EPA-650/4-75-007
February 1975
Environmental Monitoring Series
-------
EPA-650/4-75-007
SUMMARY REPORT:
WORKSHOP ON OZONE MEASUREMENT
BY THE POTASSIUM
IODIDE METHOD
by
John B. Clements
Quality Assurance and Environmental Monitoring Laboratory
ROAP No. 26AAG
Program Element No. 1HA327
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, N.C. 27711
-------
EPA REVIEW NOTICE
This report has been reviewed by the National Environmental Research
Center - Research Triangle Park, Office of Research and Development,
EPA, and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environ-
mental Protection Agency, have been grouped into series. These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These series are.
1. ENVIRONMENTAL HEALTH EFFECTS RESEARCH
2. ENVIRONMENTAL PROTECTION TECHNOLOGY
3. ECOLOGICAL RESEARCH
U ENVIRONMENTAL MONITORING
5. SOCIOECONOMIC ENVIRONMENTAL STUDIES
6 SCIENTIFIC AND TECHNICAL ASSESSMENT REPORTS
9. MISCELLANEOUS
This report has been assigned to the ENVIRONMENTAL MONITORING
series. This series describes research conducted to develop new or
improved methods and instrumentation for the identification and quantifica-
tion of environmental pollutants at the lowest conceivably significant
concentrations. It also includes studies to determine the ambient concentra-
tions of pollutants in the environment and/or the variance of pollutants
as a function of time or meteorological factors
Th"- document is available to the public for sale through the National
Technical Information Service, Springfield, Virginia 22161.
Publication No EPA-650/4-75-007
-------
CONTENTS
Page
ACKNOWLEDGMENTS iv
INTRODUCTION 1
WORKSHOP RECOMMENDATIONS 3
WORKSHOP PARTICIPANTS 4
DISCUSSIONS 5
A. Experiences with the Neutral Potassium Iodide Method . . 5
B. Evaluation of the Neutral Potassium Iodide Method ... 11
C. Absolute Measurement of Ozone 19
D. Proposals and Recommendations 26
APPENDIX A 29
TECHNICAL REPORT DATA SHEET 32
m
-------
ACKNOWLEDGMENTS
Parti-cular appreciation is expressed to Mr. E.E. Hunt of The
National Bureau of Standards for making the arrangements to host
the workshop. Especial acknowledgment is also expressed to Dr.
J.A. Hodgeson for the original suggestion to hold such a workshop,
iv
-------
SUMMARY REPORT
WORKSHOP ON OZONE MEASUREMENT
BY THE POTASSIUM IODIDE METHOD
INTRODUCTION
The Federal Reference Method for photochemical oxidants specified
in regulations on National Primary and Secondary Ambient Air Quality
Standards is based on the chemiluminescence resulting from the
reaction of ozone with ethylene. The method is calibrated by measuring
synthetically prepared standard atmospheres of ozone and developing a
method response versus ozone concentration curve. The concentration
of ozone in the standard atmospheres used for calibration is determined
by using the one percent neutral buffered potassium iodide procedure as
specified in the regulations appearing in the Federal Register
36(228):22384-22397, November 25, 1971.
Although the one percent neutral buffered potassium iodide
procedure has been in use for quite some time, it has been widely
criticized for its inconsistent and non-reproducible results. These
criticisms are of considerable concern to EPA because they call into
1
-------
question the methodological basis for one of its regulatory standards;
likewise, NBS is concerned because there seems to be a
national need associated with an important environmental measurement.
With, these concerns in mind, a workshop on the problem was held
on August 26 and 27, 1974, at the National Bureau of Standards (NBS)
facility in Gaithersburg, Maryland, to bring together individuals with
expertise in the use of the potassium iodide method for a full technical
discussion of this ozone method calibration problem.
The purpose of the workshop was for the participants to exchange
information and relate experiences on the problem and to propose
recommendations on future actions. This report presents these
recommendations and a summary of the discussions held.
-------
WORKSHOP RECOMMENDATIONS
• EPA and others should continue research and development on the
ultraviolet photometry method for measuring ozone.
• EPA and others should continue work on potassium iodide
methodology for measuring ozone.
• NBS should develop a certified ozone standard.
• EPA and others should investigate other promising methods for
measuring ozone.
-------
WORKSHOP PARTICIPANTS
National Bureau of Standards
Radford Byerly, Jr. William P. Schmidt
Ernest E. Hughes John K. Taylor
James R. McNesby
Environmental Protection Agency
Michael E. Beard John H. Margeson
Thomas A. Clark William A. McClenny
John B. Clements Larry J. Purdue
Jimmie A. Hodgeson Kenneth A. Rehme
Robert K. Stevens Karl J. Zobel
National Oceanic and Atmospheric Administration
Walter D. Komhyr
California Department of Health
Yoshiro Tokiwa
California Air Resources Board
Kiyoshi Nishikawa
Los Angeles County Air Pollution Control District
Margaret Brunell
California Institute of Technology
William B. DeMore Sigmimd Jaffe
National Aeronautics and Space Administration
Harry M. Finley J. David Kuder
(Old Dominion University)
-------
DISCUSSIONS
The discussions were organized into four topic areas
as follows: (1) Experiences with the Neutral Potassium Iodide Method;
(2) Evaluation of the Neutral Potassium Iodide Method; (3) Absolute
Measurement of Ozone; and (4) Proposals and Recommendations. Selected
participants made a presentation on some aspect of the problem, then
encouraged a full exchange of information. A brief summary of each
presentation is given below and wherever appropriate salient points
of discussion are recounted.
A. Experiences with the Neutral Potassium Iodide Method
1. Hodgeson: Dr. Hodgeson reviewed the important literature
on the use of iodometry for the measurement of ozone concentrations
(The literature references presented by Dr. Hodgeson are given as
Appendix A of this report). He pointed out that the use of one percent
neutral buffered potassium iodide is specified as the means to calibrate
the Federal Reference Method for oxidants and is an integral part of
the method. The potassium iodide calibration is & weak link as is shown
by the poor comparison between measurements of ozone made by the one
percent neutral potassium iodide procedure and by ultraviolet photo-
metry and as shown by the inability of various workers to get con-
sistent results.
2. Margeson: Mr. Margeson reported on the results of an inter-
laboratory collaborative test of the Federal Reference Method for
oxidants. The test was carried out in two parts with the first part
-------
designed to test the method's precision and the second part designed
to test the method's bias.
In the test of the method's precision, a group of nine qualified
participants were gathered together to measure the same ambient air
at a single location (Pasadena, California) for a four-day period.
The test results showed that the method has fairly good precision, in
the range of 15-203.
Tn test the method's bias, a number of ozone generators, which
had been carefully calibrated by the National Bureau of Standards,
were sent to test participants for them to make measurements of
atmospheres prepared by these generators in their own laboratories.
The test was designed so that the participants did not know the ozone
concentrations they were generating. At the end of the test, the
generators were returned to NBS for recalibration.
Nine participants made measurements at five levels of ozone;
all measurements made by the participants were less than the
corresponding calibration value, thus indicating a negative bias
some place in the method. The following table summarizes the test
results:
NBS Calibration Value (ppm) 0.05 0.08 0.13 0.28 0.50
Average Collaborator Bias -37% -31% -26% -17% -16%
Mr. Komhyr pointed out that the output from an ozone generator
is pressure dependent. The pressure dependency is not necessarily
linear over wide pressure ranges and this would affect calibration,
values at higher elevations, for example, Denver, Colorado.
6
-------
3. Beard: The apparent bias in the Federal Reference Method
led to an investigation by NBS and EPA into its causes. The
potassium iodide calibration was immediately suspected and Mr. Beard
reported on experiments which seemed to implicate the impinger used
to contain the potassium iodide as one important source of this bias.
NBS and EPA teams, using their own reagents and apparatus to sample
a common ozone atmosphere, have shown that the set of impingers used
at NBS for calibrating generators used in the collaborative test give
consistently higher ozone values than do other impingers. Mr. Beard
described an experiment in which a constant ozone atmosphere of about
0.2 ppm was established, and the bubbler-type impingers specified in the
Federal Reference Method were used by EPA for sampling and midget
impingers were used by NBS for sampling. There was also a feature in
the experiment in which the EPA worker used the midget impinger for
sampling, but used his own reagents and spectrophotometer. The follow-
ing results were obtained:
Impinger Design Ozone Found (ppm) Analyst
Federal Reference Method 0.201 EPA
0.203 EPA
0.204 EPA
NBS Midget Impinger 0.230 NBS
0.236 NBS
0.234 EPA
0.241 EPA
-------
Mr. Beard also reported an investigation in which various
levels of ozone, up to about 1 ppm, were measured using bubbler
impingers and midget impingers at each level. These experiments
indicated that the midget impingers find approximately 1.2 times
more ozone than do the bubbler-type impingers.
4. Hughes: Mr. Hughes has studied a number of midget impinger
pairs and has found one particular set, designated 1 and 1A, which
consistently gives higher ozone values than does any other set. Some
idea of the differences is shown by comparing this set with several
other sets and assuming a collection efficiency of 100% for 1 and 1A:
Impinger Designation Collection Efficiency
1 and 1A 100%
3 and 3A Ca. 90%
4 and 4A Ca. 95%
5 and 5A Ca. 95%
Mr. Hughes has studied a large number of factors, including
impinger design features and other aspects of the method, which might
cause the high results found for 1 and 1A. The following factors were
studied:
Interchanging impinger parts; varying the orifice diameter; varying
the distance from the tip of the orifice to the bottom of the container;
Doubling the solution volume from 10 ml to 20 ml; varying flow rate
from 0.6 1/min to 2 1/min; investigating iodine volatilization (by using
a modified Bergshoeff procedure); measuring absorbance immediately at the
end of the sampling and 30 minutes after end of sampling; increasing pH from
8
-------
6.83 to 8.86; varying the age of potassium iodide solution; and using
triple distilled water. Unfortunately, none of the variables studied had
a significant effect, and the reason for the bias is still not evident.
Dr. Hodgeson suggested that iodate formation caused by the catalytic
effect of surfaces at neutral pH may explain some of the differences
between impingers (see Parry and Hern, Reference 10, Appendix A).
5. Tokiwa: The Air and Industrial Hygiene Laboratory has compared
its potassium iodide ozone calibration procedure (which is also the one
used by the California Air Resources Board) with the EPA procedure. A
comparison of the two procedures follows:
EPA AIHL
Absorber 2 bubblers 2 midget impingers
Absorbing solution 1% KI: pH = 7 2% KI: pH = 7
Sampling rate 0.2-1.0 1/min. 1 1/min.
Sampling time 10 min. 5 or 10 min.
Analysis same same
It is generally agreed that the biggest difference is the concen-
tration of the potassium iodide.
Mr. Tokiwa reported that they had compared the two procedures
experimentally and find very little difference between them. He also
reported experiments in which the sampling time, collection efficiency,
and period to maximum color formation were studied. There were also
some experiments which compared impingers and frits when each system
-------
was used with two devices in series. The following tables present
the results:
Sampling time, Nominal Collection Efficiency, %
minutes 03 Cone., ppm Impinger 1 Frit 1
10 0.14 97-100 97-100
10 0.52 95 95
30 0.14 90-93 89
% Color Developed at Indicated Times
1 min. 5 rrn'n. 15 min. 20 min.
10 0.14 - 94-97 100
10 0.52 91-93 93-95 - 100
Mr. Tokiwa recommends that two impingers be used and that the analyst
wait 15 minutes for color development.
Mr. Hughes commented that in his studies, in which he uses one percent
neutral potassium iodide, approximately 10% of the ozone is collected in
the second impinger. Mr. Tokiwa said that their 2% potassium iodide
system characteristically collects at least 95% of the ozone in the first
impinger. There was general agreement that the higher potassium iodide
concentration in the AIHL system is probably the basis for its better
collection efficiency.
Mr. Tokiwa also reported that they always check their potassium
iodide for the presence of a reducing agent that is added to some grades
in order to keep them white. This additive, if not accounted for, causes
an unknown iodine demand and leads to erroneous results. They have found
10
-------
that USP grade potassium iodide does not contain this additive and
gives satisfactory results in their hands.
6. Brunelle: Mrs. Brunelle recounted the history of monitoring
for oxidants carried on by the Los Angeles County Air Pollution Control
District, pointing out that the District had used a monitoring system
based on the reaction of oxidants with unbuffered potassium iodide
reagent. For calibration, they investigated buffered and unbuffered
20% potassium iodide reagent, but found high blank values. They also
studied 2% unbuffered potassium iodide and found no blank problems and
also found that the end point in the thiosulfate titration they use
is easier to determine in the 2% system. All of these reasons led
them to select the 2% unbuffered potassium iodide system for calibration
as their standard procedure
Mrs. Brunelle stated that an absolute reference method or standard
material for ozone is urgently needed—one that would permit the
intercomparison of the various ozone measuring techniques. This view
was supported strongly by other members of the conference.
E. Evaluation of the Neutral Potassium Iodide Method
1. Hodgeson: Dr. Hodgeson reported on work carried out in his
laboratory in which various factors influencing the absorbance values
of the iodide-iodine collection reagent were investigated. He also
reported on investigations in which ozone concentrations measured by
the 1% neutral buffered potassium iodide procedure were compared with
ozone measured by ultraviolet photometry. A number of graphs presented
11
-------
by Dr. Hodgeson are reproduced in Figures 1 through 4 in a schematic
and non-quantitative fashion in an effort to summarize the major
findings.
Figure 1 presents the decay of absorbance of standard iodide-
iodine solutions with time^implying that color stability is a problem.
Figure 2 presents the change in absorbance of ozonized potassium
iodide solutions as a function of time. Mr. Tokiwa and Mr. Hughes
both remarked that they have observed the same phenomenon, and Dr. Jaffe
said that he has also observed that the decay in absorbance continues
at times beyond 60 minutes.
Figure 3 presents the change in absorbance for potassium iodide
solutions as a function of the potassium iodide concentration. Two
levels of ozone were examined.
Figure 4, which shows the ratio of ozone concentration as measured by
potassium iodide and by the Dasibi instrument as a function of sampling
time, also shows the ratio of the color in the second sampling
impinger (Ag) and the total color in both impingers (At) as a function
of sampling time. In these experiments, the ozone concentration was
0.6 ppm and sample flow rate was 0.5 A/min. Apparently, volatilization
of iodine is important as is shown by the decrease in the [03] la/COaLy
ratio and the increase in the Ag/Aj. ratio, particularly after 30 minutes
of sampling.
12
-------
A study of impinger components showed that there seems to be very
little effect on the change in ratio of ozone measured by potassium
iodide and by ultraviolet photometry with change in impinger design.
The following relationships were developed:
(1) Wide impinger orifice, poor impinger action
[03]KI = (1.070 + 0.021) [03]uv - 0.017
(2) Narrow impinger orifice, good impinger action
[03]KI = (1.087 + 0.018) [03] uv -- 0.017
The similarity of the slopes and intercepts of the two equations
shows that the features studied have little effect.
0.51
2d
60
30 40 5C
TliVIE, min.
Figure 1. Change in absorbance of iodide - iodine solution with time.
13
-------
0.26
0.25
0.24
0.23
<
DO
1 0-07
CD
0.06
0.05
0.04
0.03
r T T
0
10
50
20 30 40
TIME.min
Figure 2. Change in absorbance of ozonized solution with time.
60
14
0 2 4 G
Kl CONCENTRATION. %
Figure 3. Change in absorbance with Kl Concentration at constant
ozone concentration.
-------
1.22
[03]K, 1.10
10
20 30 40
SAMPLING TIME, min.
Figure 4. Comparison of ozone by Kl and by UV and comparison of
absorbance as a function of sampling time.
2. Jaffe: Dr. Jaffe reported experiments designed to compare
ozone measured by iodometry and by ultraviolet photometry. He
described his apparatus, which contained a photometer and a Dasibi
ozone monitor and which allowed the concurrent measurement of
ozone by ultraviolet photometry and by the potassium iodide procedure.
15
-------
The ozone concentrations studied range from 1 or 2 parts per
million to 15 or 20 parts per million. Figures 5 and 6 portray
in a schematic and non-quantitative fashion the important findings
reported by Dr. Jaffe.
Figure 5 shows a comparison of the ozone concentration as
measured by ultraviolet photometry and by potassium iodide. Figure
6 shows the relationship of the ratio of ozone measured by KI (as
shown by concentration of I3~ measured) and by UV as a function of
concentration. This experiment is evidence that the stoichiometry
of the reaction of ozone with potassium iodide changes from one to
one to a higher value, and the change seems to occur at about 2 ppm
ozone.
Dr. Jaffe offered the opinion that we could reasonably look
to a standard for ozone based on ultraviolet photometry.
16
10 20 30
KI OZONE MEASUREMENT, moles xlO8
Figure 5. Ozone by UV and KI.
40
-------
1.5
1.0
uv
0.5
10 20 3C 40 50
1
GO
Figure 6. Ozone by Kl and by UV as a function of ozone concen-
trations.
3. Rehme: Mr. Rehme reported on research in which the rapid
and quantitative reaction of nitric oxide with ozone is the basis for
making ozone measurements. The reaction of interest is:
NO + 0,
NO,
''J ^* llwrt ' ^O
This reaction is used to calibrate an ozone generator in the
following way:
A level of NO, generally about 1 ppm, is established using a
cylinder of nitric oxide, generally about 100 ppm, which has previously
been analyzed. The NO level is monitored continuously. Ozone from a
stable source is added in increments and the decrease in NO response
corresponds, according to the above equation, to the amount of ozone
added. A calibration curve can then be constructed for the ozone source,
and it then serves as a device which can produce, on demand, atmospheres
17
-------
containing known concentrations of ozone. The term gas phase
titration is used to describe the process.
Until recently a weakness in the procedure has been the measure-
ment of the nitric oxide concentration in the cylinder using gas
phase titration in a configuration somewhat as described above. In
this configuration known levels of ozone are reacted with NO and the
results used to determine the NO concentration in the cylinder. The
known levels of ozone are based on iodometry which is where the weak-
ness resides.
Mr. Rehme described one approach that has been used to obviate
this problem. The NO level of the cylinder has been related to the
output from a N02 permeation device which has been gravimetrically
calibrated. A NO monitor, in which N0? can be quantitatively con-
/k £
verted to NO, is required. Mr. Rehme reported that agreement of 3%
or better can be obtained between NO assays based on the ozone-nitric
oxide gas phase titration and NO assays based on the gravimetrically
calibrated N02 permeation device.
4. Stevens: Mr. Stevens described experiments currently under way
in his laboratory which will compare the 1% neutral buffered potassium
iodide procedure (the EPA Reference Method procedure), the 2% neutral
buffered potassium iodide procedure (the California Air Resources Board
procedure), and the 2% unbuffered potassium iodide procedure (the Los
Angeles County Air Pollution Control District procedure) by having
them all sample the same ozone atmospheres simultaneously. Several
18
-------
levels will be studied and at each level an Independent assay of the
ozone concentration will be made by gas phase titration and, he
hopes, by ultraviolet photometry.
Mr. Stevens emphasized again the acute need for an ozone
standard of some sort. In this context, Dr. Taylor of NBS expressed
confidence in gas phase titration as an approach.
C. Absolute Measurement of Ozone
1. DeMore: In their studies on reaction mechanisms, Dr. DeMore
pointed out that they have a need to know fundamental constants and in
this connection have done considerable work on the extinction co-
efficient of ozone. The coefficient has been determined for ozone at
various pressures down to a few torr and does not seem to have any
dependency on pressure. This leads to the conclusion that the
extinction coefficient at ppm levels should be the same, and Dr. DeMore
described experiments in which optical density was measured under the
conditions of a few torr and under the conditions of a few ppm ozone.
The ratios of the corresponding extinction coefficients are constant
as the following table shows:
Experiment No.
1
2
3
4
5
OD torr
0.29
0.53
0.81
1.31
1.49
OD ppm
0.0325
0.062
0.095
0.165
0.180
E torr/Eppm
0.97
1.01
1.01
1.04
1.06
19
-------
Dr. DeMore does not believe the ratio is showing a trend.
The ozone level for the first experiment is about 3 ppm; the others
are higher levels of ozone.
Dr. DeMore described experiments in which ozone measurements
made by a laboratory photometer were compared to ozone measurements
made by a Dasibi instrument. His experiments indicate that Dasibi
measurements are about 3.5% lower than the photometer measurements
at the ozone concentrations studied.
Dr. DeMore also noted that the temperature in a Dasibi instrument
runs about 10-1 5°C higher than ambient and said that this temperature
difference should be taken into account when comparing ozone measure-
ments made by a Dasibi and a laboratory photometer.
2. Komhyr: Mr. Komhyr's agency had a need to make ozone measure-
ments into the stratosphere and at clean air locations on the earth's
surface. For this purpose he developed a device, called the electro-
chemical concentration cell (ECC) ozone sensor, which is based on the
electrochemical oxidation of iodide ion to tri-iodide ion and the
measurement of the current produced. The appropriate reactions are:
31" _ } I3~ + 2e Oxidation
I. + 2e _ ^ 21" Reduction
overall cell reaction is:
31" + I - I~ + 21
20
-------
No external e.m.f. is applied to drive the cell since the highly
different electrolyte concentrations in the cell's cathode and
anode chambers (approximately 0.1 molal and 8.0 molal, respectively)
give rise to a spontaneous driving e.m.f. defined at 25 C by:
E : - 0.0591 log
--
where the a's are ionic activities.
The stoichiometric properties of the cell are dependent to a small
extent on the KI concentration of the cathode electrolyte. A
doubling of the electrolyte KI concentration, for example, increases
measured ozone amounts by 5%. Mr. Komhyr described tests in which
atmospheric total ozone amounts derived from ECC ozonesonde ascents
into the stratosphere were compared with atmospheric total ozone
amounts derived quasi-simultaneously by spectroscopic means using
Dobson ozone spectrophotometers. The tests showed agreement in
measured total ozone amounts by the two independent methods when 1.5%
KI electrolyte solutions were employed in the cathode chambers of the
ECC cells.
Some data were presented on the reproducibility of the cell's
performance over a lonq period of time. In view of the simplicity
in preparing for use and using the ECC ozone sensors, Mr. Komhyr
suggested that the devices may be usefully employed in periodic
checking of calibrated ozone generators.
21
-------
3. Hodgeson: Dr. Hodgeson described experiments carried out
in his laboratory, which related ozone concentrations as measured by
iodometry and by ultraviolet photometry. Figure 7 shows the relation-
ship of ozone as measured by neutral potassium iodide and by ultra-
violet photometry using a one-meter path-length cell. Measurements
were made between 1 and 12 ppm, and the ratio of ozone concentrations
measured by potassium iodide to the ozone concentration measured by
photometry approaches unity below 1 ppm. The regression equation for
the data is:
[03]Ki = 1.093 + 0.006 [03]uv - 0.035 + 0.021
Figure 8 shows the relationship of ozone as measured by a Dasibi
instrument and by a photometer. Concentrations determined by the
Dasibi instrument are absolute values based on a logarithmic form of
Beer's Law and an assumed cell path length of 71 cm. The regression
equation for the data is:
(Moasibi = °'954 t °-003 ^Photometer ' °'027 ± 0'008
Dr. Hodgeson described his use of NBS standard spectrophotometer
filters as a means for checking spectrophotometer calibration. This
has proven to be a very simple calibration check which may be used
to supplement the tedious chemical calibration procedure. The stability
which can be achieved with a DU-2 spectrophotometer is illustrated by
the data of Table 1. Dr. Hodgeson suggested that the determination of
the absolute absorption coefficient of iodine in 1% potassium iodide
combined with the use of NBS standard filters has the potential to
replace conventional wet chemical calibration procedures.
22
-------
2468
(03] PHOTOMETER, ppm
Figure 7. Comparative ozone measurements by neutral Kl and
UV photometry.
12
12345
(03! PHOTOMETER, ppm
Figured. Comparative ozone measurements by 1-meter photometer
and by Dasibi instrument
23
-------
Table 1. SPECTROPHOTOMETER CHECK WITH NBS STANDARD FILTERS
Filter No. 3-163
Date
1-29-74
2-6-74
2-12-74
5-7-74
8-8-74
A, nm
440
465
590
635
440
465
590
635
440
465
590
635
440
465
590
635
440
465
590
635
Labeled Abs.
0.512
0.467
0.516
0.502
0.512
0.467
0.516
0.502
0.512
0.467
0.516
0.502
0.512
0.467
0.516
0.502
0.512
0.467
0.516
0.502
Meas. Abs.
0.513
0.468
0.514
0.499
0.512
0.467
0.513
0.499
0.505
0.458
0.506
0.492
0.510
0.463
0.511
0.492
0.516
0.470
0.518
0.498
A, %
-0.20
-0.21
+0.39
+0.60
0
0
+0.58
+0.60
+1.37
+1.93
+1.94
+0.60
+0.39
+0.86
+0.97
+0.80
-0.78
-0.64
-0.39
+0.80
4. McClenny: Dr. McClenny described experiments by himself and
co-workers to make absolute measurements of ozone using ultraviolet
photometry. The goal of these experiments is to provide a simple, in-
expensive t'V photometer for use with an ozone generator. Such a
combination would provide an alternative to the KI technique, if one
is needed.
24
-------
Dr. McClenny emphasized the need to eliminate scattered light
in the photometer systems. Two UV photometers using folded optical
paths were discussed, one employing a double-pass cell and the
other a multipass "white" cell. Since the number of passes can be
changed in the "white" cell, contributions to the output signal
from scattered light can be checked. Thus, an absolute photo-
metric measurement can be made by checking the agreement at more
than one total optical pathlength, i.e.,by ensuring internal
consistency of the measurement.
Measurements were presented to illustrate the sensitivity of
the double-pass photometer. Even though typical UV sources drift,
linear drift can be eliminated by proper processing of the signal.
To illustrate this, measurements over the ambient range 50-500 ppb
were presented showing standard deviations of less than 5 ppb.
5. Nishikawa: Mr. Nishikawa recounted the history of
oxidant monitoring in California. The first method was based on
the phenolphthalin procedure, but this was abandoned in 1952
because it did not really characterize air quality. The phenol-
phthalin procedure was replaced by continuous 20% potassium iodide
instruments, which were used until 1968 when they changed to 10%
neutral buffered potassium iodide.
With the advent of the National Ambient Air Quality Standards
in 1970, they were again faced with the problem of chosing new
monitors for oxidant for the 35 stations which would be required.
25
-------
At the time when a decision was required, there was only one
manufacturer of chemiluminescence instruments on the West Coast,
and California's choice had to be between it and the Dasibi
instrument. An evaluation of both instruments led to the finding
thtit the type of chemi luminescence instrumentation available to
them would be difficult to service and would require a logistically
troublesome tank of ethylene. The Dasibi instruments did not have these
disadvantages and they chose to purchase them for their network.
They now have 19 Dasibi instruments in operation. The California
Air Resources Board is very anxious to maintain its data base, so
that trends can be kept intact. For this reason, they are doing
parallel sampling at a number of their stations with the 10% neutral
buffered potassium iodide continuous instruments and the newly
installed Dasibi instruments. The calibration system for both
instruments is based on 2% neutral buffered potassium iodide. Mr.
Nishikawa presented preliminary data, which showed excellent
correlation between the Dasibi and 10% neutral buffered potassium
iodide measurements at two sites.
D. Proposals and Recommendations
1. Taylor: Dr. Taylor discussed the NBS policy requiring that
before they certify a reference material it must have been assayed by two
independent procedures. He then presented the concept of having the
NBS provide certified ozone generators and have this certification
be based on at least two independent standards. A recalibration
26
-------
service would be provided for the certified generators. In his
concept, the ultraviolet photometer would serve as one independent
measurement for certification of the ozone generator; gas phase
titration based on a nitric oxide standard, which is nearing
certification by NBS, would serve as the other independent measure-
ment. The concept as developed by Dr. Taylor can be shown
schematically as follows:
2 PERMEATION
DEVICE
STANDARD
N02 -*• NO
CONVERTER
NO
CHEMILUWiUESCENCE
MONITOR
NO
STANDARD
GAS PHASE
TITRATION
j
,
CHEMILUITNESCEHCE
MONITOR
* I
1 *
Kl MEASUREMENT
CLEAN AIR
SOURCE
'///S/Ss S
03 GENERATOR!
///////, '/
UV MEASUREMENT
Figure 9. NBS proposed scheme for development of an ozone
standard.
Dr. Hodgeson said he would be concerned that the generator
could be transported and still retain its integrity. Dr. Taylor
agreed that this is a prime consideration and said that the
ruggedness of the device would be thoroughly tested and that
methodology would be developed which would allow the recipient
to relate his generator to one of the standards.
27
-------
Other things suggested to be Investigated were: effects of
change in temperature and pressure from the point of calibration
to the point cf use, the quality of the air (particularly the
oxygen content)tand the need for voltage stabilization.
There was some sentiment that a ultraviolet photometer would
be a viable approach to an ozone standard material. Dr. DeMore
felt that it is reasonable to expect that a photometer, suitable
as a standard, could be developed for the ambient range.
Dr. McNesby seemed to be attracted to this idea. Dr. McClenny
expressed interest in developing such a photometer and in having
a number made for evaluation by EPA personnel.
Dr. Jaffe expressed the view that,while it is true that
iodometry may have a place in ozone measurement, we should abondon
the idea that potassium iodide is a satisfactory primary standard
for ozone measurement,and we should look to other things. He
suggests that ultraviolet photometry be seriously considered as a
primary standard.
28
-------
APPENDIX A
Dr. J. A. Hodgeson's References
(Statements given with the references are the salient features of
the paper).
1. Ouster and Natelson, Anal. Chem. 21_. 1005,1949 . Absorption
spectra of trace I« in water, aqueous KI, organic solvents.
Amax in KI at 352 and 289 nm
Variation of a(352) with KI cone.
2. Birdsall, Jenkins and Spandinger, Anal. Chem. 24_:662,1952.
Compared iodometric ozone measurements with absolute ozone measured
by gas density. Approximately 1.1 agreement between neutral KI
and physical measurement (1-25% Og). Acidic KI results 50% higher
than neutral KI.
3. Byers and Saltzman. Adv. Chem. Series, No. 21, ACS (1959):93-101.
The original article on 1% neutral KI technique and basis for EPA
procedure assumes 1:1 stoichiometry based on earlier work. Alkaline
KI results were 65% of neutral KI measurements.
4. Deutsch. J. Air Poll. Contr. Assoc. 18:77, 1968. Recommends
an acid KI (pH = 3.S) technique for improved stability and freedom
from interferences. Obtained 1:1 agreement between acid and
neutral KI from 0.06 to 0.24 ppm 03. Iodine losses from aerated acidic
KI solutions noted.
5. Bergshoeff. Preprint, Pittsburgh Conference, 1970. Compared 5
spectrophotometric methods for 0., measurement (neutral KI, Di(4-pyridyl
etnylene), Diacetyl-dihydrolutidine, Dimethoxystilbene, indigotin-
29
-------
disulfonate). Deduced 1:1 stoichiometry for neutral KI based on
comparison. Modified the KI technique by adding excess thio-
sulfate to absorbing solution and performing back titration.
6. Boyd, Willis and Cyr. Anal. Chem. 4£:670, 1970. Determined
stoichiometry of neutral KI reaction by absolute UV photometric 0,
measurement. 13" formed/03 absorbed - 1.5 over range of about 50 to
1000 ppm. Used static KI measurement of 03. Prepared chemical
mechanism to account for different results in neutral and alkaline KI.
7. Hodgeson, Baumgardner, Martin and Rehme. Anal. Chem. 431:1123,
1971. Redeternrined stoichiometry of ambient 03 concentrations
by gas phase titration of 03 with known nitric oxide concentrations.
Standard nitric oxide concentrations obtained from standard Scott
cylinders, reanalyzed by Saltzman N02 technique.
I3 formed/03 absorbed = 0.98 ^ .07
8. Kopczynski and Bufalini. Anal. Chem. 43_:1123, 1971. Measured
absolute 03 concentration by IR absorption and compared with neutral
KI.
I3~ formed/03 absorbed 1 1.0
9. Dietz, Pruzansky and Smith. Anal. Chem. 45:402, 1973.
Determined absolute 03 concentration by thermal decomposition -
pressure differential measurement; measured a unity stoichiometry
factor for high 03 concentrations in neutral KI. By gas phase
dilution obtained unity factor for sub-ppm concentrations.
30
-------
Determined variations in stoichiometry with pH:
4 1.06
7 1.02
9 0.99
12 0.95
14 0.80
10. Parry and Hern, Environ. Sci. Techno!. 7^:65, 1973.
Discussed significance of I03~ formation in 03 - iodide reaction
in neutral solution. Measured I03" formation electrochemical ly.
I03" formation favored by high 03 concentrations and catalytic
effect of glass surfaces. I03" formation usually insignificant
at ambient 03 concentrations.
31
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing/
2.
1 RCPOH1 NO
EPA-650/4-75-007
1 'I'LL .NDSUBT'T! E " " ~
Summary Report: Workshop on Ozone Measurement by the
Potassium Iodide Method.
3 RECIPIENT'S ACCESSIOWNO.
5 REPORT DATE
February 1975
6 PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
John B. Clements, Ph.D., Chief, MSPEB/QAEML/NERC/RTP
8. PERFORMING ORGANIZATION REPORT NO
10 PROGRAM ELEMENT NO.] HA327
ROAP No. 26AAG
PERFORMING ORGANIZATION NAME AND ADDRESS
Methods Standardization & Performance Evaluation Branch
Quality Assurance & Environmental Monitoring Laboratory
Environmental Protection Agency, NERC
Research Triangle Park, North Carolina 27711
11 CONTRACT/GRANT NO
12 SPONSORING AGENCY NAME AND ADDRESS
Same
13. TYPE OF REPORT AND PERIOD COVERED
Summary Report
14 SPONSORING AGENCY CODE
15 SUPPLEMENTARY NOTES
16 ABSTRACT
The Federal Reference Method for photochemical oxidants specified in regulations
on National Primary and Secondary Ambient Air Quality Standards is based on the chemi-
luminescence resulting from the reaction of ozone with ethylene. The method is cali-
brated by measuring synthetically prepared standard atmospheres of ozone and develop-
ing a method response vs. ozone concentration curve. The concentration of ozone in the
standard atmospheres used for calibration is determined by using the 1% neutral buf-
fered potassium iodide procedure specified in the regulations appearing in the Federa
Register. 36, 84, Part II, 8186-8201, April 30, 1971 . "
Although the 1% neutral buffered potassium iodide procedure has been used for
sometime, it has been criticized for its inconsistent and non-reproducible results.
These criticisms are of concern to EPA because they call into question the methodo-
logical basis for one of its regulatory standards and, likewise, NBS is concerned
because there seems to be a national need associated with an important environmental
measurement.
Therefore, a workshop on the problem was held Aug. 26-27, 1974, at NBS facility,
Gaithersburg, Md., to bring together individuals with expertise in the use of the
potassium iodide method for a full technical discussion of this ozone method calibra-
tion problem. This report presents recommendations and a summary of the discussions
held at this workshop.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c COSATI 1-icLd/Group
Ozone measurement
Potassium iodide procedure
Chemiluminescent procedure
Calibration
DISTRIBUTION STATEMENT
Release unlimited
19 SECURITY CLASS (ThisReport)
None
21 NO OF PAGES
36
20 SECURITY CLASS (Thispage)
None
22 PRICE
EPA Form 2220-1 (9-73)
32
32
------- |