic activity on the surface and to reduce the number
and activity of adsorption sites on the canister's interior walls.
A5.1 POSITIVE PRESSURE SAMPLES
Samples obtained so that the final sample pressure is above atmospheric
pressure (typically 5 to 20 psig) are considered positive pressure samples. Positive
pressure samples are the least likely to be affected by the attainment of adsorption
equilibrium in the canister after sampling. The only precaution recommended in this
regard is that after sampling, no sample is withdrawn until the sample has been in the
canister for at least 24 hours to allow the adsorption equilibria to stabilize.
A5.2 DILUTED SAMPLES
Samples may be diluted by adding pressurized, clean air, N2, or other gaseous
diluent. It is recommended that at least 24 hours elapse between dilution of a sample
and removal of an aliquot for analysis.
A5.3 REPEATED ANALYSES
At least 24 hours must elapse between removal of aliquots from a sample
canister. If repeated analyses are performed on the contents of a canister, 24 hours
should elapse between successive removals of aliquots for analysis. If samples are
A-22
-------
Revision 0
OCTOBER 1991
removed from the canister without letting the adsorption equilibria within the canister
stabilize, the measured concentrations may change.
A-23
-------
Revision 0
OCTOBER 1991
APPENDIX B
LIST OF
MATERIALS, EQUIPMENT, AND VENDORS
-------
Revision 0
OCTOBER 1991
Vendor
Product
ACE Glass, Inc.
1430 N. West Boulevard
Vineland, NJ 08360
(609) 692-3333
AIITech Associates
2051 Waukegan Road
Deerfield, IL 60015
(800) 642-4667
Andersen Samplers, Inc.
4215-C Wendell Drive
Atlanta, GA 30336
(800) 241-6898
Climatronics
104 Wilbur Place
Bohemia, NY 11716
(516) 567-7300
J&W Scientific
91 Blue Ravine Road
Folsom, CA 95630
(916) 985-7888
MET ONE Instrument
479 California Avenue
Grants Pass, OR 97526
(503) 471-7111
Nutech
2806 Cheek Road
Durham, NC 27704
(919) 682-0402
Perma-Pure Products
P. O. Box 2105
Toms River, NJ 08754
(201) 244-0010
Glass Sampling Manifold Assembly
Chromatographic Grade Stainless Steel
Tubing Supplies and Standards
Sample Canisters and Canister
Sampling Systems
Meteorological Equipment
Analytical Columns
Meteorological Equipment
Sample Canisters and Canister
Sampling Systems
Permeation Drier
B-1
-------
Revision 0
OCTOBER 1991
Vendor Product
Restek Corporation Analytical Columns
Perm Euger Indust. Pk
110 Benner Circle
Bellefonte, PA 16823
(814) 353-1300
Scientific Instrumentation Sample Canisters and Canister
P. O. Box 8941 Sampling Systems
Moscow, ID 83843
(208) 882-3860
Supelco Incorporated Chromatographic Grade Stainless Steel
Supelco Park Tubing Supplies and Standards
Bellefonte, PA 16823-0048
(814) 359-3441
Waters Associates Carbonyl Sample Cartridges
34 Maple Street
Milford, MA 01757
(800) 252-4752
B-2
------- |