-------
ACKNOWLEDGMENT
The author wishes to thank Michael E. Beard of the Methods
Standardization and Performance Evaluation Branch, EPA, for his
assistance in rewriting the chemiluminescence procedure and for his
helpful suggestions during the laboratory portion of the evaluation.
m
-------
CONTENTS
Page
ABSTRACT v
CONCLUSIONS iv
I INTRODUCTION 1
II EXPERIMENTATION 3
III RESULTS AND DISCUSSION 6
IV REFERENCES 16
APPENDIX A. TENTATIVE METHOD FOR CONTINUOUS MEASUREMENT OF
NITROGEN DIOXIDE IN ATMOSPHERE (CHEMILUMINESCENCE
PROCEDURE) 20
APPENDIX B. EQUATION FOR CALCULATION OF EFFECT OF EACH
VARIABLE IN RUGGEDNESS TEST 40
APPENDIX C. RUGGEDNESS TEST DATA 41
LIST OF FIGURES
Figure
1 Comparison of NO? Generated by Gas Phase Titration and
by Permeation Tubes to Chemiluminescence Analyzer's
N02 Response 9
LIST OF TABLES
Table
1 Ruggedness Test Format 12
2 Ruggedness Test Format for Calibration of N02 Analyzers
by Gas Phase Titration 16
3 Summary of Ruggedness Test Results,
iv
-------
ABSTRACT
A detailed method write-up describing the chemiluminescence
procedure for the continuous measurement of nitrogen dioxide (NOg)
in ambient air was developed. Atmospheric concentrations of N0~
are measured indirectly by the chemiluminescent reaction of nitric
oxide (NO) with ozone (03). The NCL is first thermally reduced to
NO before it is reacted with 03-
The reliability of measurements made by a continuous or instru-
mental sampling method is strongly affected by its calibration. In
the chemiluminescence procedure, an NOo analyzer is calibrated by the
gas phase titration of NO with 0^ that produces N0« stoichiometrically.
Significant errors can be introduced into the analyzer calibration if
the standard NO cylinder is incorrectly assayed for NO as well as trace
N02 or if the converter efficiency of the analyzer is always assumed
to be equal to 1.0. Provisions are included in the method to eliminate
such errors.
The gas phase titration calibration procedure was subjected to a
ruggedness test. The results indicate that normal variations in such
factors as reaction and mixing chamber volumes, ratio of dilution
air flow to flow through the ozone generator, use of different ozone
generators with different dilution air flows and use of different
standard NO levels had little effect on the gas phase titration
procedure.
-------
CONCLUSIONS
The results of a ruggedness test have shown that the gas phase
titration procedure for calibrating N0« chemiluminescence analyzers
is insensitive to normal variations in operating parameters. Close
agreement between NCL analyzer responses to the same N02 concentra-
tions generated by both gas phase titrations and gravimetrically
calibrated permeation devices indicates that gas phase titration
is also an accurate calibration procedure.
vi
-------
EVALUATION OF GAS PHASE
TITRATION TECHNIQUE AS USED
FOR CALIBRATION OF NITROGEN DIOXIDE
CHEMILUMINESCENCE ANALYZERS
I. INTRODUCTION
It is the responsibility of the Methods Standardization and
Performance Evaluation Branch (MSPEB) of the Quality Assurance and
Environmental Monitoring Laboratory (QAEML) to standardize and to
evaluate various methods for measuring air pollutants to determine
their utility. The standardization process usually includes:
(1) a review of the procedure write-up to insure that it is clearly
written and technically accurate, (2) a laboratory evaluation to
determine if the procedure will perform according to the write-up
specifications and also if it is flexible enough to withstand minor
changes in sampling parameters, i.e., ruggedness testing^ and
(3) a collaborative test to determine the method's precision and
accuracy. The MSPEB's evaluation (items 1 and 2 above) of the con-
2
tinuous chemiluminescence procedure for N02 is the subject of this
report. The report details the results of a procedure review and a
laboratory evaluation of the calibration procedure, including a
ruggedness test. A collaborative test of the N02 chemiluminescence
1
-------
procedure has been performed under contract to MSPEB, and these
results will be reported elsewhere.
The evaluation of a continuous or instrumental sampling method
is different from that of a manual method. For example, the chemi-
luminescence procedure for N02 does not and should not specify the
analyzer design or its operating parameters. It specifies only
that the analyzer operate on the chemiluminescence principle for
detection of NC^ (as NO) and that it meet certain performance specifi-
cations. The only section of the method not inherent to the analyzer
and its operation is the calibration procedure. This evaluation of the
chemiluminescence procedure is in essence an evaluation of the cali-
bration procedure only because it would'not be meaningful to evaluate
how well one particular analyzer makes use of the detection principle
without doing the same for other N02 chemiluminescence analyzers.
-------
II. EXPERIMENTATION
A copy of "The Tentative Method for the Continuous Measurement
of Nitrogen Dioxide in the Atmosphere (Chemiluminescence Procedure)"
is included in Appendix A for easy reference. The reagents and
apparatus used in the laboratory evaluation of the qas phase titration
calibration technique are described therein.
A. APPARATUS
1. N02 Chemiluminescence Analyzer
A Bendix Model 8101-B Oxides of Nitrogen Analyzer that employs
the chemiluminescent principle for N02 detection was used in the
laboratory evaluation. This analyzer meets the performance specifi-
cations described in the method.
2. Ozone Analyzer
A Dasibi Model 1003 Ozone Monitor, calibrated by the neutral
buffered KI procedure,3 was used to monitor unreacted ozone during
gas phase titrations.
3. Gas Phase Titration Assembly
The assembly is described fully in Section 6.2 of Appendix A.
B. TEST ATMOSPHERE GENERATION
Test atmospheres containing known concentrations of N02, NO, and
03 were necessary for the laboratory study. The details for generation
of a standard atmosphere for each gas are outlined below. In each
case the dilution gas was compressed house air. It was purified by
passage through air filters for particle removal; treatment with 0
-------
for conversion of NO to NOg; silica gel for drying; and finally a
mixture of activated charcoal (6-14), molecular sieve (6-16 mesh,
type 4A), and silica gel (6-16 mesh) for removal of any NOp, hydro-
carbons, or ozone.
1. Nitrogen Dioxide
Test atmospheres containing known concentrations of N02 were
produced by the gas phase reaction of NO with 0^. Such test atmospheres
were used to calibrate the N0« response of the chenriluminescence analyzer
and to evaluate the calibration procedure. The analyzer calibration
was checked with known NOp concentrations generated from gravimetrically
calibrated NOg permeation devices.4'8
2. Ozone
9 10
Two ozone generators, both of identical design and con-
struction, were used to generate known ozone concentrations for gas
phase titration. Each generator consists of a quartz tube into which
ozone-free air is introduced and then irradiated with a stable low-
pressure mercury lamp. The level of irradiation is controlled by an
adjustable aluminum sleeve which fits around the lamp; ozone concen-
trations are varied by adjustment of this sleeve. At a fixed level of
irradiation and air flow, ozone is produced at a constant rate.
Ozone generator #14 generates ozone in the range 0-1.0 ppm at
a dilution air flow of 2.91 £/min. Generator #8 produces the same
range of ozone concentrations at a dilution flow of 5.02 £,/min. Both
generators were calibrated by the neutral buffered potassium iodide
(KI) procedure.3
-------
3. Nitric Oxide
Two cylinders of nitric oxide in nitrogen were used to
generate known concentrations of NO by flow dilution. Both cylinders
were assayed, as described in the method, by gas phase titration
with 03 from the calibrated ozone generators discussed above.
Cylinder A contains 114 ppm NO and cylinder B contains 43.0 ppm NO.
The cylinders were also analyzed for N02 impurity using the TGS
procedure ; cylinders A and B contain 1.4 and 3.0 ppm N02,
respectively.
C. FLOW MEASUREMENT
Dilution flow measurements were made with a 1 ji/rev wet test
o
meter. The NO flow was measured with a 10 cm soap bubble flow
meter. A 100 cm soap bubble flow meter was used to measure the
air flow through the ozone generators and also to measure the sample
flow of the NOp chemiluminescence analyzer. All flow measurements were
made at the beginning of each experiment, then again at the end. No
flows varied more than 2 percent from the beginning to the end of any
experiment.
-------
III. RESULTS AND DISCUSSION
A. PROCEDURE REVIEW
Improvements were made on the "Tentative Method for the Continuous
Measurement of Nitrogen Dioxide (Chemiluminescent)" published in the
Federal Register.2 The method write-up was generalized in its dis-
cussion of the chemiluminescent principle and in its directions for
analyzer adjustments during calibration so that the method would be
more applicable to any chemiluminescent NC^ analyzer. Additions were
made to the calibration procedure to minimize the sources of experi-
mental error. A brief discussion of the calibration procedure seems
appropriate at this point.
The calibration technique is based upon application of the rapid
gas phase reaction between NO and 0* to produce a stoichiometric
quantity of N02.12'13
NO + 03-*N02 + 02 k = 1.0 x 107 liter mole"1 sec"1 (1)
The quantitative nature of the reaction is used in a manner such
that, once the concentration of one component is specified, the
concentrations of the other two are determined. In this application
the NO is always kept in slight excess to the 0- concentration.
An assayed NO in N2 cylinder (50 - 100 ppm) is used to generate
known concentrations of NO as well as N02 for the respective cali-
brations of the NO and N02 detector responses of a chemiluminescence
analyzer. First the NO response is calibrated directly by an
-------
appropriate dilution of the NO gas from the assayed cylinder. Then,
to calibrate the N02 response of the analyzer, 03 is added in incre-
ments to excess NO. The incremental decrease in NO concentration
(observed by noting the NO detector response of the analyzer) is
equivalent to the N0« produced in the titration. The continuous
production of N02 in the titration provides a dynamic calibration.
One possible source of error in the calibration procedure is the
NO cylinder analysis. It is necessary to analyze the cylinder for both
NO and N02 because the-nominal NO concentration of some cylinders has
been found to be inaccurate, and some cylinders have been found to con-
tain as much as 5-10 percent of the nominal NO as N02- The NO content
of the cylinder is assayed by gas phase titration with O,. Mixtures
containing known 0^ concentration are produced by an ozone generator
2
that has been calibrated by iodometry (neutral buffered KI procedure).
The amount of Og added in this titration is equivalent to the amount of
NO consumed; thus the NO content of the cylinder is determined. Once
the cylinder is assayed, it can be used reliably for up to 6 months
(EPA laboratory findings) to prepare known concentrations of NO.
Since there is always the possibility that N02 may be present in
trace amounts in the NO cylinder, it is necessary to analyze the NO
cylinder for any N02 impurity. The concentration of N02 in the
cylinder must be determined by an independent method which has no
appreciable interference from NO; the triethanolamine-guaiacol-
sulfite (TGS) manual method is one such method. .
Another possible source of error in the calibration procedure
is the tacit assumption that the N02 converter efficiency of the
7
-------
chemiluminescence analyzer is always constant and equal to 1.0, i. e.,
one mole of NOp gas is thermally converted to one mole of NO gas In
the analyzer's converter prior to detection. This assumption is
not always valid; therefore, the NOp converter efficiency must be
determined prior to analyzer calibration. A scheme for determining
the N02 converter efficiency is included in the calibration procedure.
The efficiency of the N02 converter should be constant over the range
of operation, and for acceptable analyzer sensitivity the converter
should be replaced or rejuvenated should its efficiency drop below
0.9 (90 percent).
B. PRELIMINARY LABORATORY EVALUATION
The Bendix chemiluminescent N0« analyzer was calibrated as
described in the procedure (Appendix A) by gas phase titration of
NO with Og. The NO was generated from cylinder A (114 ppm NO + 1.4 ppm
NOp), and the CL source was generator #14. No ozone was detected by
the Disabi ozone monitor throughout the titration. The analyzer
NOp calibration was verified by checking the analyzer response against
an NOp standard generated from gravimetrically calibrated NOp permea-
tion tubes. A comparison of the analyzer's response to NOp generated
by gas phase titration and by permeation tubes is shown in Figure 1.
The two curves are indistinguishable. The least squares linear
regressions for the two curves are
A: y = 0.9998 x -0.0004
B: y = 0.9871 x +0.0018 .
8
-------
0.50
0.40
* 0.30
" 0.20
o
0.10
0.00
T
0.00
T
D -A: N02 GENERATED BY GAS PHASE TITRATION
O -B: N02 GENERATED FROM PERMEATION TUBES
0.10
0.20 0.30
N02 ANALYZER RESPONSE, ppm
0.40
O.SO
Figure 1. Comparison of NO2 generated by gas phase titration and by permeation
tubes to a chemiluminescence analyzer's 1M02 response.
-------
The gas phase titration was repeated, and this calibration was
identical to the first. Since there were no obvious problems
involved in the calibration of the one NCL analyzer by the gas
phase titration procedure, no single variable evaluations seemed
necessary and the ruggedness test was initiated.
C. RUGGEDNESS TEST
1. Theory
How is a method or procedure affected by slight changes in
operational parameters that might be expected to occur in normal
usage? To answer this question, Youden1 has described a scheme
called a ruggedness test for determining the individual effects of
variations in several parameters with a minimum number of experiments.
In a ruggedness test a series of experiments are conducted in which
selected parameters are studied at two levels: the nominal level
stated in the method write-up and a challenging level. The effect
of a particular variable is determined by comparing the test results
obtained at the two levels. Therefore, the parameters which signifi-
cantly affect the method's response, once identified, can be more
stringently controlled.
In Youden's scheme seven'3' parameters are chosen for testing,
and their nominal values are denoted with the letters A to G. A
challenging level for each parameter is then selected and denoted by
the corresponding lower case letters, a to g. A series of eight
^'Schemes for examining the effect of a larger or smaller number
of variables are available.i*U5
10
-------
experiments are then conducted using various combinations of either
nominal or challenging values for each variable. The format for
these experiments is shown in Table 1.
Each of the eight experiments produces a result, denoted as
s, t, u, v, w, x, y, and z. Examination of the format in Table 1
reveals that by summing the group of experimental results in which
a given nominal value is involved, and subtracting from it the sum
of the group of results in which the corresponding challenging
value is involved, the effect of all other variables is cancelled.
If the grouped results are divided by the number in each group,
the above subtraction will yield the average effect or difference
between the nominal and challenging conditions. For example, the
average effect of variable C is calculated as follows:
r s + u + w + y t + v + x + 2
t - c - 4 - 4
A complete set of equations for calculating the effect of A-a, B-b,
etc. are given in Appendix B.
It should be noted that the results of a ruggedness test
will not describe completely the effect of varying a given parameter.
The results will only show an effect due to the range of variation
used in the test. If an effect is shown by the ruggedness test to
be significant, further experimentation will be necessary to define
the exact relationship between variation of the parameter and the
method's response. It should also be noted that the various parameters
are assumed to be mutually independent.
11
-------
Table 1. RU6GEDNESS TEST FORMAT
Factor
value
A or a
B or b
C or c
D or d
E or e
F or f
G or g
j
Observed
result
Determination number
1
A
B
C
D
E
F
G
s
2
A
B
c
D
e
f
g
t
3
A
b
C
d
E
f
g
u
4
A
b
c
d
e
F
G
V
5
a
B
C
d
e
F
g
w
6
a
B
c
d
E
f
G
X
7
a
b
C
D
e
f
G
y
8
a
b
c
D
E
F
g
z
12
-------
2. Design
The ruggedness test described by Youden provides for evalua-
tion of seven variables. When only five or six variables seem
important, then dummy variables can be substituted for the remaining
variables. The effect from a dummy variable should be small. If
the effect of varying a dummy variable should give a large result,
some explanation should be sought. In the case of the gas phase
titration calibration procedure, five real parameters are tested;
the remaining two are labeled as dummies.
a. Dummy (No variation)
b. Reaction Chamber Volume
A kjeldahl connecting bulb with a volume of approxi-
3
mately 250 cm is used for the NO + 0., reaction chamber. Reduc-
3
tion of reaction chamber volume to 125 cm should cut down on the
residence time for the two gases in the chamber and might affect the
completeness of the reaction.
c. Mixing Chamber Volume
3
The mixing chamber is also a 250 cm kjeldahl bulb.
3
A reduction of the mixing chamber volume to 125 cm might have some
affect on the homogeneous mixing of the NO and N02 (from NO + 03)
calibration mixture with the dilution air.
d. Flow Ratio
The air stream is split upstream of the ozone generator
to allow only 1/10 to 1/15 of the total air flow to go through
13
-------
the ozone generator, with the remainder bypassing the generator to
the mixing chamber. The air stream is split in order to permit low
flows through the ozone generator to create locally high concentra-
tions of 03 and NO in the reaction chamber. At these high concentra-
tions and low flows, the reaction chamber provides sufficient resi-
dence time for quantitative reaction to occur. The nominal flow
ratio of 10:1 is challenged by a ratio of 5:1.
e. Ozone Generator
The calibration procedure specifies the use of an ozone
source capable of producing 0 to 1.0 ppm 03 at a total air flow
of 2 to 10 £/min. Ozone generator #14 with a total air flow of
2.91 a/min was used in calibrating the N02 chemiluminescence analyzer;
therefore, it is designated as the nominal generator. The challenging
ozone source is generator #8 with a total air flow of 5.02 ^/min.
f. Level of NO in Standard Cylinder
The procedure suggests either a 50 or a 100 ppm NO
cylinder for the NO calibration standard. Since the most commonly
used NO cylinder concentration is about 100 ppm, the nominal NO
standard level was set at about 100 ppm. The approximately 50 ppm
cylinder is the challenging level. The NO flow from a 50 ppm cylinder
into the reaction chamber would be twice the flow from the 100 ppm
cylinder needed to generate comparable NO atmospheres. The higher
flow from the 50 ppm cylinder might decrease the residence time of
the NO in the reaction chamber and, in turn, affect the completeness
of the NO + 03 reaction.
g. Dummy (No variation)
14
-------
3. Procedure
The selected parameters were incorporated into the rugged-
ness test format as shown in Table 2. Experiments were conducted in
random order (2, 6, 7, 1, 3, 5, 4, and 8) using the previously cali-
brated N02 chemiluminescence analyzer. Zero and span checks were made
prior to each experiment, and the points were always within 1 percent
of the original analyzer calibration. In each experiment the N0«
analyzer response was noted as the N02 concentration generated by
gas phase titration was varied from 0 to 0.5 ppm in 0.1 ppm incre-
ments. The result of each gas phase titration experiment is
expressed as the slope of the'calibration curve, i.e., NO^ concen-
tration generated vs. N02 analyzer response. All the test results
are given in Appendix C.
4. Results
The ruggedness test results listed in Appendix C were sub-
stituted into the equations of Appendix B, and the effect of each
parameter was calculated. A summary of the ruggedness test results are
shown ranked in order of absolute magnitude in Table 3. The largest
difference measured is only 2.4 percent. Since the errors involved in
reading the chemiluminescence analyzer and in generating standard NO,
N02, and 03 atmospheres are at least 2 percent, the differences measured
by the ruggedness test are insignificant.
15
-------
Table 2. RUG6EDNESS TEST FORMAT FOR CALIBRATION OF
N02 ANALYZERS BY GAS PHASE TITRATION
Factor value
Dummy
A = a, No variation
Reaction Chamber Volume
B = 250 cm3 b = 125 cm3
Mixing Chamber Volume
C = 250 cm3 c = 125 cm3
Flow Ratio
D = 10:1 d = 5:1
Ozone Generator & Air Flow
E = ~3.0K,/min e = ~5.0 A/min
Level of NO in Standard
Cylinder
F = ~100 ppm f = ~50 ppm
Dummy
G = g, No variation
Observed Result
Determination number
1
—
250 cm3
250 cm3
10:1
3.0 A/min
100 ppm
— — —
s
2
/
250
125
10:1
5.0
50
___
t
3
—
125
250
5:1
3.0
50
- — —
u
4
—
125
125
5:1
5.0
100
---
V
5
250
250
5:1
5.0
100
---
w
6
—
250
125
5:1
3.0
50
---
X
7
—
125
250
10:1
5.0
50
___
y
8
—
125
125
10:1
3.0
100
___
z
-------
Table 3. SUMMARY OF RUGGEDNESS TEST RESULTS
Rank Factor Difference. %
1 Reaction Chamber Volume (-) 2.39
250 cm3 vs. 125 cm3
2 Dummy 1.61
3 Flow Ratio (-) 0.68
10:1 vs. 5:1
4 Mixing Chamber Volume j 0.48
250 cm3 vs. 125 cm3
5 Level of NO in Standard Cylinder 0.42
~100 ppm vs. ~ 50 ppm
6 Ozone Generator (-) 0.34
Air Flow ~3.0 fc/min vs. ~5.0 a/min
7 Dummy (-) 0.12
17
-------
IV. REFERENCES
1. Youden, W. J. Statistical Techniques for Collaborative Test.
Association of Official Analytical Chemists, Washington,
D. C. 20044. p. 29-32. 1967.
2. Title 40-Protection of Environment. Federal Register. 3j3:
15174, June 8, 1973.
3. Title 40-Protection of Environment. Federal Register. 36:
22384, November 25, 1971.
4. O'Keeffe, A. E. and G. C. Ortman. Anal. Chem. 38: 760, 1966.
5. Scaringelli, F. P. et al. Amer. Indus. Hyg. Ass. J.28: 260, 1967,
6. Scaringelli, F. P. et al. Anal. Chem. 42_: 871, 1970.
7. National Bureau of Standards Technical Note 585. National
Bureau of Standards, Washington, D. C. January 1972. p. 26.
Available from: Superintendent of Documents, Government
Printing Office, Washington, D. C. 20402.
8. Rook, H. L. e t al. Operation Characteristics of N0« Permeation
Devices. (Presented at 167th American Chemical Society Meeting.
Los Angeles. March 31-April 5, 1974. Paper No. 61.)
9. Hodgeson, J. A. et al. ISA Transactions. 1J.: 161, 1972.
10. National Bureau of Standards Technical Note 585. National
Bureau of Standards, Washington, D. C. January 1972. p. 11.
18
-------
11. Fuerst, R. G. and J. H. Margeson. An Evaluation of the
TGS-ANSA Method for Measurement of N02- Copies of this
document, which includes the method write-up, can be obtained
from: Methods Standardization and Performance Evaluation
Branch, Environmental Protection Agency, Research Triangle
Park, North Carolina 27711.
12. Hodgeson, J. A. et al. Anal. Chem. 43; 1123, 1971.
13. Rehme, K. A. et al. Tentative Method for the Calibration of
Nitric Oxide, Nitrogen Dioxide, and Ozone Analyzers by Gas
Phase Titration. Environmental Protection Ag'ency, Research
Triangle Park, N. C. Publication No. EPA-R2-73-246.
March 1974.
14. Plackett, R. L. and J. P. Burman. Biometrika. 33_: 305, 1946.
15 Stowe, R. A. and R. P. Mayer. Indust. Eng. Chem. 58: 36, 1966,
19
-------
APPENDIX A. TENTATIVE METHOD
FOR" CONTINUOUS MEASUREMENT OF NITROGEN DIOXIDE
IN ATMOSPHERE (CHEMILUMINESCENCE PROCEDURE)
AUGUST 1974
a
A tentative method is one that has been carefully drafted
from availabel experimental information, reviewed editori-
ally within the Methods Standardization and Performance
Evaluation Branch, EPA, and has undergone extensive labor-
atory evaluation. The method is still under investigation
and, therefore, is subject to revision.
20
-------
1. Principle and Applicability
1.1 Atmospheric concentrations of nitrogen dioxide
are measured indirectly by the chemiluminescent reaction of nitric
oxide (NO) with ozone ((L) at reduced or near atmospheric pres-
sures. NOg is first thermally converted^ (reduced) to NO
before it is reacted with 03. A photomultiplier tube detector
measures the intensity of the light from the resulting chemilumi-
nescent reaction. Since ambient air usually contains N02 and NO
simultaneously, some means must be provided for distinguishing
between the NO concentration already in the sample and the NO con-
centration resulting from the reduction of N02 to NO. Ambient air
is drawn directly into the detector assembly to determine the NO
concentration. The NOX concentration (NO + NOp) is measured when
the air is passed through the thermal converter and then into the
detector. The N02 concentration is then determined electronically
by subtracting the NO detector response (or concentration) from the
NOX detector response. The measurement is accomplished in approxi-
mately one minute.
1.2 The procedure is applicable to the continuous meas-
urement of N02 in ambient air.
2. Range and Lower Detectable Limit
2.1 Analyzers with full scale ranges from 0 to 376 yg/m3
(0 to 0.2 ppm) to 0 to 18,800 yg/m3 N02 (0 to 10 ppm) are available.
For normal ambient air measurements the most sensitive range on the
21
-------
analyzer should be used, not to exceed a sensitivity of 0 to
940 yg/m (0 to 0.5 ppm). Separate range selectors should be made
available for NO, NOg, and NOX since NOX concentrations sometimes
exceed 1 ppm.
2.2 The lower detectable limit is approximately 20
(0.01 ppm) at the 0 to 940 yg/m (0-0.5 ppm) range.
3. Interferences
3.1 The chemiluminescent detection of NO with 03 is not
subject to interference from any of the common air pollutants such
as 03, N02, CO, NH3, and S02-1 The elimination of possible hydro-
carbon interference is achieved by means of a red sharp-cut optical
filter.
3.2 Any compounds which will be converted to NO in the
thermal converter will interfere with the measurement of the NO
A
(NO + NOp) concentration. The principal compound of concern is
ammonia; however, it is not an interferent at converter temperatures
below 300° C. Unstable nitrogen compounds, such as peroxy-
acetylnitrate (PAN) and organic nitrites decompose thermally to
form NO. However, their ambient concentrations are usually so low
that their interference can be disregarded.^
4. Apparatus
4.1 Nitrogen Dioxide Analyzer. The analyzer should be of
the chemiluminescent type with a linear response over the desired
22
-------
range. The analyzer should meet or exceed the specifications
described in Appendix A-l. See Figure A-l for a general schematic of
a typical chemiluminescence analyzer.
5. Procedure
5.1 Ambient air is drawn into the analyzer with a sample
pump or with the vacuum pump used in the detector assembly. Ozonized
oxygen or air at a constant flow is drawn into the detector assembly
to react with the sample. For exact operating procedures refer to
the analyzer instruction manual.
6. Calibration
6.1 Principle. The calibration technique is based upon
application of the rapid gas phase reaction between NO and 03 to
2 "?
produce a stoichiometric quantity of N02-
NO + 03—^N02 + 02 k = 1.0 x 107 liter mole"1 sec"1 (A-l)
The quantitative nature of the reaction is used in a manner such
that, once the concentration of one component is specified, the
concentrations of the other two are determined. The qas phase
titration is carried out in the presence of excess NO to prevent
formation of higher oxides of nitrogen through further oxidation
of N02 by 03.
Atmospheres of known 0, concentration are produced
from an ozone generator, the output of which has been calibrated by
23
-------
ro
03
GENERATOR
I I
SAMPLE INLET
SOLENOID
VALVES
DETECTOR
HOUSING TEMPER!
ATURE CONTROL
HIGH VOLT-
AGE POWER
SUPPLY
OPTICAL FILTER
REACTION
CHAMBER
PHOTOMULTIPLIER
TUBE
7/////////////////////J
THERMO-ELECTRIC
COOLED HOUSING
EXHAUST
PUMP
AMP
NO
TREND
N02
TREND
r
TREND
-•NONOx
-•GROUND
Figure A-1. Automated NO, NO2, NOx chemiluminescence analyzer.
-------
iodometry (neutral buffered KI procedure).7 The NO content of an
NO in No cylinder (50-100 ppm) is then assayed by gas phase titration
with the 03- Analysis of the NO cylinder is necessary because the
nominal NO concentration provided by some manufacturers is sometimes
inaccurate. The amount of Og added in the titration is equivalent
to the amount of NO consumed (or NOp produced), thus the NO content
of the cylinder can be determined.
Once the cylinder is assayed it can be used to prepare
known concentrations of NO. The analysis should be repeated at
6-month intervals. In addition to generating standard NO concen-
trations for calibration the NO response of an analyzer, the assayed
NO cylinder is also used to generate known N0» concentrations by a
gas phase titration with 0^. To calibrate the NOg response of a
chemiluminescence analyzer, Og is added in increments to excess NO.
The incremental decrease in NO concentration (observed by noting the
NO detector response of the analyzer) is equivalent to the NOg pro-
duced in the titration. Since the NOg is produced continually in
the titration, a dynamic calibration results.
6.2 Apparatus. A schematic drawing of the gas phase titra-
tion apparatus is shown in Figure A-2.
6.2.1 Pressure regulator for NO cylinder. A regulator to fit
the NO cylinder having internal parts of stainless steel with a
Teflon or Kel-F seat and a delivery pressure of 30 psi. Note: This
regulator as well as needle valve and delivery lines should be
thoroughly flushed after connecting to the NO cylinder. Flushing
25
-------
CLEAN AIR.
SOURCE '
ro
en
NEEDLE
VALVE
FLOWMETER.
OTO lOliters/min
CAPILLARY
RESTRICTION
REACTION
CHAMBER
MIXING BULB
ANALYZED
I NO |
CYLINDER
VENT-
SAMPLE MANIFOLD
Figure A-2. Flow scheme for calibration of NO, NO2 and NOx monitor by gas phase titration.
-------
can be accomplished by adjusting the delivery pressure to near
maximum and venting the gas for several minutes.
6.2.2 Needle valves. Stainless steel parts are required in
controlling the NO flow. Material of component parts is not
critical for control of air flow.
6.2.3 Flowmeters
6.2.3.1 NO flow. A flowmeter capable of monitoring NO flows
3 3
between 0-100 cm /min. A 25 cm capacity soap bubble flowmeter is
required for measuring absolute NO flows.
6.2.3.2 Air Flow. A flowmeter capable of monitoring flows
between 0 and 10 £/min. A large soap bubble meter or wet test
meter is required for calibration of the flow meter and for making
absolute flow measurements in this range.
6.2.4 Capillary restrictor. Glass or stainless steel capillary
of the proper I.D. and length to maintain a 1:10 to 1:15 ratio of
air flow through the generator to total air flow. A Teflon or
stainless steel needle valve can be used in place of a capillary
restrictor provided that a stable flow through the generator is
maintained.
6.2.5 Calibrated Ozone Generator. Capable of producing 0^
over the range 0 to 1 ppm at a total air flow rate of 2 to 10 £/min.
One such ozone source consists of a quartz tube into which ozone-
free air is introduced and then irradiated with a stable low-pressure
O Q
mercury lamp. ' The level of irradiation of the quartz tube is
controlled by an adjustable aluminum sleeve which fits around the
27
-------
lamp. Ozone concentrations are varied by adjustment of this sleeve.
At a fixed level of air flow and irradiation, ozone is produced at
a constant rate.
6.2.6 Reaction Chamber and Mixing Bulb. Kjeldahl connecting
3
bulbs with volumes of approximately 250 cm should be used.
6.2.7 Sample Manifold. A multiport all-glass manifold should
be used. All connections in the calibration system should be glass
or Teflon.
6.3 Reagents
6.3.1 Nitric Oxide. Cylinder containing approximately 50 to
100 ppm NO in No with less than 1 ppm NOp. In ordering cylinders
from commercial vendors, specify the use of Og free N«. This will
minimize NOg formation.
6.3.2 Clean Air Source. Compressed (house) or cylinder air
containing no more than 0.005 ppm of NO, N02, and Oj, or reactive
hydrocarbons. The air is cleaned by passing it through silica gel
for drying, treating with ozone to convert any NO to N02, and finally
by passing through activated charcoal (6-14 mesh), and molecular
sieve (6-16 mesh, type 4A), to remove any NOg, hydrocarbons, and
excess Oj. Regardless of the source of air, compressed (house) or
cylinder air, it should be purified as described above.
6.4 Procedure. The gas phase titration apparatus shown in
Figure 2 is used to assay the NO cylinder and also to generate N0«
for analyzer calibration. NO from the cylinder is diluted with a
constant flow of clean air (with the ozone source set at zero 03
28
-------
concentration) to produce an NO concentration at the sample manifold
of approximately 1 ppm. The ozone source is then adjusted to produce
incremental increases in 03 concentration which result in equivalent
incremental decreases in NO concentration down to approximately
0.05 ppm NO. During this titration NOg is generated in amounts
equivalent to the NO consumed by the 03<
Upstream of the ozone generator, the air stream is split
such that the capillary restrictor allows only 1/10 to 1/15 of the
total air flow to go through the ozone generator, with the remainder
bypassing the generator. The air stream is split in order to permit
low flows through the ozone generator to create locally high con-
centrations of 03 and NO in the reaction chamber. At these high
concentrations and low flows, the reaction chamber provides a
residence time long enough for quantitative reaction to occur.
The NO-NOg mixture is combined with the bypassed air in the mixing
chamber. The turbulent flow created assures a homogeneous sample
at the sample manifold.
6.4.1 Analysis of NO cylinder by Gas Phase Titration (G.P.T.).
Allow the chemiluminescence analyzer to sample clean air until a
stable response is obtained. After the response has stabilized,
make proper zero adjustment. Determine the NO concentration in the
cylinder as follows. With the NO flow off, set the clean air flow
at approximately 2 to 10 fc/min. (the actual flow depends on the
operating characteristics of the ozone generator, Section 6.2.5).
Measure and record the absolute air flow, F . Adjust the NO flow
29
-------
to generate approximately 0.9 ppm NO at the manifold. (Use the
NO cylinder concentration provided by the manufacturer.) Measure
and record the absolute NO flow, FMQ. Adjust the span control for
NO response on a 0 to 1 ppm range so that the NO response reads
exactly the NO concentration generated. After the NO response has
stabilized, record the reading and then add approximately 0.1 ppm
\
O.j by adjustment of the calibrated ozone source. (The ozone source
should be stable before beginning the gas phase titration.) Allow
the NO response to stabilize and record the reading. (Readings
may be taken directly from the instrument readout or from a properly
spanned recorder.) Adjust the ozone source to obtain approximately
0.2 ppm 03 and allow the NO response to stabilize. Continue this
procedure until up to 0.8 to 0.9 ppm 0^ has been added in a stepwise
fashion. A minimum of six ozone additions are required to accu-
rately define the calibration curve. Remeasure FQ and F^Q to
determine whether the flows have changed during the titration.
6.4.1.1 Calculation. Plot the 0, concentration added (y-axis)
versus the resultant NO concentration in ppm (x-axis). A linear
curve should be obtained. (See Figure A-3, for example.) The 03 con-
centration at the y-axis intercept, C'Q, is the 03 equivalent to
the initial diluted NO concentration. The cylinder NO concentration
is calculated as follows:
. F0 " C'o (A.2)
N° FNO
30
-------
1.2
1.0
0.8
< 0.6
u 0.4
m
0.2
0.0
i i i r
.EQUIVALENCE POINT. C'Q
I I I
I I
0.2
0.4 0.6
NO CONCENTRATION, ppm
0.8
1.0
Figure A-3. Gas phase titration of NO with 03.
-------
where: C». = cylinder NO concentration, ppm
».Q
3
= measured NO flow, cm /min
C'Q = 0- equivalence point, ppm
3
FO = total clean air flow, cm /min
6.4.2 N02 Impurity in the NO Cylinder. The concentration of
N02 in the NO cylinder must be determined because the error in
disregarding this impurity can be significant. Consider, for example,
a 100 ppm NO cylinder with only 1 ppm of NOg. When used to generate
1 ppm NO for G.P.T., the N02 is diluted to 0.01 ppm. Although this
is a small amount of N02, the error it would introduce relative to
typical measured NOp concentrations, 0.03 to 0.05 ppm, would be greater
than 20 percent. The N02 impurity in the cylinder must be determined
by an independent method which has no appreciable interference from
NO. The triethanolamine- guaiacol-sulfite (TGS) manual method is
recommended. However, any other method which measures N02 directly ,
and has no interference from NO may be used.
To determine the amount of N02 impurity in the NO
cylinder, sample a diluted stream from the cylinder. The exact dilu-
tion will depend on the N02 concentration in the cylinder and the
method chosen for analysis. Generally, dilutions of 1 to 100 or
1 to 50 will be adequate. Convert the N02 analyzed to ppm. Then
calculate the N02 impurity using:
'2)A X FT (A-3)
32
-------
where: C^Q conentration in cylinder, ppm
)^ = N02 concentration analyzed, ppm
3
FT = Total flow at manifold, cm /min
F = Flow from NO cylinder, cm /min
6.4.3 N02 Converter Efficiency. The efficiency of the N02
converter should be constant and linear over the range of operation.
The N02 converter efficiency may be determined in the following
manner. While the chemiluminescence analyzer is sampling clean air,
adjust the zero controls for the NO and NO responses to read zero.
A
Dilute the assayed cylinder of NO with clean air to produce approxi-
mately 0.9 ppm NO and adjust the span controls for the NO and NO
A
responses to read exactly the NO concentration generated. The
exact NO concentration is calculated as follows:
(NO) = 'NO X CNO (A-4)
FT
where (NO) = NO concentration generated, ppm
NO concentration, ppm
» cm /min
3
Fj = total flow at manifold, cm /min
Note: If N02 impurity is present in the NO cylinder, adjust the
span control for NOX response to include the N02 impurity concentra-
tion assuming a converter efficiency of 1.
33
-------
Observe and record the initial NO and NOX readings.
Next, add ozone to the NO stream to reduce the NO response to 0.8 ppm.
Observe and record the final NO and NOX readings. Repeat the above
procedure adding ozone such that the NO response is decreased to
0.7, 0.6 and 0.5 ppm successively. Calculate the concentration of
NOp generated by G.P.T. for each stepwise addition of ozone, the
total N02 concentration generated, and the concentration of N02
converted to NO as follows:
N02 Generated, ppm = N02 G.P.T., ppm * N02 Impurity, ppm
N02 Converted, ppm = N0x Final, ppm " NOFinal, ppm
Plot the N02 concentration converted (y-axis) versus
the total N0« concentration generated (x-axis). The slope of this
plot is the converter efficiency. A straight line indicates a
constant converter efficiency over the 0 to 0.5 ppm range. If
the converter efficiency is less than 0.90 (90 percent) or if it is not
constant (i.e.* nonlinear curve), the converter should be replaced
or reactivated.
6.4.3.1 Frequency of Determination. The converter efficiency
should be determined monthly when the instrument calibration is
checked (Section 6.4.4.2).
6.4.4 N02 Calibration (0 to 0.5 ppm range).
6.4.4.1 Procedure. While sampling zero air from the gas phase
titration apparatus, zero the NO, NO , and N09 detector reponses.
" £
34
-------
Adjust the NO flow rate to generate approximately 0.9 ppm NO by
flow dilution (Section 6.4.3) and adjust the span controls to
read the exact NO concentration generated. Any N02 impurity in
the NO cylinder should be included (with an appropriate correction
for dilution and converter efficiency) in the NO span adjustment.
A
Adjust the ozone source to add enough 0., to decrease
the NO response to approximately 0.45 ppm. This results in the
generation of an equivalent amount of N02 which is used to span the
N02 response. Allow the response of the instrument to stabilize
and adjust the span control to give a direct readout of the
generated N02, including any N02 impurity if present. Continue to
decrease the 03 concentration so as to produce N02 concentrations
of approximately 0.30, 0.20, 0.10, and 0.05 ppm. Using this pro-
cedure the N02 generated can be determined by the following equation:
N02 Generated = NOInitial " NOFinal + N02 Impurity, ppm "(A-6)
where: N02 Qenerate(j = Total concentration of N02 generated, ppm
^Initial = In^^a^ concentration of NO, ppm
NOp. , = Concentration of NO after 03 addition, ppm
Plot the total N02 concentration generated versus analyzer
N02 response. A straight line should be obtained.
6.4.4.2 Frequency of Calibration. A complete calibration should
be made once a month. Analyzers should be zeroed and spanned at
35
-------
80 ± 5 percent of full scale daily, then adjusted for any drift that is
detected. If the drift is more than ± 15 percent, the analyzer should
be recalibrated.
7. Calculations
7.1 Determine the N02 concentration directly from the NOg
calibration curve (Section 6.4.4.1).
7.2 Nitrogen dioxide concentrations in ppm can be converted
to yg/m at 25° C and 760 mmHg as follows:
(N02), yg/m3 = (N02), ppm x 1880 x &&- (A-7)
8. Bibliography
1. Fontijn, A, J. Sabadell, and R. J. Ronco. Homogeneous
Chemiluminescent Measurement of Nitric Oxide with Ozone. Anal.
Chem. 42 : 575, 1970.
2. Hodgeson, J. A., R. E. Baumgardner, B. E. Martin, and
K. A. Rehme. Stoichiometry in the Neutral lodometric Procedure for
Ozone by Gas-Phase Titration with Nitric Oxide. Anal. Chem. 43;
1123, 1971.
3. Rehme K. A., B. E. Martin, and J. A. Hodgeson. Tentative
Method for the Calibration of Nitric Oxide, Nitrogen Dioxide and
Ozone Analyzers by Gas Phase Titration. Environmental Protection
Agency, Research Triangle Park, N. C. Publication No. EPA-R2-73-246.
March 1974.
36
-------
4. Stedman, D. H., E. E. Daby, F. Stuhl, and H. Niki.
Analysis of Ozone and Nitric Oxide by a Chemiluminescent Method
in Laboratory and Atmospheric Studies of Photochemical Smog.
J. Air Poll. Control Assoc. 22: 260, 1972.
5. Hodgeson, J. A., K. A. Rehme, B. E. Martin, and R. K.
Stevens. Measurements for Atmospheric Oxides of Nitrogen and Ammonia
by Chemiluminescence. Preprint, Prepared at 1972 APCA Meeting, Miami,
Florida, June 1972, Paper No. 72-12.
6. Winer, A. M., J. W. Peters, J. P. Smith, and J. N. Pitts,Jr.
Response of Commercial Chemiluminescent N0-N02 Analyzers to Other
Nitrogen-containing Compounds. Envir. Sci. Tech. 8_: 1118, 1974.
7. Title 40-Protection of Environment. Part 50—National
Primary and Secondary Ambient Air Quality Standards. Federal Register.
36 (228): 22384-22397, November 25, 1971.
8. Hodgeson. J. A., R. K. Stevens, and B. E. Martin. A Stable
Ozone Source Applicable as a Secondary Standard for Calibration of
Atmospheric Monitors. ISA Transactions. 11: 161, 1972.
9. National Bureau of Standards Technical Note No. 585.
National Bureau of Standards, Washington, D. C. January 1972.
pp. 11-25. Available from: Superintendent of Documents, Government
Printing Office, Washington, D. C. 20402. '.
10. Fuerst, R. G., and J. H. Margeson. Art Evaluation of the
TGS-ANSA Method for Measurement of NO^. Copies iof this document,
which includes a copy of the method write-up, can be obtained from:
Methods Standardization and Performance Evaluation Branch, QAEML,
Environmental Protection Agency, Research Triangle Park, North Carolina
27711. '
37
-------
A-l. Performance Specifications for Atmospheric-Chemiluminescence
NO Analyzers
Parameter Unit3
Range ppm
Noise ppm
Lower Detectable Limit ppm
Zero Drift, 12-, 24-Hour ppm
Span Drift, 24-Hour ppm
Lag Time minutes
Rise Time, 95 percent minutes
Fall Time, 95 percent minutes
Specification
multiple
0-0.005
0.01
±0.02
±0.02
0.5
1.0
1.0
To convert from ppm to yg/m at 25° C and 760 mm, multiply by 1880.
3No performance test required. All other performance specifications
are tested on instrument operating in the 0-0.5 ppm range.
38
-------
A-2.
Definitions of Performance Specifications
Range - Minimum and maximum concentrations which the system shall be
capable of measuring.
Noise - Spontaneous, short duration deviations in the analyzer output
about the mean output, which are not caused by input concentration
changes.
Lower Detectable Limit - The minimum pollutant concentration which
produces a signal of twice the noise level.
Zero Drift - The change in analyzer response to zero pollutant
concentration over 12- and 24-hour periods of continuous unadjusted
operation.
Span Drift - The change in analyzer response to an upscale pollutant
concentration over a 24-hour period of continuous unadjusted operation.
Lag Time - The time interval between a step change in input concen-
tration at the analyzer inlet and the first observable corresponding
change in the analyzer response that is equal to twice the noise in
the instrument output.
Rise Time - The time interval between initial response and 95 percent of
final response after a step increase in input'concentration.
i
Fall Time - The time Interval between initial, response and 95 percent of
final response after a step decrease in input,concentration.
39
-------
APPENDIX B. EQUATION FOR CALCULATION
OF EFFECT OF EACH VARIABLE IN RUGGEDNESS TEST
i A s + t + u + v w + x + y + z
i. M - a ^ ^-*
r> a k S + t + W + X ll + V+y + Z
d.. b - D = ^ ^-"
- r - s + u + w + y _ t + v + x + z
*5 • w ~ C ™" yi " yi
4. p. d =
c F s + u + x + z t + v + w + y
5> L " e 4 4
6 F - f = s + v t w + z - t + u + x + y
7 r s + v + x + y t + u + w + z
7. G - g = - - •*- -- -
40
-------
APPENDIX C. RUGGEDNESS TEST DATA
Determination No.
8
N02 generated
by gas phase
titration,
ppm
0.082
0.192
0.309
0.419
O.b37
0.077
0.185
0.297
0.413
0.531
0.082
0.192
0.314
0.432
0.547
0.084
0.197
0.315
0.438
0.551
0.080
0.190
0.303
0.426
0.540
0.092
0.204
0.314
0.431
0.544
0.077
0.182
0.292
0.417
0.532
0.093
0.206
0.320
0.443
0.5b9
N02 analyzer
response,
ppm
0.085
0.192
0.304
0.420
0.540
0.078
0.178
0.293
0.415
0.546
0.080
0.193
0.302
0.425
0.545
0.085
0.190
0.305
0.420
0.541
0.084
0.195
0.304
0.422
0.550
0.088
0.200
0.310
0.426
0.546
0.078
0.176
0.288
0.400
0.555
0.090
0.199
0.315
0.436 '
0 .558
Least
squares
slope
0.9987
Observed
result
designation
0.9770
1.009
1.028
0.9940
w
0.9977
1.023
1.003
41
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-650/4-75-021
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Evaluation of Gas Phase Titration Technique As Used
for Calibration of Nitrogen Dioxide Chemiluminescence
Analyzers
5 REPORT DATE
April 1975
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
E. Carol Ellis and flohn H. Margeson
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Protection Agency,
Quality Assurance & Environmental Monitoring LaboratoHnT
Methods Standardization & Performance Evaluation Branch
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
1HA327
. CONTRACT/GRANT NO
12 SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection Agency
Office of 'Research and Development
Washington, D. C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A detailed method write-up describing the chemiluminescence prodedure for
the continuous measurement of nitrogen dioxide (N02) in ambient air was developed.
Atmospheric concentrations of N02 are measured indirectly by the chemiluminescence
reaction of nitric oxide (NO) with ozone (03). The M>2 is first thermally reduced to
NO before it is reacted with Og.
The reliability of measurements made by a continuous or instrumental sampling
method is strongly affected by its calibration. In the chemiluminescence procedure,
an N02 analyzer is calibrated by the gas phase titration of NO with 03 that produces
N02 stoichiometrically. Significant errors can be introduced into the analyzer cal-
ibration if the standard NO cylinder is incorrectly assayed for NO as well as trace
NOo or if the converter efficiency of the analyzer is always assumed to be equal to
1.0. Provisions are included in the method to eliminate such errors.
The gas phase titration calibration procedure was subjected to a ruggedness
test. The results indicate that normal variations in such factors as reaction and
nixing chamber volumes, ratio of dilution air flow to flow through the ozone
generator, use of difference ozone generators with different dilution air flows, and
use of different standard NO levels had little effect on the gas phase titration
procedure.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
N02 measurement
Chemiluminescence procedure
Gas Phase tritration
18 DISTRIBUTION STATEMENT
Unlimited
19 SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
47
20 SECURITY CLASS (This page)
22 PRICE
EPA Form 2220-1 (9-73)
42
-------