EPA-600/4-77-005
January 1977
Environmental Monitoring Series
EVALUATION OF 1 PERCENT NEUTRAL BUFFERED
POTASSIUM IODIDE PROCEDURE FOR
CALIBRATION OF OZONE MONITORS
iWflnmentai Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
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 quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/4-77-005
January 1977
EVALUATION OF 1 PERCENT NEUTRAL BUFFERED
POTASSIUM IODIDE PROCEDURE FOR CALIBRATION
OF OZONE MONITORS
by
Michael E. Beard, John H. Margeson, and Elizabeth C. Ellis
Quality Assurance Branch
Environmental Monitoring and Support Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and Support
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
ii
-------
CONTENTS
Page
List of Figures iv
List of Tables v
1. Introduction '. 1
2. Experimental 2
General 2
Test Atmosphere Generation 2
Sampling and Analysis 5
Test Parameters 5
3. Results and Discussion 7
Ozone Generator Calibrations 7
Evaluation Procedure 11
Improvements in the Published NBKI Calibration Procedure 23
Comparison of Improved Procedure with GPT 23
4. Conclusions 25
5. References 26
6. Appendix A: Technical Guidance for Obtaining Improved Precision
and Accuracy in Using the Ozone Calibration Procedure Given in
40 CFR Part 50, Appendix D 28
7. Appendix B. Appendix D of 40 CFR Part 50 32
iii
-------
LIST OF FIGURES
Number Page
1 Schematic Diagram of Ozone Generation System and
-Gas Phase Titration Apparatus 3
2 Schematic Diagram of Calibration Apparatus Using
N02 Permeation Device 4
3 Calibration Curve for Ozone Generator 8
4 Absorbance of I~ Equivalent to 03 Versus Time 19
5 I2 vs Absorbance Calibration With "Impure" ACS
Reagent Grade Potassium iodide _ 22
6 I« vs Absorbance Calibration with "Pure" ACS
Reagent Grade Potassium Iodide 22
7 Comparison of Neutral Buffered Potassium Iodide
Method for Ozone with Gas Phase Titration Ozone Method 24
iv
-------
LIST OF TABLES
Number Page
1 Comparison of Measured Ozone Concentration arid Ozone
Concentrations Predicted by Least Square Linear and
Parabolic Regressions (Composited Data) 9
2 Comparison of N02 Generated by GPT of NBS SRM NO
Cylinder and N0« from a Permation Device 10
3 Ruggedness Test No. 1 12
4 Ruggedness Test No. 2 13
5 Ruggedness Test No. 3 15
6 Ruggedness Test No. 4 16
7 Stability of Prepared Iodine Standards with Time 18
8 Ozone Recovery from Midget Impingers vs Modified
Saltzman Bubblers 20
-------
SECTION 1
INTRODUCTION
In the April 30, 1971, Federal Register,1 the Environmental Protection
Agency (EPA) promulgated national ambient air quality standards for six pollutants,
including a standard for photochemical oxidants (ozone). A reference method for
support of that standard was described in Appendix D in that same document. The
standards were subsequently codified as 40CFR Part 50, Appendix D, and amended on
February 18, 1975 (FR 7042). Appendix D currently describes the measurement
principle and calibration procedure. While the standard is nominally for photo-
chemical oxidants, the reference method is specific for ozone (03), and is based
on monitoring the chemiluminescence from the reaction of ozone and ethylene.
An 03 generator calibrated by the 1 percent neutral buffered potassium
iodide (NBKI) procedure is used to calibrate the chemiluminescence monitor. The
calibration procedure is based on the reaction of one mole of 0, with an excess 3
of iodide (I ) in a neutral buffered solution to produce one mote of iodine (L,)-
The iodine is determined colorimetrically by measuring its absorbance at 352 nm.
The NBKI procedure has been subjected to detailed study both prior to and .
following its promulgation as part of the reference method. Byers and Saltzman ,
and Saltzman and Gilbert described the optimum use of the 1 percent NBKI method
for measurement of 03 in ambient air. Their investigations described the method
much as it appears in the Federal Register and included a determination of the
1:1 03-I2 stoichiometry at neutral pH. TFfey confirmed the stoichiometry of 1:1
by conducting a gas phase titration (GPT) of 03 with nitric oxide (NO) to produce
nitrogen dioxide (N02) and subsequently determining the NOg generated.
Later, Altshuller and Wartburg showed the importance of using allgglass
or Teflon connections in the sampling system to avoid 03 losses. Jacobs
reported that impingers gave higher results than fritted glass absocbers. A
challenge of the 1:1 stoichiometry of 03:I2 was made by goyd et al. and later
refuted by Hodgeson et^al_. and Kopczynski and Bufalini. EP^aruT,National
Bureau of Standards sponsored a workshop on ozone analysis by NBKI at which
various researchers presented data describing their experiences with NBKI.
Despite all of these investigations, there was still considerable un-
certainty within the scientific community as to the reliability of the NBKI
procedure for analysis of ozone. Accordingly, the Quality Assurance Branch
of the Environmental Monitoring and Support Laboratory (QAB/EMSL) developed
a plan to thoroughly evaluate the NBKI procedure. This report contains the
results of the evaluation.
-------
SECTION 2
EXPERIMENTAL
A. General
The procedure used for generation and analysis of ozone used in this
study is described in the Federal Register with exceptions as described below.
GPT procedure for,calibration of N0? monitors is found in the June 8, 1973,
Federal Register.
B. Test Atmosphere Generation
The system used to generate test atmospheres for this study are shown
in Figures 1 and 2.
1. Ozone Generation
Ozone was generated continuously by ultraviolet irradiation of clean
air passing through a quartz-tube. The amount of ozone produced was controlled
by a movable sleeve surrounding the U-V lamp. A constant voltage transformer
was used with the U-V lamp power supply to eliminate fluctuations in line
voltage and, thus, assure constant output. This 0- generator ancUits operating
characteristics have been described in detail by Hodgeson ejt al_. and by the
National Bureau of Standards.
The ozone generator was calibrated for this study by GPT of National
Bureau of Standards Standard Reference Material (SRM) No. 1683 (50 ppm NO in
N2) with 03. The SRM was carefully diluted with clean air to a concentration
of about 1 ppm. The exact NO concentration was calculated from flow measure-
ments. The NO was then titrated with ozone and the decrease in the NO concen-
tration, AND, was assumed to be equivalent to the 03 produced by the generator.
A Bendix Model 8101-B Oxides of Nitrogen analyzer was used to measure the NO and
N0? produced by the GPT system. The analyzer was modified by incorporating a
Monitor Labs "Moly-Con" NO converter (Monitor Labs, Inc., San Diego, Calif.).
Ozone generator calibration curves were prepared by plotting sleeve settings vs
0- generated. The generator was calibrated in this manner prior to each experiment.
NO, generated by the 03-NO reaction was compared to the output of
gravimetricaTly calibrated N02 permeation devices to test the validity of the
calibration and the GPT system. The devices had permeation rates of 0.965
+ 0.004 pg NOp/min and 0.668 + 0.001 ug N02/min (Mean +_ Standard Deviation) at
"2~5.1 j^0.1°C. Temperature for the devices was maintained by circulating water
through a water-jacketed condenser connected to a Forma Temp, Jr. constant
temperaturegbath. Permeation devices and their use have been described in
detail.
-------
FLOW
CONTROLLER
FLOWMETER
OZONE
GENERATOR
FLOW
CONTROLLER
FLOWMETER
REACTION
CHAMBER
MIXING
CHAMBER
NITRIC
OXIDE
STANDARD
FLOW
CONTROLLER
VENT
FLOWMETER
OUTPUT
MANIFOLD
EXTRA OUTLETS CAPPED
WHEN NOT IN USE
r
TO INLET OF ANALYZER
UNDER CALIBRATION
Figure 1. Schematic diagram of ozone generation system and gas phase titration apparatus.
-------
VENT
EXTRA OUTLETS CAPPED
WHEN NOT IN USE T0 INLET OF ANALYZER
UNDER CALIBRATION
CONSTANT TEMPERATURE
CHAMBER
Figure 2. Schematic diagram of calibration apparatus using NC>2 permeation device.
-------
Air flow for the (L generator and GPT systems was effectively
controlled using two single pressure regulators in series and a needle
valve. NO flow from the cylinder was controlled by using a two-stage pressure
regulator and a needle valve. Flow measurements were made with calibrated wet
test meters and soap bubble meters.
2. Clean Air
Purified air used for (L generation and dilutions was produced by
passing compressed (house) air through silica gel for drying, by treatment with
03 to convert NO to N02, and finally by scrubbing the air with a column packed
with activated charcoal (6-A mesh), molecular sieve (6-16 mesh, type 4-A), and
silica gel to remove any NO, or hydrocarbons and traces of water. The air was
tested for completeness of N02 and 03 removal by sampling with 1 percent NBKI.
No background absorbance was observed.
C. Sampling and Analysis
Ozone was sampled from the output manifold, shown in Figure 1, by a
train of two or more absorbers. Each absorber contained 10 ml of 1 percent NBKI.
The absorbers were connected by glass tubing fitted with ground glass ball and
socket connectors. All materials between the ozone sampling manifold and the
absorbers were either glass or Teflon.
The iodine generated by ozone in the absorbers was analyzed by measuring
its absorbance at 352 nm in a 1-cm cell. A Beckman Model "B" spectrophotometer
was used for the absorbance measurements and was calibrated using iodine solu-
tions standardized against primary standard arsenic trioxide. Absorbances were
usually determined immediately (within 2 minutes after collection). However,
absorbance measurements versus time were made on some samples for periods of up
to 1 hour using a Varian Model 635 Spectrophotometer equipped with a strip chart
recorder.
D. Test Parameters
1. Absorbers
Three types of absorbers were used for the study: 1) midget impingers
(Ace Glass No. 7531), 2) modified Saltzman or "Mae West" type bubblers (Ace Glass
No. 7530), and 3) a smog bubbler (Ace Glass No. 7529-16).l7
The absorbers were normally used in trains of two units. Some experi-
ments employed 3-unit absorber trains to test the collection efficiency of the
system.
2. Temperature
The temperature of the absorbing solution was controlled by placing
the absorber in a plastic tub containing water whose temperature had been
adjusted with ice or hot water to obtain working temperatures between 25 and
32°C. Absorbers and solutions were allowed to equilibrate in the bath for
about 10 minutes before sampling.
-------
3. Light Conditions
Experiments in which the absorbers were shielded from room light were
conducted by covering theMmpingers. Normally, the absorbers were exposed to
fluorescent room lighting.
4. Flow Control
Sampling flow rate was varied by the appropriate gage hypodermic
needle (used as a limiting orifice to control flow in the method). When
needles were not available for the desired flow, a larger needle was crimped
slightly with pliers until the desired flow was obtained. The needle was
protected by a membrane filter and by a drying column following the last
absorber in the sampling train to ensure constant flow. All flows were
measured in-train with wet test meters and soap bubble meters. Rotameters
were used as flow monitoring devices.
-------
SECTION 3
RESULTS AND DISCUSSIONS
A. Ogone Genefator Calibrations
In evaluating the NBKI calibration procedure, it was imperative that
the source of ozone test atmospheres be analyzed by a procedure based on accurate
standards independent of the NBKI method.
A 6PT method referenced to an NBS SRM NO cylinder was used to analyze
the ozone atmospheres. This analysis was verified by comparison of the N0?
generated by 6PT to N02 generated independently with an SRM N02 permeation device.
The ozone generator was calibrated as described in Section 2.B.I.
Four different ozone concentrations covering a range of 0.07 to 0.7 ppm were
generated using sleeve settings of 10, 30, 50, and 70, and an air flow of 5
liters per minute (1/min). The calibration was performed 10 times throughout
the study. The data were subjected to least squares regression analysis. A
typical calibration curve for the ozone generator is shown in Figure 3 along ,
with the equation (parabolic) for the curve. Since it has been reported by NBS,
and confirmed by out data, that the output of these generators is significantly
nonlinear from sleeve settings of 0 to 10, the zero concentration-zero sleeve
setting points were not used in the regression analysis.
Table 1 compares the mean ozone concentrations with those predicted by
the parabolic regression curve derived from the 6PT data. Table 1 also shows
the ozone concentration that would be predicted if the nonlinearity of the
generator output were ignored, i.e., fitting the data to a linear curve. At
0.07 ppm the concentration would be in error by 13 percent for this generator.
At a predicted concentration of 0.05 ppm 03, an actual concentration of 0.04
ppm was measured — an error of 20 percent. Thus, users of ozone generators
should not assume the output to be linear, otherwise substantial errors may be
introduced at low (near ambient) concentrations.
The N0« generated by the GPT of the NBS SRM NO cylinder was compared
to the output o-ra gravimetrically calibrated N02 permeation device as an
independent check on the accuracy of the NO SRM. The data in Table 2 show the
results of that comparison. The least squares linear regression for the data is
[N02]GpT = 1.012 [N02]p>D + 0.0050
where:
[N02]GpT = N02 generated by GPT of NBS SRM NO cylinder
[N02]p D = N02 generated from permeation device
-------
00
0.70
0.60
O.SO
0.40
0.00
ppm03 = -0.0247 + 0.00931 (SLEEVE SETTING) + 0.000018 (SLEEVE SETTING)2
30 SO
GENERATOR SLEEVE SETTING
Figure 3. Calibration curve for ozone generator.
70
-------
TABLE 1. COMPARISON OF MEASURED OZONE CONCENTRATION
AND OZONE CONCENTRATIONS PREDICTED BY LEAST SQUARE
LINEAR AND PARABOLIC REGRESSIONS (COMPOSITED DATA)
Measured Ozone, ppm Predicted Ozone, ppm
Sleeve Setting
10
30
50
70
(Mean)
0.072
0.267
0.490
0.714
LSLRa
0.062
0.277
0.491
0.705
LSPRb
0.070
0.271
0.486
0.715
\east squares linear regression.
Least squares parabolic regression.
-------
TABLE 2. COMPARISON OF N02 GENERATED BY
6PT OF NBS SRM NO CYLINDER AND
N02 FROM A PERMEATION DEVICE
NOp Concentration, ppm
SRM NO Permeation Device
0.721 0.715
0.511 0.495
0.291 0.270
0.076 0.075
0.000 0.000
10
-------
The slope of 1.012 is not significantly different from 1.0 at the 95 percent
confidence interval. This agreement of two independent standards was con-
sidered adequate. Accordingly, the 6PT calibration technique and a calibration
curve like that in Figure 3 were used to obtain known ozone concentrations.
B. Evaluation Procedures
The NBKI procedure,was evaluated by two methods. First, ruggedness
tests as described by Youden were used to screen the method for sensitive
variables. Variables showing significant variation in the ruggedness tests,
as well-as those variables not lending themselves to ruggedness testing, were
tested individually.
1. Ruggedness Testing
One requisite for a meaningful ruggedness test is that the variables
have no interaction. Inclusion of a system blank or "dummy" variable in theg
test serves as a means of indicating whether or not the test is in control. A
low value obtained for the dummy variable indicates no significant interaction
and means all variables are in control, whereas a high dummy value indicates
interaction. The response obtained in each ruggedness test is expressed in
percent recovery and was calculated by dividing the ozone concentration
analyzed by the NBKI method by the ozone concentration generated. Normalization
of the results was necessary before the effect of the different parameters could
be determined because the method is obviously sensitive to ozone concentration.
The first ruggedness test was designed with only concentration as a
variable and the remaining six variables as dummies. Ozone concentrations of
0.043 to 0.772 ppm were generated to cover a wide range of calibration concen-
trations. The results are given in Table 3 and show that the dummies are all
low and similar. The average dummy value was 0.7 +_ 0.5 percent (one standard
deviation) and was used to provide a gauge for judging control in future
ruggedness tests. An effect due to concentration (-5.4 percent) was observed
in this test. No reason for the concentration effect was apparent at this time.
The second ruggedness test examined the effect of absorbing solution
temperature and exposure of the solution to light during sample collection at
ozone concentrations of 0.10 and 0.70 ppm. The four remaining variables were
dummies. The absorbing solution temperature was varied 5°C to approximate the
normal variation in laboratory temperature that might be experienced by users
of the method.
The results in Table 4 show that the dummies are low (2.3 +_ 1.3
percent, one s.d.). The responses due to temperature and light exposure were
both insignificant, indicating that these variables were not affecting the
method response. Again, an effect due to concentration was observed (+8.9
percent). Thus, the first two ruggedness tests suggest that ozone concentra-
tion affects the NBKI method and a separate test is in order.
The third ruggedness test examined the effect of five variables:
sampling time, sampling flow rate, impinger (absorber) type, number of impingers.
11
-------
TABLE 3. RUGGEDNESS TEST NO. 1
Variable Range Response, percent
1. Concentration, ppm 0.043 vs 0.772 -5.4
2. Dummy 1 - -1.0
3. Dummy 2 - +0.4
4. Dummy 3 - -0.4
5. Dummy 4 - +0.9
6. Dummy 5 - -0.2
7. Dummy 6 - +1.5
12
-------
TABLE 4. RUGGEDNESS TEST NO. 2
Variable Range8 Response, percent
1. 'Concentration, ppm 0.10 vs 0.70 +8.9
2. Collection Temp., °C 27 vs 32 +1.0
3. Light Exposure Light vs Dark -0.1
4. Dummy 1 - -2.1
5. Dummy 2 - -2.5
6. Dummy 3 - 0.7
7. Dummy 4 - 3.9
a
The nominal variable is listed first followed by the challenging variable.
13
-------
and delay time for determining absorbance. The ozone concentrations were again
set at 0.10 and 0.70 ppm. One dummy variable was included in the design.
The NBKI procedure allows a sample flow rate of 0.2 to 1.0 £/min and
a sampling time of 10 minutes. Sampling flow rates of 0.6 and 1.0 £/min and
sampling time of 10 and 15 minutes were chosen to detect any differences in
collection efficiency.
The use of midget impingers and modified Saltzman bubblers was compared.
The modified Saltzman bubbler is specified in the method; however, experience has
shown that a number of analysts use midget impingers. The number of impingers was
also varied by using three instead of the specified two to'see if any ozone was
passing through the train of two impingers. The time allowed for measurement of
iodine absorbance was also included as a variable. The method specifies that the
absorbance be measured "immediately" after sample col lection. The nominal and
challenging variables were set at 2 and 10 minutes, respectively.
The results in Table 5 show that the response due to number of impingers
and flow rate variation were low and similar to the dummy value of 3.6 percent.
The most significant effect was due to the elapsed time before determination of
iodine absorbance as shown by the 14.6 percent increase in response (a negative
value indicates an increase in response and a positive value a decreased response).
The use of the modified Saltzman impinger appears to reduce the method's response
relative to the midget impinger. Increasing the sampling time from 10 to 15
minutes increased the response by 6 percent.
The fourth ruggedness test (Table 6) again used ozone concentrations
of 0.10 and 0.70 ppm and one dummy variable. The high dummy value of 16.5 percent
and high responses for nearly all the variables suggested that considerable inter-
action between variables was occurring.
Reviewing the foregoing ruggedness tests showed that as the number of
variables increased, the dummy values also increased. Also, Ruggedness Test 2
showed a negligible effect (-0.1 percent) for light vs. dark sampling conditions
where Test No. 4 showed an appreciable effect (-14.3 percent) for the same
variable. Similar worsening effects were seen for other variables.
The occurrence of interactions and varying results in the ruggedness
tests emphasize the need to strictly follow the procedure. Apparent modest
changes in the procedure, as introduced in the fourth ruggedness test, produced
highly variable results.
Further investigation of the NBKI method was conducted through individual
tests of sensitive variables identified by the ruggedness tests as well as tests on
previously examined variables.
2. Single Variable Experiments
a. Absorption Maximum of I2
The absorption spectrum of a 1 percent NBKI solution containing
IP (generated by 0, was determined by scanning in the 350-nm region (using a
Varian Model 635 Spectrophgtometer). The scan confirmed the reported absorbance
maximum of I2 at 352 nm. ' '
14
-------
TABLE 5. RUG6EDNESS TEST NO. 3
Variables
1. Concentration, ppm
2.' Absorbance Det., min
3. Impinger Type
4. No. of Impingers
5. Sampling Time, min-
6. Sampling Rate,
7. Dummy
Range
0.10 vs 0.70
2 vs 10
Response, percent
+6.4
-14.6
Midget vs Modified Saltzman +6.6
2 vs 3
10 vs 15
0.6 vs 1.0
+3.0
-6.2
-4.6
-3.6
The nominal variable is listed first followed by the challenging variable.
15
-------
TABLE 6. RU66EDNESS TEST NO. 4
Variable
1. Impinger Type
2. Buffer Reagent
3. Dummy
4. Concentration, ppm
5. Light Exposure
6. Sampling Rate, Ji/min
7. Collection Temp., °C
Ranged
Midget vs Smog Bubbler
KH2P04'7H20 vs Anhydrous KH2P04
0.10 vs 0.70
Light vs Dark
1.0 vs 1.5
25 vs 30
Response, percent
23.8
19.1
-16.5
15.5
-14.3
13.5
-6.2
The nominal variable is listed first followed by the challenging variable.
16
-------
b. Absorbance Development Time
Prior to determining the effect of time on the development of iodine
absorbance, the stability of the iodine standards used for spectrophotometer cali-
bration (in the method) was determined. The absorbance was read immediately
(2 minutes) after individual preparation, and again after 30 minutes and 2 hours.
All standards were in Pyrex volumetric flasks and were capped with ground glass
stoppers. The data in,Table 7 show no significant change in absorbance.
According to Clements, data reported by Hodgeson for a similar experiment
showed a decrease in absorbance with time. However, a verbal communication with
Dr. Hodgeson revealed that his flasks were not stoppered during his experiments.
This observation points out the importance of eliminating possible losses of
iodine by volatilization or reaction with air.
The absorbance of iodine with time, generated by sampling four con-
centrations of ozone, is shown in Figure 4. These data show an increase in
absorbance with time; at concentrations above 0.1 ppm, maximum color development
occurs in about 40 to 50 minutes. The absorbance increase is about 6 to 8 percent
except at the lowest concentration, for which the absorbance is constant. These
data are similar to data reported by Saltzman and Gilbert who postulated several
mechanisms to account for the increased absorbance. Since the method calls for
"immediate" determination of the sample abscrbance, one might think that this
increase should not be a factor. However, oue to the variation of apparatus and
technique from one analyst to another, appreciable differences might be observed.
c. Effect of Type of Absorber
The effect of different types of absorbers was investigated further
by sampling common atmospheres, over the range 0.2 to 1.0 ppm, with two modified
Saltzman bubblers and two midget impingers. The ozone concentration was then
determined using the NBKI analysis procedure. The data are given in Table 8. The
least square linear regression equation for this data is ppm 03(M,-ci jmD ) = 1.164
ppm O-/... Saltzman) " °-^3^- Tnus» midget impingers collect about 16 percent more
ozone than Modified Saltzman bubblers. This finding is in good agreement with the
results of similar experiments carried out at 0.2 ppm ozone. It is assumed that
the difference is due to losses in the Modified Saltzman bubbler. The data in
Table 8 also show that the increased efficiency of the midget impinger results from
increased absorbance in the first impinger only.
A smog bubbler similar to the modified Saltzman bubbler having a deeper
collection chamber and a 1-mm capillary orifice (Ace Glass No. 7529-16) was also
compared to the midget impinger. Both of these factors should increase the Og-KI
contact time and possibly improve the collection efficiency. However, the smog
bubbler gave low results similar to those of the modified Saltzman bubbler. The
reason for the higher efficiency of the midget impinger over the bubbler is possibly
related to differences in surface area and volume.
d. Purity of Potassium Iodide
It has been reported that certain lots of ACS Reagent Grade Potassium
Iodide contain small amounts of an impurity that exhibits an iodine (ozone) demand.
The iodine demand was confirmed in this study and is demonstrated in the iodine vs
absorbance calibration curve in Figure 5. The KI used in this test has an iodine
17
20
-------
TABLE 7. STABILITY OF PREPARED IODINE STANDARDS WITH TIME
Absorbance at 352 run
Concentration3
0.975
2.44
4.88
9.75
14.6
19.5
2 min
0.058
0.133
0.262
0.524
0.782
1.035
30 min
0.055
0.130
0.259
0.521
0.779
1.032
2 hr
0.059
-
0.262
0.518
-
1.023
alodine concentrations expressed as pg
18
-------
0.85"
0.800
0.700
0.600
1 0.500
UJ
u
oc
o
U)
ta
0.400
0.300
0.200
0.100
0.000
'0.8 ppm 03
•ABSORBANCE:
0.752 INITIAL
0.813 MAXIMUM
'0.5 ppm 03
•ABSORBANCE:
0.500 INITIAL —
0.530 MAXIMUM
~ 0.3 ppm 63
•ABSORBANCE:
0.305 INITIAL
0.326 MAXIMUM
~0.1 ppm 03
•ABSORBANCE:
0.083 INITIAL
0.083 MAXIMUM-
I I I I
0 10 20 30 40 50 60 70 80 90
TIME, minutes
Figure 4. Absorbance of 12 equivalent to 03 versus time.
19
-------
TABLE 8. OZONE RECOVERY FROM MIDGET IMPINGERS VS
MODIFIED SALTZMAN BUBBLERS
Midget Implhgers
Modified Saltzman
Absorbarice
No. 1
0.142
0.318
0.495
0.798
No. 1
0.141
0.289
0.454
0.696
No. 2
0.014
0.025
0.037
0.061
Absdfbance
No. 2
0.012
0.025
0.035
0.058
Total Abs
0.156
0.343
0.568
0.859
Total Abs
0.153
0.314
0.489
0.754
00 Concentration, ppm
0.183
0.399
0.618
0.996
00 Cdiicentf ati on , ppm
0.179
0.366
0.568
0.875
20
-------
demand equivalent to 0.83 yg (L, as indicated by the large positive intercept of
the least squares linear regression equation. This amount of ozone is equivalent
to 0.05 ppm 03 at a sampling rate of 1.0 a/min and a ID-minute sampling time.
A different lot of ACS reagent grade KI was used to obtain the
calibration curve shown in Figure 6. Here the intercept is only 0.01. A
review of several similar calibration curves showed a slope of 19.22 + 0.45
M9 °3/Abs^Mean i 95 Percent confidence band) and an average intercept of -0.06
± O.T4 yg 03 (Mean +_ 95 percent confidence band). Thus, KI of acceptable
purity should yield an intercept in the range -0.20 to +0.08 95 percent of the
time. If-in preparing a calibration curve an analyst obtains an intercept
outside of this range, the KI should be rejected and a new lot of KI should
be used to prepare another calibration curve.
e. One Percent NBKI Stabllity
The effect of (1) storage in different types of containers,
(2) exposure to room light and temperature, and (3) exposure to elevated
temperature on the absorbance of the 1 percent NBKI reagent and its ability
to collect ozone was tested. The containers were as follows:
B. Blank (reference) low actinic flask
1. Clear Lurex 500-ml volumetric flask
2. Low actinic 500-ml Pyrex flask
3. Clear Pyrex reagent bottle
4. Dark brown Pyrex reagent bottle
5. Clear glass reagent bottle
6. Dark brown glass reagent bottle
7. Teflon bottle
8. Clear Nalgene bottle
9. Dark brown Nalgene bottle
The containers were washed with cleaning acid, rinsed thoroughly with distilled
water, and rinsed and then filled with freshly prepared 1 percent NBKI. They
were then placed on the laboratory counter where they were exposed to existing
conditions of light and temperature. A reagent blank "B" was stored in the dark
for use as a reference.
Stability was measured by comparing the absorbance of the different
solutions with the blank value over an 8-week period and by using the same
solutions (and the blank) to sample an atmosphere containing 0.10 ppm ozone seven
timp<: over the same time period. No significant change in absorbance or ability
to collect ozone was observed over the 8-week test period.
Portions of the solutions exposed to room conditions from containers
1, 2, 8 and 9 and the blank were then heated to 38°C (from 25°C) and held at this
temperature for 6 hours. The solutions were then cooled to 25°C, their absorbance
was compared to the unheated blank, and they were used to sample the 0.10-ppm
ozone atmosphere, as in the previous experiments. The solutions subjected to
temperature cycling showed a decrease in their absorbance background, but no loss
in their ability to collect ozone when referenced to the blank sample that had
been heated to 38°C also.
21
-------
o
N
O
3
20
15
10
ppm-
f 19.25 (ABS) + 0.83
9.82 (ABS) + 0.42
0.2
0.4 0.6
ABSORBANCEAT352nm
0.8
1.0
Figure 5. \2 versus absorbance calibration with impure ACS reagent grade potassium iodide.
20
15
o
IM
O
S
10
«K)3 = 19.22 (ABS) + 0.01
n 9.80 (ABS) + 0.01
Ppm03 = y-
0.2
0.4 0.6
ABSORBANCE AT 352 nm
0.8
1.0
Figure 6. \2 versus absorbance calibration with "pure" ACS reagent grade potassium iodide.
22
-------
From these experiments it is concluded that the stability of
the 1 percent NBKI reagent does not appear to be an important variable.
C. Improvements in the Published NBKI Calibration Procedure
Based on the findings of this study, a document was prepared
which gives the essential details for an improved NBKI procedure. The
document is given in the Appendix. The main points are (a) use of a con-
stant voltage transformer with the ozone generator to ensure the constancy
of the ozone being sampled, (b) use of midget impingers instead of modified
Saltzman bubblers, (c) specifications and a procedure for determining KI
purity, and (d) stressing the importance of measuring the iodine absorbance
within three minutes. The errors associated with the method should be re-
duced if one follows the instructions and clarifications in this document.
D. Comparison of Improved Procedure with GPT
The improved NBKI procedure was used to analyze ozone atmos-
pheres generated by a GPT-calibration ozone generator over the range 0.07 to
0.72 ppm. The results are shown in Figure 7. The regression equation slope is
°3NBKi = T-O^O) °-034
These data indicate an average positive bias in the NBKI procedure of 7.5
percent with a range of 4 to 11 percent. This result compares favorably
with similar experiments reported by Paur ejt al_. , which showed slopes from
1.0 to 1.11.
23
-------
o
N
°3 NBKI = t-075 <°3 GPT' ± °-034
R2 = 0.999
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Figure 7. Comparison of neutral buffered potassium iodide method for ozone with gas phase titration
ozone method,
24
-------
SECTION 4
CONCLUSIONS
The results of this study show that the NBKI procedure is vulnerable to
error in three major areas: 1) the time allowed for measurement of iodine
absorbance, 2) purity of potassium iodide reagent, and 3) type of impinger used.
The reasons for some of these errors are not fully understood and, therefore,
are difficult to control. Erratic results brought about by modest changes in
operating procedure, such as those introduced in the ruggedness tests, suggest
that widely varying results may be obtained in actual use.
Under ideal use conditions, the NBKI procedure has a small but significant
positive bias.
Since calibration procedures that appear more promising than NBKI are now
available (gas phase titration and UV-photometry), QAB/EMSL has decided that
further work on improving the current NBKI procedure is not warranted. A verbal
communication with Dr. Daniel Flamm of Texas A&M University indicates an alter-
nate iodometric procedure may be developed to overcome some of the problems with
the NBKI procedure. Accordingly, a program has been initiated to evaluate the
above procedures as candidates to replace the NBKI procedure. In the interim
period, the NBKI procedure should be used as specified in the clarification
memorandum to improve data quality.
25
-------
SECTION 5
REFERENCES
1. National Primary and Secondary Ambient Air Quality Standards, Federal
Register, 36_(84):8186-8187, 8195-8200, April 30, 1971.
2. Hodgeson, J.A., B.E. Martin, and R.E. Baumgardner. . Laboratory Evaluation
of Alternate Chemiluminescent Approaches for the Detection of Atmospheric
Ozone. U.S. Environmental Protection Agency, Research Triangle Park, N.C.
(Presented at meeting of American Chemical Society, Chicago, 111.
September 13-18, 1970.)
3. Saltzman, B.E. and N. Gilbert. lodometric Microdetermination of Organic
Oxidants and Ozone. Anal. Chem. 31_(11): 1914-1920, 1959.
4. Byers, D.H. and B.E. Saltzman. Determination of Ozone in Air by Neutral
and Alkaline Iodide Procedures. J. Am. Ind. Hyg. Assoc. 1^:251-257, 1958.
5. Altshuller, A.P. and A.F. Wartburg. The Interaction of Ozone with Plastic
and Metallic Materials in Dynamic Flow System. Int. J. Air Water Pollut.
4_:70-78, 1961.
6. Jacobs, M.B. The Chemical Analysis of Air Pollutants. New York,
Interscience Publishers, 1960. p. 218-220.
7. Boyd, A.W., C. Willis, and R. Cyr. New Determination of Stoichiometry of
the lodometric Method for Ozone Analysis at pH 7.0. Anal. Chem. 42_(6):670,
1970.
8. Hodgeson, J.A., R.E. Baumgardner, B.E. Martin, and K.A. Rehme. Stoichio-
metry in the Neutral lodometric Procedure for Ozone by Gas Phase Titration
with Nitric Oxide. Anal. Chem. 43(8):1123-1126, 1971.
9. Kopczynski, S.L. and J.J. Bufalinf. Some Observations on the Stoichiometry
of the lodometric Analyses of Ozone at pH 7.0. Anal. Chem. 43_( ):1123,
1971.
10. Clements, J.B. Summary Report: Workshop on Ozone Measurement by the
Potassium Iodide Method. U.S. Environmental Protection Agency. Research
Triangle Park, N.C. Report No. EPA-650/4-75-007. February 1975.
11. Reference Method for Determination of Nitrogen Dioxide. Federal Register.
38(110):15174-15180, June 8, 1973.
26
-------
12. Hodgeson, J.A., R.K. Stevens, and B.E. Martin. A Stable Ozone Source
Applicable as a Secondary Standard for Calibration of Atmospheric Monitors.
In: Analysis Instrumentation Symposium. Houston, Instrument Society of
America, April 1971.
13. Taylor, J.K. (Editor). National Bureau of Standards. Washington, D.C.
NBS Technical Note No. 585. January 1972. p. 11-25.
14. Scaringelli, P.P., S.A. Frey, and B.E. Saltzman. Evaluation of the Teflon
Permeation Tubes for Use with Sulfur Dioxide. J. Amer. Ind. Hyg. Assoc.
28:260, 1967.
15. Scaringelli, P.P., A.E. O'Keefe, E. Rosenburg, and J.P. Bell. Preparation
of Known Concentrations of Gases and Vapors with Permeation Tubes Cali-
brated Gravimetrically. Anal. Chem. 42_(8):871, 1970.
16. Rook, H.L., E.E. Hughes, R.G. Fuerst, and J.H. Margeson. Operation
Characteristics of N02 Permeation Devices. U.S. Environmental Protection
Agency, Research Triangle Park, N.C., and National Bureau of Standards,
Washington, D.C. (Presented at the Division of Environmental Chemistry,
American Chemical Society, Spring Meeting, Los Angeles, California,
March 31-April 5, 1974.)
17. Catalog No. 600. Ace Glass, Inc. Vineland, N.J.
18. Lodge, J.P., Jr., J.B. Pate, B.E. Ammons, and G.A. Swanson. Use of
Hypodermic Needles as Critical Orifices in Air Sampling. J. Air Pollut.
Contr. Ass. 1_6_:197, 1966.
19. Youden, W.J. Statistical Techniques for Collaborative Tests. Association
of Official Analytical Chemists. Washington, D.C. 1967. p. 29-32.
20. Mueller, P.K., Y. Tokiwa, E.R. deVera, W.J. Wehrmeister, T. Belsky,
S. Twiss, and M. Imada. A Guide for the Evaluation of Atmospheric
Analyzers. Prepared for the U.S. Environmental Protection Agency, under
Contract No. 68-02-0214, by: Air and Industrial Hygiene Laboratory,
California State Department of Health, Berkeley, Calif. 1973.
21. Paur, R.J., R.E. Baumgardner, W.A. McClenny, and R.K. Stevens. Status
of Method for the Calibration of Ozone Monitors. U.S. Environmental
Protection Agency, Research Triangle Park, N.C. (Presented at Division
of Environmental Chemistry, American Chemical Society, Spring Meeting,
New York City, N.Y. April 1976.)
27
-------
APPENDIX A
28
-------
APPENDIX A: TECHNICAL GUIDANCE FOR OBTAINING IMPROVED PRECISION AND ACCURACY
IN USING THE OZONE CALIBRATION PROCEDURE GIVEN IN 40 CFR PART 50, APPENDIX D.
'The reference method calibration procedure specified in Appendix D of
40 CFR Part 50 allows a moderate degree of flexibility in a number of its
provisions. This flexibility arises because some of the equipment and pro-
cedural specifications are given in terms which tend to be general or subject
to interpretation, rather than highly specific. Such general-type specifi-
cations permit variations to accomodate operator preferences and available
equipment. However these variations compromise, to some extent, the precision
and accuracy of the resulting ozone measurements.
Within certain somewhat general specifications prescribed in Appendix D
of 40 CFR Part 50, it is possible to obtain improved precision and accuracy by
following more detailed and restrictive procedures. These augmentative
procedures and instructions are set forth below for those who wish to obtain
improved results. While use of these additional instructions is not required,
EPA recommends that they be followed where improved uniformity and accuracy
are desired.
1. Section 5.10.1 indicates that "all-glass impingers as shown in
Figure D4 are recommended". It is obvious that the intent of this specifi-
cation is not to exclusively limit the type of absorber used to the exact
type or shape shown in Figure D4. A number of other all-glass absorbers are
available and are permissable under the intent of section 5.10.1. EPA now
believes that the best uniformity and accuracy are obtained by the use of a
type of absorber referred to as a "midget impinger". Specifications for this
midget impinger are given in Figure 1.
2. Section 6 indicates the reagents which are required to carry out
the procedure. Greater uniformity and accuracy may be obtained if the
following reagents are specified as "ACS Reagent Grade": potassium iodide
(KI), potassium dihydrogen phosphate (KH?POJ, disodium hydrogen phosphate
(Na?HPO.), sodium hydroxide (NaOH), sulfDrie acid 95% to 98% H2S04), starch
(soluble), mercuric iodide (Hgl?), iodine (resublimed I2), and sodium
bicarbonate (NaHCOj).
3. Section 6.5 specifies the use of either anhydrous disodium hydrogen
phosphate (Na-HOP.) or the dodecahydrate salt (Na2HOP/12H20). However the
heptahydrate form of this compound Na?HOP/7H20) Ts mdre stable and this
provides improved precision. Since all three of these forms of the compound
are absolutely equivalent chemically, the heptahydrate form is recommended.
The equivalent quantity of Na2HOP4'7H20 is 26.8 grams.
29
-------
4. Section 6.5 specifies the use of potassium iodide (KI). Some
sources of KI -- even ACS Reagent Grade -- have been reported to contain
small amounts of reducing agents. Such impurities can cause an iodine
demand which could cause a significant measurement error. Best accuracy is
thus obtained with KI which has no significant iodine demand. To determine
if this specification is being met, plot the I2 absorbance (y-axis) versus
total jig Oo (x-axis) calibration curve in section 8.1.2 as follows:
Plot the five points obtained for the 0.5, 1, 2, 3, and 4 ml of
I,- Do not include zero as a point. Draw the curve or use
linear regression analysis. If the intercept is significantly
different from zero, the KI has an I2 demand and should be
discarded. Typical intercepts obtained at NERC/RTP are between
+0.08 and -0.20 pg 03. Typical slopes obtained for-the I« vs
Absorbance are between 19.67 and 18.77 ug 03/Absorbance Unit.
5. Section 6.2 specifies "cylinder air, dry grade". This specification
is rather ambiguous, but it seems clear that the intent is to require air of a
clean, dry, and uniform quality. Air from sources other than compressed gas
cylinders can certainly meet and exceed those requirements. Thus, this speci-
fication can be interpreted to include any source of clean, dry, uniform air.
But again, greater precision can be realized if additional specifications are
applied to the air. In particular, the air must also be essentially free of
ozone, NO, N02, reactive hydrocarbons, and any other interferent which may
cause a positive or negative bias. While it is not practical to attempt to
analyze the air for presence of these various interferents, air meeting the
above additional requirements may be produced by appropriate treatment as
described below. Either compressed ambient air or cylinder air may be treated.
The air is first passed through silica gel for drying, then it is treated with
ozone to convert any NO to N02- Finally, the air is passed through 6-14 mesh
activated carbon and 6-16 mesn type 4A molecular sieve to remove N02 and
hydrocarbons.
NOTE: The oxygen content of cylinder air may vary from 18% to
22%, which can cause changes in the calibration of ozone generators. If
cylinder air is used, recalibration of the ozone generator with each new
cylinder of air is recommended.
6. Sections 8.1.1, 8.2.2.1, and 8.2.2.2 all require measurement of the
iodine absorbance "immediately". Best precision is obtained when these
measurements are made within 3 minutes after obtaining the material to be
measured.
30
-------
S mm 1.0.-J U-
INSIDE
CLEARANCE
3TO 5mm
10 mm 0.0.
? 24/40, CONCENTRIC WITH
OUTER PIECE AND WITH
NOZZLE
GRADUATIONSAT5-ml
INTERVALS. ALL THE
WAY AROUND
.NOZZLE 1.0. EXACTLY
1mm;PASSES O.C9 TO O.It
-dm AT 12 in. H20 VACUUM.
PIECES SHOULD BE INTER-
CHANGEABLE, MAINTAINING
NOZZLE CENTERING AND
CLEARANCE TO BOTTOM
INSIDE SURFACE
Figure 1 •
31
-------
APPENDIX B
33
-------
40
Protection of
Environment
PARTS 50 TO 69
Revised as of July 1, 1975
CONTAINING
A CODIFICATION' OF DOCUMENTS
OF GENERAL APPLICABILITY
AND FUTURE EFFECT
AS OF JULY 1, 1975
With Ancillaries
Published by
the Office of the Federal Register
National Archives and Records Service
General Services Administration
as a Special Edition of
the Federal Register
35
-------
Chqffter I—Environmental Protection Agency
§50.11
APPENDIX C—MEASI)BEM£NT
PRINCIPLE
FOR THE
AND/
CALIBRATION/PROCEDURE
TINUOUS MEASUREMENT or CARBON
OXIDE IN^HE ATMOSPHERE (NoN-D
STVE INTOARED SFECTBOMETRT)
1. Principle and applicability.
1.1 yrfhls principle is based on/he absorp-
tion/or Infrared radiation by caroon monox-
in a non-dispersive photometer. Both
i pass into matched cells, each contain-
ing a selective detector and CO. The CO in
the cells absorb infrared'radlation only at its
characteristic frequencies and the detector is
sensitive to those frequencies. With a non-
absorbing gas in joe reference cell, and
no CO In
both detector? are balanced electronlpfilly.
Any CO Introduced into the sample cpQ will
absorb radiation, which reduces the/temper-
ature and pressure in the detector cell and
dlaphram. This displacement la
. electronically and amplified to pro-
output
this principle
e method only
as a reference
method in accordance* with Part 53 of this
chapter.
2.—6. [Rese:
7.
7.1 Calibrate the instrument as described
In 8.1. All ^ases (sample, zero, calibration,
aust be Introduced into jme en-
system. Figure -Cl^fthowB—6
ow_ diagram. For speci&ir operating
fictions, refer to the manufacturer's
iual.
Calibration.
8.1 Calibration
linearity
operating flow rate
pare a calibration <
furnished with
zero gas i
a recorder rearfing of zero. Introduce sySn
gas and adjust the span control to indicate
the proper/value on the recorder scale (e.g.
on 0-58>mg./m.» scale, set the 4^mg./m.1
at 80 percent of the recorder
Recheck zero and span/until adjust-
are no longer necesalry. Introduce
idlate calibration gpses and plot the
lues obtained. If a safooth curve is not
obtained, calibration,-' gases may need
replacement. /
9. Calculations.
9.1 DetermlnXthe concentrations directl
from the calibration curve. No calculatli
are
monoxide concentrations in
mg./m.y6r« converted to p.pjn. as/TOllows:
p.p.m. CO=mg. CO/m.»xyS73
rfO. Bibliography. s
The Intech NDIH-CO A^ilyzer by Frank
McElroy. Presented at .the llth Methods
Conference in Air PoUtitlon. University of
California, Berkeley. .Calif.. April 1, 1970.
Jacobs. M. B. et'ftl., JJU.C.A. 9, No. 2
110-114, August K)59.
/
Curve/ Determine the
response at the
temperature. Pre-
•e and check the curve
Instrument. Introduce
zero control to
MSA LIRA Infraredydas and Liquid Ana-
lyzer Instruction Bodk, Mine Safety Appli-
ances Co., Pittsburgh. Fa.
Beckman Instruction 1G35B, Models 215/
315A and 415AJuu*rareu Analyzers. Be
Instrument Company, Fullerton, Calif.
Continuous CO Monitoring System,
A 5611, In&rtech Corp., Princeton, :
X—UNOR Infrared Gas Analyzers.
Ronce/erte, W. Va.
[36/
22384, Nov. 25. 1971,
'.1043. Feb. 18.19751
amended at
APPENDIX D—MEASUREMENT PRINCIPLE AND
CALIBRATION PROCEDURE FOR THE MEASURE-
MENT OP PHOTOCHEMICAL OXIDANTS COR-
RECTED FOR INTERFERENCES DUE TO NITRO-
GEN' OXIDES AND SULFUR DIOXIDE.
1.1 Ambient air and etbylene are de-
livered simultaneously to a mixing zone
where the ozone In the air reacts with the
ethylene to emit light which Is detected by
a photomultlplier tube. The resulting photo-
current Is amplified and Is either read di-
rectly or displayed on a recorder.
1.2 An analyzer based on this principle
will be considered a reference method only
If it has been designated as a reference
method in accordance with Part 53 of this
chapter.
2.—I. |Reserved]
5. Apparatus.
5.1—5.9 [Reserved]
5.10 Apparatus for Calibration
5.10.1 Absorber, All-glass implngera as
shown in Figure D4 are recc nmended. The
implngers may be purchased from most ma-
jor glassware suppliers. Two absorbers in
series are needed to Insure complete collec-
tion of the sample.
5.10.2 Air Pump. Capable of drawing 1
liter/minute through the absorbers. The
pump should be equipped with a needle valve
on the inlet side to regulate flow.
5.10.3 Thermometer. With an accuracy
of ±2« C.
5.10.4 Barometer. Accurate to the nearest
Tntn Hg.
5.10.5 Flawmeter. Calibrated metering de-
vice for measuring flow up to 1 liter/minute
within ±2 percent. {For measuring flow
through Implngers.)
5.10.6 Floiameter. For measuring airflow
past the lamp; must be capable of measuring
flows from 2 to 15 liters/minute within ±5
percent.
5.10.7 Trap. Containing glass wool to pro-
tect needle valve.
6.10.8 Volumetric Flasks. 35, 100, 600.
1.000 ml.
5.10.9 Buret. 50 ml.
5.10.10 Plpets. 0.5. 1, 2, 3. 4. 10, 25, and
SO ml. volumetric.
5.10.11 Erlenmeyer Flasks. 300 ml.
6.10.12 Spectrophotomcter. Capable of
measuring absorbance at 352 nm Matched
1-cm. cells should be used.
6. Reagents.
6.1 Ethylene. C. P. grade (minimum).
36
-------
§50.11
Title 40—Protection of Environment
6.2 Cylinder Air. Dry grade.
6.3 Activated Charcoal Trap. For filtering
cylinder air.
6.4 Purified 'Water. Used for nil reagents.
To distilled or delonlzed water In an all-glass
distillation apparatus, add a crystal of potas-
sium permanganate and a crystal of barium
hydroxide, and redistill.
6.5 Absorbing Reagent. Dissolve 13.6 g.
potassium dlhydrogeu phosphate (KH.PO.).
14.2 g. anhydrous disodlum hydropen phos-
phate (NaJtPO,) or 35.8 g. dodecahydrate
salt (Na.HPO412H._,O), and 10.0 g. potassium
iodide (KI) In purified water and dilute to
1.000 ml. The pH should be 6.8 ±0.2. The
solution Is stable for several weeks. If stored
In a glass-stoppered amber bottle in a cool.
dark place.
6.6 Standard Arseniovs Oxide Solution
(0.05 N). Use primary standard grade arse-
nious oxide (As,O,). Dry 1 hour at 105* C.
Immediately before using. Accurately weigh,
to the nearest 0.1 mg., 2.4 g. arsenlous oxide
from a small glass-stoppered weighing bottle.
Dissolve in 25 mL 1 N sodium hydroxide In a
-Bask or beaker on a steam bath. Add 25 ml.
1 N sulfurtc acid. Cool, transfer quantita-
tively to a 1.000-ml. volumetric flask, and
dilute to volume. NOTE: Solution must be
-neutral to litmus, not alkaline.
Normality Ia= •
ml. AssO»X Normality As:O«
ml. Ii
Normality AsaOs=-
wt As,0, (g.)
49.46
6.7 Starch Indicator Solution (0.2 per-
cent). Triturate 0.4 g. soluble starch and ap-
proximately 2 ing. mercuric Iodide (preserva-
tive) with a little water. Add the paste slowly
to 200 ml. of boiling water. Continue boiling
until the solution is clear, allow to cool, and
transfer to a glass-stoppered bottle.
6.8 Standard Iodine 'Solution (0.05 N).
6.8.1 Preparation. Dissolve 5.0 g. potas-
sium iodide (KI) and 3.2 g. resublimed Iodine
(Ij) in 10 ml. purlfled water. When the iodine
•dissolves, transfer the solution to a 500-ml.
glass-stoppered volumetric flask. Dilute to
mark with purified water and mix thor-
oughly. Keep solution in a dark brown gloss-
stoppered bottle away from light, and re-
standardize as necessary.
6.8.2 Standardization. Pipet accurately 20
ml. standard arsenlous oxide solution into a
300-ml. Erlenmeyer flask. Acidify slightly
with 1:10 sulfurlc acid, neutralize with solid
sodium bicarbonate, and add about 2 g. ex-
cess. Titrate with the standard Iodine solu-
tion using 5 ml. starch solution as indicator.
Saturate the solution with carbon dioxide
near ths end point by adding 1 ml. of 1:10
.sulfuric acid. Continue the tltratlon to the
first appearance of a blue color which per-
sists for 30 seconds.
6.9 Diluted Standard Iodine. Immediately
before use, plpet 1 ml. standard Iodine solu-
tion into a 100-ml. volumetric flask and
dilute to volume with absorbing reagent.
7. Procedure.
7.1 Instruments can be constructed from
the components given here or may be pur-
chased. If commercial Instruments are used,
follow the specliic Instructions given In the
manufacturer's manual. Calibrate the In-
strument as directed In section 8. Introduce
samples Into the system under the same con-
ditions of pressure and flow rate as are used
In calibration. By proper adjustments of zero
and span controls, direct reading of ozone
concentration is possible.
8. Calibration.
8.1 KI Calibration Curve. Prepare a curve
of absorbance of various iodine solutions
against calculated ozone equivalents as
follows:
8.1.1 Into a series of 25 ml. volumetric
flasks, plpet 0.5, 1, 2, 3. and 4 ml. of diluted
standard iodine solution (6.9). Dilute each
to the mark with absorbing reagent. Mix
thoroughly, and immediately read the ab-
sorbance of each at 352 rim. against unex-
posed absorbing reagent as the reference.
8.1.2 Calculate the concentration of th«
solutions as total fig. O> as follows:
Total flg. 03= (N) (96) (V,)
N=Normallty L (see 6.8.2), meq./ml.
V,—Volume of diluted standard I, added,
ml. (0.5,1,2.3,4).
Plot absorbance versus total jig. Or
8.3 Instrument Calibration.
8.2.1 Generation of Test Atmospheres. As-
semble the apparatus as shown in Figure D3.
The ozone concentration produced by the
generator can be varied by changing the po-
sition of the adjustable sleeve. For calibra-
tion of ambient air analyzers, the ozone
source should be capable of producing ozone
concentrations in the range 100 to 1,000
Ag./m.1 (0.05 to 0.5 p.pjn.) at a flow rate ot
at least 5 liters per minute. At all times the
airflow through the generator must be great-
er than the total flow required by the sam-
pling systems.
8.2.2 Sampling and Analyses of Test At-
mospheres. Assemble the KI sampling train
as shown In Figure D4. Use ground-glass
connections upstream from the Implnger.
Butt-to-butt connections with Tygon tubing
may be used. The manifold distributing the
test atmospheres must be sampled simul-
taneously by the KI sampling train and the
Instrument to be calibrated. Check assem-
bled systems for leaks. Record the Instru-
ment response in nanoamperes at each
concentration (usually six). Establish these
concentrations by analysis, using the neu-
37
-------
Chapter I—Environmental Protection Agency
§50.11
tral buffered potassium Iodide method as
follows:
8.2.2.1 Blank. With ozone lamp off, flush
the system lor several minutes to remove
residual ozone. Plpet 10 ml. absorbing re-
agent Into each absorber. Draw air from the
ozone-generating system through the sam-
pling train at 0.2 to 1 liter/minute Tor 10
minutes. Immediately transfer the exposed
solution to a clean 1-cm. cell. Determine the
Bbsorbance at 352 run. against unexposed
absorbing reagent as the reference. If the
system blank gives an absorbance. continue
flushing the czone generation system until
no abscrbance Is obtained.
8.2.2.2 Test Atmospheres. With the ozone
lamp operating, equilibrate the system for
about 10 minutes. Plpet 10 ml. of absorbing
reagent Into each absorber and collect sam-
ples for 10 minutes In the concentration
range desired for calibration. Immediately
transfer the solutions from the two absorb-
ers to clean 1-cm. cells. Determine the ab-
sorbance of each at 352 nm. against unex-
posed absorbing reagent as the reference. Add
the absorbances of the two solutions to ob-
tain total absorbance. Read total /ig.Oi from
the calibration curve (see 8.1). Calculate to-
tal volume of air sampled corrected to ref-
erence conditions of 25* C. and 760 mm. Hg.
as follows:
P 298
Va=VX X X10-*
760 t+273
V» = Volume of air at reference condi-
tions, m."
V =Volume of air at sampling condi-
tions, liters.
P = Barometric pressure at sampling
conditions, mm. Hg.
t =Temperature at sampling conditions,
"C.
10-»= Conversion of liters to m.1
Calculate ozone concentration In p.pjn. as
follows:
p.pja.O>=
X6.10X10-*
Vm
8.2.3 Instrument Calibration Curve. In-
strument response from the photomulUpller
tube Is ordinarily in current or voltage. Plot
the current, or voltage If appropriate,
(y-axls) for the test atmospheres against
ozone concentration as determined by the
neutral buffered potassium Iodide method.
In p.p.m. (x-axls).
9. Calculations.
9.1 If a recorder Is used which has been
properly zeroed and spanned, ozone concen-
trations can be read directly.
9.2 If the DC amplifier Is read directly,
the reading must be converted to ozone
concentrations using the Instrument calibra-
tion curve (8.2.3).
9.3 Conversion between p.pjn. and
-------
§ 50.11
Title 40—Protection of Environment
SAMPLE AIR IN
EXHAUST
6mm
ETHYLENE IN
6mm
I"
2 mm
>— 10 mm
— 6 mm
2 mm
PYREX CONSTRUCTION
-
lo
-in O.D.
PHOTOMULTIPLIER TUBE
EPOXY SEALED OPTICALLY FLAT
PYREX WINDOW ON END
Figure D1. Detector cell.
[36 PR 22384, Nov. 25,1971, as amended at 40 FR 7043, Feb. 13,1975]
39
-------
$ 50.11
6-in. PEN-RAY
LAMP
Chapter I—Environmental Protection Agency
»
S!m»v6»»>:^jpgpB!<^^
Figure 02. Ozone source.
FLOW METER
(0-10 liters/min)
5 litersynin
NEEDLE
VALVE
FLOW
CONTROLLER
MICRON
FILTER
CYLINDER
AIR
OZONE
SOURCE
VENT
SAMPLE
tLJtLJtl
MANIFOLD
Figure D3. Ozone calibration air supply, source, and
manifold .system.
40
-------
§50.11
Title 40—Protection of Environment
RUBBER TUBING
FLOWMETER
Figure 04. Kl sampling train.
Time—The time Ihteafel '&6m"a"iUp-
A. Suggested Performance Specifications
for Atmospheric jmalyzers for Hydrocarbo
Corrected for Methane:
Range (:
turn) _-
Minimum detectable sen-
• sitlvity. ,
Zero drift (maximum) ___
(
Span drift (maximum)
Precision (minimum) —
Operational period (mini-
mum).
Operating temperature
range (minimum).
Operating humidity range
(minimum).
Linearity (maximum) —y
03 mg./m
p.pjn.)
0-3 mg/in.' (0-5
p.p>£.) CH4
0-ljy mv. full
scale.
/6.1 p.pon. THC.
0.1 p.p.m. CH4.
Not to exceed
1 percent/24
hours.
Not to exi
1 percent.
hours.
±0.5 peri
3 days.
0 percent.
1 percent of full
scale.
B. Suggested Definitions of Performance
Specifications:
Range—The minimum and maximum measX
urement limits. /
Output—Electrical signal which Is propor-
tional to the measurement; intended for
connection to readout or data processing
devices. Usually expressed as millivolts or
mllliamps full scale at a given Jcnpedence.
Full Scale—The maximum measuring limit
for a given range.
Minimum Detectable Sensitivity—The small-
est amount of Input concentration that
can be detected as the concentration ap-
proaches zero. ,
Accuracy—The degree of agreement between
a measured vaXie and the true value; usu-
ally expressed at ± percent of full seal*.
41
change in Input concentration at the In-
strument Inlet to tbVfirst corresponding !
change In the Instrument output.
Time to 90 Percent/Response—The time In-
terval from a step change in the Input
centratlon at/The Instrument Inlet tof a
reading of 90 percent of the ultlma^rre-
corded concentration.
Rise TimX (90 percent)—The intaAral be-
tweegnnltlal response time and/Cme to 90
percent response after a step/decrease In
the Inlet concentration.
Zjwo Drift—The change In instrument output
, over a stated time period: usually 24 hours.
of unadjusted continuous operation, when
the Input concentration is zero; usually
expressed as percent full scale.
Span Drift—The orfange In instrument out-
put over a staged time period, usually
hours, of unadjusted continuous opera
when the input concentration Is a B
upscale yflue; usually expressed as nfrcent
full
Precision—The degree of agreement between
repeated measurements of thy'eame con-
centration. It Is expressed a* the average
deviation of the single results from the
• mean.
Operational Period—The/period of time over
which the Instrument can be expected to
operate unattended%lthln specifications.
Noise—Spontaneous deviations from a mean
output not caused by Input concentration j
changes.
Interference-/^ undeslrod positive or t&jf- ,
tlve output caused by a substance qlher •
than the one being measured. /
Interference Equivalent—The portion or in-
dicated input concentration date to the
presence of an Interferent. .< ;
Operating Temperature Range^-The range of
ambient temperatures ovsf which the In-
strument will meet all performance specifi-
cations.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/4-77-005
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
EVALUATION OF 1 PERCENT NEUTRAL BUFFERED POTASSIUM
IODIDE PROCEDURE FOR CALIBRATION OF OZONE MONITORS
5. REPORT DATE
January 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
M. E. Beard, J. H. Margeson and E. C. Ellis
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Research and Development
Environmental Monitoring and Support Laboratory
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
1HD621
11. CONTRACT/GRANT NO.
12.SPQNSORING AGENCY NAME AND ADDRESS
Environmental Mom ton ng and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16.
percent neutral buffered potassium iodide procedure, as specified
for calibration of ozone monitors in the Environmental Protection Agency (EPA)
reference method for measurement of photochemical oxidants, was evaluated.
The procedure was found to be vulnerable to error in three major areas:
1) the time allowed for measurement of iodine absorbance, 2) purity of
potassium iodide reagent, and 3) type of impinger used. Variations in results
produced by minor changes within the specifications of the procedure suggest
that the method is difficult to control.
Improved specifications and procedures were documented to minimize the
effect of these parameters on results and to aid the user in proper use of
the procedure. The improved procedure shows a positive bias of 7.5 +_ 3.4
percent.
The Environmental Monitoring and Support Laboratory (EMSL), EPA, has
decided that further work to improve the procedure is not warranted and has
initiated work on a program to evaluate candidate procedures to replace the
neutral buffered potassium iodide procedure.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Ozone
Calibrating
lodometry
Reference Method
Gas Phase Titration
Ozone-Ethylene
Chemi1umi nescence
Ozone-Nitric Oxide
Chemi1umi nescence
13B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
JkSS (This Report)
21.
. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
42
------- |