RESEARCH     TRIANGLE     INSTITUTE
                                      Final Report
                        CONCEPTS FOR DEVELOPMENT OF FIELD USABLE

                           TEST ATMOSPHERE GENERATING DEVICES

                                           by

                               Research Triangle Institute

                      Research Triangle Park, North Carolina 27709
                                 Contract No. 68-02-1242
                                 Program Element 1HA327
                           EPA Project Officer:   J.  F.  Walling
                 Quality Assurance & Environmental Monitoring Laboratory
                          National Environmental Research Center
                                 Research Triangle Park
                                 North Carolina  27711
                                     December 1973
                                      Prepared  for

                            Office  of  Research  and  Monitorinp,
                          U.S.  Environmental  Protection  Agency
                                 Washington,  D.C. 20460
RESEARCH  TRIANGLE   PARK,  NORTH  CAROLINA
27709

-------
                               ABSTRACT

     The purpose of this project was to examine possible physical,  physical
plus chemical, and biological concepts applicable to the development of
field usable test atmosphere generating devices.  The primary activity for
application of these devices is the round-robin survey to assess instrument/
operator performance on a routine periodic basis.  Ten concepts (i.e.,
desorption, effusion, thin film evaporators, novel permeation, radiolysis-
photolysis, plasma discharge, thin films of dissolved reactants, electro-
chemical films of laminated reactants, and biological generation) were
examined for each of the five required pollutants (i.e., sulfur dioxide
nitrogen dioxide, carbon monoxide, ozone, and 1-butene) and a total of
fifty individual discussions prepared.  The program was divided into three
distinct phases:  (1) comprehensive literature search and preparation of
technical discussions; (2) panel review of 50 discussions; and  (3) compar-
ative evaluation.  Of the concepts investigated, the effusion method is
recommended for generating test atmospheres for sulfur dioxide, carbon
monoxide, and 1-butene, while for ozone, UV generation modified for more
reliable and easier operation is recommended.  Gas phase titration of
nitric oxide with ozone modified for field use is recommended for
generating test atmospheres of nitrogen dioxide.
     This report was submitted in fulfillment of Contract No. DU68-02-1242
by Research Triangle Institute under the sponsorship of the Environmental
Protection Agency.  Work was completed as of December 1973.
                                  iii

-------
                         TABLE OF CONTENTS


                                                                   Page

ABSTRACT                                                           ill

ACKNOWLEDGMENT                                                       v

Section

  1.0    INTRODUCTION                                                1

         1.1  Background/Objectives                                  1

         1.2  Approach                                               2

  2.0    OVERALL SYSTEM CONCEPTS                                     5

         2.1  Optimal Surveillance System Application of Test
              Atmosphere Generation Methods                          5

         2.2  Classification of Generation Concepts                  7

         2.3  Constraints and Objectives                             8

         2.4  Modular Approaches                                     9

  3.0    SPECIFIC METHODS                                           11

         3.1  Contemporary Test Atmosphere Generation Methods       11

              3.1.1  Sulfur Dioxide                                 11

              3.1.2  Nitrogen Dioxide                               13

              3.1.3  Carbon Monoxide                                15

              3.1.4  Ozone                                          16

              3.1.5  Butene-1                                       18

         3.2  Physical Methods                                      18

              3.2.1  Desorption                                     18

                     3.2.1.1  Sulfur Dioxide                        19

                     3.2.1.2  Nitrogen Dioxide                      24

                     3.2.1.3  Carbon Monoxide                       26

                     3.2.1.4  Ozone                                 28

                     3.2.1.5  Butene-1                              31
                                    vi

-------
Section                                                            Page
              3.2.2  Effusion Methods                               34
                     3.2.2.1  Sulfur Dioxide                        37
                     3.2.2.2  Nitrogen Dioxide                      40
                     3.2.2.3  Carbon Monoxide                       43
                     3.2.2.4  Ozone                                 46
                     3.2.2.5  1-Butene                              47
              3.2.3  Thin Film Evaporation                          49
                     3.2.3.1  Sulfur Dioxide                        53
                     3.2.3.2  Nitrogen Dioxide                      56
                     3.2.3.3  Carbon Monoxide                       56
                     3.2.3.4  Ozone                                 56
                     3.2.3.5  Butene-1                              57
              3.2.4  Novel Permeation Methods                        57
                     3.2.4.1  Sulfur Dioxide                        62
                     3.2.4.2  Nitrogen Dioxide                      65
                     3.2.4.3  Carbon Monoxide                        67
                     3.2.4.4  Ozone                                 68
                     3.2.4.5  Butene-1                              68
         3.3  Chemical Methods                                      70
              3.3.1  Radiolysis,  Photolysis, and Thermolysis         70
                     3.3.1.1  Sulfur Dioxide                        71
                     3.3.1.2  Nitrogen Dioxide                      74
                     3.3.1.3  Carbon Monoxide                        75
                     3.3.1.4  Ozone                                 76
                     3.3.1.5  Butene-1                              80
              3.3.2  Plasma Discharge                               84
                     3.3.2.1  Sulfur Dioxide                        84
                     3.3.2.2  Nitrogen Dioxide                      87
                                    vii

-------
Section                                                            Pa8e

                     3.3.2.3  Carbon Monoxide                       93

                     3.3.2.4  Ozone                                 94

                     3.3.2.5  Plasma Generation of 1-Butene
                              Test Atmospheres                     100

              3.3.3  Thin Films of Dissolved Reactants             100

                     3.3.3.1  Sulfur Dioxide                       100

                     3.3.3.2  Nitrogen Dioxide                     103

                     3.3.3.3  Carbon Monoxide                      104

                     3.3.3.4  Ozone                                104

                     3.3.3.5  Butene-1                             104

              3.3.4  Electrochemical Methods                       105

                     3.3.4.1  Sulfur Dioxide                       105

                     3.3.4.2  Nitrogen Dioxide                     108

                     3.3.4.3  Carbon Monoxide                      112

                     3.3.4.4  Ozone                                113

                     3.3.4.5  1-Butene                             114

              3.3.5  Films of Laminated Reactants                  114

                     3.3.5.1  Sulfur Dioxide                       1U

                     3.3.5.2  Nitrogen Dioxide                     117

                     3.3.5.3  Carbon Monoxide                      117

                     3.3.5.4  Ozone                                117

                     3.3.5.5  Butene-1                             118

         3.4  Biological Generation                                119

              3.4.1  Sulfur Dioxide                                119

              3.4.2  Nitrogen Dioxide                              119

              3.4.3  Carbon Monoxide                               119

              3.4.4  Ozone                                          123
                                   viii

-------
Section                                                            Page





              3.4.5  1-Butene                                      123




         3.5  Other Methods                                        132




              3.5.1  Foamed Plastic                                132




              3.5.2  Microsyringe                                  133




              3.5.3  Burn Stick                                    133




              3.5.4  Bubbler Systems                               133




              3.5.5  Exponential Dilution                          134




  4.0  COMPARATIVE EVALUATION                                      136




         4.1  Selection of Panel                                   136




         4.2  Evaluation Technique                                 136




         4.3  Panel Study                                          144




         4.4  Results                                               145




  5.0  RECOMMENDATIONS                                              153
                                     ix

-------
                             LIST OF FIGURES


Figure

3-1    SO- test atmosphere generator                           23

3-2    Adsorption isotherms at 0° and 25°C for ozone in
       oxygen on Davison silica gel                            29

3-3    Test atmosphere generator                               36

3-4    Constant low pressure effusion source                   40

3-5    On-off valves for CLOPES source                         42

3-6    Storage unit with metering valve                        45

3-7    Test generator with metering valve                      48

3-8    Contact column test generator                           51

3-9    Time lag vs pressure curve                              60

3-10   Permeation device                                       61

3-11   Permeation test atmosphere generator                    64

3-12   Thermolysis test atmosphere generator                   74

3-13   Ozone source                                            78

3-14   Ozone test atmosphere generator                         79

3-15   Aerosol generator                                       81

3-16   Pyrolysis gas generator                                 81

3-17   Pyrolysis test atmosphere generator                     83

3-18   Plasma test atmosphere generator                        88

3-19   Electrolytic cell  for quantitative NO  generation        89

3-20   Nitrogen dioxide test atmosphere  generator              92

3-21   Instrument ozonizer                                     95

3-22   Plasma test atmosphere generator  for ozone.             98

 3-23   Reactant  test  atmosphere generator                      101

 3-24   Canned  dopant  gas  with  electrolytic  drive              109

-------
Figure                                                        I28S.

3-25   Electrolytic test atmosphere generator for
       nitrogen dioxide

3-26   Dissolved gas test atmosphere generator                117

3-27   Biological test atmosphere generator for
       carbon monoxide                                        122

3-28   Effect of enzyme concentrations on reaction rate,
       assuming substrate concentrations at saturation
       amounts                                                124

3-29   Effect of temperature on reaction rate for
       enzymes                                                125

3-30   Effect of pH on ethylene production                    126

 3-31   Substrate vs product profile  at  constant  enzyme
       concentration                                          127

 3-32   Enzyme  test  atmosphere  generator for  ethylene         130
                                     xi

-------
                           LIST OF TABLES

Tables                                                         Page

1-1    Classification of New Field Usable Test Atmosphere
       Generating Devices                                         3

2-1    Field Usable Test Atmosphere Generating Methods
       Evaluation Criteria                                        9

4-1    Summary of Methods—Sulfur Dioxide                      137

4-2    Summary of Methods—Nitrogen Dioxide                    138

4-3    Summary of Methods—Carbon Monoxide                     139

4-4    Summary of Methods—Ozone                               140

4-5    Summary of Methods—1-Butene                            141

4-6    Method Ranking System                                   142

4-7    Comparative Ranking of Methods for 	                 146

4-8    Comparative Ranking of Methods for S02                  147

4-9    Comparative Ranking of Methods for NO-                  148

4-10   Comparative Ranking of Methods for CO                   149

4-11   Comparative Ranking of Methods for 0_                   150

4-12   Comparative Ranking of Methods for 1-Butene             151

4-13   Summary of Comparative Evaluation                       152
                                   xii

-------
1.0  INTRODUCTION

1.1  Background/Objectives
     Air pollution surveillance networks are an integral part of the total
national effort to control air pollution.  Regional control, source identi-
fication, episode reaction, and other elements of the control program are
keyed to air quality measurements by surveillance networks.  The generation
and promulgation of regulatory actions are keyed to air quality data.  Thus,
the data from the air quality measurements have very important economic and
social as well as scientific meaning.  If the data are erroneous, the
consequences may be costly and public confidence in the control program
will be eroded.
     Surveillance networks are required to monitor routinely the common
pollutants which are considered as measures of air quality.  These include
sulfur dioxide, carbon monoxide, nitrogen dioxide, ozone, and nonmethane
hydrocarbons.  Because of the difficulties in measuring ambient concentra-
tions of these pollutants, a variety of chemical and physical methods have
been applied in commercially available instrumentation.  These include
chemiluminescent, flame photometric, infrared absorptive, gas
chromatographic/flame ionization, dual-isotope-fluorescence spectrometric,
coulometric, colorimetric, and other detection methods.  Obviously, air
quality data from these diverse instruments will be subject to many different
variables with the result that the data will be of variable quality.
     The validation of accuracy and hence the usefulness of data from air
surveillance networks depends to a large extent upon the calibration of
instruments under actual operating conditions.  For continuous automated
measurements, a dynamic calibration is required whereby the measuring
instrument is calibrated with a known quantity of the gaseous pollutant in
a test atmosphere while the system is operational in the field.
     In addition to providing a basis for standardization of data,  test
atmospheres are required for quality control of the air quality  surveillance
network.  The great diversity in the types of instruments,  in the  abilities
of personnel, and in the types of organizations which are involved makes  some

-------
form of uniform quality control necessary.   One way of accomplishing this
is to provide, on a regular basis, samples of unknown test atmospheres to
each station in the surveillance network.  Each station will measure
pollutant levels in the test atmosphere and report the results.   Analysis
of the data thus generated will provide a confidence level for air quality
data from the total network, quality control on data from individual
stations, and indicate points where remedial action is required.

1.2  Approach
     The availability of test atmospheres thus fulfills three important
functions for air quality surveillance—calibration, standardization, and
network quality control.  To fulfill this role, methods for test atmosphere
generation must be identified and developed.  These methods must (1) have
a high degree of reliability, (2) be relatively immune to operator errors,
and (3) be sufficiently economical to allow broad application.  Several
test atmosphere generation methods have received much use in instrument
calibration.  These include permeation tubes, quantitative gas mixtures,
and UV ozone generators.  While these may be prime candidates for providing
broad test atmosphere generation capabilities, each has limitations  that
make the examination of other methods desirable.
     In Section 2.0, an overall system concept for assessing the performance
of an air quality surveillance network through use of test atmospheres is
discussed.  This includes a framework for classification of the various
methods for generation of test atmospheres so that comparative evaluation
can be accomplished.  The operational requirements and constraints for
viable test atmosphere generating methods and for an operational system
employing these methods are also discussed.  The modular approach in which
commonalty is attained in the components of methods employed for the various
pollutants is also described.
     In Section 3..0, the various specific methods proposed for test
atmosphere generation are described.  To insure comprehensiveness, it was
required that a method be considered for each of the five pollutants in each
of the classifications indicated by an asterisk in Table 1-1.  Even when  the

-------
            Table 1-1:  Classification of New Field Usable
                        Test Atmosphere Generating Devices

        Physical methods
             Desorption
             Effusion
             Thin film evaporates
             Permeation (novel methods)
        Chemical methods
             Radiolysis/photolysis/thermolysis                   *
             Plasma discharge                                    *
             Thin film evaporator                                *
             Electrochemical                                     *
             Laminated reactants                                 *
        Biological

        Other methods
      Required
     """Optional
method appears trivial for a specific pollutant, that fact must be stated
and substantiated.  This requires a basic set of 50 methods.  In
addition, to assure comprehensiveness in the subsequent analysis,
conventional methods of test atmosphere generation are also discussed and
additional methods that do not conform to the classifications are included.
     In order to provide common bases for evaluation, the various methods
are compared in Section 4.0.  The parametric measures—costs, weight,
operator training requirements, logistical demands, developmental status,
predicted accuracy and reliability—are employed to rank the various
methods.
     To identify the most promising methods and to specify the one that
is recommended for development, a panel of experienced scientists and

-------
engineers was given the data and information contained in Sections 2, 3,
and 4.  The results of the panel's deliberations are given in Section 4.4.
     In Section 5, recommendations compounded from the panel and the
project studies are presented.  These consist of a recommendation of a
method for each of the pollutants as well as the one method that is
recommended for immediate development because of its potential present
value to the air quality surveillance program.

-------
2.0  OVERALL SYSTEM CONCEPTS
     The impetus for development of field usable test atmosphere generation
methods is for standardization and quality control for air surveillance
networks.  Emphasis is on quality control because it represents the more
challenging problem and, if solved, will provide standardization capa-
bilities for air quality data.  In order to provide a basis for evaluation
of test atmosphere generation methods, the operation of an optimal quality
control system is described in this section.  Then a basis is given for
classification of the various test atmosphere generation methods and the
desirable attributes of such methods are given.

2.1  Optimal Surveillance System Application of Test Atmosphere
     Generation Methods
     The air quality surveillance system is assumed to consist of from
1,000 to 5,000 measurement stations grouped in a nonuniform pattern by
area, State, and region and providing air quality data to various control
agencies and data banks.  In the foreseeable future, it is expected that
the number of stations providing continuous reliable records will be small
but growing, and that there will be a large number of stations coming on
line, many of which will have sporadic records and low quality data.  The
problem is to provide test atmosphere generating devices, whereby the
quality of the data from any monitoring station can be accessed.
     Because the operation of air quality surveillance stations has been
decentralized, the degree of uniformity in personnel, resources, and
procedures will be minimal.  This leads to considerable variation in the
nature of the data that are produced, and this disparity is increased by
the variety of instrumentation that is employed in data collection.  Any
attempt to attain uniformity in station capabilities in the near future
will be difficult.  Thus, the only method that can attain meaningful surveillance
network operation is to determine the quality of the data produced by an
Independent testing procedure.  The development of field usable test atmosphere
generation methods that can be used to ascertain the data quality from each

-------
of the stations is a step toward attaining a viable independent testing
procedure.
     In order that test atmosphere generation be accomplished by a
uniform, reproducible method at each of the measuring stations, it is
desirable that both a basic test atmosphere generation accessory kit and
the individual test samples be provided by a central agency.  The kit
might consist of an air purification system, a manifold, a blower, and
certain flow measuring and control devices.  The samples, dependent on the
method, might consist of small containers to be attached to the manifold
by prescribed procedures.  It would be intended that the kit would be
assembled such that samples and instruments could readily be attached to
it to accomplish a given test.  A supply of materials such as might be
used in the air purifier would also be supplied with well-defined procedures
and controls for their use.
     Operation, at the measuring site, would begin with receipt in the
mail of a sample container and directions for its use.  The operator would
                  /
schedule the test at which time he would activate the air purification
system and blower, attach the sample container and instrument to the
manifold,  and, following the procedures given, would perform the measure-
ment.   The results would be recorded on forms provided and returned to the
central agency for evaluation.
     The central agency, on a scheduled basis, would procure or prepare the
necessary number of test samples and distribute them to the operating
stations.   At the end of each test the results would be recorded and
provided to the air quality control data collection center to use for
quality control.  Each station would be provided with a record of its
performance so that, if remedial action were required, it could be taken.
In addition, summary data could be provided to the administrative control
points of the air surveillance network for their use.
     In order that such a quality control system be operable, certain
restrictions on the methods for test atmosphere generation become obvious.
Before examining these restrictions, however, a framework for the classi-
fication of test atmosphere generation methods will be described.

-------
2.2  Classification of Generation Concepts
     A practical test atmosphere generation system involves at least three
basic functions.
     1)  Storage - The pollutant, either in pure or dilute form,  must
         be stored in some enclosure for distribution and retention
         until the test is made.  Storage may be as a gas, liquid,
         or adsorbed phase or may be in a chemically different form.
     2)  Dispensing - The stored form of the sample must be dispensed
         from storage at a known rate.  This may be by a metering
         valve; by a controlled reaction; by evaporation, effusion,
         or permeation; or by batch techniques such as with a gas
         syringe.
     3)  Dilution - It is necessary to mix the dispensed pollutant
         gas with a known quantity of air to obtain the desired test
         atmosphere.  This is accomplished in a mixing manifold in
         which the pollutant is mixed with air.  The air must be free
         of interferences.  This usually involves elimination of
         particulates and moisture, and reducing pollutant levels in
         the air below detection limits.
     Each of the above functions is required in test atmosphere generation
but, in a given method, the functions can often be combined.  Thus it is
desirable to examine other bases for test method classification.  Since
each method embodies a basic scientific principle, it is possible to use
these principles for classification.  Physical, physical plus chemical, and
biological methods form the basis for this type of classification.
     Most test atmosphere generation methods are based on physical
principles.  Physical storage may involve pressurized gas, entrapment of
gas in a solid material, or adsorption of the  gas on a surface.  Physical
methods for dispensing gas involve effusion, diffusion, evaporation, or
permeation.
     Chemical methods are based  on dissociation of a compound containing
the desired entity, combination  of materials to form a desired compound,  or
an activation process.  Gases such as carbon monoxide can be  formed by

-------
dissociation but characteristically in such reactions conditions must be
carefully controlled to obtain the desired product.  More often dissociation
is applicable as one step in a two-step process wherein an elemental form
is first produced and then reacted so as to combine with a second material
to produce the desired product.  Combining reactions are typified by
oxidation of carbon, sulfur, or nitrogen to produce CO, S02, or NC^.
Activation is most strongly associated with the production of ozone from
oxygen but also may be used as a driving force for other reactions.
     There is less variety in the availability of biological methods for
test atmosphere generation largely because of the complexity of these
processes and the resulting difficulty in releasing the desired gaseous
component from the biological system.
     Because of the nature of this survey of possible test atmosphere
generation methods, the latter classification scheme based on scientific
principles has been adopted.  It is given in Table 1-1 and is the basis for
the subsequent discussion of specific methods.  It must be recognized that
seldom does a generation method described in this classification stand
alone.  More often, a combination of methods is required wherein several
physical processes are combined with a chemical process to obtain a viable
generation method.  Then too, one process may fulfill several functions.
It is not uncommon to find the dispensing and dilution functions, for
example, combined in one piece of apparatus.  Thus, as with any classifica-
tion method, the boundaries are indistinct, and one must be alert to these
shortcomings in order to realize the objective—to provide a comprehensive
framework for surveying and comparing all possible generation methods.

2.3  Constraints and Objectives
     The evaluation criteria for field usable test atmosphere generation
methods are listed in Table 2-1.  Emphasis must be on an economically
viable system wherein test samples originate at a  central point, the
tests are run with minimum effort and minimum opportunity for error at a
large number of test sites, and the results are analyzed at the central
point to assess the quality of the data being generated at each station.
                                  8

-------
          Table 2-1:  Field Usable Test Atmosphere Generating
                      Methods Evaluation Criteria
1)  Simplicity


2)  Portability



3)  Cost
4)  Capacity


5)  Range



6)  Operating mode



7)  Accuracy



8)  Safety
The method should be usable by an untrained operator
with little opportunity for error.

The method should be applicable in a mode that
requires no complex logistics in providing test
atmospheres at many sites at frequent intervals.

The material and labor costs must be minimal to allow
wide use.  These include costs at the central distri-
bution point, the cost of the distributed item and
costs at test sites.  Where possible, costs should be
incurred in the central distribution point in pref-
erence to the test site in order to amortize the
costs over a much larger population of tests.

The generation method must be capable of providing
1-3 £/min of test atmosphere for periods up to 2 hr.

The pollutant levels in the test atmosphere must range
from the minimum detectable levels with the reference
test methods to alert levels.

The test atmosphere must be usable in a "blind" mode;
i.e., one in which the instrument operator does not
know the pollutant concentration.

A deviation of the actual concentration from the
specified concentration of +20 percent at the time of
use is acceptable.

The apparatus and components used to generate test
itmosphere must include provisions for operator safety
*v% rl 4*i*'w«r*>r«/'h**^*^^-{ r*f+ -»»«**i 11 A ^*s>wt A*^ 4- e*
                     and transportation requirements.
2.4  Modular Approaches

     In a modular approach, one attempts to integrate methods for
generation of the various types of test atmospheres so that  the number of
components associated with each individual type is minimized.  This  allows
a large investment in the components that are common to all  of the methods

with potentially beneficial effects on the overall system.

-------
     An obvious component that is common to many generation methods is the
supply of clean air for dilution.  Whether this is a compressed air source
or an air scrubbing apparatus (the trade-offs between these should be
analyzed), an air supply suitable for all types of atmospheres can be
obtained with no or relatively little additional effort over that required
to supply a method for one type of atmosphere.
     Additional opportunities for common modules include the following:
          1)  Mixing chambers
          2)  Compressed gas storage containers
          3)  Effusion modules
          4)  Thermostatically controlled chambers
          5)  Power supplies for electrochemical or
              plasma generation methods
          6)  Flowmeters and valves
          7)  Blowers and pumps
          8)  Instrument manifold
          9)  Dispensing valves
         10)  Dilution systems
     Component and module commonality may be sufficiently important so as
to favor selection of one method over an otherwise superior method in order
to enhance  the capability and availability of the overall system.
                                  10

-------
3.0  SPECIFIC METHODS

     Specific methods or devices for test atmosphere generation that are
described in this chapter are not necessarily optimized with respect to the
constraints given in Section 2.0.  Rather the approach was to describe methods
based on specified physical and chemical principles.  The subsequent compara-
tive evaluation and panel analysis is directed toward examining each method
with respect to the defined constraints.  In particular, the methods described
in Section 3.1 are those currently employed for calibration of air quality
instrumentation and are based primarily on technical requirements.  The
operational constraints of test atmosphere applications in quality control
of a surveillance network have not been considered in the evolution of these.

3.1  Contemporary Test Atmosphere Generation Methods
     Contemporary test atmosphere generation methods are referred to as
dynamic calibration systems.  These are available commercially for each of
the specific pollutants being considered and are employed for the field
calibration of air quality analyzers.

3.1.1  Sulfur Dioxide
A.  Principle
    Permeation tubes offer a simple method of preparing low level concen-
trations of S02 and are currently recommended and used to calibrate S02
analyzers both in the laboratory and  for field applications.  Permeation
tubes are prepared by sealing liquefied sulfur dioxide in fluorinated
 ethylene-propylene copolymer (FEP Teflon).   The permeation process  depends
primarily on dissolution of gas  in Teflon, permeation  through  the Teflon
wall, and evaporation from  the  outer  surface.  The  main driving  force  is
the difference in partial pressure between the inner and  outer walls of
the tube.  Permeation tubes are  highly  temperature  sensitive and require
that  the  temperature be controlled  to +0.10C.  A permeation tube can be
accurately  calibrated by  gravimetric  or volumetric  methods.  After  initial
equilibration  the permeation  rate for S02 tubes  is  essentially constant
                                   11

-------
until all the liquid has been consumed.  By proper selection of temperature,
tube length, wall thickness, and diluent flow rate, accurately known
concentrations of S0_ can be generated in the parts-per-billion to parts-
per-million range.  Sulfur dioxide permeation tubes are available from
several vendors, including the National Bureau of Standards, which offers
a certified SO- permeation tube as a primary standard.

B.  Physical Characteristics
    A typical system is 30.5 x 45.7 x 40.6 cm and weighs 22.7 kg-

C.  Operational Requirements
    Permeation tube test atmosphere generators require skilled operators,
a supply of clean air, and require several hours of warmup time.

D«  Maintenance Requirements
    Little maintenance is required.

E.  Shipping Requirements
    The permeation tube is small and weighs less than 10 g but a
dynamic calibration system is difficult to ship because it is a precision
electromechanical assembly and is easily breakable.

F.  Estimated Accuracy
    On the order of +2 percent under optimum conditions.

G.  Estimated Cost
    Permeation tube calibration systems are commercially available from
a number of vendors (i.e., Bendix, Tracer, Analytical Instruments Development
Inc., etc.) and cost from $1,000-$2,500, depending on the options.
                                   12

-------
H.  Advantages/Disadvantages
    Permeation tube calibration systems are inherently expensive, require
skilled operators, require precise temperature controls, lack ruggedness
and portability and require several hours temperature equilibration of
the bath (i.e., air or water) and the permeation tube.  Advantages of
S02 permeation tubes are accuracy (once calibrated), reliability, long
life (6 to 12 months), availability,  and ease of transport of the permeation
tube itself.

3.1.2  Nitrogen Dioxide
A.  Principle
    Due to problems associated with field use of N02 permeation tubes, an
alternate procedure (gas phase titration was developed for dynamic cali-
bration of chemiluminescent NO-NO-NO, analyzers.  The technique is based
                                 X   £t
upon application of the rapid gas phase reaction between NO and ozone to
produce a stoichiometric quantity of N02.

                            NO + 03 -*• N02 + 02

     Nitric oxide (50-100 ppm in N2) from a pressurized cylinder is diluted
with a constant flow of clean air to provide NO concentrations in the range
                              3
0.05 to 1 ppm (94 to 1880 yg/m ) and is then used to calibrate the NO and
NO  cycles of chemiluminescent NO-NO -N00 analyzers.  By incorporation of
  X                                 3C   fc
a calibrated ozone generator in the calibration apparatus upstream from the
point of NO addition, precise N02 concentrations can be generated by oxidation
of NO to NO. with 0~.  As long as a slight excess of NO is present, the
amount of 0, added is equivalent to the amount of NO consumed and is equivalent
to the concentration of N0« generated.  Field usable calibration systems based
on the gas phase titration method are commercially available and have the
capability for calibration of both 0- and N02 analyzers.
                                  13

-------
B.  Physical Characteristics
    In its commercial form, this is a 30.4  x 40.6 x 45.7 cm instrument
weighing about 22.7kg.

C.  Operational Requirements
    A calibrated supply of nitric oxide in  an inert gas is required in
addition to the normal manifold and diluent airstream.

D.  Maintenance Requirements
    Once operational little maintenance is  required.

E.  Shipping Requirements
    This is a bulky electromechanical assembly which must be carefully
packed for shipping.

F.  Estimated Accuracy
    Accuracy can be +3 percent.

G.  Estimated Cost
    This unit costs approximately $2,000.

H.  Advantages/Disadvantages
    The major advantage of the gas phase titration system is that the
system can be used to generate known concentrations of 0~, NO, and N0».
Major disadvantages are that the system is  designed for calibration of
chemiluminescent NO-NO -NO- analyzers (i.e., excess NO is present), Is
expensive, requires the services of a skilled operator, and requires
the use of a cylinder of NO in N~ and an expensive stainless steel
regulator.  At the present time the gas phase titration system is
considered to be more reliable for field use than NO, permeation tubes.
                                  14

-------
3.1.3  Carbon Monoxide
A.  Principle
    Calibration of CO analyzers is limited to the use of calibration gases
(CO in air or nitrogen), purchased in pressurized cylinders from various
manufacturers (i.e., Scott, Matheson, Linde, etc.) and furnished with a
certificate of analysis as to the CO concentration.

B.  Physical Characteristics
     The physical characteristics are related to the size and weight of the
compressed gas cylinder, 1.8 kg, 5.1 cm in diameter and 38.1 cm long for
a lecture bottle; 61.4 kg, 22.9 cm-diameter and 132.1 cm-length for a
size 1A cylinder.

C.  Operational Requirements
     The cylinder can be turned on and used at any time.  Most operators
have familiarity with gas cylinders and regulators.  A calibration and
diluent airflow may be required, depending on the cylinder gas concen-
tration and the instrument calibration inlet system.

D.  Maintenance  Requirements
    These  are nominal.

E.  Shipping Requirements
    Shipping of  cylinders  of  compressed  gas is  required.   These  may vary
from  the small lecture bottles  to large  cylinders depending on usage and
operational decisions.

F.  Estimated Accuracy
    Analysis  of  component concentration certified to an accuracy of +2
percent can be  obtained from the manufacturer.
                                  15

-------
 G.  Estimated Cost
     This is dependent on cylinder size but is approximately $100.  For
 each location about $150 is required for regulators and flowmeters.

 H.  Advantages/Disadvantages
     Advantages of using compressed gases are simplicity of operation,
 minimum setup time and immediate usage on demand.   Disadvantages  include
 acceptance of manufacturer's certification as to CO content or recali-
 bration against the NBS standard, lack of portability, instability of
 CO at low concentrations (<25 ppm) in pressurized tanks (i.e., stable
 for approximately 6 months), safety aspects, and the need for several
 cylinders of different concentrations for multipoint calibrations.

 3.1.4  Ozone
 A.  Principle
     The production of a stable and reproducible concentration of ozone
 in air has always been a major problem in the operation and calibration
 of ozone analyzers.  In the absence of a primary ozone standard,  the
 calibration of ozone instrumentation has been accomplished by reference
 to the neutral-buffered potassium iodide method.  A dynamic calibration
 system using an ultraviolet ozone generator is recommended for generation
 of test atmospheres for calibration of ozone-oxidant instrumentation.
 This system is currently being evaluated by the National Bureau of
 Standards as a secondary ozone calibration source.  Briefly, the ozone
 source consists of an 8-inch current-controlled ultraviolet mercury arc
 lamp which irradiates a 5/8-inch quartz tube through which clean (compressed)
airflows at 5 i/min.  Ozone levels from low parts-per-billion to 1.0
 parts-per-million (1960 yg/m ) can be generated by variable shielding of
 the lamp envelope.  Recent field evaluation studies have shown that once
 initially calibrated (i.e., referenced to neutral-buffered potassium
 iodide method) the precision of generating ozone concentrations via
 the ozone generator is equal to or better than the precision of the
 measurement by the manual iodometric method.
                                   16

-------
     Ozone calibration systems utilizing the low pressure mercury arc lamp
are commercially available in various configurations from several vendors
(i.e., Bendix, MacMillan, etc.).

B.  Physical Characteristics
    These are dependent on the particular design employed but can range in
weight from 9.1 to 22.7 kg and in dimensions up to 30.5 x 40.6 x 45.7 cm.

C.  Operational Requirements
    Electrical power (115 VAC) and clean air are required.

D.  Maintenance Requirements
    These are nominal once the flow rates are calibrated.

E.  Shipping Requirements
    This is a 22.7 kg assembly.

F.  Estimated Accuracy
    Under ideal conditions, accuracies  can be about ^ 3  to  5 percent.

G.  Estimated Cost
    A commercial unit will cost between $1,000  and  $2,000.

H.  Advantages/Disadvantages
    Advantages of  this  ozone  generation system  are  the inherent short-term
and long-term stability,  portability,  reliability,  and ease of operation,
once  calibrated.   Disadvantages  include the  requirement' for reference to
the manual  iodometric method  and the effect   of flow rate and line voltage
fluctuations.
                                   17

-------
3.1.5  Butene-1
       Generation of test atmospheres of butene-1 for field applications
can be accomplished by several methods—permeation tubes, dynamic dilution,
or static dilution of a fixed volume or weight of contaminant into a fixed
volume of diluent air.   Static dilution systems are usually employed when
comparatively small volumes are required, as would be the case for produc-
tion of gas-phase standards for calibration of gas chromatographs, mass
spectrometers, and infrared spectrometers.  Here, microliter quantities of
butene-1 would be injected into a Tedlar bag and diluted with clean air to
sufficient volume to produce the desired parts-per-million concentration.
Dynamic dilution would involve single or double stage dilution of butene-1
and clean air to produce the desired parts-per-million concentration.
Permeation tubes which have previously been described are also available
for generation of parts-per-million calibration concentrations of butene-1.

3.2  Physical Methods
     The test atmosphere generation methods described in this section are
based on physical processes; i.e., desorption, effusion, evaporation, and
permeation.   The release of the pollutant species into an air stream is
determined by actual control of a flow rate occurring due to a pressure
gradient or by control of the vaporization rate from a condensed form of
the pollutant.  It is obvious that the physical properties of the five
pollutant gases are sufficiently different so as to produce a large variety
in the best apparent methods.

3.2.1  Desorption
       Probably no other technique has as much theoretical and practical
information in the literature as the adsorption/desorption technique.  In
the field of chromatography In which the two phenomena are Intimately
involved, the following serials are published:  Advances in Chromatography.
Separation Science, Chromatographic Reviews, Chromatographia. Journal of
Chromatography, Journal of Chromatographic Science, and Journal of Gas
                                  18

-------
Chromatography.  While both physical and chemisorption processes are
possible for the adsorbent-adsorbate pairs considered here, physical
adsorption is thought to predominate.  We base this conclusion on the
chromatographic work on the considered gases and the fact that relatively
low temperatures (below 100°C) are required for their total desorption.

3.2.1.1  Sulfur Dioxide
A.  Principles Involved
    Experimental measurements of the amount of a gas adsorbed as a function
of pressure and temperature may conveniently be plotted in the form of
adsorption isotherms.  One such isotherm is the so-called BET isotherm
which is written as follows:

                    N             k P/Po
                    Nffl   (1 - P/P0) (1 - (l-k)p/po)

where
          N = loading, mmol/g
         N  = monolayer loading
          k = a constant
          p = equilibrium pressure
         p  = vapor pressure of the bulk liquid.

     If only physical adsorption on nonporous surfaces is involved, then
desorption will follow the same isotherm curve upon decreasing of the
pressure.
     Physical adsorption and desorption per se is instantaneous; however,
with adsorbents containing many fine pores or capillaries, the effective
rate is governed by one or more of several diffusional steps.  Broadly
speaking, these may be categorized into the fluid phase and the solid
phase mass transport.  The fluid phase mass transport is governed by the
molecular (and/or ionic) diffusivity and also eddy diffusivity in turbulent
flow.  The solid phase mass transport is governed by molecular diffusion,
                                   19

-------
Knudsen diffusion, and surface diffusion.  Thus, when capillary condensation
occurs, the adsorption isotherm exhibits hysteresis on desorption.
     For desorption five basic techniques exist (Ref. 1):
          1)  thermal swing desorption;
          2)  pressure swing desorption;
          3)  purge gas stripping;
          4)  displacement desorption with a fluid less
              strongly adsorbed than the material being
              adsorbed;
          5)  displacement desorption with a fluid that is
              more strongly adsorbed than the material being
              desorbed.
     Often a combination of some of the above basic desorption techniques
is also used.  In the thermal swing desorption technique,  one raises the
temperature of the adsorbent, and the desorption rate is governed by the
rate of heat transfer to the adsorbent.  Knowing the final desorption
temperature and adsorbate vapor pressure permits one to predict the final
adsorbate loading from equilibrium data.  Pressure swing desorption is
accomplished by conducting desorption under reduced  total system pressure.
It has the advantages of simplicity and adaptability to rapid cycling.
Purge gas stripping technique operates by sweeping the desorbate away from
the adsorbent, thus maintaining the partial pressure of the desorbate
vapor below its equilibrium value.  The amount of desorption for a given
amount of purge gas depends primarily upon the equilibrium vapor pressure
at the temperature and total pressure of operation.  Higher operating
temperature and lower total operating pressure will  increase the overall
desorption rate.  Desorption with a purge that is less strongly adsorbed
than the adsorbate is accomplished by both displacement and purge.
Desorption with a purge that is more strongly adsorbed than the material
being removed results in the formation of a steady-state desorption front.
Behind the front the adsorbent is loaded with the adsorbable purge, and
ahead of the front the desorbate, uncontaminated with  the displacement
purge, is available.
                                   20

-------
     Of the different adsorbents that have been used In the chromatographic
analysis of sulfur dioxide, we suggest the use of 40-60 mesh Teflon coated
with a polyphenyl ether containing phosphoric acid (Ref. 2) for the
adsorption-desorption function intended here.  This adsorbent is suggested
because it has been used in the chromatography of sulfur dioxide at 50°C;
it apparently involves no chemisorption since it has been used in the analysis
of as low as a few ppb amounts; and the materials are commercially
available.

B.  Physical Characteristics
     Tha calibration output requirements for SC>2 have specified as
1 ppm as the upper limit with the field usable test atmosphere generating
device delivering the doped air at 3 &/min (upper rate) for periods up to
                                                                          o
2 hr.  Thus, the maximum volume (360 £) of doped air would require 0.36 cmj
(0.016 mmol) of adsorbed SCL.  The adsorption capacity data and sorption
kinetic data for the proposed system are not available.  Batchwise delivery
of a known amount of SO- would be possible as all of the dopant can be
desorbed at 50°C without apparent loss.  Feasibility of a continuous
delivery at the required rates would have to be established first.  The
chromatographic work indicates that the effects of temperature cycling upon
the adsorbent efficiency are minimal.
     At the instrument site, a supply or source for zero (clean) air is
required.  This may be either a compressed tank of clean air or an air
scrubbing system.   Alternatively,  zero air source can be a part of the
portable test atmosphere generating unit proposed.
     It is proposed that stainless steel be used to fabricate the tube as
this material has  been shown not to adsorb sulfur dioxide (Ref. 3).  For
the purposes of setting upper portable unit dimensions, we assume that 5 g
of the adsorbent is sufficient for storing the dopant.  This amount could
be packed in a tube 1 cm in diameter x 7 cm long.  The entire unit including
compressed zero air lecture bottle, dopant source tube, and storage bag
if any (up to 360 Si capacity) would have the following physical characteristics:
                                   21

-------
          Total weight:  3.6 kg

                         air lecture bottle - 2.3 kg
                         dopant tube  - - - - 0.2 kg
                         valve + tubing - 	 0.2 kg
                         storage bag  - - - - 0.9 kg

 Total unit dimensions:  45.7 x 30.5 x 7.6 cm

                         air lecture bottle - 38.1 x 5.1 cm diameter
                         dopant tube  ----7.6x0.6 cm diameter
                         storage bag	12.9 x 12.7 x 2.5 cm
                                                   when collapsed;
                                              inflated -88 cm diameter
                                                sphere for 360 a volume.


C.   Operational Requirements

    Any one of the five basic desorption techniques or a combination of them
can be used to deliver sulfur dioxide for dilution.  Thus, depending upon
the choice of the desorption technique, a source of desorbing material, heat
or vacuum would be needed.   In case a conducting adsorbent were used, then
only a source of electric power would be needed for direct coupling with

the adsorbent.
     If the continuous delivery of SO- were found  to be feasible, then the

operation of the unit would consist of plugging in the heater of the
adsorption tube set to produce a predetermined temperature (50°C for

example) inside the tube (probably less than 10 min to reach it) and adjust-
ing of the precision valve on the compressed air bottle and/or the S02

adsorption tube for the desired rate of doped air  delivery.  This could be

directed directly into the instrument manifold or  in a storage bag.


D.  Maintenance Requirements

    No special maintenance is required.
                                   22

-------
E.  Shipping Requirements
    The tubes would be expected to be shock resistant  and no special
shipping containers would be needed for the monolayer  adsbrbed tubes.   For
the highly loaded tubes a gastight, low-pressure resistant shipping
container would be needed.
F.  Estimated Accuracy
    The accuracy depends on the ability to measure a quantity of sulfur
dioxide or on the precision of a metering valve.  Differential screw-driven
diaphram valves with conductances that can be adjusted to less than
10 ~^ &/s are available.  This rate adjustment capability is some
six orders of magnitude below that required for the minimum SO. doping
              —8
rate (1.7 x 10   £/s).  Thus, we believe an accuracy of + 10 percent  should
be achievable.

G.  Estimated Cost
    1)  Development - $50,000
    2)  Production - The material costs would be minor and commercial
                     sources of SQ^ are available.  However, the initial
                     costs at each installation could be several
                     thousand dollars.

H.  Advantages/Disadvantages
    This type of test atmosphere benerator could be simple to operate and
have a minimum of support requirements. There are, however, a number of
parameters that require investigation before the merits of this method
can be evaluated.
 I.   Design Sketch
    carrier gas
                           heater
adsorbent +
    SO,,
                                                          ® ) instrument
Tedlar
  bag
               Figure 3-1.   S02 test atmosphere generator.
                                   23

-------
J.  Recommendations
    Determine experimentally the feasibility of continuous doping prior
to development efforts.

K.  References

1.  Lee, M. N. Y. Jin Recent Developments in Separation Science. Vol 1,
    Cleveland, Ohio, Chemical Rubber Co. Press, 1972, p. 75.
2.  Stevens, R. K., and A. E. b'Keeffe. Anal. Chem. 42zl43A (1970).
3.  Wohler, H. C., H.  Newstein, and D. Dannls. J. Air Pollution Control Assn.
    17_:753 (1967),

3.2.1.2  Nitrogen Dioxide
A.  Principles Involved
    All of the general and specific considerations advanced for SO. are
equally applicable to NO,.  In addition, because of its great reactivity
and tendency to undergo compositional changes, a most challenging problem
in its handling is presented.  The following equilibria, among others,
are pertinent:
             3N20^ + 2H20 (moisture)        v?Nn + 4HNO
                        N02 + NO
Irreversible adsorption of N02 and interaction with water are two problems
that have been solved in certain systems (Ref. 1).  As a general rule,
systems handling NO, require prior conditioning with the gas.
     Of the different adsorbents that have been evaluated in the chroma-
tographic analysis of NO- we suggest the use of Fluoropak-80 coated with
10% SF96.  This adsorbent is suggested because it has been used in the
chromatographic analysis (Refs. 1, 2) of a few ppm of N02 at room tempera-
ture and the materials are available commercially.

-------
B.  Physical Characteristics
    The calibration output requirements for NO, are approximately those of
                            3
S02, thus  requiring 0.36 cm  (0.016 mmol) of adsorbed N02< Prior
conditioning of the adsorbent with N0» is a requirement and neither the
adsorption capacity nor sorption kinetic data are available for the system.
The same delivery considerations as for S0_ apply also for NO-.
     Stainless steel is proposed as the material to be used in the fabri-
cation of the tube as it has been found to have adequate resistance to
NO2 (Ref. 3).  Because of practically the same volumes of adsorbed gases
required, the same weight and dimensional considerations as for SO  also
apply to the N0_ unit.

C.  Operational Requirements
    Same as those for sulfur dioxide except that most of the nitrogen
dioxide chromatographic work has been done at room temperature.

D.  Maintenance Requirements
    Same as those for sulfur dioxide.
E.  Shipping Requirements
    Same as those for sulfur dioxide.

F.  Estimated Accuracy
    Same as that for sulfur dioxide.

G.  Estimated Cost
    1)  Development - $30,000.
    2)  Production - Since the material costs would be mino.r and
                     nitrogen dioxide is available, the unit cost
                     should be small for adsorption sources.  The
                     initial costs for each using installation would
                     be several thousand dollars.
                                   25

-------
H.  Advantages/Disadvantages


    Same as those for sulfur dioxide.



I.  Design Sketch


    Same as that for SO.*



J.  Recommendations


    Determine experimentally the feasibility of continuous doping prior to

development effort.



K.  References



1.  Lawson, A., and H. G. McAdie. J. of Chromat. Sci. 8;731 (1970).


2.  Morrison, M. E., R. G. Rinker, and W. H. Corcoran. Anal. Chem. _36_:2256  (1964).


3.  Wright, C.M., and A. A. Orr. Anal. Chem. 40:29 (1968).



3.2.1.3  Carbon Monoxide

A.  Principles Involved


    The same considerations advanced for SO  are also applicable to CO.

Chromatographic studies on carbon monoxide and an extensive bibliography on

its analytical methods in general are available (Refs. 1-7).  Of the

different adsorbents that have been studied in the chromatographic analysis

of CO, we suggest the use of 30-60 mesh Molecular Sieve 13X (Ref. 3).  We

are suggesting this adsorbent since it was successfully used in the gas

chromatography of CO at 258C and the material is available commercially.



B.  Physical Characteristics


    The maximum calibration output requirements for CO can be 50 times those
                               3
of SO  , thus  requiring 18.0 cm  (0.8 mmol) of absorbed CO.  A conservative
                                                  2
molecular sieve monolayer equivalent area is 500 m /g.  The area of a
                                                              «2
CO molecule in the formation of a monomolecular layer is  16.8 A   (Ref. 2).
                                   26

-------
                                          20
The 0.8 mmol of CO corresponds to A.8 x 10   molecules which would be
                                                      2
expected to form a nomomolecular layer of about 80.6 m  area.  Thus, a
gram of the molecular sieve would provide a surface area in excess of that
required by 0.8 mmol of CO to form a monomolecular layer.  The sorption
capacity and kinetic data are not available for this system.
     Stainless steel is proposed as the material to be used  in the
fabrication of the tube.  The bulk density of Molecular  Sieve 13X is 0.61.
Thus, a tube 1 cm in diameter x 3 cm long would have a sufficient volume
to hold the absorbent.  The weight and size of the entire unit would still
be the same as that of the SO™ unit.

C.  Operational Requirements
    Same as those for sulfur dioxide except that the unit can be operated
at room temperature.

D.  Maintenance Requirements
    Same as those for sulfur  dioxide.

E.  Shipping Requirements
    Same as those for sulfur  dioxide.

F.  Estimated Accuracy
    Same as that  for  sulfur dioxide.

G.  Estimated Cost
    1)  Development - $35,000.
    2)  Production -  Unit  cost  of $25  and set-up cost of
                      approximately $2,000.

H.   Advantages/Pis advantages
    Same  as those for sulfur  dioxide and nitrogen dioxide.
                                    27

-------
 I.  Design  Sketch
    Same as  that for  sulfur dioxide.

 J.  Recommendations
    Determine experimentally the possibility of continuous doping prior
 to development effort.

 K.  References
 1.  Kyriacos, G., and C. E. Boord. Anal. Chem _29_:787 (1957).
 2.  Greene, S. A. Anal. Chem.  _M:480 (1958).
 3.  Trowell, J. M.  Anal. Chem. 37:1152 (1965).
 4.  Wilhite, W. F., and 0. L.  Hollis. J. Gas Chromat. 6:84 (1968).
5.  Altshuller, A.  P., S. L.  Kopszynski, W. A.  Lonneman, T. L. Becker, and
    R. Slater. Environ. Sci.  Technol. 3.:899 (1967).
6.  Dubois,  L., A.  Zdrojewski, and J. L. Monkman.  J. Air Pollution Control
    Assoc.  L6:135 (1966).
7.  Cooper,  A. G.,  Carbon Monoxide.   A Bibliography with Abstracts. Washington  nC
    HEW,  PHS, 1966.                                                            » D<

 3.2.1.4  Ozone
 A.  Principles  Involved
                                                     *
    Ozone has  been  separated  from oxygen by adsorption on refrigerated
 silica gel,  followed  by desorption,  either in  pure form at reduced pressure,
 or diluted  by  air,  nitrogen,  argon,  or other gas  not strongly adsorbed on
 silica gel.   This  is  a practical  method free from"hazards when correctly
 performed.   A pilot plant has  been  operated successfully to  demonstrate the
 removal of  ozone from oxygen  and  its transfer  to  other gases, using the
 silica gel  adsorption/desorption  technique (Ref.  1).
      When the  quantities of ozone adsorbed do  not have to be large,
 adsorption  can  be  conducted at room temperature.   The  adsorption isotherms
 at 0° and 25°C  for  ozone in oxygen  on  a silica gel are shown in Figure 3.2
 (the  partial pressure of ozone is 10.3 torr when  the ozone-oxygen mixture
 is at atmospheric pressure.
                                   28

-------
                            -I   •  f
                         3.C
                         3.4
                         3.2
                         3.0
                          =
                        S 9.0
                        f,
                          1.4

                          1.2
                          1.0
                          0.4
                         '•0.1
                                                   I     I
                                W   20   30    40   SO
                                OZONt MRTIAL PftCSSUAC,MM Of HO
        Figure 3-2:  Adsorption isotherms at 0° and 25°C
               for ozone in oxygen on Davison silica  gel.
     Both small- and large-scale safety tests have been made  on silica gel
loaded with adsorbed ozone.  With up  to 9  g of  ozone  per  100  g of silica
gel, no explosion could be initiated  at -78°C by an  ignited iron wire
imbedded in the gel.
     Stability of small amounts of silica  adsorbed ozone  (monomolecular
layer) does not appear to have been investigated.  The stability of large
amounts of adsorbed ozone (4.5 k-g and more per  45.5 kg of silica gel) is
determined by the temperature at which it  is kept since the decomposition
appears to be due only to that of the gas  phase ozone. Thus, at -78°C
about 9% of ozone had decomposed in 2 weeks while at  liquid nitrogen
temperature no decomposition was expected.  However,  ozone adsorbed only to
the extent of a monomolecular layer converage would  have  a very low vapor
                                   29

-------
pressure and consequently improved stability.   Furthermore,  no studies
appear to have been reported in the literature on other adsorbents  for
ozone and, thus, this area needs exploration.
B.  Physical Characteristics
    The maximum calibration output requirements for ozone are the same as
those for SO , thus requiring 0.36 cm  (0.016 mmol) of adsorbed ozone.
Using the adsorption isotherms given in Ref.  1, one can calculate that 1 g
of the commercial grade silica gel Grade 40 will adsorb 0.008 mmol
of ozone at 25°C.  Thus, 2 g of such a silica gel would be sufficient
to adsorb the required 0.016 mmol of ozone.  Although no mention  was
made whether this was a monolayer adsorbed 0_ or not, it is possible
that it was more than a monolayer adsorbed 0-.  On the basis of general
adsorption characteristics of silica gel considerations, we estimate that
probably 2 to 3 times of the above calculated amount of silica gel would
be required for the monolayer adsorption.  No sorption kinetic data are
available on monolayer adsorbed ozone; thus,  the feasibility of continuous
delivery of ozone still needs to be established.
     A glass tube of 1 cm diameter x 7 cm long would be sufficient to pack
the required amount of silica gel as indicated above.  The weight and size
of the entire unit would be the same as that  of the S02 unit.

C.  Operational Requirements
    If a continuous delivery of 0, were found to be feasible, its desorption
would be accomplished by a purge gas (nitrogen, for example) stripping at
room temperature.  The operation would simply consist of a precision valve
adjustment on the compressed nitrogen cylinder.

D.  Maintenance Requirements
    Maintaining the tubes at room temperature or preferably below.

E.  Shipping Requirements -
    No special shipping requirements besides  adequate packing to guard
against breakage.
                                   30

-------
F.  Estimated Accuracy
    Needs to be established; in principle, as good as the technique for
adsorbing ozone on the gel.

G.  Estimated Cost
    1)  Development - $45,000.
    2)  Production - Source cost of $40 and setup costs of $2,000.

H.  Advantages/Disadvantages
    Generation of stable and reproducible concentrations of ozone may be
possible requiring no skilled personnel in the operation.

I.  Design Sketch
    Same as that for SO. unit.

J.  Re commendations
    Undertake monomolecular layer adsorbed ozone stability studies.

K.  References
1.  Cook, G. A., A. D. Kiffer, C. V. Klumpp, A. H. Malik, and L.  A. Spence.
    Advances in Chemistry  Series, No. 21. Washington, D.C. , American
    Chemical Society, 1959, p.44.

3.2.1.5  Butene-1
A.  Principles Involved
    Probably more different adsorbents have been used for hydrocarbon  gas
chromatographic analysis ttfan for any other  class of materials.   Among the
many  references in this field, a number deal directly with the C, hydro-
carbon analysis:
    1)  silica gel (Refs.  1-3)
                                   31

-------
     2)   modified  silica  gel  and  alumina with  3,  3'-oxydipropionitrile
         with  20 M carbowax  (Ref.  5) with  dibenzyl ether  (Ref.  6)
     3)   partially graphitized  carbon  coated with glycerol  (Ref. 7)
     4)   dimethyl  sulfoxide on  60- to  80-mesh  Gas Chrom RZ  (Ref. 8)
     5)   activated coconut shell  carbon (Type  JD-1)  (Ref. 9)
     By  selection of appropriately modified silica  gel or  alumina, for
example, it is claimed that  separation of various butane isomers is
possible.  We propose the use  of  a 20 M carbowax modified  activated alumina
adsorbent  (Ref. 5).  This adsorbent has been  used successfully at ambient
temperatures  in the analysis of C, hydrocarbons  of  concentrations as low
as 0.004 ppm  and  the materials are commercially  available.

B.   Physical  Characteristics
     The maximum calibration output requirement for  butene-1 is the same as
that for N02, thus requiring 1.80  cm  (0.08 mmol) of adsorbed butene-1.
Neither  the sorption capacity nor  kinetic data are  available for this system.
     Stainless steel is proposed as the material for the fabrication of the
tube.  The same tube and entire unit  dimensional considerations as for the
SO- unit apply.

C.  Operational Requirements
    Same as those for sulfur dioxide.

D.  Maintenance Requirements
    Same as those for sulfur dioxide.

E.  Shipping Requirements
    Same as those for sulfur dioxide.

F.  Estimated Accuracy
    Same as those for sulfur dioxide.
                                  32

-------
G.  Estimated Cost

    Same as for sulfur dioxide.


H.  Advantages/Pisadvantages

    Same as for sulfur dioxide, nitrogen dioxide, and carbon monoxide.


I.  Design Sketch

    Same as that for SO .


J.  Recommendation

    Determine experimentally the feasibility of continuous doping

prior to development effort.


K.  References

1.  Bellar, T. A., and J. E. Sigsby. 151st American Chemical Society  National
    Meeting, Pittsburgh, Pa., March 22-31, 1966.

2.  Bruk, A. I., D. A. Vjakhirev, A. V. Kiselev, Y. S. Nikitin,  and N.  M.
    Olefirenko, Neftekhimiya 7.:145  (1967).

3.  Vyakhirev, D. A., M. V. Zuyeva, and A. I.  Bruk. Trudy  Khimii Khimicheskoi
    Teknologii Corky 2_:268  (1964) .

4.  Jeung, E., and H. L. Helwig, Symposium on  Air Pollution at the 144th
    National Meeting, American  Chemical Society, Los  Angeles,  Calif., April 1963,

5.  Kuley, C. J.  Anal. Chem.  35_:1472  (1963).

6.  Bellar, T. A., and J. E. Sigsby, Jr. Environ. Sci. Technology  1^:242 (1967).

7.  DiCorcia, A., D. Fritz, and F.  Bruner. Anal. Chem. 42_:1500 (1970).

8.  Trowell, J. M. Anal. Chem.  3]_:ll52  (1965).

9.  Turk, A., J. I. Morow,  S.  H. Stoldt, and W.  Baecht.  J. Air Pollution
    Control Assoc. 16:383  (1966).
                                   33

-------
3.2.2  Effusion Methods
       Effusion is descriptive of the process wherein a gas escapes from a
container through a small opening.  For the generation of test atmospheres,
the container must contain the dopant gas (NO^s SCKj 0,, CO, or C,Hfi)
either in pure form or in a mixture for dilution with air to obtain a test
atmosphere.  The factors that will affect the ease and precision in
obtaining a test atmosphere include the following:
     1)  The precision with which a specified mixture can be
         obtained in a container.
     2)  The chemical and physical stability of the gas mixture.
     3)  The precision with which small discharge holes may be
         obtained and reproduced.
     4)  The range of operating temperatures.
     5)  Calibration of the doping rates for different gases and
         conditions, which will vary with viscosity of the doping
         mixture.
     6)  The control of airflow in the test manifold.
     7)  Dead space in the dispensing arrangement.
     8)  Pressure variations in the storage container.
     Considering storage, dispensing, and mixing as the three steps in
obtaining test atmospheres by effusion,  storage is discussed first.  In its •
simplest form, storage consists of a container filled with the gas or with
a noninteractive mixture containing the gas.  Examples include low pressure
cans and high pressure cylinders or lecture bottles.  More complex storage
systems may involve condensed storage either as an adsorbate or as a liquid
as well as storage in other chemical species to be released as needed by
reaction.  When considering test atmosphere generation by effusion, the

-------
maintenance of a storage pressure in excess of atmospheric pressure is
required.  Each of the above storage mechanisms is compatible with this
requirement.  Since the pressure in the storage container affects the
dispensing rate, it is necessary to take such pressure changes into
account, to maintain a constant pressure, or to provide sufficient storage
so that pressure changes little with use.
     The temperature effect on storage pressure must be taken into account.
Although, for a compressed gas at constant volume, pressure is linearly
dependent on absolute temperature, the actual number of molecules per unit
volume remains constant.  Thus, if the dispensing rate is also a linear
function of pressure, the rate at which the dopant gas is supplied to a
manifold will remain constant, if measured in terms of number of molecules
or mass per unit time.
     Stability of gas mixtures in a storage container is important in
determining accuracy of the test atmosphere generator.  The five gases—
sulfur dioxide, carbon monoxide, nitrogen dioxide, ozone, and 1-butene—are
available commercially in cylinders, the first three in pure form or  in gas
mixtures while ozone is provided in a cold condensed gas mixture and
butene-1 is a liquid at the storage pressure.  This and other information
indicate that stability in storage is possible but must be examined
carefully for each gas in order to determine  applicable conditions and
duration.
     Leaving the dispensing step until  last,  because it involves diffusion,
the mixing  of gases  in a manifold or a  holding volume  is  the  third step in
obtaining test atmospheres.   Considering only the dynamic process  involving
a manifold  as shown  in Figure 3~3(A),  this  consists  of a  flow of clean air
through  a tube  that  provides  taps  for  a monitoring  instrument for  insertion
of a dopant gas to create a test  atmosphere.   Assuming that  the effusion
rate from the dopant source is known,  then  the accuracy achieved in  the
dopant  content  of  the test  atmosphere  depends on control  of  the manifold
flow rate.  Manifold flow is  maintained with an  air pump  or  blower on the
discharge end of  the manifold or  with  a compressed  air supply at the input
                                    35

-------
           FLOWMETER
          , MIXING
ORIFICE
         «

         I
         co
AIR
IN




L_



f '
,



w






/ /
\
\









AIR
PUMP




EXH/


             DISPENSING
H
CO
       MEASURING
       INSTRUMENT
      BEING TESTED
        FLOW RATE
        1-3 £/min
                                 (A)
   COMPRESSED
   ZERO
   AIR
   SUPPLY
                                PRECISION  PRESSURE REGULATOR
                                 CALIBRATED FLOW CONTROLLER
                                         -."_!	* EXHAUST


                                         MEASURING
                                          INSTRUMENT
                                         BEING TESTED
    DOPANT GAS SOURCE
    WITH EFFUSION CONTROLLED
    DISCHARGE
                                (B)
              Figure 3-3. Test atmosphere generator.
                                36

-------
end  (see Figure 3-3(B).  Flow can be controlled by means of a critical orifice
that will maintain constant volume flow with variable pressure at the low
pressure side of the orifice.
     The interface between the mixing manifold and storage volume is the
dispensing rate control device—in this discussion only effusion methods
are being considered.  Since all effusion devices depend on flow from high
pressure storage to a low pressure manifold through a small channel that
restricts the flow to some predetermined rate, the variety of effusion
sources are limited although any one basic type may be implemented in a
variety of ways.  Representative implementations include:
     1)  A fixed orifice or capillary flow controller to obtain
         a continuous constant volumetric flow rate from
         essentially constant pressure sources, disposable
         storage, and dispenser.
     2)  Similar to 1) above, but with on-off capability to
         provide longer useful life.
     3)  Similar to 2) above, but with adjustable effusion rate.
     4)  Compressed lecture bottle sources with conventional
         regulations and flowmeters operative for any gas or
         dopant level.
     Each of these techniques is applicable to various of the dopant
pollutant gases.  However, they are discussed in the following subsections
in terms of a particular pollutant.

3.2.2.1  Sulfur Dioxide
A.  Principles Involved
    The method described in this section provides for a  low pressure
pollutant gas source, maintained at almost constant pressure, and with  a
continuous constant effusion rate, referred to as CLOPES  (Constant  Low
Pressure Effusion Source).
     This method requires the pollutant gas to be supplied in low pressure
cans,  similar to aerosol consumer products, in which the pressure is
                                   37

-------
maintained by inserting in the can a collapsed plastic bag containing
condensed Freon.  The supply pressure is thus maintained at the Freon vapor
pressure level until the contents of the can are exhausted.  The plastic
bag prevents dilution of the pollutant gas with Freon.
     Dispensing is controlled by a plastic capillary tube on the top of
the can.  In use the source is activated by breaking off the tip of the
plastic capillary allowing gas to effuse from the capillary at a continuous
constant rate until exhausted.

B.  Physical Characteristics
            3
    A 500 on  can containing a pure pollutant gas will supply  sufficient
pollutant to provide a 50 ppm test atmosphere at a 3 £/min rate for
approximately 5000 hours.  This number will be reduced proportionally if
dilute gas mixtures are employed to provide a large leak rate.
     Approximate calculations using the  Poiseuille equation for flow
indicate that for capillaries of 0.1 mm inside radius, the dopant gas must
be in a dilute mixture on the order of 0.1 to 1 percent pollutant to
obtain the required test atmospheres.  This will reduce the useful life of
the source to between 5 and 50 hours which is in excess of that required.
     The weight of the "canned gas" should be on the order of 0.5kg, and
its dimensions would be those of an aerosol paint can.  It requires the
use of a mixing manifold similar to those shown in Figure 3-2.

C.  Operational Requirements
    A facility is required for fabricating and filling the cans with gas.
User requirements are minimal.

D.  Maintenance Requirements
    The only required maintenance is  that associated with the mixing
manifolds.
                                   38

-------
E.  Shipping Requirements
    Mailable can.

F.  Estimated Accuracy
    Dependent on pressure variations of Freon with, variations in ambient
temperature, calibration accuracy of capillary, mixing manifold flow rate,
and supply gas mixture.  Ten percent should be achievable.

G.  Costs
    Use cost is highly dependent on volume requirements.  If thousands of
cans are required, cost per can should be under $5.  Development costs are
estimated to be approximately $50,000.

H. Advant ages /Pis advant ages
    Once operational, this method provides a minimum cost, minimum effort
capability for a large number of users and is adaptable to various gases.
    The prime disadvantage is the startuu costs.  There is an  additional
problem in determining when the source is exhausted.

I.  Design Sketch
    See Figure 2-4.

j.  Recommendat ions
    This should  be given  serious  consideration  for  providing test
atmospheres.
                                    39

-------
              BREAKOFF
              NOTCH
PROTECTIVE
CAP
                                                      POLLUTANT
                                                      GAS
                                                      MIXTURE
                                                     FREON
        Figure 3-4.  Constant  low pressure  effusion  source.

3.2.2.2  Nitrogen Dioxide
A.  Principles Involved
    In order to increase the useful life of the pollutant source described
in Section 3.2.2.1, it may be provided with an on-off capability.
     Any on-off valve used in a low flow rate line must not introduce dead
                                  3
space.  At a flow rate of 0.003 cm /min such as required for 1 ppm doping
of a 3 ^/min gas stream, a dead space of 1 cm  would require over 16 hours
to flush.  At these flow rates, it is also difficult to discern when there
is flow; i.e., a valve that is stuck closed would be almost undetectable.
                                  40

-------
B.  Physical Characteristics
    Two methods are suggested, one of which is to plug the capillary with
a needle and the other is to energize a heater to allow gas to form.
In the first case, the capillary discharge tube would be capped with a
solenoid-operated needle which, when unenergized, plugs the end of
the capillary.  With current in the solenoid, the needle is withdrawn and
gas flows.  A sketch of this is shown in Section I (Figure 3-5(A)..
     In the second example, a metal capillary discharge tube is employed
but it is plugged with a rod of a lower thermal expansion material at the
discharge end.  Heat applied through a small coil would be sufficient to
expand the capillary relative to the plug allowing gas to flow in the space
thus provided as shown in Section I  (Figure 3-5(B).
     In neither instance will more than a few ounces be added to the weight,
and a negligible dimensional change will be entailed.

C.  Operational Requirements
    Same as in Section 2.1 except for small amount of electrical power.

D.  Maintenance Requirement
    No additional maintenance required.

E.  ghipping Requirements
    Same as for the CLOPES.

F.  Estimated Accuracy
    Same as for the CLOPES source.

G.  Costs
    Additional $1 per can  estimated  over  the  CLOPES  source in volume usage.
Little additional development  costs.

-------
H.  Advantages/Disadvantages

    Additional useful life due to on-off capability.  Otherwise same as for
the CLOPES source.
I.  Design Sketch
      MAGNETIC
      SLUG
        NEEDLE
                                                         COIL
                                                         PORTS  FOR
                                                         GAS  FLOW
                                                         TO  STORAGE
                                                         CAN
       (A) Solenoid-operated needle valve (closed position)
      HEATING
      COIL
                    (B) Thermally operated valve
            Figure 3-5.  On-off valves for CLOPES source.
                                                           SLUG OF
                                                           LOW THERMAL
                                                           EXPANSION
                                                            METAL WITH
                                                            HIGH THERMAL
                                                            EXPANSION
                                  42

-------
J.  Recommendations
    On the basis of present data, the on-off feature does not appear to
be required.  However, the simple plug cut-off will increase the useful
life by a large factor and may be cost-effective.

3.2.2.3  Carbon Monoxide
A.  Principle Involved
    This method provides for a storage arrangement similar to the
CLOPES source except that the Freon is omitted from the storage volume, and
an adjustable metering valve is used at the output.
     A can similar to that in Figure 3-4 is filled to a specified pressure
with the appropriate pollutant mixture but without Freon so as to illustrate
the case with a variable storage pressure.  Various precision valves are
available.  Valves with sapphire seats made for vacuum applications can be
controlled to as low as 10    &/min with one atmosphere across them while
relatively inexpensive needle valves have maximum flow rates of 0.5 £/min
                              _3
but are controllable to 5 x 10   8,/min or less.
     Since it is obvious that almost any flow rate one desires can be
obtained but at a rapidly increasing cost for lower flow rates, the high
flow rate valve will be considered.  At 5 x 10~  A/min, it  is necessary to
have a pollutant concentration of about 1000 ppm in the storage volume to
obtain 5 ppm in a gas stream flow at 1 Jl/min.  At  this rate 0.38. at a
pressure of 1 atmosphere will be used each hour.   If a 1-Jl  storage volume
is at a pressure of 3 atmospheres initially, after 1 hour  the storage
pressure will be reduced to 2.7  atmospheres with a consequent reduction in
                     *• ^
flow rate to A.5 x 10   Jl/min.   If no correction  is provided, this  introduces
an error of 10 percent in  the pollutant concentration  in  the test atmosphere.
If, however, a controlled  flow rate  of 5  x  10~  Jl/min  is  obtainable, then
the error becomes 1 percent.  The viability of this method  thus depends
on the availability of a low flow-rate valve  at an acceptable cost.
                                   43

-------
B.   Physical  Characteristics
     The  storage  unit  is a  lightweight can with a  1-&  capacity with a
valve  assembly mounted on  the  top.  The valve assembly has a screwdriver
locked metering  valve and  a toggle on-off valve  (see  Figure 3-6).
     The overall weight of the unit is 0.9 kg, its dimensions are about
7.6  cm o.d. by 20.3 cm high.  The cans are reusable.

C.   Operational  Requirements
     The canned gas is attached to a controlled source of clean air and its
output is attached directly to the instrument manifold.  The airflow is
started and the  toggle valve is opened 60 a  before a reading is taken in
order  to flush the valve assembly.

D.   Maintenance  Requirements
     Because of the cost of the valve, the assembly is returned to the
issuing organization  for refilling and adjustment of the metering valve.
Metering valves  are calibrated volumetrically by displacement of a liquid.

E.   Shipping Requirements
    Capability to ship a 0.9  kg can.   A clean air supply is required at the
test site.

F.  Estimated Accuracy
    Ten-percent  accuracy should be readily obtainable.

G.  Estimated Costs
    A realistic  cost for identifying and characterizing an optimum system
should be about  $50,000.  Once established, the actual costs of each
valve/container  assembly should be of the order of $100 and refilling
should be accomplished at no more than $10.
                                   44

-------
               LOCK
               SCREW
Ul
                                  \    /
                                                 MICROMETER
                                                 METERING
                                                 VALVE
                                                 ADJUSTMENT
            AIR
            IN
TEST
ATMOSPHERE
OUT
                         POLLUTANT
                          MIXTURE
                                        Figure 3-6. Storage unit with metering valve.

-------
H.  Advantages/Disadvantages
    The requirement for reuse of the test atmosphere generator is a
disadvantage over some other methods.  The requirement for a controlled
flow of clean air is common to most methods.  The low shipping weight and
relative ease of use is a definite advantage.

I.  Design Sketch
    See Figure 3-6.

J.  Recommendations
    This should be compared to throw-away containers in terms of overall
cost and effectiveness.

3.2.2.4  Ozone
         The primary problem for effusion sources of ozone is that of
stability of the ozone in a gas mixture.  While there are indications that
long-term stability of ozone can be obtained in dilute mixtures and
compressed gas mixtures containing ozone are available, these data are not
sufficient to serve as the basis of an effusion source.  One method that
might be feasible is to provide a small source of ultraviolet radiation in
a gas mixture containing oxygen.  This would provide some equilibrium
concentration of ozone in the mixture which would have to be determined.
At this time, no data exist on which to base such a method.
     In order to ascertain the possibility of a CLOPES source for ozone,
the equilibrium ozone concentration in a low pressure can which contains a
small ultraviolet radiation source should be determined for various oxygen-
containing gas mixtures, materials, and pressures.
                                   46

-------
3.2.2.5  1-Butene
A.  Principles Involved
    Butene-1 is a gas at atmospheric pressure, but in storage containers  it
                                                     2
is a liquid with a vapor pressure at 70°F of 24 Ib/in  g.   Thus, it is  ideal
for a CLOPES-type source since it will not require a Freon reservoir to maintain
a constant storage pressure.  In this present method, however, a lecture
                        3            3
bottle containing 2.7 ft  (28, 317 cm ) of 1-butene is used.
     A lecture bottle of 1-butene is provided with a regulator to give a
                            2
delivery pressure of 4 Ib/in  g.  This is coupled to the instrument manifold
through a fine metering valve (Nupro "5" series cross-pattern metering
                                             -3   3
valve, for example) to give a flow rate of 10   cm /min.  At these rates,
direct insertion in a manifold in which a flow rate of 1 SL/min is maintained
will give a concentration of 1 ppm.  Dead space in the plumbing must be at
a minimum.  The valve may be preset to give "blind" detection levels.

B.  Physical  Characteristics
    The weight of this apparatus approximates 4.5 kg and its dimensions are
of the order  of 5.1 by 50.8 cm.  It consists of a small gas cylinder with a
regulator and valve assembly connected directly to the instrument manifold
as shown in Figure 3-7.
C.  Operation Characteristics
    The metering valve would be preset to a specific flow rate for a given
inlet pressure.  Operation  consists of attaching the lecture bottle and
valve assembly  to  the manifold and adjusting  the regulator to the specified
inlet pressure  with  a specified flow  rate for air  in the manifold.
 D.   Maintenance  Requirements
     Only a minimum of maintenance  is  required.   Lecture bottles  of  gas
 are available commercially.
                                    47

-------
oo
                             CLEAN
                             AIR
                                                                TEST ATMOSPHERE
                                                                       METERING
                                                                       VALVE
                                                                                           INSTRUMENT
                                       Figure 3-7. Test generator with metering valve.

-------
E.  Shipping Requirements
    The shipping assembly will weigh about 5.5 kg and will require return.
The unit is reusable a large number of times.

F.  Estimated Accuracy
    The direct introduction of undiluted pollutant air into an airstream
requires accurate control of flow rates that are difficult to maintain.  It
is estimated that + 20% is the best accuracy that is possible in a
practical environment.

G.  Estimated Costs
    The valve and cylinder assemblies may cost $200 each but refilling and
regeneration, after 100 uses, is only about $10.  Development cost to prove
the system out should be less than $30,000.

H.  Advantages/Disadvantages
    This method is relatively easy to implement and components are
commercially available.  It may not be possible to obtain sufficient
accuracy but since components are common, personnel may be trained to use
them with a minimum effort.

I.  Design Sketch. See Figure 3-7.

J.  Recommendations
    Since this utilizes standard components and techniques, it may serve as
a baseline for comparison of other potential methods in the evaluation.

3.2.3  Thin Film Evaporation
       Thin films may be either gaseous, liquid, or solid.  A gaseous thin
film is formed by adsorption on a solid surface and is released by desorp-
tion.   This process for generation of test atmospheres is considered in
Section 3.2.1.  A solid thin film may be prepared for the pollutant gases
                                   49

-------
of interest but this requires cumbersome refrigeration techniques that are
impractical for field applications.  Liquid films may be used for test
atmosphere generation but the pure pollutants would also require either
refrigeration or high pressures to maintain them in a liquid state.
However, the pollutants may be dissolved in various solvents and thin liquid
films of the solutions employed as controlled pollutant sources for test
atmosphere generation.
     In chemical engineering, solvents are routinely employed in adsorption
or contact columns for removal of components from gas streams.  The reverse
process (i.e., injection of a gas into a gas stream by release from solu-
tion) is an equally viable process.  The solution containing the pollutant
gas must be maintained in a sealed container until use since it will release
the solute in order to attain equilibrium with the atmosphere above it.  In
use, the solution is exposed to a flowing airstream and a steady state is
achieved between solution flow, airflow, and rate of outgassing since there
is insufficient pollutant to achieve equilibrium with the airstream.  A
thin liquid film of solution is employed to achieve a rapid rate of
outgassing.  The concentration in the airstream will thus depend on:
     1)  Airflow rate
     2)  Concentration of solute in solution
     3)  Solution flow rate.
     A schematic representation of a test atmosphere generation system
employing thin films of pollutant gas solutions is shown in Figure 3-8.
     An inert collapsible bag containing the solution is maintained at a
known height to give a constant rate of flow of solution through a
capillary  to the column.  In the column, a thin liquid film flows over
the glass  beads and out  the bottom.  An airstream flows up through the
column and, after condensation of  any solvent vapors, constitutes the
test atmosphere.
                                   50

-------
     SOLVENT CAN
     BE CONDENSED
     IF NECESSARY
GLASS BEADS:
KNOWN NUMBER,
KNOWN DIAMETER


        AIR IN
                        AIR OUT
PRESSURE
HEAD ON
CAPILLARY
                                                      INERT
                                                    COLLAPS-
                                                       ABLE
                                                       BAG
DISSOLVED
POLLUTANT
SOLVENT
                                              -SUITABLE
                                               CAPILLARY
        COLUMN OF
        KNOWN VOLUME
                                               SOLVENT OUT
      Figure 3-8. Contact column test generator.
                           51

-------
     The quantity of liquid flowing through the capillary with a moderate
pressure head is determined by Poiseville's law (Ref. 1)

                              q = irpr /8&n

where
        q = flow rate, cm /s
                                         9
        p = differential pressure, dyn/cm
        r = capillary radius, cm
        H = capillary length, cm
                                             2
        n = liquid viscosity, poise (dyn-s/cm ).

The column length must be sufficient to essentially complete stripping of
the solute pollutant from the liquid film.  The amount of solute from the
solvent can be estimated from the relationship
where
                                                 ' f-
                   AC = C  - C = C  [1 - exp  ~~
                         o        o             V
        AC = concentration loss per unit volume of solution
             at any time, t
         C = concentration of solute in solution at any time
        C  = initial concentration of solution
         o           -42
         k - 2.5 x 10   mol min/cnr/atm specific rate of
             release (estimated for a system of this general size
             from standard j-factor versus Reynolds number correlations
             (Ref. 2)
         A = surface area of film of solution
        . » _ 	AP in atmospheres	
             AC in same units as C and C
                                        o
         V * volume of liquid, and
         t = time in minutes since solution was introduced into column.

The rate of pollutant introduction into the gas stream is thus given by AC
times the volumetric flow rate of solution through the column.  In actual
                                   52

-------
practice,  this is an upper limit to the rate of introduction but for a given
column  and set of conditions, the actual rate would be a reproducible
function of  this maximum rate.
      The use of thin film evaporation for the generation of test atmospheres
is  thus based on the release of dissolved pollutants from  thin films of
liquids obtained in a  column packed with glass beads.  Distribution of the
solution is  from a central laboratory sealed in plastic bags.  The column,
associated apparatus,  and air supply  (and possibly a cascade  air dilution
apparatus), would  be required at the test site.   Excess  test  atmospheres
must  be properly vented.

3.2.3.1 Sulfur Dioxide

A.  Principles Involved
    A solution of sulfur dioxide  in carbon  tetrachloride should  be suitable.

B.  Physical Characteristics
     Since  18.45 cm3 of S02 will dissolve in CCL^  at  25°C and  760 torr
 (Ref. 3),  a  more  than  adequate  amount of S0« will be available.   A dilute
solution  is  therefore  required.   This may be prepared either  by  first
putting CC1, in the  plastic  container,  then adding a gas mixture containing
S09,  or preparing  the  solution  first, then  introducing  it  into the bag and
sealing the  container.

C.  Operational Requirements
     Operation should be  simple  once the apparatus is assembled.   In
sequence:
      1)  Liquid  flow is  turned  on by a  toggle  valve, by opening
          a clamp,  or by  opening a regular  valve all the way.   Flow
          is  controlled by the capillary.
      2)  Airflow is adjusted to a specified rate.
      3)  When liquid begins to  accumulate  in the bottom of the
          contact  column,  the instrument to be tested can start
          sampling the test atmospheres.
                                    53

-------
     4)  Care should be taken to maintain the liquid at the preset
         level to insure constant solution flow.
     If the solvent is removed from the airstream by a cold trap, the
refrigerant and the trap should be placed in the Dewar flask before the
test begins.

D.  Maintenance Requirements
    The device should be cleaned and checked for correct flows before and
after field use.  Broken parts should be replaced.  The whole system
should be calibrated in the home laboratory prior to being used in the
field.  Solutions of pollutants will have to be prepared and analyzed in
the home laboratory.

E.  Shipping Requirements
    For the apparatus, the case should be rigid and the parts should be
strapped in.  When shipped, the interior should be stuffed with shock
absorbant filler of some kind.
     The solution of pollutant may be shipped sealed in a plastic bag
that is inert to the pollutant.  A second material may be sealed around
the basic bag container if its permeation characteristics are unfavorable.
The bag of test solution should be shipped in a rigid container.
     The apparatus,  as diagrammed in Figure  3-8, including a cold trap
and a Dewar flask,  can be contained in a case 61 x 46 x 20 cm in size.  The
weight of the apparatus less the case should be less than 2.3 kg.  A case
could be built to weigh about 6.8 kg total.

F.  Estimated Accuracy
    Delivery of an accurately known concentration of sulfur dioxide will
depend on measurements in the central laboratory which relate concentration
to the various parameters of the apparatus and to temperature.
     Precision of delivery will depend on control of temperature and the
accuracy of the hydrostatic head and the airflow setting.  A single point

-------
flowmeter can help the latter.  The hydrostatic head can be designated by
fixed marks on a stick but will require frequent adjustment.

G.  Estimated Costs
    The provision of plastic bags containing the solution that will serve
as the pollutant source is a significant cost advantage.  They should
require little effort to prepare and should cost less than $5 each.
     The apparatus required at the test site will cost approximately $300.
     Since this is an unproven method, approximately $35,000 in developmental
costs would be required before it is used.

H«  Advantages/Pisadvantages
    With proper quality control, there is no need to rely on outsiders for
delivery of a known concentration of sulfur dioxide to the instrument to
be tested.
     Accuracy and precision should be acceptable.
     The only disadvantages might arise from transportation and handling
of the solution.  If it is not kept enclosed, the dissolved pollutant will
escape.

I.  Design Sketch
    See Figure  3-8.

J.  Recommendations
    This method should be  developed only  if some advantage now not apparent
can be obtained by using  it.   It is,  however, a reasonable approach.

K.  References
1.  Weast, R. C., and S. M. Selby  (Eds.). Handbook of Chemistry and Physics.
    48th Ed. Cleveland, Ohio,  Chemical Rubber Co., 1968, p. F-33.
2.  Perry, J. H. Chemical Engineer's Handbook.  3rd Ed. N.Y., McGraw-Hill  Book
    Co., 1950, p. 547.
3.  Stephen, H., and T. Stephen  (Eds.). Solubilities of  Inorganic  and
    Organic Compounds. Vol. 1. N.Y., The MacMillan Co.,  1963,  p.960.
                                    55

-------
 3.2.3.2  Nitrogen Dioxide

         A device for generating a test atmosphere on NO- would be exactly
 as described  for SO  .  N0? will dissolve in a number of substances,
 including 1,3,5-trioxane  (Ref. 1).

     The various costs could also be essentially the same as for SO-.


     References
1.   Stephens, H., and T. Stephens  (Eds.). Solubilities of Inorganic and
     Organic Compounds. Vol. I. N.Y., MacMillan and Co., 1963, pp. 922-925.
3.2.3.3  Carbon Monoxide
         A test atmosphere of CO can be generated in virtually the same
manner as the test atmosphere for S0«.  Carbon monoxide is soluble in a
number of solvents including chloroxane and acetone (Ref. 1).
     References

1.   Stephens, H., and T. Stephens (Eds.).Solubilities of Inorganic and
     Organic Compounds. Vol. I. N.Y., MacMillan and Co., 1963, pp. 10*52-1053.


3.2.3.4  Ozone

         Ozone would have to be generated in the laboratory.  A concentration
of about 1 percent 0~ can be generated with commercial ozonizers.  In
                                 3                                3
contact with liquid CC1,, 0.03 cm  gaseous 0- will dissolve per cm  of CC1
                                3                                         ^
(Ref. 1).  At flow rates of 1 cm /min solution and lA/min air, 0. will be
                                                   3
delivered to the airstream at an estimated 0.002 cm /min.

     The cost of the 0, device should be increased by about $250 over that
for SO^ to cover the price of a suitable 0_ generator.


     References
1.   Stephens, H. and T. Stephens (Eds.). Solubilities of Inorganic and
     Organic Compounds. Vol. I. N.Y., MacMillan and Co., 1963, p. 576.
                                  56

-------
3.2.3.5  Butene-1
         Butene-1 could be dispensed in the same manner as  SO..   Butene-1
is soluble in a number of solvents to which the flame ionization detector
does not respond (e.g., CC1,).   For generation of test atmospheres  for
hydrocarbons, evaporation of  a  liquid hydrocarbon such as n-hexane  might
be better than the use of a solution of Butene-1.

3.2.4  Novel Permeation Methods
       Permeation tubes for the preparation of known concentration  of air
pollutants were originally suggested by O'Keefe and Ortman  (Ref. 1);
since then the performance characteristics of such tubes (Ref. 2) and
their critical evaluation for sulfur dioxide (Ref. 3) have  been reported.
     Permeation tubes offer a convenient and relatively simple means of
conducting dynamic calibration of gas detection devices.  However,  accurate
and precise temperature control is essential in producing exact sample
emission rates.  Depending upon the sample gas involved, the temperature
change, and the permeation material and wall thickness, steady-state
operation of a permeation tube may require equilibrium saturation times of
anywhere from 2 to over a 100 hours.  This places a severe limitation
on the utility of permeation tubes since it is impractical to wait for many
hours for the tube emission rate  to stabilize after only moderate temperature
changes.
     Although precalibrated permeation tubes are obtainable from commercial
sources, the following practical  drawbacks have been experienced with them:
(Ref. 2 and 4)
     1)  Tubes, which are sealed with balls of glass or stainless
         steel, may leak;
     2)  The permeating gas may be contaminated with solid particles
         and/or water;
     3)  Water may permeate into  the tube  during storage and shipping;
     4)  The permeation  rate for  NO   in  a  single-walled tube  is
         inconveniently high;
                                   57

-------
     5)  The thicker the tube wall, the longer the equilibrium
         saturation time;
     6)  If the tube wall thickness is not sufficient,
         dimensional tube changes occur which in turn affect
         permeation rate;
     7)  When liquid phase permeant is in contact with the
         area of permeation, the permeation rate is affected;  and
     8)  Accumulation of grease and moisture on tube exterior
         affect the permeation rate.
     We propose a new design for permeation tubes adapted from a recently
patented gas sensing device used in the medical field (Ref .  5) which should
obviate most if not all of the above mentioned drawbacks.  Of  all of the
problems and drawbacks mentioned, the lost time waiting for the tube emission
rate to stabilize is the most serious.  Although the equilibrium saturation
time is dependent upon a number of factors (including the type of gas and
the kind of tube material used), two factors are important for achievement
of fast equilibrium saturation regardless of the type of gas and tube material
 used.  These are the operating temperature and the wall thickness.   Thus
the thinner the wall, the faster the equilibrium saturation is reached.
Our proposal provides for these two factors .
     For the case of gaseous penetrants that are sparingly soluble in the
membrane substance, and that do not chemically associate with  one another
or the membrane material, the membrane molecular structure is  not perturbed
by the dissolved molecules.  Thus, diffusivity of a penetrant  is essentially
constant throughout the membrane.  Moreover, the solubility of such species
within the membrane is essentially directly proportional to its activity in
the equilibrium gas phase, and Henry's Law applies:
                    ci
                                                                      (2)
                                    58

-------
or
                                         ~ P^
                    J.  - k,D,
                     i
     Where c. is the concentration of (i) in the surface of the membrane,
k. is the solubility coefficient in the membrane, p  is the pressure at the
                                   22
surface, J. is the mass flux (gm/cm  s), D. is the diffusivity  (cm /s) ,
the subscripts (1) and (2) refer to the "upstream" and "downstream" surfaces
of the membrane, and £ is the membrane thickness.  The product  (k.D )  is
usually termed the "permeability," P., of the membrane to  (i) ;  it is equal
to the flux of (i) through a membrane of unit thickness when exposed to a
unit pressure-difference across it.  Most gaseous penetrants and membranes
obey equation (3); the permeability is thus independent of the  total pressure
of the penetrating gas.
     In the decision on membrane thickness, the following  considerations are
Involved.  When gases diffuse through a solid or liquid they take a definite
time, the magnitude of which is related  to  the diffusion  rate.   If one
measures and plots the rise in pressure in  a constant volume with time that
would result on one side of a membrane from a discontinuous change in
pressure on the other at time t =  o, and then draws an asymptote to the
steady state conditions, the asymptote will cut the time axis at the
coordinate T.  This time lag T is  related  to D and  the thickness of  the
membrane,  £, as follows:
                                 T "  6D  '

The  time  to  reach  a  steady  state condition is  3i ,  as shown in Figure 3-9
          — f\   9                 — ^
For  D - 10   cm /s ; I -  2.5  *  10  cm; T -  1 s.  The permeation can be
measured  in  torr-«,/s and  the  following relationship relates  all the
membrane  parameters:

                                     P   ' P
 where:   A is the membrane area, t is time and the other parameters as
         defined before.
                                    59

-------
                          THE TIME LAG T
                                               STEADY-
                                               STATE
                                               SLOPE
                                               GIVES p
                                                   -> t
FOR D - 10~6cm2/s~1;
                                      2.5  *  10~3cm; T  -  1 s
                 Figure 3-9. Time lag vs pressure curve.

     If the permeability constant is expressed in units of cm3 (STP)
      2           -1
cm (cm s cm of Hg)  , then it is found that the permeability constant falls
                  —6      —13
in the range of 10   to 10    for most pure polymeric membranes and
permeants.  A review of the available permeability data permits one to make
a further generalization insofar as the different polymers (used as membranes)
are concerned.  The different classes of polymers may be arranged in the
following order of decreasing permeabilities:  silicone rubbers > silicone
rubber copolymers > hydrocarbon rubbers > none las tome rs .  Thus, for
example, nitrogen and carbon dioxide permeabilities for polydimethyl
siloxane, a silicone rubber, and FEP (fluorinated ethylene/propylene
copolymer) compare as follows:
                  P x 1Q10 (cm3STP) (cm)/(cm2 s cmHg)
    FEP
N2
iiloxane 227
1.6
co2
355
13
                                                                   —6  2
     Typical values of D reported for silicone rubber are around 10  cm
               —Q   2
and for FEP, 10   cm /s, for hydrocarbon permeants.
     The permeation device that we are proposing is adapted from the
patent of Van der Grinten on a "gas sensing device" (Ref. 5) and is
depicted in Figure 3-10 below.
                                   60

-------
                       Figure 3-10. Permeation device.

     The unit consists of a stainless steel cylinder (28) capped with a
screw cap (31) on one end and a discrete blend of a solid polymer in the
performations of a perforated metal plate (32), heat pressed to a thin
film of a polymer (10), at the other end.  Such a membrane in the patented
                                        — A       ft
device permitted the maintenance of a 10~  to 10   torr vacuum in the system
over a range of temperatures.  Thus, apparently obtaining a pinhole-free
membrane was not a problem.  The entire unit was operated at a constant
temperature, maintained by means of a thermoregulated heater (27).  As this
device has not been used before in a permeation application that we require,
nothing definite can be said about the membrane cross sectional area
requirements at this time.
     While in all of the permeation tubes designed up to now, one is
limited to a certain tube wall thickness below which one cannot go for
structural or mechanical strength reasons, the polymer-perforated metal
membrane permits much thinner membranes and, as a result, drastically
reduces the equilibrium saturation time required for the membrane.  In
addition, none of the currently designed permeation tubes provide for
                                   61

-------
operation at isothermal conditions without using isothermal chambers.
The proposed device should not require such chambers.   Thirdly, current
permeation tubes must use nonelastomeric materials as the permeating
medium even though they are not ideal with respect to permeability and
practicality.  The polymer-perforated metal membrane device does not require
nonelastomers for the membrane, it can just as satisfactorily use an
elastomer as the membrane material.  In fact, one of the patented polymeric
impregnants was polydimethyl siloxane.
     On the basis of the extensive permeability background in our
laboratories and some specific recent work in the literature (Ref. 6)
polydimethyl siloxane appears to be the best material for use as the
membrane material in the perforations of the metal plate.  This silicone
rubber showed a minimum change in permeability with temperature change (5%
for a 10°C increase in temperature) and the permeation rate of the membrane
attains 90% of the steady-state value within 10 min.,  thus eliminating one
of the more serious drawbacks of current permeation tubes.

3.2.4.1  Sulfur Dioxide
A.  Principles Involved
    The calibration output requirements for SO- have been specified as 1 ppm
as the upper limit with the field usable test atmosphere generating
device delivering the doped air at 1 to 3 £/m for periods up to
2 hr.  Thus, the maximum flow rate of S0? required would be 0.001 to
        3
0.003 cm /min which is a several orders of magnitude lower rate than that
achievable with a pure polydimethyl siloxane membrane.  The use of a
polymer-perforated metal composite will have a lower permeability than pure
polydimethyl siloxane; however, it would be expected to be in the range
required.  Furthermore, the permeation rate can be regulated by the membrane
area and the temperature at which the device is operated.

B.  Physical Characteristics
    At the instrument site a supply or source for zero  (clean) air is
required.  This may be either a compressed tank of clean air or an air
                                   62

-------
scrubbing system.  Alternatively, zero air source can be a part of the
portable test atmosphere generating unit proposed.  For the purposes of
setting upper portable unit dimensions the following is suggested:
          Total weight:  3.5 kg
                       - air lecture bottle 2.3 kg
                       - permeation tube    0.5 kg
                       - valve + tubing     0.5 kg
                       - manifold system    0.2 kg
      Total dimensions:
                       - air lecture bottle 38.1 x 5.1 cm. diameter;
                       - permeation tube    12.7 x 2.5 cm. diameter
                             (estimated - will require experimental
                              confirmation);
                       - manifold system

C.  Operational Requirements
    The operation of the unit would consist of plugging in the heater of the
permeation tube set to produce the desired temperature and adjusting the
precision valve on the compressed air bottle for the desired rate of
diluent air delivery which could be directed into the instrument manifold.

D.  Maintenance Requirements
    Compressed air bottle - none;
    Permeation tube - none;
    Entire unit requires minimal maintenance.

E.  Shipping Requirements
    The membrane end of the permeation  tube could be fitted with  threads
and a gastight cover for shipping and storing purposes and when the  cover
is removed for screwing the tube into the manifold.
                                   63

-------
F.  Estimated Accuracy
    Accuracy of the existing permeation tubes is claimed to be as good as
+_ 2%.  The same would be expected for the proposed device.

G.  Estimated Cost
    1)  Development costs of approximately $45,000.
    2)  Initial Setup Costs
           Zero air in cylinder with regulator - $150
           S0~ in cylinder with regulator        $150
           Precision valve                       $200
           Permeation tubes                      $200
           Plumbing and bag                      $100

    3)  Operation costs are negligible

H.  Advantages/Disadvantages
    Known and constant concentrations of gases may be conveniently prepared
by using permeation devices.  Simple to operate with minimum of support
requirements.  Can be operated by unskilled personnel with high precision.
If replaceable membrane discs are designed into the tube, costs would
be minimal.
I.   Design Sketch

   TO INSTRUMENT
   MANIFOLD
          BAG
                      TO HEATER
                                                   PERMEATION TUBE
                                                        PRECISION VALVE
                                        -ZERO AIR
          Figure 3-11. Permeation test atmosphere generator.
                                   64

-------
 J.   Recommendation
     Undertake  development  of  the permeation  device.

 K.   References
 1.   O'Keeffe,  A.  E.,  and G. C.  Ortman.  Anal.  Chem.  3£ (6):760  (1966).
 2.   Lucero, D. P.   Anal. Chem.  43.  (13):1744  (1971).
 3.   Scaringelli,  F. P., S. A.  Frey,  and B. E.  Saltzman.  Amer.  Ind. Hyg.
     Assoc. J.  Z8/260,  1967.
 4.   Lindquist, F.,  and R.  W.  Lanting.  Atm.  Environment. i6:943  (1972).
 5.   Van  der Grinten,  W. J.  Ger. Offen. 2,237,793  (Feb. 22,  1973).
 6.   Reiszner,  K.  D.,  and P. W.  West.  Environ.  Sci.  &  Technol.  7:526  (1973).

3.2.4.2  Nitrogen Dioxide
A.  Principles Involved
    The general and specific considerations given for SO  apply equally
well to N02>   Because of the greater reactivity of NO-, apparently the only
membrane materials that have sufficient oxidative stability are the
perfluoro materials.  Thus, materials like Teflon, FEP, or Kel-F are
suggested for impregnation of the perforated stainless steel sheet.  Con-
ventional melt or solution processing is possible with FEP and Kel-F.   For
Teflon, in situ radiation-initiated polymerization of the monomer is
required.  A method for making pinhole-free composites of porous graphite,
bronze, or silica with tetrafluoroethylene or chlorotrifluoroethylene has
been patented recently (Ref. 1).

B.  physical Characteristics
    The calibration output requirements for N09 are  approximately those for
                                       3
S02, thus requiring a 0.001 to 0.003 cm /rain flow rate of N02<  This is a
rate achievable with  FEP and possibly with Kel-F, although the  latter material
has not been investigated as much.
     The portable unit dimensions and weight would be  the same  as those for
the S02 unit,  since the only 'difference in the  unit  is the membrane material.
                                    65

-------
C.  Operational Requirements
    The same as those for  the S0» unit.

D.  Maintenance Requirements
    The same as those for  the S0» unit.

E.  Shipping Requirements
    The same as those for  the S02 unit.

F.  Estimated Accuracy
    The same as that of the S02 unit - + 2%.

F.  Estimated Cost
    Development costs would be higher than for the sulfur dioxide source
because of the in situ polymerization technique used to form the membrane.
Our estimate is $50,000.   Other costs would be the same as for the S0_
source.
     2)  Production - same as those for the SO- unit.

H.  Advantages/Disadvantages
    The same as those for  the S02 unit except for possible NO,, compositional
changes with time and the possibility of more rapid deterioration of the
membrane.

I.  Design Sketch
    The same as that for the SO. unit as shown.

J.  Recommendation
    Undertake development of the permeation device.
                                  66

-------
K.  References
1.  Roesinger, S., W. Ulsamer, and G. Pietzka.  Ger. Offen. 2,108,363,
    Sept. 7, 1972.

3.2.4.3  Carbon Monoxide
A.  Principles Involved
    The same method needed for SO  is also applicable to CO.  Polydimethyl
siloxane is proposed as the membrane material.  Because of its high
permeability characteristics, a wide latitude in design and operating
temperatures is possible.

B.  Physical Characteristics
    The calibration output requirements for CO are  10 times those of N02»
thus requiring a  0.05 to 0.15 cm  /min flow rate of  CO.  With the use of
polydimethyl siloxane,  this rate  is  still lower than that  possible with the
pure polymer.
     The portable unit  dimensions and weight would  be the  same as those
for the S02 unit.

C.  Operational Requirements
    The same as those for  the S02 unit.

D.  Maintenance Requirements
    The same as those for  the S02 unit.

E.  Shipping Requirements
    The same as those for  the SO^ unit.

F.  Estimated Accuracy
    The  same as that of the  S02 unit - +  2%.
                                    67

-------
G.  Estimated Cost
    1)  Development - the same as that for the SO- unit.
    2)  Operation - the same as that for the SO- unit.
H.  Advantages/Disadvantages
    The same as those for the SO- unit.

I.  Design Sketch
    The same as that for the S02 unit.

J.  Recommendation
    Undertake development of the permeation device.

3.2.4.4  Ozone
         Ozone at temperatures above liquid nitrogen temperature is not
stable.  Thus, at -78°C about 9% of ozone decomposes in two weeks and a
permeation tube approach to preparation of known and constant concentrations
of ozone is not possible.

3.2.4.5  Butene-1
A.  Principles Involved
    The same considerations advanced for S02 are also applicable to butene-1,
Polydimethyl siloxane is proposed as the membrane material.  However, this
material may have flow rates much in excess of those required, and a
silicone copolymer such as silicone-carbonate may be a better candidate.

B.  Physical Characteristics
    The calibration output requirements for butene-1 are the same as those
                                            2
for NO-, thus, requiring a 0.005 to 0.015 cm /min flow rate of butene-1.
This is several orders of magnitude lower rate than achievable with pure
dimethyl siloxane.
                                    68

-------
     The portable unit dimensions and weight are the same as those for the
S02 unit.
C.  Operational Requirements
    The same as those for the SO- unit.
D.  Maintenance Requirements
    The same as those for the S02 unit.
E.  Shipping Requirements
    The same as those for the SO- unit.
F.  Estimated Accuracy
    The same as that of  the SO. unit - +  2%.

G.  Estimated Cost
    1)  Development - the  same as  that for  the SO-  unit.
    2)  Operation - the  same  as that for  the  SO-  unit.

H.   Advantages/Disadvantages
    The same as those for  the SO-  unit.

I.  Design Sketch
    The same as that  for the S0_  unit.

j.  Recommendation
    Undertake development of the  permeation device.
                                    69

-------
 3'3  Chemical Methods
      The  derivation  of  test  atmospheres  from chemical  changes  is  competi-
 tive  with physical techniques.   The most prominent example of  svich  a
 method  is the photolysis  of  oxygen to  ozone  stimulated by ultraviolet
 radiation.   In  considering possible methods  for  chemical generation of
 test  atmospheres, the following  categories are apparent:  radiation
 stimulated   changes  (e.g., photolysis  and radiolysis) , thermally stimulated
 decomposition,  energetic  reactions in  plasmas, electrolytic decomposition,
 and reactions by mixing.  In chemical  reactions, when  the only gaseous
 product is the  desired pollutant gas, then,  following the model employed
 for physical methods, the chemical species act as a  storage media for the
 pollutant, and the reaction, when it is rate controlled, performs the
 function  of a dispensing mechanism.  When more than  one gaseous product
 results from the reaction, then  the effect of the other gases  on instru-
 ment  calibration must be considered.   The presence of  other gaseous
 reaction  products may interfere  with the attainment  of suitable test
 atmospheres and thus negate  the  use of the method.
      Generally,  it is difficult to identify  reaction processes  controllable
 at  the rates required in test atmosphere generation.  Thus chemical methods
 tend  toward greater complexity than do physical methods.  A possible
 exception is the application of  electrochemical decomposition, not  to
 generate a pollutant but to  control the rate of discharge of a  pollu-
 tant  from a storage volume.

 3.3.1  Radiolysis, Photolysis, and Thermolysis
     Both photolysis and radiolysis involve the breaking up or activation
of molecular species by high energy radiation such that the subsequent
 recombinations result in the derivation of the desired pollutant species.
The only difference between the  two techniques is in the source of radia-
tion;  photolysis employs ultraviolet light quanta, while radiolysis
employs emanations from nuclear reactions.  Thermolysis achieves the same
result but with lower energy consisting primarily of decomposition by
                                     70

-------
heating.
     In general, thermolysis is more readily applied and provides less
complexity than do radiolysis or photolysis.  Particularly for high
molecular weight compounds, the results of the high energy methods are
complex mixtures of reactant products.  For thermolysis, specific bonds
are broken at fixed temperatures thereby providing well-defined decompo-
sition products.
     Chemical species can act as an efficient storage medium for a
pollutant gas; however, usually the control of the dispensing of
the gas is not possible in a decomposition reaction.   Characteristically,
the rate of reaction is a logarithmic function of temperature.  Typically,
the decomposition can be a simultaneous radical chain reaction as well
as a molecular reaction.  Thus, a batchwise injection of the generated
gas in a mixing/dilution bag is indicated for most cases.

     Many reactions that produce a selected contaminant  also  produce
additional products or generate unwanted side reactions.  These byproducts
in certain cases may have to be removed which complicates the system.
Finally, if a batch process is the only way to do it, because of the small
amounts of dopant gases involved, very small amounts of  the starting
material (a few milligrams or less) are required, which may affect the
precision of the measurements.

3.3.1.1  Sulfur Dioxide
     A.  Principles involved - The compound that we are proposing to use
in a batchwise process, in the absence of  the knowledge  about its decom-
position rate characteristics for a continuous process,  is a  coordinated
iridium complex of the  following structure (Ref.  1):
               [(S02)2 Ir  (2P-CH2CH2-P$2)2]Cl     $ - phenyl
This is only one member in a family of coordinated iridium complexes  that
form adducts with a very wide variety of gases including oxygen  and
hydrogen.  It decomposes  cleanly according to the reaction below:
                                     71

-------
                                     238°C
      [(S02)2 Ir  (2P-CH2CH2-P$2)2]Cl	>  [Ir^P-CH^-P^^ld + 2 SO^
Furthermore, It  is a compound with a high molecular weight (1152), is pre-
pared easily in  high yield (85-95%) (Refs. 1, 2), is stable in the solid
state and does not require a too high temperature (238°C) for its complete
decomposition.   Once it has been decomposed to generate the pollutant, it
can be regenerated with more SCL and used again.
     B.  Physical characteristics - The calibration output requirements for
                                               3
S02 have been specified as 0.005 ppm (13.1 ug/m ) as the minimum limit with
the field usable test atmosphere generating device delivering the doped air
at 1 £ per minute (lowest rate) for periods up to 2 hr.  Thus, the minimum
                                                  3          -5
volume (120 £ of doped air would require 0.0006 cm  (2.7 * 10   mmol) of
S02.  This small quantity of SO- could not be obtained directly in a precise
manner.  Thus, several dilution steps of conveniently doped, air would have
to be involved.  For example, for the purposes of illustration the following
steps could be taken:  1) decompose 15.552 mg of the iridium compound into
1.2 £ of zero air, which will result in a 5000 ppm SO. concentration; 2) dilute
this in three successive 1 to 100 dilution steps to obtain the required
0.005 ppm (13.1 ug/m ) SO. doped air.   If a Cahn Electrobalance is used
in weighing out the sample, one of the dilution steps could be eliminated by
weighing out a proportionately smaller sample.  Cahn Electrobalances are
accurate to ± 0.001 of a milligram in a milligram range.  The compound would
be weighed out in a quartz tubing (0.6 cm diameter x 15.2 cm length) and
positioned in the middle of the tubing and plugged with glass wool at both
ends.  For decomposition purposes the center part of the tubing would be
fitted with a heating tape or a microfurnance.
     At the instrument site, a supply or source for zero (clean) air is
required.  This may be either a compressed tank of clean air or an air
scrubbing system.  Alternatively, zero air source can be a part of the
portable test atmosphere generating unit proposed.
     The entire unit including compressed zero air lecture bottle, dopant
source tube and mixing/dilution bag (up to 360 £ capacity) would have the
following physical characteristics:
                                     72

-------
          Total weight:   3.4 kg
                       - air lecture bottle,  2.3 kg
                       - dopant tube, valve,  tubing,  0.2 kg
                       - mixing/dilution bag, 0.9 kg
          Total unit dimensions:  45.7 x 30.5 x 7.6 cm.
                       - air lecture bottle,  38.1 x 5.1  cm diameter
                       - source tube, 0.6 x 15.2 cm
                       - mixing/dilution bag  ~ 12.7 x 12.7 x 2.5 cm when
                                             collapsed;  inflated ~ 88 cm
                                             diameter sphere for 360 I
                                             volume.
     C.  Operational requirements - The operation of the unit consists of
connecting the source tube by means of Teflon connectors to the clean air
line; plugging in the heater, set to produce a predetermined temperature
(240°C) inside the tube; and subsequently delivering the required amount
of air through the source tube into the mixing/dilution  bag.  If several
dilution steps are involved, additional metering in successive dilution
bags would be Involved.
     D.  Maintenance requirements - Little maintenance should be required
other than regular flushing of the mixing bags after use.
     E.  Shipping requirements - None beyond the normal  parcel post packag-
ing.  The source compound is air stable.
     F.  Estimated accuracy - The estimated accuracy in a batch operation
will be determined by the accuracy of weighing out the solid sample and
the accuracy of dilution steps.  A sample on a Cahn Electrobalance can be
easily weighed out ± 0.01% in the sample weight range required.  Thus,
the accuracy with which the dilution steps are conducted becomes the
limiting factor; however, an accuracy of ± 10% should be achievable.
     6.  Estimated cost - Development costs  to assemble, test, and
calibrate this method are estimated  to be $40,000.  In operation, the
generation of test atmospheres would be under $10  for materials.
     H.  Advantages/Disadvantages - This method is relatively straightforward
and has a minimum of support requirements.  Until  found otherwise the approach
nay be capable of only a batchwise operation.
                                     73

-------
      I.   Design Sketch -
          ZERO  AIR
                                                     ADDITIONAL
                                                     SOURCE  TUBES
                                      METERING
                                      VALVE
                                      SOURCE TUBE
                                       HEATER
                                                    BAG
           Figure 3-12. Thermolysis test atmosphere generator.

      J.   Recommendat ion - Investigate decomposition kinetics for possible
 development  of  continubus doping process.
      K.   References
 !•   Vaska, L. ,  and D,  L.  Catone.   J.  Am Chem.  Soc.  JJ8:22,  5324 (1966).
 2-   Vaska, L. ,  and J.  W.  DiLuzio.   J.  Am.  Chem.  Soc.  83_:2784 (1964).

 3.3.1.2   Nitrogen Dioxide
     A.   Principles involved -  The same general  and specific considerations
 advanced  for S02 generation apply  to  N02 generation.   The  coordinated
 iridium complex for N02 has the following  structure:
                    [(N02)2 Ir  (<(>2P-CH2CH2-P<(>2)2]C1
 The compound is prepared  from  [Ir  (^-P-CH.CH.-PO.JCl and NO in air  which
                                    £•    £  t,    £ £            ^^^ "•""«•
 leads to  the nitrogen  dioxide adduct.   The compound  decomposes according
 to the following reaction:
                                      163°C
Ir
                                             [Ir
     B.  Physical characteristics - The  calibration  output  requirements for
N09 have been specified as 0.01 ppm 18.8 ug/m   at  the  lower limit with the
                                     74

-------
field usable test atmosphere generating device delivering the doped air at
1 i/min (lowest rate) for periods up to 2 hr.  Thus, the minimum volume
                                            3          -5
(120 £) of doped air would require 0.0012 cm  (5.4 x 10   mmol) of N02.
The same dilution approach would have to be used for NO. as for S02»  The
amount of the irldium compound required to give 10,000 ppm NO. in 1.2 £
of zero air is 30.132 mg.  This amount could be decreased so as to
eliminate one of the dilution steps.  The weight and dimensional consid-
erations are the same as for the S02 unit.  Because of the greater
reactivity of N02, stainless steel and Teflon are the preferred materials
for handling it.  The mixing/dilution bags, for example, would have to be
made of FEP Teflon.  As a general rule systems handling N02 require prior
conditioning with the gas.
     C.  Operational requirements - The same operational requirements as
for the SO. unit apply except that the decomposition temperature of the
iridium compound is lower (163°C).
     D.  Maintenance requirements - The same as those for the  S0_ unit.
     E.  Shipping requirements - The same as those for the S02 unit.
     F.  Estimated accuracy - The same considerations as those for  the
S02 unit apply—± 10%.
     G.  Estimated cost - 1) Development - the same as that of the  S02
unit.    2) Operation - the same  as that of  the S02 unit.
     H.  Advaneagea/Pisadvantages - The same as those for the  S02 unit.
     I.  Design sketch - The same as that for the S02 unit.
     j.  Recommendation - Investigate  decomposition kinetics  for possible
development of continuous doping  process.

3.3.1.3  Carbon Monoxide
     A.  Principles  Involved -  The  same  general and  specific  considerations
advanced  for  S02  generation apply also for  CO generation.   The coordinated
iridium complex  for  CO has  the  following structure:
                      [(CO)  Ir  «>2P-CH2CH2-P(|>2)2]  Cl
The compound  decomposes  according to the following reaction:
                                      75

-------
                                 260°C
   [(CO) Ir  ($2P-CH2CH2-P2)2]Cl	> [Ir  (
-------
     According to Hodgeson, Stevens, and Martin (Ref. 1) the most promising
approach to ozone generation is the ultraviolet photolysis configuration
as practiced by Regener (Ref. 2) and Tommerdahl (Ref. 3).  In their designs,
an extremely stable, low-pressure mercury arc lamp irradiated a quartz tube
through which the airatream flowed.  A shutter or variable aperture
between the lamp and quartz tube was used to vary the ozone output.  By
changing geometry, diluent flow rate, and aperture size, different ozone
mass flow requirements can be met.
     B.  Physical characteristics - An example of the ozone source unit is
that of Hodgeson, Stevens, and Martin (Ref. 1) and is depicted in Figure 3-13.
The photolysis source is a low pressure mercury arc lamp.  The lamp current
is regulated at 18 mA with a constant current power supply and a constant
voltage transformer.  The quartz tubing is type T-20 Suprasil and has high
transparency at 185 nm.  The box containing lamp and tube is an enclosed
system of 0.6 cm aluminum plate.  A hollow aluminum tube fits around the lamp
envelope and can be adjusted continuously to provide from zero to 100%
shielding of the radiation incident on the tube.
     The maximum output of an ozone source of the configuration shown in
Figure 3-14 depends upon:  1) lamp size, 2) tubing, 3)  inner surface
reflectivity, 4) airflow rate, and 5) applied lamp voltage.  By using
an 8-inch lamp and 25 mm o.d.tubing, the maximum concentration of 2.1 ppm
          3
(4116 ug/m ) of ozone at 10 £/min. airflow rate can be  obtained.  To
vary the concentration of ozone produced, it was found  best to keep the
airflow constant at the desired rate and vary the shielding of the lamp
envelope.  Thus, the generation of 0.005 to 1.0 ppm  (9.8 to 1960 ug/m )
ozone doped air at 1 to 3  £/min flow rate presents no problems.
     At the instrument site, a supply or source for zero  (clean) air is
required.  This may be either a compressed tank of clean air or an air
scrubbing system.  The dimensions  of the ozone source are 35.6 x 10.2 x
10.2 cm and the entire unit would  require also a diaphragm pump, drying
columns, needle valves, flowmeters,  and the associated  plumbing.
     C.  Operational requirements  - The operation of the unit requires  a
clean air supply and normal  electric power.   The operation would  consist
                                      77

-------
oo
              8 in. PEN-RAY
                  LAMP
                              J       3/8  in.
              AIR
                      I

                                                                         ALUMINUM
                                                                      BOX ENCLOSURE
                                                           14  in.
                                                     ADJUSTABLE  SLEEVE
                                           QUARTZ TUBE, 25 mm  o.d.
                                                                                  4  in.
                                                                                   i
                                                                                        J—-J.
                                                                                                           COLLAR
                                               Figure 3-13. Ozone source.

-------
of starting the diaphragm pump,  adjusting for the desired airflow rate,
plugging in the electrical supply for the generator and adjusting the
shielding of the lamp envelope for the generation of the desired ozone
concentration.  Stabilization of the instrument is usually achieved within
10 minutes.
     D.  Maintenance requirements - The unit maintenance would consist
primarily of periodic calibration of the instrument since the output of  the
source decreases with time.  Short-term stability of the unit is excellent
once provisions for constant airflow rate and elimination of fluctuations
in line voltage are made.
     E.  Shipping requirements - Should the shipment of the ozone source,
the diaphragm pump, and the associated plumbing be desirable, a box of
the approximate dimensions of 0.6 * 0.3 x 0. 3 m would be required.
     F.  Estimated accuracy - The accuracy of the described instrument is
in the area of + 3% for the high concentrations and + 3-5% for the low
concentrations .
     G.  Estimated cost - The cost of the instrument depends upon the
design features one wants but would be expected to be in the general area
of $500.  The cost of operating the instrument would be minor.  Develop-
ment would cost approximately $25,000.
     H.  Advantages /Disadvantages - The ultraviolet ozone generator provides
a dynamic test atmosphere generator that is reliable and simple to operate.
It is, however, bulky and needs periodic recalib ration.  The cost may forbid
wide use for the atmosphere generation.
     I.  Design sketch
                                                 ADJUSTABLE SLEEVE
                                                 OR FIXED APERTURE
                              UV LAMP
     DIAPHRAGM
     PUMP
V
  AIR
                                                        x:
     MOLECULAR
     SIEVES
                                  INSTRUMENT

                              QUARTZ TUBE
                    Figure 3-14. Ozone test atmosphere generator,
                                     79

-------
     J.  Recommendation - For ozone, the ultraviolet generator appears to
be the best choice at this time.
     K.  References
1.  J. A. Hodgeson, R. K. Stevens, B. E. Martin, A paper on "A Stable
    Ozone Source Applicable as a Secondary Standard for Calibration of
    Atmospheric Monitors" presented at the Analysis Instrumentation
    Symposium, Instrument Society of America, Houston, Texas, April 1971.
2.  Regener,  V. H.,  J. Geophys. Res. 69^:3795  (1964).
3.  Tommerdahl, J. B.   (Pasearch Triangle Institute, Research Triangle
    Park, N. C.), Air Pollution Control Office, Environmental Protection
    Agency, Final Report—Part I, Contract No. CPA-22-69-7 (1969).

3.3.1.5  Butene-1
     A.  Principles involved - Gas-phase reactions that have been studied
in some detail are of the following type:

                             CH-CX 	>- C=C   +    HX
                            /    \        /   \

where X - halogen.  Many of these reactions are known as the four-center
reactions because presumably, the transition state involves a structure
in which four adjacent atoms are sharing their valence electrons:
                                    •••••• C
                                          .\
                                          . \
                                  H
Thus, if n-iodobutane is used in the pyrolysis reaction, the expected
product would be butene-1
           CH3-CH2-CH2-CH2-I 	=9- CH3-CH2-CH -  CH2  t + HI
                                      80

-------
     B.  Physical characteristics - The calibration output requirements for
butene-1 have been specified as 0.05 to 5.0 ppm with the field usable test
atmosphere generating device delivering the doped air at 1 to 3 A/min for
periods up to 2 hr.   Since the concentration of the dopant gas  required
is low, either the turbulence of the diluent gas (nitrogen, air, etc.) can
cause the required vaporization of the liquid or, alternatively, it can be
atomized into an aerosol in the atomizer shown in Figure 3-15.
                  (7*
' / / / / / / /
      GAS
  LIQUID
      GAS
                                 NOZZLE
                            MIXING
                            CHAMBER
                                        \>
                     Figure 3-15. Aerosol generator.

Furthermore, many atomizer systems are available commercially which will
produce known concentrations of gas or vapors for calibration work in the
1 to 2*000  ppm  range.   Such systems are  accurate to +  1% if a precision
flowmeter is used for the diluent gas.  After the liquid has been atomized*
the mixed gas enters a quartz tube containing a ceramic core heater where
the pyrolysis takes place, Figure 3-16.,
                  GAS
            CERAMIC CORE HEATER
            GAS
            VARIABLE TRANSFORMER
             Figure 3-16.  Pyrolysis  gas  generator.

 and  from there through a soda lime trap into the instrument.  The
 kinetics of HI el mination, the pyrolysis  temperature, and the iodobutane
 vaporization  technique all need to be worked out.
                                     81

-------
      At the Instrument  site,  a supply or  source  for  zero  (clean) air or
 nitrogen is required.   This may be  either a  compressed tank of clean air
 or an air scrubbing system.   Alternatively,  zero air source can be a
 part  of the portable test  atmosphere  generating  unit proposed.
      The entire unit including compressed zero air lecture bottle,
 dopant source  tube,  and mixing/dilution bag  (up  to 360 £.  capacity)
 would have the following physical characteristics:
           total weight:  4.3  kg
                      — air  lecture  bottle    2.3 kg
                        - atomizer,  iodobutane  0.2 kg
                        - Pyrolyzing tube       0.5 kg
                        - soda lime  scrubber    0.5 kg
                        - storage bag           0.9 kg
           total unit dimensions:  45.7 x  30.5 x  10.2 cm
                        - air  lecture bottle  38.1 x  5.1 cm diameter
                        - atomizer 7.6 x 2.5 cm diameter
                       - pyrolyzing tube   25.4 x 2.5 cm diameter
                       - soda  lime  scrubber  5.1 x 2.5 cm diameter
                       - storage bag  ~ 12.7 x 12.7 x 2.5 cm when collapsed-
      C.  Operational requirements - The operation of the unit would
 require  iodobutane, clean air  supply>  and normal electric power.  The
 operation would consist of plugging in of  the electrical supply for the
 pyrolyzing unit and adjusting  for the desired air and iodobutane flow
 rates.
     D.  Maintenance requirements - The pyrolyzing tube and soda lime
 scrubber would require periodic replacement.
     E.  Shipping requirements - None beyond the normal parcel post packag-
 ing.  Iodobutane is a high boiling liquid that would require a glass con-
 tainer .
     F.  Estimated accuracy - No kinetic pyrolysis data are available.
Flow rate adjustments are capable of better than ± 10% so that an estimate accur
of ± 10% is reasonable.   However, experimental confirmation is required.
     G.   Estimated cost - A development effort is required that  would
                                    82

-------
cost on the order of $50,000 to assess the applicability of this method.
In operation, costs would be about $500 for each generator.
     H.  Advantages/Pisadvantages - This could be the basis of a dynamic
test atmosphere generator that is simple to operate.  However developmental
work is required.
     I.  Design sketch
         ZSRO AIR
       TO CERAMIC
       HEATER
       SODA LIME
       SCRUBBER
 EEDLE VALVE

 PRECISION FLOWMETER

 IODOBUTANE
 ATOMIZER

BAG
 STOPCLOCK
             Figure 3-17. Pyrolysis test atmosphere generator.
      J.   Recommendation - Investigate  decomposition kinetics  for possible
 development of continuous doping process.
                                      83

-------
3.3.2  Plasma Discharge
     A plasma is a gas consisting largely of charged particles.  Plasmas
may be produced by energetic particles or by quanta with sufficient energy
to ionize and excite neutral gas molecules and atoms.  Plasma physics
concentrates on very high temperature systems but plasma chemistry is more
concerned with low temperature plasmas such as are produced in discharge
tubes.  These are of interest because of the unique chemical reactions
found in the presence of free neutral atoms, ions, and free radicals.
Up until the last two decades, most plasma chemistry studies employed DC
or low frequency electrode or electrodeless'discharges.   Now the RF and
microwave discharge are predominately used to produce plasmas.
                                                                 •*.
3.3.2.1  Sulfur Dioxide
     A.  Principles involved - Although no studies have been made of
sulfur atoms produced in a discharge, there has been extensive study of
reactions in oxygen plasmas.  These have included the oxidation of
elemental sulfur and investigation of SO, SO-, and S0_ reactions.  The
only apparent application of plasma generation to the preparation of S02
test atmospheres is through the oxidation of elemental sulfur or the
reaction of sulfur compounds with atomic oxygen to produce S02.  One such
method is suggested here.
     A strip of filter paper impregnated with a predetermined amount of
the sulfur or a sulfide (polysulfide) serves as the storage medium for- the
pollutant.  In order to generate the test atmosphere, the filter paper
is subjected to a plasma discharge in clean air such that all of the sul-
fide is oxidized thereby producing a known quantity of sulfur dioxide.
This is diluted by mixing with zero air and employed as a test atmosphere.
The preparation and distribution of sulfur sources require a minimum of
effort and equipment.  The filter paper impregnation technique, although
not tested, should be easy.  It requires that between 1 and 500 yg of
releasable sulfur be available for the required dilution.
     At the instrument site, a supply or source for zero (clean) air is
required.  This may be either a compressed tank of clean air or an air
                                    84

-------
scrubbing system.  Since these are common to a number of generation con-
cepts, they are discussed elsewhere.
     The plasma source will consist of a small RF generator which drives
a coil.  The coil surrounds a glass tube in which is placed the calibrated
sulfur source.  The tube is flushed with zero air and its pressure reduced
to below 10 torr by means of a diaphragm pump.  The tube is closed off
and the RF generator turned on.  A discharge is maintained in the tube
for 10 minutes.  This creates active oxygen atoms which react rapidly with
the sulfur containing source to produce sulfur dioxide.  After completing
the oxidation, the plasma tube is filled with zero air and the mixture,
further diluted with air, is stored in a plastic bag.  This constitutes the
test atmosphere.
     For a 360 liter bag, the source must contain 516 yg of sulfur for a
                3
1 ppm (2617 yg/m ) concentration level.  Other pollutant concentrations
are readily calculable from this number.   It may be more efficient to
vary the concentration by providing different lengths of filter paper
with a standard sulfur content per unit length.
     B.  Physical characteristics - The sulfur source is a strip of impreg-
nated filter paper weighing about 20 g, and a test atmosphere generator is
required at field site consisting of the following components:
     1.  Zero air supply - Available in cylinders containing over 6000 £.
     2.  RF generator - Electronic cabinet with approximate dimensions of
         a cube with 0.5 m edges requiring 115 VAC supply at 100 W and
         with a mass of about 9 kg.
     3.  Plasma tube, plumbing and holding bag - The collapsible holding
         bag will have a volume of 360 S, (if spherical, a radius of 44
         cm).  The plasma tube and plumbing will have a small volume and
         mass.  A standard wet test meter is required to measure the total
         volume of dilutant air.
     C.  Operational requirements - Source preparation consists of process-
Ing of a quantity of impregnated filter paper, testing the batch for
sulfur content and uniformity, and storage in a passive environment until
used.  For test gas generation, the sulfur impregnated paper is oxidized
to SO- in a plasma and used to prepare a dilute test atmosphere.  Apparatus
                                     85

-------
and supplies  include
      1.   Clean  air  source  (360  £ @ STP),
      2.   An RF  generator,
      3.   115  VAC supply,
      4.   A wet-test meter  or rotameter,
      5.  Tubing, valves, and plasma tube, and
      6.  A plastic bag.
      D.  Maintenance requirements -
      1.  RF generator - No maintenace other than that occasioned by
         reduced output or failure.
      2.  Plumbing, plasma  tube, and bag - Flush after use and hold for
         reuse.  Check integrity on reassembly.
      3.  Wet test meter or rotameter - Normal use requires little or no
         maintenance.
      E.  Shipping requirements - These are minimal consisting of one
large brown envelope.
      F.  Estimated accuracy - Estimate ± 10% (requires experimental con-
firmation)
     G.  Estimated costs - Development costs are estimated to be approxi-
mately $50,000.  Source cost depends on volume of use.  The total cost of
100 samples will be approximately the same as for 1000 samples.  Setup
and testing costs are major items as follows:
                        Initial setup       $2,000
                        Cost per run        $2,000
     Assuming 1000 samples per run, the cost, including handling and dis-
tribution, is approximately $5 per test.  The apparatus required at the
test site is as follows:
                        RF generator        $500
                        Plasma tube          200
                        Wet test meter       500
                        Plumbing and bag     100
                        Zero air in cylinder
                          with regulator     150
                              Total cost   $1450
                                     86

-------
     H.  Advantages/Pisadvantages - The method has as its strongest point
the simple sample preparation and distribution.  However, it requires an
RF source at each test site and is subject to possible interferences from
other active plasma species.
     I.  Design sketch - See Figure 3-18.
     J.  Recommendations - This technique merits consideration but since it
hasn't been used or tested, experimental confirmation is required.  Con-
sideration should be given to possibilities for use of station apparatus
for other types of test atmospheres.
     K.  References
1.  Halstead, C. J., and B. A. Thrush.  Nature. 204;988  (1964).
2.  Rolfee, T. R., R. R. Reeves, and P. Harteck.  J. Phys. Chem. 69;849
    (1965).
3.  Kaufman, F.  Progress in Reaction Kinetics. Chapt. 1. N.Y., Pergammon
    Press, 1961.
4.  McTaggart, F. K.  Plasma Chemistry in Electrical Discharge. N.Y., Elsevier
    Publishing Co., 1967.
3.3.2.2  Nitrogen Dioxide
     A.  Principles involved - Production of N02 directly in a plasma dis-
charge is complicated by the various reactions producing NO, NO   NO,
and active nitrogen and oxygen.  These various species are found to exist
In various proportions with no well-defined stability criteria.  However,
if a controlled amount of NO is generated quantitatively by another method
and is mixed with an excess of plasma-produced ozone in a reaction chamber,
then NO2 can be generated quantitatively.
     Nitric oxide is produced quantitatively in an electrolytic cell con-
taining nitrosyl hydrogen sulfate in concentrated sulfuric acid as
Illustrated in Figure 3-19.  The NO is generated at the platinum cathode
while oxygen is generated at the anode.  In the electrolyte,  the reaction
is
                             N0+ + e" •» NO
Since only one electron is required per NO molecule,  the rate  of NO
                                     87

-------
  ROLL OF
  GLASS FIBER
  FILTER PAPER
                                 HOT DRY AIR
                                   DRYING TUBE
                      SULFIDE
                      BATH
 ROLL OF
 SULFUR
 IMPREGNATED
 PAPER
                           SAMPLE PREPARATION
                                                            TO INSTRUMENT
                                                                 MANIFOLD
        WET-
        TEST
        METER
CLEAN
AIR
SUPPLY
                               n n
                      RF GENERATOR
        FLUSH
/   N   LINE
                                            VACUUM
                                            PUMP
                                                              PLASTIC
                                                                BAG
              Figure 3-18.  Plasma test atmosphere generator.
                                    88

-------
  N2 + NO -«-
        Nr
                               H
                              r^
                                1   w
                             /
                          !^
                                       '/
                                   /  /
NOHSO. 'H.SO,
     4  2  4
Figure 3-19. Electrolytic cell for quantitative NO generation.
                                                       PLATINUM
                              89

-------
generation is given by
where                           q,, - rate of gas generation, mole a
                                 SL
                                I  - electric current, A
                                z  = number of electrons required for each
                                     generated gas (z - 1)
                                F  = Faraday's constant (9.65 x 10  C mol"1)
Since this is also the rate determining reaction for NO- production, and
                                                                   3
a concentration of N02 in the range 0.05 to 5 ppm (9.4 to 9400 yg/m ) is
required at flow rates from 1 to 3 £/min, one may calculate the cell
current required.  This is in the range from 3.59 to 1.08 mA, the former
                                                             3
current corresponding to a concentration of 0.05 ppm (94 yg/m ), and a
flow rate of 1 fc/min and the latter to 5 ppm and 3.4 min.
     The generated NO would be mixed with ozone, produced in a silent
discharge, in a reaction chamber and passed over hot glass beads to
remove excess ozone and to decompose the small amounts of N?0_ produced.
Nitrogen pentoxide results when the excess of ozone reacts with N0? by the
reaction
                              2N02 + °3 •* N2°5 + 02
while the thermal decomposition reaction is
                                2N205 -»• 4N02 + 02
     This ozonization of NO produces a dynamic test atmosphere for NO.
measuring instruments.  The critical control parameters are the cell
current and the test atmosphere flow rate.
     B.  Physical characteristics - Because of the nature of this test
atmosphere generation system, a "suitcase" design is described.  In this
Ojt N-, zero air, and electric power connnections are made to a generation
unit, the unit is turned on, and in ten minutes it is generating the test
atmosphere with a continuous flow.  It will continue to generate for 200
hours for each gram of dissolved NOHSO,.  Since the NO- generation rate is
precisely controlled within the instrumentation, only the total flow of
the test atmosphere requires measurement.
                                     90

-------
     Referring to Figure 3-20, the "suitcase" contains:
          a.  an electrolytic cell with a preset constant current power
              supply,
          b.  an ozone generator consisting of a discharge tube and a
              stabilized high voltage source,
          c.  a reaction tube,
          d.  a glass-bead loaded heated column with an automatic tem-
              perature controller, and
          e.  associated flow lines and a mixing chamber.
     Since only 0.01 g of the solute, NOHSO,, will be required for a 2
hour run at a maximum concentration level, the electrolytic cell will
require infrequent recharging.  The N.-O.-NO- doping gas flow rate of
about 100 ml/min will necessitate a 10-minute period to reach the
equilibrium level if the column and reaction cell have volumes of about
250 cc.
     C.  Operational requirements - Cylinders of nitrogen and oxygen are
required.  A lecture bottle of prepurified nitrogen or oxygen contains
56 & which is sufficient for 18 hr of test atmosphere generation.  (Cost
^ $26, gross wt. - 33 kg, size 20.3 x 68.5 cm, 2 stage regulator with
flowmeter cost » $150.
     Zero air is also required.  A No. 2 cylinder of zero air contains
2400 I which is sufficient for six hours of operation at the maximum rate
(Cost *> $26, gross weight * (33 kg), Size 20.3 x 68.5 cm, 2-stage regulator
with flowmeter cost ^ $150.
     A 117 VAC power source is required.
     D.  Maintenance requirements - The electrolytic cell must be cleaned
and the electrolyte replaced after every 100 hours of operation and must
be constructed so as to be spill-proof.  Otherwise no routine maintenance
is required.
     E.  Shipping requirements - If all gases are to be carried with the
unit, it is estimated that the total weight will be close to 63.6 kg.  If
gases are available at the site, the unit should weigh about 18.2 kg.  The
electrolytic cell and gas cylinders must be  separately packed  for shipment.
                                    91

-------
VO
ro
                       IN-*- 50 mi min
                                    -1
                           N2 IN
                         50 ml min
-1
                    ZERO AIR IN 1 t min
                                      -1
                                                           TEMPERATURE CONTROLLER
                                                               T   loo o o o o o o o o o, o oo         ^
                                                                I   I ooo00ooQ e o o  o o o I   I    _s.
                                                                  ss/sss/s
                                                                                     ••••••• I/'TT
                                                                                    ^ -^ fSSSS///\
                                                                   GLASS BEADS - 150°C
                                                                    REACTION TUBE
                                                                   OZONE GENERATOR
                                                     PLASMA HIGH VOLTAGE
                                                          SUPPLY
                                                                                              CONSTANT CURRENT
                                                                                              POWER SUPPLY
                                                                                                                   TO INSTRUMENT
                                                                                                                   MANIFOLD
                                      Figure 3-20. Nitrogen dioxide  test  atmosphere generator.

-------
     F.  Estimated accuracy - Capable of < 10% error but will realistically
provide 20% with untrained personnel.
     G.  Estimated costs - Development cost are estimated to be approxi-
mately $50,000.  The costs of a complete generation set-up are as follows:
         Gases and regulators, valves, and flowmeters $375
         Ozone generator                               500
         NO generator and power supply                 200
         Plumbing and glass bead column                200
         Oven and controls                             300
         Cabinet case                                  300
                                                     $1875
This cost is for one unit capable of  frequent repetitive use.
     H.  Advantages/Disadvantages
              Advantages
                1.  Self contained
                2.  Continuous flow
              Disadvantages
                1.  Weight and cost
                2.  Untested
     I.  Design sketch  - See  Figure  3-20.
     j.  Recommendations - Consider  trade offs  between a complex generator
like this and  a predoped gas  mixture with mixing and flow controls.
     K.  References
\.  T.  Singh,  R.  F.  Sawyer,  E. S.  Starkman,  and L.  S.  Caretto, Rapid Con-
     tinuous Determation of Nitric Oxide Concentration in Exhaust Gases",
     J.  Air  Pollution Control Assoc., 18, 102, Feb.  1968.
2.   G.  0. Nelson, Controlled Test Atmospheres,  Ann Arbor Science Publish-
     ers, Inc., Ann Arbor,  Mich.  1971.
 3.   J.  T.  Shaw,  "The Measurement of Nitrogen Dioxide in the Air", Atmos.
     Enviro., Pergamon Press _! 81-85, 1967.

 3.3.2.3  Carbon Monoxide
      Although there is some evidence that acetylene reacts in an oxygen
                                      93

-------
plasma to produce carbon monoxide, the further oxidation of this to carbon
dioxide has also been observed.  This prohibits the use of this reaction
for obtaining test atmospheres of carbon monoxides.  It may be possible,
however, to find an organic compound which reacts in a plasma to produce
carbon monoxide.  Preferably this would be accomplished by impregnating
filter paper with a known quantity of the organic compound and activating
it in the plasma of an inert gas.
     An alternative suggestion is to expose filter paper with a known
releasable carbon content to a carbon dioxide plasma such as to reduce a
known amount of the CO- to CO.  Although there is no experimental evidence
that this would be successful, it appears promising.  It would require the
same procedures and equipment as employed for S02 plasma generation except
for the addition of a source of pure C02 with which to fill the plasma
tube.  Costs, if successful, would be of the same magnitude as those for
S02 atmospheres.

3.3.2.4  Oaone
     A.  Principles involved - Most of the ozone required for industrial
applications is generated by means of plasma discharges.  For these
purposes a large variety of ozonizers have been designed and some of them
are commercially available.  These units have efficiencies as high as
        -5  J~^
3.8 X 10   g    (approximately 140 g of ozone per kw-hr of energy) when
pure oxygen is employed—less when air is used.  Where generation
efficiency is required, dry air and low operating temperatures are
employed.
     An ozone test atmospheric generating device based upon plasma dis-
charge can be provided.  An ozonizer tube similar to the type that would
be used in such a generator is shown in  Figure  3-21.  When an oxygen
containing gas is allowed to flow through such a tube and a high voltage,
(approximately 16kV)  is applied between the inner and outer electrodes,
a discharge takes place and ozone is generated.
     Low flow rate ozonizers of this type are not usually employed for
applications requiring a calibrated source.  Stabilization and calibration
                                     94

-------
                                                   SPACED
                                                                                             ELECTRODE
vo
Oi
OZONE
 OUT
^ OXYGEN
ZJ   IN
                                              SILVER CONDUCTIVE
                                              PAINT
                                                            "0" RINGS
                Note:  Approximately 15 kv is applied between the silver conductive paint and the electrode
                       in the form of pulses at a 16 Hz rate.
                                           Figure 3-21.  Instrument  ozonizer.

-------
are therefore a prime problem to be encountered in providing a plasma
generator as a source for test atmospheres.  There is no identificable
reason however why such a source cannot be obtained.
     The principle to be employed in test atmosphere generation of ozone
is that employing an ozonizer tube as described above.  The generation
rate will be controlled by two parameters; the oxygen content of the gas
in the discharge and the power supplied to the discharge.  With respect
to the first of these, an oxygen-inert gas mixture will be employed in
order to avoid complex reactions which are possible for more active
constituents in a plasma.  Since high voltage is commonly supplied to the
discharge tube in the form of pulses from an ignition coil, the most
obvious method for controlling the energy input is to control the fre-
quency with which pulses are supplied.  This can readily be accomplished
with a minimum of complex circuitry.  For a given plasma energy and
partial pressure of oxygen, the efficiency of ozone generation will
depend on the reaction cross section between the atomic oxygen produced
in the discharge and oxygen molecules.  Testing is necessary to determine
this efficiency for the various plasma conditions and to determine the
efficacy of the approach.
     Alternative techniques for quantitative generation of ozone in a
plasma could involve; 1) the use of pressures lower than atmospheric
in order to stabilize the plasma, 2) the use of chemical sources to supply
oxygen to the plasma in order to provide quantitative control, or 3) the
use of a pure oxygen supply with additional dilution.
     B.   Physical characteristics - The general principle for providing
a calibrated plasma source for ozone test atmospheres is described in the
preceding section.  Ozone will be extracted from an oxygen-inert gas
plasma.   In a 500 cm  plasma tube operating at atmospheric pressure with
                                               2   -2
the oxygen partial pressure at 1 torr (1.3 X 10  Nm  ) and assuming 100%
conversion (2 ozone molecules produced for each 3 oxygen molecules supplied)
then the ozone produced would give a concentration of 1.2 ppm (2352 ug/m )
in a 360 H test atmosphere.  If the conversion efficiency is 8%, then the
                                               o
concentration would be about 0.10 ppm (196 yg/m ).  Low concentrations are
                                     96

-------
obtainable by reducing the oxygen partial pressure,  by changing the dilution,
or by changing the plasma energy.  While these numbers apply to a static
atmosphere generator, they equally apply to a dynamic system which pro-
vides continuous generation.  Also it is important to note that the numbers
given are representative but do not represent the actual efficiencies which
will be obtained in a calibrated ozonizer with variable energy input.  These
numbers would be obtained through testing.
     The ozone test atmosphere generator will consist of an ozonizer tube
and its associated supply plus the plumbing and accessories necessary to
control gas flow rates and provide dilution.  A diagram for a proposed
system is shown in Figure 3-22.  Air will be supplied to the mixing chamber
through a rotometer and valve.  It can be atmospheric air supplied through
a pump with filtering and scrubbing of air from a compressed storage con-
tainer.  The ozonizer requires a supply of an inert gas-oxygen mixture at
a controlled low flow rate, approximately 100 ml/min.  At this rate, a
lecture bottle containing the gas mixture would last for over 6 hr.
     C.  Operation requirements - Samples on which to base the ozone content
of the test atmospheres are distributed in the form of lecture bottles con-
taining inert gas-oxygen mixtures in various controlled proportions.  The
generator in addition requires a supply of clean air and normal electrical
power.  The generator itself can be a small module not much bigger in
size than a lecture bottle and either distributed with the gas mixture or
provided as part of the test site equipment.
     Operation of the generation apparatus would require the provision
and attachment of the 2 gas supplies, the plugging in of the electrical
supply for the test atmosphere generator, and  its attachment to the  instru-
ment.  The only adjustment which is critical is that determining  the flow
rate of the dilutent air.   It  is not expected  that the ozonizer would be
strongly dependent on the flow rate of  the oxygen-containing gas  mixture.
     After the test atmosphere generator  is assembled and turned  on, it
will require a short period of time, approximately 10 minutes,  to stabilize.
After  this it may remain  in operation for 5 or 6 hr without  further
attention.
                                      97

-------
                 AIR
                             ROTAMETER
                             1-3 A/mln
                 OXYGEN
                 GAS
                 MIXTURE
00
                         FLOW
                         RESTRICTOR
                                                           OZONIZER
MIXER
                                                                                    IGNITION
                                                                                    COIL
                                                                                                       TO
                                                                                                    INSTRUMENT
                                                                        VARIABLE
                                                                        FREQUENCY
                                                                        PULSER
                                  Figure 3-22.  Plasma test  atmosphere generator for ozone.

-------
     D.  Maintenance requirements - Aside from preparation of the oxygen
mixtures, the generation system required little maintenance.   Operating at
relatively low input powers, the ozonizer tubes can be expected to have
relatively long lives and, properly used, the generation apparatus should
not be susceptible to contamination.
     E.  Shipping requirements - Depending upon the specific test scenario
adopted, a given test will require provisions of one lecture bottle con-
taining a gas mixture or alternatively provision of both the gas mixture
and the ozonizer which has been preset to a specific plasma energy level.
     F.  Estimated accuracy - An accuracy of ± 10% should be readily
obtainable using the test atmosphere generation system.  This could be
reduced to 5% or could be as much as 20% depending almost completely on
detailed cost-accuracy tradeoffs in the design.
     G.  Estimated cost - Development cost is estimated to be approxi-
mately $50,000.  The oxygen-inert gas mixture prepared in sufficient
quantities should cost under $5 for each such test atmosphere.  The exact
cost for the ozonizer depends almost completely upon the design and cost
of the ozonizer tube.  An estimate of $400 for the generator appears
reasonable but would be dependent upon its detailed design, required
precision, and number manufactured.  It would be required to carry out
a development project in order to test the efficacy of this technique
and to provide calibration and operational data.
     H.  Advantages/Disadvantages
              Advantages
                1.  Sample preparation and distribution is relatively easy.
                2.  A dynamic test atmosphere generator is provided.
                3.  The generator is relatively independent of operator
                    proficiency.
              Disadvantages
                1»  A development and testing program  is required.
     I.  Design sketch -  See Figure 3.22.
     j.  Recommendations  - While a plasma generator for ozone  can no doubt
be obtained and will operate satisfactorily, it must be compared to the
ultraviolet ozone generator for which considerably more experience has
                                     99

-------
been obtained.
     The logistics of this testing technique are more involved than in
some others suggested.
     K.  References
1.  F. K. McTaggert, Plasma Chemistry in Electrical Discharges, Elsevier
    Publishing Co., New York 1967.
2.  "Operation and Service Manual for Bendix Model 8101-B Oxides of
    Nitrogen Analyzer", Bendix Corporation, Ronceverte, West Virginia.
3.  A. P. Altshuller and A. F. Wartburg, Intern., J. Air Water Pollution
    4_ 70 (1961).
4.  S. Kaye and J. E. Koency, Anal. Chem. 41, 1491 (1969).
5.  S. Kaye and J. E. Koency, Rev. Sci. Instr. 4£ 505 (1969).
6.  B. L. Richards, 0. C. Taylor, and G. F. Edmunds, J. Air Pollution
    Control Assoc. 18 73 (1968).

3.3.2.5  Plasma Generation of 1-Butene Test Atmospheres
     Similarly to the situation with carbon monoxide, the chemistry of
active organic species in a plasma discharge is so poorly controlled as
to prohibit a viable approach to a 1-butene test atmosphere generator.  In
essence, while it may be possible to obtain such a technique, the research
cost that would be involved in attaining this capability is prohibitive.
3.3.3  Thin Films of Dissolved Reactants

3.3.3.1  Sulfur Dioxide
     A.  Principles involved - The same general type of apparatus as is
described in Section 3.2.3 can be utilized, except that two solutions
would have to be mixed, the reaction generating S02>  Many of the
reactions which generate S02 must be heated, so the contact column (See
Figure 3.23)  would require a water or steam jacket.  Solutions to be
used could be transported in glass containers and the hydrostatic head
could be maintained using an open glass leveling bulb.
                                    100

-------
REACTANT  	
SOLUTION

GAGE
METER     	
(HYDROSTATIC
HEAD)
CONTACT COLUMN
     AND
GLASS BEADS
CARRIER AIR IN
HOT WATER
    OR
STEAM JACKET
                             TEST AIR
                                OUT
                         HOT WATER
                            IN
          REACTANT
          SOLUTION
                                            HOT WATER
                                               OUT
MIXED SOLUTION
     OUT
      Figure 3-23.   Reactant test  atmosphere generator.
                             101

-------
      At about 50 to 100°C a solution of sodium dithionate (Na.S.O,)  with
                                                              226
 concentrated acid will produce SO..   The reaction is:
                                4.  950°r             4.
                    Na2S2°6 + 2H       s  H2S04 + 2Na  + S02'
 The acid should be in excess (Ref.  1)
      Other reactions may be used,  but  many are unsuitable because  of
 extreme reaction conditions (e.g.,  excessive temperature needed).
      B.   Physical characteristics  - Figure 3-23 is a diagramatic
 representation of such an exchange  system.   The two solutions would  be
 delivered at a known constant rate.  The exact rate of release  of  S00
 would best be determined empirically.
      C.   Operational requirements -  Operation should not be too complex.
 Once the apparatus is assembled:
      1.   Liquid flow is turned on by opening a valve for each of the
          solutions with the flow rate  controlled by capillary tubes.
      2.   Air flow would be controlled  at the specified rate.
      3.   When liquid begins to accumulate in the bottom of the  contact
          column the instrument to be tested can start  sampling  the test
          atmospheres.
      A.   Care should be taken to maintain the liquid at the preset
          level to insure constant solution fldw.
     D.  Maintenance requirements - The device should be cleaned and
checked for correct flows before and after field use.  Broken parts should
be replaced.  The whole system should be calibrated in the home laboratory
prior to being used in the field.  Solutions of reactants would be pre-
pared and analyzed in the home laboratory.
     E.  Shipping requirements - A case for the apparatus should be rigid,
the parts should be strapped in.  When shipped the interior should be
stuffed with shock absorbant filler of some kind.  It's weight should
not exceed 6.8 kg.
     The reactants should be shipped in inert sealed containers.
     F.  Estimated accuracy - Delivery of an accurately known concentra-
tion of the pollutant depends on measurements in the home laboratory which
relate concentration to the various settings of the apparatus and to
temperature.
                                   102

-------
     Precision of delivery will depend on control of temperature and the
accuracy of the hydrostatic head and the air flow setting.  A single
point flow meter can help the latter.  The hydrostatic head can be desig-
nated by fixed marks on a stick, but probably will require frequent adjust-
ment.
     6.  Estimated cost
         Estimated Development costs      $35,000
         Supplies - Each test setup         1,000
                      Each test                25
     H.  AdvantaRes/Disadvantages - With proper quality control the technique
should be adequately accurate and precise.  The advantage of this technique over
that of using dissolved gases is that no special care need be exercised to
prevent the loss of any reactant from solution before mixing in the contact
column.
     The major disadvantage is that hazardous reactants such as strong
acids are needed.
     I.  Design sketch - See Figure 3-23.
     J.  Recommendations - This  is not  too attractive, largely because it
is more complex and somewhat more elaborate than many
added necessity of having to maintain a constant temperature may also make
it less desirable.  There is a possibility that this technique could
prove attractive, but much development work would be required.
     The recommendation is not to develop this technique unless experience
shows none other to be completely satisfactory.
     K.  Reference
1.  Kleinberg, J., W. J. Argersinger, Jr., and E. Griswold, Inorganic
    Chemistry, D. C. Health, Boston, p.  453 (1960).

3.3.3.2  Nitrogen Dioxide
     A device for generating a test atmosphere of NO. could be constructed
exactly as described for SO, in  the previous section except that the hot
water jacket would not be needed.
     A solution containing sodium nitrate and sodium nitrite in equimolar
concentrations can be mixed with concentrated acid to produce NCL:
                                    103

-------
NaN0  + NaN0  + 2H+ -»• 2Na+
                 2 + NaN03       -»•   a     2       2       (Ref _  1)
     The cost would be similar to that for the S02 device.  Other
considerations will be the same as for the SO 2 device.
     Reference
1.  Cotton, F. A. and G. Wilkinson, Advanced Inorganic Chemistry, Interscience
    N. Y., p. 258 (1962)

3.3.3.3  Carbon Monoxide
     The same device and the same considerations apply in the case of
CO as for SO-.  The chemical reaction at 100 °C is between formic acid
(H2COOH) and concentrated H2S04
               H2COOH + H2S04 -»• H2S04 • HZ + CO.       (Ref . 1)
     Reference
1.  Cotton, F. A. and G. Wilkinson, Advanced Inorganic Chemistry,  Inter-
    science, N. Y. p 225.

3.3.3.4  Ozone
     No reaction system producing ozone is known  to exist.

 3.3.3.5  Butene-1
      The same considerations hold for Butene-1 as for SO- including
 estimated costs and recommendation.  The reaction would be dehydration
 of butyl alcohol with concentrated H-SO, :
            C4H9OH + H2S04 -»• H2S04H20 + C^g.       (Ref. 1)
 Since butyl alcohol also activates FID, a cold trap with suitable re-
 frigerant or an adsorbent of polar components such as ascarite should
 be placed between the generator and the instrument under observation.
      References
 1.  Whitmore, F. C., Organic Chemistry (Reproduction of 2nd Van Nostrand
     and Co. Edition 1951) Dover, Inc., New York, p. 39.
                                    104

-------
3.3.4  Electrochemical Methods
     There are two general techniques for employing electrochemical processes
In the generation of test atmospheres.  The first of these uses the con-
trollable quantitative generation of gases as a driving force for an
impurity injection system.  This merits careful consideration and is dis-
cussed below.  The second technique involves the direct generation of
the impurity gases in the electrochemical reaction.  This depends on the
existence of a suitable reaction and is therefore limited in its
applicability.  No suitable electrochemical reaction of the generation
of sulfur dioxide has been identified, however in  subsequent sections
application  of this technique to the  generation  of ozone .and nitrogen
dioxide  test atmospheres will be described.
      The versatility  of electrochemical processes  and  the quantitative
relation of  reaction  rates to the  controllable  flow of electric  current
indicates the need  for more extensive consideration than  is given here.
One would like to consider a solid electrolyte-gas generation  system for
example.  Perusal of  the  literature,  however,  indicates that phase
changes  involving gas generation are  unstudied.   Emphasis has  been on
reversible electrochemical processes  in which no phase change  occurs.
No data  exists on gas generation in such  processes. Other  potential
test  gas atmosphere generation  methods based on electrochemical processes
might be derived from release of entrapped gases in electrodeposited
films, by-products  of miniature fuel-cells,  the release of  entrapped
pollutant gas by electrochemical dissolution of a metal electrode, or
one of the more  complex  electrochemical processes.

3.3.4.1  Sulfur  Dioxide
      A.  Principles involved  -  The test  atmosphere generation scheme
using the electrochemically evolved gases as a driving force is described
in  this  action  for  sulfur dioxide  but is equally applicable to the injec-
tion  of  other impurity gases.   In  particular it will be referred to as a
competitive  technique in each of  the following sections dealing with
electrochemical test atmosphere generators.
                                   105

-------
     Test atmosphere generators based upon use of generated gases as a
driving force for injection of impurity gases into a flow stream can
take many forms.  Gas syringes using electrochemical gas generation as
the driving force are available commercially.  These are capable of
delivering from 0.04-4 ml/hr and can operate for 10 days without re-
charging the electrolytic cell.  They are battery driven.
     Since the number of molecules per unit volume of gas is independent
of the type of gas, the relationship between the rate of gas generation
in an electrochemical cell and the electrical current required does not
depend upon the type of gas generated except as that relates to the
number of electrons required to liberate one molecule of gas.  Assuming
2 electrons are required per molecule as occurs for example with the
generation of oxygen, then doping levels from .05 to 5 ppm (130-13,085
yg/m ) in gases flowing at rate from 1 to 3 Jl/min require cell currents
in the range from about 7 yA to 2 mA.  Since it is inadvisable to
operate electrochemical cells below 10 yA when quantitative generation
is required, the use of electrochemical cells for impurity injection
at the lowest concentration and lowest flow rate would not be advisable.
This is a very minor limitation on the operating range.
     The most convenient way to provide test atmosphere generation using
electrochemical reaction as the driving force appears to be in the form
of a small container.  The container, about the size of an aerosol paint
can, would contain the dopant gas, in this case sulfur dioxide.   In the
bottom of the can but separated from it by an extensible diaphragm would
be placed an electrolytic cell.  The diaphragm may actually consist of
either an elastic material or a collapsed bag of a selected plastic.
The electrolyte need be nothing more than salt water.  The electrodes
would be brought through the bottom of the can.  Upon receipt, the tip
of the sealed plastic capillary tube in the top of the can would be
broken off and the can attached to the instrument manifold.  Using a
small constant current power supply or a battery, the electrolytic
cell would be energized.  The electrolysis gases will expand into the
bottom of the can forcing the original contents of the can out through
the capillary at a constant rate.  This rate could be adjusted from 0.1
to 10 ml/min.  It would be possible to adjust the impurity concentration
                                   106

-------
in the gas by varying the current, by varying the dilutent flow rate,  or
by varying the dopant concentration in the canned gas.
     B.  Physical characteristics - This test atmosphere generation device
consists of a can of gas, the weight of which should be under one pound and
with no dimension greater than 15.2 cm.  While a more sophisticated power
supply can be employed, it is more convenient to employ a standard dry cell
as a source of power for the electrolytic cell.  This could require use of
a current limiting .resistor in order to obtain the accuracy required.
      C.  Operational requirements - At the test site,  the operator would
 take a can of gas as received and attach it to the intake manifold.  In
 doing this he would break off at a notch the tip of the output tube of
 the can.  After attaching the can to its manifold he would attach wires
 from a dry cell to the electrodes provided on the can and adjust the flow
 rate of air through the manifold to a specified standard flow rate.  He
 would then be ready to operate his instruments.
      D.  Maintenance requirements - There are no maintenance requirements
 associated with this method of test atmosphere generation.
      E.  Shipping requirements - There would be no extraordinary or
 complex shipping requirements associated with the use of canned gas.  Pre-
 filled cans of gas with internal pressure approximating normal atmospheric
 pressure and containing sulfur dioxide in a preselected mixture will be
 required.  Because of the low pressures, there would be no particular
 hazards associated with these shipments.
      F.  Estimated accuracy - Sulfur  dioxide mixtures can be provided in
 the can such that little deterioration would occur.  There would be little
 room for error in attaching a dry  cell to  the  terminals provided on the
 can, therefore the prime source  of error would be in the determination of
 tne flow rate in the measuring instrument manifold.  This however  is part
 of the normal operating requirements  on the measurement  facility and should
 Oot be considered as part of the test atmosphere generation  inaccuracy.   It
 is estimated that 5% accuracy would be readily obtainable as determined  by
 the can dimensions,  filling pressure, temperature,  and mode  of operation.
 Temperature equilibration would  be required.
                                     107

-------
     G.  Estimated cost - It is estimated that the cost of manufacturing
and filling a can with a specified SCL mixture would be less than $15 per
test when quantities on the order of hundreds are required.  This cost
would Include the gas, the built-in electrolytic cell, the separation dia-
phragm, and provision of a dry battery.  The can would be discarded after
use.  Development costs of $50,000 would not be unreasonable.
     H.  Advantages/Disadvantages
              Advantages
                1.  This is a simple and readily employable technique with
                    little opportunity for error.
                2.  The electrolytic Injection system exhibits cost
                    advantages.
              Disadvantages
                1.  In order to obtain a variety of concentrations and types
                    of gases it may be necessary to purchase machinery to
                    prepare and seal the cans.  More likely a commercial
                    supplier can be identified.
     I.  Design sketch - The concept of using a canned gas should be
given  careful consideration because of its  logistical simplicity and
its relative freedom  from error.  See Figure 3-24.

3.3.4.2  Nitrogen Dioxide
     A.  Principles involved - As described in the previous section for
sulfur dioxide, nitrogen dioxide can also be provided to a test atmosphere
by electrochemical injection.  However, nitrogen dioxide is one of the
gases which can be supplied as the product of an electrochemical reaction
and consideration will be given to this method for providing a test atmo-
sphere.  The reactivity of nitrogen dioxide makes this generation tech-
nique attractive compared to methods requiring long-time storage.
     Nitrogen dioxide cannot be produced directly from an electrochemical
reaction.  However nitric oxide is generated quantitatively by the elec-
trolysis of nitrosyl hydrogen sulfate (NOHSO^) dissolved in concentrated
sulfurlc acid at a platinum cathode under an atmosphere of nitrogen.  At
the same time the anode reaction produces oxygen that, when allowed to
mix with the nitric oxide before dilution, provides a quantitative yield
of nitrogen dioxide.  This same basic process is described in section
3.3.2.2 wherein the nitric oxide is reacted with plasma produced ozone to
give quantitative nitrogen dioxide.
                                    108

-------
PROTECTIVE
CAP
ELECTROLYTE
                                                 BREAKOFF POINT
                                                 CAPILLARY
                                                 OUTLET
                                                 TUBE
                                                   STORED  GAS
                                                   EXTENSIBLE
                                                   DIAPHRAGM
                          ELECTRODES

   Figure 3-24. Canned dopant gas with electrolytic drive.
                             109

-------
     In one electrolytic generator nitric oxide from the cathode and the
oxygen from the anode were led through a capillary tubing to a mixing
bulb where the nitrogen dioxide was formed.  The output of the mixing
bulb was measured with a nitrogen dioxide analyzer.  At 800 yA of cathode
current, sufficient nitrogen dioxide was produced to obtain a 12 ppm
concentration in a 1 fc/min dilutent stream.  Experiments indicated that
it would be possible to control the nitrogen dioxide production by con-
trolling either nitric oxide or oxygen evolution rates.
     B.  Physical characteristics - The design of the electrochemical
nitrogen dioxide test atmosphere generator will require ingenuity in order
that it satisfy the generator requirements.  One problem is  the low generation
rate for nitric oxide (12 v£/min) which is required.  This must be
mixed with the oxygen flow before dilution in order that it be converted
to nitrogen dioxide.  This means that the cell must have a minimum of
dead space or, at the low flow rate, it will require an unacceptable
long period of time for the generated gases to replace those in the dead
space.  Another problem is associated with the handling of the cell.  In
order that the cell be useful for test atmosphere generation, it must be
capable of shipment and handling.  This requires that the cell be designed
so that the liquid electrolyte will not enter the gas flow system thereby
creating barriers to the gas flow and depleting the electrolyte supply.
     Some solutions to these problems are evident.  For example, in the
cell design it is unnecessary to separate the anode and cathode.  This
would allow the cell itself to serve as the mixing chamber.  Provision
must be made for the fact that non-stoichiometric quantities of nitric
oxide and oxygen are produced.
     Ingenuity applied to this design problem should allow the develop-
ment of a self-contained generation unit which is no more than 15.2 cm
on a side and which contains the required cell or cells batteries and
other devices as needed.  The outlet of this unit when properly attached
to an instrument manifold would be capable of supplying nitrogen dioxide
test atmospheres over the full range of those required.
                                     110

-------
     C.  Operational requirements - An electrolytic nitrogen dioxide
generator which is properly designed should operate for a minimum of 10
hr without battery replacement.  Replacing the dry cells at regular
intervals should allow the generator to be used for 50 or more hrs without
regeneration of the electrolyte.  The unit would be relatively easy to
handle and would require a minimum of operator skills.  Optimally, no
external supplies of gas or electrical energy would be necessary.  It may
be advantageous to consider a non-reusable unit.
      D.   Maintenance requirements - See previous subsection.
      E.   Shipping requirements - The electrolytic nitrogen dioxide
 generator is in many respects similar to small laboratory wet cells employed
 for calibration of electrical equipment.  Thus, in a suitable shipping con-
 tainer,  they should be able to survive the rigors of transportation without
 damage and would not be limited in their transportability.  The small
 amount of concentrated sulfuric acid which is contained within the cell
 would be completely absorbed by the packing material if an accident should
 occur.
      F.   Estimated accuracy - In a laboratory environment with ample time
 for achieving stability, accuracies of better than 5% are readily obtainable
 with this generator, 1% accuracies have been claimed.  In a field test
 environment, with limited time for achieving stability, accuracies of
 approximately 10% should be expected.
      G.   Estimated cost - The largest cost associated with this test
 atmosphere generation system is that for development.  There are a number
 of laboratory measurements which must be made and the generator design
 requires considerable effort.  A sum of $50,000 could readily be utilized
 in accomplishing this.  Once the design cycle is completed, it would be
 expected that test atmosphere generators would cost on the order of $30
 in small production lots.  Periodic cell regeneration should cost no more
 than  $10 per cell.
                                     Ill

-------
    H.  Advantages/Disadvantages

             Advantages
               1.  The electrolytic cell is potentially capable of provid-
                   ing test atmospheres utilizing the flow rates in the
                   instrument manifold such that no additional external
                   gas or energy requirements exist.

               2.  The basic principle of electrolytic generation of
                   dopants for test atmospheres is its highly quantitative
                   nature which lends itself to accuracy.

             Disadvantages
               .  A considerable development effort is necessary to achieve
                  an acceptable operating generator.

    I.  Design sketch
                                                TO INSTRUMENT
                                                    MANIFOLD
 Figure  3-25. Electrolytic  test atmosphere generator for nitrogen dioxide.
3.3.4.3  Carbon Monoxide

     No electrochemical process has been identified from which carbon

monoxide may be derived.  Therefore, the only test atmosphere generation
                                   112

-------
technique which will result in a carbon monoxide test atmosphere is that
based upon electrochemical injection as described in Section 3.3.4.1.
This involves the provision of a gas mixture containing carbon monoxide
in known quantities such that it may be metered into an instrument mani-
fold flow by means of the electrochemical injection system.  This utilizes
the quantitative evolution of gas to exert a force on an isolated volume
of carbon monoxide which is then injected into the flow stream at a known
controlled rate,
3.3.4.4  Ozone
     The generation of ozone in an electrochemical reaction has been
studied extensively.  The cell utilizes sulfuric acid as an electrolyte
and two platinum electrodes.  Hydrogen is evolved at the cathode and a
mixture of oxygen and ozone at the anode.  This process is however not
sufficiently quantitative to serve as the basis of a test atmosphere
generator.
     Another method for quantitative generation of ozone could be con-
sidered based upon quantitative evolution of oxygen from an electrolytic
cell and the subsequent plasma conversion of the oxygen to ozone.  The
efficiency of plasma ozonization is however small and, coupled with the
low evolution rate of electrochemical processes, will be an unlikely
candidate for an ozone test atmosphere generator.
     Barring further research on either of the above two methods, it appears
that an ozone test atmosphere generator based upon electrochemical reactions
is only possible based on the injection technique described in 3.3.4.1.
This would depend upon the provision of ozone in a gas mixture with suffi-
cient concentration and stability to serve as a source for subsequent
dilution.  Although dilute ozone gas mixtures have been known to be very
stable under carefully prescribed conditions  (i.e.,  up to  8 percent ozone),
this technique appears to be a second choice which should be considered
only if other test atmosphere generation schemes encounter problems.  A
generator based upon electrochemical injection would be very similar
to and have the same general attribute as that described for sulfur dioxide.
                                    113

-------
3.3.4.5  1-Butene
     This non-methane hydrocarbon cannot be evolved from an electrochemical
reaction nor can it be derived from gases from such reactions.  Complex
organic species of this type require more sophisticated synthesis tech-
niques.  Thus the only test atmosphere generation scheme utilizing
electrochemical reactions which can be applied to 1-butene test atmospheres
is that based upon injection as described in Section 3.3.4.1.  One must be
careful to take into account the condensation of 1-butene at room
temperature and moderate pressure.

3.3.5  Films of Laminated Reactants

3.3.5.1  Sulfur Dioxide
     A.  Principles involved - Laminar films for generation of test
atmospheres could include:
         (1)  Films of Laminar material containing pure pollutant gas.
         (2)  Films consisting of or containing reactants which would
              be brought together mechanically in the test atmosphere.
     Items (1) and (2) above could be "sandwiches" in which the material
of interest is held between inert layers or they could be a layer or
layers encapsulated within the body of the inert material.
     Mechanically, the most easily delivered and disrupted film might be
a paraffin wax with bubbles of pollutant in it.  Such a paraffin wax
could be obtained or prepared with a desirable melting point, e.g., ~30°C,
(C16 to C2Q normal alkanes melt at ~18°C to ~37°C) [Ref. 1],  At a
temperature slightly above melting for the alkane wax, the pollutants could
be stirred into the wax.  The wax could be chilled into a film of con-
venient thickness, width, and length.   Some experience in quality control
would allow a fairly uniform product to be made.
     The film could be run into a chamber with a warm surface (Teflon
covered), the paraffin wax melted at a desired rate,  and the SO  released.
     A temperature of ~30°C would not generate a reaction between the SO
and the paraffin wax nor would it cause any other thermal reactions, such
as degradation of the pollutant to take place.
                                    114

-------
     Sulfur dioxide at one atmosphere pressure could be stirred or whipped
into melted paraffin wax as small bubbles and the material cooled rapidly
by being extruded as a thin film into cold water.  The pollutant content
would, as stated, depend on the condition of preparation.
     After mixing, while the wax is still easily malleable it can be
rolled into desired dimensions or even shaped into uniform sticks.
Alternately  solutions  could be prepared  as  films,  introduced into  the
test  air,  and melted simultaneously.  Higher melting wax could  be  used
for the  reactions
                Na0S.O, +  2H+ -> H0SO.  + 2 Na+ S00         (Ref.  2)
                  L  2. O           24            i
as it requires  a temperature of  50"C  or  higher.
      B.  Physical characteristics -  The  physical characteristics of  the
system are described above.
      C.  Operational requirements -  The  characteristics  of movement  (films
and air) should be determined by the central  laboratory, so  that only
mechanical assembly and  setting  of control  to  previously designated
readings is  necessary  by  the  field operator.
      D.  Maintenance  requirements  -  Production of paraffin wax films
containing known quantities  of pollutant,  plus quantitative  testing  of
the content  must be performed  at the central  laboratory. Calibration of
the heating  system must  be maintained.   Cleaning the apparatus and
removing melted wax would be necessary.
      Care  must  be taken in storage and  in transit to maintain  a cool
temperature  to  prevent melting or undue cold flow of the film.
      E.   Shipping characteristics -  The entire apparatus should weigh
4.5 kg or  less  (less case) and the case should weigh no more than 6.8  kg.
The overall  dimensions of the case should be no more than 38.1 x 30.4  x
 30.4 cm.
      A rigid wooden or metal container  and cushioning packing material
 should be sufficient.   Inside the shipping container the wax film should
be packed in some insulating material or surrounded by a refrigerant.
      F.   Accuracy - The accuracy would  depend almost entirely on  the
 quality control performed by the central laboratory.  The only judgment
 needed by the field operator would be in setting  flow rates of air.
                                    115

-------
     G.  Estimated cost and supplies

     Zero gas                                                 $60
     Sulfur dioxide                                            15
     Material for case                                         50
     Custom glassware                                         500
     Other reagents including wax                              60
                                                              685

     Equipment
     Heating element                                           25
     RPD motor                                                 30
     Heated-blender                                            90
     Flow meter air                                            50
     Gas regulators                                           430
     Ozone generator                                          190
                                                              815

     Service

     Teflon coating                                           500
     Machine shop                                             150
                                                              650
     Development cost                                     $40,000


     Cost of kit
     Gases and chemicals including wax film prepared           20
     Zero gas                                                  60
     Custom glassware                                         250
     Case                                                      50
     Teflon coating                                           250
                                                             $640

     H.  Advantages/Pisadvantages - If assembly of the apparatus and central

 laboratory quality control is good, it represents a very simple operation

 for  the  field person.
     Set air flow desired, turn on constant speed RPD motor and turn on

heating element.
     I.  Design sketch. See Figure 3-26.

     J.  Recommendations - The development of such a device should be

worth pursuing.

     K.  References

1.  Whitmore, F. C.  Organic Chemistry (Reproduction Of 2nd Van Nostrand
    and Co. Edition 1951) Dover, Inc., New York, p. 39.

2.  Kleinberg, J., W. J.  Argensinger,  Jr., and E.  Griswold, Inorg-arfic Chem.
    D.  C. Heath & Co., Boston p. 453 (1960).
                                   116

-------
   FILM WITH TEST GAS 	y                    /	 TEFLON  COVERED
                                                        HEATING ELEMENT
CLEAN AIR IN
            RPD MOTOR              >	 FILM CARRIAGE
                                          ASSEMBLY
          Figure 3-26.  Dissolved gas  test  atmosphere  generator.

 3.3.5.2  Nitrogen Dioxide
      The same considerations including cost and recommendations apply to
 N02 as to S02.
      If bubbles of solution are used instead of gaseous NO-,  the reactant
 solutions could be NaNO? + NaNO. and a strong acid:
                  NaN02 + NaN03 4- 2H+ + H20 + 2Na+ -I-
      Reference
 1.  Cotton, F. A. and G. Wilkinson, Advanced Inorganic Chemistry, Inter-
     science, N.Y. p 258 (1968)

 3.3.5.3  Carbon Monoxide
      The same design and considerations apply to CO as to N02«  If reactant
 solutions are used they could be liquid formic acid and concentrated I^SO^
 at 100°C:
                   H2COOH + H2SOA •»• H2S04H20 + CO .
      Reference
 1.  Cotton, F. A. and G. Wilkinson, Advanced Inorganic Chemistry Interscience,
     N. Y. p. 225 (1962).

 3.3.5.4  Ozone
      The same technique could be used for 0~.  This would  call  for  an
 additional  $250 for an 0. generator.  The temperatures  involved would not
                                    117

-------
cause a destruction of 0~.  The partial pressure of 0_ in the entrapped
bubble would  be only of the order of 0.01 atm.
     No thermal reactions of liquids or solutions are known to generate
0_, so the alternative is not open for the production of a test atmosphere
containing 0,.

3.3.5.5  Butene-1
     Butene-1 might dissolve in the paraffin wax, but the technique should
be a valid one for producing a butene-1 test atmosphere.  The same con-
siderations apply as for the S02 system.
     If the alternate system of generating butene-1 from two reactant
solutions is desired the reactant would be liquid butanol and concentrated
providing any excess butanol is removed by condensation on adsorption.
     Reference
1.  Whitmore, F. C., Organic Chemistry (Reproduction of 2nd Van Nostrand
    and Co. Edition 1951) Dover, Inc., New York, p. 4.
                                   118

-------
3.4  Biological Generation
     The metabolic diversity of biological systems and the application of
biological processes to the production of various hydrocarbons encourages
the consideration of such processes for test atmosphere generation.   The
practicality of application restricts consideration to microbiological
organisms of which there are many.  Micro-organism may be found that attack
all forms of organic matter as well as some inorganic matter.  In examining
the many products of such reactions, only a few are found that offer poten-
tial in test atmosphere generation.
3.4.1  Sulfur Dioxide
     Many biological systems are capable of producing sulfite ions as
an intermediate during sulfur and sulfate metabolism.  However, none
have been reported in the literature that are capable of producing
gaseous sulfur dioxide.
3.4.2  Nitrogen Dioxide
     Similarly to sulfur dioxide, the synthesis of the NO   ion is
also commonly found as an intermediate in nitrogen metabolism but the
production  of nitrogen dioxide has not been reported.
3.4.3  Carbon Monoxide
     A.  Principles involved - Biological production of  carbon monoxide
(CO) has been reported in bacteria  (Ref. 1,2), yeast  (Ref.  1), algae  (Ref. 3-8),
fungi  (Ref.  1,9), higher plants  (Ref.  10,11) in siphonophores (Ref.  12)  and
in man (Ref. 13,14).  It is well established that during the  catabolism
of hemoglobin in  senescent erythocytes in human beings,  the heme  ring  is
cleaved by  removal  of the  ot-methyne  bridge  carbon, which is then  oxidized
to CO.  Recently, it has been  reported that during formation  of phycocyan-•
obilin, equimolar quantities of  CO  are similarly  produced (Ref.  15).   Phyco-
cyanin is produced  by the  unicellular  alga  Cyanidium caldarium.   If  the  alga
is grown  in the  dark, it laclis  the  photosynthetic pigments  but synthesizes
chlorophyll £ and phycocyanin when  placed in the  light  producing 1 mole  CO
for  each  mole of  pigment.  The  proposed  technique for generating CO  as a
test gas  for field  use  is  based upon the above phenomenon.
                                   119

-------
     The amount of CO produced is a function of the number of cells of
Cyaniduim caldarium.  The number of moles of CO produced is directly
related to the rate of phycocyanobilin synthesis in £. caldarium.  The
rate of phycocyanobilin synthesis can be determined spectrally (Ref. 15);
thus the CO rate is also known.
     C_. caldarium cells are grown in the dark in a MF 114 New Brunswick
fermentator to yield about 140 g fresh weight.  The cell suspension is
aereated with 95 percent 0. - 5 percent CO..  The alga is illuminated (light
intensity to be experimentally determined) during which time phycocyanobilin
and CO are produced.  The cells are harvested and resuspended in new
media.  The cell density and phycocyanobilin content are determined spectrally
and a known number of cells are placed in a glass generating system to yield
known desired rates of CO.  The gas reservoir (headspace) is purged with
compressed air at rates to yield the final desired CO/air mixture.
     B-  Physical characteristics - An algal cell culture as described has
been reported to produce CO up to 72 hr after illumination has begun (Ref. 6)
Thus, it is proposed that CO generation be initiated at the laboratory
site.  The generating system is then transported to the field within these
time frame conditions.   The CO/air mixture should be scrubbed to remove CO
using a Baralyme or Ascarite cartridge.
     The physical parameters of this system are estimated as follows:
     Weight
     Glass culture container -                      11.3
       (with media
     Illuminating lamps  -                            1.4
     Mixing device -                                 2.3
     Compressed air bottle -                         2.3
       (serves  also as  aereator for culture)
     Teflon storage bag                               0.2
     Precision  valve and tubing                      0.2
                                       Total weight 17.7 kg
                                 120

-------
     Unit Dimensions
     Stirrer motor                          7.6 x 10.1 cm
     Air lecture bottle                     38.1 x 5.1 cm
     Culture container                      15 £ capacity
                   Total dimensions: 61 x 61 x 91 cm
     C.  Operational requirements - If the continuous delivery of CO is
found to be feasible from these algal cells, then the unit will require a
light source for illumination of the cells, a mixture motor for stirring
the cells, and the maintenance of a constant temperature in the culture
medium.  Adjustment of the vessel purge rate would provide a range of
concentrations of CO/air mixtures.
     D.  Maintenance requirements - The following maintenance is required:
     Compressed air bottle. Refill.
     CO-generation vessel. Cleaned and replenished with new culture of cells
     Storage bag.  Flush after using.
     Other.  As deemed necessary from development studies.
     E.  Shipping requirements - The vessel should be constructed so as
to prevent shattering.  Constant illumination is required during transport
DC power provided).  This method requires an airtight generating system.
     F.  Estimated accuracy - The accuracy of this method is unknown so
it must be determined experimentally; 10 percent should be possible.
     G.  Estimated cost
         (1) Development
             Cell fermentator                    $1,000
             Lamps                                  100
             Algal cells                            100
             Plumbing and bag                       100
             Zero air bottle                        150
             Development                         35,000
                                         Total  $36,450
         (2) Production
             Estimate of $200 per cell plus $50 per test.
                                  121

-------
     H.  Advantages/Disadvantages - This should be simple to operate
in the field, requiring no prior startup in field operation since
generation is initiated in the laboratory.  Calibration will be done in
the laboratory.  The method requires experimental development and verifi-
cation.
     I.  Design sketch.
                                                 LIGHT SOURCE
                                                          STORAGE BAG
                                         TO INSTRUMENT
             THERMALLY CONTROLLED AT
             OPTIMUM CO GENERATION
  Figure 3-27.  Biological test atmosphere generator for carbon monoxide.
     J.  Recommendation - This method requires considerable experimental
development to determine feasibility but may ultimately be a competitive
method.
     K.  References
1.  Junge, C., W. Seller, R. Bock, K. D. Greese and F. Radler.
    Naturwissenschaften. 58^:362-363  (1971).
2.  Junge, C., W. Seiler, U. Schmidt, R. Bock, K. D. Greese, F. Radler, and
    H. J. Rueger.  Naturwissenschaften. 59_(11) :514-515 (1972).
3.  Troxler, R. F.  Plant Physiol. 51/1):72-75 (1971).
4.  Crespi, H. L., D. Huff, H. F. DaBall, and J. J. Katz.  U.S. Nat.
    Tech Inform. Service, P. B. Rep. No. 213914/4, 29 pp.
5.  Wilson, D. F., J. W. Swinnertan, and R. A. Lamontague.  Science.
    168:1577 (1970).
6.  Troxler, R. F.  Plant Physiol. 48_:376-378  (1971).
                                  122

-------
 7.   Gafford,  R.  D.,  and C.  E.  Croft.   U.S.A.F.  Report.  58-128,  School
     of Aviation  Medicine, Randolph A.F.B.,  Tex.
 8.   Loervus,  M.  W.,  and C.  C.  Deliviche.  Plant Physiol. .38:371-374 (1963).
 9.   Westlake, D. W.  S., J.  M.  Roxburgh,  and G.  Talbot.  Nature.  189:510-511
     (1961).
10.   Siegel,  S. M. ,  G. Renwick, and L.  A.  Rosen.   Science.  137:683-684  (1962).
11.   Wilks, S. S.  Science.  U9_:964-966 (1959).
12.  Wittenberg, J. B.  J. Exp.  Biol. 3>7.:698-705  (1960).
13.  Sjostrand, T.  Acta Physiol. Scand. 26.:  338-344 (1952).
14.  White, P.  D.  et  al.  J.  Clin.  Invest.  46:1989-1988 (1967).
15.  Troxler,  R. F.,  R. Lester,  A.  Brown,  and P.  D. White.   Science.  167:192-193
    (1970).
3.4.4  Ozone
     There is no  recorded method by which ozone is produced in a biological
process.
3.4.5  1-Butene  (Nonmethane Hydrocarbon)
     No biological method is known for the production of 1-butene in a
biological process but if the objective is to obtain test atmospheres of
nonmethane hydrocarbons, then a process is available.  Ethylene is the
gas and it is available from an enzyme reaction.
     A.  Principles involved - Plants readily produce ethylene; e.g. bean
cotyledons, apple tissue, pea stems, green bananas, and cauliflower
florets (Refs. 1-5).  Several investigators have examined possible immediate
precursors for ethylene, including methionine, methional, propanol,
acetaldehyde, and 3-alanine (Refs. 6-8).
     Thompson and Spencer (Refs. 9,10) provided evidence which supported the
idea that 3-alanine was indeed the precursor responsible for ethylene
production in bean cotyledons.  Subsequently, Stinson and Spencer (Ref. 11)
reported the preparation of a soluble enzyme system from the subcellular
particulate fraction  of wax bean cotyledons which produces ethylene using
3-alanine as a substrate.  The test generating system proposed here is
based upon their enzymatic method  (Ref. 11).
        1.  Principles  of enzyme chemistry - The  generation of ethylene
for use as an atmospheric test gas will be based  upon a soluble  enzyme
system isolated  from bean cotyledons.  It is appropriate here  to review
                                   123

-------
the basic enzyme reactions responsible for synthesis of ethylene as well
as some enzyme chemistry prior to outlining the method proposed for field
use.
     Ethylene has been shown to evolve from the following enzymic reactions:
C02H
CH2 <—
3-alanine
^ ^°2H + 2
— cn2 * —
CHO
Malonate
semialdehyde
:K+. (f02li -H?.O ^
• 
-------
It follows that this relation can be expressed as
                   Product Concentration
or
               E
                 .III
                   Kt
where E is the enzyme concentration, K is a constant of proportionality,
and the reaction rate is measured and expressed as the amount of product
[P] formed per unit time, t.
     The enzyme concentration is therefore directly proportional to the
amount of product formed when all other conditions are held constant.
For example, if 0.2 yg/hr of ethylene is produced in one experiment and
0.6 yg/hr in another, the second experiment involves three times as much
enzyme.  Furthermore, the rate of reaction for enzymes can be expressed
in terms of specific activity which is the number of mmol of product
produced per unit time per mg of enzyme (S.A. » mmol ethylene/hr/mg enzyme)
On this basis calibrated amounts of ethylene can be produced by knowing
the S.A. of the enzyme and using the desired amount (weight) of enzyme at
the saturating amounts of substrate.
     Several other factors affect enzymic reaction rates.  As the temper-
ature is raised, the reaction rate of the enzyme reactions increases until
a maximum is reached (Figure 3-29).
                                           i37°C
           Q
           3
              W
            P-i P-J
                        10     20      30       40
                            TEMPERATURE (°C)

    Figure 3-29. Effect of temperature on reaction rate  for enzymes.
                                    125

-------
     The temperature coefficient (Q10) f°r an enzymic reaction is approxi-
mately 2; that is, the reaction rate is doubled for each 10°C increase in
temperature until a maximum rate is reached.  The maximum is usually at
37°C; higher temperatures denature the enzyme and the rate decreases.  The
apparent temperature for maximum reaction rate is greatly influenced by
other experimental conditions.
     The hydrogen ion concentration of an enzyme solution also has a
marked effect on reaction rate (Figure 3-30).  This involves a change in
the degree of ionization of functional groups at the active site of the
enzyme.  The enzymes display the greatest activity for ethylene evolution
in a solution at pH 7 which presents the greatest percentage of the
molecules in the above sequence in the proper ionic form.
S
o
I
w
,J
         H
         W
             20
             18
            16
            12
            10
                            7    8
                              PH
                                    10
           Figure 3-30. Effect of pH on ethylene production.
     Another factor that affects enzymic activity (reaction rate) is
addition of compounds called cofactors.  The addition of MgCl^, pyridoxal
phosphate, or thiamine pyrophosphate will enhance catalytic activity and
give greater rates of ethylene production.  Cofactor (and substrate) are
generally supplied at constant optimum concentrations for enzymes.  Thus,
the reaction rate is controlled primarily by the amount of enzyme present.
                                   126

-------
     Finally, the last important  parameter affecting reaction  rates  is
substrate (B-alanine) concentration.   With a constant enzyme concentration,
an increase of substrate produces at  first a very rapid rise in reaction
rate (Figure 3-31).  As the substrate, $-alanine concentration continues
    6-i
Q
W   5
|
iw 4-1
       w
  a 3 j
  s
  £2J
       w
                                   V
                                 /*  max
                                 \y
               ZERO-ORDER KINETICS
 MIXTURE OF ZERO
 AND FIRST-ORDER KINETICS
                      FIRST ORDER KINETICS
                      1/2
              0Klu 10   20
30   40  50   60
mmol g-ALANINE
                                         70  80   90.
Figure 3-31. Substrate vs product profile at constant enzyme concentration.

to increase, the rate of reaction begins to slow down until with a large
substrate concentration, no further change in rate  (zero-order kinetics)
is observed.  This reaction rate at this substrate  concentration is defined
as the maximum velocity, V    , of the enzyme-catalyzed reaction for ethylene
                          UlajC
under constant ethylene is constant and controlled  solely by enzyme availa-
bility.
     2.  Experimental parameters - This section describes the prepara-
tion of materials that are required for a system of producing ethylene  as
a test atmospheric gas in the field.
     In order to obtain cotyledons containing the enzymes for ethylene
production, seeds of Phaseolus vulgaris 1. var. Kinghorn wax are planted
in horticultural grade vermiculite and grown at 26°C in the dark.   Cotyledons
are picked  4-9 days  (should be experimentally determined) after planting.
                                   127

-------
     The enzyme  fraction for ethylene production Is prepared by grinding
 cotelydons  in  a  mortar containing 2 ml buffer/g cotyledons (buffer TES »
 0.05M N-trls-(hydroxymethyl)methyl-2-aminoethanesulfonic acid) at 0°C.
 The brei Is  filtered  through cheesecloth and the filtrate is subjected to
 differential centrifugation.  The supernatant from the first spin at
 2500 x  g for 10  min is again centrifuged at 32,000 x g for 15 min.  The
 pellet  formed  is resuspended in 0.01M TES, pH 7.6 at 0° and then freeze-dried.
 The lyophilized  fraction is used for ethylene production.
     The production of ethylene occurs when the following are all mixed:
     (a)  TES  buffer  (optimum pH for activity) is added to a known weight
          of lyophilized enzyme which will produce the desired rate of
          ethylene  (based on S. A. which is experimentally determined for
          a  batch of  enzyme).
                                                       I j
     (b)  Specific Concentrations of Cofactors (ATP, Mg  , Pyridoxal PO
                                                                       4*
          Thiamine pyrophosphate) are added to the enzyme solution.
     (c)  Temperature is brought to 37°C.
     (d)  The  reaction is started immediately upon the addition of an excess
          of substrate-B-alanine.
     B.  Physical characteristics - The calibration output requirement for
 ethylene should  be as high as 1 ppm with a constant generating system deliver-
 ing the sparged  air at 3 £/min (maximum rate) for up to 2 hr.  The maximum
 volume of test gas (360 1) would require a total of 360 pJl of ethylene.
 Thus, the rate of ethylene evolution should be 3 uH/min.   Based upon these
 criteria, the weight of lyophilized enzyme to use would be a direct function
 of the S. A. of  the enzyme (mmol ethylene produced/min/g  enzyme).
     At the test site, a supply of zero grade air (or inert gas) is required
 to dilute the ethylene to the final desired concentration.   The compressed
 air may be an integral part of the portable system.   The  air should be scrubbed
 of C02  and H20 resulting from the purging of the ethylene generating vessel.
This could be done after the test gas has been diluted to the appropriate
 concentration by employing in tandem cartridges of Lithasorb  (-CO ) and
 "drierite" (-H2)).
                                  128

-------
     The portable generating system conceptually consists  of:
                                                              Weight  (kg)
     (1)  Zero air lecture bottle                                  2.3
     (2)  Lyophilized enzyme, buffer,  cofactors, container         2.3
     (3)  Valve and tubing                                         0.2
     (4)  Storage container for synthetic ethylene/air mixture
          (teflon bag)                                             0.5
                                                     Total weight  4.8 kg
     Total Unit Dimensions:  61 x 45.7 x 15.2 cm.
     C.  Operational requirements - In order to generate the ethylene  at a
constant rate, its generation would involve adding measured volumes of buffer
and cofactors to a closed glass container at 37°C containing a known weight
of lyophilized enzyme.  The mixture would have to be gently stirred throughout
the test period.  The surface of the generating solution would be purged with
air (or inert gas as desired) at a calibrated fixed flow rate (controlled by
a precision valve) to yield the final synthetic air/ethylene mixture.   The
sparged effluent would be transported directly to an instrument manifold or a
teflon storage bag.  The system would require a heating jacket to maintain
the enzymic generating system at optimum temperature for ethylene evolution
(~37°C).
     D.  Maintenance requirements - Maintenance is required as follows:
         (1)  Compressed gas bottle.  Refilled
         (2)  Glass ethylene generating system.  Washed and dried
         (3)  Enzyme powder.  Prepared in -capsular form containing known
              desired amounts for ethylene production at known rates
         (4)  Cofactors.  Fresh solutions prepared, must be refrigerated
              prior to use at 0°C.
         (5)  Other.  As deemed necessary from experimental development.
     E.  Shipping requirements - Cofactors need to be refrigerated.
     F.  Estimated accuracy - The ethylene enzymic generating system should
produce it at a rate within + 5 percent of desired amount.  The overall re-
producability and accuracy could be + 10 percent or less but requires
experimental conformation.
                                    129

-------
     6.
Estimated coat
Initial development
- Compressed gas plus precision valve and plumbing      $300
- Cofactors (ATP, NADH, Thiamine pyrophosphate,
  pyrrdoxal phosphate)                                 1,200
- Seeds of Phaseolus vulgaris                             50
- Development
                                                               40.000
                                                       Total  $41,550
         Production
         Estimated to be $300 per unit plus $50 per test.
     H.  Advantages/Pisadvantages - This generator should  be simple to operate
in the field.  Large batches of lyophilized enzyme can be  prepared and
employed for many tests.
     There is a possible variability in the S.A. of enzyme preparation with
time (over 12 month period) or from batch to batch.  After each batch is
produced, S.A. would have to be determined experimentally  in the laboratory
(by GLC) for the rate of ethylene production per mg of lyophilized enzyme
powder.  Thus, initial calibration is required.  Variability of enzyme storage
must be determined during experimental development.
     I. Design Sketch of Portable System

1
61





COMPRESSED
GAS



cm






,


f.









n<
A








0 .
D




j
A









t
4.


(
(
—




5.7


1




(BA<

a

L





D

n

/

1
r


\



/



<
^

ooc>



GLASS VESSEL
WITH ENZYMIC
SYSTEM



LUrALluKo
(REFRIGERATED-
DRY ICE)
	 FM7VMT
                    TO INSTRUMENT     \
                                 MAGNETIC STIRRER
                                                               CAPSULES
                PRECISION VALVE
        Figure 3-32.  Enzyme test atmosphere generator for ethylene.
                                   130

-------
      J.   Recommendations - This method should be subjected to more careful
 searching before undertaking a development effort.
      K.   References
 1.   Burg, S. P., and C. 0. Clagett.  Biochem. Biophys. Res. Commun.
     27.: 125-130 (1967).
 2.   Ku,  H. S. , S. F. Yang, and H. K. Pratt.  Arch Biochem. Biophys.
     118:756-758 (1967).
 3.   Lieberman, M., A. T. Kunishi, L. W. Mapsan, and D. A. Wardale.
     Plant Physiol.  41.:376-382 (1966).
 4.   Mapson, L. W., and A. Mead.  Biochem. J. 108:875-881  (1968).
 5.   Mapson, L. W., and D. A. Wardale.  Biochem. J. 102_:574-85  (1967).
 6.   Mapson, L. W., and D. A. Wardale.  Biochem. J. W]_:433-442  (1968).
 7.   Lieberman, M. , and A. T. Kunishi.  Science. ].48_:938  (1967).
 8.   Shimokawa, K., and Z. Kasai.  Plant Cell Physiol  (Tokyo). 2:1-9  (1967),
 9.   Thompson, J. S.  ,and M. Spencer.  Nature. 210:595-597  (1966).
10.   Thompson, J. E., and M. Spencer.  Can. J. Biochem.  ^5_:563-571  (1967).
                                     131

-------
3.5  OTHER METHODS
     In the preceding paragraphs, one specific method is discussed for
each pollutant for each class of methods.  As pointed out in Section 2.2,
no one classification system can accomodate the varied methods which can
be proposed.  Even within the classification system employed, a method
may be presented which is superior by certain criteria but there may exist
other methods which are better by other criteria.  This section includes
methods not previously described due to the above reasons but which must
be considered to assure comprehensiveness.

3.5.1  Foamed Plastic
     Expanded plastics are employed for many common applications—packing
material and insulation are prominent examples.  In these a relatively
inert material such as polyethylene is fabricated with numerous air cells
resulting in a structually strong form which has a very low density when
compared to the unexpanded material.  It is suggested that if an appropriate
mixture of air and a pollutant gas is employed to form the expanded
structure then reproducibility could be obtained in the gas mixture
contained in the material.  If then, the material is mechanically crushed,
the gas mixture would be released for generating a test atmosphere.
     To apply this concept, apparatus would be required to form the expanded
plastic.   The expanded form could be a continuous strip of the material
which would contain a calibrated releasable amount of pollutant per unit
length.  It would be released by crushing.  For a dynamic test atmosphere
generator, this could be accomplished with a motor operated roller.
     This method while appealing from a cost and simplicity viewpoint
would have to be tested.  The stability of the pollutant mixture in the
plastic matrix, the degree of control in the release rate, and other
parameters would have to be examined before it could be applied to test
atmosphere generation.
                                   132

-------
3.5.2  Mlcrosyringe
     Very small quantities of gases are injected into chromatographic
columns and other analytical instruments using calibrated microsyringes.
Sample sizes of fractions of a microliter may be injected and motor
driven constant rate syringes are available.  Thus, it is a simple
procedure to provide an appropriate gas mixture that can be withdrawn
from a container through a septum and injected through another system
into an instrument manifold.  Through choice of the pollutant concentration
being injected, almost any test atmosphere concentration can be obtained
with sufficient accuracy.  The only apparent problems are associated with
the cost of the syringe and accessories and the design of the generation
system so as to minimize operator errors.

3.5.3  Burn Stick
     Various pyrolysis methods are adaptable to the use of a burn stick
in test atmosphere generation.  Using, as an example, the generation of
an SO2 test atmosphere, a suitable slow burning fuse or burn stick may be
impregnated with a suitable sulfur compound.  With a slow controlled burn
rate, in an air stream, the sulfur is oxidized in the combustion to give
S02 at a predetermined rate.
     Slow burning fuses are available and provide sufficient temperature to
oxidize sulfur.  The primary problems are associated with the production of
other gases in the combustion process.  Various pollutant levels in the
test atmosphere can be obtained by control of burn rate and the chemical
concentration in the stick.  In operation, the prepared stick in a suitable
holder would be ignited and inserted through a port into the air stream in
the instrument manifold.  It would be designed to burn for the required
period.  This method is suited to S02 and may be possible for N02 and CO.

3.5.4  Bubbler Systems
     Control of low concentration gas mixtures has been achieved using
bubblers in laboratory applications for many years.   In brief, the  flow  of
                                   133

-------
a gas or gas mixture is controlled by employing a fine valve to release
gas through an opening into a non-reactive liquid such that uniform size
bubbles are obtained.  The bubbles travel up through the liquid at a
rate dependent on bubble size and liquid viscosity and are released into
the air flow above the liquid.  Measuring the size of the bubbles immediately
before they escape and the rate at which they are released provides the
flow rate for the gas.
                                                                3
     While it is easy to obtain dilution factors of 1 part in 10  or
better in a bubbler system, provision must be made for thorough mixing in
the gas stream because of the non-continuous supply of the pollutant gas.
One must also be aware of the solubility of the gases in the liquid used;
it should be low in order to avoid errors.
     There is no doubt that a bubbler dilution system can be employed for
test atmosphere generation of some of the five pollutants.  There may be
difficulty in using it with ozone because of the stability problem.  It
does require a supply of a gas mixture containing the pollutant gas and is
thus a dispensing system.
     The same operation can be obtained with a rotating stopcork which
releases one stopcork bore volume for each turn.

3.5.5  Exponential Dilution
     If a predetermined amount of pollutant is released, perhaps by crushing
a vial, into a fixed volume of air and clean air flow into the volume at
a known fixed rate while polluted air flows out at the same rate, then the
concentration of pollutant in the effluent decreases from a high known
initial concentration at an exponential rate.  This rate is given by

                              C - C0 exp  (- ££)

where
     C   is the concentration at time, t
     CQ  is the initial concentration
     F   is the flow rate
     v   is the volume
                                  134

-------
     Vials of S02, NO-, CO,  or C^H_ mixtures could be readily prepared
for this method and the apparatus and operation of the system could be
simplified by design to obtain reliable performance.   This system can be
made to work and has a number of advantages.
                                    135

-------
4.0  COMPARATIVE EVALUATION

4.1  Selection of Panel
     A panel was selected for the purpose of examining the attributes of the
various proposed methods for generation of test atmospheres and to recommend
based on the results of the evaluation technique to be presented in
Section 4.2, the candidate method for generation of each pollutant gas
(i.e., S02, N02, CO, 0.,, and 1-butene).   Panel members were selected to
bring together individuals with expertise in the fields of physics,
chemistry, engineering, and air pollution.  The panel members were as
follows:
     1)  Dr. Raimond Liepins (Research Triangle Institute) -
              Analytical Chemist
     2)  Dr. L. K. Monteith (N. C. State University) - Professor
              of Electrical Engineering
     3)  Dr. L. A. Ripperton (Research Triangle Institute) - Air
              Pollution/Atmospheric Chemist
     4)  Dr. R. M. Burger (Research Triangle Institute) - Physicist
     5)  Mr. C. E. Decker (Research Triangle Institute) - Air
              Pollution Chemist
4.2  Evaluation Technique
     To assist the members of the review panel in the comparison of and
ultimately the choice of recommended methods for the generation of. the
required test atmospheres, summary tables of the fifty methods were
prepared and a methods ranking system developed with the special constraints
of the project in mind.  A summary of eleven methods is presented in
Tables 4-1 through 4-5 for each pollutant and contains a brief description
of each method.  The ranking system developed for the comparative evalu-
ation is presented in Table 4-6 and includes the methods evaluation
criteria described in Section 2.1 (Table 2-1) for field usable test
atmosphere generating methods, as well as other constraints deemed perti-
nent to the requirements for which the device/devices were to be used.
Briefly, these criteria include the following:  simplicity of method;
equipment/processing requirements; operator skill requirements; portability
                                  136

-------
            TABLE 4-1.   SUMMARY OF METHODS— SULFUR DIOXIDE
     Class
Method
  #
Description
Contemporary                1

Desorptlon                  2


Effusion                    3


Thin film evaporation       4


Novel permeation            5

Photolysis/thermolysis      6


Plasma                      7


Reactive films              8


Electrochemical             9



Laminated reactants        10

Biological                 11
           Conventional permeation tube

           Heating of loaded adsorption tube with
           subsequent dilution

           Low constant pressure can using Freon
           with calibrated capillary dispenser

           Column desorption from CC1, solvent
           with carrier gas

           Perforated metal-polymer permeator

           Thermolysis of irridium complex with
           dilution

           Polysulfide impregnated filter paper
           in RF discharge containing oxygen

           Column reaction of liquid films of
           sodium dithronate with sulfuric acid

           Electrolytic generation of gas to
           obtain calibrated dispensing rate
           for gas mixture

           Paraffin wax containing S02 bubbles

           No method recommended
                                  137

-------
           TABLE 4-2.  SUMMARY OF METHODS - NITROGEN DIOXIDE
     Class
                          Method
                      Description
Contemporary


Desorption


Effusion



Thin film evaporation


Novel permeation


Photolysis/thermolysis


Plasma


Reactive films



Electrochemical


Laminated reactants


Biological
10
11
Gas-phase titration of NO from
calibrated source with ozone

Carrier gas through loaded adsorption
tube with subsequent dilution

Low constant pressure can using Freon
with calibrated capillary and on-off
valve

Column desorption from 1,3,5-tioxane
solvent with carrier gas

Perforated plate-polymer permeator
with perfluoro materials

Thermolysis of irridium complex with
dilution

Mixing of electrolytically generated
NO with plasma generated ozone

Column reaction of liquid films of
sodium nitrate and sodium nitrite with
acid

Electrolytic NOHSO^ cell and NO + 0
reaction

Paraffin wax containing sodium nitrate
sodium nitrite, and acid bubbles.

No method recommended
                                  138

-------
          TABLE 4-3.   SUMMARY OF METHODS - CARBON MONOXIDE
     Class
Method
  I
Description
Contemporary                1

Desorption                  2


Effusion                    3


Thin film evaporation       4


Novel permeation            5


Photolysis/thermolysis      6


Plasma                      7

Reactive films              8


Electrolytic                9



Laminated reactants        10


Biological                 11
           Calibrated gas mixture

           Carrier gas through loaded adsorption
           tube with subsequent dilution

           Variable pressure can with metering
           valve

           Column desorption from CC1, solvent
           with carrier gas

           Perforated plate-polymer permeator
           using poly (dimethylsiloxane)

           Thermolysis of irridium complex with
           dilution

           Reduction of C0« in plasma

           Column reaction of liquid films of
           formic acid and sulfuric acid

           Electrolytic generation of gas to
           obtain calibrated dispensing rate
           of gas mixture

           Paraffin wax containing bubbles of
           formic acid and sulfuric acid

           Algal cell
                                 139

-------
                TABLE 4-4.  SUMMARY OF METHODS - OZONE
     Class
Method
Description
Contemporary                1

Desorption                  2



Effusion                    3

Thin film evaporation       4


Novel permeation            5

Photolysis/thermolysis      6

Plasma                      7

Reactive films              8

Electrolytic                9

Laminated reactants        10

Biological                 11
           UV generator referenced to KI

           Carrier gas through calibrated
           adsorption tube loaded with ozone on
           silica gel

           None recommended

           Column desorption from CC1, solvent
           with carrier gas

           No method recommended

           UV photolysis (conventional)

           Plasma ozonizer tube

           No method recommended

           No method recommended

           No method recommended

           No method recommended
                                  140

-------
               TABLE 4-5.  SUMMARY OF METHODS - 1-BUTENE
     Class
                          Method
           Description
Contemporary                1

Desorption                  2


Effusion                    3


Thin film evaporation       4

Novel permeation            5

Photolysis/thermolysis      6

Plasma                      7

Reactive films              8


Electrolytic                9



Laminated reactants        10


Biological                 11
Dilution or permeation tube

Carrier gas through loaded adsorption
tube with subsequent dilution

Lecture bottle with fine metering
valve

Column desorption from solvent

Perforated-plate polymer permeator

Thermolysis of n-iodobutane

No method recommended

Column reaction of liquid films of
butyl alcohol and sulfuric acid

Electrolytic generation of gas to
obtain calibrated dispensing rate
for gas

Wax containing bubbles of butanol and
sulfuric acid

Enzyme generation of ethylene as an
alternative to 1-butene
                                  141

-------
                   TABLE 4-6.  METHOD RANKING SYSTEM
1)  Simplicity of Method
        A     Relatively uncomplicated
        B     Moderate
        C     Complex
2)  Equipment/Processing Requireemnts
        A     Minor requirements
        B     Moderate requirements
        C     Complex requirements
3)  Operator Skill Requirements
        A     Low
        B     Moderate
        C     High
4)  Portability
        A     Unique requirements for test weight under 2 kg
        B     Unique requirements for test weight 2 to 20 kg
        C     Unique requirements for test weight over 20 kg
5)  Cost
    Development                Apparatus             Per Test
        A     Under $20,000       A     Under $100      A    Under $10
        B     $20,000-$50,000     B     $100-$500       B    $10-$30
        C     Over $50,000        C     Over $500       C    Over $30
6)  Safety
        A     No obvious safety problem
        B     Moderate hazards present
        C     Specific safety precautions required
7)  Range
        A     Full test range capability
        B     Perhaps full test range capability
              but some uncertainty
        C     Partial test range capability
                                  142

-------
           TABLE 4-6.  METHOD  RANKING SYSTEM (Cont'd)
 8)   Blind Mode
         A    Capable of  full  blind mode  operation
         B    Subject to  operator  compromise
         C    Incapable of blind mode operation
 9)   Accuracy
         A    Error ^ to  5 percent
         B    Error 5 to  10 percent
         C    Error 10 to 20 percent
10)   Ease of Implementation
         A    Ready for assembly and use
         B    Requires development, but principles proven
         C    Depends on unproven principles
11)   Confidence  of Success
         A     High
         B     In between
         C     Low
12)   Maintenance
         A     Little
         B     Moderate
         C     Much
13)  Operational Capability
         A     Dynamic  (Continuous)
         C     Batch
 Weighting Factor
         A - 4
         B - 3
         C - 2
                                    143

-------
of equipment required for method; cost (development, apparatus, per test);
safety; range; capability for blind mode of operation; estimated accuracy;
ease of implementation; confidence of success; maintenance requirements;
and operational capability.  Bach of the selected criteria was  assigned
an equal weight in the ranking system for comparison of methods.  Usually,
three ranks are included as A, B, or C for each of the thirteen
criteria selected and are assigned a numerical factor of 4, 3,  and 2,
respectively.  In cases where only two ranks are used, A and C
are used with the same numerical values.   The criteria and rank
descriptions become obvious upon examination of Table 4-6 and will not be
further discussed.  The utilization of the methods ranking system to
select the recommended method for each pollutant is further described  in
Section 4.3 (Panel Study).   Use of the methods ranking system should result
in the selection of the best method or concept for generation of test
atmospheres at field locations.  This statement is based solely on the
assumption that the selection criteria and constraints imposed  are correct
for this application.  It is also assumed that an external source of
diluent air is available or will be provided at each field location.

4.3  Panel Study
     The panel met on December 14, 1973 to examine in detail the fifty
discussions prepared for the generation of test atmosphere at field
locations.  The panel discussion began with a statement of the  objective
of the project, which was to examine various classes of physical, physical
plus chemical, and biological methods applicable to devices for generation
of test atmospheres for use in assessing instrument/operator performance.
The constraints and requirements applicable to such a quality assurance
program were reviewed.  The panel meeting continued with a brief discussion
of each class of methods (i.e., physical, physical plus chemical, and
biological) with particular emphasis being given to advantages, disadvan-
tages, and cost trade-offs for each method within the three classes.
Technical aspects, regarding the theory of operation, were reviewed and
practical considerations such as the relative merits of each method were
discussed.
                                  144

-------
     After  completion of  the discussion session, the evaluation technique
designed  to obtain a consensus opinion for the recommended method for
generation  of each test atmosphere was presented and discussed.  Basically,
the evaluation procedure  consisted of a ranking system based on numerical
considerations for the thirteen sets of criteria previously described in
Section 4.1.  Five work sheets identical to that shown in Table 4-7 were
distributed to each panel member, along with a summary of methods
(Section  4.1 - Tables 4-1 to 4-5) and the methods ranking system
(Table 4-6).  Each of the thirteen sets of criteria were reviewed by the
panel and discussed.  In several cases pertinent comments were raised
regarding the criteria and constraints presented for the methods ranking
system.   These comments were reviewed, modified as needed, and a consensus
set of criteria developed.  These criteria are identical to those presented
in Table  4-6.  For the most part few revisions were necessary.  Each
panel member was then requested to review, in detail each of the fifty
discussions and to complete the five tables needed to obtain a consensus
opinion for the recommended method for generation of each pollutant.  Each
panel member was given three working days to complete this assignment.
The panel meeting was then concluded.
     Upon receipt of the five data sheets from each panel member,  a
consensus comparative ranking of methods for each pollutant was determined
and is presented in Tables 4-8 through 4-12.

4.4  Results
     Since each of the thirteen sets of criteria presented in Table 4-6
was given an equal weight, the results of the comparative evaluation were
computed by totalling the numerical equivalent of the sum of the rankings
for each method and dividing by thirteen, thus giving the average numerical
ranking for each method.  The results of consensus opinion of the panel
regarding the relative ranking of the eleven methods for each pollutant gas
are summarized in Table 4-13.    The method that received the highest
numerical number was, in the opinion of the panel, the recommended method.
                                  145

-------
TABLE 4-7.  COMPARATIVE RANKING OF METHODS  FOR
Criteria
Simplicity of Method
Equipment/Processing
Requirements
Operator Skill
Requirements
Portability
Cost
Development
Apparatus
Per Test
Safety
Range
Blind Mode
Accuracy
Ease of Implementation
Confidence of Success
Maintenance
Operational Capability
1



•












2
















3
















4
















5
















6
















7
















8
















9















10


•• in
— —™™»^— »

™™— ««BMl
"
•^"••""•^^W

• .—


11


~
	
"
_
™ •
"

'
• ••
—
                     146

-------
         TABLE 4-9.  COMPARATIVE RANKING OF METHODS FOR NO,
Criteria
Simplicity of Method
Equipment /Process ing
Requirements
Operator Skill
Requirements
Portability
Cost
Development
Apparatus
Per Test
Safety
Range
Blind Mode
Accuracy
Ease of Implementation
i
Confidence of Success
Maintenance
Operational Capability
1
B
C
C
C

A
C
B
A
A
B
A
A
B
A
A
2
B
B
B
A

C
B
A
A
B
B
B
B
B
A
C
3
A
B
A
A

C
B
A
A
A
A
B
B
A
A
A
4
C
C
C
B

C
C
A
B
B
B
B
C
C
B
A
5
B
B
B
B

B
B
B
A
A
B
A
B
B
A
A
6
B
C
C
B

B
B
A
A
B
A
B
C
C
B
C
7
C
C
C
C

C
C
B
B
B
B
B
C
B
B
A
8
C
C
C
C

C
B
A
C
A
B
C
C
C
B
A
9
B
B
B
A

C
B
B
A
A
B
B
B
B
B
A
10
A
C
B
A

B
C
B
C
A
A
C
C
C
C
A
11
-
-
-
-

-
-
-
-
-
-
-
-
-
- .
"™
- Method not recommended
                                   148

-------
           TABLE 4-8.   COMPARATIVE RANKING OF METHODS  FOR SO,
Criteria
Simplicity of Method
Equipment /Process ing
Requirements
Operator Skill
Requirements
Portability
Cost
Development
Apparatus
Per Test
Safety
Range
Blind Mode
Accuracy
Ease of Implementation
Confidence of Success
Maintenance
Operational Capability
1
B
B
B
B

A
B
B
A
A
B
A
A
A
A
A
2
B
B
B
A

C
B
A
A
B
A
B
B
A
A
C
3
A
B
A
A

C
B
A
A
A
A
B
B
A
A
A
4
C
C
C
B

C
B
A
C
A
B
B
C
C
B
A
5
A
B
B
B

C
C
B
A
A
A
A
B
B
A
A
6
B
C
C
B

C
B
A
A
B
A
B
C
C
B
C
7
C
C
B
C

C
C
A
B
B
A
B
C
C
B
C
8
C
C
B
B

C
B
A
B
B
B
B
C
B
B
A
9
A
B
B
A

B
B
— ^^— ^— i
B
A
^™^^^^™»«
A
A
B
•• ••^
B
A
A
A
10
A
B
B
"^ .—
A

B
B
"
B
B
•
A
B
C
i
C
— ' •—
C
.
B
A
11
-
-
11 •...

T
— ••^•^^
-• 1 1.
-
	
•*™W»»»^_
•' ....
-
— Method not recommended
                                  147

-------
TABLE 4-10.  COMPARATIVE RANKING OF METHODS FOR CO
Criteria
Simplicity of Method
Equipment /Process ing
Requirements
Operator Skill
Requirements
Portability
Cost
Development
Apparatus
Per Test
Safety
Range
Blind Mode
Accuracy
Ease of Implementation
Confidence of Success
Maintenance
Operational Capability
1
A
A
A
C

A
B
A
A
B
B
A
A
A
A
A
2
B
B
B
A

C
B
B
A
B
B
B
B
B
A
C
3
A
B
A
A

C
B
A
A
A
A
B
B
A
A
A
4
C
C
C
B

C
B
B
B
B
B
B
C
C
B
A
5
A
B
B
B

C
B
A
A
A
A
A
B
B
A
A
6
B
C
B
B

C
B
A
A
B
A
B
C
B
B
C
7
C
C
C
C

C
C
B
B
B
B
C
C
C
C
A
8
C
C
C
B

C
B
A
B
B
B
C
C
B
B
A
9
A
B
B
A

B
B
B
A
A
B
B
B
A
B
A
10
B
B
B
B

B
C
B
B
A
B
C
C
C
B
A
11
C
C
C
C

C
C
C
A
A
A
C
C
C
C
A
                           149

-------
         TABLE 4-11.  COMPARATIVE RANKING OF METHODS FOR 0,
Criteria
Simplicity of Method
Equipment /Process ing
Requirements
Operator Skill
Requirements
Portability
Cost
Development
Apparatus
Per Test
Safety
Range
Blind Mode
Accuracy
Ease of Implementation
Confidence of Success
Maintenance
Operational Capability
1
B
B
B
C

A
B
B
A
A
A
B
A
A
A
A
2
C
C
C
C

C
C
B
B
B
A
C
C
C
B
C
3
A
B
A
B

B
B
B
A
A
A
C
C
C
A
A
A
C
C
C
C

C
C
B
B
B
B
B
C
B
B
A
5
-
-
-
-

-
-
-
-
-•
-
-
-
-
-
-
6
B
B
B
C

A
B
B
A
A
B
A
A
A
A
A
7
C
C
C
C

C
C
B
B
A
B
C
B
B
B
A
8
-
-
-
-

- .
-
-
-
-
-
-
-

-
-
9
-
-
-
-

-
-
-
-
-
-
-
-
-
-
-
10
-
- •
-
-

-
-
-
-
-
- •
-iii •.
•*"•— — ^— .
-

11
^
-
-
—

-
-
-
-
-
-
•• • —
• i-
™— ^^••^
—  Method not recommended
                                 150

-------
         TABLE  4-12.   COMPARATIVE  RANKING  OF METHODS  FOR  1-BUTENE
Criteria
Simplicity of Method
Equipment /Process ing
Requirements
Operator Skill
Requirements
Portability
Cost
Development
Apparatus
Per Test
Safety
Range
Blind Mode
Accuracy
Ease of Implementation
Confidence of Success
Maintenance
Operational Capability
1
A
B
B
B

A
B
A
A
B
B
A
A
A
A
A
2
A
B
B
A

B
B
A
A.
B
A
B
B
B
A
C
3
A
A
A
A

B
B
A
A
B
A
C
B
A
A
A
4
B
C
B
B

C
B
B
A
B
B
B
C
C
B
A
5
B
B
B
B

C
C
B
A
A
A
A
B
C
A
A
6
B
B
B
B

B
C
B
B
B
A
B
C
C
B
C
7
-
-
-
-

-
-
-
-
-
-
-
-
-
-•
-
8
C
C
C
B

C
B
A
C
B
B
C
C
B
B
A
9
A
il
*J
B
A

B
B
B
A
A
B
B
B
A
B
A
10
B
C
C
A

C
C
B
B
A
A
C
C
C
C
A
11
C
C
C
C

C
C
B
A
B
A
C
C
C
C
A
— Method not recommended
                                  151

-------
TABLE 4-13.  SUMMARY OF COMPARATIVE EVALUATION
Method
1
2
3
4
5
6
7
8
9
10
11

S02
3.5
3.3
3.5
2.8
3.4
3.0
2.7
2.9
3.4
3.1
— —
N02
3.3
3.2
3.6
2.8
3.4
3.0
2.7
2.7
3.3
3.0
—
Pollutant
CO
3.4
3.2
3.5
2.7
3.4
3.0
2.6
2.9
3.4
3.0
2.6
o3
3.6
2.6
3.2
2.7
	
3.6
2.9
	
	
	
___
C4H8
3.4
3.3
3.4
2.8
3.3
3.0
—
2.9
3.3
3.1
2*. 8
                       152

-------
5.0  RECOMMENDATIONS
     The comparative evaluation described in the preceding section focuses
attention on certain methods for each pollutant that obtained a high
ranking.  In particular, for S02, N02, CO, and C4Hg, the effusion, electro-
chemical drive, and novel permeation methods rate very high.  This
illustrates some of the weaknesses of both the overall organization of the
methods and the ranking system.  For those cases where the electrochemical
source rated highest—SO,,, CO, and C^—the method utilizes electro-
chemical generation to force gas from a container at a known rate through
a capillary.  For those cases where effusion rates high—S02, CO, and C,Hg—
a gas mixture is maintained in a storage volume and dispensed through a
capillary or valve.  These methods are very similar differing only in
detail.  This similarity leads to a basic recommendation as follows:
     Development effort directed towards the provision of S02, CO,
     and C^Hg test atmosphere generators should be focused on low
     pressure storage cans to ship and store prepared mixtures of
     the required pollutants.  Three dispensing mechanisms—a Freon
     based system, electrolytic drive, or a valved method are
     competitive.  The freon based system is recommended due to its
     simplicity.
Primary problems which must be looked into during the development of such a
method are the stability of the stored gases, the materials to be employed,
the detailed procedures which are necessary, and the various sources of
error.  The effusion method is not recommended for nitrogen dioxide at this
time due to its questionable stability of moderate concentrations.  Thus,
the gas phase titration and novel permeation methods are the only other two
choices for generating concentrations of nitrogen dioxide.  Of these, the
gas phase titration system appears to be the most attractive, based on
anticipated development costs and laboratory-proven reliability.
     With respect to the one remaining pollutant, ozone, the UV generator
modified for more reliable and easier operation is almost the only method
which survives critical examination.  The plasma discharge source received
a fairly high ranking but appears to present no distinct advantages when
compared to the UV generator.  Since the UV ozone generator is an integral
                                   153

-------
part of the gas phase titration system, the following recommendation is

made:

     Development effort directed toward the provision of test
     atmosphere generators for ozone and nitrogen dioxide should be
     combined to incorporate an "easy-to-use" UV ozone generator
     into the required apparatus to yield a single device capable
     of producing test atmospheres for both ozone and nitrogen
     dioxide.

     It is noteworthy that certain attributes of the novel permeation method
are attractive as evidenced by its consistently high ranking for all pollu-
tants other than ozone.  The fact that it ranked behind some of the other
methods stems from the more complex nature of the pollutant source.  As a
calibration source it would have received a higher ranking.  From an overall

standpoint of having the capability for both calibration and quality control,
the novel permeation method is attractive.
                                  154

-------
                                  TECHNICAL REPORT DATA
                           /Please read /uuructi»nf on tlic reverse before completing)
 PEPCRT NC.
    EPA-650/4-74-ni6
                            2.
                                                          3. RECIPIENT'S ACCESSION-NO.
 TITLE AND SUBTITLI:
    "Concepts  for Development of Field Usable Test
    Atmosphere Generating Devices."
             5. REPORT DATE
              December 1973
            6. PERFORMING ORGANIZATION CODE
 AUTHOR(S)

    N/A
                                                          8. PERFORMING ORGANIZATION REPORT NO.
 PERFORMING ORGANIZATION NAME AND ADDRESS
    Research  Triangle Institute
    Research  Triangle Park, N. C. 27709
                                                          10. PROGRAM ELEMENT NO.
             11. CONTRACT/GRANT NO.
                                                           68-02-1242
12. SPONSORING AGENCY NAME AND ADDRESS
    Environmental Protection Agency, NERC,  Analytical
    Quality Control Section, Quality Control  Branch,
    Quality Assurance and Environmental  Monitoring Lab
    Research  Triangle Park, N. C. 27711
             13. TYPE OF REPORT AND PERIOD COVERED
              7/1  - 12/30/73    Final
             14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
      The puroose of this project was  to  examine possible physical, physical  plus
 chemical,  and biological concepts  applicable to the development of field  usable  test
 atmosphere generating devices.  The  primary activity for application of these  devices
 is  the round-robin survey to assess  instrument/operator performance on a  routine
 periodic basis.  Ten concepts  (i.e.,  desorption, effusion, thin film evaoorators,  novel
 permeation, radiolysis-photolysis,  plasma discharge, thin films of dissolved reactants,
 electro-chemical films of laminated  reactants, and biological generation) were examined
 for each of the five required  pollutants (i.e., sulfur dioxide, nitrogen  dioxide,
 carbon monoxide, ozone, and 1-butene)  and a total of fifty individual discussions  pre-
 oared.  The program was divided into three distinct phases:  1) comprehensive  litera-
 ture search and preparation of technical discussions; 2) panel review of  50  discussions
 and 3) comparative evaluation.  Of the concepts investigated, the effusion method  is
 recommended for generating test atmosoheres for sulfur dioxide, carbon monoxide,  and
 1-butene, while for ozone, UV  generation modified for more reliable and easier
 ooeration is recommended.  Gas phase titration of nitric oxide with ozone modified for
 field use is recommended for generating test atmospheres of nitrogen dioxide.
      This report was submitted  in fulfillment of Contract No.
 Triangle Institute  under  the  sponsorship of the Environmental
     completed as of December  1973.	
                   DU68-02-1242  by  Research
                   Protection Agency.  Work
 7.
                               KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSAT; I idd (".mup
19. DISTi-HbUTION.STATbMENT
 Unlimited
19. SECURITY CLASS < Tins Krpurtf
   Unclassified
                                                                         21. NO. -'JF PAGES
                                                                           163
20. SECURITY CLASS (Tlii

   Unclassified
                                                                         22. PRICE
EPA Form 2220-1 (9-73)
                                            155

-------