»EPA
United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
Research Triangle Park NC 2771 1
EPA-.600/4-33-027
June 1983
Research and Development
Technical Assistance
Document for
Sampling and
Analysis of Toxic
Organic Compounds in
Ambient Air
-------
SECTION 1
INTRODUCTION
This Technical Assistance Document (TAD) is intended to serve as a guide to those
persons responsible for designing and implementing ambient air monitoring programs for
toxic organic compounds. Presumably technical personnel within regional, state, and local
environmental regulatory agencies will be the primary users of this document, although
the scope is sufficiently broad that other persons may find it useful for the design of
monitoring programs.
This document is not intended to serve as a single source of information from which
all necessary technical input concerning toxic organic monitoring can be obtained.
Rather, the TAD should be used as an overview document to be consulted during the
initial phases of a monitoring program, with the user being directed to the referenced
literature or other specific, detailed information as required. The structure of the TAD is
such that the reader will normally recognize points at which such detailed information is
required, and a substantial amount of this information can be obtained from the secondary
sources referenced throughout the document.
Obviously this document does not eliminate the need for sound technical judgement
on the part of those persons planning a given monitoring program (i.e., the TAD is not a
"cookbook"). In general, persons planning such programs must have a strong chemistry
background with specific knowledge and expertise in instrumental sampling and analysis
techniques. If such persons are not available to a given program the probability of success
will be extremely low and the information obtained inaccurate. Indeed, the process of
ambient air monitoring for trace levels of organic compounds is highly sophisticated and
fraught with difficulties which are frequently not recognized by nontechnical personnel.
On the other hand, most ambient air monitoring programs are devised in response to
a public or regulatory need which is frequently nontechnical. Consequently, the technical
personnel responsible for planning and implementing the details of the program must
consult with the appropriate policy personnel to ensure that the technical objectives of
the program are directed towards meeting that regulatory or public need. This
technical/policy interface is extremely important and should be implemented in a very
early stage of development of a monitoring program in order to ensure that the regulatory
and/or public objectives can be technically achieved and data presented in a useful
format.
The TAD is presented in four major sections, including this introductory section.
The second section deals with regulatory policy and public issues which lead to the need
for ambient air monitoring programs. Special emphasis is given to the examination of
regulatory or public objectives as they relate to the technical design and implementation
of monitoring programs.
-------
The third section presents a set of detailed guidelines for development of an
ambient air monitoring program. The stepwise procedure of defining objectives, com-
piling existing information, selecting sampling and analysis methods, selecting a sampling
strategy, specifying QA and safety procedures, and defining data format is presented.
The intended purpose of this section is to give the reader an appreciation for the factors
to be considered in designing a monitoring program at each stage of development. This
section again emphasizes the importance of the technical/policy interface, especially in
the definition of program objectives. Although the development process is presented as a
stepwise process, it is actually an iterative process wherein conflicts between overall
objectives and sampling and analysis limitations must be reconciled through modification
and/or technical refinement of the program plan. The iterative nature of the develop-
ment process is described at appropriate points in the document.
The final section is a detailed review of sampling and analysis state-of-the-art
wherein both screening and detailed (specific) techniques are described. This section is
intended to allow the user to select specific methods for use. In addition, the physical and
chemical properties of compounds to be considered in selection of appropriate methods
are discussed. The latter aspect of the document is extremely important since in most
monitoring situations concentrations of certain compounds must be determined for which
specific methods have not been devised. Furthermore, combining several compounds into
one analysis will often be desirable from a cost standpoint (as opposed to monitoring each
compound by a specific method) and knowledge of chemical and physical properties is
necessary in this "method consolidation" process.
Figure 1 diagrams the topic areas of the document. This figure as well as the index
at the end of the document are provided to aid the reader in locating portions of the TAD
of specific interest.
-------
LIST OF ABBREVIATIONS
AFD — alkali flame detector (same as NPD)
ASTM — American Society for Testing and Materials
DFTPP — decafluorotriphenylphosphine
BCD — electron capture detector
EPA — U.S. Environmental Protection Agency
ESP — electrostatic precipitator
eV — electron volts
FED — flame ionization detector
FTIR — Fourier transform infrared spectrometry
FPD — flame photometric detector
GC — gas chromatography
HECD — Hall electrolytic conductivity detector
HPLC — high performance liquid chromatography
IR — infrared spectrometry
IS — internal standard
KV — kilovolts
LRMS — low resolution mass spectrometry
m^ — cubic meters
mg — milligrams
m/e — mass to charge ratio
MID — multiple ion detection
MS/MS — Tandem mass spectrometry (e.g., triple quadrupole mass spectrometry)
MS — mass spectrometry
NBS — National Bureau of Standards
NIOSH — National Institute of Occupational Safety and Health
NOX — oxides of nitrogen
NPD — nitrogen-phosphorus selective detector
PAH — polynuclear aromatic hydrocarbons
PCB — polychlorinated biphenyis
PCN — polychlorinated naphthalene
PID — photoionization detector
pKa — logarithm (base 10) of the dissociation constant for an acidic compound
pKb — logarithm (base 10) of the dissociation constant for a basic compound
ppbv — parts per billion on a volume basis
ppmv — parts per million on a volume basis
QA — quality assurance
RCRA — Resource Conservation and Recovery Act
SIM — selected ion monitoring
SRMs — standard reference materials
TAD — technical assistance document
TCDD — tetrachlorodibenzodioxin
TEA — thermal energy analyzer
TSCA — Toxic Substances Control Act
-------
UV — ultraviolet spectrometry
UV-VIS — ultraviolet-visible spectrometry
WCOT — wall coated open tubular capillary column
xn
-------
TABLES
Number Page
1 Organic Compounds Being Considered for Regulation Under
Clean Air Act Amendments 6
2 Technical Objectives Requiring Definition 10
3 Technical Information to be Used in Development of a
Monitoring Plan 15
4 Factors to be Considered in the Selection of Sampling and
Analysis Methods 20
5 Quality Assurance (QA) Activities to be Specified in
Program Plan 29
6 Sampling Techniques for Gas Phase Organic Components 39
7 Particle Sampling Approaches 45
8 Typical Commercially Available Screening Techniques for
Organics in Air 50
9 Colorimetric Screening Techniques for Determining Classes of
Organic Compounds 57
10 Commonly Used GC Stationary Phases 63
11 Common GC Detectors 64
12 Useful Dual GC Detector Combinations 67
13 Comparison of Reversed and Normal Phase HPLC 73
14 HPLC Detectors 75
15 Classification of Organic Compounds for Ambient Air
Monitoring Studies 77
16 Measured Concentrations of Organic Compounds in Urban Air 78
-------
TABLES
(Continued)
Number Page
17 Summary of Sampling and Analysis Methods for Organic
Compounds in Ambient Air 81
18 Summary of Sampling and Analysis Methods for Selected
Toxic Organic Compounds 89
19 Calibration Requirements for Sampling and Analysis
Instrumentation 97
20 Typical Sampling/Analysis Frequencies for QC Samples 99
-------
CONTENTS
Foreword ill
Preface IV
Abstract V
Figures Vl i i
Tables i X
Abbreviations xi
1. Introduction 1
2. Regulatory and Related Issues Concerning Toxic
Organic Monitoring 4
3. Guidelines for Development of a Monitoring Plan 8
Definition of Objectives 8
Compilation and Evaluation of Available Information 14
Selection of Sampling and Analysis Methods 18
Selection of Sampling Strategy 24
Specification of Quality Assurance Protocols 27
Definition of Data Reporting Format 35
Safety Considerations 35
4. Sampling and Analysis State of the Art 37
Overview of Sampling Methods 37
Overview of Analytical Methods 47
Methods for Specific Compounds and Compound Classes 74
Quality Assurance Procedures 94
References 103
Appendix 108
Topic Index 119
VII
-------
FIGURES
Number
1 Topical Flowchart for Technical Assistance Document 3
2 Flowchart for Development of a Monitoring Plan 9
3 Flowchart for Selection of Sampling and Analysis Methods .............. 21
4 Quality Assurance Organization 28
5 Typical Sampling Data Sheet 32
6 Chain of Custody Record 33
7 Block Diagram of a Typical Gas Chromatograph 60
8 Capillary and Packed Gas Chromatographic Columns 61
9 Approximate Gas Chromatographic Detector Ranges 65
10 Block Diagram of a Typical GC-MS System 69
11 Typical Configuration of an HPLC System 72
12 Method Validation Scheme 95
vm
-------
ABSTRACT
This project had as an objective the development of a Technical Assistance
Document (TAD) for sampling and analysis of toxic organic compounds in ambient air.
The primary users of this document are expected to be regional, state, and local
environmental protection personnel who are faced with the need to determine ambient air
quality for regulatory or information gathering purposes.
The TAD consists of the following four sections:
(1) Introduction
(2) Regulatory Issues Related to Toxic Organic Monitoring
(3) Guidelines for Development of a Monitoring Plan
(4) Sampling and Analysis State of the Art.
A topical index is included to assist the reader in locating the pertinent subject
areas within the document.
-------
-------
FOREWARD
Measurement and monitoring research efforts are designed to anticipate potential
environmental problems, to support regulatory actions by developing an in-depth under-
standing of the nature and processes that impact health and the ecology, to provide
innovative means of monitoring compliance with regulations, and to evaluate the
effectiveness of health and environmental protection efforts through the monitoring of
long-term trends. The Environmental Monitoring Systems Laboratory, Research Triangle
Park, North Carolina, has responsibility for: assessment of environmental monitoring
technology and systems; implementation of Agency-wide quality assurance programs for
air pollution measurement systems; and supplying technical support to other groups in the
Agency, including the Office of Air, Noise and Radiation, the Office of Toxic Substances,
and the Office of Enforcement.
Determination of toxic organic compounds in ambient air is a complex task, primari-
ly because of the wide variety of compounds of interest and the lack of standardized
sampling and analysis procedures. Prepared at the request of the Office of Air Quality
and Standards, this document provides overview information and technical guidance con-
cerning the determination of toxic organic compounds in ambient air. This information
should be of value to many federal, state, regional, and local environmental protection
agencies.
Thomas R. Hauser, Ph.D.
Director
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina
m
-------
PREFACE
The determination of toxic organic chemicals in ambient air is an area of concern
for many environmental protection personnel. There exist intense regulatory pressures to
gather data concerning human exposure to such chemicals, however, sampling and analysis
methodology for performing such measurements is not standardized (unlike inorganic and
"criteria" pollutants for which specific methods have been issued.) Consequently each
group performing such measurements must assess the methodology available and establish
appropriate sampling and analysis protocols.
This technical assistance document (TAD) has been written for the purpose of
providing such personnel with a basis for developing specific monitoring plans for toxic
organic chemicals. However, this document does not circumvent the need to have highly
skilled technical personnel organizing such monitoring efforts and the TAD should not be
used as a rigid, step-by-step guide for developing monitoring plans.
Individual readers will find only certain subject areas within the TAD to be of
interest and therefore a "Topical Index" has been provided to assist in locating relevant
information.
-------
EPA-600/4-83-027
June 1983
TECHNICAL ASSISTANCE DOCUMENT FOR
SAMPLING AND ANALYSIS OF TOXIC ORGANIC
COMPOUNDS IN AMBIENT AIR
by
R. M. Riggin
Battelle-Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Contract No. 68-02-3745 (WA-1)
EPA Project Officers:
L. J. Purdue
Quality Assurance Division
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
and
H. G. Richter
Air Management Technology Branch
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
.-born Ur
,, 60504
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
F:T,-:. onmental Protection Agency
11
-------
General Topic of Interest*
Regulatory
Issues
I
Monitoring
Strategies
1
Specific Information
Concerning Sampling and
Analysis Techniques
I
SECTION 2 SECTION 3
-Monitoring Objectives
pp. 8-14
-Sources of Monitoring
Information, pp. 14-18
-Selection of Sampling &
Analysis Methods, pp. 18-24
-Sampling Strategy (e.g.,
site selection), pp. 24-26
-Quality Assurance
Considerations, pp. 27-35
-Data Format, p. 35
-Safety, p. 35
-Sources of Sampling &
Analysis Methods, p. 18
-Sources of Monitoring
Data, p. 14
-M eteorological
Considerations, p. 17
-Method Performance Data
for Sampling & Analysis
Methods, p. 18
SECTION 4
-Sampling Methods,
Overview, pp. 37-47
-Analytical Methods,
Overview, pp. 47-74
-Screening Methods,
Field, pp. 49-56
-Screening Methods,
Laboratory, pp. 56-58
-Compound Specific
Methods (GC, HPLC,
GC/MS, etc.), pp. 58-74
-Compound Classes,
Definition, pp. 76-80
-Specific Sampling &
Analysis Methods, pp. 80-94
-Methods for Determining
Compounds Listed in
Table 1, pp. 88-94
-Quality Assurance, pp. 94-102
Figure 1. Topical flowchart for technical assistance document*
*Consult Index for specific topics.
-------
SECTION 2
REGULATORY AND RELATED ISSUES CONCERNING
TOXIC ORGANIC MONITORING
The primary motivations for conducting ambient air monitoring of toxic organics
are: (a) regulatory requirements, (b) emergency situations (i.e., chemical spills and fires)
and nuisance complaints, and (c) air pollution research activities. The types of activities
classified as "air pollution research activities" are listed in the last paragraph of this
section. Each of these topical areas have distinct requirements for ambient air
monitoring which are important to recognize.
The regulatory needs for ambient air monitoring are obviously diverse and
constantly being modified. At the federal level the regulations having the greatest
impact on ambient air monitoring are the Resource Conservation and Recovery Act
(RCRA), the Clean Air Act, and the Toxic Substances Control Act (TSCA). Although the
Occupational Safety and Health Act requires monitoring of workplace air, this situation is
quite distinct from ambient air monitoring and is not considered in this discussion.
RCRA governs the disposal of hazardous waste and a significant impact of this
legislation is the need to monitor toxic organics in ambient air in the vicinity of landfills,
abandoned dumpsites, etc. Proposed regulations require landfill operators to establish an
ambient air monitoring system, although this requirement may be modified or eliminated
in the future. A more common and certain monitoring activity related to RCRA is the
need to monitor a wide variety of toxic organics in the areas around abandoned dumpsites.
Such monitoring programs are necessary in order to determine whether such facilities
represent a hazard to persons living in the area and are especially critical during cleanup
or related construction activities during which hazardous materials may be released to the
atmosphere.
The existing portions of the Clean Air Act most directly affecting ambient air
monitoring for toxic organics are Sections 108-110, 111, and 112. Section 111 provides
for the control of new and existing sources of non-criteria pollutants which may reason-
ably be anticipated to endanger public health. To date regulations have been proposed for
several chlorinated solvents, including trichlorethylene, perchloroethylerie, methylene
chloride, methyl chloroform, and chlorofluorocarbon 113 using Section 111.
Section 112 of the Act is considered to be more directly applicable to toxic
organics. This section requires EPA "to publish, and from time to time revise, a list of air
pollutants which cause or contribute to air pollution which may reasonably be anticipated
to result in an increase in mortality or an increase in serious irreversible, or incapaci-
tating reversible, illness". To date benzene and vinyl chloride are the only organic
-------
compounds listed as hazardous air pollutants under this section of the Act. The develop-
ment of a more complete listing of hazardous air pollutants is slowed by the lack of
reliable health effects data needed to establish casual relationships between specific
compounds and health impairment.
EPA is reviewing a number of organic chemicals to determine the need for regula-
tion. Among those with the highest priority is the group listed in Table 1. If regulated
or as part of the process of review, specific ambient air monitoring programs for these
chemicals may be undertaken.
Sections 108-110 of the Act are intended to apply to those pollutants which can be
reasonably anticipated to endanger the public's health or welfare and for which emission
to the air result from numerous and diverse sources. As such these sections could be
used in certain instances to establish an ambient air quality standard for a toxic
pollutant. Establishment of an ambient air quality standard would set in motion
development of state implementation plans (SIPs) for control of existing sources, and
new source performance standards under Section 111 for control of new sources. With
regards to SIPs one of the major focuses is ambient air monitoring. Minimum monitoring
requirements for SIPs are specified for the purpose of determining compliance with the
National Ambient Air Quality Standards (NAAQS) as well as determining the degree of
control required of emitting sources.
However, these are not the only purposes for conducting ambient air quality moni-
toring under the Clean Air Act. For example, monitoring may be performed for source
characterization studies in response to an episode situation or research related
activities.
TSCA gives EPA authority to secure information on all new and existing chemical
substances and to control any of these substances determined to cause unreasonable risk
to the public health or environment. While TSCA does not directly require ambient air
monitoring for regulatory purposes, it does require that information on environmental
behavior and fate as well as likely sources of environmental discharges and exposed
populations be made avilable. To this end EPA's Office of Toxic Substances and various
other agencies and industries have undertaken source, "fenceline", and ambient air moni-
toring programs for selected organic chemicals.
In addition to federal regulations, a variety of state and local regulations have
required the development of ambient air monitoring programs. Discussion of all such
regulations is beyond the scope of this document. Many of these regulations have been
the result of public concern over localized air pollution problems, especially in highly
industrialized areas. In many cases regional and local air monitoring programs have been
implemented to determine the scope of the problem as well as the need for further or
modified regulations. High instances of cancer and other chronic illness in certain
geographical regions has led to the implementation of detailed monitoring programs for
organic pollutants in air, wherein trace (ppb and below) levels of a wide variety of
chemicals are determined over a wide area.
Emergency response activities are primarily related to chemical spills (e.g., during
rail or highway transport of materials) and fires at chemical handling facilities. In such
situations there exists an immediate need to conduct ambient air monitoring in the
-------
Table 1. Organic Compounds Being Considered for Regulation
Under Clean Air Act Amendments
Acetaldehyde
Acrolein
Acrylonitrile
Allyl chloride
Benzyl chloride
Carbon tetrachloride
Chlorobenzene
Chloroform
Chloroprene
o-,m-,p-Cresol
p-Dichlorobenzene
Dimethyl nitrosamine
Dioxin (2,3,7,8 tetra-
chlorodibenzodioxin)
Epichlorohydrin
Ethylene dichloride
Ethylene oxide
Formaldehyde
Hexachlorocyclopentadiene
Maleic anhydride
Methyl chloroform (1,1,1 trichloroethane)
Methylene chloride (dichloromethane)
Nitrobenzene
Ni t r oso mor pholine
Perchloroethylene (tetrachloroethylene)
Phenol
Phosgene
Polychlorinated biphenyls
Propylene oxide
Toluene
Trichloroethylene
Vinylidene chloride (1,1-dichloroethylene)
o-,m-,p-Xylene
-------
vicinity of the activity as well as in remote areas where transport of hazardous
compounds is possible. Emergency monitoring situations differ from most regulatory
related activities in several important ways. Since acute rather than chronic effects are
of concern, the required detection sensitivity is less stringent. However, in many cases
the exact compounds of interest may not be immediately known and hence a broad range
of monitoring techniques must be used to ensure detection of the toxic compounds. In
addition, the regulatory related monitoring activities are generally not time critical and
development of a monitoring strategy can be accomplished over a several day or week
period, whereas emergency situations obviously require immediate response without
development of a detailed monitoring strategy. These differences in scope of the various
monitoring situations will be discussed in more detail at appropriate points in the next
section.
Nuisance complaints arising from the general public are primarily related to noxious
odors, eye irriation, or, in some cases, more serious illness. Monitoring activities in this
regard are similar to emergency response activities in that the chemicals of interest are
often not known. However, the levels of interest may be a great deal lower in this situ-
ation due to the highly sensitive nature of the human senses.
Air pollution research activities require a wide variety of monitoring strategies. In
general such research activities are concerned with one of the following topics:
• Identification of toxic organics
• Validation and refinement of sampling and analysis techniques
for toxic organics
• Development and validation of air pollution models
(e.g., dispersion, transport, etc.)
• Fate and transport studies of specific organic compounds
• Correlation of human health effects data with ambient air
monitoring data.
In general such research activities are characterized by (a) a great deal of time and effort
in development of a monitoring plan and (b) the need to detect trace levels of organic
compounds, if present. While many research activities are directed towards the develop-
ment and support of proposed regulations, they are distinguished from regulatory moni-
toring activities by having a more general scope and greater flexibility.
-------
SECTION 3
GUIDELINES FOR DEVELOPMENT OF A
MONITORING PLAN
Probably the most critical stage of an air monitoring program is the development of
an appropriate monitoring plan. Such planning activities must be initiated well in advance
of the desired completion time. The important steps in this development process are
diagramed in Figure 2. While these steps can be listed sequentially, the development
process is iterative and cannot be approached in a rigid stepwise format. Consequently,
the following sections in which the development steps are discussed in detail should be
viewed as an integral approach rather than a series of separable efforts.
DEFINITION OF OBJECTIVES
While definition of the specific objectives of a monitoring program is logically a
necessary step, the importance of this stage of the development process is frequently
underestimated. In most cases two levels of objectives, both of considerable importance,
must be defined as illustrated in Figure 2. Broad, general objectives (e.g., to detect and
identify toxic organics released from a particular type of process) normally are defined
prior to initiation of a monitoring effort. These objectives are usually nontechnical and
serve as guidelines for the development of specific technical objectives. At a minimum
the technical objectives must define the aspects of the program listed in Table 2.
In general, the overall objectives will not automatically lead to the definition of a
single set of technical objectives, i.e., several possible approaches must be considered.
Consequently, the specific objectives stated by the technical personnel must be examined
carefully by those persons responsible for defining the general program objectives to
ensure that both parties understand the goals of the program as well as any technical
limitations in meeting those goals. It must be recognized that at the present time sensi-
tive methods do not exist for many organic compounds. Therefore, detection of all
compounds of interest at the desired level may not be feasible. Regulatory requirements
often precede technical feasibility.
An obvious first technical objective to be defined is the site or source of interest.
In most cases the site of interest will be well defined (e.g., an abandoned dump, chemical
spill, etc.). However, in some cases the definition of a specific site will require a great
deal of effort and will not be accomplished until the detailed sampling strategy is
developed, as discussed later in this section of the TAD. For example, if one has the
general objective of identifying or studying transport characteristics of toxic organics in a
-------
Outline (Define)
general objectives and restraints
Review by technical
personnel
Redefine
objectives
if required
Define specific
technical objectives
Review by
management
personnel
Compile and evaluate
existing information
Reconcile
conflicts between
sampling and
analysis state-
of-the-art and
project objectives
Select sampling and analysis
methods
Select sampling strategy
Specify QA and safety procedures
Specify data requirements
Figure 2. Flowchart for development of a monitoring plan
-------
Table 2. Technical Objectives Requiring Definition
Site of interest
Compounds of interest
Qualitative specificity (class or compound specificity)
Method performance parameters
Physical state of interest (gas, particle, or total)
Cost restraints
Data quality objectives (requirements)
10
-------
particular city, the definition of the sites to be monitored will properly await the
selection of a detailed sampling strategy.
The compound(s) of interest in a particular program appear to be an easily defined
objective. However, one must recognize that selection of too broad a range of compounds
will lead to excessive cost whereas selection of too few compounds may result in non-
attainment of the general program objectives. In most cases the selection of a specific
set of target compounds will represent a compromise between technical feasibility and
environmental significance. For example, if a particular compound is emitted from a
source in the region and the program objective is to evaluate the environmental impact of
that source, then monitoring of this compound is a requirement regardless of technical
limitations.
On the other hand, if a program has the objective of describing transport of toxic
organics over a wide region, selection of the compounds of interest must heavily consider
technical feasibility. In this case, methods capable of detecting a wide range of
compounds simultaneously should be employed and compounds requiring specialized pro-
cedures should be included only if the environmental significance of the compounds in the
region is substantial.
A further consideration is the stability of a compound in the atmosphere. If a
compound is rapidly degraded in the atmosphere, through hydrolysis, oxidation, photolysis,
or other processes, then one must consider the option of monitoring degradation products,
rather than the compound itself. Likewise one must consider the possibility of formation
of a particular compound of interest in the atmosphere. For example, formaldehyde is a
common oxidation product of olefinic compounds in the atmosphere. Therefore, if one
wishes to describe the impact of a source emitting formaldehyde, the formation rate of
formaldehyde from olefinic materials must also be considered.
In many cases, the objectives of a program can be attained using class specific
screening procedures rather than more expensive compound specific methods. In many
cases class specific or nonspecific (total organic) techniques can be conducted in the field,
often in "real time". Consequently, the use of a two-tiered (or "pilot" study) approach is
often advisable, wherein these procedures can be used to rapidly define emission "hot
spots", approximate concentrations, and types of compounds emitted. Subsequently,
appropriate compound specific techniques can be employed to gain more detailed
information.
The two-tiered monitoring approach is a simple example of using two sets of techni-
cal objectives to meet a single set of general overall objectives. There are many other
cases where overall objectives will be best attained through the sequential accomplish-
ment of specific technical objectives. This approach allows the knowledge gained in one
monitoring effort to facilitate planning of subsequent efforts. In many cases, this
sequential approach may be the only viable alternative due to information gaps or techni-
cal limitations.
Method performance characteristics should be carefully considered in the planning
stage since these characteristics have a very profound effect on the cost and probable
success of a project. The specified ranges of the performance characteristics should be
compatible with the intended use of the data. In certain cases, regulatory requirements
may dictate the use of a specific sampling and analysis method. In such cases, project
11
-------
personnel should still evaluate the method characteristics and state any limitations
impacting the success of the project.
The major method characteristics of concern are:
Detection limit
Accuracy (absolute and relative)
Precision
Interferences (selectivity)
Analysis time.
Obviously the more stringently one defines these requirements, the more costly a program
becomes, often without a direct increase in useful information. Conversely, if one does
not specify sufficiently rigid performance criteria, the data obtained may be of little
value. Unfortunately, these performance goals are difficult to define because of the large
number of determining variables (e.g., sampling location, sampling time, source compo-
sition, etc.). In addition, some of these parameters (e.g., accuracy) are difficult to
determine experimentally because the wide variation in source characteristics can influ-
ence these parameters. Accuracy is particularly difficult to assess because of the lack of
organic Standard Reference Materials. Consequently, these parameters should be initially
defined as guideline "ranges" and should be rigidly defined only in cases where the exact
requirements are known.
Required detection limits will vary considerably depending on the compound of
interest, the sampling location relative to the source, meteorological conditions, and
available sampling time. Detection limits should always be defined in terms of component
concentration in air (yg/m-* or ppbv), rather than a method detection limit (e.g., in micro-
grams), since the latter value is subject to a great deal of interpretation (e.g., sampling
time, etc.). In general, the range of detection limits required for toxic organics will vary
between 0.1 ppbv and 10 ppmv. Detection in the 0.1-10 ppbv range will be required for
ambient air studies where sites are remote from emission sources, whereas 1-10 ppmv
detection limits may be sufficient for monitoring in the vicinity of chemical spills, dump-
sites, and other concentrated sources. Obviously, highly toxic compounds will require
more stringent detection limits in most cases.
Certain classes of compounds (e.g., halogenated compounds) can be detected at very
low levels because of unique characteristics and the availability of special detectors.
Consequently, one typically will obtain data on these compounds at low levels irrespective
of the project requirements. However, alternative approaches in such cases include the
use of shorter sampling periods or more general procedures capable of detecting other
compounds of interest. These alternatives may reduce costs without losing the desired
information. Therefore, one should not set detection limit criteria based solely on the
available detection limit since flexibility is lost in this process.
Some compound classes (e.g., highly polar materials such as alcohols and acids)
cannot be detected at low levels. For these compounds one must decide whether or not
the program objectives can be met using available analytical methods, toxicological
properties being the primary consideration. If not, then one is left with the difficult
choice of either (a) undertaking a frequently costly development program to improve the
methodology, (b) reducing the program requirements to allow partial fulfillment of the
objectives, or (c) not undertaking the program at all.
12
-------
Accuracy and precision requirements are normally more flexible than detection
limit requirements and usually do not result in a major revision of program objectives.
However, several aspects of the accuracy and precision requirement need to be addressed.
The required accuracy is defined primarily by the method of data comparison. If data
collected on one program are to be compared to data collected on another program,
perhaps conducted at a different time, then absolute accuracy is important. In other
words, one should know the accuracy for each program.
However, if one simply desires to know the ratio of component concentration at two
sites or at the same site at different times of the day then relative accuracy may be more
important than absolute accuracy. In this case, the only requirement is that the accuracy
be the same for each data point. The relative accuracy concept is especially useful when
using class specific screening techniques since the absolute accuracy will be known only if
the specific compounds giving rise to the response are known, whereas the relative accu-
racy is known so long as the relative proportion of the various compounds being detected
is constant. In a strict sense absolute accuracy for a method can be determined only when
a suitable NBS (or NBS traceable) standard is available and can be analyzed under con-
ditions duplicating the actual sampling and analysis progress.
Precision requirements are somewhat related to accuracy requirements since for a
given atmospheric concentration the measured quantity is a function of both accuracy and
precision. If one desires a certain degree of confidence that a measured signal is repre-
sentative, then an appropriate number of replicate determinations must be made. The
more precise the method, the fewer measurements need be made to attain a given degree
of confidence in the data. Precision and accuracy will both become poorer as analyte
concentrations approach the detection limit.
Specificity, degree of interference by other materials, for a method must be con-
sidered in the definition of project objectives since this aspect influences data accuracy
as well as the range of compounds to be evaluated. If large quantities of innocuous
compounds are present at the sampling location the selection of compounds of interest, as
well as sampling and analysis methods, should be based on circumventing potential inter-
ference problems whenever possible. This problem often limits usefulness of nonspecific
screening procedures since the response from background components may overwhelm any
response due to trace levels of toxic materials. If such interferences cannot be avoided
by judicious selective of the compound of interest, then alternative sampling and analysis
procedures which minimize the interference problem must be selected. In any case,
compounds which represent potential interferences in a given program should be listed, in
addition to the compounds of interest to aid in the selection of appropriate sampling and
analysis techniques.
A frequently ignored program objective is the physical state of interest. Organic
compounds can be present in the gas phase or bound to various types of particles (e.g.,
dust, flyash, water droplets, etc.). Relatively nonvolatile materials such as
benzo(a)pyrene exist predominately in the particle phase at ambient temperatures.
However, most compounds are distributed partially into each phase and this distribution is
influenced by atmospheric conditions. Furthermore, the distribution is altered by most
sampling techniques to some extent and hence the data obtained are influenced by the
sampling approach selected.
In most cases, the total concentration (gas as well as particle bound) of a component
is sufficient information. However, if the particle bound concentration is of interest
13
-------
(e.g., for inhalation toxicology considerations) one must recognize that the data will be
influenced heavily by the choice of sampling technique.
A final, and often overriding program objective (restraint) is the allowable cost.
Frequently, the cost is specified, within limits, prior to any other specific program
objectives. While this state of affairs must be accepted, one should not allow such limi-
tations to be used as an excuse to collect low quality data. If the specific program
objectives cannot define a technically feasible and scientifically valid program, then a
redefinition of scope, an expansion of available funding, or termination of the program are
the only valid alternatives. The impact of limitations recognized by technical project
personnel must be clearly and openly stated to appropriate management personnel in order
to avoid implementation of an ill-conceived monitoring program.
COMPILATION AND EVALUATION OF
AVAILABLE INFORMATION
The efficiency with which an air monitoring program can be developed and imple-
mented is highly dependent on the extent and quality of technical information available
during the development phase of the program. In most cases this information will be used
in an iterative fashion to develop specific technical objectives for the monitoring program
(as depicted in Figure 2), since such objectives must be realistic and achievable.
Information most useful in this process can be roughly categorized into the areas listed in
Table 3.
Probably the most useful information is data from previous air monitoring programs
in the region. Although such information is seldom directly relevant (e.g., due to
differences in the compounds monitored, sites used, or other program objectives) some
aspects of the data normally will be useful. For example, total hydrocarbon data from the
region may be useful in predicting sites and/or times having maximal pollutant concen-
tration. Data for a compound of interest may be available from a site similar to the one
of interest, thereby giving an "order of magnitude" estimate of pollutant concentration to
be anticipated.
Additional useful information which could be gained from such studies includes
compound stability in air under various conditions, degradation products formed, other
compounds present in conjunction with the compound of interest, phase distribution of the
compound in ambient air, and performance data for sampling and analysis methods of
interest. As with all sources of information, critical evaluation is necessary in order to
determine the quality and relevance of the data reported.
The acquisition of air monitoring data is not particularly straightforward because of
the large number of sources of such data and the wide variety of program objectives.
Some primary sources of such data include:
4 U.S. Environmental Protection Agency,
Research Triangle Park, N.C.
• Regional offices of the U.S. EPA
• State and local environmental protection (or equivalent) offices
• Contractors and universities performing work through contracts or
grants with EPA or related agencies
• Peer review journal articles and government reports.
14
-------
Table 3. Technical Information to be Used in Development
of a Monitoring Plan
Previous Air Monitoring Data
Composition of Emission Sources in the Region
(especially organic compound sources)
Meteorological Conditions
Sampling and Analysis Methods and Performance Data
for Compounds of Interest
15
-------
A wide variety of other government and related research organizations may have
available data in specialized areas (e.g., Electric Power Research Institute, Palo Alto,
CA; Department of Energy; National Oceanic and Atmospheric Administration).
Although a few data compilations are available (71), no comprehensive source of
monitoring data exists. Consequently, the best approach towards obtaining relevant data
is to contact several sources (local as well as regional or national) and to describe the
type of monitoring information desired. Since data specifically relevant to the program
are probably not available, one should seek to obtain any potentially useful monitoring
information. In many cases references within these documents may lead to additional
useful data. Table 16 provides some measurement data for selected toxic organics in
urban air.
Critical evaluation of available data should address the following points:
• Documentation of sampling and analysis techniques in terms of detection
limit, precision and accuracy, specificity and analysis speed
• Sampling strategy in terms of reference points and spatial and
temporal resolution
• Documentation of atmospheric conditions
• Influence of emission sources in the region.
The purpose of this evaluation should be to assess the relevance of the data to the
program being developed; therefore aspects of the data having no relevance to that
program need not be evaluated.
Since most ambient air monitoring programs will be impacted by emission sources in
the region, data concerning the composition of such emission sources are extremely
useful. Programs having the objective of assessing the impact of a particular source on
the environment obviously require knowledge of the source composition, whereas other
programs need source emission information to avoid biasing the data as a result of con-
tamination by the source. Typical sources of concern include:
Chemical production, processing or handling facilities
Fuel sources (gasoline stations, storage tanks, etc.)
Mobile sources (automobiles, etc.)
Chemical waste landfills, lagoons, etc.
Miscellaneous sources (dry cleaners, sewers, residential
heaters, etc.).
Information concerning emission components from chemical production or related
manufacturing facilities may be available from the plant operator. Discharge permits
represent another source of information. Since such facilities differ widely, depending on
the chemical processes contributing to the emissions, one can seldom rely on information
available from similar facilities to any great extent. Furthermore if the program ob-
jective is to assess the impact of the source on regional air quality, simultaneous
measurement of the source emissions and the ambient air is highly desirable since the
composition of a single source can vary considerably with time. Knowledge of production
schedules and similar factors is extremely useful in predicting the degree of variability.
In the case of fuel sources, mobile sources, dry cleaners, etc. the qualitative nature
of the emissions can be predicted fairly well since these sources have been extensively
16
-------
characterized and in the case of drycleaners only one or two compounds (e.g.,
perchloroethylene or Stoddard solvent) are in use. Consequently, one can predict in
advance whether or not these sources will interfere with the ambient air monitoring
effort. In general, a useful monitoring approach assuming these source contributions are
not being evaluated, is to locate sampling stations at sites remote and/or upwind from
such sources, using screening techniques such as flame ionization or photoionization
detection to locate the points of high source contribution. This approach requires that
only the location of the source be determined, not the detailed composition. However, if
the contribution of such sources to regional air pollution is being evaluated, knowledge of
the detailed source composition will be required and should be made part of the overall
monitoring program.
Knowledge of meteorological conditions is important to ambient air monitoring
programs in several regards. The impact of a particular source on regional air quality will
be highly influenced by meteorological conditions since dilution rates, transport rates, and
compound stability are dependent on these conditions. Overall regional air quality,
representing the summation of source contributions, will be influenced by these conditions
in the same manner.
A detailed discussion of meteorological considerations in air quality monitoring is
given in the Air Pollution series by Sternal) and is beyond the scope of this document.
Meteorological parameters of primary concern include the following:
Wind speed and direction
Temperature
Humidity
Barometric pressure
Solar radiation intensity
Precipitation
Mixing height.
While these data should be collected during the course of the monitoring program, prior
knowledge of typical conditions can be extremely useful in selecting sampling points.
Regional meteorological data can be obtained from the National Weather Service in
the locality of interest. However, one must recognize that localized meteorological
conditions, not apparent from the National Weather Service data can exist. For example,
wind speed and turbulance can be affected by obstructions such as tall buildings in the
area. Fog may occur selectively in marshes or in hilly areas. Surveillance of the sampling
site and discussions with persons familiar with the area can be useful in selecting sampling
sites yielding favorable meteorological conditions.
Availability of suitable sampling and analysis techniques is often the limiting factor
in the development of a comprehensive air monitoring program, in spite of the large
number of techniques which exist. Since virtually every air monitoring program has a
different list of compounds of interest, a variety of techniques will be required to cover
this range of components, resulting in a significant cost impact on the program. Further-
more, unique requirements in terms of specificity, detection limit, analysis time, and
other performance characteristics may require modifications to existing methods.
Consequently, a highly desirable approach is to develop program objectives which are in
consonance with the capabilities of available sampling and analysis techniques wherever
possible. In order to use this approach without severely limiting the program objectives
17
-------
one must have available as much relevant information on sampling and analysis techniques
as possible.
Primary sources of sampling and analysis methods include the following:
• National Institute for Occupational Safety and Health (NIOSH)
Methods^)
• American Society for Testing and Materials (ASTM)
Methods^3)
• Methods of Air Sampling and Analysis'^)
• Peer review and journals and government reports
• EPA Reference Methods (Federal Register).
Methods for a number of toxic organic compounds of particular concern are presented in
Section 4. Contacts with various persons active in air quality monitoring is also an
effective means of gathering recent information on sampling and analysis methods.
Unfortunately, no compendia of analytical methods for toxic organics at low ambient
concentrations have been developed. The NIOSH and many ASTM methods are designed
for workplace or source concentrations and may require substantial modification to detect
ambient concentrations.
Available sampling and analysis methods should be evaluated with regard to the
following performance characteristics:
Selectivity (range of compounds detected and potential interferences)
Detection limit
Precision and accuracy (often unknown)
Analysis time
Cost
Quality of performance evaluation data.
The methods should be evaluated from two standpoints; (a) the degree to which method
performance has been documented under conditions similar to those expected and (b) the
adequacy of the performance characteristics for achieving the program objectives.
If method performance is not well documented one must decide whether or not to
include further documentation as part of the monitoring program with the resultant
increase in data quality as well as cost. A similar decision must be made when existing
methods are slightly modified (e.g., extended to similar compounds, etc.). One must
recognize that some assumptions concerning method performance will inherently be part
of any monitoring program due to cost and time limitations (i.e., documentation of all
performance character-istics under each set of monitoring conditions is prohibitively time
consuming and expensive). Consequently, those assumptions which are made should be
based on strong technical logic and should be carefully documented.
SELECTION OF SAMPLING AND ANALYSIS METHODS
Once a set of program objectives, including specification of the compounds of
interest, have been developed and information concerning sampling and analysis tech-
niques has been evaluated, the next logical step is selection of the sampling and analysis
procedures to be used in the monitoring program. Factors to be considered in selecting
18
-------
sampling and analysis methods are outlined in Table 4. A useful stepwise approach to this
selection process is depicted in Figure 3.
The important chemical and physical properties to be considered include thermal
stability, volatility, polarity, ionic character, phase distibution, and chemical composition.
The manner in which these properties determine the suitability of analytical as well as
sampling procedures is discussed in detail in Section 4.
Volatility, thermal stability, polarity, and ionic character determine the extent to
which commonly used gas chromatographic procedures are suitable for the compounds.
These properties also must be considered in the selection of sample preparation pro-
cedures since solvent extraction and concentration procedures may result in loss or degra-
dation of volatile or thermally labile compounds.
Phase distribution (gas/particle) of a compound is obviously dependent on the other
physical and chemical properties but must be considered separately since sample prepa-
ration procedures for particle bound compounds will usually be different than for gas
phase compounds. Compounds which are likely to be present in both phases require
special consideration. Sampling methods can sometimes be employed to drive the equi-
librium into one phase or the other if only the total concentration is to be determined
(e.g., by using a heated filter).
Chemical composition, specifically the presence of heteroatoms, is often the most
important factor in assigning compounds to a particular analytical group. Specialized
chromatographic detection systems tend to be specific for a particular class of organic
compounds and selective colorimetric procedures are available for certain classes of
compounds as well. However, one must avoid oversimplication of the grouping process by
automatically placing compounds having a particular functional group in a single class.
Individual compounds may have unique properties making placement in another class more
suitable. For example, gas chromatography with electron capture detection (GC/EC) is
sensitive and selective for detection of many halogenated compounds, but does not
respond well to vinyl chloride, chlorobenzene, or 1,2-dichloroethane. Hence these
compounds should not be grouped with other halogenated compounds targeted for GC/EC
analysis.
Familiarity with the analytical procedures available is an obvious requirement in the
compound grouping process. One must recognize that analytical procedures include both
sample preparation and compound determination steps, each of which must be considered
in the assignment of a compound to a particular analytical group. Simply specifying gas
chromatography/mass spectrometry (GC/MS), for example, as the analytical technique
does not define an analytical procedure.
The anticipated relative concentrations of the various compounds of interest should
be considered in the grouping process. Compounds present at high levels may be candi-
dates for direct analysis in the field (e.g., by portable GC) since sample concentration
techniques might not be necessary. At the other extreme trace components may require
sample concentration as well as cleanup steps to obtain sufficient sensitivity and
selectivity.
A final consideration in the grouping process is the relative importance of the
various compounds to the program objectives. While a formal priority ranking of the
compounds normally is not required one should consider relative importance in the
19
-------
Table 4. Factors to be Considered in the Selection of
Sampling and Analysis Methods
Physical and chemical properties of compounds
Relative and absolute concentrations of compounds
Relative importance of various compounds to program
objectives
Method performance characteristics
Potential interferences present at site
Time resolution requirements
Cost restraints
20
-------
Revise Compound
Groups and/or
Objective as
Required
Subdivide Compounds into
Logical Groups Based on
Physical and Chemical
Properties, for Analytical
Purposes
Select Appropriate Analytical
Procedures Based on
Required Performance
Characteristics
Select Appropriate Sampling
Procedures Based on
Analytical Requirements
Specify Sampling and Analysis Procedures
in Terms of Standard Operating
Procedures (SOPs)
Figure 3. Flowchart for selection of sampling and analysis methods
21
-------
following qualitative manner. Compounds of special concern, either because of high
toxicity or known emission sources, should be grouped so as to optimize analytical
performance for these compounds. In certain cases this process may include placing
compounds of high priority in unique groups, targeted for specialized, specific analysis
techniques. Compounds of lesser interest should be placed in the appropriate groups
selected for optimal detection of the compounds of primary concern, usually with some
comprise in analytical performance.
The selection of analysis procedures is largely guided by the compound grouping
process in which the available analytical techniques were heavily considered. However,
the definition of analytical procedures in terms of specific operating parameters should
normally await completion of the grouping process.
Relatively few standardized analytical procedures for toxic organics are available.
Consequently, program technical personnel will normally have the responsibility of
defining specific sample preparation procedures and chromatographic operating con-
ditions, as well as writing the detailed standard operating procedures for the monitoring
program. Factors to be considered in this selection process include required detection
limit, specificity, accuracy, precision, analysis time, and cost, all of which were con-
sidered in the development of project objectives. The task of specifying standard oper-
ating procedures for analytical methods must be accomplished in conjunction with the
specification of sampling protocols, since detection limits and other method performance
characteristics are a function of both steps.
Accuracy, precision, analysis time, and cost are determined by the choice of
sampling and analysis techniques and in many cases, cannot be substantially improved by
refinement of the methods. Consequently one must weigh these factors carefully during
the method selection process to ensure that the program requirements are achievable
using the technique selected. Method specificity can be altered to some extent by the
addition of cleanup steps to the analytical methods but is still highly dependent on the
type of detection system used.
Detection limits, on the other hand, can be substantially adjusted by the selection of
specific operating parameters within the sampling and analysis methods. Parameters
contributing to the detection limit include:
• Detector sensitivity (minimum quantity of material introduced into the
instrument which gives a detectable response)
• Proportion of sample introduced into the instrument
• Volume of sample collected
• Recovery of component through the entire sampling and
analysis procedure.
These parameters can be related to the detection limit by the following equation:
DL
Ps (%) x Vs (m) x R
22
-------
where:
SD = Detection sensitivity
Ps = Percentage of total sample introduced into the detection system
Vs = Volume of sample
R = Percent recovery.
This equation is useful as a guide for the selection of appropriate sampling and
analysis parameters. However, one must realize that many of these parameters are in
turn dependent on other factors. Detection sensitivity is dependent on the concentration
of background components in the sample (i.e., practical detection sensitivity is limited by
interference from background components, not by absolute instrument sensitivity). The
proportion of sample which can be introduced into the instrument is also dependent on the
quantity of background components present and can be substantially increased by the use
of more effective cleanup techniques. Component recovery, on the other hand, can be
adversely affected by the incorporation of additional cleanup techniques into the method.
Because of the limited number of sampling methods available, the selection of
specific sampling protocols is best accomplished after the definition of analytical require-
ments, based on compound groups of interest, etc. However, the specification of detailed
analytical procedures obviously requires knowledge of the sampling scheme (e.g., form of
sample presented for analysis, sampling volumes, etc.).
Primary factors to be considered in the selection of sampling procedures include
phase distribution, stability, and time resolution requirements. Compounds entirely con-
tained in the particle phase can be readily sampled using filtration whereas gas phase
components require more elaborate techniques such as resin sorption, impinger collection
or cryogenic trapping. Stability of the components of interest during the sampling process
must be considered. If necessary, samples can be removed from the sampling apparatus
and stabilized (e.g., frozen) to prevent degradation. Stability can also be affected by the
collection of reactive components components (SO^, NOX, Cl^, etc.). Therefore sampling
techniques which minimize the collection of such components should be selected whenever
possible.
Time resolution requirements may prevent the use of certain techniques which
cannot acquire sufficient sample in the required time period. If extremely fine time
resolution is required, continuous monitors (e.g. flame ionization or photoionization
detectors) may be the only acceptable alternative. Techniques of this type are usually not
specific for particular organic components, and this limitation must be weighed against
the importance of time resolution.
After sampling and analyis procedures have been selected, the specific details of
these methods (e.g., sampling volumes, cleanup steps, etc.) can be selected, considering
the detection limit requirements in particular. Subsequently one must examine the
program objectives to determine the capability of the sampling and analysis scheme to
accomplish these goals. Once any differences between sampling and analysis capabilities
and program objectives have been resolved, the sampling and analysis procedures should
be written into detailed "standard operating procedures" as a final step in the method
specification process.
The preceding discussion has presented the important parameters to be considered in
the selection of appropriate sampling and analysis procedures. This discussion has
23
-------
necessarily centered on the complex interdependency of the various parameters to be
considered in the selection process. The summary provided below and diagrammed in
Figure 3, will provide the reader with a workable approach to the method selection
problem.
The primary steps in the selection process include:
• Subdivide compounds into logical groups based on similar analytical
characteristics. This process will result in a preliminary selection
of generic analysis techniques, since analytical techniques must be
considered in the grouping process.
• Specify detailed analytical procedures based on the guidance provided
by the grouping process, considering accuracy, precision, time, and
cost requirements. Incorporate additional cleanup schemes based on
selectivity requirements, if necessary.
• Select sampling procedures based on analytical requirements and
physical properties of the compounds.
• Specify sampling and analysis parameters (sample volumes, final
extract volumes, etc.) based on detection limit requirements.
• Reconcile conflicts between program objectives and sampling and analysis
capabilities. Redefine objectives and/or undertake procedure development
efforts as required.
• Write detailed "standard operating procedures" for the sampling and
analysis methods.
SELECTION OF SAMPLING STRATEGY
Selection of a sampling strategy capable of meeting the program objectives is vital
to the success of any monitoring program. Parameters which must be specific include:
• Sampling locations
• Numbers, sampling volumes, and time points of samples to be taken
at each location
• Meteorological parameters to be monitored and dependency of sampling
strategy on these parameters.
Obviously the development of a sampling strategy can be extremely simple or
extremely complex, depending on the program objectives. Programs involving characteri-
zation of the pollutant contribution from a point source tend to be more straightforward
whereas programs investigating fate and transport characteristics of components from
diverse sources require more elaborate sampling strategies. An excellent discussion of
the factors to be considered in developing a sampling strategy is provided in the Air
Pollution series by Stern(5).
The selection of an optimal sampling siting plan must take into account the
following factors:
24
-------
• Locations of stationary as well as mobile sources
• Transport characteristics of pollutants from these sources and influences
of meteorology and topography on these characteristics
• Spatial resolution required to meet program objectives
• Availability of space and utilities for operating sampling equipment
at potential sites.
Relevant information on optimum site selection criteria are available in various
EPA documents^"'. One should consult such documents in developing a siting plan.
The necessity for locating stationary and mobile sources prior to site selection has
been previously discussed and should be readily apparent. One must not only have
knowledge of the existance of these sources but also the trajectory of emitted pollutants
under the existing meteorological and topographical conditions. Avoiding sample con-
tamination by these sources or, conversely, accurately measuring the pollutant contri-
bution of these sources may require continuous monitoring of meteorological parameters
and adjustment of sampling location or timing accordingly. In some cases the siting
process can be aided through the use of air quality models, in particular disperson models
which predict concentration profiles for pollutants emitted from point sources under
various meteorological conditions. The placement of reference (control) sites upwind of
known emission sources is included in most monitoring schemes to determine the contri-
bution of those sources. Additionally wind direction must be continuously monitored in
these situations to ensure the validity of the reference samples.
Spatial resolution requirements obviously are highly dependent on the program
objectives. If one is attempting to estimate pollutant concentrations in the vicinity of a
point source (e.g., hazardous waste landfills) spatial resolution of a few tens or hundreds
of meters may be required whereas programs monitoring pollutant transport through an
urban area may require resolution of a few miles or tens of miles. In determining spatial
resolution requirements one must realize that a sampling site provides only the concen-
tration at a given point and that data point is used to represent the pollutant concen-
tration over the entire area between the sampling site and adjacent sampling sites.
Consequently factors which lead to spatially inhomogeneous pollutant concentrations must
be carefully considered in the siting process (i.e., spatial resolution requirements will be
more restrictive in areas of inhomogeneous concentrations). Such factors include low
wind velocities, presence of tall buildings and other obstacles, and source emission
variability.
In certain specialized cases vertical as well as horizontal resolution is required. In
most cases vertical sampling requires use of an aircraft acquiring samples at various
altitudes depending on air flow characteristics. Since this requirement is infrequently
encountered in toxic organic monitoring, further discussion in this document is not
warranted.
The use of mobile sampling facilities to provide greater flexibility (e.g., in response
to changes in wind direction, etc.) is an attractive option in many cases and can be a cost
effective means of providing greater spatial resolution. In many cases a continuous
screening monitor, such as a photoionization detector, may be used to define the need for,
and optimal placement of, fixed monitoring stations.
A final consideration in the siting process is the availability of space and utilities
for operating sampling equipment. This requirement is well recognized and does not
Z5
-------
require a great deal of emphasis in this document. Obviously the selection of sites should
be done with the knowledge of available locations and utilities. However, in cases where
utilities are not readily available portable utilities (e.g., generators, heaters, or air
conditioners) must be provided. In many cases these portable units represent a potential
pollution source and hence contamination of samples may result if their placement is not
carefully considered.
After selection of appropriate monitoring locations one must consider the sampling
requirements in terms of timing and numbers of samples. The reference mentioned
above^' provides a useful discussion of time resolution considerations. In many respects
time resolution and spatial resolution requirements place similar demands on the sampling
scheme. Continuous screening monitors can be used to determine time dependency of
pollutant concentrations, and hence define time resolution requirements more accurately.
As with spatial resolution, time resolution requirements may not be uniform for a given
program. Certain time periods may require finer time resolution than during other time
periods (e.g., related to traffic patterns, production schedules, etc.).
In order to achieve adequate time resolution one must be able to collect sufficient
sample in the allowed time. In some cases the sample quantity requirement may result in
the need to use multiple sampling devices at each sampling location, although modern
analytical instruments tend to provide sufficient sensitivity to obviate this situation.
The numbers of samples to be collected at each site/time point is obviously
dependent on the variability of the sampling and analysis methods, the precision require-
ments of the program, and cost limitations. In many cases replicate analysis of a single
sample, rather than replicate sampling will be employed, in which case only analytical
precision is determined. In other cases samples collected at different times and/or sites
may be treated as replicate samples due to homogeneous spatial and temporal distribution.
of pollutants, hence overcoming the need to take multiple samples at each site/time
point.
Sampling volume, specified in the previous section of this document may become a
variable in the monitoring strategy as well. If pollutant concentrations varying over a
wide range (not predictable in advance), or if pollutants of greatly differing concen-
trations are to be monitored, one should consider the possibility of collecting multiple
samples of varying volumes at each site/time point. In this maanner one has some
assurance that at least some of the samples will contain levels of analyte within the
operating range of the analytical instrument. This approach is vital in certain cases
where the entire sample is introduced into the analytical instrument (e.g., Tenax
resin/thermal desorption techniques) and is of lesser importance when a selectable
proportion of the sample is analyzed.
Meteorological parameters to be monitored usually will include temperature,
barometric pressure, wind speed, wind direction, relative humidity, and precipitation.
Wind speed and direction is probably the most important variable impacting the sampling
strategy. In many cases one may wish to sample only when the wind is in a certain
direction, due to favorable source transport characteristics. In some cases meteorlogical
parameters may impact upon spatial and temporal resolution requirements in a manner
such that variations in these parameters will change the sampling strategy with time.
26
-------
SPECIFICATION OF QUALITY ASSURANCE PROTOCOLS
Overview
The term quality assurance (QA) refers to an overall system designed to monitor,
document, and control the technical performance of a program. While the need for good
QA protocols is widely recognized, the design and implementation of these protocols is
frequently treated as a secondary part of the overall monitoring program. If the QA
protocols for a monitoring program are to serve a useful purpose, they must (a) be readily
implemented within the cost and time constraints of the program, (b) be appropraite in
terms of providing useful control of technical performance, and (c) be well understood by
the project personnel. Preparation of the QA plan for a monitoring program should be
undertaken after the sampling strategy and sampling and analysis methods have been
defined. This section of the TAD describes the important aspects of QA to be addressed
prior to implementation of the monitoring program. A more detailed discussion of
specific QA approaches for various sampling and analysis methods is given in Section 4.
A simplified view of an overall QA system is given in Figure 4. QA activities to be
specified in the program plan are given in Table 5. The QA management function
monitors and controls the various QA efforts including (a) design of the QA system prior
to program implementation, (b) sampling, (c) analysis, and (4) data reduction. The
functions of QA management as well as the specific QA requirements for sampling,
analysis, and data reduction are discussed in the following sections of the TAD.
A series of volumes entitled Quality Assurance Handbook for Air Pollution Measure-
ment Systems^'', available from EPA in Research Triangle Park, serves as a useful
detailed guidance document in the QA area. In particular Volume I -Principles, and
Volume n - Ambient Air Specific Methods may be useful in the field of toxic organic
monitoring. Specific guidance for preparation of QA plans is provided in another EPA
document™'. QA practices are also discussed in Methods of Air Sampling and Analysis^).
QA Management
The functions and responsibilities of QA management are a critical part of the
overall monitoring program. These functions and responsibilities are listed in Table 5.
Designing a QA system involves specification of QA requirements for each aspect of
the program, preparation of a program QA manual describing these requirements, and
specification of auditing and other control procedures for ensuring proper implementation
of the QA system. Although the design of the QA system for a specific program is the
responsibility of QA management, input from the technical personnel responsible for
implementing the program is required. Technical personnel must identify areas of tech-
nical performance requiring validation, documentation, or control, estimate costs for
implementation of specific quality control functions, and critically evaluate QA docu-
ments (e.g., the program QA manual) prepared by QA management. Document control
involves (a) specification of revision numbers and dates on QA manuals and related
materials, (b) inspection and storage of calibration and maintenance logbooks for field and
laboratory instrumentation, (c) inspection and control of laboratory record notebooks, and
(d) control of chain of custody forms documenting sample deposition.
QA management is responsible for the evaluation of QA data in a timely manner.
Failure to review the data immediately prevents implementation of timely corrective
action procedures and may result in poor data quality.
27
-------
QA
Management
QA
System
Design
Sampling
QA
Analytical
QA
Data
Reduction
QA
Figure 4. Quality assurance organization
28
-------
Table 5. Quality Assurance (QA) Activities to be
Specified in Program Plan
QA Management
QA System Design
Document Control
Data Evaluation and Storage
Audit Procedures
Corrective Action
QA Reports to Program Management
Training
Sampling QA
Site Selection
Instrument Calibration and Maintenance
Collection of Routine Quality Control Samples
Data Recording
Sample Labeling, Preservation, Storage and Transport
Chain of Custody Procedures
Analytical QA
Method Validation Requirements
Instrument Calibration and Maintenance
Quality Control Sample Analysis
Data Recording
Data Reduction QA
Merging Sampling and Analysis Data Files
Storage of Raw and Intermediate Data
Data Validation
29
-------
Audit procedures include both performance and system audits. A performance audit
involves introduction of a reference sample (blank, spike, standard reference material,
etc.) into the analysis system in a blind fashion with subsequent assessment of the data.
System audits involve a review of program documentation such as record notebooks, data
files, and instrument logbooks to assess whether or not the QA system is operating
properly.
In many cases data review or audit procedures will result in the need for corrective
action. Corrective action may involve repeating certain aspects of the work or simply
providing more detailed documentation for work already performed. In either case QA
management will be responsible for documenting the need for, type of, and imple-
mentation of corrective actions.
QA management is responsible for providing scheduled as well as nonscheduled
reports to program management. Scheduled reports include descriptions of the QA system
prior to program implementation, QA data reports, and audit reports. Unscheduled
reports generally describe corrective actions required and the impact of these actions on
the program.
A final responsibility of QA management is to provide training to technical
personnel. In particular, personnel need to be given a detailed view of the QA system and
their responsibilities for its implementation.
Sampling Quality Assurance
Aspects of sampling to be addressed in the QA plan are shown in Table 5. Site
selection considerations have been discussed in the section on sampling strategy develop-
ment. However, the QA plan should specify factors which could result in a modification
of the siting plan during the course of the monitoring effort (e.g., changes in source
location or characteristics) and provisions for documenting any such modifications.
Instrument maintenance and calibration procedures should be specified to the extent
possible in the QA plan. Any maintenance or calibration activity, scheduled or non-
scheduled, should be recorded in an appropriate logbook in order to determine any effects
on the data obtained. Typical calibration data obtained should include:
Flow measurements
Volume measurements
Temperature measurements
Pressure measurements
Determination of response factors, precision, and accuracy for
continuous monitors using span gases and zero gases.
In general the QA plan should specify routine calibration checks at several time points
during the program.
Quality control samples to check overall system performance may include replicate
or split samples, spiked samples, standard reference materials, blanks, and backup samples
(e.g., series impingers or resin cartridges). Split or replicate samples are useful checks on
sampling and analysis precision and should be included with each group of samples. Field
blanks, in which the sampling activity is duplicated exactly except that no air is sampled,
should also be routinely collected. Backup samples should be collected whenever the
30
-------
recovery performance of a particular sampling medium has not been documented or is
subject to wide variations depending on ambient conditions. Spiked samples should be
included wherever an accurate spiking procedure is available, provided that the spiked
material reasonably simulates the physical and chemical state of the native material.
Standard reference materials (SRMs) for organic analysis are rather sparse. The
National Bureau of Standards (NBS) has certified an urban dust (SRM 1649) for selected
polynuclear aromatic hydrocarbons and various gas suppliers will prepare certified
standards of many organic components as dilute gas mixtures. These gas mixtures should
be checked against NBS standard reference materials (e.g., propane in air) to ensure
accuracy of the gas mixture. Routine calibration of continuous monitors using these
certified gas mixtures is highly advisable. SRMs containing selected organic compounds
at trace levels in air or nitrogen are presently being developed and should be available in
the near future from NBS.
Data recording procedures to be specified in the sampling QA plan include (a)
periodic readings of the temperature, flow, volumes, and other parameters, (b) docu-
mentation of meteorological conditions at appropriate time points, (c) documentation of
instrument operating variables (e.g., resin cartridge number), (d) documentation of any
upset conditions such as sudden leakage or pressure surges, and (e) documentation of
calibration or maintenance activities. A logbook for the overall sampling program in
which sampling descriptions, meteorological data, and upset conditions are documented
should be maintained. In addition a sampling data sheet, such as the example in Figure 5,
should be prepared for each sample or set of samples in which the periodic readings and
instrument parameters are recorded. Certain measurements such as filter numbers and
weights or impinger volumes which are required for analytical purposes can be recorded
on a separate sheet with provisions for recording subsequent analytical data on the same
sheet. Separate maintenance and calibration logbooks should be maintained for each
instrument. In most cases, sampling data forms specific for a given program must be
prepared because of differences in the sampling design between programs.
Sample labeling, preservation, storage, and transport procedures should be specified
in the QA plan and these procedures should be carefully explained to field personnel prior
to sampling to ensure proper implementation. Sample labels, prepared in advance, should
include sufficient information to associate the sample with a particular data sheet as well
as the overall program record notebook. In general each sample should be given a unique
identification number with a prefix describing the type of sample.
Sample preservation, storage, and transport procedures must be appropriate for the
type of analyses required. Particulate samples generally should be placed in air tight
containers and stored in the dark to minimize analyte degradation. Resin cartridges and
impingers generally require more attention, because of analyte instability in the matrix,
and should be shipped to the laboratory for analysis within a relatively short time period
(e.g., a few days). These sample types should be placed in airtight, glass containers and
stored at subambient temperatures until analysis. Exposure to solvents must be avoided
for resin cartridges during all stages of handling in order to avoid sample contamination.
Chain of custody forms are required for certain programs having direct legal impli-
cations. The objective of the chain of custody procedures is to document the movement
of a sample from collection until analysis to ensure its integrity. A typical chain of
custody form is shown in Figure 6. Formal chain of custody requirements place a sub-
stantial burden on the field as well as laboratory personnel and should be employed only
31
-------
co
ro
Sample Sample Start Stop
I.D. Description Date Time Time Location
Flow Rate,
liter/min.
Init. Final
Atmospheric Calibration
Pressure Temp. Data
mmHg °C Operator Reference Comments
Figure 5. Typical sampling data sheet
-------
CHAIN OF CUSTODY RECORD
Sample Number:
Shipper Name:
Address:
number street city state zip
Collector's Name Telephone: ( )
signature
Date Sampled Time Sampled hours
Type of Process Producing Waste
Field Information
Sample Receiver:
1.
name and address of organization receiving sample
2.
3.
Chain of Possession:
1.
signature title inclusive dates
2.
signature title inclusive dates
3.
signature title inclusive dates
Figure 6. Typical Chain of Custody Form
33
-------
when the program objectives specifically require such measures. However, if the data
obtained are to be used in litigation, the use of chain of custody procedures is mandatory.
Analytical Quality Assurance
Aspects of the analytical work to be addressed in the QA plan include:
• Method validation requirements
• Instrument maintenance and calibration
• Quality control sample analysis
• Data recording.
Most monitoring programs will use new methods or modifications of existing
methods to some extent. The QA plan must address the validation requirements for each
of these methods. Typical requirements will include determination of precision, accuracy,
detection limit, and specificity through the analysis of laboratory standards, and whenever
possible, representative samples. The validation requirements should be appropriate for
the program objectives and should simulate the actual sampling and analytical situation as
nearly as possible. Validation data should be included as part of the monitoring report and
method writeups and any limitations of the data in terms of defining the performance
characteristics under the actual use conditions should be documented.
Instrument maintenance and calibration requirements for laboratory instruments
will be similar to those for field instruments, including the need to document any
activities of this type. To the extent possible calibration and preventive maintenance
schedules should be included in the QA plan. The format for recording calibration data
(e.g., injection of standards of known concentration) should be specified prior to initiation
of the monitoring effort.
Quality control samples for evaluating analytical performance should include blanks,
spiked process blanks, spiked samples, standard reference materials, and replicate (or
split) samples. Standard reference materials and replicate or split samples should
generally be included as part of field QA and need not be additionally included at the
analysis stage. However additional blanks, spiked process blanks, and spiked samples
should be included at the analysis stage since problems with sample instability and con-
tamination during sampling storage or shipment can be determined separately from
laboratory related problems. Both spiked process blanks and spiked samples should be
included since this practice allows matrix effects to be distinguished from analytical
losses.
Data recording requirements during analysis require a great deal of attention to
ensure that all necessary raw data are available for inspection should unexpected results
occur. The advent of computerized data handling tends to "hide" raw data from the
analyst. Hence the QA plan should specifically state which raw data are to be recorded,
the manner of presentation, and storage procedures. Laboratory data notebooks should
include all raw data or a clear reference as to where the data are recorded (e.g., 9-track
magnetic tape, etc.), equations used in performing intermediate calculations, and final
results. Equations used for calculations, including units for all parameters, should be
presented as part of the method writeups or program QA plan.
34
-------
Data Reduction Quality Assurance
Since sampling and analytical data processing occurs independently, in most cases,
the QA plan should address the manner in which data from the two activities are to be
treated and validated during the reduction process. The actual presentation of data is
described in the following portion of the TAD and will not be discussed here.
Aspects of data reduction to be treated in the QA plan are shown in Table 5. The
extent of documentation and verification required will be greatly dependent on the
program objectives, the nature of the raw data, and the complexity of the computational
process.
DEFINITION OF DATA REPORTING FORMAT
Many air monitoring programs are undertaken without prior definition of the data
reporting format. In some cases this appraoch is justified because of the unexpected
nature of the data obtained (e.g., unexpected compounds detected or previously unknown
sources identified). However, to the extent possible the format for data presentation
should be defined prior to initiation of the monitoring effort. This practice helps to
identify limitations on the available data and further clarifies the extent to which tech-
nical and management or policy personnel understand the program objectives. Stern's Air
Pollution series^' contains an excellent discussion of procedures for analyzing and
presenting air quality data.
The optimal format for data presentation obviously is highly dependent on the
program objectives and the quantity of data obtained. In cases where only a few data
points are obtained around a point source (e.g., a hazardous waste landfill) tabular
presentation of data (compound concentrations at each site) may be appropriate.
However, in most monitoring situations the quantity and complexity of the data set will
require graphical presentation. This type of data format requires definition of the
important variables to be considered (e.g., source locations, sampling times, sampling
sites, meteorological effects, etc.). Statistical methods for evaluating correlation
between the important variables are usually required to obtain meaningful conclusions
from the data set. Typical methods for statistically evaluating and displaying air quality
data are given in Stern's book''/ and therefore are not presented here.
SAFETY CONSIDERATIONS
Safety considerations in air monitoring are similar to those for other chemically
related occupations but should be considered for each air monitoring program since
unusual hazards may be present in these situations. A discussion of general safety con-
siderations is available^. Potential safety hazards can be subdivided into the following
broad categories:
• Chemical hazards
• Electrical equipment
• Mechanical equipment.
Chemical hazards include toxic chemicals such as carcinogenic compounds,
corrosive chemicals such as concentrated acids or bases, and explosive hazards such as
35
-------
compressed gases. Protective equipment should be employed to minimize direct exposure
to such hazards. Since most air monitoring programs require working with concentrated
standards of toxic organic compounds, special emphasis should be placed on minimizing
exposure to these materials. Programs involving investigation of concentrated or
potentially concentrated sources of hazardous organic compounds require additional
safety protocols to protect workers in the field as well as laboratory workers who could be
unexpectedly exposed to concentrated samples collected at such sites.
Hazards from corrosive chemical, compressed gases, glassware, mechanical
equipment, and electrical equipment are presented in the reference given above'^' and do
not require special emphasis here. However, these hazards should be addressed in the
monitoring plan.
36
-------
SECTION 4
SAMPLING AND ANALYSIS STATE OF THE ART
This section of the TAD provides both general and specific information regarding
sampling and analysis techniques. The section is divided into six parts as follows:
Overview of sampling methods
Overview of analysis methods
Discussion of specific sampling and analysis methods for
various compounds
Quality assurance procedures
Instrument calibration procedures
Data handling.
OVERVIEW OF SAMPLING METHODS
Important Chemical and Physical Properties
In order to select optimal sampling techniques for a given compound or group of
compounds one must consider the important properties of the compounds(s) of interest
having an effect on the sampling process. Physical properties to be considered include
boiling point, vapor pressure, water solubility, and solubility in various organic solvents.
If a compound has sufficiently low volatility to be associated with particulate matter, one
must also consider the size distribution of the particles and heat of adsorption of the
compounds of interest.
Chemical properties of concern include thermal stability, reactivity with water or
other common constituents in ambient air, and photochemical reactivity. The dissociation
constant (pKa or pKb) for ionizable compounds should also be considered. The exact
manner in which the various chemical and physical properties impact the sampling scheme
is discussed below, with the discussion of individual sampling methods.
There are several useful handbooks which can be consulted to obtain information
concerning the chemical and physical properties of organic compounds. These include the
Handbook of Chemistry and Physics^"), the Merck Index^D, and the Handbook of
Environmental Data on Organic chemicals^^T. The latter reference also contains useful
information on environmental effects and regulatory limits for various industrial
chemicals.
Methods for Gas Phase Components
Compounds which are predominantly in the gas phase at ambient temperature and
pressure are generally sampled by passing the air sample through a filtration device to
37
-------
remove participate material prior to capture of the gaseous components. The sampling
techniques commonly used for gas phase components are summarized in Table 6. In
selected cases, direct analysis of the filtered gas stream is possible, circumventing the
need for the capture process (e.g., direct GC with photoionization or electron capture
detection). This situation is rare in ambient air monitoring for toxic organics because the
low concentrations generally make preconcentration of the sample a necessity. A general
discussion of the various sampling approaches for gas phase components is provided in
Solid Adsorbents-
Solid adsorbents are the media most commonly employed for sampling gas phase
organics. The primary advantage of this sampling approach is the large volume of air
which can be sampled relative to other techniques such as impingers or cryogenic
sampling. These media can be generally divided into three categories as follows:
• Organic polymeric adsorbents
• Inorganic adsorbents
9 Carbon adsorbents.
Organic polymeric adsorbents include materials such as Tenax^ GC and XAD-2. These
materials have the important feature that water is not collected in the sampling process
and hence large volumes of air can be collected. Another advantage of the organic
polymeric adsorbents is the absence of "active sites" which can lead to irreversible
adsorption of certain polar compounds. A major disadvantage of these materials is their
inability to capture highly volatile materials (e.g., vinyl chloride) as well as certain polar
materials (e.g., methanol, acetone).
Inorganic adsorbents include silicagel, alumina, Florisil^, and molecular sieves. The
materials are considerably more polar than the organic polymeric adsorbents, leading to
the efficient collection of polar materials. Unfortunately, water is also efficiently
captured leading to rapid deactivation of the adsorbents. Consequently, these materials
are seldom used for sampling trace organic components in air, except in cases where
relatively high concentrations of certain polar materials are of concern.
Carbon adsorbents are relatively nonpolar compared to the inorganic adsorbents and
hence water adsorption is a less significant problem, although the problem may still
prevent analysis in certain applications. The carbon based materials tend to exhibit much
stronger adsorption properties than organic polymeric adsorbents, hence allowing efficient
collection of volatile materials such as vinyl chloride. However, the strong adsorption on
carbon adsorbents can be a disadvantage in cases where recovery by thermal desorption of
less volatile materials such as benzene or toluene is desired because of the excessive
temperatures required (e.g, 400 C).
There are a variety of carbon based adsorbents available with widely varying
adsorption properties. The commonly available classes of carbon adsorbents include:
• Various types of conventional activated carbons
• Carbon molecular sieves
• Carbonaceous polymeric adsorbents.
Conventional activated carbons have a microporous structure which leads to difficulty in
recovering adsorbed materials and hence this material is rarely used in trace organic
sampling.
38
-------
Table 6. Sampling Techniques for Gas Phase
Organic Components
Solild Adsorbents
- Organic Polymers (Tenax, XAD-Z)
- Inorganic (Silica gel, Florisil)
- Carbon (Activated Carbon, Carbon Molecular Sieves)
Cryogenic Trapping
Impingers
Whole Air Collection (Cannisters, Glass Bulbs)
Derivatization Techniques
39
-------
Carbon molecular sieves have a spherical, macroporous structure which leads to
better recovery of adsorbed materials (relative to conventional activated carbons). These
materials are sold under various tradenames (e.g., Spherocarb, Carbosphere, Carbosieve,
and Ambersorb XE-347) and have been used to a limited extent for the determination of
volatile organics such as vinyl chloride and methylene chloride^4). Much more work
needs to be done in this area before the usefulness of such adsorbents can be established.
Carbonaceous polymeric adsorbents are described as hard, nondusting spheres with
properties intermediate between activated carbon and organic polymeric adsorbents.
These materials are available from Rohm and Haas Chemical Company under the
tradename "Ambersorb" (XE-340, XE-347, XE-348). Of the three materials, Ambersorb
347 is classified as a carbon molecular sieve, XE-348 is most similar to activated carbon,
and XE-340 is most similar to organic polymeric absorbents. XE-340 appears to have
some attractive features in terms of trace organic sampling and has been shown to be
useful for volatile compounds in the C± to C& boiling point range(^,16)? a volatility range
not covered by organic polymeric adsorbents such as Tenax.
In selecting particular adsorbent materials for sampling trace organics one must
consider both the capture process and the compound recovery process. In general, one of
two recovery processes are employed, thermal desorption or solvent extraction. Thermal
desorption is most useful for compounds having boiling points less than 300 C whereas
solvent extraction is most useful for compounds boiling above 150 C. Thermal desorption
is an attractive approach in many cases since the entire sample can be introduced onto
the analytical instrument compared to solvent extraction where only a small fraction of
the extract (e.g., 1-10 percent) can be introduced. Furthermore, the thermal desorption
process is more readily automated and does not require disassembly of the sampling
cartridge prior to analysis.
Solvent extraction offers the advantage of being able to adjust the concentration of
analyte introduced into the analytical system (i.e., to remain within the working range of
the instrument) and also allows replicate analysis of a sample, features not available for
most thermal desorption systems. Use of solvent extraction also avoids the problem of
thermal decomposition of labile compounds during the analysis step.
Although a wide variety of approaches employing solid adsorbents may be feasible
for any given monitoring situation the following summary offers useful guidance for
generalized cases:
• Thermal desorption of organic polymeric adsorbents (especially Tenax GC)
is useful for compounds boiling between 60-300 C, exclusive of highly
polar compounds such as methanol and acetone.
• Solvent extraction of organic polymeric adsorbents (especially XAD-2)
is most useful for compounds boiling above 150 C and can in some
cases be extended to more volatile compounds depending on solvent
and mode of concentration.
• Thermal desorption of carbon adsorbents (especially carbon molecular sieves
or Ambersorb XE-340) may be useful for compounds boiling in the range 0-70 C,
including vinyl chloride. However, the high desorption temperatures
required (350-400 C) may lead to degradation of certain labile compounds.
40
-------
If medium or high concentractions of compounds are of concern, solvent
desorption of carbon adsorbents may be a useful alternative.
«, Inorganic adsorbents such as silica gel or alumina are not generally useful
for ambient air monitoring because of the water deactivation problem.
Cryogenic Collection—
The collection of atmospheric organics by condensation in a cryogenic trap is an
attractive alternative to adsorption or impinger collection. The primary advantages of
this technique include:
• A wide range of organic materials can be collected
• Contamination problems with adsorbents and other collection media are avoided
• The sample is immediately available for analysis without further workup
• Consistent recoveries are generally obtained.
However, an important limitation of the technique is condensation of large
quantities of moisture, and lesser amounts of certain reactive gases (SC>2, Nx* etc.). The
principles of cryogenic sampling are described in Stern's Air Pollution
Cryogenic sampling can be accomplished in a variety of ways, depending on the
desired detection limit and compounds of interest. The important parameters to be
specified include:
• Choice of cryogen
• Trap design
• Method of sample recovery
• Method of analysis.
Useful cryogens include liquid oxygen, liquid argon, dry ice-solvent systems, and ice
water. Liquid nitrogen is not an acceptable cryogen because large quantities of air will
condense. Ice water is not cold enough for collecting organics in ambient air, except for
relatively nonvolatile compounds. Dry ice-solvent systems should be employed with
caution because of the high probability of contaminating the sample with relatively large
quantities of solvent. Liquid oxygen or argon appear to be the most generally applicable
cryogens, although the safety hazards associated with liquid oxygen make its use less
attractive.
The design of a suitable sample trap is extremely important in cryogenic sampling.
One must ensure that the air residence time in the trap during sampling is great enough to
allow cooling of the gas stream and condensation of the analyte of interest. The trap
material must be able to withstand the wide temperature range involved in the sample
process. The trap design must also be appropriate for the sample recovery step, allowing
efficient recovery without loss or contamination of the sample. For highly volatile
materials, the inclusion of an adsorption medium such as silica gel may be necessary to
obtain good collection efficiency.
Cryogenically collected samples can be recovered either by flash evaporation into
an analytical instrument or by solvent flashing of the trap. The former approach is
preferable in most cases since it allows more sensitive detection and avoids contamination
(e.g., from solvents). However, the solvent flushing approach can be accomplished with
less elaborate equipment in the field and may be preferable for the analysis of higher
concentrations of material.
41
-------
A generally applicable cryogenic sampling approach for monitoring trace organics in
ambient air has been described" '>. This system involves the use of a small (3 mm ID)
freeze out trap held at liquid argon or liquid oxygen temperature and packed with
silanized glass beads. A 50-500 ml volume of ambient air is drawn through the trap using
an evacuated tank/manometer assembly as the pumping and volume measurement system.
The trap is directly connected to a six-part stainless steel valve and following sampling
collection the condensed material is flash evaporated onto a capillary GC/FID system for
analysis. Detection limits on the order of 0.5 to 1 ppbv can be achieved using this
approach. Even lower detection limits can be achieved for certain compound with the use
of selective detectors, such as electron capture (ECD).
This technique requires transport of the analytical system to the monitoring site, or
collection of a whole air sample in an evacuated cylinder with subsequent laboratory
analysis. For higher boiling compounds adsorption onto container walls may represent a
significant problem if the latter approach is employed.
Impinger Collection—
Impinger or "bubbler" collection involves passing the gas stream through an organic
solvent or other suitable liquid and capture of the analyte by partioning into the solvent
(special derivatization impinger systems are discussed later in this document). This
technique is not generally applicable for trace organic analysis because large volumes of
air cannot be sampled, due to solvent evaporation during the sampling process. However,
this technique is relatively simple and may be useful in high level (ppm) monitoring
situations.
The impinger should be designed so that contact between the air and solvent is
maximized, either by the use of fritted diffusers or capillary jets. Some typical impinger
designs are given in Stern's book^°'. In general, the impinger systems should be cooled in
ice water so as to reduce solvent loss during sampling.
Whole Air Collection-
Collection of whole air samples using evacuated glass bulbs, stainless steel
canisters, or similar devices is probably the simplest sampling approach, and can be useful
in many situations. An obvious limitation of this approach is that the sample components
of interest may be adsorbed or decomposed through interaction with the container walls.
At very high analyte levels (e.g., several ppm) condensation may be a problem.
Consequently, this approach is most useful for relatively stable, volatile compounds such
as hydrocarbons and chlorinated hydrocarbons with boiling points less than 150 C.
However certain compounds within these classes pose storage stability problems (e.g.,
carbon tetrachloride interacts with stainless steel surfaces and is lost). Preconditioning
surfaces (e.g., formation of an oxide coating) or selection of alternate container materials
can circumvent these problems in many cases. In all cases, the container must be flushed
(with moderate heating if possible) with zero grade nitrogen or air prior to sampling in
order to remove trace contaminants.
Containers used for whole air sampling can be roughly categorized as rigid or non-
rigid devices. Nonrigid devices include Tedlar^- or Teflon^ plastic bags. Generally such
devices are used to collect samples for analysis within a few hours, since the rate of
leakage and/or permeation of materials into and out of the bag is relatively high.
Rigid containers have an advantage in that leakage and/or permeation rates are
generally low and samples can be stored for several days or even several months for
certain compounds. Examples of rigid containers include:
42
-------
• Glass bulbs
• Gas-tight syringes
• Stainless steel cylinders or cannisters.
Sampling into such devices can be accomplished either by evacuating the container in
advance and then allowing sample to enter the container or by having both inlet and outlet
valves on the container and pumping sample through the container until equilibrium is
obtained (e.g., after 5-10 container volumes of sample has been flushed through the
system). The first approach has the advantage that no sampling pump is required in the
field. However, the latter approach has the advantage that equilibrium with the container
walls is more readily achieved and high vacuum seals are not required.
An alternate sampling approach is to pressurize a whole air sample into a stainless
steel cylinder. This approach has proven useful in ambient air monitoring studies^''
wherein electropolished 6-liter stainless steel cans were pressurized with ambient air to
approximately 15 psi. Under these conditions, relatively volatile materials having boiling
points less than 120 C had stable concentrations for one to two weeks. This approach has
the advantage that a relatively large volume of air can be collected and transported to
the laboratory for analysis. However, condensation of material will become a significant
problem as the container pressure is increased.
Derivatization Techniques—
A fundamental limitation of the various sampling techniques discussed so far is the
decomposition of reactive compounds during collection or transport of the samples. Two
approaches can be used to circumvent this problem. One obvious approach is to use a
direct analysis technique in which the analyte concentration is determined without
physically isolating the air sample (e.g., total hydrocarbon analysis).
An alternate approach, discussed in this section, is to stabilize reactive compounds
by combining them with a derivatizing reagent during the sampling process. In many
cases derivatizing reagents can be chosen which not only stabilize the compound but
enhance its detectability. Such schemes are available for determining aldehydes,
phosgene, and certain other reactive compounds.
In certain cases, such as the 4-aminoantipyrene method for phenols and the
ninhydrin method for amines the derivatization step is used solely to enhance detect-
ability and may be done in the laboratory rather than the field. In this section, only
derivatization techniques for field use are considered. Derivatization schemes for
specific compounds or compound classes are discussed later.
Derivatization reagents for field use can be held either in an impinger or on a solid
adsorbent. The impinger approach is most convenient since reagents can be prepared and
stored as liquid solutions whereas solid adsorbent systems require more elaborate
preparation and storage procedures. However, the solid adsorbent approach is more
sensitive in many cases because larger volumes of air can be sampled. A study of both
these approaches for aldehyde derivatization has been reported^'). In this particular
example, the impinger approach appears most useful since it provides adequate sensitivity
and is less susceptible to humidity effects on analyte recovery.
All derivatization techniques require the use of high purity reagents. Normally
fresh stocks of reagent should be prepared for each sampling episode, or more frequently
if degradation or contamination (through passive diffusion of materials into the reagent) is
43
-------
a problem. One must be aware of potential interference in the method, especially from
reactive gases present at high levels in the sample (e.g., water, SC>2, NC>2, 03, etc.) and
must incorporate QA procedures to ensure such interferences are not a problem.
Direct Analysis—
Probably the most attractive sampling and analysis approach is the direct analysis of
atmospheric contaminants without collection of a sample. Methods fitting this definition
for organic compounds include gas phase optical techniques, total flame ionization
detection, and related techniques. The specific techniques are discussed in detail in the
description of "field screening techniques" in the analytical section of the TAD. A major
limitation of these techniques is sensitivity since no sample preconcentration is obtained.
Direct analysis techniques do not require sample collection and hence avoid
problems with analyte degradation or adsorption during collection and storage. However,
most of the techniques require delivery of the air sample to the analytical device through
some type of tubing. Generally such tubing should be inert (e.g., teflon or passivated
stainless steel) and heated (>100 C) to avoid wall losses. Generally a heated filter will be
placed upstream of the analytical device to avoid interference from particulate matter.
Sampling Methods for Particle Bound Components
Relatively nonvolatile (boiling point greater than 200 C) and polar compounds may
exist partially or exclusively on the surface of suspended particulate matter. The distri-
bution of organic compounds between the gas and particle phase is greatly affected by the
sampling conditions chosen. Certain components may be slowly driven from the captured
particles during sampling whereas certain gas phase components may adsorb to filter
media, hence appearing in the particle sample. Some types of particles (e.g., water mists)
may volatilize during sampling. Consequently the "particle bound" concentration of an
analyte is an operationally defined term and must be chosen so as to satisfy the program
objectives. An important aspect of particle sampling for trace organic analysis is the
need for a gas phase collection device (e.g., impinger, adsorbent, etc) downstream of the
particle sampler since virtually all organics will be present in the gas phase or volatilize
from the particle sampler to some extent. A detailed discussion of sampling consider-
ations for particulate matter is given in Stern's book(20). A listing of available particle
sampling approaches is given in Table 7.
Filtration—
The most common method for collection of particulate matter is filtration. If only
the total quantity of analyte is of interest the filter may be operated at an elevated
temperature to drive material into the gas phase collection device, hence simplifying the
analysis requirements. In general, compounds more volatile than anthracene will not be
retained on the filter at a temperature of 150 C. If the particle bound concentration is to
be determined the filter should be maintained at ambient temperature and the volume of
gas sampled should be minimized to reduce volatilization of the particle bound com-
ponents. Unfortunately, the low levels of ambient particulate matter usually require the
collection of several hundred cubic meters of air (e.g., with a high volume sampler).
Under these conditions even relatively nonvolatile compounds (e.g., 3 and 4 ring PAHs)
may be lost to some extent. Chemical transformation of some adsorbed materials through
the interaction with atmospheric gases (e.g., NOX, SC>2, 03) analyses can also be a
problem in such situations.
44
-------
Table 7. Particle Sampling Approaches
Filtration
- Cellulose Fiber
- Glass or Quartz Fiber
- Teflon Coated Glass Fiber
- Membranes
Centrifugal Collection (Cyclones, etc.)
Impaction
Electrostatic Precipitation
45
-------
A variety of filter media are available for air sampling, are described in Katz1
The most commonly used filter media for sampling organics are:
• Cellulose fiber
• Glass or quartz fiber
• Teflon coated glass fiber
• Membrane (cellulose nitrate or teflon).
Cellulose fiber filters do not have uniform pore sizes and tend to have unpredictable
collection efficiencies for fine particles. However, these filters have high mechanical
stength and are inexpensive. Glass or quartz fiber filters have the advantage of with-
standing higher temperatures and yielding better collection efficiencies than cellulose
fiber filters. However, these filters tend to be mechanically fragile and require careful
handling. An additional problem, especially with glass fiber filters is irreversible
adsorption of organics onto the filter. This problem is less severe for quartz and is
circumvented to a large extent by using teflon-faced glass fiber filters.
Membrane filters have a well defined pore size distribution and have very predicta-
ble sampling properties. However these filters are expensive, have a relatively high
resistance to flow, and a rather low sample capacity. In addition, most of the membrane
filters (with the exception of teflon) are soluble in organic solvents, hence complicating
the analyte recovery process for organic compounds.
Centrifugal Collection and Impaction—
Most particle sampling approaches other than filtration are employed for the
purpose of size classification of particulate matter. The primary reason for collecting
size fractionated particle samples is to distinguish respirable (<15ym diameter) from larger
particles. Although the most common objective is to compare particle mass in each size
range, the chemical composition is of interest in certain cases. Methods for collecting
size fractionated particulate samples generally rely on the difference in momentum (mass
times velocity) of the various size particles. Devices operating on the principle include
cyclones (centrifugal collectors) and impactors. Both of these devices collect particles of
a specific size range, based on the design of the apparatus. Cyclones tend to have a
rather broad size cut whereas impactors have a much sharper size cut. These devices are
useful for particles greater than approximately lym in diameter and hence are often used
as the basis for "respirable" particle determinations.
In a typical configuration two or more cyclones or impaction stages will be placed in
series followed by a high efficiency filter to collect fine particles. In this scheme one
could collect, for example, particle size fractions of >15ym (cyclone 1), 1-15 ym (cyclone
2), and <1 pm (filter) diameter.
A device commonly employed for collecting size fractionated ambient particulate
matter is the dichotomous sampler. This device consists of an impactor stage followed by
a high efficiency filter. Particles greater than 15 urn diameter are excluded from the
device. Particles with diameters less than 15 ym are collected on filters in two size
fractions (2.5-15 ym and<2.5 ym). A recent study has compared the quantity of ambient
particulate matter collected using the widely used "Hi-Vol" sampler (filtration collection
only) and the "dichotomous sampler"(22). Since the Hi-Vol sampler can collect particles
of >15 ym this device will normally give higher particle loadings than the dichotomous
samples.
46
-------
A major problem in sampling ambient particulate matter for organic analysis is the
quantity of sample which can be collected. This problem is especially acute for size
fractionation type sampling procedures, due to flow rate limitations. A typical ambient
particle loading of SOyg/rn^ requires that 1000 m^ of sample be collected in order to
obtain 50 mg of particulate matter (a typical amount required for organic analysis). A Hi-
Vol sampler requires 20 hours sampling time to collect this amount of material and a
dichotomous sampler, typically operating at 50 liters per minute, would require 15 days.
A device has been developed by EPA for sampling large volumes of size fractionated
ambient particulate matter^). This device called the "Megavolume Respirable
Particulate Sampler" consists of a cascade impactor followed by an electrostatic precipi-
tator. The absence of a filter reduces the pressure drop and allows higher sample flow
rates (up to 15 m-Vminute). Particles of greater than 20 ym diameter are excluded from
the sampling system and size ranges of 2.5-20ym, 1.7-1.5 ym, and<1.7ym are collected.
This device has the disadvantage of being very bulky (4 ft x 2 ft x 2 ft) and the effect of
electrostatic precipitation on chemical composition has not been thoroughly investigated.
Electrostatic Precipitation (ESP)—
This method is capable of capturing particles of greater than 0.1 ym and serves (for
example in the sampler described above) as an alternative to filtration for the collection
of fine particles. A high electrical field (12-45 KV) is maintained between a series of
parallel plates. Particles entering the device become charged by ions present in the gas
phase and are attracted to the plates.
This sampling approach allows large sample volumes to be collected, due to the
lower pressure drop, compared to filtration. However, the effect of the corona discharge
and associated reactive species (e.g., ozone) generated in the ESP on organic chemical
composition is of concern and has not been thoroughly investigated.
OVERVIEW OF ANALYTICAL METHODS
In this section of the TAD, an overall description of analytical techniques is
presented and the chemical and physical properties governing the applicability of the
various methods for organic compounds are discussed. Analytical techniques have been
subdivided into the following general categories:
• Field screening techniques
• Laboratory screening techniques (compound class methods)
• Compound specific techniques.
Field screening techniques (referred to as direct analysis techniques in the sampling
section) are defined as relatively inexpensive techniques which can be accomplished
rapidly in the field. Laboratory screening procedures are defined as techniques such as
infrared spectroscopy which can determine functional group concentrations as opposed to
concentrations of specific compounds. Relatively expensive compound specific techniques
for laboratory use, and in selected cases for field use (e.g., portable MS system) are
discussed in the final category.
Physical and Chemical Properties of Concern
In many respects the physical and chemical properties to be considered in the
selection of sampling methods are of similar importance in the selection of analytical
47
-------
methods. However, important differences in emphasis exist and must be recognized.
Properties of primary concern include:
Physical state (gaseous or adsorbed)
Volatility
Polarity
Water solubility
Ionic character
Reactivity
Thermal stability
Chemical composition.
Obviously the properties of concern are those of the compounds in the form delivered to
the laboratory (e.g., derivatives of reactive compounds, trapped on solid adsorbents, etc.).
One of the most important properties governing the selection of analytical methods
is volatility. Sample preparation procedures, wherein the analytes of interest are
delivered from the sample matrix into a form suitable for analysis, generally involve
either solvent extraction/concentration or thermal desorption/cold trapping approaches.
Highly volatile compounds (e.g., boiling points below 100 C) cannot be readily analyzed by
solvent extraction/concentration approaches because of losses during the concentration
process and interference from the solvent on the analytical system. On the other hand,
compounds with boiling points above 250 C usually are not candidates for the thermal
desorption approach.
Volatility, polarity, and thermal stability must also be considered in the selection of
determinative techniques since GC procedures are not suitable for nonvolatile, highly
polar, or thermally labile compounds unless special precautions are taken. Generally
polarity and thermal stability are limiting factors for GC determination of organic
compounds in ambient air since these compounds tend to be sufficiently volatile (i.e.,
boiling point less than 400 C).
Ionic character, water solubility, and reactivity are primarily of concern during
sample preparation procedures involving liquid extraction or adsorption chromatography
as well as for determinative procedures such as high performance liquid chromatography
(HPLC). lonizable compounds (acids or bases) require careful control of pH when aqueous
systems are employed.
Generally ionizable compounds are converted to the nonionized form during
processing by adjusting the pH two to three units below the pKa for acidic compounds and
two to three units above the pKb for basic compounds. Highly reactive compounds may
present problems in the sample preparation procedure through reaction with solvents or
adsorption chromatography supports. Phosgene, for example, reacts readily with water
and hence water must be carefully excluded from the sample preparation scheme.
Chemical composition is important in the selection of determinative techniques.
Whenever possible a detection system selective for the compound class should be used.
Suitable selective detection systems for GC analysis are available for halogen, nitrogen,
sulfur and phosphorous containing compounds. HPLC analysis is appropriate if UV
absorbing (usually aromatic), fluorescent or electrochemically active (e.g., phenol,
aromatic amines) functional groups are present in the molecule.
-------
The chemical and physical properties of major compounds (other than those of
interest) present in the sample should be considered, since these compounds represent
potential interferences. Whenever possible the analytical and sampling techniques should
take advantage of differences in properties between compounds of interest and other
compounds to minimize interference in the methods.
Field Screening Techniques
The term "screening technique" can imply different approaches to different people
and therefore an introductory definition is necessary. Within this document "field
screening techniques" will refer to techniques which can be accomplished rapidly in the
field with relatively inexpensive instrumentation. This definition excludes sampling and
analysis approaches which involve conventional sampling techniques and subsequent
application of group specific methods in the laboratory (e.g., EPA Level 1 methodology).
This definition also excludes devices such as mobile mass spectrometers (MS) and triple
quadruple MS/MS systems because these systems are relatively expensive.
This limited definition of screening technique should not be viewed as an implication
that the sampling and analysis approaches excluded are not useful. Indeed, many circum-
stances will arise where such techniques are extremely useful and should be employed in
preference to (or in addition to) the screening techniques described in this document. The
purpose in limiting the definition of "screening technique" in this way is to present the
reader with a discussion of alternative approaches which can be used to rapidly evaluate
whether or not hazardous organic compounds are present at a particular site. This evalu-
ation should be accomplished rapidly (e.g., field operable) and with a minimum of expense
(since a multitude of sites may need to be investigated). Such techniques need not be
selective for a particular compound but should be capable of distinguishing organic
components typically found in uncontaminated air (e.g., methane) from hazardous
substances (e.g., chlorinated and aromatic hydrocarbons). If possible, the technique should
indicate the class of organic compound present so that subsequent detailed investigations
of a site can be directed toward a limited set of target compounds, thereby minimizing
the cost and time associated with these detailed investigations.
For the purpose of this document the available screening approaches can be broadly
subdivided into the following four categories:
• Colorimetric methods
• Spectroscopic devices
• lonization devices (with or without gas chromatographic separation)
• Photometric devices (with or without gas chromatographic separation).
Table 8 presents a listing of some typical commercially available devices useful as
screening techniques. This listing is certainly not complete and is simply intended to
present the reader with an overview of the specifications for such instruments. One must
be aware that most of these devices have been designed to meet monitoring needs of the
industrial hygiene community, where part per million levels are usually of concern.
Consequently, if one must detect part per billion levels in ambient air, many of these
devices will not be sufficiently sensitive.
In the following sections the operating principles, advantages and limitations of the
various screening approaches are discussed.
49
-------
TABLE 8. TYPICAL COMMERCIALLY AVAILABLE SCREENING TECHNIQUES
FOR ORGANICS IN AIR
en
o
Technique
Compounds
Manufacturer Detected
Approximate
Detection
Limit
Comments
Gas Detection Tubes
Draeger
Matheson (Kitagawa)
Continuous Flow Colorimeter CEA Instruments, Inc.
Colorimetric Tape Monitor
Infrared Analysis
FID (Total Hydrocarbon
Analyzer)
GC/FID (portable)
MDA Scientific
Foxboro/Wilkes
Beckman
MSA, Inc.
AID, Inc.
Foxboro/Century
AID, Inc.
Various organics
and inorganics
Acrylonitrile,
Formaldehyde,
Phosgene, and
various organics
Toluene, diiso-
cyanate, dinitro-
toluene, phosgene,
and various
inorganics
Most organics
Most organics
Same as above
except that polar
compounds may not
elute from the
column
0.1 to 1 ppmv
Sensitivity and selectivity
highly dependent on
component of interest
0.05 to 0.5 ppmv Sensitivity and selectivity
similar to detector tubes
0.05-0.5 ppmv
1-10 ppmv
0.5 ppmvC
0.5 ppmvC
Same as above
Some Inorganic gases
(H20, CO) will be detected
and therefore are potential
interferences
Responds uniformly to most
organic compounds on a
carbon basis
Qualitative as well as
quantitative information
obtained
-------
TABLE 8. (Continued)
Technique
Manufacturer
Compounds
Detected
Approximate
Detection
Limit
Comments
PID and GC/PID (portable)
GC/ECD (portable)
GC/FPD (portable)
HNU, Inc.
AID, Inc.
Photovac, Inc.
AID, Inc.
AID, Inc.
Most organic compounds 0.1 to 100 ppbv
can be detected with
the exception of
methane
Halogenated and nitro 0.1 to 100 ppbv
substituted compounds
Sulfur or phosphorus- 10-100 ppbv
containing compounds
Selectivity can be adjusted
by selection of lamp
energy. Aromatics most
readily detected
Response varies widely from
compound to compound
Both inorganic and organic
sulfur or phosphorus
compounds will be detected
Chemlluminescent
Nitrogen Detector
Antek, Inc.
Nitrogen-containing
compounds
0.1 ppmv(as N)
Inorganic nitrogen
compounds will interfere
-------
Colorimetric Detection—
Colorimetric devices involve drawing an air sample through an adsorbent or solution
containing a reagent which reacts with the cotnpounds(s) of interest to yield a colored
material. The specificity of such methods is highly dependent on the compound being
monitored and one must always be cognizant of the fact that some airborne components
may create negative or positive interferences in the method. Humidity, temperature, and
sample flow rate can affect the results obtained, and consequently these variables must
be known and carefully controlled whenever possible.
The most common type of Colorimetric device available for air monitoring is the gas
detection tube (e.g., Draeger or Kitagawa tubes). Tubes of this type are available for
long-term (e.g., 8-hour worker exposure) or short-term (e.g., 5-10-minute grab sample)
collection. The tubes consist of an adsorbent such as silicagel impregnated with a color
developing reagent. When contaminated air is drawn through the tube at the specific rate
and sampling time, the length of the stain (color) developed will be proportional to the
concentration of the contaminant.
At present, gas detection tubes are available for the determination of approximately
40 different organic components commonly found in the workplace. The sensitivity and
selectivity of the various tubes are highly dependent on the compound of interest.
Consequently the user must carefully review the performance specifications in terms of
interferences and should understand the chemical reaction occurring in the detection
process before assuming a particular gas detection tube is suitable for a particular
monitoring effort.
Detection limits for organic compounds using the detector tube are generally 0.1
pptnv or higher and often will not be adequate for monitoring ambient levels of the
compounds. However, this level of sensitivity may be sufficient for monitoring air in the
vicinity of hazardous waste dumpsites and similar sources.
Colorimetric devices other than gas detection tubes are available. One such device
(available from CEA Instruments Incorporated) uses an integrated liquid pumping system
to trap the component of interest, react it with a color developing reagent, and measure
the absorbance continuously. This device, while suffering from the same sensitivity and
selectivity limitations as the detector tube, offers a continuous monitoring capability
which can be useful for rapidly surveying an area for sources of contamination.
Currently, this device can detect formaldehyde, acrylonitrile, and a variety of inorganic
species with detection limits on the order of 0.05 to 0.5 ppmv. However, this basic
system could conceivably be applicable to many more compounds using alternate color
developing reagents.
Another Colorimetric approach involves impregnating a continuous paper tape reel
with a color reagent and passing air through the tape. The tape is slowly driven from the
air sampling region to an optical detector which measures the absorbancy at a specified
wavelength. Devices of this type are available from Universal Environmental Instruments
and MDA Scientific, Inc. Currently, devices for the detection of isocyanates,
dinitrotoluene, phosgene, and a variety of inorganic species are available. Detection
limits of a few ppfav are achievable for selected compounds (e.g., toluene diisocyanate),
although the detection limits and selectivity of this approach are generally similar to the
other Colorimetric approaches. Nevertheless, this approach can be used for area
monitoring on a continuous basis to observe temporal changes in concentration of selected
components.
52
-------
Spectroscopic Devices—
Although a wide variety of spectroscopic techniques (e.g., infrared spectroscopy
(IR), ultraviolet absorption (UV), fluorescence) could potentially be applied to air
monitoring, only infrared (IR) monitoring has been employed to a significant extent. UV
spectroscopy is too nonspecific to be of any real value and fluorescence detection is
applicable only to a few selected compounds, although either of these techniques could be
useful in selected cases.
IR has the advantage that virtually all organic compounds can be detected and give
characteristic spectra. Hence, both selectivity and a wide range of applicability are
achievable with this approach. Major limitations of the IR approach are (1) many
inorganic gases (e.g., water, CO, NO) are also detected at certain wavelengths, and (2) IR
absorbance is not as strong as UV absorbance, hence giving rise to poor detection limits
(1-10 ppmv). The latter limitation is minimized by employing long, folded-path optical
cells (20 meters). Microcomputer control of commercial IR screening devices has
partially overcome the first limitation by allowing automated correction for inter-
ferences. A complete line of portable IR devices is available from Foxboro Analytical
(Wilkes Division). Analytical parameters (reference and analysis filter wavelength) for
monitoring approximately 100 organic compounds are available.
lonization Devices—
lonization techniques operate on the principle of subjecting the analyte to a high
energy source with subsequent detection of the ionized species by collection at an
appropriately polarized electrode. The most widely used techniques of this type are
(a) flame ionization detection (FID), (b) photoionization detection (PID), and (c) electron
capture detection (ECD). Each of these techniques can be used in a "stand-alone" mode or
can be coupled with a gas chromatograph to provide greater specificity.
Flame lonization Detection—By far the most widely used device in the category is
the FID, wherein the gas sample is introduced into a hydrogen/air flame. The technique is
specific for organic compounds and gives relatively uniform response for the various
compounds. Consequently, this device serves as a basis for most of the commercially
available "total hydrocarbon" and "nonmethane hydrocarbon" analyzers. The detection
limits for such devices typically are about 0.5 ppmv (as carbon).
In addition to the commercially available total hydrocarbon monitors (e.g., Beckman
Instruments, Mine Safety Appliance, Inc.), several portable GC/FID systems are available
(e.g., Foxboro/Century, Analytical Instrument Development). These GC/FID systems
provide qualitative, as well as quantitative data in the field — an invaluable feature in
many cases. In particular, the GC/FID systems separate methane and many other ubiqui-
tous organic compounds from the more hazardous compounds, thus allowing more sensitive
and selective determination of the compounds of interest.
In addition, some of these GC systems (in particular the Analytical Instrument
Development device) can utilize alternate detectors (e.g., flame photometric, ECD) to
provide even greater qualitative information, as well as better sensitivity for selected
groups of compounds.
Photoionization Detection—Photoionization detection (PID) involves subjecting the
gas phase compounds to a high-intensity beam of UV radiation of a particular energy.
Absorption of the radiation by a gas molecule leads to formation of a positive ion and free
electron, provided the ionization potential for the compound is less than the radiation
53
-------
energy. The ion is collected at an electrode and the resultant current (proportional to
analyte concentration) is monitored.
Since the ionization potential of a particular compound must be less than the
radiation energy, compounds having high ionization potentials will be less easily detected
than those of lower ionization potential. Consequently, the choice of lamp energy will
have a profound effect on detector specificity. This aspect of PID detection is attractive
from the standpoint that using a higher energy lamp will provide a relatively nonselective,
highly sensitive detector; whereas, a lower energy lamp will yield a selective detector
which can sensitively detect certain readily ionized compounds (e.g., aromatic hydro-
carbons) but will not detect aliphatic hydrocarbons.
Several PID devices are commercially available (e.g., HNU, Analytical Instrument
Development, Photovac). A variety of lamp energies are available ranging from 9.5 to
11.7 eV. In general, the ionization potential for important groups of organic compounds
increases in the order: aromatics
-------
Photometric Devices-
Photometric techniques involve monitoring radiant energy emitted by a species
which is subjected to physical or chemical excitation processes. The two most common
techniques in this category are flame photometric detection (FPD) and
chemiluminescence.
Flame Photometric Detection (FPD)—The FPD uses a hydrogen/air flame to decom-
pose the compound and excite certain elements. The light emitted by certain elements
(e.g., sulfur and phosphorus) is monitored. The element to be monitored is selected by
placing an appropriate optical filter in front of the photomultiplier tube to allow only
light of the particular wavelength to be detected. While several elements could be
monitored by this approach, only sulfur and phosphorus compounds (separately) are
generally analyzed.
The FPD is used as a total sulfur or phosphorus detector in a stand-alone mode
(similar to FID for total hydrocarbons). However, inorganic as well as organic species will
be detected and common inorganic gases, such as SO2, will interfere with organic
analysis. Consequently, use of GC/FPD is advantageous for monitoring sulfur or
phosphorus-containing organics. Such a GC/FPD system is available for field use (from
Analytical Instrument Development). Detection limits on the order of 10 to 100 ppbv (as
S or P) are achievable in many cases.
Chemiluminescence—Chemiluminescence detection involves the measurement of
light emitted from a compound which has been activated by a chemical process. The only
chemiluminescent process to be discussed here is the detection of nitrogen compounds by
conversion of the compound to NO and subsequent reaction with ozone. The chemical
reactions involved are as follows:
(pyrolysis)
R - N + O2 * CO2 + H2O + NO
NO + O3 » NO2* + O2
NO2* * NO 2 + light
While this approach is highly selective for nitrogen, it does not distinguish between
inorganic and organic nitrogen and, hence, cannot be used to detect trace quantities of
organic nitrogen compounds in air. As yet, no commercially available systems involving
GC preseparation are available. In order to use this technique for determining trace
organic nitrogen compounds in air, a method for removing the NOX and other inorganic
nitrogen compounds must be found.
Summary—
While a large number of screening devices and techniques are available, as described
in this document, one must carefully evaluate the sensitivity, selectivity, and types of
interferences of these techniques before deciding whether or not they are suitable for use
in a particular monitoring situation. Many of the available techniques are designed for
workplace monitoring where the levels of interest are greater and the environment is
better known and controlled than is generally the case for ambient air monitoring.
In many cases none of the available screening techniques will be suitable to
accomplish a given monitoring objective, and, hence, a more sophisticated analytical
approach will be required. In selected cases the existing screening devices may be
suitable if some means of sample preconcentration or cleanup is employed. However,
extensive sample workup in the field is generally cumbersome and probably offers little
advantage over conventional laboratory analysis of the sample.
55
-------
The ionization devices (particularly FID and PID) probably are the most generally
useful in terms of sensitivity for ambient air monitoring, although colorimetric methods
can be quite good in favorable cases. A further advantage of the ionization and FPD
approaches is the availability of these techniques in combination with gas chromatography
which extends the qualitative as well as quantitative capabilities of the techniques. The
IR spectroscopic approach can give useful qualitative information not available for the
other techniques but will not be useful for monitoring ambient levels of many trace
components because of spectral overlap from major organic and inorganic components.
The lack of a suitable GC/IR for field use is a further limitation of this approach.
Laboratory Screening (Compound Class) Techniques
The techniques discussed in this section give primarily compound class information,
rather than data for specific compounds. Many techniques discussed in the last section fit
this description, although the current discussion will center on techniques primarily
designed for laboratory rather than field use.
Techniques in this category include the following:
• Colorimetric techniques
• Infrared spectroscopy
• Fluorescence spectroscopy
• Low resolution mass spectrometry.
These techniques can be used to obtain compound class information as well as to
determine concentrations of specific compounds. Consequently compound specific
methods based on these techniques are presented later in the TAD to avoid confusion.
Colorimetric Techniques-
Table 9 provides information concerning colorimetric techniques useful for moni-
toring toxic organic compound classes. Relatively few compound class methods for
ambient air have been developed in the last few years because of the advances in chro-
matographic procedures and the intense interest in determining specific organic
compounds. However, colorimetric detection principles exist for many organic compounds
and ambient air monitoring techniques based on these principles could be developed if a
particular need arose.
Most colorimetric techniques involve collection of the compound(s) of interest in an
impinger with subsequent formation of a colored derivative, in either the field or labora-
tory. The content of the impinger solution should be chosen to (a) stabilize the analytes
during collection and transport, and (b) selectively capture the analytes of interest,
allowing potential interferences to pass through.
Infrared Spectroscopy (IR)—
IR is widely used as a screening tool for determining organic compound classes from
source emissions in the EPA Level 1 protocol^). However, this approach is generally
less sensitive than available colorimetric approaches and will be of value in ambient air
monitoring only when high (ppm) levels of semivolatile or nonvolatile compounds are
present.
The Level 1 IR approach cannot be used for volatile compounds (boiling point 100 C)
because of the need for solvent extraction. Gas phase IR approaches such as those
56
-------
TABLE 9. COLORIMETRIC SCREENING TECHNIQUES FOR DETERMINING
CLASSES OF ORGANIC COMPOUNDS
Analyte
Aldehydes
Detection
Principle
Reaction with 3-methyl-2-
Detection
Limit
0.03 ppmv
Reference
(4)
Comments
Mercaptans
Phenols
Amines
(primary and secondary)
Polynuclear
Aromatic
Hydrocarbons
(PAH)
benzothiazolone
hydrazone (MPTH). Further
reaction with ferric
chloride-sulfamic acid to
form an azo dye. Colorimetric
detection at 628 nm.
(25 liter sample)
0.002 ppmv
(25 liter sample)
Mercaptans trapped with
aqueous mercuric acetate-
acetic acid solution.
Determined colorimetrically
at 500 nm after reaction
with N,N-dimethyl-p-
phenylenediamine.
Phenols are captured in an
alkaline solution, then
reacted with 4-aminoanti-
pyrine. Colorimetric
determination at 400 nm.
Amines are trapped in
acidified isopropanol
solution & subsequently
reacted with ninhydrin.
Colorimetric determination
at 575 nm.
PAH are collected on"Hi-Vol" 10 ng/m3
filter & extracted with
benzene. Benzene extract
is evaporated to dryness &
partitioned by alumina
chromotography. PAH are
determined colorimetrically.
(4)
0.001 ppmv
(1 m3 sample)
0.1 ppmv
(25 liter sample)
(4)
(4)
(2)
NIOSH
Method
184
\\2$ will give a slight
response.
(200 fold lower)
Phenols with substi-
tuents at the 4-
(p-) position will
not be detected.
Fractions from aluminia
column are analyzed
to obtain individual
PAH values or combined
to indicate total PAH
content.
-------
described for field screening could be employed in the laboratory but the problems
associated with sample transport make this approach less attractive. Long path IR
approaches for determining specific organic compounds are discussed later in the TAD.
Low Resolution Mass Spectrometry (LRMS)—
LRMS is a second technique used in the EPA Level 1 protocol(24). This approach
suffers the same limitations as IR, when applied to ambient air monitoring. While the
LRMS approach is somewhat more sensitive than IR, the data obtained are more difficult
to interpret and the instrumentation required is more expensive. Consequently, this
approach is of little value for ambient air monitoring.
Fluorescence Spectroscopy—
Fluorescence is a relatively inexpensive and highly sensitive technique useful for
determining selected classes of organic compounds. The principles of fluorescence spec-
troscopy, as applied to air monitoring are discussed in Katz' book'"), Polynuclear
aromatic hydrocarbons (PAH), aromatic amines, and phenols are the primary classes of
organic compounds in ambient air which can be determined using fluorescence
spectroscopy.
A practical limitation of fluorescence spectroscopy is the quenching phenomenon,
wherein large quantities of UV absorbing material will reduce the fluorescence response.
Consequently, large quantities of UV absorbing material must be removed from the
sample extract (e.g., by adsorption chromatography) before a fluorescence determination
can be made. A second limitation is the wide variability in fluorescence response from
compound to compound, even within a given class. This limitation makes development of
a quantitative fluorescence method for compound classes difficult. As a result of these
two limitations most available fluorescence methods involve chromatographic isolation of
a particular compound prior to its determination. Methods of this type are discussed later
in the TAD.
Compound Specific Techniques
A listing of the techniques available for determining specific organic: compounds is
given below:
Gas chromatography (GC)
Gas chromatography-mass spectrometry (GC-MS)
High performance liquid chromatography (HPLC)
Thin layer chromatography (TLC)
Column chromatography
Infrared spectroscopy (IR)
Ultraviolet or visible absorption (UV-VIS)
Fluorescence spectroscopy.
The operating principles as well as advantages and limitations of the various techniques
are discussed below.
Gas Chromatography (GC)—
GC is by far the most widely employed technique in ambient air monitoring of toxic
organic compounds. The sensitivity, specificity, and versatility of GC, coupled with the
relatively volatile nature of most compounds in ambient air, make this a very attractive
technique.
58
-------
The principles of GC are described in Katz' book(26). The discussion given below is
designed to present those aspects of GC operation important to ambient air analysis. One
should consult the above reference or chromatography text books for more detailed infor-
mation. Although the operation of GC is rather simple, many practical problems can
arise which are avoided or corrected only by experienced operators. Consequently the
availability of such personnel is essential to the success of a monitoring program.
Obviously, this statement applies to other instrumental analysis techniques as well.
Basically GC is a separation technique wherein components of a sample are sepa-
rated by differential distribution between a gaseous mobile phase (usually helium,
nitrogen, or hydrogen carrier gas) and a solid or liquid stationary phase held in a glass or
metal column. Sample is injected into the carrier gas as a sharp plug and individual
components are detected as they elute from the column at characteristic "retention
times" after injection. Both column temperature and carrier gas flow must be carefully
controlled to obtain uniform response and retention time characteristics. The basic
components of a gas chromatograph are the following:
Gas flow system
Injector system
Stationary phase (column)
Oven and associated temperature controller
Detector
Data recording device.
Only columns, detectors, and injectors will be discussed in detail in the TAD. A typical
arrangement for the various GC components is shown in Figure 7.
Columns—In order to understand the requirements for the various GC components,
one must recognize that two different types of GC columns (illustrated in Figure 8) are
widely used. The conventional type of GC column is termed a packed column and consists
of a solid support coated with a liquid stationary phase (gas/liquid chromatograph) or
simply a solid adsorbent (gas/solid chromatography). The second type of column is a wall-
coated open tubular (WCOT) or capillary column and has been widely adopted for environ-
mental analysis because of its superior resolution and broader applicability, compared to
packed column GC. As shown in Figure 8, the capillary column consists of a liquid
stationary phase coated or bonded to the specially treated glass or fused silica tubing.
Fused silica tubing is most commonly used because of its physical durability. Bonded (or
crosslinked) columns are used in preference to coated columns because of the greater
operating temperatures which can be obtained. Columns of this type are available from
various suppliers, including Hewlett-Packard and J & W Scientific. The principles and
applications of capillary column GC are discussed in the literature^ljSZ).
The typical linear flow velocity through both capillary and packed columnns is about
30 cm/sec. However, the much narrower cross sectional area of the capillary column
translates to a volumetric flow rate of 1-2 cm-^/minute versus 30-50 cm^/minute for
packed column. Th much lower flow rate of the capillary column requires (a) that dead
volume in the injector and detector be minimized and (b) that the effective sample
injection volume be reduced. These requirements have been fulfilled satisfactorily in
most modern GC instruments through the use of low volume detector cells and
split/splitless type injection systems. Consequently, modification of older packed column
instruments for capillary column operation is not recommended unless no alternative is
available.
59
-------
INJECTION
PORT
HEATED
DETECTOR
(HEATED)
COLUMN
OVEN
DATA
RECORDING
DEVICE
Figure 7. Block diagram of a typical gas chromatograph
60
-------
PACKED
.••.«,.'/':••.. -. •>?•,«*.-•. :i \\
:-. -•.•;;. *./ .-.• i» iV.-*>-«Vv
Glass Liquid Stationary
Phase Coated on
Stationary Support
Typical Dimensions
6mm O.D. x 2mm I.D. x 1.8m
CAPILLARY
Liquid Stationary
Phase Coated on
Fused Silica or
Glass Wall
Typical Dimensions
0.50mm O.D. x 0.22mm I.D. x 50m
Figure 8. Capillary and packed gas chromatographic columns
61
-------
Table 10 lists the most commonly employed packed and capillary columns and their
respective applications. While capillary column GC is superior for most situations, highly
volatile compounds (boiling points <50 C) are best determined on various packed columns
capable of retaining such compounds at ambient temperature.
Detectors — The GC detection system chosen will determine, to a large extent, the
specificity and sensitivity of an analytical method. Consequently one should carefully
review the program requirements when selecting a detection system for use in a given
study. Table 11 provides an overview of the characteristics of available detection
systems and Figure 9 graphically illustrates the approximate sensitivity range for various
GC detectors. The principles of operation for many of the detectors are briefly described
in the section on field monitoring techniques.
By far the most widely used GC detector is the flame ionization detector (FID)
which provides good sensitivity and uniform response for a wide variety of compounds.
The uniformity in response of the FID allows fairly accurate quantification of compounds
for which specific response factors have not been determined. This is probably the most
advantageous feature of the FID and is useful for ambient air monitoring studies in which
the complete range of compounds present cannot be anticipated. None of the other
detectors give nearly as uniform a response as FID.
The electron capture (ECD) and photoionization (PID) detectors are becoming more
widely used in ambient air monitoring because of their greater sensitivity, compared to
FID. PID is widely viewed as a substitute for FID, although its response varies consider-
ably from compound to compound. PID is particularly useful for monitoring aromatic
compounds. ECD is selective for electron deficient materials, especially polyhalogenated
and nitrosubstituted compounds. In favorable cases, picogram (10~^ g) levels can be
detected, but response varies widely from one compound to the next. Several literature
references are available which describe the detection sensitivities of various compounds
on ECD and
The flame photometric (FPD) detector is specific for either phosphorus or sulfur
compounds, depending on the optical filter employed and gives sensitivity comparable to
the FID. The response obtained is logarithmically related to the amount of sulfur or
phosphorus injected (log sulfur amount = response). This detector is of particular
advantage when trace levels of phosphorus or sulfur compounds are to be detected in the
presence of much greater quantities of aliphatic hydrocarbons.
The alkali flame (AFD), thermionic specific (TSD), or nitrogen-phosphorus (NPD)
detector is selective for nitrogen and phosphorus compounds and is considerably more
sensitive than either FPD or FID. However, response varies considerably from compound
to compound. This detector is preferred over FPD for determining phosphorus compounds
in many cases because of its greater sensitivity.
The Hall electrolytic conductivity detector (HECD) can be operated in either the
halogen, sulfur, or nitrogen specific modes, although the halogen specific mode is the
most widely employed. This system gives poorer sensitivity than ECD for polyhalogenated
compounds, but better response for monohalogenated compounds, such as vinyl chloride,
since the HECD response is proportional to the amount of halogen injected. Corre-
spondingly, the HECD gives a more uniform response for polyhalogenated compounds than
ECD. An additional advantage of HECD is its temperature stability which allows temper-
ature programming (ECD drifts considerably under temperature programmed conditions).
62
-------
TABLE 10. COMMONLY USED GC STATIONARY PHASES
Liquid Phases
SE-30, OV-1 (Methyl Silicones)
OV-17, SE-54 (Methyl/Phenyl
Silicones)
Carbonax 20M (Polyethylene Glycol)
FFAP, SP-1000 (Polyethylene Glycol
Terephthalate)
Solid Phases (Packed Column Only)
Chromosorb, Porapak Series
(Styrene/Di vi nylbenzene
Polymers)
Carbon Molecular Sieves
(Carbosphere, Spherocarb,
Carbosieve)
Porous Silica (Unibeads, Pprasil A)
Applications
Hydrocarbons, Chlorinated
Hydrocarbons
PAHs, Chlorinated Pesticides,
Hydrocarbons
Polar Compounds; Esters,
Alcohols, etc.
Phenols, Volatile Acids
Volatile Alcohols, Ketones,
Hydrocarbons, and Halocarbons,
(B.P. 30 to 100 C)
CI-GS Hydrocarbons
Hydrocarbons
63
-------
TABLE 11. COMMON GC DETECTORS
Detector
Flame lonization (FID)
Compounds
Detected
Organics
Advantages
Response is relative con-
Limitations
Not as sensitive as many
CTl
Photoionization (PID)
Electron Capture (ECD)
Alkali Flame (NPD or
AFD)
Flame Photometric
Detector (FPD)
Hall Electrolytic
Conductivity (HECD)
Thermal Energy
Analyzer (TEA)
Most organics except
methane
Polyhalogenated and nitro
compounds
Nitrogen and phosphorus
compounds
Sulfur or phosphorus com-
pounds (separately)
Halogen, sulfur, or
nitrogen compounds
(separately).
Nitrosamines
stant from one compound
to another.
Response selectivity can be
varied by choice of lamp
energy.
Highly sensitive and
selective.
of the other detectors.
Response varies from com-
pound to compound. Less
rugged than FID.
Subject to contamination.
Response varies widely
from compound to compound.
Response drifts during
temperature programming.
Subject to contamination.
Response varies more from
day to day than for FID
or ECD.
Response varies from day to
day. Not as sensitive
as NPD for phosphorus
compounds.
Highly selective. Response Only halogen mode has been
relatively constant from employed extensively. Re-
compound to compound. Stable quires considerable operator
baseline during temperature attention. Not as sensitive
programming. as ECD for polyhalogenated
compounds (e.g., PCBs).
Highly sensitive and
selective.
Highly selective.
Highly sensitive and
selective.
Very expensive.
(•v-$35K)
-------
Detector
FID
PID (a) 10.2 eV
(b) 11.7 eV
ECD (Ni63) ,
HECD
NPD
FPD
I 1
10-13 iQ-11 1Q-9 10~7 1(T5 1(T3
Quantity of Material Injected, g
Figure 9. Approximate Gas Chromatographic Detector Ranges
65
-------
HECD is not widely used in air monitoring programs because the greater sensitivity of
ECD is highly desirable and tends to outweigh other considerations.
A continuing challenge in GC analysis of organic compounds in ambient air is the
qualitative identification of components detected by the various detection systems. Use
of mass spectrometry as a GC detector (GC-MS) is a highly effective tool in this regard,
as discussed later in this section. However, GC-MS is a very expensive instrument and is
not always available. Attractive alternatives to GC-MS include analyzing samples twice
using dissimilar columns or using dual (parallel or series) detectors. The latter approach is
most attractive since only a single injection of sample is required and retention times can
be accurately matched from the two outputs. If series detectors are employed, the first
detector must be sealed, nondestructive towards the sample, and have a low dead volume
to reduce peak broadening. On the other hand, parallel detector operation requires a
precise stream splitter. Such a system has been described for simultaneous FID/PID
detection using capillary columns GC'"*'.
Some detector combinations found to be most useful in terms qualitative infor-
mation are listed in Table 12. Note that each detector combination contains one selective
(ECD, HCD, or NPD) and one relatively nonselective (PID or FID) detector. This approach
allows the selective and sensitive detection of compounds of special interest as well as
the detection of any large, unanticipated components. The ratio of response for a given
peak on the two detectors helps confirm its identity.
Injection Systems—Injection systems are of particular importance in GC analysis of
ambient air, because of the wide range of forms in which the sample may be collected.
The primary injection modes, based on the manner of sample collection include:
• Liquid injection (particulate or resin extract)
• Whole air injection (syringe or valve)
• Thermal desorption (e.g., of Tenax)
• Cryogenic trapping of whole air sample.
Liquid and whole air injections are relatively straightforward and employ con-
ventional techniques. In the case of capillary column GC a well designed splitless injector
is required in order to obtain optimal resolution and prevent solvent tailing. Generally the
capillary column temperature is maintained well below the elution temperature of the
analyte in order to reduce peak broadening.
Thermal desorption of resin samples involves (a) heating the resin cartridge to
suitable temperature in an inert gas stream, (b) passing the inert gas stream through a
cryogenically cooled trap to condense the desorbed analytes, and (c) flash evaporating the
analytes onto a GC column held at low temperature to "focus" the materials at the head
of the column. This injection process is used for Tenax desorption work, using a semi-
automatic device available from Nutech, Inc. (Durham, N.C.) and has been described in
the literature^27).
The whole air cryogenic sampling procedure is similar to the process described
above except that thermal desorption of the resin is omitted. As described in the
literature^) a 30-500 ml volume of air is passed through a cryogenically cooled trap
(liquid argon) and the condensed material is flash evaporated onto a GC column held at
low temperature (e.g., -50 C). In both cases, the desorbed components are separated by
temperature programing the GC column.
66
-------
Table 12. Useful Dual GC Detector Combinations
Detector Combination
ECD-PID (Series)
PID-FID (Series)
ECD-FID (Series)
PID-HECD (Series)
NPD-FID (Parallel)
FPD-FID (Parallel)
PID-NPD (Series)
HECD-FID (Parallel)
Application
Volatile Aromatics and
Chlorinated Solvents
Aromatic Hydrocarbons
Chlorinated Hydorcarbons
Aromatic and Halogenated
Volatile Compounds
Nitrogen or Phosphorous
Compounds
Sulfur Compounds
Amines
Hydrocarbons and
Halogenated Hydrocarbons
Reference
Unpublished data
61
67
Unpublished data
65,69
65
66
64, 68
67
-------
Both the resin and cryogenic thermal desorption approaches require the use of non-
adsorptive stainless steel or teflon valves and fittings capable of withstanding the
extreme operating temperatures without leaking or contaminating the system. Capillary
columns should be directly coupled to the injection valve using teflon and stainless steel
fittings, as opposed to using conventional splitless injectors.
Gas Chromatography-Mass Spectrometry (GC-MS)—
In principle GC-MS can be viewed as another detection mode for gas chromato-
graphy. However, in practice GC-MS is a highly sophisticated and powerful technique
requiring more specialized skills and detailed knowledge than for other forms of GC. A
detailed discussion of GC-MS theory and practice is beyond the scope of this document. A
comprehensive discussion of GC-MS is available in textbook form'^°% although this
reference does not discuss the most recent advances in the rapidly expanding field of
capillary column GC/MS.
Instrumentation—A diagram of a typical GC-MS system is shown in Figure 10. In
addition to the GC system the major components include:
Interface (between GC column and MS ion source)
Ion source
Mass analyzer
Vacuum system(s)
Data acquisition and instrument control system.
Mass spectrometric analysis involves the following sequential processes:
(a) lonization of the sample introduced into the ion source
(b) Acceleration of the ions from the ion source into the mass
analyzer region
(c) Separation of the ions in the mass analyzer region on the basis
of the mass to charge (m/e) ratio
(d) Detection of the ions at each mass.
In order to minimize ion-molecule reactions and interference from air the ion source,
mass analyzer, and ion detector must be maintained at low pressure (10~^ mm of Hg or
less). Consequently, the high vaccum system is a critical component of the GC-MS
system and air leakage through valves, fittings, and/or flanges, is a common problem in
GC-MS.
Since a GC operates at ambient pressure and above, whereas the MS must be main-
tained at low pressure, the GC-MS interface ideally will discard excess carrier gas flow
and quantitatively deliver the sample components into the ion source for subsequent
determination. Since typical GC-MS systems can accept approximately Z-3 ml/minute of
gas flow without significant spectral distortion (due to elevated pressure in the MS
system) the use of capillary column GC allows the entire column effluent to be delivered
to the ion source. If fused silica columns are used, the column can actually be placed
directly into the entrance of the ion source. However, packed column GC-MS operation,
in which column flow rates of 30-50 ml/minute are common, requires the use of mem-
brane or jet separators which remove a large portion of the light carrier gas (usually
helium) delivering the majority of the sample to the ion source.
68
-------
DATA ACQUISITION
AND INSTRUMENT
CONTROL SYSTEM
Figure 10. Block diagram of a typical GC-MS system
69
-------
The avoidance of membrane and/or jet separators and associated fittings, in addition
to the other inherent advantages of capillary column GC (i.e., better resolution, less
irreversible adsorption) make capillary column GC-MS preferable to packed column GC-
MS for most environmental applications. In fact many of the newer commercial GC-MS
systems are designed primarily for capillary column operation.
Although a wide variety of powerful, specialized forms of MS ionization modes,
mass analysis schemes, etc. are available, the majority of ambient air monitoring appli-
cations use electron impact (70 eV) ionization and quadruple mass analyzers. One must
bear in mind that low resolution mass spectral identifications (e.g., from quadrupole mass
analyses) are considered tentative and for litigation purposes high resolution spectral data
may be required. A discussion of the various optional approaches is provided in the previ-
ously referenced text^S).
GC-MS analysis of complex environmental samples results in the acquisition of large
amounts of data. A typical capillary column GC-MS run might result in the acquisition of
2000 mass spectra (e.g., one scan per second) over a range of 40-400 amu, representing
100-200 individual components. Obviously the acquisition and reduction of this volume of
data requires the use of a laboratory computer system operating with a set of sophisti-
cated software. Typical GC-MS systems use laboratory mini-computers with high density
magnetic disks for mass storage. Archival storage of data is usually accomplished using
9-track magnetic tape, a more convenient medium than disks for long-term storage.
A new form of mass spectrometry, termed MS-MS or triple quadruple MS has
recently become available and is receiving much attention™). This technique involves
direct analysis of an atmospheric sample by (a) ionizing the sample components in an
atmospheric pressure ion source, (b) separating the ions in the first quadruple mass
analyzer, (c) passing an ion of one specific mass into a second quadrupule containing a
collision gas such as argon, and (d) separating the product ions resulting from the ion
molecule reaction in Step c using a third quadruple. The use of two mass analyzers in
series introduces great selectivity into the analysis, hence eliminating the requirement for
GC separation in many cases. In practice the MS-MS approach is highly sophisticated and
expensive thus limiting its availability for most ambient air monitoring programs.
However, in selected cases MS-MS may be the only technique capable of detecting trace
levels of contamination.
Applications—The power of GC-MS analysis has resulted in a wide variety of appli-
cations, especially in the environmental area. Reviews'30) and research articles
describing these applications are available throughout the scientific literature and no
attempt will be made to summarize that information here.
GC-MS analysis can be used effectively for ambient air monitoring in two gener-
alized operating modes. The first mode involves surveying collected samples to determine
what compounds are present and the approximate concentrations. A typical example of
this operating mode is the widely used Tenax resin collection/GC-MS analysis method for
ambient air surveys^7). This operating mode takes advantage of the powerful identifi-
cation capability of GC-MS and is often used in conjunction with other GC detection
schemes, wherein the sample components quantified by FID, ECD, etc. are identified by
GC-MS.
A second mode of GC-MS operation involves searching a sample for trace levels of
specific compound(s) and accurately quantifying this material. This operating mode
70
-------
typically employs a form of GC-MS termed selected ion monitoring (SIM) or multiple ion
detection (MID) wherein only a few selected ions are monitored, rather than scanning a
broad mass region as in the first operating mode. The SIM approach is more sensitive and
accurate than full spectrum scanning approaches.
High Performance Liquid Chromatography (HPLC)—
HPLC is a technique primarily designed for the determination of polar and non-
volatile compounds not readily determined by gas chromatography. Consequently HPLC is
not widely used for ambient air monitoring, in which most compounds of interest are
relatively volatile. Two specialized cases where HPLC is useful for ambient air moni-
toring are:
(a) Determination of nonvolatile material (e.g., PAHs) in ambient
participate material
(b) Determination of reactive components (e.g., aldehydes) by reacting
the compounds with a derivatizing reagent.
The theory and application of HPLC has been described in detail in textbook form (33).
A typical configuration of an HPLC system is shown in Figure 11. The important
components include:
Pump
Injection valve
Column
Detector
Data recording device.
HPLC separation involves a process similar to that described for GC (i.e., separation of
sample components on the basis of the differential distribution between mobile and
stationary phases). However, GC separations are performed primarily on the basis of
volatility, whereas HPLC separation mechanisms are much more varied and complex. The
liquid mobile phase composition has a profound effect on component retention and a
change in mobile phase composition can lead to reversal of the elution order for two
components.
The most commonly employed forms of HPLC are termed (a) normal phase and (b)
reversed phase chromatography. Normal phase HPLC uses a polar stationary phase (e.g.,
silica gel, alumina) and a relatively nonpolar mobile phase (e.g., hexane, methylene
chloride) and separation is based primarily on an adsorption mechanism, with polar
components eluting later than nonpolar components. Reversed phase HPLC was
developed later than normal phase HPLC and uses a nonpolar (usually chemically bonded)
stationary phase and polar (e.g., water, methanol, acetonitrile) mobile phase. Reversed
phase HPLC separates primarily on a partition mechanism, with hydrophilic components
eluting earlier than hydrophobic components.
At the present time reversed phase HPLC is more widely used than normal phase
HPLC because of the greater versatility of this technique. Table 13 compares the
principles and relative advantages of the two approaches.
In order to achieve rapid, efficient separation of sample components, HPLC tech-
niques employ microparticulate packing materials, typically spherical materials of 5 Mm
diameter. HPLC requires a pumping system capable of delivering a precise, pulse-free
71
-------
SOLVENT
RESERVOIR
PUMP
tNJECTION
VALVE
PACKED
COLUMN
(Typically
25cm x 4.6mm I.D.)
I
DETECTOR
DATA
RECORDING
DEVICE
Figure 11. Typical configuration of an HPLC system
72
-------
TABLE 13. COMPARISON OF REVERSED AND NORMAL PHASE HPLC
Typical Stationary
Phase
Typical Mobile
Phase
Elution Order
Advantages
Typical
Environmental
Applications
Reversed
Phase
Normal
Phase
bonded to silica
Water, methanol, acetonitrile
Silica gel, cyanopropyl
bonded to silica
Hexane, ether, methylene
chloride
Water soluble compounds first Nonpolar compounds first
Usually not deactivated by
samples. Avoids problems with
irreversible adsorption.
Readily equilibrated during
gradient elution.
PAHs, aromatic amines, phenols,
pesticides
Useful for water labile
compounds. Can separate
isomeric compounds more
readily.
PAHs, nitrogen heterocycles
(especially for isomer
separation), pesticides
73
-------
flow (e.g, 1-2 ml/minute) of mobile phase at pressures of up to 400 atmospheres.
Obviously, high pressure injection valves, columns, and fittings are also required. Because
of the relatively low flow rates involved (compared to GC) "dead volume" in fittings,
injection valves, and detector cells must be minimized to maintain good separation
efficiency.
A variety of detection systems are available for HPLC. The readily available
detectors of value in ambient air monitoring are listed in Table 14. By far the most
widely used HPLC detector is the UV detector (either fixed or variable wavelength).
Unfortunately no sensitive, universal detector (equivalent to GC/FID) is available for
HPLC. This limitation tends to reduce HPLC to the role of a specialized analytical tool
for determining selected classes of compounds, at least in the ambient air monitoring
area.
Thin Layer and Column Chromatography—
Column chromatography and thin layer chromatography (TLC) are widely used as
"cleanup" techniques prior to GC-MS, GC, or HPLC determination and are no longer
widely used as determinative techniques in ambient air monitoring. The few methods
where these techniques have been used for compound determination, such as PAH
analysis(34)? have been largely replaced by HPLC methods. However, in some cases (e.g.,
benzo(a)pyrene analysis in ambient air) TLC has remained a viable technique because of
its lower cost compared to HPLC. A review of TLC techniques is provided in Katz'
book(35). Both TLC and column chromatography are forms of adsorption chromatography.
Spectroscopic Techniques—
Spectroscopic techniques of potential value for compound specific analysis include
infrared spectroscopy (IR), fluorescence spectroscopy, and ultraviolet-visible absorption
(UV-VTS) spectroscopy. However, UV-VIS and fluorescence spectroscopy are not suf-
ficiently selective to be of value except in combination with chromatographic procedures.
Consequently these two techniques are used as online HPLC detectors as well as offline
determinative procedures in combination with GC, TLC or HPLC. PAH methods using
these approaches have been described^").
IR is also useful as an online or offline determinative tool. The advent of rapid
scanning Fourier transform IR (FTIR) techniques has improved instrumental sensitivity
and analysis speed, making online GC/FTIR techniques feasible^?). However, GC/FTIR is
still undergoing considerable development and cannot be considered a routine technique at
this time.
Long path (kilometer) FTIR techniques have been used to detect individual
compounds (e.g., formaldehyde) in ambient air(38). This technique has the advantage that
sample collection is not required. However, this is not a readily available technique and
hence is of limited value in ambient air monitoring programs.
METHODS FOR SPECIFIC COMPOUNDS AND COMPOUND CLASSES
This section of the TAD presents information concerning specific analytical methods
for selected compounds and compound classes of particular concern in ambient air
monitoring. The first portion of this section defines the classification of organic
compounds into specific groups based on chemical structure and the second part presents
the sampling and analytical methods of primary importance.
74
-------
Table 14. HPLC Detectors
Detector
Ultraviolet (UV),
Fixed or Variable
Wavelength
Fluorescence
Electrochemical
Refractive Index
Approximate
Sensitivity,
Amount Injected
l(T6
g
Application
PAH, phenols,
nitrogen
heterocycles
PAH, aromatic
amines, phenols
Aromatic amines,
phenols
Too insensitive
for environmental
applications
75
-------
Definition of Compound Classes
In order to minimize the number of methods required for a given monitoring
program, one should attempt to use methods which apply to multiple compounds, or even
multiple classes of compounds. As an aid in this selection process, Table 15 presents a
broad classification of organic compounds of particular concern in ambient air monitoring
studies. Table 16 presents data indicating the range of concentration of selected organic
compounds in ambient urban air''*).
As shown in Table 15 organic compounds can be roughly divided into three
categories - neutral, basic, and acidic. Basic compounds are readily protonated to form
cations whereas acidic compounds readily deprotonate to form anions. lonizable (acidic or
basic) compounds usually require specialized analytical methods (e.g., control of pH) and
hence represent a more challenging problem in ambient air monitoring programs. These
compounds generally are less volatile than neutral compounds and have a tendency to
adhere to surfaces, including atmospheric particles, through adsorptive, ionic, or con-
densation processes.
Neutral Compounds—
Neutral compounds are more commonly encountered than ionic compounds in
ambient air monitoring programs because of greater volatility. Hydrocarbons, compounds
containing only carbon and hydrogen, are generally the most abundant compounds in
ambient air. C} (methane) to C\Q hydrocarbons are present at detectable concentrations
in most urban as well as rural sites. Such compounds are derived from natural sources
such as trees and other vegetation as well as from petroleum processing and utilization.
In areas of high gasoline emissions, the aromatic compounds, benzene and toluene, are
among the most abundant hydrocarbons (other than methane which has a global back-
ground concentration of ~2 ppm).
Halogenated hydrocarbons are commonly encountered in urban, but not rural, areas
since these compounds are derived exclusively from anthropogenic sources. Solvents from
manufacturing, and drycleaning operations (methylene chloride, trichoroethylene,
perchloroethylene, etc.) represent the most common volatile halogenated hydrocarbons.
These compounds are of considerable toxicological importance and hence are commonly
determined in ambient air monitoring programs. PCBs, PCNs, and halogenated pesticides
represent semivolatile halogenated hydrocarbons of considerable environmental
significance.
Oxygenated compounds are present through direct emission into the air as solvents
(e.g., methane, ethanol, acetone, etc.) as well as through the atmospheric oxidation of
olefinic hydrocarbons. For example, the interaction of olefins and ozone leads to the
formation of aldehydes, especially formaldehyde.
Organic nitrogen compounds are primarily introduced into the atmosphere from
chemical production or processing operations although interaction of nitrogen oxides with
other organic compounds can occur. For example, interaction of PAHs with nitrogen
dioxide can lead to the formation of nitro PAHs, which are believed to be highly car-
cinogenic. Compounds in this class have diverse properties, thereby complicating the
analysis process.
Sulfur and phosphorus containing organic compounds are present in ambient air
almost exclusively through direct emission into the atmosphere. Significant levels of
76
-------
Table 15. Classification of Organic Compounds for Ambient
Air Monitoring Studies
Compound Class
Neutral Compounds
Hydrocarbons
Halogenated Hydrocarbons
Oxygenates (alcohols, aldehydes)
Nitrogen Compounds
Sulfur Compounds
Phosphorous Compounds
Basic Compounds
Aromatic Amines
Aliphatic Amines
Nitrogen Heterocycles
Acidic Compounds
Phenols
Sulfonic Acid
Carboxylic Acids
Typical Compound(s)
Hexane, benzene
Trichloroethylene, chlorobenzene, PBCs
Methanol, formaldehyde
Nitrobenzene, acrylonitrile
dimethylnitrosamine
Thiophene, dimethylsulfide
Triphenylphosphate
Aniline, toluenediamine
Dimethylamine
Pyridine
Phenol, cresol, nitrophenol
p-toluene sulfonic acid
Acetic acid
77
-------
Table 16. Measured Concentrations of Organic
Compounds in Urban Air*
Compound
Concentration (Parts Per Trillion)
MIN MAX
Chlorofluorocarbons
Trichlorofluoromethane (Fll)
Dichlorofluoromethane (F12)
Trichlorotrifluoroethane (F113)
Dichlorotetrafluoroethane (F114)
Halomethanes
Methyl chloride
Methyl bromide
Methyl iodide
Methylene chloride
Chloroform
Carbon tetrachloride
Haloethanes and halopropanes
Ethyl chloride
1,1 Dichloroethane
1,Z Dichloroethane
1,2 Dibromoethane
1,1,1 Trichloroethane
1,1,2 Trichloroethane
1,1,1,2 Tetrachloroethane
1,1,2,2 Tetrachloroethane
1,2 Dichloropropane
Chloroalkenes
Vinylidene chloride
(cis) 1,2 Dichloroethylene
Trichloroethylene
Tetrachloroethylene
Allyl chloride
Hexachloro-1,3 butadiene
Chloroaromatics
Monochlorobenzene
a-Chlorotoluene
o-Dichloro benz ene
m-Dichlorobenzene
1,2,4 Trichlorobenzene
200
380
22
12
430
7
0.2
49
19
110
<4
21
5
34
<5
1
9
<5
1
1
1
1,900
3,100
2,200
62
2,300
1,000
7.0
9,400
5,100
2,900
10
8
45
8
130
<5
2
2
11
1,200
150
2,500
370
2,700
900
89
77
250
220
600
2,500
7,600
<5
150
2,800
110
230
55
40
78
-------
Table 16. (Continued)
Aromatic hydrocarbons
Benzene 110 37,000
Toluene 110 65,000
Ethyl benzene 50 19,000
m/p Xylene 110 23,000
o-Xylene <10 9,700
4-Ethyl toluene 60 7,400
1,2,4 Trimethyl benzene 50 15,000
1,3,5 Trimethyl benzene 30 5,300
Oxygenated species
Formaldehyde 6,600 41,000
Phosgene <20 <20
Peroxyacetylnitrate (PAN) <10 5,800
Peroxypropionylnitrate (PPN) <10 900
*Summarized from Reference 71.
79
-------
sulfur or phosphorus organic compounds are generally encountered only in the vicinity of
emission sources.
Basic Compounds—
Basic compounds include strongly basic aliphatic amines, weakly basic nitrogen
heterocycles, and aromatic amines which tend to have intermediate basicities. In this
document, the term "aromatic amine" applies only to compounds in which the amine group
is bonded directly to the aromatic ring (e.g., aniline). Compounds containing an amine
group attached to an aliphatic carbon are termed "aliphatic amines" and will be more
basic than aromatic amines.
Aromatic amines, especially aniline, are widely used in dyestuff and related chemi-
cal industries and are present in the atmosphere primarily through direct emission.
Nitrogen heterocycles, especially pyridine, are used as a solvents and for other appli-
cations in the chemical industry. These compounds can be present at trace levels in
petroleum products and liquids derived from coal. Aliphatic amines are used for a wide
variety of applications and are present in the atmosphere through direct emission, either
from man-made or biological sources.
Acidic Compounds—
Acidic compounds include strong acids, carboxylic and sulfonic acids, as well as
weak acids - phenols. Phenols are widely used as intermediates in the chemical industry
and as disinfectants. Phenols are emitted from various combustion sources and through
the atmospheric oxidation of aromatic compounds. Trace levels of phenol as well as other
phenolic compounds (e.g., cresols) are commonly encountered in ambient air.
Carboxylic acids are derived from natural as well as anthropogenic sources, whereas
sulfonic acids are primarily derived from anthropogenic sources. Carboxylic acids can
also be produced by the atmospheric oxidation of organic compounds. These compounds
would be expected to be readily removed from the atmosphere through dissolution in
precipitation or adsorption onto particles. Analytical difficulties have hampered the
determination of strongly acidic compounds in air.
Discussion of Specific Sampling and Analysis Methods
In this portion of the TAD, an attempt has been made to select the most useful
sampling and analysis methods for the organic compounds of primary concern in ambient
air monitoring programs. An attempt has been made to reference methods presented as a
formalized writeup (e.g., NIOSH or ASTM methods) whenever possible, since these
methods require the least amount of interpretation on the part of the analyst.
Unfortunately such formal methods are available for only a few compounds at the present
time. A compendum of method writeups, in a standardized format is to be prepared
within the next year(39).
In cases where formal methods are unavailable, literature articles have been cited.
Unfortunately, most of the referenced articles do not contain complete information con-
cerning the sampling and analysis methodology and will require considerable interpre-
tation on the part of the user.
Table 17 provides a listing of the most widely used sampling and analysis approaches
for ambient air monitoring. An attempt has been made to consolidate the wide variety of
approaches for hydrocarbons and halogenated hydrocarbons into a limited number of
80
-------
TABLE 17. SUMMARY OF SAMPLING AND ANALYTICAL METHODS FOR
ORGANIC COMPOUNDS IN AMBIENT AIR
oo
He t hod
Designation
Hydrocarbon*
A. ASTH D2820
B. —
C. —
D. --
E. AST* D3686/
D3687
or NIOSH
P4CAM 127
Compounds
Determined
CJ-CJQ Hydrocarbons
-100 to 175*C
C$-Ci2 hydrocarbon*
orgsnlcs with B.P.
and other nonpolar
B.P. 60-200'C
Honpolar
voUtilea
(B.P, 0-150'C)
Sampling and
Analysis
Approach
Injection onto
GC/FID
Whole air
cryogenic
CC/F1D analysis
Adsorption on
desorptlon
CC/KS analysis
Tenax; thermal
Into canisters;
CC/FID or CC/MS
analysis
Adsorption on
charcoal;
deaorptlon with
€82; analysis
Detection Relative Cost
0.1 ppbv -- +101 Moderate Moderate
(100 mL
sample)
1-200 pptv 70-951 +10-40Z High Moderate
sample)
(20 liter to hlRh,
on CC
detector
0.1-1 ppmv ~ +10Z Low Low
(10 liter
•ample)
Losses due to
«*y occur In some
cases, especially
compounds.
than a week are not
recommended .
are not readily
analyzed.
27 Blank levels usually
Artifacts due to
reactive component*
(e.g. Oj, NOX) can
be a problem.
Sample can be
analyzed only once.
gives poor recovery
adsorption onto
metal surface.
Blanks and arti-
fact problems
game a* above
2 3 Sensitivity much
poorer than for
thermal desorp-
tlon approaches.
Comments
CC/MS can be used
CC/PID is useful
CC/MS can be used
for component
Identification.
If possible, a
should be used to
•void sample
•forage problems.
CC/FID can be used to
cost, if component
identification
la not required.
3-4 times, thus
quantification and
GC/MS Identification
Charcoal adsorb* highly
volatile compound*
more efficiently
than Tenax.
by CC/FID
-------
TABLE 17. (Continued)
oo
ro
Method Compounds
Designation Determined
r. — Notipolar volatile*
(B.r. 0-100*0
C. NIOSH PAH
P4CAH 206
Sampling and
Analysis Detection
Adsorption on 0.01-1 ppbv 80-1 OCX
•rbon vole- (20 liter
hermal
eaorptlon in-
o canister;
nalysl* by
GC/FID or CC/MS
Collection of 3 ng/»3 951
partlculate (1300 •'
matter on high sample)
voluaw filter;
ultrasonic
extraction with
eyclohexane/allica
powder; analysis by
normal phase HPLC
Relative Cost
+20X Moderate Moderate U
to high,
on CC
detector
+5X Mode r a te Moderate 2
Limitations
High temperature
(350'C) required
•ay decompose
labile com-
pounds.
PAHs wore volatile
than benzo{a)pyrene
•ay be lo*t by
volatilization
during the sampling
per lod . Approaches
reference 40 nay be
required for such
compounds.
Comments
See Method 0.
PAH and other particle
bound component* may
be determined using
GC/MS.
Halogenated Hydrocarbona
Methods A-F can be used for volatile halogenated hydrocarbons as well, except that GC/ECD should be used In place
of GC/FID in most casea.
ng/m3 60-1001 +ZOZ
PCB», PCN«.
organochlorlne
pesticides, and
other aemivoJa-
tll« organlce
(B.P. UO-450'C)
Adsorption on
solid adsorbent (15OO «
such as poly- sample)
urethane foaa
(PUF), XAD-2,
or ChromoBorb
102); solvent
deaorptlon; CC/
ECD, CC/MS. or
GC/PID analysis
Moderate Moderate
to high,
depending
on CC
detector
40
Lower volume
approaches
(41,42,50)
Similar approaches ualng
low volume sampling
trains may be more use-
ful for detecting
higher levels (1-10 ug/« ).
Hydrocarbons with boiling
points >140'C (>Cg) can be
determined.
-------
TABLE 17. (Continued)
Method
Designation
Compo
Deter
unds
mined
Sampling and
Analysis
Approach
Detec
tlon
Relative
Cost
Comments
Tetrachloro-
dibenzodloxlnea
Collection on
high volume
filter;
extraction
(Soxhlet)
with methy-
lene chloride;
column chroma to-
graphic cleanup;
high resolution
CC/HS analyals
••-I Pg/«
(1500 m3
sample)
+10-201 Very high High
TCDDs occur primarily TCDD is highly toxic
in the particle
bound state, but
backup adsorption
approaches auch
as reference 40
may be required t
ensure no loss
occurs through
volatilization.
hence the need for
high sensitivity.
These analyses are
generally conducted
In a special labora-
tory equipped to
handle such
materials.
Oxygenated Neutral Compounds
J. — Aldehydes
oo
CO
Aldehydes
-------
TABLE 17. (Continued)
Method Compounds
Designation Determined
L.
K10SH Alcohols
PfcCAM 556.
-------
TABLE 17. (Continued)
Method Compounds
Designation Determined
0. ~ Aliphatic Amines
(C.-C. )
14
P. HIOSH Aromatic Amines
PtCAM 168
Q. — Aromatic Amines
Sampl Ing and
Analysis
Approach
Adsorption of
alkali treated
Porasil A;
thermal desorp-
tion CC/NPD
analysis
Adsorption on
silica gel;
elutlon with
ethanol; GC/FID
analysis
Adsorption on
Tenai CC; ther-
mal desorption
CC/F1D or GC/HP1
analysis
rwtertion Relative Cost
1-5 ppbv 80-100* +5-101 Moderate Moderate 4.4
(60 liter
sample)
100 ppbv — +10t Low Low 2
(100 liter
sample)
1 ppbv — — Low Moderate 45
(100 liter
sample)
D
Requires careful
conditioning of
CC system to
obtain adequate
performance
High humidity levels
reduce adsorption
capacity.
Desorption of high
boiling compounds
is not complete.
"
Sensitivity could be
improved using CC/HPD,
HPLC can also be used
(46).
Solvent desorption
approaches may be
more ef f ec t ive in
some cases.
00
nitrogen Hetcrocycles
These compounds can be determined bv the method* described for hydrocarbons and halogenated hydrocarbons,
depending on volatility, except CC/BPD or HP1XI detection of these compounds may be required. Polar CC
phase* (e.g. Carbovu 20 H) •bould be used In moat cases.
-------
"generic" approaches since an enormous number of individual literature references are
available for these compounds.
Methods A-F (Table 17) represent sampling and analysis approaches for volatile
hydrocarbons and halogenated hydrocarbons with boiling points less than 200 C. The
simplest approach (Method A) involves direct injection of a gas sample onto a GC/FID or
other detection system and is useful for compounds more volatile than benzene. Higher
boiling compounds can be determined in some cases, although condensation onto the
container surface is a more significant problem as volatility decreases. Use of sensitive
and/or selective detectors such as ECD for halocarbons and PID for aromatics(52) or
certain olef ins can be of great value.
Method B involves cryogenic concentration of a whole air sample. Preferably this
approach is used as a field method so as to avoid sample transport and storage problems,
although laboratory analysis has been used successfully^''. This approach is more time
consuming but also much more sensitive than the direct GC injection approach and can be
used at the part per trillion level in favorable cases (e.g., using GC/ECD or PID).
Compounds in the C2~C\Q volatility range can be determined using cryogenic trapping. A
limited volume of air (<500 ml) can be sampled since condensed moisture will plug the
trap if greater air volumes are collected.
Methods C and D involve the sampling of ~20 liters of air using Tenax (a porous
polymer adsorbent), thermal desorption of the adsorbed components, and GC or GC/MS
analysis. Method C is more commonly used and involves direct desorption of the Tenax-
adsorbed components onto the GC or GC/MS system. Method D involves desorption of the
components into an evacuated cannister and subsequent analysis of the cannister contents.
Method D offers the advantage of replicate analysis of a single sample, but is somewhat
less sensitive than the former approach. Adsorptive losses of higher boiling compounds
onto the cannister surface is another potential problem. An advantage of the Tenax
approach, relative to cryogenic trapping, is that water and other inorganic atmospheric
components are not retained. Storage of the organic components in the resin-adsorbed
state also tends to circumvent problems with adsorption on container surfaces.
Major disadvantages of the Tenax adsorption approach is the potential for artifact
formation (e.g., by interaction of ozone with the resin) and/or contamination. These
factors tend to increase the background (blank) levels relative to cryogenic trapping,
thereby adversely affecting the detection limit. The widely varying retention volumes for
organic compounds on Tenax requires careful consideration for each monitoring situation.
A recent report describes the Tenax adsorption process and gives retention volume data
for many organic compounds of interest.(°™
Methods E and F involve adsorption of volatile, nonpolar compounds on carbon based
adsorbents followed by solvent extraction or thermal desorption and GC or GC/MS analy-
sis. Method E, adsorption of charcoal followed by extraction with CS£ or other solvents,
is the basis for many of the NIOSH methods. Unfortunately this approach is not suf-
ficiently sensitive for most ambient air monitoring programs, although the use of se-
lective detectors can improve sensitivity for certain compounds.
Method F has been used to some extent for ambient air monitoring of vinyl chloride
and other volatile compounds. This method is similar to Method D, except that carbon
molecular sieves are used in place of Tenax. Direct thermal desorption/GC analysis
(analogous to Method C) could be employed for carbon molecular sieves as well. A major
86
-------
advantage of this approach is the strong retention of vinyl chloride and other highly
volatile materials. However, the high temperatures required for thermal desorption from
carbon based adsorbents can lead to degradation of strongly adsorbed, nonvolatile and/or
polar materials.
Methods G, H, and I represent approaches useful for the determination of semi-
volatile or nonvolatile compounds (i.e., boiling points greater than 140 C). Method G,
while referenced as a PAH analysis method, is readily adapted for the determination of
many nonvolatile materials adsorbed on atmospheric particulate material. Method I is
referenced as a separate method for tetrachloro-dibenzodioxins (TCDDs) which are
currently of great environmental concern. Highly specific cleanup and GC/MS steps are
used to gain selectivity for TCDDs using the method.
Method H represents as useful approach for PCBs, PCNs, organochlorine pesticides
and other semivolatile compounds which can occur in both the particle and vapor state.
Such compounds are not retained using conventional high volume filtration techniques
hence adsorbents such as XAD-2 or polyurethane form (PUF) must be used in back of the
filtration device. The semivolatile components are then recovered from the adsorbent by
solvent extraction. Tenax is less useful than XAD-2 or PUFs in this method since many
solvents will partially dissolve Tenax.
The remaining methods in Table 17 represent specialized techniques for selected
groups of compounds. Methods J and K are used for determination of volatile aldehydes
(GI to C&). Method J involves formation of the DNPH derivatives of the various
aldehydes followed by reversed phase HPLC analysis, whereas Method K captures the
aldehydes as bisulfite addition products and then employs a variety of colorimetric and
GC analytical procedures for the individual aldehydes. Method J should generally be used
if a wide variety of aldehydes are to be determined or if interferences with the colori-
metric methods are likely to be encountered. The determination of formaldehyde using
chromatropic acid, as in Method J, is a simple procedure and is very useful for screening
pruposes, although negative and positive interferences can occur.
Detection of alcohols in ambient air has not been of widespread interest and no
sensitive methods exist. Method L refers to a group of NIOSH methods employing carbon
adsorption and solvent desorption/GC analysis. These methods are not useful below 100
ppm. Alcohols are not retained on porous polymer adsorbents such as Tenax and are likely
to decompose if carbon adsorption/thermal desorption approaches are attempted. An
analytical method involving silylation of alcohols followed by cryogenic trapping of the
silylated derivative (as in Method B) has been reported'-*").
Nitrosamines are of considerable environmental significance because of the toxico-
logical hazard and potential for formation in various combustion sources. Method M
refers to an approach involving collection of the nitrosamine on a specially treated nylon
adsorbent (thermosorb/N) and subsequent analysis using GC with MS or thermal energy
(TEA) detectors. The latter detector is relatively specific for nitrosamines^^). Since
both MS and TEA are relatively expensive detectors, the less expensive GC/NPD approach
may be preferable in relatively "clean" environments in which ultimate selectivity is not
necessary. Thermosorb/N minimizes the formation of nitrosamines on solid adsorbents
during sampling, which has proven to be a problem on Tenax under certain sampling
conditions.
87
-------
Methods N-Q represent the best available methods for determining aliphatic and
aromatic amines. Nitrogen heterocycles containing no other polar functional groups can
usually be determined by one or more of the methods described earlier for hydrocarbons,
using nitrogen selective detectors. Methods N and P represent NIOSH procedures for
determining aliphatic and aromatic amines, respectively. These methods involve ad-
sorption of the compounds on silicagel, elution with acid and GC/FID analysis. The
adsorption of water reduces the capacity of the silica gel and limits the practical
sampling volume. Therefore, these methods cannot determine low ppb levels of amines,
although use of GC/NPD should increase the sensitivity somewhat. HPLC with fluo-
rescence, electrochemical, or UV detection^) can improve the detection limit for
aromatic amines.
Method O involves determination of volatile aliphatic amines (C\ -Cq) by adsorption
on alkali-treated silica (Porasil A) followed by thermal desorption and GC/NPD analysis.
This method is reported to achieve 1-5 ppb sensitivity if the GC system is carefully con-
ditioned(44). Method Q is a similar approach using Tenax/thermal desorption for aromatic
amines. However the thermal instability of many aromatic amines must be considered
and may limit the usefulness of this approach.
Relatively few analytical methods are available for determining acidic compounds
(with the exception of phenols) in ambient air. Phenol itself and possibly cresols can be
determined using a Tenax adsorption approach such as Method D. Higher boiling phenols
can be determined using a resin adsorption/solvent extraction approach such as Method H.
Method S represents a standard method for phenols wherein the compounds are collected
in a dilute sodium hydroxide impinger and then steam distilled and analyzed by GC/FID or
GC/MS. Less volatile phenols can be analyzed (without steam distillation) by HPLC with
fluorescence, UV or electrochemical detection.
Volatile carboxylic acids such a formic and acetic acid can be detemined (Method R)
by collection in a dilute sodium carbonate impinger followed by ion chromatographic (1C)
analysis^), ic is a special form of ion exchange HPLC wherein conductance detection is
employed (mobile phase buffer is removed prior to detection using a stripper column).
Suitable methods for determining sulfonic acids in air have not been reported.
However, collection in an aqueous impinger followed by HPLC analysis^?) appears to be a
viable approach.
In order to provide the reader with some useful examples of appropriate method-
ology, Table 18 lists a group of toxic organic compounds of concern in ambient air mon-
itoring programs. The most appropriate methods for determining these compounds,
specific literature references, and additional analytical considerations are presented in
Table 18. This information should be used as guidance and will not be accurate for every
monitoring situation.
The majority of the compounds in Table 18 represent volatile hydrocarbons, volatile
halogenated hydrocarbons, or semivolatile halogenated hydrocarbons which can be deter-
mined using conventional cryogenic trapping, Tenax adsorption/thermal desorption, or
resin adsorption/solvent extraction approaches. A few compounds (e.g., acrylonitrile,
allyl chloride, ethylene oxide) are too volatile to be captured on Tenax and require use of
carbon adsorbents or cryogenic trapping. Several of the compounds can be determined
using the procedures for aldehydes (e.g., formaldehyde, acetaldehyde, acrolein).
88
-------
TABLE 18. SUMMARY OF SAMPLING AND ANALYSIS METHODS FOR
SELECTED TOXIC ORGANIC COMPOUNDS
Compound
Most Appropriate
Methods From
Table 17
Specific
References
Comments
00
Acetaldehyde
Acrolein
Acrylonitrile
Allyl Chloride
(3-Chloro-l-propene)
Benzyl Chloride
Carbon Tetrachloride
J,K
J,K
B,F
B,E,F
B,C or D
B,C or D
4,19
4
2
27
27
Compound very unstable, requires
immediate analysis or derivatization.
Compound very unstable, required
immediate analysis or derivatization.
NIOSH Method Si 56 uses methanol as
desorption solvent. GC/NPD will
give improved sensitivity.
Compound is too volatile to allow use
of Tenax/thermal desorption approach.
NIOSH Method SI 16 uses benzene
solvent desorption with GC/FID
analysis.
Cryogenic trapping or Tenax adsorption
appear to be the best approaches.
Adsorption on XAD-2 may also be use-
ful. GC/PID is a useful determinative
method.
Cryogenic trapping or Tenax adsorption
appear to be the best approaches.
Storage of samples in contact with
stainless steel surfaces can result
in rapid loss of this compound.
GC/ECD is a useful method.
-------
TABLE 18. (Continued)
Compound
Most Appropriate
Methods From
Table 17
Specific
References
Comments
Chlorobenzene
Chloroform
B,C or D
B,C or 0
Chloroprene B,C or D
(2-Chloro-l,3-butadiene)
o,m, or p-cresol
(2,3, or 4-methyl
phenol)
1,4-dichlorobenzene
C,S,H (low
volume)
B,C or D
Dimethylnitrosamine M
(N-nitrosodimethylamine)
Epichlorohydrin
(l-Chloro-2,2-
epoxypropane)
B,C or D
14,27,50
14,27
4,54
27,50
49
27
Cryogenic trapping or Tenax adsorption
appear to be the best approaches.
GC/PID is a useful method.
Cryogenic trapping or Tenax adsorption
appear to be the best approaches.
6C/ECD is a useful determinative method.
Cryogenic trapping or Tenax adsorption
appear to be the best approaches,
although very little data is available.
GC/PID is a useful determinative
method.
Either collection in sodium hydroxide
impinger, Tenax adsorption, or resin
adsorption/solvent extraction can be
used. GC/PID is a useful determinative
method.
Cryogenic trapping or Tenax adsorption
appear to be the best approaches.
Resin adsorption/solvent desorption
approaches can be used. GC/ECD or
PID are useful determinative methods.
Adsorption on Thermosorb N/thermal
desorption appears to be the best
approach. GC/NPD may provide
sufficient selectivity in many cases.
Relatively little ambient air data are
available in the literature. Cryo-
genic trapping or Tenax adsorption
appear to be viable approaches.
-------
TABLE 18. (Continued)
Compound
Most Appropriate
Methods From
Table 17
Specific
References
Comments
Ethylene dichloride
(1,2-dichloroethane
Ethylene Oxide
Formaldehyde
Hexachlorocyclopenta-
i£ diene
Maleic Anhydride
Methyl Chloroform
(1,1,1-trichloroethane)
Methylene Chloride
Nitrobenzene
B,C or D
A,B,E
J,K
C or D,H
(low volume
approach)
B,C or D
B,F
14,27
B,C or D
4,19
53
27
14
14
Cryogenic trapping or Tenax adsorption
appear to be the best approaches.
GC/ECD is a useful determinative
method.
Compound too volatile to use Tenax
adsorption. Cryogenic trapping is
probably the best approach.
Compound very reactive. Requires
immediate analysis or stabilization.
Very little data available for this
compound in ambient air.
No suitable methods could be found.
Cryogenic trapping or Tenax adsorption
appear to be the best approaches.
GC/ECD is a useful determinative
method.
Cryogenic trapping, appears to be the
most useful approach. This com-
pound is not retained well by Tenax
or other polymeric adsorbents.
Adsorption on carbon molecular sieves
in place of Tenax is a useful approach.
Laboratory contamination with methylene
chloride is a common problem.
Tenax adsorption is probably the best
approach. GC/ECD, PID, or NPD are
useful determinative methods.
-------
TABLE 18. (Continued)
Compound
Ni trosomorphol i ne
Perchloroethylene
Most Appropriate
Methods From
Table 17
M
B,C or D
Specific
References
55
14,27
Comments
See Dimethylnitrosamine. HPLC may be use-
ful in place of GC for this compound.
Cryogenic trapping or Tenax adsorption
ro
(Tetrachloroethylene)
Phenol
Phosgene
C or S
4,19
51
PCBs
40
Propylene Oxide
B,C or D
27
appear to be the most useful approaches,
GC/ECD is a useful determinative
method.
Trapping in basic impinger solution or
Tenax adsorption appear to be the best
approaches. GC/PID is a useful
determinative method.
This compound is highly unstable and
hence field determination is desirable.
A manual colorimetric method using
4-nitrobenzyl-pyridine appears to be
the best approach for routine analysis
(detection limit ^0.05 ppm for 25
liter sample).
Adsorption of XAD-2 or polyurethane foam
followed by solvent extraction and GC/
ECD analysis appear to be the best
approaches. PCB formulations are com-
posed of many individual compounds
and the method of quantification
required careful consideration.
Cryogenic trapping or Tenax adsorption
appear to be the best approaches,
although the Tenax approach should be
used with caution because of the low
breakthrough volume for this compound.
-------
TABLE 18. (Continued)
Compound
Most Appropriate
Methods From
Table 17
Specific
References
Comments
10
CO
Toluene B,C or D
Trichloroethylene B,C or 0
Vinylidine Chloride B,C or D
(1,1-dichloroethane)
o,m,p-xylene B,C or D
(1,2; 1,3; or 1,4-
dimethyl benzene)
14,17,27, Cryogenic trapping or Tenax adsorption
41,52 appear to be the best approaches.
GC/PID is a useful determinative
technique.
14,27 Cryogenic trapping or Tenax adsorption
appear to be the best approaches.
GC/ECD is a useful determinative
technique.
14,27 Cryogenic trapping or Tenax adsorption
appear to be the best approaches.
GC/ECD is a useful determinative
technique.
14,17,27, Cryogenic trapping or Tenax adsorption
41,52 appear to be the best approaches.
Adsorption on XAD-2 and solvent
extraction is also possible. GC/PID
is a useful determinative method.
-------
Compounds in Table 18 representing significant problems for sampling and analysis
include maleic anhydride, dimethylnitrosamine, nitrosomorpholine, and phosgene. The two
nitrosamines have been determined in air'49,55) bu^ require special adsorption techniques
and GC analysis approaches to avoid artifact formation. Phosgene methods have been
evaluated in air and a manual colorimetric technique has proven to be most suitable for
routine use'-*l). The instability of phosgene precludes sample collection and hence field
analysis or stabilization (e.g, derivatization) approaches must be employed.
No suitable methods were found for determining maleic anhydride in air. The ion
chromatographic approach described for formic and acetic acids'43) possibly could be
adapted for this compound (i.e., determination as maleic acid collected in sodium
carbonate impinger).
QUALITY ASSURANCE PROCEDURES
The purpose of this part of the TAD is to address quality assurance and related
needs specific for the sampling and analysis approaches described above. The overall QA
requirements of ambient air monitoring programs have been described in Section 3 of the
TAD and references therein(^>^'.
Method Validation
Validation of method performance is important in all sampling and analysis programs
but is of special significance for trace organic monitoring because of the large number of
compounds of interest and variables affecting method performance. In many situations
time, cost, or technical limitations will preclude rigorous method validation and certain
assumptions will be required. Ideally any such assumptions will be based on sound tech-
nical judgement and/or prior experience.
Aspects of method performance requiring validation include the following:
Accuracy
Precision
Blank or background level
Detection limit
Interferences
Ruggedness (effect of important variables on method performance).
In the ideal situation each of these aspects of method performance will be evaluated using
the entire sampling and analysis scheme to monitor an atmosphere containing constant,
known amounts of the analytes under conditions identical to the field. Two technical
limitations prevent the accomplishment of this "ideal" method validation strategy in most
cases.
The most severe limitation is that duplication of field conditions is impossible
because of the wide variability in field conditions. The second limitation is that gener-
ating atmospheres of known constant composition is relatively difficult, especially for
unstable components which also pose the greatest problem for sampling and analysis.
A typical approach used to partially overcome these limitations is shown in Figure
12. In this scheme a laboratory validation effort is conducted wherein the emphasis is
94
-------
Generate atmosphere under laboratory conditions
containing known, constant amount of analyte.
For unstable compounds use dry inert gas and/or
reference analysis methods to ensure a known,
stable concentration.
Laboratory
Determine method detection limit, and accuracy
and check interference from known materials
using the laboratory generated atmosphere.
Determine method "ruggedness" by varying
temperature, humidity, sampling volumes, etc. in
cases where a stable concentration of analyte can be
maintained. Use series samples to check breakthrough
when appropriate.
Determine method precision by collecting
parallel samples.
Field
Estimate method accuracy by comparing data using
test and reference methods, running series
samples (check on breakthrough), and/or
spiking field sample with known amounts
of analyte.
Figure 12. Method Validation Scheme
95
-------
placed on generating a known, stable concentration of analyte. Unstable compounds may
require the use of dry, inert gas (e.g., nitrogen) to make up the atmosphere. In some cases
reference methods may be available as a check on compound stability (such methods need
not be highly specific since a relatively clean atmosphere is being employed).
Method detection limit, precision, accuracy, interferences, and background can be
determined using the laboratory generated atmosphere. Method ruggedness can be tested
provided the variables do not affect the analyte stability; unstable compounds often will
not meet this requirement.
The field method validation efforts include a check on precision using parallel
sampling. Accuracy can be estimated through the use of reference or alternate methods,
and/or spiking field samples with known quantities of analyte. The latter approach must
be accomplished in such a manner that all of the spiked material enters the sampling
device. Alternatively, one can choose to spike the collected sample (e.g., impinger
liquid, resin cartridge, etc.) as a check only on the sample transport and analytical
procedures. Series samplers can be used as a check on capture efficiency (analyte
breakthrough) but will not determine accuracy. In many cases, it will be useful to
collect series samples using various sampling volumes to further document component
breakthrough characteristics.
Instrument Calibration
Instrument calibration requirements for sampling and analysis equipment are
outlined in Table 19. Specific calibration and maintenance procedures will vary somewhat
from one manufacturer to another, hence the user should consult the instrument manual
for more specific information.
Sampling equipment calibration procedures for toxic organic monitoring are similar
to other types of monitoring and adequate information on this subject can be found in the
literature^). Continuous analyzers require that a suitable calibration standard be
available in the field. Ideally the calibration standard is a dilute mixture or series of
dilute mixtures of the analyte at stable concentrations. Methods for generating such
atmospheres can be static (e.g., dilution flasks, compressed gas cylinders) or dynamic
(e.g., permeation tubes, diffusion tubes, syringe delivery systems). In all cases, a method
of generating clean air must be available in order to set the baseline level on the con-
tinuous monitor as well as for the generation of calibration standards. Methods for gener-
ating clean air as well as static and dynamic calibration methods are discussed in the
literature^).
In general, static systems are most convenient to use and are the preferrred cali-
bration methods, provided the analytes are stable in the dilution system. Generally light
hydrocarbons and other stable, volatile compounds (e.g., halocarbons) are suitable for
static calibration. Dynamic calibration systems, while more complex for field use, are
often required for reactive materials (e.g, phosgene) which may be degraded in the static
systems.
Less volatile materials such as PCBs, organochlorine, pesticides, or PAHs are rarely
of interest for continuous monitors since the ambient concentrations of these materials
are usually not detectable using this approach. If atmospheres of such compounds are to
be generated probably the best approach is a heated dynamic dilution system wherein a
dilute solution of the material in a volatile solvent is delivered at a constant rate into the
gas stream using a syringe pump.
96
-------
TABLE 19. CALIBRATION REQUIREMENTS FOR SAMPLING
AND ANALYSIS INSTRUMENTATION
Device
Sampling Instrumentation
Sampling pump and
controller
Parameter
Calibrated
Flow rate
Method of
Calibration
Wet or dry test
meter or calibrated
Approximate
Frequency
Weekly
Comments
Sample volume measurement
device (usually a dry
test meter)
Analytical Instruments
Continuous monitors
(e.g., FID, PID, FPD,
etc.)
Chromatographic
instruments
ChromatograpMc
instruments
GC/MS
GC/MS
Total volume
Response
Column performance
and retention
time for each
analyte
Response for
each analyte
Response and
retention time
for each analyte
Mass spectral
resolution and
turning parameters
rotometer
Wet test meter
Weekly
Generation of test
atmosphere of
known concentration
Injection of
standard using
the same process
as for sample
Injection
Same as above
Daily or
more
frequently
if required
Daily or
more
frequently
1f required
Must be determined at
known atmospheric oressure
and temperature. Flow
rate should be similar to
that used for sampling.
Test atmosphere should be
referenced to a primary
standard (e.g., NBS benzene
in air). Flow/pressure
conditions should duplicate
sampling process.
Standard composition
should be checked against
primary standards if
available.
Same as above Same as above
Same as for other chromatographic instruments.
(a) Introduction
of nerfluoro-
compound directly
into MS
(b) Injection of
tuning standard
(e.g., bromofluoro-
benzene) into GC
Dally
Selection of tuning
standards will be
dependent on type of
of analysis being
performed.
-------
Chromatographic instrumentation requires calibration of both the retention time
and response characteristics of the analytes under the conditions used for the analysis of
samples. Samples delivered to the instrument in liquid form (e.g., XAD or PUF extracts,
impingers, etc), represent no particular problems since liquid standards of the analytes
may be readily prepared. However, techniques such as whole air injection, cryogenic
trapping, and thermal desorption (e.g., Tenax) require the generation of a known, stable
gas phase calibration standard, using methods similar to those described for continuous
monitors. In the case of the Tenax thermal desorption method standards have been suc-
cessfully prepared using a heated static dilution flask wherein the sample is injected onto
a clean Tenax cartridge using a gas-tight syringe.
The performance of GC columns, especially capillary columns should be checked
periodically in terms of column efficiency (theoretical plates) and peak asymmetry
(especially for polar compounds). This performance check can be done using liquid cali-
bration standards and is a useful tool for determining when a column needs to be replaced.
Mass spectrometers require various calibration steps, in addition to the normal
chromatographic calibration requirements. These include calibration of the mass spectral
relative intensities and mass resolution. Quadrupole MS systems are greatly effected by
such tuning parameters and hence data performance checks are required to ensure the
usefulness of the mass spectrum for peak identification.
In practice, two levels of MS tuning calibration are usually performed. First a
volatile perfluoro-compound (e.g., perfluorokerosine or perfluorotributylamine) is intro-
duced into the ion source and the MS tuning parameters are adjusted to yield certain
spectral characteristics. This tuning process is usually described in detail in the instru-
ment manual. A second level of tuning involves the injection of a particular compound
EPA methodology (70) prescribes bromofluorobenzene for volatiles and decafluorotri-
phenylphosphine (DFTPP) for semivolatile compounds onto the GC/MS system. If the
mass spectral characteristics for the reference compound are not correct, the parameters
are adjusted and the calibration process repeated.
Routine Quality Control
In addition to method validation and instrument calibration processes, the mon-
itoring program should include certain processes for periodically documenting per-
formance of the sampling and analysis procedures. Typical frequencies for sampling/
analysis QC samples are shown in Table 20. In the case of continuous monitors the cali-
bration process itself serves as a periodic indicator of method performance. Other
sampling and analysis systems require the collection or acquisition of QC samples to
check method performance. The types of QC samples of primary value include:
Blanks (both field and laboratory)
Spiked samples
Internal standards
Replicate parallel samples (or split samples)
Series samples
Reference samples.
Blanks should be processed exactly as the samples, except that no air is drawn
through the sampler. If samples are transported to the laboratory for analysis then labo-
ratory as well as field blanks should be included. In the case of resin samples (e.g., Tenax)
98
-------
Table 20. Typical Sampling/Analysis Frequencies
for QC Samples
Type of Sample
Field Blanks
Laboratory Blanks
Spiked Samples
Duplicate (parallel) Samples
Instrument Calibration Standards
Reference Samples
Series (Backup) Samples
Typical Frequency
Each Sample Set; at least 10% of
total number of samples.
Daily; at least 10% of total number
of samples. Each batch of samples.
Each sample set; weekly
10% of total number of samples; each
sample set.
Daily
Weekly
Each sample set.
99
-------
both laboratory and field blanks should be routinely analyzed since contamination during
transport is a common problem. Material blanks such as Tenax cartridges, impinger so-
lutions, etc. should be routinely analyzed prior to sampling to avoid wasting valuable field
efforts due to contamination. Blanks for whole air collection or cryogenic trapping
systems will simply be clean air as used for instrument calibration.
Spiked samples must be prepared in such a manner that (a) the sample form is not
altered by the spiking process and (b) all of the spike material is available to the sampling
system. In many cases, the air stream itself cannot be spiked because of technical practi-
calities. Consequently the most common approach is to spike the collection matrix (e.g.,
Tenax resin, impinger, etc.) either before or after sampling. Resin cartridges can be
spiked using the static dilution system described above. In some cases (e.g., for volatile,
stable compounds) a whole air sample can be collected in a cylinder or Teflon bag and
spiked with a known amount of analyte.
The use of internal standards (IS) is advisable for chromatographic methods for
which the IS can be placed into the sample without altering it. The use of an IS helps to
track instrument sensitivity and to compensate for losses during sample processing.
Internal standards are most commonly employed for liquid injection or thermal desorption
methods but are not advisable for whole air injection or cryogenic trapping in most cases.
Internal standards are especially useful for GC/MS techniques since the stable isotopically
(e.g., deuterium or *-3Q labeled analytes can be placed into the sample and compensate
for any losses during processing or changes in instrument response. This approach has
been used extensively for Tenax thermal desorption procedures.
Method precision can be routinely monitored by collecting replicate parallel
samples. The precision of the analytical step alone can be determined using split samples,
provided homogeneous splits can be prepared without contaminating or otherwise altering
the sample. In practice impinger solutions and other liquid samples can be split, but resin
samples cannot due to inhomogeneous distribution of the analytes in the resin.
Collection of replicate samples of varying sample volumes can be useful for deter-
mining sampling volume effects on the method. Series samples are recommended as
routine checks on breakthrough for resin samplers, since changes in atmospheric con-
ditions may alter the breakthrough characteristics of a given analyte. Collection of series
samples using different sample volumes, while time consuming, may be a necessary diag-
nostic tool in certain cases. This type of consideration is discussed in a recent report for
solid adsorbent sampling procedures'^^'.
Reference samples, especially standards available from NBS, are useful as a routine
check on method accuracy although only a limited number of such samples are available.
NBS gas standards for organics (e.g., benzene) will become available in the near future
and will be of great value in the toxic organic monitoring area. Secondary standards sup-
plied by various manufacturers and calibrated against an NBS reference (e.g., propane in
air) are currently in widespread use both of instrument calibration and routine quality
control. NBS SRM 1649, an urban dust, is available as a reference standard for particu-
late phase PAH determinations.
Data Handling
This section of the TAD discusses specific calculation processes for converting raw
data to a meaningful form. Methods for data presentation and QA requirements with
respect to data processing have been presented in Section 3 and references therein^''.
100
-------
Most of the NIOSH and ASTM methods contain equations for converting raw ana-
lytical data to final concentrations, many of which are applicable to other sampling and
analysis methods. Raw data obtained during the sampling and analysis process should
include the following:
• Volume of sample collected (unconnected for ambient pressure and
temperature KVcjA
• Temperature of sample volume measuring system-Tg
• Ambient atmospheric pressure at sampling site-Pg
• Quantity of analyte in total sample-Q^.
In almost all cases, the final data should be presented in terms of quantity of
analyte per unit volume, under standard conditions (25 C and 760 mm Hg pressure).
Actual sample volume can be converted to standard sample volume Vgg using the
following equation:
v
Vss
.7 i ^
= VSA (mj) x
298
Ts (°C) + 273
PS(mmHg)
-
760
The concentration of analyte
readily calculated as follows:
in the sample under standard conditions can then be
„
CA
vss
In many cases, one may wish to convert concentrations from yg/m3 to parts per
billion (ppbv), although these units are only meaningful if the material is present entirely
in the gas phase (i.e., nonvolatile compounds such as PAHs should always be reported in
Pg/m3 or similar units). The following equation can be used to convert ug/m3 to ppbv for
gas phase components at 25 C and 760 mm pressure:
CA (ppbv) = CA (yg/m3) x
Z4.4
where M\YA - molecular weight of analyte.
In many cases, hydrocarbon concentrations are most useful when reported in terms
of ppbC (parts per billion carbon) since this unit is relatively proportional to the output of
an FID detector for such compounds. The following equations can be used to calculate
ppbC:
CA (ppbC) = CA (ppbv) x Nc
101
-------
where NQ = the number of carbons in the analyte molecule. In a strict sense the term
ppbC should be reserved for hydrocarbons since the presence of O, Cl, N, etc. greatly
affects the per carbon response of the FID.
102
-------
REFERENCES
1. Air Pollution. A. C. Stern, ed., 2nd Edition, V. H, Academic Press, New York, 1968.
pp. 329-391.
2. NIOSH Manual of Analytical Methods, Parts 1-3. 2nd Edition,National Institute for
Occupational Safety and Health, Cincinnati, Ohio, 1977.
3. Annual Book of Standards. Part 26, Gaseous Fuels; Coal and Coke; Atmospheric
Analysis. American Society for Testing and Materials, Philadelphia, Pennsylvania.
(Published Annually)
4. Methods of Air Sampling and Analysis, M. Katz, ed., 2nd Edition, American Public
Health Association, Washington, D.C., 1977.
5. Air Pollution, A. C. Stern, ed., 3rd Edition, V. m, Academic Press, New York, 1976.
pp. 344-368.
6. Ambient Monitoring Guidelines for Prevention of Significant Deterioration (PSD).
EPA-450/4-80-012 (PB 81-153231). U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, 1980. 45 pp.
7. Quality Assurance for Air Pollution Measurement Systems. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina, January 1976.
Vol I - Principles, EPA-600/9-76-005. Vol H - Ambient Air Specific Methods,
EPA-600/4-77-027a.
8. Interim Guidelines and Specifications for Preparing Quality Assurance Project
Plans, QAMS-005/80, U.S. Environmental Protection Agency, Washington, D.C.,
December 29, 1980.
9. Reference 5. pp. 453-484.
10. Handbook of Chemistry and Physics, Chemical Rubber Company, Cleveland, Ohio.
(Published Annually)
11. The Merck Index, Merck and Co., Rahway, New Jersey. (Published Annually)
12. Handbook of Environmental Data on Organic Chemicals, K. Verschueren, ed.,
Van Nostrand and Co., New York, 1977. 659 pp.
13. Reference 4. pp. 38-48.
14. Kebbekus, B. B., and J. W. Bozzelli. Collection and Analysis of Selected Volatile
Organic Compounds in Ambient Air. Proc. Air Pollution Control Assoc., Paper No.
82-65.2. Air Poll. Control Association, Pittsburgh, Pennsylvania, 1982.
15. Holzer, G., H. Shanfield, A. Zlatkis, W. Bertsch, P. Juarez, H. Mayfield, and
H. M. Liebich. Collection and Analysis of Trace Organic Emissions from Natural
Sources. J. Chromatog. 142, 1977. pp. 755-764.
103
-------
16. Reference 5. pp. 156-158.
17. Holdren, M., S. Humrickhouse, S. Truitt, H. Westberg, and H. Hill. Analytical
Technique to Establish the Identity and Concentration of Vapor Phase Organic
Compounds. Proc. Air Poll. Control Assoc., Paper No. 79-52.2, Air Pollution
Control Association, Pittsburgh, Pennsylvania, 1979.
18. Reference 5. pp. 152-154.
19. Fung, K., and D. Grosjean. Determination of Nanogram Amounts of Carbonyls as
2,4-Dinitrophenylhydrazones by HPLC. Anal. Chem. 5_3, 1981. pp. 168-171.
20. Reference 5. pp. 3-47.
21. Reference 4. pp. 191-205.
22. Kolak, N. P., and J. R. Visalli. Comparison of Three Methods for Measuring
Suspended Particulate Concentrations. Env. Sci. Tech, 15, 1981. pp. 219-224.
23. Mitchell, R. I., W. M. Henry, and N. C. Henderson. "Mega-volume Respirable
Particulate Sampler (Mark n)". Proc. Air Poll. Control Assoc., Paper No. 77-35.1,
Air Pollution Control Association, Pittsburgh, Pennsylvania, 1977.
24. IERL-RTP Procedures Manual: Level 1 Environmental Assessment, 2nd Edition,
EPA 600/7-78-201. U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, October 1978.
25. Reference 4. pp. 184-186.
26. Ibid. pp. 88-98.
27. Krost, K., E. D. Pellizzari, S. G. Walbun, and S. A. Hubbard. Collection and Analysis
of Hazardous Organic Emissions. Anal. Chem. 54, 1982. pp. 810-818.
28. McFadden, W. H. Techniques of Combined Gas Chromatography Mass Spectrometry,
John Wiley and Sons, New York, 1973.
29. Lane, D. A. Mobile Mass Spectrometry. Env. Sci. Tech. 16, 1982. pp. 38A-46A.
30. Burlingame, A. L., A. Dell, and D. H. Russel. Gas Chromatography Mass
Spectrometry. Anal. Chem. 54, 1982. pp. 363R-409R.
31. Gas Chromatography with Glass Capillary Columns, Jennings, W., ed., Academic
Press, New York, 1978.
32. Recent Advances in Capillary Gas Chromatography, Bertsch, W., G. Jennings, and
R. E. Kaiser, ed., Verlag, Heidelberg, 1981.
33. Introduction to Modern Liquid Chromatography, Snyder, L. R., and J. J. Kirkland,
2nd Edition, John Wiley and Sons, New York, 1979.
34. Reference 4. pp. 248-256.
104
-------
35. Ibid. pp. 128-136.
36. Ibid. pp. 216-219.
37. Shafer, K. H., M. Cooke, F. DeRoos, R. J. Jakobsen, O. Rosario, and J. D. Mulik.
WCOT Capillary Column GC/FT-IR and GC/MS for Identifying Toxic Organic
Pollutants, Applied Spectroscopy, 35, 1981. pp. 469-472.
38. Tuazon, E. C., A. M. Winer, R. A. Graham, and J. N. Pitts. Atmospheric
Measurements of Trace Pollutants by Kilometer-Pathlengths FT-IR Spectroscopy.
In Advances in Environmental Science and Technology, 10, 1980. pp. 259-300.
39. Compendium of Sampling and Analysis Methods for Organic Compounds in Ambient
Air. U.S. Environmental Protection Agency, North Carolina, in preparation (1983).
40. Jackson, M.D., and R. G. Lewis. Polyurethane Foam and Selected Sorbents as
Collection Media for Airborne Pesticides and PCBs. In Sampling and Analysis of
Toxic Organics in the Atmosphere, ASTM STP 721, American Society for Testing
and Materials, Philadelphia, Pennsylvania, 1980. pp. 36-47.
41. VanTassel, S., N. Amalfitano, and R. S. Norang. Determination of Arenes and
Volatile Haloorganic Compounds in Air at Microgram per Cubic Meter Levels by Gas
Chromatography. Anal. Chem., 5JJ, 1981. pp. 2130-2135.
42. Erickson, M. D., L. C. Micheal, R. A. Zweidinger, and E. D. Pellizzari. Development
of Methods for Sampling and Analysis of Polychlorinated Naphthalenes in Ambient
Air. Env. Sci. Tech., 1£, 1978. pp. 927-931.
43. Bodek, I., and K. T. Menzies. Ion Chromatographic Determination of Formic Acid in
Diesel Exhaust and Mine Air. In Chemical Hazards in the Workplace, G. Choudhary,
ed., Symposium Series 149, American Chemical Society, Washington, D.C., 1981.
pp. 599-613.
44. Kuwata, K., Y. Yamazaki, and M. Uebori. Determination of Traces of Low
Aliphatic Amines by Gas Chromatography. Anal. Chem., 5_2, 1980. pp. 1980-1982.
45. Bowen, B. E. Determination of Aromatic Amines by an Adsorption Technique with
Flame lonization Gas Chromatography. Anal. Chem., 48, 1976. pp. 1584-1587.
46. Lores, E. M., D. W. Bristol, and R. F. Moseman. Determination of Halogenated
Anilines and Related Compounds by HPLC with Electrochemical and UV Detection.
J. Chrom. Sci., 16, 1978. pp. 358-362.
47. Knox, J. H., and G. R. Laird. Soap Chromatography-A New HPLC Technique for
Separation of lonizable Materials. J. Chrom., 122, 1976. pp. 17-34.
48. Harvan, D. J., J. R. Hass, J. L. Schroeder, and B. J. Corbett. Detection of
Tetrachlorodibenzodioxins in Air Filter Samples. Anal. Chem., 5j, 1981.
pp. 1755-1759.
105
-------
49. Rounbehler, D. P., J. W. Reisch, and D. H. Fine. Nitrosomine Air Sampling Using a
New Artifact-Resistant Solid Sorbent System. In Sampling and Analysis of Toxic
Organics in the Atmosphere, ASTM STP 721, American Society for Testing and
Materials, Philadelphia, Pennsylvania, 1980. pp. 80-91.
50. Langhorst, M. L., and T. J. Nestrick. Determination of Chloro-benzenes in Air and
Biological Samples by Gas Chromatography with Photoionization Detection. Anal.
Chem., 251, 1979. pp. 2018-2025.
51. Ruggle, R. M., G. G. Esposito, T. L. Guivan, T. L. Hess, D. Lillian, G. Podolak,
K. G. Sexton, and N. V. Smith. Field Evaluation of Selected Monitoring Methods for
Phosgene in Air. Amer. Ind. Hyg. Assoc. J., 40, 1979. pp. 387-394.
52. Hester, N. E., and R. A. Meyer. A Sensitive Technique for Measurement of Benzene
and Alkylbenzenes in Air. Env. Sci. Tech., 13, 1979- pp. 107-109.
53. Dillon, H. K. Development of Air Sampling and Analytical Methods for Toxic
Chlorinated Organic Compounds. NTIS Report No. PB80-193279, National Institute
for Occupational Safety and Health, 1980. 84 pp.
54. Iwansiya, Y., and T. Nishishita. Determination of Phenols in the Atmosphere by
Concentration Equilibrium-Sampling Gas Chromatography. Bunseki Kagaku, 28,
1979. pp. 26-31.
55. Goff, E. U., J. R. Coombs, D. H. Fine, and T. M. Baines. Determination of
N-Nitrosamines from Diesel Engine Crackcase Emissions. Anal. Chem., 52, 1980.
pp. 1833-1836.
56. Osman, M., H. H. Hill, M. W. Holdren, and H. Westberg. Vapor-Phase Silylation of
Alcohols for Air Analysis. In Advances in Chromatography, A. Zlatkis, ed.,
Chromatography Symposium - University of Houston, Houston, Texas, 1979.
pp. 301-312.
57. Rasmussen, R. A., D. E. Harsch, P. H. Sweany, J. P. Krasnec, and D. R. Cronn.
Determination of Atmospheric Halocarbons by a Temperature Programmed Gas
Chromatographic Freezeout Concentration Method. J. Air Poll. Cont. Assoc., 27,
1977. pp. 579-581.
58. Air Sampling Instruments for Evaluation of Atmospheric Contaminants, 5th Edition,
American Conference of Governmental Industrial Hygienists, Cincinnati, Ohio, 1978.
59. Reference 4. pp. 16-26.
60. Walling, J. F., R. E. Berkley, D. H. Swanson, and F. J. Toth. Sampling Air for
Gaseous Organic Chemicals Using Solid Adsorbents-Application to Tenax.
EPA-600/7-54-82-059, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, 1982. 32 pp.
61. Cox, R. D., and R. F. Earp. Determination of Trace Level Organics in Ambient Air
by High Resolution Gas Chromatography with Simultaneous Photoionization and
Flame lonization Detection. Anal. Chem., 54, 1982. pp. 2265-2270.
106
-------
62. Pellizari, E. D. Electron Capture Detection in Gas Chromatography. J. Chrom., 98,
1974. pp. 323-361.
63. Freedman, A. N. Photoionization Detector Response. J. Chrom., 236, 1982.
pp. 11-15.
64. Dahlgran, J. R. Simultaneous Detection of Total and Halogenated Hydrocarbons in
Complex Environmental Samples. J. High Resolut. Chrom. and Chrom. Commun., 4,
1981. pp. 393-397.
65. Parliment, T. H., and M. D. Spencer. Applications of Simultaneous FID/NPD/FPD
Detectors in the Capillary Gas Chromatograph Analysis of Flavors. J. Chrom., 19,
1981. pp. 435-438.
66. Towns, B. D., and J. N. Driscoll. Detection and Identification of Amines Using
GC-PID/NPD in Series. American Laboratory, 14, 1982. pp. 56-62.
67. McCarthy, L. V., E. B. Overton, C. K. Paschke, and J. L. Laseter. Analysis of Trace
Levels of Volatile Organic Contaminants in Municipal Drinking Water by Glass
Capillary Gas Chromatography Using Simultaneous Flame lonization and Electron
Capture Detection. Anal. Lett., 1JI, 1980. pp. 1417-1429.
68. McCarthy, L. V., E. B. Overton, M. A. Maberry, S. A. Antoine, and J. L. Laseter.
Glass Capillary Gas Chromatography with Simultaneous Flame lonization (FID) and
Hall (HECD) Detection. J. High Resolut. Chrom. and Chrom. Commun., 4, 1981.
pp. 164-168.
69. Oomens, A. C. Experience with a Dual Detector Headspace Gas Chromatograph for
Acrylonitrile Analysis. In Applications of Headspace Gas Chromatography, B. Kolb,
ed., Heyden, London, 1980. pp. 11-116.
70. GC-MS Methods 624 and 625 for the Determination of Organic Priority Pollutants in
Water. Available from the U.S. Environmental Protection Agency, Cincinnati, Ohio,
1982.
71. Singh, H. B., L. J. Salas, A. Smith, R. Stiles, and H. Shigeishi. Atmospheric
Measurements of Selected Hazardous Organic Chemicals. EPA-600/53-81-031
(PBS 1-200-628). U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, 1981.
107
-------
APPENDIX
DEFINITION OF TERMS
absorbance: the logarithm to the base of 10 of the reciprocal of transmittance.
accuracy: the degree to which a measured value agrees with the true or accepted
reference value (e.g., pollutant concentration), usually expressed as the percentage
of the true or reference value represented by the difference between the two
(true and measured) values.
acidic compound: a compound which dissociates in water to form a hydrogen ion (proton)
and the corresponding anion (e.g., acetic acid dissociates into a hydrogen ion and
the acetate anion).
adsorbate: material that has been retained by the process of adsorption.
adsorbent: solid material on the surface of which adsorption takes place."
adsorption: a physical process in which molecules of gas, of dissolved substances, or of
liquids, adhere in an extremely thin layer to the surfaces of solid bodies with which
they are in contact.
aerosol: a dispersion of solid or liquid particles in gaseous media.
air at normal conditions (standard air): air at 50 percent relative humidity, 70 F and
29.92 in. Hg (21 C and 760 mm Hg). These conditions are chosen in recognition of
the data which has been accumulated on air-handling equipment. They are
sufficiently near the 25 C and 760 mm Hg commonly used for indoor air
contamination work that no conversion or correction ordinarily need be applied.
air pollution: the presence of unwanted material in the air. The term "unwanted
material" here refers to material in sufficient concentrations, present for a
sufficient time, and under circumstances to interfere significantly with comfort,
health, or welfare of persons, or with the full use and enjoyment of property.
aliquot: a representative portion of the whole.
alkali flame detector (AFD): see nitrogen phosphorus detector.
atmosphere, an: a unit of pressure equal to the pressure exerted by a vertical column of
mercury 760 mm high, at a temperature of O C, and under standard gravity.
atmosphere, synthetic: a specific gaseous mass containing any number of constituents and
in any proportion produced by man for a special purpose.
108
-------
atmosphere, the: the whole mass of air surrounding the earth and being composed largely
of oxygen and nitrogen.
basic compounds: compounds which protonate (add a hydrogen ion) in water to form a
cation (e.g., amines).
breathing zone: that location in the atmosphere at which persons breathe.
bubbler: a sampling device consisting of a gas disperser immersed in an absorbing liquid.
bubbler, fritted: a bubbler having a frit as the gas disperser.
calibration: establishment of a relationship between the response of a measurement
system obtained by introducing various calibration standards into the system.
The calibration levels should bracket the range of levels for which actual
measurements are to be made.
chemiluminescence detection: the measurement of emitted visible or ultraviolet
radiation resulting from the reaction of a pollutant with a reactive gas
(e.g., detection of nitric oxide by reaction with ozone).
column chromatography: a separation technique wherein the stationary phase is a solid
(e.g., silica gel) and the mobile phase is a liquid, usually an organic solvent or
mixture of solvents. The technique is usually performed using a tubular column
operating an ambient pressure, and is most commonly used for the removal of
potential interferences prior to determination of the compound of interest.
collection efficiency: the percentage of a specified substance retained by a gas cleaning
or sampling device.
colorimeter: an instrument used for color measurement based on optical comparison with
standard colors.
condensate: liquid or solid matter formed by condensation from the vapor phase.
In sampling, the term is applied to the components of an atmosphere which have
been isolated by simple cooling.
condensation: the process of converting a material in the gaseous phase to a liquid or
solid state by decreasing temperature, by increasing pressure, or both. Usually in
air sampling only cooling is used.
contaminant: a material added by human or natural activities which may, in sufficient
concentrations, render the atmosphere unacceptable.
cryogenic collection (trapping): a sampling process wherein an air sample is passed
through a cooled trap (usually using liquid argon or similar material as the cryogen)
to collect organic compounds.
density: the mass per unit volume of substance.
derivatization: a sampling and analysis process wherein a compound to be monitored is
converted to another more stable and/or readily detectable compound via chemical
reaction during the sampling or analysis step.
109
-------
desorption: the process of freeing from a sorbed state.
detection limit: the minimum quantity of a compound which yields a "measureable
response". Many statistical definitions of "measureable response" are in use.
One must be careful to differentiate "instrumental detection limit", which refers to
the minimum quantity of material introducible into a measurement system which
can be detected from "method detection limit" which refers to the minimum
concentration of compound in the sample which, when carried through the entire
sampling and analysis process, can be detected.
diffusion, molecular: a process of spontaneous intermixing of different substances,
attributable to molecular motion and tending to produce uniformity of
concentration.
dispersion: the most general term for a system consisting of particulate matter
suspended in air or other gases.
diurnal: recurring daily. Applied to (variations in concentration of) air contaminants,
diurnal indicates variations that follow a distinctive pattern and which recur from
day to day.
droplet: a small liquid particle of such size and density as to fall under still conditions
but which may remain suspended under turbulent conditions.
dust: a loose term applied to solid particles predominantly larger than colloidal and
capable of temporary suspension in air or other gases. Dusts do not tend to
flocculate except under electrostatic forces; they do not diffuse but settle under the
influence of gravity. Derivation from larger masses through the application of
physical force is usually implied.
efficiency: a measure of the performance of a collector. Usually it is the ratio of the
amount collected to the inlet loading, expressed in percentage.
efficiency, fractional: the mean collection efficiency for specific size fractions of a
contaminant. Commonly this term has been applied to the performance of air
cleaning equipment towards particulate matter in various size ranges.
electron capture detector (ECD): a detection device for gas chromatography which
responds sensitively and selectively to electron deficient (e.g., halogenated,
nitrosubstituted) compounds.
elute: to remove sorbed materials from a sorbent by means of a fluid.
emissions: the total of substances discharged into the air from a stack, vent, or other
discrete source.
filter: a porous medium for collecting particulate matter.
filter, controlled pore: a filter of various plastics or metals having a structure of
controlled uniform pore size. Sometimes referred to as a membrane or molecular
filter.
110
-------
flame ionization detector (FID): a detection device for gas chromatography which
responds to most organic compounds.
flame photometric detector (FPD): a detection device for gas chromatography which
responds selectively to sulfur- and phosphorus-containing compounds.
flowmeter: an instrument for measuring the rate of flow of a fluid moving through a
pipe or duct system. The instrument is calibratd to give volume or mass rate of
flow.
fluorescence spectrometry: the measure of ultraviolet or visible radiation emitted by a
compound after excitation with radiation of a lower wavelength. The technique is
widely used for the determination of polynuclear aromatic hydrocarbons.
fly ash: the finely divided particles of ash entrained in flue gases arising from the
combustion of fuel. The particles of ash may contain incompletely burned fuel. The
term has been applied predominantly to the gas-born ash from boilers with spreader
stoker, underfeed stoker, and pulverized fuel (coal) firing,
fog: a loose term applied to visible aerosols in which the dispersed phase is liquid.
Formation by condensation is usually implied. In meterorology, a dispersion of
water or ice.
fractionation: the process of separating a mixture into components having different
properties (as by distillation, precipitation, or screening).
freezing out: see sampling, condensation or cryogenic collection.
frit: a porous material permeable to gas flow usually made by sintering microbeads of an
appropriate material.
fume: properly, the solid particles generated by condensation from the gaseous state,
generally after volatilization from melted substances, and often accompanied by a
chemical reaction such as oxidation. Fumes flocculate and sometime coalesce.
Popularly, the term is used in reference to any or all types of contaminant, and in
many laws or regulations with the added qualification that the contaminant have
some unwanted action.
gas: one of the three states of aggregation of matter, having neither independent shape
nor volume and tending to expand indefinitely.
gas chromatography: a separation technique for organic compounds wherein the
stationary phase is a solid, liquid coated on a solid, or liquid coated or bonded to the
interior column wall (capillary column) and the mobile phase is an inert gas.
gas meter: an instrument for measuring the quantity of a gas passing through the meter.
gasometer: an apparatus employing a calibrated volume which is used to calibrate
gas-measuring devices.
grab sample: see sampling, instantaneous.
Ill
-------
gustiness: now referred to as intensity of turbulence which is defined as the ratio of the
root mean square of wind velocity fluctuations to the mean wind velocity.
Hall electrolytic conductivity detector (HECD): a detector device for gas
chromatography which responses selectively to halogenated, sulfur containing
compounds, or nitrogen containing compounds (depending on operating mode
selected).
halogenated compound: a compound containing chlorine, bromine, or iodine (chlorinated
compounds being the most commonly encountered in ambient air).
high performance liquid chromatography: a separation technique wherein the mobile
phase is a liquid and the stationary phase is a solid, usually having a particle
diameter of 10 urn or less. HPLC is similar to column chromatography except that
small particle diameter stationary phases and high pressures are used to achieve
faster analysis and greater resolution.
impaction: a forcible contact of particles of matter, a term often used synonymously
with impingement.
impactor: a sampling device that employs the principle of impaction (impingement).
The "cascade impactor" refers to a specific instrument which employs several
impactions in series to collect successively smaller sizes of particles.
impingement: the act of bringing matter forcibly in contact. As used in air sampling,
impingement refers to a process for the collection of particulate matter in which
the gas being sampled is directed forcibly against a surface.
impingement, dry: the process of impingement carried out so that particular matter
carried in the gas stream is retained upon the surface against which the stream is
directed. The collecting surface may be treated with a film of adhesive.
impingement, wet: the process of impingement carried out within a body of liquid, the
latter serving to retain the particulate matter.
impinger: see bubbler.
internal standard: a known quantity of a reference compound added to a collected sample
for use in the quantification of other compounds.
inversion: a reversal of the normal atmospheric temperature gradient, thus an increase
of temperature of the air with increasing altitude.
ionic or ionizable compound: a compound which dissociates in water to give ionic species
(i.e., acidic or basic compounds).
isokinetic: a term describing a condition of sampling, in which the flow of gas into the
sampling device (at the opening or face of the inlet) has the same flow rate and
direction as the ambient atmosphere being sampled.
lapse rate: the rate of change of the absolute value of any meteorological element with
increase of height. vWhen used without modifier, it refers to the rate of decrease of
temperature with increase of height).
11Z
-------
mass concentration: concentration expressed in terms of mass of substance per unit
volume of gas or liquid.
mass spectroscopy: a widely used analytical instrument capable of identifying and
quantitifying organic materials on the basis of the mass fragmentation pattern.
Most commonly used for organic analysis in combination with gas chromotography
(i.e., GC-MS).
mass median size: a measurement of particle size for samples of particulate matter,
consisting of that diameter such that the mass of all larger particles is equal to the
mass of all smaller particles.
meteorology: the science dealing with the atmosphere and weather conditions.
micro-climatology: the science that deals with the climate of restricted areas and
investigates their phenomena and causes.
micro-meteorology: the study of the meteorological characteristics of a local site that
is usually small and often is confined to a shallow layer of air next to the ground.
mist: liquid, usually water in the form of particles suspended in the atmosphere at or near
the surface of the earth; small water droplets floating or falling, approaching the
form of rain, and sometimes distinguished from fog as being more transparent or as
having particles perceptibly moving downward.
mobile phase: in chromatography, the separation medium which is in motion.
month: for reporting analyses of outdoor air on a monthly rate results are calculated to a.
base of thirty days.
neutral compound: a compound which does not ionize in water (e.g. not acidic or basic).
nitrogen-phosphorous detector (NPD): a detection device for gas chromatography which is
sensitive and selective for nitrogen- and phosphorous-containing organic compounds.
opacity rating: a measurement of the opacity of emissions, defined as the apparent
obscuration of an observer's vision to a degree equal to the apparent obscuration of
smoke of a given rating on the Ringelmann Chart.
orifice meter: a flowmeter, employing as the measure of flow rate the difference
between the pressures measured on the upstream and downstream sides of the
orifice (that is, the pressure differential across the orifice) in the conveying pipe
or duct.
particle: a small discrete mass of solid or liquid matter.
particle concentrations: concentration expressed in terms of number of particles per unit
volume of air or other gas.
Note - On expressing particle concentration the method of determining the
concentration should be stated.
113
-------
particle fall: a measurement of air contamination consisting of the mass rate at which
solid particles deposit from the atmosphere. A term used in the same sense as the
older terms dust fall and soot fall but without any amplication as to nature and
source of the particles.
particle size: an expression for the size of liquid or solid particles expressed as the
average or equivalent diameter.
particle size distribution: the relative percentage by weight or number of each of the
different size fractions of particulate matter.
particulate: solids or liquids existing in the form of separate particles.
phase distribution: the relative amounts of compound associated with the particle and gas
phases in the atmosphere.
photochemical reaction: any chemical reaction that is initiated as a result of absorption
of light.
photochemical smog: a type of air pollution resulting from photochemical reactions.
photoionization detector (PID): a detection device for gas chromatography which detects
aromatic, halogenated, and olefinic compounds but is relatively insensitive for
aliphatic compounds. The selectivity can be adjusted by the choice of lamp energy.
ppb: a unit of measure of the concentration of gases in air expressed as parts of the gas
per billion (10^) parts of the air-gas mixture, normally both by volume (ppbv).
ppm: a unit of measure of the concentration of gases in air expressed as parts of the gas
per million parts of the air-gas mixture, normally both by volume (ppmv).
precipitation, electrostatic: a process consisting of the separation of particulate matter
from air or other gases under the influence of an electrostatic field.
precipitation, meteorological: the precipitation of water from the atmosphere in the
form of hail, mist, rain, sleet and snow. Deposits of dew, fog and frost are
excluded.
precipitation, thermal: a process consisting of the separation of particulate matter from
air and other gases under the influence of a relatively large temperature gradient
extending over a short distance. In the "Thermal Precipitator" (a sampling
instrument), the air or gas is drawn slowly through a narrow chamber across which
extends a heated wire, particulate matter being deposited upon the adjacent
collecting surface.
precipitation, ultrasonic: a process consisting of the separation of particulate matter
from air and other gases following agglomeration induced by an ultrasonic field.
precipitator, electrostatic: apparatus employing electrostatic precipitation for the
separation of particles from a gas stream. The apparatus may be designed either for
sampling or for cleaning large volumes of gas.
114
-------
precision: the degree of agreement of repeated measurements of the same property,
expressed in terms of dispersion of test results about the mean result obtained by
repetitive testing of a homogeneous sample under specified conditions. The
precision of a method is expressed quantitatively as the standard deviation
computed from the results of a series of controlled determinations.
pressure, gage: the difference in pressure existing within a system and that of the
atmosphere. Zero gage pressure is equal to atmospheric pressure.
pressure, static: the pressure of a fluid at rest, or in motion, exerted perpendicularly
to the direction of flow.
pressure, total: the pressure representing the sum of static pressure and velocity pressure
at the point of measurement.
pressure, velocity: that pressure caused by and related to the velocity of the flow of
fluid; a measure of the kinetic energy of the fluid.
probe: a tube used for sampling or for measuring pressures at a distance from the actual
collection or measuring apparatus. It is commonly used for reaching inside stacks
and ducts.
quality assurance: a system of activities designed to provide assurance that the quality
control system is performing adequately.
quality control: a system of specific efforts designed to test and control the quality of
the data obtained.
radiosonde: a miniature radio transmitter with instruments that is carried aloft (as by an
unmanned balloon) for broadcasting by means of precise tone signals or other
suitable method the humidity, temperature, pressure, or other parameter every few
seconds.
resin: a porous polymer adsorbent such as Tenax^ for collection of gas phase organic
compounds.
rotometer: a device, based on the principle of Stoke's law, for measuring rate of fluid
flow. It consists of a tapered vertical tube having a circular cross section, and
containing a float that is free to move in a vertical path to a height dependent upon
the rate of fluid flow upward through the tube.
sample, cumulative: a sample obtained over a period of time with (1) the collected
atmosphere being retained in a single vessel, or (2} with a separated component
accumulating into a single whole. Examples are dust sampling in which all the dust
separated from the air is accumulated in one mass of fluid: the absorption of acid
gas in an alkaline solution; and collection of air in a plastic bag or gasometer. Such
a sample does not reflect variations in concentration during the period of sampling.
sample, running: withdrawal of a portion of the atmosphere over a period of time with
continuous analysis or with separation of the desired material continuously and in a
"linear" form. Examples are continuous withdrawal of the atmosphere accompanied
by adsorption of a component in a flowing stream of absorbent or by filtration in a
115
-------
moving strip of paper. Such a sample may be obtained with a considerable
concentration of the contaminant but it still indicates fluctuations in concentration
which occur during the period of sampling.
sampling: a process consisting of the withdrawal or isolation of a fractional part of a
whole. In air or gas analysis, the separation of a portion of an ambient atmosphere
with or without the simultaneous isolation of selected components.
sampling, condensation: a process consisting of the collection of one or several
components of a gaseous mixture by simple cooling of the gas stream in a device
which retains the condensate.
sampling, continuous: sampling without interruptions throughout an operation or for a
predetermined time.
sampling, instantaneous: obtaining a sample of an atmosphere in a very short period of
time such that this sampling time is insignificant in comparison with the duration of
the operation or the period being studied.
sampling, intermittent: sampling successively for limited period of time throughout an
operation or for a predetermined operiod of time. The duration of sampling periods
and of the intervals between are not necessarily regular and are not specified.
sampling, source: withdrawal, with or without simultaneous isolation of specific
components, of a portion of the offgases from a source of pollutants. Sometimes
referred to as stack sampling when withdrawal is from a chimney, duct, or stack.
sampling train: the assemblage of equipment necessary to sample atmospheres.
scrubber: a type of apparatus used in sampling and in gas cleaning in which the gas is
passed through a space containing wetted "packing" or spray.
sensor: a device designed to respond to a physical stimulus (as temperature, illumination,
and motion) and transmit a resulting signal for interpretation, or measurement, or
for operating a control.
series collection: an operation involving the use of two or more collectors joined in
series.
settling velocity: the terminal rate of fall of a particle through a fluid as induced by
gravity or other external force; the rate at which frictional drag balances the
accelerating force (or the external force).
smog: a term derived from smoke and fog, applied to extensive atmospheric
contamination by aerosols, these aerosols arising partly through natural processes
and partly from the activities of human subjects. Now sometimes used loosely for
any contamination of the air.
smoke: small gas-borne particles resulting from incomplete combustion, consisting
predominantly of carbon and other combustible material, and present in sufficient
quantity to be observable independently of the presence of other solids.
116
-------
soot: agglomerations of particles of carbon impregnated with "tar", formed in the
incomplete combustion of carbonaceous material.
sorbent: a liquid or solid medium in or upon which materials are retained by absorption
or adsorption.
Soxhlet apparatus: an apparatus for use in extracting fatty or other material with a
volatile solvent (as ether, alcohol, or benzene) in which the solvent is recirculated
by evaporation and subsequent condensation.
specific gravity: the ratio of the density of the substance in question to the density of a
reference substance at specified conditions of temperature and pressure.
spectrometry: a method of identification of a compound by identification of the spectrum
produced.
spectrophotometry: a method for identification of substances and determination of their
concentration by measuring light transmittance in different parts of the spectrum.
standard operating procedures (SOP): a detailed description of the operation of a
sampling or analysis system for a specific application.
stationary phase: in chromatography, the separation medium which is held fixed.
temperature, absolute: (a) temperature measured on the thermodynamic scale, designated
as Kelvin (K). (b) temperature measured from absolute zero (-273.18 C or
-459.58 F). The numerical values are the same for both the Kelvin scale and the
ideal gas scale.
Tenax^: a polyphenyloxide polymer prepared as a porous adsorbent for determination of
gas phase organic compounds. The relatively good thermal stability of Tenax^-
allows recovery of adsorbed organics by thermal desorption.
thermionic detector: see nitrogen-phosphorus detector.
thin layer chromatography (TLC): a chromatographic technique wherein the stationary
phase is a solid coated as a thin layer on a glass plate and the mobile phase is a
liquid. The technique is commonly used for the determination of polynuclear
aromatic hydrocarbons, using fluorescence detection.
validation, data: a systematic effort to review data to identify outliers or errors and
thereby cause deletion or flagging of suspect values to assure the validity of the
data for the user.
validation, method: the process of documenting the performance characteristics of a
method through the analysis of blanks and replicate samples of known analyte
concentration. The analyte concentrations tested should cover the range likely to
be encountered in the actual monitoring situation.
vapor: the gaseous phase of matter that normally exists in a liquid or solid state.
117
-------
volume concentration: concentration expressed in terms of gaseous volume of substance
per unit volume of air or other gas usually expressed in parts per million (ppmv) or
parts per billion (ppbv).
year: for reporting analyses of outdoor air on a yearly rate twelve 30-day months are
to be used.
118
-------
TOPIC INDEX
Adsorbents, solid
Calibration techniques
Centrifugal collection
Chemiluminescence detection
Classes of organic compounds, definition
Colorimetric detection
Column chromatography
Cryogenic trapping
Data reporting
Derivatization techniques
Electron capture detection
Electrostatic precipitation
Filtration
Flame ionization detection
Flame photometric detection
Fluorescence spectroscopy
Gas chromatography (GC)
Gas chromatography-mass spectroscopy (GC-MS)
High performance liquid chromatography (HPLC)
Infrared spectroscopy
Pages
38-41
96-98
46-47
55
76-80
52, 56
74
41-42
35, 100-102
43-44
54, 62-65
47
44-46
53, 62-65
55, 62-65
58
58-68
68-71
71-74
53, 56
119
-------
Pages
'.. *•'
Impingers 42
Impaction 46-47
lonization devices 53-54
Mass spectrometry, low resolution 58
Meteorology 17
Monitoring objectives, definition 8-18
Photoionization detection 53-54, 62-65
Photometric detection devices 55
Quality assurance 27-35, 97-102
Quality control, routine 98-100
Regulatory background 7 » " 4-7
4
Safety considerations ' , • . 35-36
t ' *-
Sampling and analyses methods, literature sources 18
Sampling and analysis methods, selection process 18-24
Sampling and analysis, specific methods 74-94
Sampling methods, gas phase components 37-44
Sampling methods, particle bound components 44-47
Sampling strategy, selection of 24-26
Screening techniques, field 49-56
Screening technique, laboratory 56-58
Spectroscopic devices 53, 74
Thin layer chromatography 74
Ultraviolet/visible spectroscopy 53, 74
Whole air collection 42-43
120
------- |