United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 27711
EPA-600/2-79-051
March 1979
Research and Development
Current
Methods to
Measure
Atmospheric Nitric
Acid and Nitrate
Artifacts
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development. U S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring' '
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-79-051
March 1979
CURRENT METHODS TO MEASURE ATMOSPHERIC NITRIC ACID
AND NITRATE ARTIFACTS
edited by
Robert K. Stevens
Inorganic Pollutant Analysis Branch
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH 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 Sciences Research
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.
11
-------
FOREWORD
The Environmental Sciences Research Laboratory (ESRL) of the U.S. Environ-
mental Protection Agency at Research Triangle Park, North Carolina (EPA-RTP)
conducts an intramural and extramural research program to detect, define, and
quantify the effects of air pollution on urban, regional, and global atmo-
spheres (and the subsequent impact on water quality and land use). Within
ESRL, the Inorganic Pollutant Analysis Branch develops and conducts research
in chemistry, physics, and engineering related to development of instrumenta-
tion for field measurement and characterization of atmospheric contaminants.
In performance of this role, the Inorganic Pollutant Analysis Branch
conducted a workshop on "Measurement of Atmospheric Nitrates" in Southern
Pines, North Carolina, on October 3 and 4, 1978. This report represents one
important product of that meeting.
Aubrey P. Altshuller, Director
Environmental Sciences Research
Laboratory
111
-------
ABSTRACT
Presentations given at a workshop on "Measurement of Atmospheric Nitrates"
(Southern Pines, North Carolina, October 3-4, 1978) are documented. The
authors consider various analytical methods to measure ambient concentrations
of nitric acid and artifact nitrate formation.
In an Introduction, R. Stevens and W. McClenny summarize the workshop
presentations and outline a proposed field experiment to evaluate and compare
nitric acid and particulate nitrate measurement methods. E. Tuazon and col-
leagues discuss measurements of nitric acid in the California South Coast Air
Basin by Fourier transform long path infrared spectrometry. C. Spicer de-
scribes measurement of nitric acid by coulometric, chemiluminescence, and ion
chromatographic techniques. Chemiluminescence measurements of nitric acid at
Boulder, Colorado are reported by T. Kelly and D. Stedman. A. Lazrus and
colleagues describe reduction to ammonium ion of fixed inorganic nitrogen
collected on nylon filters, followed by the indophenol ammonia test. Col-
lection of nitric acid on sodium chloride impregnated filters at high volume
flow rates is discussed by J. Forrest and coworkers. R. Hare and colleagues
describe electron capture gas chromatographic techniques for analysis of
nitric acid. J. Tesch and R. Sievers report particulate nitrate and gaseous
nitric acid measurements by electron capture gas chromatography. Use of the
Denuder Difference Experiment to obtain separate deterrd-nations of particulate
nitrate and gaseous nitric acid without interference of positive or negative
nitrate artifacts is discussed by R. Shaw and coworkers.
IV
-------
CONTENTS
FOREWORD iii
ABSTRACT iv
FIGURES vii
TABLES ix
ABBREVIATIONS AND SYMBOLS xii
INTRODUCTION 1
R. K. Stevens and W. A. McClenny
Overview of Methods 1
Planned Field Experiment 4
References 6
MEASUREMENTS OF AMBIENT HNO IN THE CALIFORNIA SOUTH COAST AIR BASIN
BY KILOMETER PATHLENGTH FOURIER TRANSFORM INFRARED SPECTROMETRY. . 9
E. C. Tuazon, A. M. Winer, R. A. Graham, and J. N. Pitts, Jr.
Introduction 9
Instrumentation and Methods 10
Results 16
Discussion 21
Conclusions 22
Acknowledgments 25
References 25
MEASUREMENT OF GASEOUS HNO BY ELECTROCHEMISTRY AND CHEMILUMINESCENCE . 27
C. W. Spicer
Introduction 27
Measurement of HNO by Coulometry 27
Measurement of HNO by Chemiluminescence 29
Measurement of HNO by Filtration 31
v
-------
CHEMILUMINESCENCE MEASUREMENTS OF HNC>3 IN AIR 37
T. J. Kelly and D. H. Stedman
Introduction 37
Instrumentation and Methods 38
Results and Discussion 42
Acknowledgments 42
References . 42
INDOPHENOL AMMONIA TEST IN MEASUREMENT OF HNO and N0~ 45
A. L. Lazrus, B. W. Gandrud, and J. P. Greenberg
Introduction 45
Instrumentation and Methods « 45
Results and Discussion 47
References 50
DETERMINATION OF ATMOSPHERIC HNO WITH NaCl-IMPREGNATED FILTERS AT
HIGH VOLUME FLOW RATES 51
J. Forrest, R. L. Tanner, D. Spandau, T. D'Ottavio, and L. Newman
Introduction 51
Instrumentation, Methods, and Results 51
Acknowledgments 61
References 61
SELECTIVE COLLECTION AND MEASUREMENT OF GASEOUS HNO IN AMBIENT AIR . . 63
R. J. Hare, M. T. Wininger, and W. D. Ross
Introduction 63
Instrumentation and Methods 65
Results and Discussion 65
SELECTIVE COLLECTION AND MEASUREMENT OF PARTICULATE NITRATE AND
GASEOUS HNO IN AMBIENT AIR 67
J. Tesch and Robert Sievers
Introduction „ 67
Instrumentation and Methods 68
Discussion 72
Acknowledgments 77
References 77
THE DENUDER DIFFERENCE EXPERIMENT 79
R. W. Shaw, T. G. Dzubay, and R. K. Stevens
Introduction 79
Instrumentation and Methods 80
Discussion 81
Acknowledgments 83
References 83
VI
-------
FIGURES
Number
MEASUREMENTS OF AMBIENT HNO IN THE CALIFORNIA SOUTH COAST
AIR BASIN BY KILOMETER PATHLENGTH FOURIER
TRANSFORM INFRARED SPECTROMETRY
E. C. Tuazon
A. M. Winer
R. A. Graham
J. N. Pitts, Jr.
1 Eight-mirror multiple reflection system 11
2 Kilometer pathlength FTS-LPIR system 12
3 Time-concentration profiles obtained from 900-m pathlength
FTS-LPIR spectra of Riverside ambient air on October
14, 1977 20
4 Reference spectra at 0.5 cm spectral resolution and
1080-m pathlength 23
5 Riverside ambient air spectrum at 0.5 cm spectral
resolution, 900-m pathlength, with HgCdTe detector 24
MEASUREMENT OF GASEOUS HNO BY
ELECTROCHEMISTRY AND CHEMILUMINESCENCE
C. W. Spicer
1 Microcoulometric monitor 28
2 Comparison of HNO analysis by FTS-LPIR and coulometric
methods 28
3 Photooxidation of propylene/nitrogen dioxide/sulfur dioxide
in FTS-LPIR chamber 30
4 Schematic of chemiluminescent instrument modified for
simultaneous NO /HNO monitoring 30
X j
5 Coulometric vs. chemiluminescent response to gaseous HNO . . . 32
vii
-------
Number Page
6 Chemiluminescence response 32
7 Chemiluminescence response under varying conditions 33
8 O and HNO during urban plume traverse 33
9 Vertical profiles downwind of Philadelphia 34
CHEMILUMINESCENCE MEASUREMENTS OF HNO IN AIR
T. J. Kelly
D. H. Stedman
1 Instrumentation for measurement of NO by Chemiluminescence. . 39
x *
2 O , CO, NO , and HO levels at Boulder, Colorado, as
measured by Chemiluminescence 40
3 HNO and NO levels at Boulder, Colorado, as measured by
Chemiluminescence 41
INDOPHENOL AMMONIA TEST IN MEASUREMENT OF HNO AND NO
A. L. Lazrus
B. W. Gandrud
J. P. Greenberg
1 NH and total inorganic fixed N manifold 48
DETERMINATION OF ATMOSPHERIC HNO WITH
NaCl-IMPREGNATED FILTERS AT HIGH VOLUME FLOW RATES
J. Forrest
R. L. Tanner
D. Spandau
T. D'Ottavio
L. Newman
1 Adsorption of HNO by quartz filters 57
Vlll
-------
TABLES
Number Page
INTRODUCTION
R. K. Stevens
W. A. McClenny
Summary of HNO Detection Techniques
MEASUREMENTS OF AMBIENT HNO IN THE CALIFORNIA SOUTH COAST
AIR BASIN BY KILOMETER PATHLENGTH FOURIER
TRANSFORM INFRARED SPECTROMETRY
E. C. Tuazon
A. M. Winer
R. A. Graham
J. N. Pitts, Jr.
1 Calculated Detection Limits for Several Pollutant Species. . . 15
2 Maximum Pollutant Concentration Measured in
Riverside Air 17
3 Pollutant Concentrations in Riverside Air,
July 21, 1977 18
4 Pollutant Concentrations in Riverside Air,
July 25, 1977 19
MEASUREMENT OF GASEOUS HNO BY
ELECTROCHEMISTRY AND CHEMILUMINESCENCE
C. W. Spicer
Evaluation of Interferences to Coulometric HNO Monitoring . . 29
IX
-------
Number Page
INDOPHENOL AMMONIA TEST IN MEASUREMENT OF HNO AND NO
A. L. Lazrus
B. W. Gandrud
J. P. Greenberg
1 Precision as a Function of Concentration 49
2 Organic Nitrogen Compound Analysis for NH 49
DETERMINATION OF ATMOSPHERIC HNO3 WITH
NaCl-IMPREGNATED FILTERS AT HIGH VOLUME FLOW RATE
J. Forrest
R. L. Tanner
D. Spandau
T. D'Ottavio
L. Newman
1 Adsorption of NO by NaCl Filters in Quartz-Nad-NaCl
Filter Pack 53
2 Conversion of NO to Volatile HNO by Quartz Filters in
Quartz-NaCl-Quartz Filter Pack 54
3 Conversion of NO to Retained NO by Quartz Filters in
Quartz-NaCl-Quartz-NaCl Filter Pack 55
4 Loss of Particulate NO by Reaction with Aerosol H?SO
on Preloaded Quartz Filters 58
5 Adsorption of SO2 by K2CO3 and NaCl Filters 58
6 Comparison of Particulate and Gaseous Nitrates 59
7 Precision of HNO Analyses 60
SELECTIVE COLLECTION AND MEASUREMENT OF GASEOUS
HNO IN AMBIENT AIR
R. J. Hare
M. T. Wininger
W. D. Ross
1 Houston Field Study Measurements of HNO , August 1978 .... 64
-------
Number Page
SELECTIVE COLLECTION AND MEASUREMENT OF PARTICULATE NITRATE
AND GASEOUS HNO IN AMBIENT AIR
J. Tesch
R. E. Sievers
1 Collection Efficiency of Filters and Sorbents for HNO
Vapor and NO_ 74
2 Representative Measurements of Particulate Nitrate and
HNO in Ambient Air 76
XI
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
approx.
ARB
ASTM
atm
BNL
°C
cm
DDE
ECGC
EPA
ESRL
op
FEP
ft
FTS
g
hi-vol
hr
i.d.
in
IR
K
km
LPIR -------
m
mi
vg
approximately
Air Resources Board
American Society for Testing and Materials
atmosphere
Brookhaven National Laboratory
degree Celsius
centimeter
Denuder Difference Experiment
electron capture gas chromatography
U.S. Environmental Protection Agency
Environmental Sciences Research Laboratory
degree Farenheit
fluorinated ethylene propylene
foot
Fourier transform spectrometry
gram
high-volume
hour
inside diameter
inch
infrared
thousand
kilometer
long path infrared
mat£.r
mile
microgram
XII
-------
ym
min
ml
mm
N
NCAR
run
NOAA
NSF
PDT
PH
ppb
ppm
RTF
QAST
SAPRC
USP
SYMBOLS
Ar
C
CH4
CO
co2
C6H5N02
C6H6
C7H8
Cu
CuCl
EDTA
FeS04
H
HCHO
HCOOH
micrometer
minute
milliliter
millimeter
normality
National Center for Atmospheric Research
nanometer
National Oceanic and Atmospheric Administration
National Science Foundation
Pacific Daylight Time
negative logarithm of hydrogen ion concentration
part per billion
part per billion by volume
part per million
Research Triangle Park, North Carolina
light transmission cutoff
Statewide Air Pollution Research Center
U.S. Pharmacopoeia
argon
carbon
methane
carbon monoxide
carbon dioxide
nitrobenzene
benzene
toluene
copper
cuprous chloride
ethylenediaminetetraacetic acid
iron sulfate
hydrogen ion
formaldehyde
formic acid
Xlll
-------
HC1
H2°2
H02N°2
HgCdTe
InSb
M
NH_
63Ni
NMHC
NO
N02
N0~
N0~
NO
x
N2°5
NaCl
NaOH
°3
PAN
so2
S°4
Zn
(CN)
hydrochloric acid
nitrous acid
nitric acid
water
hydrogen peroxide
per(oxy)nitric acid
sulfuric acid
mercury cadmium telluride
indium antimonide
potassium carbonate
third body molecule (usually N or O_) in a three-body
atmospheric collision
ammonia
ammonium ion
radioactive isotope of nickel (beta transmitter)
ammonium nitrate
total nonmethane hydrocarbons
nitric oxide
nitrogen dioxide
nitrite ion
nitrate ion
total oxides of nitrogen
nitrogen pentoxide
sodium chloride
sodium nitroprusside
sodium hydroxide
ozone
peroxyacetyl nitrate
sulfur dioxide
sulfate ion
zinc
- 2.718
xiv.
-------
INTRODUCTION
Robert K. Stevens
William A. McClenny
Inorganic Pollutant Analysis Branch
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
A workshop entitled "Measurement of Atmospheric Nitrates" was held at
Southern Pines, North Carolina, on October 3 and 4, 1978, to discuss analytical
methods to measure ambient concentrations of nitric acid (HNO ) and artifact
nitrate formation. Attending the workshop were scientists directly involved
in the development of methods to measure HNO , including representatives from
the academic, private, and public sectors. The first day was devoted to
discussions of measurement methods and factors causing formation of nitrate
artifacts on a variety of collection surfaces. On the second day, a field
sampling experiment was designed to test HNO_ measurement methods and to
determine factors that cause positive and negative nitrate artifacts.
OVERVIEW OF METHODS
Methods described in the workshop sessions included the use of continuous
(real-time) and semi-continuous monitors as well as integrative collection
of HNO on adsorbing materials. The continuous methods were the following:
(1) microcoulometry (Spicer, Battelle Columbus Labs); (2) chemiluminescence
(Spicer, Battelle Columbus Labs; Stedman, University of Michigan); (3) Fourier
transform long path infrared spectrometry (FTS-LPIR) (Winer, University of
California at Riverside). Methods involving preconcentration were: (1)
-------
collection of HNO on nylon or cotton followed by extraction, conversion to
nitrobenzene (C-.H._NO_), and analysis by gas chromatography (Hare, Monsanto
o D 2
Research; Sievers, University of Colorado); (2) reduction to ammonium ion
(NH.) of fixed inorganic nitrogen collected on nylon filter, followed by the
indophenol ammonia test (Lazrus, National Center for Atmospheric Research);
(3) collection of HNO on sodium chloride (NaCl) impregnated filters, followed
by extraction and hydrazine reduction-diazotization analysis of nitrate (New-
man, Brookhaven National Labs).
A tabulated summary of these techniques is presented in Table 1. Each
technique is associated with a procedure consisting of a number of experimental
steps. "Sampling" refers to the transport of ambient air to the place of
measurement; as used in Table 1, the term includes prefiltering to eliminate
particulates. "Scrubbing" is the elimination of some interfering gaseous
component of the ambient air; "conversion" occurs in those cases when the HNO
is thermally converted to nitric oxide (NO) prior to measurement. "Collection"
and "extraction" refer to sampling onto a filter followed by a wet chemical
extraction in which HNO is changed in form. "Measurement" is by the tech-
nique stated, whereas "subtraction" refers to the identification of a signal
increment contributed by HNO . "Minimum detectable levels" are taken from the
investigators' reports (this volume) or are estimated from available informa-
tion.
Each technique has unique features. FTS-LPIR is ideal for detection of
HNO , since measurement takes place in the atmosphere and identification is
made unambiguously by the recognition of characteristic infrared absorptions.
However, the equipment is not portable, and the method has a minimum detection
level of 5 ppb. Chemiluminescence has the inherent sensitivity of a rate
sensor, but involves the measurement of small differences in a signal that is
frequently large and time-varying due to interferences (e.g., total oxides of
nitrogen, NO ; peroxyacetyl nitrate, PAN; and organic nitrates).
Collection techniques involve tedious documentation of collection and
release efficiencies, maximum loading, and interferences; preconcentration, of
course, reduces the requirements for sensitivity. The formation of nitrate
-------
TABLE 1. SUMMARY OF HNO DETECTION TECHNIQUES
1
Technique
chemiluminescence
(Spicer)
chemi luminescence
(Stedmrin)
FTS-LPIR
(Winer)
microcoulometry
(Spicer)
color ime try
( lozrus)
colovimetry
(Newman)
ECGC
(Hare, Sievers)
ion chromatography
(Spicer)
Minimum
Detectable
Procedure Level
sample scrub convert measure subtract 5. 0
sample scrub convert measure subtract 0. 3
measure subtract 5.0
sample scrub measure subtract 5. 0
sample scrub measure subtract •'0. 1
sample collect extract measure O.I
sample collect extract measure 0.1
sample collect extract measure 0.1
Temporal
Status
cyclic
cyclic
real-time
cyclic
periodic
periodic
periodic
periodic
Tested
Interference (s)
NO, NO , PAN,
organic nitrates
NO, NO , PAN,
organic nitrates
most qaseous
species in
normal ambient
air
SO , NO , TAN
NH. , par tic-
ul ate nitrate
NO- , partic-
iilate nitrate
NO., partic-
ulflte nitrate
bromide
Reference (s)
Joseph rind
Spicer, 1978
Kelly and
Stedman, this
volume1; Likens ,
Tuazon et at. ,
this volume
Miller and
Spicer, 1975;
Spicer ct cl. ,
1978b
I,azrus et a/-. ,
1968
Ok it* ct at. ,
1976
Ross ct
-------
artifacts during collection of nitrates on filter media can occur because:
(1) filters may collect HNO and nitrogen dioxide (NO ) , producing a nitrate
ion (NO ) artifact; (2) nitrate present as ammonium nitrate (NH NO ) may evapo-
rate during collection; and (3) nitrate may be lost from the filter if acid
aerosols react with solid nitrate on the filter to release HNO . Shaw et al.
(this volume) describe the Denuder Difference Experiment in which — assuming
the nylon collection surface quantitatively collects only HNO — the positive
and negative artifact nitrate on a variety of filter media is measured. This
technique is considered to be the best proposal for quantifying the "negative
nitrate artifact" — the effect of acid gases (especially sulfuric acid, H^SO.)
reacting with particulate nitrate on the filter surface to release HNO .
PLANNED FIELD EXPERIMENT
On the second day of the workshop, the participants divided into three
groups to consider the protocol for a field experiment to evaluate and compare
methods to measure HNO and nitrates. Group I considered the pollutant gases
that need to be measured. Group II considered HNO methods and testing pro-
cedures, and Group III designed experiments to determine nitrate artifact
levels.
It was decided that a 7-day field study should be conducted at Harvey
Mudd College in the Los Angeles Basin during early June, 1979, for the purpose
of intercomparing methods to measure HNO and particulate nitrate. This site
was selected because an FTS-LPIR instrument is located there. (FTS-LPIR
equipment is not portable; therefore, it is necessary to conduct the initial
comparison study at an FTS-LPIR site.) The group also recommended that a
second study be conducted in the eastern part of the United States to evaluate
methods to measure HNO and nitrates in an environment in which nitrate and
HNO levels are relatively low and in which other acid aerosols (H9SO ) are
,3 ~ — - — ^ - — -.._,_.__ _ ^ ~r
proportionately higher.
A final discussion among the three groups produced the following outline
for the proposed experiment:
-------
Objectives
(1) To test and compare measurement systems for HNO_ vapor and nitrate
aerosol; and
(2) To determine the extent to which measurement systems are susceptible to
artifact nitrate formation or loss and, if possible, to determine the
cause of artifacts.
Methods and Rationale
(1) The various measurement systems are now available; their comparison
can best be achieved by side-by-side evaluation under field conditions,
similar to the EPA Charleston Aerosol Sampler Comparison Study (see
Camp et al., 1978).
(2) In order to determine the absolute reliability of the instruments, each
must be run side-by-side with FTS-LPIR, which is the only technique estab-
lished at this time to be interference-free.
(3) In order to determine the cause of nitrate artifacts, the following
possible parameters (in addition to nitra
simultaneously with the instrument study:
possible parameters (in addition to nitrate and HNO-) should be monitored
carbon monoxide (CO)
hydrogen ion (H )
hydrochloric acid (HC1)
water (HO)
hydrogen peroxide (HO.)
NO
NO,,
ammonia (NH )
ammonium ion (NH )
nitrogen pentoxide (N_O_)
ozone (O3)
PAN
sulfur dioxide (S02)
sulfate ion (SO.)
Because of the large number of interested participants who already possess
equipment, it seems that all (or nearly all) of these can be measured.
-------
(4) In order to eliminate possible confusion concerning the role of large
aerosol particles, all participants making aerosol measurements will
use 15 ym cutoff inlets.
(5) Long- and short-term sampling will be performed to investigate possible
artifact-time relationships and diurnal patterns.
(6) The Denuder Difference Experiment (Shaw e~b al., this volume) is expected
to provide unambigious value
losses (negative artifacts) .
to provide unambigious values for HNO., vapor and measurements of nitrate
To summarize, this is an ambitious project that will require considerable
work to arrive at a fruitful conclusion. We believe, however, that such an
effort can bring to an end the aerosol nitrate ambiguity that plagues evalua-
tion of air monitoring data. The experiment will establish the best techniques
for HNO and nitrate measurements, and the conditions under which the remaining
techniques yield valid results.
The balance of this report consists of workshop presentations relating
to HNO measurement techniques.
REFERENCES
Camp, D. C., A. L. VanLehn, and B. W. Loo. 1978. Intercomparison of Samples
Used in the Determination of Aerosol Composition. EPA 600/7^78-118,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina.
Joseph, D. W., and C. W. Spicer. 1978. Chemiluminescence Method for Atmo-
spheric Monitoring of Nitric Acid and Nitrogen Oxides. Anal. Chem. 50:
1400-1403.
liazrus, A. L., E. Lorange, and J. P. Lodge, Jr. 1968. New Automated Micro-
analyses for Total Inorganic Fixed Nitrogen and for Sulfate Ion in
Water. Adv. Chern. 73.
Likens, G. E. 1976. Acid Precipitation. Chem. Eng. News. 54 (48):29-44.
Miller, D. F., and C. W. Spicer. 1975. Measurement of Nitric Acid in Smog.
J. Air Poll. Control Assoc. 25:940-942.
-------
Okita, T., S. Morimoto, M. Izawa, and S. Konns. 1976. Measurement of Gaseous
and Particulate Nitrates in the Atmosphere. Atmos. Environ. 10:1085.
Ross, W. D., G. W. Butler, T. G. Duffy, W. R. Rehg, M. T. Wininger, and R. E.
Sievers. 1975. Analysis for Aqueous Nitrates and Nitrites and Gaseous
Oxides of Nitrogen by Electron Capture Gas Chromatography. J. Chrom.
112:719-727.
Spicer, C. W., P. M. Schumacher, J. A. Konyoumjian, and D. W. Joseph. 1978a.
Sampling and Analytical Methodology for Atmospheric Particulate Nitrates.
EPA 600/2-78-067, U.S. Environmental Protection Agency.
Spicer, C. W., G. F. Ward, and B. W. Gay, Jr. 1978b. A Further Evaluation
of Microcoulometry for Atmospheric Nitric Acid Monitoring. Anal. Lett.
All(l):85-95.
-------
MEASUREMENTS OF AMBIENT HNO IN THE CALIFORNIA SOUTH COAST
AIR BASIN BY KILOMETER PATHLENGTH FOURIER TRANSFORM INFRARED SPECTROMETRY
Ernesto C. Tuazon
Arthur M. Winer
Richard A. Graham
James N. Pitts, Jr.
Statewide Air Pollution Research Center
University of California
Riverside, California 92521
INTRODUCTION
The current lack of systematic and reliable data on ambient concentrations
of a host of nitrogenous and oxygenated compounds, known or suspected to be
present in polluted urban atmospheres, is generally recognized. To address
this problem, EPA has supported an ambient air monitoring program based on
FTS-LPIR. In this program the Statewide Air Pollution Research Center (SAPRC)
of the University of California has operated a kilometer pathlength FTS-LPIR
instrument, originally designed and assembled by Hanst (EPA-RTP). This system
provides an i-n-situ ppb detection capability for many oxygenated and nitro-
genous pollutants for which reliable, alternative analytical methods are
presently unavailable. A description of the FTS-LPIR instrument and the
initial measurements made in Riverside during the fall of 1976 have been
reported (Tuazon et al,, 1977, 1978); these earlier results included the first
spectroscopic detection of HNO and formaldehyde (HCHO) in the polluted tropo-
sphere. The present paper briefly describes the SAPRC experimental facility,
its specific application to ambient measurements of HNO , and some results
obtained frorr. monitoring carried out in the California South Coast Air Basin
9
-------
during the smog seasons of 1977 at Riverside (a downwind receptor site) and
1978 at Claremont (a mid-basin site). A more comprehensive account of this
program and the data it has produced is currently in press (Tuazon e~k at. ,
1979).
INSTRUMENTATION AND METHODS
Kilometer Pathlength Cell and FTS-LPIR Instrument
Figure 1 illustrates the eight-mirror multiple reflection cell employed
in this study. The details of its design have been described by Hanst (1971).
The in-focus (nesting) mirror assembly consists of four rectangular mirrors,
while the out-of-focus assembly (collecting mirrors) consists of four 30-cm
diameter mirrors and is 22.5 m (the eight-mirror average radius of curvature)
from the nesting mirror assembly. All mirrors are optically polished and
gold-coated for maximum reflectivity in the infrared (IR).
The prime consideration in the choice of the operating pathlength is
signal attenuation accompanying an increased number of reflections. The
average reflectivity coefficient of the cell mirrors was determined to be
~0.98. The 47 mirror reflections at a pathlength of 1080 m correspond to a
transmitted intensity of 0.39 times the incident radiation. This value is
slightly greater than 1/e (0.37), which has been established as the criterion
for maximizing the signal-to-noise ratio for IR systems that are detector
noise limited (Stephens, 1958, 1961).
The cell consists of a sectional rectangular aluminum frame (total
dimensions: 0.81 x 0.84 x 23 m) with a lining of 50 ym fluorinated ethylene
propylene (FEP) Teflon film. Teflon provides a relatively inert, replaceable
surface and transmits solar radiation efficiently when the cell is used as a
reaction chamber for synthetic systems.
Figure 2 depicts the multiple-reflection cell coupled to the FTS-LPIR
system, which is housed in an air-conditioned building (3.7 x 3.7 m) . The
collimated IR beam from a Nernst glower is modulated by a Digilab Model 296
10
-------
o o
20
9 Q ©
8 16 24
© O
7\0
48
© ®
y
O 0
o
9
0
22
O
IB
O
14
®
10
©
6/
/
tJ
NESTING MIRRORS
N2
N4
N3
Nl
COLLECTING MIRRORS
Figure 1.
Eight-mirror multiple reflection system. Numbered dots indicate sequence of reflections
during six complete passes through the cell; I = entrance aperature; O = exit aperature.
-------
MICHELSON
INTERFEROMETER
Figure 2. Kilometer pathlength FTS-LPIR system.
-------
Michelson IR interferometer (resolution >0.5 cm ). The beam exiting from
the cell is sent to either of two liquid nitrogen cooled detectors: a photo-
voltaic indium antimonide (InSb) detector for the 2000-3900 cm range or a
photoconductive mercury cadmium telluride (HgCdTe) detector for the 600-2000
-1
cm region.
Data collection and processing are performed by a Data General Nova 1200
computer with 4096 16-bit words of core. A disk unit (128 K words) is used
for software storage and partial data storage. Archival data is stored on
a Kennedy Model 9700 magnetic tape unit. The Digilab system is also equipped
with a pen plotter, an oscilloscope, and a teletype.
Routine data collection is carried out at pathlengths of 900 and 1080 m.
Data collection, transform computation, routine plotting of a portion of the
spectrum, and magnetic tape storage can be accomplished within 10 min for each
detector. Hence, a time resolution of three complete spectra (600-4000 cm )
per hour is available to characterize moving smog fronts.
To measure the best performance of the system, experiments are conducted
at night (when thermal effects and vibrations are minimal) and with the cell
purged to low relative humidities with dry nitrogen gas. At pathlengths near
1 km, a signal to peak-to-peak noise ratio of 400 is attained. This cor-
responds to a signal-to-noise ratio of approximately 1000 for a single mirror
reflection. From the analysis of Kanst et al. (1973), a theoretical detec-
tion sensitivity of 0.8 ppb has been calculated for a compound with an absorp-
tivity of 30 cm atm (e.g., PAN). In practice, the theoretical detection
sensitivity is not usually achieved due to effects such as vibrations, con-
centration fluctuations with time, interfering absorptions, and variations in
the refractive index of the air caused by temperature gradients in the cell.
Additional details of optical alignment procedures, multiple reflection cell
operation, and signal-to-noise considerations can be found elsewhere (Tuazon
et al. , 1979).
13
-------
Data Collection and Processing
Strong absorptions due to the HO and carbon dioxide (CO ) content of
the atmosphere render certain regions of the IR unsuitable for detection of
pollutant molecules. At pathlengths of 1 km and longer, the spectral windows
for IR detection are generally limited to the 760-1300 cm , 2000-2230 cm ,
and 2390-3000 cm regions. Fortunately, the majority of pollutant molecules
have "fingerprint" absorptions in the accessible IR regions. For data anal-
ysis the spectra are retrieved from storage and ratioed against a clean
background spectrum with the same HO content in order to eliminate the
majority of interfering water absorptions.
The sampling procedure consisted of drawing air through the cell at a
rate of 330 liters sec for a minimum of 4 min before the start of an inter-
ferogram averaging. This corresponded to a displacement of the previous air
sample by a minimum of five volumes of fresh sample. Total pathlengths of 900
or 1080 m and a spectral resolution of 0.5 cm were employed. Actual data
collection of 30 co-added interferograms took no more than 6 min. Several
experiments were performed in which spectra of smoggy air were recorded; alter-
nately for air being drawn continuously through the cell and for the system in
the stopped flow mode. These experiments established that key pollutant
species (e.g., HCHO, HNO , formic acid [HCOOH]) were not significantly lost to
or off-gassed from the cell walls during the brief scanning period.
FTS-LPIR Absorptivities and Detection Limits
Initial estimates of pollutant concentrations in the 1976 study {Tuazon
et al., 1977, 1978) were based on literature values of absorptivities measured
at lower spectral resolutions (S2 cm ) than those employed in this program.
New measurements (Tuazon et-at.,._ 1979) of absorptivities for HCOOH, HCHO,
KH , £.nd KNC have since been made in the SAPRC evacuable chamber, and in some
cases significant disagreements with the literature values were found.
Table 1 gives the newly measured absorptivities for HCOOH, HCHO, NH ,
and HNO , as well as their estimated detection levels in the present system at
14
-------
TABLE 1. CALCULATED DETECTION LIMITS FOR SEVERAL POLLUTANT SPECIES
Compound
' °3
PAN
NI!^
HNO
HCHO
IICOOH
"N02
H2°2
H02N°2
Measurement Absorptivity*
Frequency , -1 -1
(cm atm ,
(cm) RTF )
1055
1162
931
967.5
993
896
2779
2781.5
1105
791
(trans)
853
(cis)
1251
803
9.7
32
27
35
21
12*
16*
70*
~30
71
9±3
27
*Absorptivity = log e(I /I)/pl where p is the
.and I and I are the transmitted intensities
J23°C, 760 torr.
''Measured from the intensity of the Q branch
Resolution
(cm'1)
1-2
2-4
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
2
0.5-1
Approximate
Detection Limit
Reference at 1 km Pathlength
Pitts et al., 1976 10
McAffee et al. , 1976
Stephens, 1964 ' 3
Tuazon et al., 1979 4
Tuazon et al., 1979 3
Tuazon et al., 1979 4
Tuazon et al., 1979 6
Tuazon et al., 1979 6
Tuazon et al., 1979 2
Chan et al., 1976 10§
(cis + trans)
Chan et al. , 1976
K
Hanst et al. , 1975 40
Graham et al. , 1977 8
concentration (atm) of the compound, 1 is the pathlength (cm)
with and without the compound present, respectively.
only. (See text.)
See text for discussions of atmospheric interferences for the spectral analysis of HNO-, H2°2' and
HO NO concentrations.
-------
a 1-km pathlength. Nominal uncertainties in measured concentrations are one
half of the detection limits. For the analytical bands of HCOOH at 1105 cm
and HCHO at 2780 cm , no more than 10% difference exists between the absorp-
tivity value of the Q branch height alone and that of the total value as
measured from the true baseline. For HNO , however, the total absorptivity at
896 cm is approximately twice that of the value for the Q feature alone due
to the significant contribution of the broad P and R branches. Nevertheless,
only the Q branch height absorptivities were used for all of the above com-
pounds; this procedure minimized baseline interferences caused by possible
absorptions from unidentified species. These Q branches were verified by
experiment to obey Beer's law for our program's specific conditions.
The detection limits given in Table 1 are, in most cases, based on a 1%
absorption. For the 993 cm line of NH and the 896 cm Q branch of HNO ,
however, the detection sensitivities are slightly better (~0.75% absorption)
because these features are virtually free of atmospheric H?O and CO absorp-
tions. Moreover, spectra analyses made in 1977 (Tuazon e~t a.1., 1979) facili-
tated the establishment of operational minimum signal levels at which HO,
HNO?, and per (oxy)nitric acid (HO?NO ) can be positively detected.
RESULTS
Table 2 summarizes the maximum concentrations of NH , HNO , HCOOH, HCHO,
O , and PAN measured in 1977 in Riverside between 1 a.m. and 7 p.m. (PDT)
during 10 episode days in July, August, and October. Detailed pollutant
concentrations as a function of time are presented in Tables 3 and 4 for
representative 1977 air pollution episodes in Riverside on July 21 and 25,
respectively. Figure 3 shows time concentration profiles for nine species
observed on October 14, 1977.
Included in the data for July 21 are the corresponding s-min concentra-
tion averages for NO, NO , total nonmethane hydrocarbons (NMHC, expressed as
total carbon [C]), and CO, obtained by SAPRC staff using the California Air
Resources Board (ARB) mobile laboratory located adjacent to the km cell facil-
ity. In this SAPRC-ARB facility, NO/NO/NO were measured with a Bendix
16
-------
TABLE 2. MAXIMUM POLLUTANT CONCENTRATION (PPB) MEASURED
IN RIVERSIDE AIR IN 1977
Date
Jul.
Aug.
Oct.
21
25
10
11
12
3
4
11
14
17
°3
264
291
234
287
249
241
195
242
260
240
PAN
11
10
10
13
11
10
10
18
17
14
™3
41
47
40
39
46
57
37
132
102
81
HNO
12
20
*
13
16
13
15
12
9
11
HCOOH
5
4
8
10
11
8
8
9
13
11 .
HCHO
20
19
19
27
25
29
25
38
36
38
*Below detection limit.
-------
TABLE 3. POLLUTANT CONCENTRATIONS* (PPB) IN RIVERSIDE AIR,
JULY 21, 1977
Time
1140
1202
1231
1258
1336
1423
1444
1507
1534
1552
1610
1632
1653
1711
1743
1800
1818
1835
1857
°3
173
180
176
161
152
198
216
214
245
229
253
264
238
239
239
225
200
206
181
PAN
10
8
4
3
—
3
3
6
5
8
9
10
9
9
9
11
9
9
9
NH3
—
—
—
4
25
41
38
29
24
19
13
11
5
6
6
—
9
8
8
HNO HCOOH
- 3
- 3
- 4
- 2
- 2
- 2
— —
- 2
10 2
7 3
9 5
10 4
- 4
10 3
- 4
- 3
12 4
- 2
- 2
HCHQT
15
12
11
13
13
11
13
15
15
19
20
19
19
18
18
17
19
N02
40
36
35
28
23
36
43
35
37
53
37
54
65
49
66
59
55
64
83
NO
1
1
1
2
2
2
2
2
3
3
2
2
2
2
2
2
2
2
2
NMHC
(ppmC)
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.3
0.3
0.4
0.4
0.5
0.5
0.6
0.4
0.4
0.4
0.4
0.4
CO
(ppm)
1.2
1.4
1.1
0.8
0.7
1.0
0.9
1.1
1.2
1.4
1.5
1.4
1.5
1.6
1.3
1.3
1.3
1.3
1.3
*Dashes indicate below detection limit. Blanks indicate no measurement.
t:
,PDT.
THCHO measurements were made with the InSb detector approx. 10-15 min after
the times listed in the table for the HgCdTe detector spectra.
18
-------
TABLE 4. POLLUTANT CONCENTRATIONS* (PPB) IN RIVERSIDE AIR, JULY 25, 1977
Time
1058
1128
1210
1229
1308
1350
1500
1542
1603
1623
1642
1704
1722
1755
°3
137
145
158
142
130
193
51
189
256
291
247
197
191
184
PAN
3
—
—
—
—
—
—
—
9
10
8
5
5
3
NH3
—
—
8
7
29
47
14
24
13
18
19
13
14
13
™°3
8
8
8
9
—
—
—
9
15
20
17
15
12
14
HCOOH
2
3
—
2
—
—
—
—
3
3
4
3
2
2
HCHO
11
11
12
16
15
13
9
16
19
16
13
19
13
N02
51
31
55
85
59
42
*Dashes indicate below detection limit. Blanks indicate no measurement.
-------
150
150
CVJ
O
50
1200 1400 1600 1800
PACIFIC DAYLIGHT TIME
ex
Q.
O
O
O
2000
Figure 3. Time-concentration profiles obtained from 900-m pathlength FTS-
LPIR spectra of Riverside ambient air on October 14, 1977.
20
-------
chemiluminescent analyzer, and CO and total NMHC were measured with a Beckman
6800 gas chromatograph. The NO^ concentrations have been corrected for inter-
ferences due to PAN and HNO (Winer et al., 1974; Spicer and Miller, 1974).
DISCUSSION
Reliable data on the ambient concentrations of HNO- are essential in
understanding the fate of nitrogen oxides emitted into the atmosphere and in
establishing an adequate nitrogen balance for photochemical smog. The impor-
tant reactions for the formation of HNO are
OH + NO2 + M -»- HNO3 + M (1)
and
NO- + HO -»• 2HNO (heterogeneous) (2)
Since HNO has a low photolysis rate in the lower atmosphere, it serves as a
sink for gas phase nitrogen oxides.
The first detection of HNO_ in polluted urban air was reported by Spicer
and Miller (1974) and Miller and Spicer (1975). Using a modified Mast micro-
coulomb meter adapted for sensing acids rather than oxidants, the authors
reported a daily average concentration of 3 ppb and an average hourly peak of
10 ppb HNO in West Covina, California, during the period August to September
in 1973. A later report (Spicer, 1977) detailed the results of the West
Covina study and the July-August 1973 measurements in St. Louis, in which both
the continuous Mast analyzer and an integrated analysis by a modified chroma-
tropic acid method were used. Their tabulated data showed the highest 1-hr
maximum HNO concentration to be 80 ppb for St. Louis and 40 ppb for West
Covina. The West Covina results contrast sharply with the almost concurrent
FTS-LPIR study in Pasadena by Hanst et al. (1975). Hanst and co-workers did
not observe HNO , even in severe smog episodes (i.e., oxidant ~0.6 ppm), and
they reported a detection limit of approximately 10 ppb for their FTS-LPIR
system. A subsequent FTS-LPIR study reported the first spectroscopic
21
-------
detection of HNO , at concentrations up to 6 ppb, in Riverside during the fall
of 1976 (Tuazon et al., 1977, 1978).
As seen in Figure 4(c), the absorption band of HNO at ~880 cm is
J -1
characterized by three distinct Q branches at 879, 885 and 896 cm . Improve-
ments in operating parameters provided spectra of higher quality during our
1977 and 1978 studies, and they verified conclusively our previously reported
detection of HNO (Tuazon et al. 1977, 1978). Figure 5 is a portion of a
spectrum recorded at 12:32 p.m. on August 12, 1977. Although the atmospheric
HO absorptions are not exactly cancelled in this ratio spectrum, the three Q
branches mentioned above clearly appear in their proper intensity ratios and
establish the presence of HNO in this sample at a 16 ppb concentration.
The maximum HNO concentration observed in our monitoring studies was 20
ppb and occurred during an episode in Riverside on July 25, 1977, in which the
O concentration reached 300 ppb. Measured NH levels were low during this
episode (<30 ppb most of the time). Smog chamber (Spicer and Miller, 1976)
and ambient air (Spicer and Miller, 1974; Miller and Spicer, 1975) data in-
dicate that O and PAN levels are strong predictors of HNO concentrations.
Indeed, our data indicate that HNO concentrations of 10-20 ppb coincide with
high O levels (200-300 ppb) for several cases, particularly when the NH
-J -J
levels are low (<30 ppb). Unusually high concentrations of NH were recorded
on October 11, 14, and 17, 1977, with maxima of 80-130 ppb. Shortly after
these maxima occurred, the HNO levels dropped below the detection limit of 6
ppb, even though the O, levels remained near or above 200 ppb. During the
earlier hours of these episodes, when the NH concentrations were low, 8-12
ppb HNO levels were recorded at oxidant levels of 110-140 ppb. In general,
the HNO concentration was near or below our detection limit when high NH
concentrations were present.
The data cited above, as well as those reported elsewhere in more com-
plete form (Tuazon et al., 1979), support the current view on the NH role in
stabilizing HNO by forming particulate nitrates. However, the observation in
22
-------
100 r-
80
£60
(a)
750
860
790
900
910
980
830 870 910
FREQUENCY (cm-')
1020
950
1060
2840
990
Figure 4. Reference spectra at 0.5 cm spectral resolution and 1080-m path-
length: (a) NH ~200 ppb; (b) HCHO ~270 ppb; (c) HNO3 ~150 ppb.
23
-------
CO,
K)
100 r
£90
z
<
h-
h-
2 80
z:
o:
h-
^ 70
;
60
8(
-1
i
1 I
, |
T T
| 1 1
. j
HNO,
6 15 ppb
16
•
30
ppb
AUG. 12 77
12:32 PM
i i i
900 940 980
FREQUENCY (cm-1)
-i
Figure 5. Riverside ambient air spectrum (August 12, 1977 at 12:32 p.m.) at 0.5 cm spectral resolu-
tion, 900-m pathlength, with
=15 ppb.
detector.
concentration = 16 ppb; NH concentration
-------
this study of the HNO and NH coexistence, with HNO present in concentra-
tions of ~10-20 ppb simultaneously with NH levels as high as ~50 ppb, provides
potentially important data in quantitatively elucidating the relationship
between NH NO aerosol and gaseous HNO and NH in the atmosphere. Further
studies of this relationship are thus being carried out in our laboratory
(Doyle et al., in preparation).
ACKNOWLEDGMENTS
The authors gratefully acknowledge the support of this work by EPA (Grant
No. 804546-01) and Dr. Philip Hanst, the Project Monitor and developer of the
km pathlength FTS-LPIR facility. We also thank the California ARE for use of
the air monitoring data obtained in the ARE laboratory operated by SAPRC under
Contract No. A6-171-30. The contents of this paper do not necessarily reflect
the views and policies of the EPA or ARE, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
REFERENCES
Chan, W. H., R. J. Nordstrom, J. G. Calvert, and J. H. Shaw. 1976. Kinetic
Study of HONO Formation and Decay Reactions in Gaseous Mixtures of HONO,
NO, NO , HO, and N . Environ. Sci. Tech. 10:674-682.
Doyle, G. J., E. C. Tuazon, A. M. Winer, R. A. Graham, and J. N. Pitts, Jr.
1978. The Equilibrium Relationship of Ammonia and Nitric Acid to Partic-
ulate Ammonium Nitrate in Polluted Atmospheres. In preparation.
Graham, R. A., A. M. Winer, and J. N. Pitts, Jr. 1977. Temperature Dependence
of the Unimolecular Decomposition of Pernitric Acid and Its Atmospheric
Implications. Chem. Phys. Lett. 51:215-220.
Hanst, P. L. 1971. Spectroscopic Methods for Air Pollution Measurement. Adv.
Environ. Sci. Tech. 2:91-213.
Hanst, P. L., A. S. Lefohn, and B. W. Gay, Jr. 1973. Detection of Atmospheric
Pollutants at Parts-Per-Billion Levels by Infrared Spectroscopy. Appl.
Spectres. 27:188-198.
Hanst, P. L., W. E. Wilson, R. K. Patterson, B. W. Gay, Jr., L. W. Chaney,
and C. S. Burton. 1975. A Spectroscopic Study of California Smog.
EPA-650/4-75-0006, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina.
25
-------
McAfee, J. M. , E. R. Stephens, D. R. Fitz, and J. N. Pitts, Jr. 1976. Infra-
red Absorptivity of the 9.6 ym Ozone Band as a Function of Spectral
Resolution and Abundance. J. Quant. Spectres. Radiat. Transfer. 16:
829-837.
Miller, D. F., and C. W. Spicer. 1975. Measurement of Nitric Acid in Smog.
J. Air Poll. Control Assoc. 25:940-942.
Pitts, J. N., Jr., J. M. McAfee, W. D. Long, and A. M. Winer. 1976. Long-Path
Infrared Spectroscopic Investigation at Ambient Concentrations of the 2%
Neutral Buffered Potassium Iodide Method for Determination of Ozone.
Environ. Sci. Tech. 10:787-793.
Spicer, C. W. 1977. The Fate of Nitrogen Oxides in the Atmosphere. Adv.
Environ. Sci. Tech. (J. N. Pitts, Jr., and R. L. Metcalf, Eds.). 7:163-
262.
Spicer, C. W., and D. F. Miller. 1974. Nitrogen Balance in Smog Chamber
Studies. Presentation (Paper No. 74-247) before the 67th Annual Meeting
of the Air Pollution Control Association, Denver, Colorado, June 9-13.
Spicer, C. W., and D. F. Miller. 1976. Nitrogen Balance in Smog Chamber
Studies. J. Air Poll. Control Assoc. 26:45-50.
Stephens, E. R. 1958. Long-Path Infrared Spectroscopy for Air Pollution
Research. Appl. Spectros. 12:80-84.
Stephens, E. R. 1961. Long-Path Infrared Spectroscopy for Air Pollution
Research. Infrared Phys. 1:187-196.
Stephens, E. R. 1964. Absorptivities for Infrared Determination of Peroxy-
acetyl Nitrates. Anal. Chem. 36:928-929.
Tuazon, E. C., R. A. Graham, A. M. Winer, R. R. Easton, J. N. Pitts, Jr.,
and P. L. Hanst. 1978. A Kilometer Pathlength Fourier-Transform
Infrared System for the Study of Trace Pollutants in Ambient and Syn-
thetic Atmospheres. Atmos. Environ. 12:865-875.
Tuazon, E. C., A. M. Winer, R. A. Graham, and J. N. Pitts, Jr. 1977. Appli-
cation of a Kilometer Pathlength FT-IR Spectrometer to Analysis of Trace
Pollutants in Ambient and Simulated Atmospheres. In: Proceedings of
the 4th Joint Conference on Sensing of Environmental Pollutants (November
6-11). pp. 798-802.
Tuazon, E. C. , A. "M." "Wirier"," R." A. Graham, and J. ,N. Pitts^ Jr. 1979. Atmo-
spheric Measurements of Trace Pollutants by Kilometer Pathlength FT-IR
Spectroscopy. Adv. Environ. Sci. Tech. In press.
Winer, A. M., J. W. Peters, J. P. Smith, and J. N. Pitts, Jr. 1974. Response
of Commercial Chemiluminescent NO-NO Analyzers to Other Nitrogen-
Containing Compounds. Environ. Sci. Tech. 8:1118-1121.
26
-------
MEASUREMENT OF GASEOUS HNO BY
ELECTROCHEMISTRY AND CHEMILUMINESCENCE
Chester W. Spicer
Battelle
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
INTRODUCTION
Gaseous and particulate nitrates are important constituents of the atmo-
sphere due to their potential impact on health, visibility, and precipitation
chemistry. This report describes three methods we have employed for measure-
ment of atmospheric HNO . The first method to be discussed involves
coulometry, the second employs chemiluminescence, and the last is a filtra-
tion technique.
MEASUREMENT OF HNO BY COULOMETRY
Instrumentation and Methods
A schematic of the microcoulometric monitor is shown in Figure 1.
Table 1 presents a list of potential interferences we have investigated. At
atmospheric concentrations, none of the species studied will significantly in-
terfere with the HNO measurement.
27
-------
TIMER
ACID
DETECTOR
RECORDER
SAMPLE
INLET
C,H INLET
Figure 1. Microcoulometric monitor.
[HNO,], PPM
10 20 30
40 50 60
TIME, MIN
100
Figure 2. Comparison of HNO analysis by FTS-LPIR and coulometric methods.
28
-------
TABLE 1. EVALUATION OF INTERFERENCES TO COULOMETRIC HNO MONITORING
Species Comment
O No interference up to 0.8 ppm
NO . No interference up to 1 ppm
SO No interference up to 1.1 ppm
H SO. No interference up to 100 yg/m
HC1 No interference up to 0.4 ppm
HCHO No interference up to 0.5 ppm
PAN No interference up to 0.2 ppm
HCOOH Variable interference between 0-10%
of HCOOH concentration
HNO2 No interference up to 0.4 ppm
Results
A comparison of the coulometric instrument with FTS-LPIR for gaseous HNO
is shown in Figure 2. Agreement between the two methods is quite good. A
similar comparison under simulated smog conditions (in the EPA smog chamber)
also showed good agreement, even in the presence of high concentrations of
potential interferences such as SO?, H SO , NO , O , PAN, and HCOOH. This
comparison is shown in Figure 3.
MEASUREMENT OF HNO BY CHEMILUMINESCENCE
Instrumentation and Methods
While the coulometric instrument can be used for many HNO_ monitoring
applications, it is not well suited for all sampling requirements. Its
detection limit (~5 ppb) and relatively long response time make it partic-
ularly unsuited for either clean air measurements or aircraft sampling in
plumes. To accommodate these situations we have developed two other methods
29
-------
2.0
1.5
CONCENTRATION, PPM
1.0
0.5
Block lamps
kdlNO/
lamps on
i 1 1 1 r
LPIR-FTS
COULOMETRIC
- Propylene
- Ozone
s- uzone
' X
\ /-Nitrogen dioxide \ Formic °cld7
V/r Sulfur dioxide *_•• J..
" \—
*- Peroxylocetyl nitrate
,...^ ^s- Nitric acid
' £>• :«*.*.-" «* s
t-°^ I r—-•---4. I I
10:15 11:00 12:00
TIME
1:00
Figure 3. Photooxidation of propylene/nitrogen dioxide/sulfur dioxide in
FTS-LPIR chamber.
SompK in*l
Pwmp
Figure 4. Schematic of chemiluminescent instrument modified for simultaneous
NO /HNO, monitoring.
30
-------
for measuring HNO . One of these methods is a simple modification of the
chemiluminescence procedure used for monitoring NO. A schematic of the
instrument is shown in Figure 4.
Results
The chemiluminescence instrument agrees quite well with the coulometric
for gaseous HNO , as seen in Figure 5. The performance of the chemilumines-
cence monitor when exposed to NO, NO_, and HNO is shown in Figure 6.
A study of the response of the instrument to varying converter temperature
is shown in Figure 7. For molybdenum converters, 300-325° C is the optimum
operating range. At these temperatures, complete conversion of NO and HNO
X O
is obtained without conversion of NH .
Some actual atmospheric data from our recent study of NO reactions in
urban plumes are shown in Figures 8 and 9. A traverse of the Boston urban
plume approximately 40 km downwind of the city (east over the ocean) is pic-
tured in Figure 8. Both the HNO and O concentrations are shown. Figure 9
shows the vertical profiles of several variables downwind of Philadelphia
during inversion conditions. These data demonstrate the utility of the
chemiluminescence instrument for airborne studies.
MEASUREMENT OF HNO BY FILTRATION
Low concentrations of gaseous HNO can be quantitatively collected on
nylon filters without significant interference from other atmospheric con-
stituents. Analysis of the nylon filter for nitrate provides an integrated
measure of the HNO concentration. We employ Teflon prefilters to separate
particulate nitrate from gaseous HNO . Thus the analysis of the two filters
yields a simultaneous estimate of the particulate nitrate and HNO concentra-
tions .
The dual filter procedure has been tested in the laboratory and recently
used during a field study. The results of the field test are not available.
31
-------
COULOMETRIC
RESPONSE, PPM
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
CHEMILUMINESCENCE RESPONSE, PPM
Figure 5. Coulometric vs. chemiluminescent response to gaseous HNO_
; 4
i
NO, « MHO,
NO,
|
w
K
s
I§J
H
O
z
X
Figure 6. Chemiluminescence response.
32
-------
Figure 7. Chemiluminescence response under varying conditions.
BOSTON PLUME TRAVERSE(40 km)
13:30 8-14-78
O 03
• HN03
ISO
160
140
o 120
Q.
0.
«-i 100
80
60
40
10
8 m
6 £
0
0 9 10 15 20 25 30 35
DISTANCE THROUGH PLUME, MILES
Figure 8. O and HNO during urban plume traverse.
33
-------
DOWNWIND PHILADELPHIA, 8-14-78
8000
6000
ALTITUDE,
FT (MSL)
4000
2000
11 15 19 23 27
TEMPERATURE, °C
0369 30 70 110
90 "° 17° NO., PPB « 3 6 9
OZONE, PPB HONO»PPB PART./CC <» 10>)
Figure 9. Vertical profiles downwind of Philadelphia.
34
-------
Manufacture of nylon filters has been discontinued by Millipore Corpora-
tion. However, similar nylon filters will soon be available from Ghia Corpora-
tion (7071A Commerce Circle, Pleasanton, California 94566).
35
-------
CHEMILUMINESCENCE MEASUREMENTS OF HNO IN AIR
Thomas J. Kelly
Donald H. Stedman
Departments of Chemistry and Atmospheric and Oceanic Science
University of Michigan
Ann Arbor, Michigan 48109
INTRODUCTION
Knowledge of the concentration of HNO in the atmosphere is important
for an understanding of the chemistry of both polluted and unpolluted air.
HNO is the major end product in the tropospheric photochemical transforma-
tion of NO . In regions of high emissions, HNO formation results in in-
X -J
creased acidity of precipitation, causing economic and ecological damage.
HNO has been measured in polluted air by a coulometric technique (Miller
and Spicer, 1975} and by an LPIR method (Tuazon et al., 1978), and in the
clean troposphere by a selective filter technique (Huebert and Lazrus, 1978).
Drawbacks of these methods include poor sensitivity and/or long analysis time.
The chemiluminescence method of Joseph and Spicer (1978) does not distinguish
the components of NO (other than HNO ).
X .3
We have measured HNO by adaptation of a sensitive chemiluminescence
NO monitor described elsewhere (Ritter e~t aZ., 1978). Measurements were made
X.
at the site of the National Oceanic and Atmospheric Administration (NOAA)
Boulder Atmospheric Observatory 300-m tower, in a rural area about 15 mi east
of Boulder, Colorado. All measurements were made near ground level, during
the last two weeks of February 1978. The site is sufficiently far from local
37
-------
sources that, though it is often heavily polluted, the pollutants are well
mixed. In addition to the measurements of HNO , NO and NO were measured by
the same chemiluminescence instrument used for HNO,, CO was measured by in-
frared absorption, and O was measured by chemiluminescence with ethylene.
INSTRUMENTATION AND METHODS
A schematic diagram of the HNO detection system is presented in Figure
1. The single sample inlet point is a glass frit connected to a glass and
Teflon plumbing system. The detector has four sample flow channels corre-
sponding to four solenoid valves in the inlet plumbing that are controlled
automatically by the detector electronics. When VI is opened, sample air
passes directly into the reaction chamber where any NO in the sample is de-
tected. Alternatively, a background reading can be taken by diverting the
sample flow to a prereactor where it is mixed with the O flow and the chem-
iluminescent reaction completed before passage into the detection chamber.
Opening V2 passes the sample air through a pot of granular iron sulfate
(FeSO ), which converts NO_ to NO. Thus, the signal from Channel 2 indicates
the sum of NO + NO_, and NO is obtained by subtraction of the Channel 1 NO
signal. Channel 3 passes the sample air through a trap of loose cotton or
nylon fiber, selectively removing HNO from the sample (Miller and Spicer,
1975). The stream then passes through the HNO converter. Channel 4 goes
directly into the converter (a quartz tube at 350° C packed with 4 mm diameter
Pyrex beads), which quantitatively converts HNO to NO . The FeSO completes
the conversion to NO. The quartz tube is hot enough to also cause detection
of organic nitrates or PAN, but does 'rwb oxidize organic amines or ammonia.
Any difference in signal between Channels 2 and 3 is a measure of the "total
organic nitrate" present, though the individual compounds in that total are
not identified. HNO is thus seen as the difference in signal between Chan-
nels 3 .and 4. .,„......
The detector electronics switch sequentially through the four channels
and a background reading in 1 min. The system has been calibrated directly
with ppb quantities of HNO in air as well as with standard mixtures of NO or
NO . We have found no loss of HNO in the inlet plumbing, and have shown
^ ~J
38
-------
GLASS FRIT
SOLENOID
HNO, TRAP
HN03
CONVERTOR
FeSO,
REACTOR
Figure 1. Instrumentation for measurement of NO by chemiluminescence.
X
-------
3U
40
O
2
20
10
E
s 4
O 3
0 5
2
1
i | i i i i i | i | i i i i i i i i
\ ! ' !
- NOX '| I ft "r
~"~~~" r\y Uo I ' In I
I f JV ' •
1! * /; li if
II 1 I 1 n\ II
1 'N ^ I f
1 1 * '\ I 1
/ 'L/ \ 1 ^ l^1 iv ! "
/^v * / « \ i JsX/"\ \ /
y PV£./ ^^ // ^ v • / v^ N-^V
^^ — ' N. /
— o3 h
_. f\ f^L 1 « ~"
r\ ~~~— ^\j j
'V'si A ^ /i
\ A / I / \. 4-
1 , 1 f '\_.^ ; i / 'A._ a
i Ali/V \! \ 1 , / v"\>
\ AM y v V-' \ r- / \?
W V\ ' V \ ;k I 4 . \h
!i ' A «•• !*:\ S A'
-:: h :\ /\ j( /\- -''"N J l»; '"'*'
..; ?•••••!—• 'C*/ V* "i"" f*":* i '"*! I** '*T*"T" \ Li t 1 i 1 i
2.0
o.
a.
M
j_
1.0
60
a.
40 Q-
10
O
20
O
MNMNMNMNMN
2/20/78 2/21/78 2/22/78 2/23/78
Figure 2. O , CO, NO , and HO levels at Boulder, Colorado, February 20-23,
j X ^ ^
1978, as measured by chemiluminescence.
40
-------
HNO, -NO, DATA 2/17-2/26/78
ERIE TOWER
50
40
jo 30
ex
Q.
CM
O
20 -
10;
o 2/17/78
• 2/18
02/19
• 2/20
Q2/2I
• 2/22
A 2/23
A 2/24
x2/25
+ 2/26
00
8
HN03 ppb
Figure 3. HNO and N0» levels at Boulder, Colorado, February 17-26, 1978, as measured by chemilumines-
-J £1
cence.
-------
HNO to be efficiently converted by the quartz tube and trapped by the fiber
trap. The detection limit of the system is presently about 0.3 ppb.
RESULTS AND DISCUSSION
Figure 2 shows data for 0 , CO, NO (NO + NO9) , and HO for several days
j X ^ £ £.
in February 1978. During this period, concentrations of HNO were frequently
below the detection limit. Figure 3 shows HNO and simultaneous NO data for
those periods when HNO was relatively constant for half an hour or more. In
general, higher NO levels correlated with higher HNO , with HNO values of
0.5 to 5 ppb. NO was always much greater than HNO_. Our-NO /HNO.. ratios
r x 3 x 3
were mostly greater than 10, with some readings of >20 ppb NO_ with no detect-
able HNO present. The NO levels clearly indicate tropospheric pollution,
•J X
since background NO levels are less than 0.5 ppb in the Boulder area (Ritter
X
et al. , 1978). These NO /HNO ratios are in agreement with those of other in-
X .3
vestigators for the polluted troposphere (Spicer, 1977; personal communication,
1978) . In clean air, HNO is expected to be greater than NO by as much as a
•J X
factor of 10 (Liu, 1977; Fishman and Crutzen, 1977; Chameides, 1978), but
measurements to test this prediction have not been accomplished. "Total
organic nitrate" levels of a few ppb were commonly observed.
ACKNOWLEDGMENTS
We wish to thank the NOAA personnel of the Boulder Atmospheric Observatory
(particularly Dr. C. Kaimal) for cooperation, many National Center for Atmo-
spheric Research (NCAR) scientists and staff for assistance, and the National
Science Foundation (NSF) Atmospheric Science Section for financial support.
REFERENCES
Chameides, W. L. 1978. The Photochemical Role of Mropospheric Nrcrogen
Oxides. Geophys. Res. Lett. 5:17-20.
Fishman, J., and P. J. Crutzen. 1977. A Numerical Study of Tropospheric
Photochemistry Using a One-Dimensional Model. J. Geophys. Res. 82:5897-
5906.
42
-------
Huebert, B. J., and A. L. Lazrus. 1978. Presentation before the American
Chemical Society, Environmental Chemistry Division, Anaheim, California,
12-17 March.
Joseph, D. W., and C. W. Spicer. 1978. Chemiluminescence Method for Atmo-
spheric Monitoring of Nitric Acid and Nitrogen Oxides. Anal. Chem. 50:
1400-1403.
Liu, S. C. 1977. Possible Effects on Tropospheric O and OH Due to NO
Emissions. Geophys. Res. Lett. 4:325-328.
Miller, D. F., and C. W. Spicer. 1975. Measurement of Nitric Acid in Smog.
J. Air Poll. Control Assoc. 25:940-942.
Ritter, J. A., D. H. Stedman, and T. J. Kelly. 1978. Ground Level Measurements
of NO, NO , and O in Rural Air. Presentation before the American Chemical
Society Symposium on Atmospheric Nitrogen Compounds, Anaheim, California,
12-17 March.
Spicer, C. W. 1977. The Fate of Nitrogen Oxides in the Atmosphere. Adv.
Environ. Sci. Tech. (J. N. Pitts, Jr., and R. L. Metcalf, Eds.). 7:163-
261.
Tuazon, E. C., R. A. Graham, A. M. Winer, R. R. Easton, J. N. Pitts, Jr.,
and P. L. Hanst. 1978. A Kilometer Pathlength Fourier-Transform
Infrared System for the Study of Trace Pollutants in Ambient and Synthetic
Atmospheres. Atmos. Environ. 12:865-875.
43
-------
INDOPHENOL AMMONIA TEST IN MEASUREMENT OF HN03 AND
Allen L. Lazrus
B. W. Gandrud
J. P. Greenberg
The National Center for Atmospheric Research
Post Office Box 3000
Boulder, Colorado 80307
INTRODUCTION
Lazrus e~t al. (1968) described a technique for a total inorganic fixed
nitrogen test utilizing a rapid reduction of NO and nitrite ion (NO«) to NH
by a zinc-copper (Zn-Cu) reduction column, followed by an indophenol ammonia
test. Since then, many improvements in the indophenol NH test (Harwood and
Huyser, 1970a,b) have been incorporated. The test, in automated form, is
performed on the Auto-Analyzer II, and is capable of yielding total inorganic
fixed nitrogen and NH separately, depending on whether the sample passes
through the reduction column. Using the selectivity of the sampling system,
the analysis may be used to measure HNO vapor and NO aerosol. A combinatior
of a Teflon prefil-ter and nylon filter is suitable for this purpose.
INSTRUMENTATION AND METHODS
Reagents
Sodium Phenate—
Dissolve 85.4 g phenol and 200 ml 5 N sodium hydroxide (NaOH) in deionized
HO, diluting to 1 liter.
45
-------
Nitroprusside—
Dissolve 1.32 g sodium nitroprusside (Na Fe(CN)^NO) • 2H O in deionized
HO, diluting to 500 ml.
Sodium Hypochlorite—
Dilute 62.8 ml commercial bleach (with 3.5% available chlorine) to 500 ml
with deionized H_O.
Ethylenediaminetetraacetic Acid-
Dissolve 37.4 g tetrasodium salt of ethylenediaminetetraacetic acid (EDTA)
in 1 liter deionized HO. Do not heat.
Electrolyte-
Dissolve 262.5 g NaCl in deionized HO. Add 40 ml of 1 N HC1 and dilute
to 1 liter.
Zn-Cu (Lazrus et al., 1968)—
Dissolve 1.0 g cuprous chloride (CuCl) • 2H O in 80 ml HO containing 12
drops of 1 N HC1. To 40 g 20-mesh Zn granules add 40 ml concentrated HC1
(diluted 1 to 5 with HO), and etch the Zn for 1 min. Pour out the acid and
wash the An four or five times with HO. Quickly add the CuCl solution to the
Zn, and shake vigorously until the blue color just disappears. Immediately
swirl the alloy into a filter in a buchner funnel under suction. Wash once
with water and then once with methyl alcohol. Let air be drawn past the
granules until they no longer stick together; spread the Zn on a large piece
of filter paper to dry. Place the dry alloy in a glass-stoppered test tube,
and shake vigorously to remove loosely adhering Cu. Separate the Cu powder
from the alloy granules. (A sieve to eliminate the fines works nicely.) Fill
a glass tube (4-mm i.d., 5-in length) with the alloy, and put a glass wool
plug on each end of "the tube to contain the granules.
Manifolds
The procedure requires a manifold that satisfies the conditions deter-
mined by Harwood and Huyser (1970a,b) and the reduction of Lazrus et al. (1968)
46
-------
The manifold schematic is shown in Figure 1. Removal of the Zn-Cu reducing
column allows the determination of NH.. With the reducing column in place,
NO and NO are reduced to NH , and the color developed is proportional to
NO~-NH*, NO~-NH* and NH*-NH* (The convention NO~-NH* signifies that the
j~r^4t 4i 4r ^ ~x
nitrogen is in the form of a nitrite ion but is expressed as NH.)
Interference filters that transmit 630 ym light are used in the color-
imeter. The reference side of the colorimeter uses only air in the sample
path.
Ghia nylon membrane filters, 47-mm diameter, 1.2-um pore size, are used in
the procedure. The filters are extracted with 0.1 N NaOH in a nonionic deter-
gent (5 ppm Brij 35). The extracting solution and filter are subjected to
ultrasonic treatment for 15 min, neutralized with HC1 using an automatic
pipette, and analyzed for NO and NH .
The detection limit of the test in its present configuration is 5 ppb N
in HO, utilizing 0.6 ml HO at a rate of 30 samples per hour.
£* £•
RESULTS AND DISCUSSION
In measuring NH , the test is effective between pH 2 and pH 8.5. NH
begins to disappear from solutions more basic than this (D'Elia e'b al., 1977).
Samples containing only nitrate nitrogen show no variability between pH values
2 and 10. Since the automated system is closed and the sample is acidified
during reduction, no NH can be lost. At pH 1, the test results are erratic.
As shown in Table 1, the precision of the test appears to be a function of
concentration.
The test appears to be free of interferences from organically-bound
nitrogen at concentrations in which they typically occur. The flight inter-
ferences occur in the indophenol NH test, and they appear to be unaffected by
the presence of the reduction column. The interferences may result from NH-
47
-------
/I\ Waste air (free)
Sample® Sampler
^0.056 (yellow)
Zn-Cu reducing column
only for total N
(black)
for NH4 only
(orange/white)
V
Return®0.065 (blue)
00
(orange)
(black)
Figure 1.
1" '
, ) cell
\ v^v - A vy
CD
Waste thru
— . Dumn 6ft
0.081
(orange/yellow)
(orange/yellow) ,
(purple)
NH+ and total inorganic fixed N manifold. Range: 0.05-2 ppm for NH ; after dilution, 0.05-
10.0 ppm NH4. Wavelength: 610-18 nm. Sample rate: 20/hr. A, B, C: Five-turn mixing coils.
D: 15-20 min delay coil. •: Solvaflex tubing. Reagents: (1) Sodium phenate 0.035, (2) nit-
roprusside 0.020, (3) sodium hypochlorite 0.02, (4) EDTA tetrasodium salt 0.025, (5) electro-
lyte 0.040, (6) air bubbled through 50% H SO , 0.030. Tubing size is color-coded.
-------
TABLE 1. PRECISION AS A FUNCTION OF CONCENTRATION
Concentration
as NH (ppm)
0.05
0.10
0.25
0.50
5.0
Coefficient of Variance
NH+ NO:;
5.9 2.0
3.13 2.4
0.46 1.1
0.66 0.69
0.36 0.22
(1%)
NO~
2.1
2.7
1.6
0.54
0.43
TABLE 2. ORGANIC NITROGEN COMPOUND ANALYSIS FOR NH
Solution (100 ppm NH ) Response as NH.
ethyl amine HC1 0.94
diethylamine HC1 0.1
hydrazine 2HC1 0.1
aminophenol 1.1
alanine 0.1
aniline sulfate 4.3
dZ.-phenylalanine 0.15
-------
contamination of the organic reagents. Solutions of organic compounds equiv-
alent to 100 ppm NH were analyzed. The NH respc
Hydroxylamine reacts as an inorganic constituent.
alent to 100 ppm NH were analyzed. The NH response is shown in Table 2.
The Ghia nylon membrane filters efficiently collect HNO vapor. No NH
O TT
blank was observed.
REFERENCES
Lazrus, A., E. Lorange, and J. P. Lodge, Jr. 1968. New Automated Microanalyses
for Total Inorganic Fixed Nitrogen and for Sulfate Ion in Water. Adv.
in Chem. Series #73:164-171.
Harwood, J. E., and D. J. Huyser. 1970a. Automated Analysis of Ammonia in
Water. Water Res. 4:695-704.
Harwood, J. E., and D. J. Huyser. 1970b. Some Aspects of the Phenol-
hypochlorite Reaction as Applied to Ammonia Analyses. Water Res. 4:
501-515.
D'Elia> C. F., P. A. Stendler, and N. Corwin. 1977. Determination of Total
Nitrogen in Aqueous Samples Using Persulfate Digestion. Limnol. and
Oceanog. 20(4):760-764.
50
-------
DETERMINATION OF ATMOSPHERIC HNO
WITH NaCl-IMPREGNATED FILTERS AT HIGH VOLUME FLOW RATES
Joseph Forrest
Roger L. Tanner
Daniel Spandau
Ted D'Ottavio
Leonard Newman
Brookhaven National Laboratory
Upton, New York 11973
INTRODUCTION
The existence in the atmosphere of gaseous nitrates, particularly HNO ,
has been well documented (Spicer, 1977; Okita et ai., 1976; Okita and Ohta,
1978). Moreover, it appears that the concentration in some localities of
gaseous nitrates exceeds that of particulate nitrates. Any attempt to under-
stand the formation mechanisms, transformations, and relationships with acid
precipitation of nitrogen oxides must necessarily involve the accurate deter-
mination of atmospheric HNO .
INSTRUMENTATION, METHODS, AND RESULTS
Experiments were conducted to explore the possibility of extending the
standard Brookhaven National Laboratory (BNL) 5-in diameter high-volume (hi-
vol) SO_ and SO filter pack to permit measurement of plume HNO concentrations.
The nylon filters described by Spicer were no longer commercially available
(and in all probability would not have had the proper flow characteristics for
hi-vol employment). Efforts were therefore concentrated on the NaCl-impregnated
51
-------
cellulose paper technique of Okita, using Schleicher and Schuell Fast Flow 2W
filters to achieve maximum flow rates. Sampled papers were extracted in warm
HO and nitrate measured by a hydrazine reduction-diazotization auto analyzer
technique.
When placed downstream from acid-pretreated quartz particulate filters in
a hi-vol sampler, single NaCl-impregnated filters collected HNO vapor at
efficiencies of about 90%. By combining two filters on a single screen,
efficiency was improved to >95% with negligible decrease in flow rate.
Some adsorption of N
-------
TABLE 1. ADSORPTION OF NO BY NaCl FILTERS IN QUARTZ-NaCl-NaCl FILTER PACK*
w
Date
4/20
4/20
4/24
4/26
4/26
7/5
7/11
Sampling
Time,
min
60
60
60
90
90
60
60
N02
Concentration ,
ppm
0.91
0.91
0.91
0.53
0.53
1.0
2.0
Total
N02,
mg
41
41
41
36
36
16
32
Flow
Rate,
ft min
14
14
14
14
14
5
5
Temperature,
op
70
50
70
70
45
70
70
Relative
Humidity,
%
52
95-60
52
52
70
52
52
NO in NaCl
Backup
Filter, yg
3
5
3
4
7
3
3
*Double 4-in diameter NaCl filters on each screen.
-------
TABLE 2. CONVERSION OF NO TO VOLATILE HNO BY QUARTZ FILTERS
IN QUARTZ-NaCl-QUARTZ FILTER PACK
Sampling
Time,
Date hr
6/27-28
6/28
£ 6/28
7/27-28
7/5-6
8/7-8
7/11
7/11
16
2
2
16
16
16
1
1
Concentration, 2'
ppm mg
ambient —
ambient —
ambient —
ambient —
0.010 6.3
0.0053 4.4
1.0 16
2.0 33
Relative
Temperature , Humidity ,
°F %
72
84
84
72
61
74
72
72
97
60
60
98
86
100
70
70
Artifact
HNO in
Third NaCl
Filter, yg
77
6.8
7.9
35
50
13
3.4
2.3
NO
2 Artifact HNO..,
Converted 3
to HNO Total HNO , *
- 3.6
- 3.0
- 5.9
- 7.3
0.57 7.9
0.15 11.6
0.02 -
0.007 -
*HNO concentration range = 0.10-2.9 ppb.
-------
Ul
Cn
TABLE 3. CONVERSION OF NO TO RETAINED NO BY QUARTZ FILTERS IN
QUARTZ-NaCl-NaCl-QUARTZ-NaCl FILTER PACK
Sampling
Time,
Date
6/27-28
6/28
6/28
7/27-28
7/5-6
8/7-8
7/11
7/11
hr.
16
2
2
16
16
16
1
1
N02
Concentration
ppm
ambient
ambient
ambient
ambient
0.010
0.0053
1.0
2.0
Total „ . ^.
NO Relative
, 2' Temperature, Humidity,
mg °F
- 72
- 84
- 84
- 72
6.3 61
4.4 74
16 72
33 72
%
97
60
60
98
86
100
70
70
Artifact
NO in
Second
Quartz
Filter, yg
21
3
8
22
31
28
2
4
N02
Converted , ^ . _ ^ -
Artifact NO /
t0 N°3' Total Particu-
% late NO ,* %
- 10
5.0
- 4.8
- 4.0
0.37 3.6
0.47 8.9
0.01 -
0.01 -
*Particulate NO concentration range = 0.7-4.4 pg/m .
-------
At high humidities, adsorption of HNO on the quartz prefilter became
significant, suggesting that relative retention varies inversely with HNO
concentration and directly with relative humidity (Figure 1). However,
adsorption losses could be minimized by desorption in a post-collection, 10-
min sampling period in which drier air is provided by a heat gun.
Loss of collected particulate nitrate during sampling by on-filter
reaction with ambient H2SO. to form volatile HNO has long been suspected.
The extent of such loss was evaluated by preloading quartz filters with atmo-
spheric particulates (16-hr hi-vol sampling). Opposite quarters of the filters
were masked and the remaining areas exposed to generated H SO aerosol.
Circles (1-in diameter) were cut from each quarter and analyzed for NO and SO..
Differences between exposed and covered sections ranged from 10 to 25 yg NO
3 -
per mg of collected H SO or (for 16-hr samples of ~1 yg/m NO ) from 4 to
16% of the particulate nitrate per mg of H SO (Table 4) . Some or all of the
NO losses may possibly have arisen from the reaction of H?SO. with adsorbed
HNO . After deposition of the aerosol H SO., the permitted reaction time
prior to extraction and analysis of the filters was not carefully controlled.
Both of these factors will require further consideration in future experiments.
Cellulose filters impregnated with potassium carbonate (K CO ) were
placed in series downstream from quartz and NaCl papers and the filter pack
subjected to 0.07 to 1 ppm SO atmospheres. No losses of SO were detected on
the preceding filters, indicating the integrated filter pack could be safely
used for SO , SO_, and HNO analyses (Table 5).
Ambient measurements of particulate NO and HNO concentrations were made
.3 -J
over 2- to 16-hr periods. In only 1 of 13 periods was the particulate NO
level decidedly greater than that of HNO ; in 2 cases, the levels were approxi-
mately equal (Table 6). In the remaining 10 periods, the fraction of gaseous
nitrate to total nitrate ranged from 0.59 to 0.92, with an average ratio of
0.71 for the 13 measurements (Table 6). Standard deviation for eight dupli-
cate analyses (including errors in hi-vol flow measurements) at HNO levels of
56
-------
Ul
_20
z
LJ 18
o
5 16
o" 14
UJ
S l2
O
g I0
< 8
IO
o
13-35 ppb HN03
HN0
1-7 ppb HN03
100-300 ;jg HNO-
10 20 30 40 50 60
RELATIVE HUMIDITY, PERCENT
70
80
90
100
Figure 1. Adsorption of HNO by quartz filters.
-------
TABLE 4. LOSS OF PARTICULATE NO BY REACTION WITH
AEROSOL H SO ON PRELOADED QUARTZ FILTERS
t
No.
1
2
3
4
5
6
7
8
N0~,
vg
231
235
241
182
217
365
127
123
H2S°4
Added ,
mg
3.14
1.00
1.70
1.94
2.54
2.31
1.56
4.06
N0~,
Lost,
%
12.9
0.90
17.6
10.5
20.9
14.1
24.8
39.5
N0~,
Lost per
mg H2SO4, pg
9.6
2.5
24.9
9.5
17.6
22.0
21.5
11.6
TABLE 5. ADSORPTION OF SO BY K CO AND NaCl FILTERS
No.
1
2
3
4
Sampling
Time,
nij.n
30
30
7
15
SO Concentration,
ppm
0.07
0.30
0.80
1.07
so2
K2C°3
Filter
2.14
11.0
6.69
1S.O
Adsorbed , mg
NaCl
Filter
0.042
0.029
0.050
0.051
58
-------
TABLE 6. COMPARISON OF PARTICULATE AND GASEOUS NITRATES
Concentration,
Particulate
NO~
1.16
0.28
0.82
2.27
0.55
2.24
3.94
0.55
0.66
1.30
0.88
2.63
3.59
yg/m
Gaseous
HN03
1.21
1.62
1.74
3.27
5.07
5.94
1.99
4.25
7.69
5.76
3.89
2.48
16.0
Gaseous HNO /NO *
j X
0.34
0.85
0.68
0.59
0.90
0.73
0.34
0.89
0.92
0.82
0.82
0.49
0.82
*Average = 0.71.
59
-------
TABLE 7.
PRECISION OF HNO ANALYSES*
Date
6/14
6/19
6/26
6/27
7/5
7/10
7/11
7/18
Sampling
Time,
hr
18
17
16
16
16
2
2
2
HNO Concentration,
Sample A
0.68
1.97
1.66
2.88
0.74
2.28
1.39
5.86
ppb
Sample B
0.68
2.17
1.28
2.99
0.96
2.24
1.51
6.29
*Relative standard deviation = 12%.
60
-------
0.7-6 ppb was ~12% (Table 7). Minimum detectable concentration (defined as 3
x blank) for a 2-hr sample at flows of 20 ft min was ~0.1 ppb HNO .
ACKNOWLEDGMENTS
This work was performed under Contract No. EY-76-C-02-0016 with the U.S.
Department of Energy.
REFERENCES
Spicer, C. W. 1977. Photochemical Atmospheric Pollutants Derived from
Nitrogen Oxides. Atmos. Environ. 11:1089.
Okita, T., S. Morimoto, M. Izawa, and S. Konns. 1976. Measurement of Gaseous
and Particulate Nitrates in the Atmosphere. Atmos. Environ. 10:1085.
Okita, T., and S. Ohta. 1978. Measurement of Nitrogen Compounds in the
Atmosphere. Presentation before the American Chemical Society Symposium
on Atmospheric Nitrogen Compounds, Anaheim, California, 12-17 March.
61
-------
SELECTIVE COLLECTION AND MEASUREMENT OF GASEOUS HNO IN AMBIENT AIR
Richard J. Hare
M. T. Wininger
William D. Ross
Monsanto Research Corporation
Post Office Box 8, Station B
Dayton, Ohio 45407
INTRODUCTION
Nitrates were found to accumulate on glass fiber filters when hi-vol air
samplers were used to collect particulates; these accumulations were caused by
interactions of gaseous nitrogen compounds with the filters. NO» and HNO
£• J
vapor are both capable of interacting with the glass fiber filter under sam-
pling conditions normally used for determination of particulate nitrate in
air. Consequently, most earlier measurements of particulate nitrate are
probably in serious error.
Several techniques have been used to determine atmospheric HNO . One
such technique is chemiluminescence monitoring; another is the Mast coulo-
metric method, which requires exacting control of conditions, is expensive,
and is therefore impractical as a widespread monitoring program.
In the present study, an automated sampling technique was developed for
the collection of HNO vapor. In this technique, ambient air is analyzed for
HNO vapors by quantitatively collecting the HNO on nylon fibers, extracting
•J -3
the fibers with an NaOH solution, and converting the nitrate in an acid solu-
tion of benzene (C.H-) to form C(,HCNO0. The C^H^NO,, is then measured by
DO O J ^ D D Z
electron capture gas chromatography (ECGC).
63
-------
TABLE 1. HOUSTON FIELD STUDY MEASUREMENTS OF HNO (PPB, BY VOLUME) AUGUST 1978
Date
Sample Time, Flow Rate
Tube Position
23456
Thursday 8/24
35 min @ 3 liter/min
Friday 8/25 ,
35 min @ 3 liters/min
I,
Saturday 8/26
65 min @ 4.9 liters/min
Monday 8/28
65 min @ 4.9 liters/min
Tuesday 8/29
65 min @ 4.9 liters/min
Wednesday 8/30
65 min @ 4.9 liters/min
2.8 3.3 2.9 2.4 5.3 10.1 14.9
14.4 9.6 2.9 8.1 5.8 2.9 7.7 15.4
2.6 1.0 2.5 2.3 2.5 2.3 3.4 3.4
2.3 1.5 2.1 2.0 3.4 2.9 3.7 3.4
4.6 0.9 2.6 5.8 3.4 4.0 3.1 4.3
1.8
0.8
1.5
*Sample lost.
Sample equal to blank level.
-------
INSTRUMENTATION AND METHODS
Nitric acid vapors in air are collected on 100 mg of nylon fiber stretched
out in a 6 in x 1/4 in Teflon tube. All HNO is retained at concentrations as
low as 0.5 ppb with air flows up to 10 liters/min. There are no interferences
from exaggerated concentrations of NO and NO , even at elevated temperatures
and humidity. A Teflon particulate filter, 5 ym pore size, is required in
advance of the nylon fiber to remove particulate nitrates which, if unscreened,
can interfere with chemical analysis. Laboratory research has indicated no
retention of HNO on Teflon filters.
Desorption of HNO from the nylon fiber is accomplished by passing a 1%
aqueous solution-of NaOH through the filter tube, opposite the direction of
air flow sampling. Complete removal of as little as 0.1 \ig HNO can-be
achieved. Reacting the desorbing solution with C..H^ and concentrated H^SO.
bo 24
quantitatively converts the HNO_ to CfiH NO for ECGC analysis. Minimum detect-
-12
ability of C,.H..NO_ is 1 x 10 g with a linear response range covering five
6b 2
orders of magnitude.
RESULTS AND DISCUSSION
The sampling system and analysis method described above were recently
field tested during August 1978 in Houston, Texas. The results of the test
are shown in Table 1. The HNO ranged from 1.4 to 14.9 ppb during the course
of the study. Collection time for the first 2 days was 35 min; thereafter it
was 65 min.
The gas chromatographic method for measuring HNO is highly selective and
has excellent sensitivity. The field test measurements of HNO .were reason-
able and within expected limits. In addition, the instrumentation needed to
perform the analysis is widely available and not highly specialized.
65
-------
SELECTIVE COLLECTION AND MEASUREMENT OF PARTICULATE NITRATE
AND GASEOUS HNO IN AMBIENT AIR
John Tesch
Robert Sievers
Department of Chemistry
University of Colorado
Boulder, Colorado 80309
INTRODUCTION
The ever increasing use of fossil fuels is accompanied by an increase in
the emission of NO into the air. The fate of NO involves a number of products,
including NO , HNO vapor, and particulate nitrate. NO and particulate
^ j ^
nitrate have been measured for years, while measurement of HNO vapor has
frustrated investigators and is only now beginning to be accomplished. These
new measurements give evidence that HNO vapor is a more important member of
this family of pollutants than previously believed; this evidence also suggests
that a large number of earlier measurements of particulate nitrate might be
positively biased by accidental HNO vapor collection. This bias is confirmed
by Spicer et at. (1974) and by studies in this laboratory.
Chemiluminescence appears to be a highly reliable method for the analysis
of NO^ and NO. Recently, Joseph and Spicer (1978) reported the use of a
Chemiluminescence monitor for the determination of atmospheric HNO . Precise
control of conditions is needed for the Mast coulometric method for HNO
measurements; this technique is therefore probably unreliable for any extensive
monitoring program. A method described by Lazrus and Gandrud (1974) for the
determination of HNO vapor in the stratosphere was based on the collection of
HNO vapor by IPC-1478 filters impregnated with tetrabutyl ammonium hydroxide
67
-------
and bis-2-butoxy ethyl phthalate. Okita et al. (1976) have collected partic-
ulate nitrate on Fluoropore filters and HNO vapor on a second stage cellulose
filter impregnated with NaCl. The nitrate collected was extracted with warm
water and measured by a hydrazine reduction-diazotization colorimetric tech-
nique (Okita et al., 1976; Newman, this volume).
In the present study, we developed a sampling technique for serial col-
lection of particulate nitrate, HNO_ vapor, and/or NO_. The species collected
was analyzed by a modification of the technique described in our earlier paper
(Tesch et al., 1976) with the important difference that Teflon filters were
substituted for the previously used glass fiber filters. The technique in-
cludes conversion of the collected species to a nitroaromatic compound, which
is separated and measured readily, even in very small amounts by ECGC. Hare
et al. (this volume) recently used a partially automated modification of this
technique, substituting nylon sorbent for cotton with excellent success. Used
in an EPA field test in Houston, this technique was sufficiently sensitive to
permit measurement of particulate nitrate and of gaseous HNO. every 30 min.
No potentially interfering peaks were seen in the chromatograms.
INSTRUMENTATION AND METHODS
Collection of Particulate Nitrate
Air samples are drawn through a series of a filter and two sorbents by a
vacuum pump. Flow rates vary from 8 to 15 liters of air per min. Typically,
a 1000 liter volume of air is sampled; however, samples as small as 100 liters
are analyzed with acceptable sensitivity and precision for both particulate
nitrate and HNO vapor. Particulate nitrate is collected on 25 mm diameter
Fluoropore filters.
Collection of
The determination of HNO vapor in air is based on selective collection
by cotton. Interference from particulate nitrate is eliminated by removing
68
-------
particulate matter with a Fluoropore filter before passing the air through the
cotton. The cotton filter consists of cotton fiber loosely packed to a depth
of 3 cm in 0.8-mm i.d. glass tube. Collection studies indicate that >95% of
the first 7 yg of HNO vapor passing through this type of cotton filter is
collected on the first 0.05 g (~15 mm) of cotton. The amount collected is
sufficient to allow accurate determination of HNO vapor in air.
A mixture of concentrated HNO and H SO (1 to 3, by volume) recommended
J ^ ~r
by Fehsenfeld et at. (1975) is placed in a 1-cm bulb blown at one end of a
heavy-walled capillary tube (2.0-mm i.d., 15-cm length). HNO diffuses from
this source at ~1 ug/min at 25° C. Nitrogen gas is passed at 60 ml/min through
copper coils in a constant temperature bath and through a constant temperature
chamber containing the HNO diffusion tube. The system is allowed to equili-
brate overnight to saturate the sites on the glass walls of the tubing.
NO and NO from standard sources are passed through cotton filters; <5%
is retained. When a cotton filter is placed in the inlet of a chemilumines-
cence detector, the concentrations of NO_ and NO read by the detector are vir-
tually unchanged from those measured without the cotton filter. Repetitions
of the experiment at several NO» concentrations have indicated little or no
detectable retention of NO_ or NO by cotton filters.
Analysis of Collected Sample
For each nitrogen species collected, analysis begins with conversion to
a nitroaromatic compound such as C..HCNO (Ross et ai., 1976; Tesch et al.,
b J 2
1976). The nitroaromatic is separated and quantitated by ECGC. Investigations
carried out after completion of the present study indicate that toluene (C_H )
7 o
can be substituted for the more toxic C^H. satisfactorily, so long as account
b b
is taken of nitrotoluene isomers (Wizner and Sievers, unpublished data).
For determination of particulate nitrate, the Fluoropore filter is rolled
and placed directly into a 2-dram borosilicate glass reaction vial; 0.20 ml
H_O, 1.00 ml C^H.., and 1.00 ml concentrated H_SO. are added. The vial is
2 bo 24
69
-------
capped and shaken for 10 min to allow the nitration reaction to proceed to
essential completion. The upper C..H,. layer is removed, and a 1-yl aliquot is
b b
injected into the ECGC instrument. The peak height of the C H NO generated
O J £•
is compared with a calibration plot generated from injection of standard
concentrations of CtH.-NO_ in C H . Chromatographic conditions are described
b D 2. bo
below.
The first 0.05 g (~4 mm) of cotton are removed from the inlet of the
cotton fiber filter, placed directly into a vial, and analyzed by the same
method as earlier described for the Fluoropore filters. A second 0.05-g
portion of the sampled cotton and an unsampled portion also are analyzed to
assure complete collection by the first piece of cotton and to allow for
measurement of a blank concentration due to contaminant nitrate on the cotton.
If more than 0.08 g cotton is placed in the reaction vial, incomplete nitra-
tion results.
The larger volume of the molecular sieves requires that they be analyzed
in a 50-ml Erlenmeyer flask with a ground glass stopper. Reagent amounts were
increased tenfold; HO is replaced by 0.1 N HO. Addition of HO is a
necessary oxidizing process because NO is disproportionated in HO to form
equal amounts of NO and NO , and the latter would not be converted to C H,.NO .
j ^ D J ^
Occasionally, up to 5 ml HO is added after the reaction is complete to dis-
place the C^H. from the molecular sieves for analysis. Because of the larger
b b
amount of reagents required, sensitivity is significantly decreased. For this
reason, chemiluminescence detection is probably a preferred method for NO
measurement.
Apparatus and Reagents
Gas Chromatographic Conditions—
Chromatographic analyses are performed on Hewlett-Packard Models 402 and
5750 gas chromatographs equipped witn NJ. and citaniura tritide ^lociro..
capture detectors (fixed frequency, variable current), respectively. The
columns contain 1.5% SE-30 polydimethylsiloxane coated on 80-100 mesh acid-
washed Chromosorb G, packed in glass tubing (76 cm x 4-mm i.d.) or heavy-walled
70
-------
Teflon tubing (51 cm x 3-mm i.d.). Isothermal column temperatures between
100° and 190° C are used for various experiments, and carrier gas flow rates
of 40 and 60 ml/min of 90/10 Ar/CH are employed. The detector temperature
most commonly is 180° C; the injection port temperature is 200° C.
Hi-Vol Air Sampler-
Airborne particulate samples are collected over 24-hr periods with a
Precision Scientific hi-vol sampler meeting American Society for Testing and
Materials (ASTM) standard D2009.
Chemiluminescent Analyzer—
Chemiluminescent measurements of NO and NO? are made using a Model 14D
dual chamber NO-NO«-NO analyzer from Thermo Electron Corporation, 85 First
^ X
Avenue, Waltham, Massachusetts 02154.
Reagents-
Reagent grade H^.SO. and thiophene-free, reagent grade C-H- are available
£. 4 DO
from both Fisher Scientific and J. T. Baker, Inc. Precautions noted by Ross
e~t at. (1976) with regard to H SO purity should be observed. Both the CrHr
2 4 66
and the H SO may be used without additional purification. Other liquid
organic reagents are supplied by commercial sources and are purified by frac-
tional distillation at least once prior to use. All other reagents are used
without further purification.
Filters-
Glass fiber filters (8 in x 10 in) meeting ASTM D2009 requirements are
supplied by Fisher Scientific Company (Catalog No. 1-037-048). Quartz filters
(8 in x 10 in, 2500 QAST) are obtained from Pallflex Products Corporation,
Kennedy Drive, Putnam, Connecticut 06260. U.S. Pharmacopoeia (USP) sterilized
absorbent cotton (1 Ib. roll, seamless, No. 96-9116) is available from Swans-
down Corporation, New Haven, Connecticut 06503. Nylon fiber wool is a pro-
duct of Atlas Electric Company. Pyrex wool filtering fiber (Catalog No. 3950)
and porous glass beads may be obtained from Corning Glass Works, Corning, New
York.
71
-------
Uni-Pore polycarbonate membrane filters (25-mm diameter, 0.8-ym pore
size, Catalog No. 312-0059) are supplied by Bio-Rad Laboratories, 2200 Wright
Avenue, Richmond, California 94804. Duralon (nylon, 25-mm diameter, 1.0-ym
pore size, Catalog No. NRWPO2500), Mitex (Teflon, 25-mm diameter, 5.0-ym pore
size, Catalog No. LSWPO2500), and Fluoropore (reinforced Teflon, 25-mm diameter,
0.5-ym pore size, Catalog No. FHLPO2500) filters are available from Millipore
Corporation, Bedford, Massachusetts 01730.
Molecular sieves, Linde Type 5A calcium aluminosilicate (pore size 5
angstroms, 40-60 mesh) may be obtained from Matheson, Coleman, and Bell, Inc.,
Norwood, Ohio.
Zeolon 200H molecular sieve particles have physical dimensions of 1/16-in
diameter x 1/4- to 1/2-in length and a bulk density of 47 Ib/ft . The material
is a synthetic crystalline aluminosilicate with a 10 to 1 ratio of silica to
alumina, and is very resistant to acid attack. The crystalline material has
an effective pore size of 8-9 angstroms.
Precision Wet Test Meter—
A Precision Scientific wet test gas meter (Catalog No. 63125) meeting
ASTM D1071 is used for measuring high gas flow rates and volumes when samples
are collected with other than the hi-vol sampler. Mallinckrodt CfiH NO_ (Chem
Services, West Chester, Pennsylvania) is employed to prepare calibration
standard solutions.
DISCUSSION
When standard sources of HNO vapor and NO are passed through the
Fluoropore filters, not more than 5% (the detection limit under the conditions
tested) of each is collected, even when extended sampling times and slow flow
rates (60 ml/min) are employed. In general, collection efficiencies of partic-
ulates on these filters are reported (GkiLa ^ z.1., 1975) >95p-. ?t flovj rates
up to 20 liters/min., but the collection efficiency of nitrate salts is dif-
ficult to estimate owing to the low but significant volatility of NH.NO . It
72
-------
should be borne in mind, however, that the suspended particles being collected
are already bathed in the same air which is being drawn through the filters
over the collected particles. Consequently, provided that sampling times are
kept short, losses due to volatilization of NH NO will be minimal. Further-
more, the air being drawn through contains some gaseous HNO , and sometimes
significant concentrations of NH . Therefore, loss of NH NO from the filter
is much less than would be predicted from the vapor pressure, assuming per-
fectly clean air is being drawn through the filter.
Filter/Sorbent Comparison for HNOa Vapor and NO2 Collection
Several filters and sorbents were tested for collection efficiency with
both HNO vapor and NO . Filters tested were 25-mm in diameter and held in a
polycarbonate aerosol filter holder (Bio-Rad No. 342-0004). Nitrogen streams
containing HNO vapor and NO_ at concentrations varying from a few ppb to 1
•3 £
ppm were passed through the filters at flow rates from 30 ml/min to 15 liters/
min. Filter analysis was accomplished by placing the entire filter into the
reaction vial prior to adding the HO, C^H , and H0SO.. To insure complete
Z b b z 4
recovery of C^H_NO« with the filters present, samples containing various
bo 2.
concentrations of NO in HO were added to unsampled filters and"reacted.
Percent recoveries were computed for HNO by bubbling the gas through a frit
into 50 ml of 0.1 N NaOH. The NaOH was analyzed by the general microanalysis
procedure described in an earlier paper (Tesch et a.1., 1976). Percent recoveries
for NO_ were calculated, based on the weight loss of the permeation tube.
The results of these studies are listed in Table 1. Teflon filters did
not retain detectable amounts of HNO vapor or NO?, and therefore were selected
as the material for the first stage of the collection system. Nitric acid
vapor was collected on several of the other filters and sorbents, but the
amount collected varied greatly with flow rate and concentration. These data
also show that a significant error is introduced when collecting particulate
nitrate on glass fiber filters. We attribute the high collection efficiency
of the IPC-1478 filter to its high cotton content. It should be noted that
both cotton and nylon collected HNO vapor selectively and quantitatively over
73
-------
a wide range of flows and concentrations. This remained true even when air
and nitrogen were bubbled through H^O prior to passing through the HNO or
^ J
NO sources, indicating that N0_ was not converted to HNO and collected in
the presence of HO vapor when passed through nylon or cotton at these con-
centration levels.
TABLE 1. COLLECTION EFFICIENCY OF FILTERS AND SORBENTS* FOR HNO VAPOR AND NO
Substance
Fluoropore filter
Mitex filter
Dura Ion filter
Unipore filter
Quartz filter
Glass fiber filter
IPC-1478 filter
Whatman-41 filter
Porous glass
Cotton
Nylon
Molecular sieves
Drierite
Percent
f
f
47-66
6-30
56
38-92
89-94
10-81
35-95
95-100
95-100
$
75
Percent
Recovered
t
f
f
t
t
0-22
1-5
t
32-64
t
t
55
*
*Based on 3-cm collection patn.
,No data.
tested.
Sampling Problems and Differentiating Particulate Nitrate and Gaseous HNO3
It the air contains high concentrations of H_SO. aerosols, some partic-
ulate nitrate may be displaced from the filter and collected as HNO on the
second stage cotton sorbent. Although perhaps less likely, the opposite
effect might be seen if basic sites exist in the particulates on the Teflon
74
-------
filter. If large volumes of air are passed thorugh particulates already
collected, the particulate matter may collect some HNO, vapor by reaction with
basic constituents or by adsorption. For this reason, the volume of air
passed over particles already on the filter must be kept to a minimum. It
seems clear, however, that removal of gaseous HNO on collected particulates
on Teflon filters is not nearly as great as that unavoidably collected on
glass fiber filters, which appear to have basic sites on the glass fiber
surfaces. Even relatively acidic quartz fiber filters retain some gaseous
HNO . To minimize possible artifacts, one must shorten sample times and/or
reduce linear velocities of air drawn through the filters.
Measurements of particulate nitrate and HNO_ vapor in air sampled at the
University of Colorado campus were made intermittently during the spring of
1977 using a combined two-stage filter. Sampling time was in general from 30
min to 2 hr at 8 to 15 liters/min air flow rate. Each filter was analyzed as
previously described. Results of these analyses are listed in Table 2. As
can be seen, HNO vapor is a significant contributor to air pollution, even in
relatively clean areas. These values are about a factor of 10 higher than
those found at Fritz Peak, Colorado, in a remote area high in the mountains
near the Continental Divide (Wizner and Sievers, unpublished data).
Nylon fibers have been shown by a number of workers to collect HNO vapor
(Spicer e~b al. , 1974; Hare e~k al., this volume). Although no significant
advantages are apparent for its use in preference to cotton, tests were
conducted to insure that recovery of HNO and its conversion to C H NO could
be accomplished quantitatively when nylon was the collection medium. No
interferences were found with nylon; however, the cotton fibers were finer,
giving more surface area, and total collection requied less fiter material
with cotton than nylon. Consequently, it would appear that nylon fibers can
be substituted for cotton without any compromise in the efficacy of the in-
tegrated selective measurement method described here.
Short sampling times and reduced volume throughout can be achieved only
with a highly sensitive method, such as that described here based on ECGC. In
75
-------
TABLE 2. REPRESENTATIVE MEASUREMENTS OF PARTICULATE NITRATE
AND HNO IN AMBIENT AIR*
Date
10 May
12 May
13 May
16 May
17 May
23 May
24 May
2 June
3 June
7 June
8 June
14 June
Time
1013-1113
1150-1330
1230-1350
0830-0930
1944-2109
0825-0910
0925-0956
1258-1333
1549-1632
1415-1628
1352-1609
1055-1456
1340-1545
1055-1210
1348-1518
1003-1203
Concentration
Particulate
Nitrate
2.56
2.07
0.82
0.92
0.30
0.63
0.43
0.75
0.49
0.29
0.23
0.90
0.48
0.32
0.19
1.05
(yg/m )
HN03
Vapor
0.57
1.22
0.80
0.49
0.51
0.74
0.30
0.27
0.98
0.31
0.13
0.42
0.63
0.30
0.20
0.79
*All measurements were made at the University of Colorado, Boulder, Colorado,
in 1977.
76
-------
the EPA field test (Hare et al., this volume), samples were taken every 30
min, and the signal to noise (and blank) ratio was sufficient to allow measure-
ments to be taken with much lower volumes of sampled air than required for
other methods of analysis.
ACKNOWLEDGMENTS
Nylon fiber wool was a gift from William D. Ross, Monsanto Research
Corporation. Molecular sieves (Zeolon 200H) were a gift from Gulf Oil
Corporation.
REFERENCES
Fehsenfeld, F. C., C. J. Howard, and A. L. Schmeltekopf. 1975. Gas Phase Ion
Chemistry of Nitric Acid. J. Chem. Phys. 63:2835-2841.
Joseph, D. W., and C. W. Spicer. 1978. Chemiluminescence Method for Atmos-
pheric Monitoring of Nitric Acid and Nitrogen Oxides. Anal. Chem.
50:1400-1403.
Lazrus, A. L., and B. W. Gandrud. 1974. Distribution of Stratospheric Nitric
Acid Vapor. J. Atmos. Sci. 31:1102-1108.
Okita, T., S. Morimoto, I. Michio, and S. Konno. 1976. Measurement of Gaseous
and Particulate Nitrate in the Atmosphere. Atmos. Environ. 10:1085-1089.
Ross, W. D. , G. W. Butler, T. G. Duffy, W. R. Rehg, M. T. Wininger, and R. E.
Sievers. 1975. Analysis for Aqueous Nitrites and Nitrates and Gaseous
Oxides of Nitrogen by Electron Capture Gas Chromatography. J. Chromatog.
112:710-727.
Spicer, C. W. , P. M. Schumacher, J. A. Kovyovmjian, and D. W. Joseph. 1978.
Sampling and Analytical Methodology for Atmospheric Particulate Nitrates.
Battelle Memorial Institute: Final Report. EPA-600/2-78-067, U.S.
Environmental Protection Agency.
Tesch, J. W. , W. R. Rehg, and R. E. Sievers. 1976. Microdetermination of
Nitrates and Nitrites in Saliva, Blood, Water, and Suspended Particulates
in Air by Gas Chromatography. J. Chromatog. 126:743-755.
77
-------
THE DENUDER DIFFERENCE EXPERIMENT
Robert W. Shaw
Thomas G. Dzubay
Robert K. Stevens
Inorganic Pollutant Analysis Branch
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
INTRODUCTION
Nitrates in the atmosphere may occur as solids (e.g., ammonium or metal
nitrates in particles) or as gases (e.g., HNO ) , and a difficulty has existed
in distinguishing between them because of possible nitrate artifact formation
during sampling. The problems of atmospheric nitrate artifacts are now thought
to be of two kinds: (1) the formation of particulate nitrates from gaseous
nitrates on a collection filter (positive artifacts), and (2) the loss of
particulate nitrates as gaseous nitrates due to chemical reactions on the
filter (e.g., H SO aerosol and particulate nitrates reacting to particulate
sulfates and gaseous HNO ) or evaporation. The occurrence of positive nitrate
artifacts has been reported by Stevens et at. (1978a) and Spicer et at.
(1978), and the loss of particulate nitrate by chemical reaction on the col-
lecting filter has been demonstrated in the laboratory by Harker e~t al.
(1977). The objective of the Denuder Difference Experiment (DDE) is separately
to determine particulate nitrate and gaseous HNO without influence of either
type of artifact.
79
-------
INSTRUMENTATION AND METHODS
The DDE requires two sampling assemblies. Both assemblies consist of a
Teflon particle filter followed by an HNO collection tube (Hare e~t al. , this
volume). For one of the assemblies, however, the particle filter is preceded
by an acid gas diffusion denuder (Stevens e'k al., 1978b) coated with a strong
base; under typical operating conditions, the residence time of a particle in
the denuder is approximately 0.4 seconds. This denuder will remove acid gases
(in particular, HNO ) and pass aerosol particles. Thus, the difference be-
tween the amounts of nitrate collected by the two assemblies is due only to
the removal of gaseous HNO in one assembly. The measured quantities are
expressed as follows:
Quantities of Interest
N = total atmospheric nitrate (particulate nitrate and gaseous
HN03)
N = particulate nitrate
N = gaseous HNO
H - j
Observed Quantities
F = nitrate measured on filter behind denuder
F = nitrate measured on filter in assembly without denuder
T - nitrate measured on collection tubs bshir.d dar.udcr
T = nitrate measured on collection tube in assembly without
denuder
Thus
NH = NT -
N = F + T
NP = FD + TD
80
-------
Therefore
N = (F + T) - (F + T )
ti L) U
Notice that the experiment depends on the following characteristics of
the apparatus:
(1) The diffusion denuder removes all gaseous HNO and passes all particles;
(2) The combination of particle filter and HNO collection tube collects
all gaseous HNO and particulate nitrate; and
(3) The two assemblies sample equal amounts of air.
DISCUSSION
The results of the DDE are not affected by artifacts due to loss of
particulate nitrate as HNO from the filter or gain of particulate nitrate
from reactions of HNO- on the filter. In fact, the DDE results may be used to
determine the amount of nitrate loss or gain due to artifacts. From the
definitions above, we see that T is equal to the amount of nitrate lost as
HNO from the filter, since the diffusion denuder prevents gaseous HNO from
,j -J
entering the filter-collector assembly. On the other hand, particulate ni-
trate gain from HNO is given by the difference F - F . Since 'nitrate forma-
tion from HNO is negligible for Teflon filters (Spicer et al., 1978), any
measurable gain would presumably be due to gaseous conversion on the collected
aerosol.
In this discussion it has been assumed that amounts of gaseous nitrates,
other than HNO , are negligible. If amounts of other gaseous nitrates are
significant, their effect on the DDE can be determined by considering the
event matrix:
81
-------
Collector Efficiency Denuder Efficiency
100% 0%
100%
0%
T = 0
r^
T = X
T = T = 0
— V
D
T = X
T = T = 0
Here X is the amount of gaseous nitrate (not HNO ) that enters the system; T
and T have been previously defined. For simplicity, we assume that no
particulate nitrates are present. If the diffusion denuder removes the gaseous
nitrate and the tubes are capable of collecting it, the amount of gaseous
HNO determined will be too high by the amount X. This is demonstrated by
the relation
N == (F + T) ™" (F + T )
In this case (as represented by the upper left matrix element)
N— T _ rp == Y
~~ i j. ^\
In all other cases, the values of N are not affected. If the denuder
H
does not remove the gaseous nitrate and the tube collects it (as represented
by the upper right matrix element), the value N will be too high by the
amount X. For all other cases, the values of N are not affected.
As mentioned above, loss of particulate nitrate may occur by evaporation
from the collecting filter. Significant losses of NH.NO have been observed
in the laboratory in an extreme experiment — loading a filter with pure
NH NO aerosol and subsequently passing clean air through it (D. Reutter,
private communication). Because of their much lower vapor pressures, metal
nitrates are expected to show much lower evaporative losses than NH.NO,.
82
-------
Losses of NH NO by evaporation from urban aerosols may be suppressed
by: (1) presence of gaseous NH and HNO' in the atmosphere in equilibrium
with the particles, and (2) reduction of vapor pressure of the particles due
to the presence of other materials. If evaporative loss from the filter occurs,
however, we expect the NH NO to decompose according to
NH NO -> NH + HNO
Because the HNO collection tubes follow the particle filters, DDE results are
not affected by evaporative loss. If more than one artifact process becomes
significant, however, interpretation of the DDE may become complicated.
To summarize, the DDE provides the following measures of atmospheric
nitrates:
N,
H
= gaseous HNO
N = particulate nitrate
N = total atmospheric nitrate
T = artifact due to loss of particulate nitrate by reaction
on the filter
F - F = artifact due to gain of particulate by transformation of
gaseous nitrate
ACKNOWLEDGMENTS
We thank Dr. Charles Lewis for his many helpful criticisms and suggestions.
REFERENCES
Harker, A., L. Richards, and W. Clark. 1977. Effect of Atmospheric SO
Photochemistry Upon Observed Nitrate Concentrations. Atmos. Environ.
11:87-91.
83
-------
Spicer, C. W., P. M. Schumacher, J. A. Kovyovmjian, and D. W. Joseph. 1978.
Sampling and Analytical Methodology for Atmospheric Particulate Nitrates.
Battelle Memorial Institute: Final Report. EPA 600/2-78-067, U.S.
Environmental Protection Agency.
Stevens, R., T. Dzubay, R. Burton, G. Russwurm, and E. Tew. 1978a. Compari-
son of Hi-Vol and Dichotomous Sampler Results on Nitrates and Sulfates.
Preprint No. 98 for meeting of the Division of Environmental Chemistry,
American Chemical Society, September.
Stevens, R., T. Dzubay, G. Russwurm, and D. Rickel. 1978b. Sampling and
Analysis of Atmospheric Sulfates and Related Species. Atmos. Environ.
12:55-68.
84
-------
TECHNICAL REPORT DATA
(Please read Instru^ rions on the reverse before completing)
1. REPORT NO. 2. 3. RECIF
EPA-600/2-79-051
4. TITLE AND SUBTITLE 5. REPO
CURRENT METHODS TO MFASURF ATMfKPHFRTT NTTRTP ^ar
ACID AND NITRATE ARTIFACTS
7. AUTHOR(S) 8. PERF
R.K. Stevens, Editor
"» Tl \/ 1 l^fl luTlM Fi T3 n ' "^ii i^P Mi^O € f?O C O9 Y*f* R 1 «3 it ft V*3 1 ft 1O\/ 1 Kl t Q
Liiviiwiiiiidiwcii ^i* i c i I\H»C o rxcwCCii^ii LdUUidisVJiy ^ii QO
Offirp nf Rp^e^fch and Development
U.S. 'Environmental Protection"^ Agency H.CON
Research Triangle Park, N.C. 27711
*9 SP£HUSDaU\LQ.AQEr»C.Yi NAME ANkO_ADQFlESfi -^ 13 TYP
trnvrrunmentai StienCer KesreaTcTl Laboratory — RTP, N.C. in-n
O'f'f'ipc* nf Rpcpa v*ph a nH Hox/ol Anmont1 ' _ _
U.S. Environmental Protection Agency EPA/
Research Triangle Park, N.C. 27711
15. SUPPLEMENTARY NOTES
IE NT'S ACCESSION-NO.
RT DATE
ch 1979
ORMING ORGANIZATION CODE
ORMING ORGANIZATION REPORT NO.
8fiMEli^rPr-79)
TRACT/GRANT NO.
= OF REPORT AND PERIOD COVERED
JU 5c
^SORING AGENCY CODE
500/09
16. ABSTRACT
Presentations given at a workshop on "Measurement of Atmospheric Nitrates"
(Southern Pines, N.C. October 3-4, 1978) are documented. The authors consider
various analytical methods to measure ambient concentrations of nitric acid and
artifact nitrate formation.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS b. IDENTIFIERS/OPEN ENDE
*Air pollution Artifact nitrate
*Nitric acid
"Inorganic nitrates
"Chemical analysis
Meeti ngs
18. DISTRIBUTION STATEMENT 19. SECjy.Ri'LY^CLASSL/TTlij J
ppi PAQF Tn PIIRI TP
KtLtAit IU 1 UHL1L 20. SECURITY CLASS r77,«,
UNCLASSIFIED
D TERMS c. COSATI Field/Group
5 13B
07B
07D
05B
Report) 21. NO. OF PAGES
99
lage) 22. PRICE
EPA Form 2220-1 (9-73)
85
------- |