EPA-600/2-78-019
February 1978
Environmental Protection Technology Series
SPECTRAL MEASUREMENTS
OF GASEOUS SULFURIC ACID
USING TUNABLE DIODE LASERS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development. U S. Environmental
Protection Agency, have been grouped into 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-600/2-78-019
February 1978
SPECTRAL MEASUREMENTS OF GASEOUS SULFURIC ACID
USING TUNABLE DIODE LASERS
by
Richard S. Ehg, Kenneth W. Nill and Jack F. Butler
Laser Analytics, Inc.
Lexington, Massachusetts 02173
Contract No. 68-02-2482
Project Officer
Roosevelt Rollins
Emission Measurement and Characterization Division
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 Research
Science Laboratory, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation
for use.
11
-------
ABSTRACT
Using a tunable diode laser spectrometer with a spectral
resolution of about 10 cm~l, the important central portions of
the two infrared absorption bands of H2S04 at 8.2 urn (1222 cm"1)
and 11.3 ym (880 cm"1) have been scanned at low pressure (-0.67
Torr of H^SC^) and at atmospheric nitrogen pressure. Maximum
absorption coefficients have been measured to be 6.5 cm~1/atm and
6.9 cnrVatm at the 8.2 ym (1222 cm'1) and 11.3 vim (880 cm"1)
bands respectively. Interference spectra of 502, ^^2 an£ ^0
near the H^SC^ absorption peaks at 1222 cm"1 and 880 cm"1 were
scanned using a 1.1 m cell at 200°C to determine interference-
free regions. A spectroscopic method was used to measure the
partial pressures of HjSO^, SO^ and 1^0 vapors above azeotropes
of H2S04 at 107°C, 150*C and 200°C. The expected performance
characteristics of an J^SO^ tunable diode laser stack monitor
are considered on the basis of the above results.
iii
-------
CONTENTS
Abstract ill
Figures vi
Tables viii
1. Introduction 1
2. Summary 4
3 . Conclusions 6
4. Recommendations 8
5. Experimental Apparatus, Methods and Results 11
High resolution absorption measurements
of low pressure ^SC^ vapor 10
High resolution absorption measurements of
HoS04 vapor in atmospheric pressure of
nitrogen 23
High resolution absorption measurements of
interferants - H20, SC-2 and C02 32
Spectroscopic determination of partial vapor
pressures above hot azeotropic H2SO^
solution 42
-------
FIGURES
Number Page
1 Low-pressure H2S04 absorption measurement setup 12
2 Low-pressure H2SO. absorption cell 14
3 Absorption spectra above hot H2S04 solution
(T~165°C) 16
4 H2S04 absorption near 1223 cm"1 17
5 Low-pressure absorption H2S04 absorption near
1224 cm'1 19
6 Low-pressure H2SC>4 absorption near 880.5 cm" 20
7 Strip chart recording of low pressure I^SO.
absorption vs. time 21
8 Low pressure H2S04 absorption (P-0.67 Torr
T=170°C) ; 22
9 Low pressure H2SO.. absorption (P~0.67 Torr
T=170°C) 24
10 Experimental setup for H2S04 absorption at atmos-
pheric N2 pressure 25
11 Absorption cell for ^SO* absorption at atmospheric
nitrogen pressure and H2S04 vapor generator 26
12 H2SC>4 absorption at atmospheric nitrogen pressure.... 28
13 H2SO4 absorption at atmospheric nitrogen pressure.... 29
14 H2SC>4 absorption at atmospheric nitrogen pressure.... 30
15 H2SC>4 absorption at atmospheric nitrogen pressure.... 31
16 Location of the 2 cm"1 atmospheric scan region with
respect to the 1210-1240 citT1 low-pressure scan.... 33
VI
-------
Number Page
17 H2S04 absorption at atmospheric nitrogen pressure.... 34
18 H2S04 absorption at atmospheric nitrogen pressure.... 35
19 H2SC>4 absorption at atmospheric nitrogen pressure.... 36
20 H2S04 absorption at atmospheric nitrogen pressure.... 37
21 H2S04 absorption at atmospheric nitrogen pressure.... 38
22 Location of the 2 cm"1 atmospheric scan region with
respect to the 870-895 cm"1 low pressure scan 39
23 S02 absorption near 1222 cm 40
24 S02 absorption 41
25 H20 absorption 43
26 H20 absorption 44
27 C02 absorption 45
28 H20 absorption 46
29 Absorbance per unit pressure (ocL/Torr) vs. tempera-
ture of cell body for the EUO line at 1330 cm"1
(L=64 cm) 48
30 H20 line at 1330 cm'1 above a 150°C H2S04 azeotrope
L=64 cm, Tbody=170«C ? 49
31 Absorbance per unit pressure (otL/Torr) vs. tempera-
ture of cell body for the SOo line near 1354 cm"1
(L=64 cm) 50
32 803 absorption lines near 1354 cm"1 51
33 Total pressure measurement apparatus 54
34 Dissociation constant Kp of H2S04 gas vs.
temperature 58
35 Partial pressure of H2S04, H20 and 803 above azeo-
tropic H2S04 solution vs. temperature 59
36 Total vapor pressure above H2S04 solution vs. azeo-
tropic composition 62
vn
-------
TABLES
Number Page
1 H20 Partial Pressure vs. Temperature of
H2S04 Azeotrope .................................. 52
2 863 Partial Pressure vs. Temperature of H2S04
Azeotrope ........................................ 52
3 Total Pressure vs. Temperature of
Azeotrope ........................................ 55
4 H2S04 Partial vs. Temperature of ^804 Azeotrope... 55
5 Azeotropic Composition vs. Temperature of
Azeotrope .................................. 57
Calculated Partial Pressures Including the
Effects of Dissociation .......................... 61
Vlll
-------
SECTION 1
INTRODUCTION
The central purpose of this program was to determine the
high resolution spectral properties of HoSO^ and of possible in-
ter ferants H20, S02, S03 and C02« This information is needed to
evaluate the potential sensitivity of and design a sensitive
stack gas monitor based on tunable laser techniques.
There is considerable environmental interest in sulfuric
acid because of its damaging effect on the ecologic system and
strongly suspected adverse health effects on humans (1). The
presence of sulfuric acid in air is mainly due to the burning of
sulfur-bearing fuels, particularly coal, which is projected to
become a widely-used, major fuel in the coming years. While the
primary combustion product is fulfur dioxide, 1 to 10% may be
emitted as trioxide, which reacts rapidly with moisture in the
atmosphere to produce H2S04. Furthermore, S02 may be oxidized
to form H2S04, and this is an efficient process in the presence
of catalysts such as platinum, vanadium pentoxide and oxides of
nitrogen (most commercial methods of producing H2S04 rely on the
catalytic oxidation of 802).
Monitoring H2S04 by spectroscopic techniques require that
the relevant parameters of H2S04 be known. A number of investi-
gators (2-4) have published low-resolution, qualitative band
spectra of 1^504 vapor in the infrared spectral region.
1. West, P. W., A. D. Shendrikar and N. Herrara. Analytica
Chimica Acta 69, 111 (1974).
2. Stopperka, K. "The Infrared Spectrum of the Water-free H2S04
and the Composition of the f^O-I^SO^ System." Zeit fur
Anorgan und Alleg. Chem. 344, 263-278 (1963).
3. Chackalackal, S. M. and F. E. Stafford. "Infrared Spectra of
the Vapors above Sulfuric and Deuteriosulfuric Acids." Journ.
Amer. Chem. Soc., 88:4, 723-728, Feb. 20, 1966.
4. Stopperka, K. and F. Kilz. "Die Zusammersetzung der
Gasphase viber dem fliissigen System H20-H2S04 in Abhangigkeit
von der Temperatur." Z. anorg. u allgem. Chem., Band 370,
49-66, 1969.
-------
According to these published works, there are four infrared bands
of H2SO/ centering at 2.8 ym, 6.9 ym, 8.2 ym and 11.3 ym. Be-
cause of very strong interference by H20 for the first two bands,
only the last two bands are worth considering for monitoring pur-
poses.
Besides the low resolution absorption data mentioned above,
high resolution spectra of H2S04 vapor from 1312 to 1549 cm"1
have been published by Alpert (5). Since they lie in the 6.5 ym
band of H20, they are of little use for monitoring purposes.
However, the author did identify many individual lines as being
H2SO^. On this basis, it might be expected that there are simi-
lar individual HoSO^ absorption lines in the 8.2 ym and 11.3 ym
bands which would be suitable for monitoring purposes.
_i
Most recently, quantitative low resolution (~1 cm ) spectra
of the 1220 and 880 cm"1 bands of EjSO^ have been obtained inde-
pendently by Burch and co-workers (5) at Aeronutronic Ford and
R. Majkowski (7) at the General Motors Research Laboratory. The
maximum absorption strength observed with low resolution in these
bands is approximately 2 to 5 citT1 atm"1 (.0002 to 0.0005 m"1/
ppm). While these absorption strengths appear to be adequate for
infrared monitoring H2SO^ in stack gas, the lack of spectral
structure (e.g., isolated lines) at low resolution seriously
complicates the monitoring task. Strong interferences further
complicate the monitoring since there are few, if any, 1 cm'1
spectral intervals in the H2S04 bands free of absorption due to
H2O, CO2 and SO2. High resolution (<0.01 cm"1) spectra are re-
quired to identify the strongest H2S04 absorption lines and re-
veal any absorption features which would be useful in the design
of a monitoring instrument. High resolution spectra of C02, S02
and H20 at elevated temperatures are also needed to determine the
severity of the interference problem for a low resolution
(1-5 cm~l) monitoring instrument and to locate spectral windows
free of interference for a laser monitoring instrument.
The primary difficulty in carrying out high resolution spec-
troscopy of hot H2SO^ is in the area of gas cell design and gas
handling. Reading of the above mentioned EPA report (6) and the
5. Alpert, B. D. "Studies in High Resolution Infrared Spectros-
copy." Ph.D. Dissertation. (Appendix B contains H2S04 data.)
Ohio State University, Columbus, Ohio, 1970.
6. Burch, D. E., F. J. Gates and N. Potter. "Infrared Absorption
by Sulfuric Acid Vapor." FINAL REPORT NO. EPA-600/2-76-191,
July 1976.
7. Majkowski, R. F. "Infrared Absorption Coefficient of H2S04
Vapor from 1190 to 1260 cm"1." J. Opt. Soc. Am. 67, 624,
1977.
-------
recently published journal article (7) indicate that there have
been many problems encountered in fabricating and using suitable
infrared gas cells. Neither Aeronutronic Ford nor General Motors
has obtained high resolution spectra of the pure azeotropic 112804
at a known (measured) temperature and pressure. All of General
Motors' spectra were obtained in a flowing system at atmospheric
pressure using N2 as a carrier gas. The temperature of the H2S04
vapor studied by Aeronutronic could only be inferred from pres-
sure measurements. This report describes an approach to cell de-
sign to obtain the desired spectral absorption data at the high-
est resolution available with tunable diode lasers.
In this report, the experiments and the results of high re-
solution scans of HoSC^ vapor at low pressure and at atmospheric
nitrogen pressure will be presented for the spectral regions cen-
tering at 8.2 ym and 11.3 ym. In addition, high resolution scans
of H20, CC>2 and S02 gases conducted to determine the lowest in-
terference regions for H2S04 vapor monitoring will be discussed.
In order to monitor the concentration of H2S04 vapor in a
smoke stack gas, which has a temperature range from about 150°C
to 250 °C, it is necessary to determine the absorption coeffi-
cients of H2SO/ vapor at this temperature range. Because of the
dissociation of P^SO^ into 1^0 and 803 (8) , partial pressures are
known only for solutions with rather low H2S04 concentration and
for the atmospheric azeotrope (9) . The solutions with low I^SC^
concentration are not azeotropes and are therefore difficult to
be reproduced reliably in the laboratory for absorption measure-
ments. The partial pressures at different temperatures of an
atmospheric H2SC>4 azeotrope have been measured (9) . However,
since the vapor phase and liquid phase are not in equilibrium at
any temperature other than 326°C, the partial pressures will de-
pend on the temperature and volume of the vessel or container.
Based on the above reasoning, it is best to work with azeotropes
in the stack temperature range of 150 to 250°C for reliable re-
producibility.
Tunable diode lasers offer a convenient method of measuring
the partial pressure of H20 and S03 vapors directly. The partial
pressure of H2S04 vapor can then be determined by subtracting
the SO-j and f^O partial pressures from the total vapor pressure.
In this report, a spectroscopic method and results of partial
pressure determination will be presented.
8. Gmitro, J. I. and Vermeulen. "Vapor-Liquid Equilibria for
Aqueous Sulfuric Acid." A. I. Ch. E. Journal 10, 740 (1964).
9. Luchinskii, G. P., Zhur. F12 Khim, 30, 1207 (1956).
3
-------
SECTION 2
SUMMARY
-4 -1
Using the highest spectral resolution (-10 cm ) presently
available with tunable diode lasers, important central portion of
the H2S04 bands at 8.2 ym (1222 cm"1) and 11.3 ym (880 cm"1) have
been scanned at low pressure (~0.67 Torr of E^SO*) and at atmos-
pheric pressure with nitrogen as the main gas. Because of the
extremely corrosive nature of hot H2S04 and SG>3/ cell design and
construction were carefully carried out. The result was that
there were no visual or spectrally-detectable contamination in
the cell after many hours of cell operation near 200°C. Due to
the complex thermodynamic nature of the vapors above a hot I^SO,
solution, only azeotropes of H2S04 solution prepared from stan-
dard solution by a distillation method were used throughout in
the experiments.
In a low pressure scan of a sealed-off H2SC>4 cell from 1210-
1240 cm"1, two peaks were observed. They are located near
1222 cm"1 and 1235 cm"1, the former being sharper and stronger
than the latter. Similar low-pressure scan in the range 870-
895 cm"1 shows only one peak at 880 cm"1. The absorption coeffi-
cients are found to be 7.2 and 7.0 cm'-^/atm at 1222 cm"1 and
880 cm"1 respectively. Besides these rather broad (>2 cm"1)
spectral peaks, there were no sharp individual I^SO^ lines
throughout nearly the whole of the two spectral regions.
Using a flowing-gas cell with dry nitrogen as the carrier
gas, atmospheric scans were performed at liquid H2S04 reservoir
temperatures of 180°C and 200°C over 2 cm"1 intervals near
1222 cm"1 and 880 cm"1. Absorption coefficients of 6.5 cm'-Vatm
(6.5 x 10"4 ppm'i-m"1) and 6.9 cnrVatm (6.9 x 10~4 ppm-i-m"1)
have been obtained at 1222 cm~^ and 880 cm"1 respectively.
Using a 1.1 m cell operated at 200°C, interference spectra
due to SC>2, H20 and C02 were scanned over the same two cm"1 in-
tervals. In the 1222 cm"1 peak region, the SOj spectra exhibit
a very high line density. The H^O absorption is much stronger at
1222 cm"1 than at 880 cm"1. In the 880 cm"1 region, C02 absorp-
tion at 100 Torr pressure was not detectable at a cell length of
l.lm.
Spectra of 1^0 and 803 and SC>2 were all easily observed with
a 50 cm low pressure cell, when the liquid H2S04 reservoir was
heated to about 100°C or higher.
-------
The partial pressures of 803 and H20 above H2S04 azeotropes
at several temperatures (107°C, 150°C and 200°C) were determined
spectroscopically. This was done by measuring the absorption
strengths of t^O and 803 lines and comparing them against cali-
brated absorption strengths, the calibrated strengths being de-
termined using either 1^0 or 803 alone in the absorption cell.
A "u" tube manometer in which the fluid in the "U" tube was an
H?S04 azeotrope was used to measure the total vapor pressure.
The partial pressure of H2S04 vapor was then given by the dif-
ference between the total pressure and the sum of the 803 and
H20 partial pressures. The 112864 partial pressures above azeo-
tropes at different temperatures obtained in the present experi-
ment are about 30 percent smaller than those reported by
Luchinskii (9) for the atmospheric azeotrope at the correspond-
ing temperatures. The present results were used to compute the
absorption coefficient of 112804 vapor at different temperatures.
-------
SECTION 3
CONCLUSIONS
At low pressure, the H2S04 vapor spectrum has a dense, fine
structure near the 8.2 ym band center, -1223 cm""-'-. The width of
this fine structural band is only about 0.5 cm~^. Since the ob-
served line spacing is about 0.01 cnT1, it was expected and ex-
perimentally confirmed that this fine structure would be smoothed
out at atmospheric pressure. For stack monitoring of I^SO^ vapor,
it is necessary to use the broad smooth peak (>2 cnT1) observed
at atmospheric pressure. The maximum percentage change in ab-
sorption strengths occurs on the side of a peak at 1222 cm~l. It
has a slope of about 35%/cm~l at atmospheric pressure, whereas
the absorption signal at the peak is about 1% for a 15 ppm-m
1^504 concentration. The detection of this level of absorption
change can be met using state-of-the-art tunable diode laser
techniques. The partial pressure of H7S04 obtained in the pres-
ent experiment is about 30 percent smaller than those reported
by Luchinskii (9). As a result, the detection limit for H2S04
vapor is correspondingly lower, ~2 ppm-m. The intensity of
individual 803 lines have been measured, for the first time, as
a function of temperature for the V3 band centering at 1391 cm~l.
It appears that the present technique of partial pressure mea-
surements can be extended to include study on the thermodynamics
of 2S03 t 2S02 + 02. This last catalytic reaction has been known
to exist both in stack gas and in automobile exhaust.
Based on the present experimental result, a monitoring in-
strument utilizing the 880 cm~l absorption peak should be a dual
wavelength absorption type. One wavelength would be fixed near
880 cm~l, where there is strong H2S04 absorption coupled with
low interference from H^O and C02« The other wavelength would
be at a region about 1000 cm~^-, where there is very little ab-
sorption by H2S04, S02, CO? or H20. Subtracting the absorptions
measured at these two wavelengths will produce a signal propor-
tional to the H2S04 concentration. The two wavelength technique
has been used in long-path monitoring of atmospheric pollu-
tants (10). Comparison with this earlier work indicates that
4 ppm-m of H2S04 can be measured, neglecting the difference in
10. Menzies, R. T. and M. S. Shumate. "Remote Measurements of
Ambient Air Pollutants with a Bistatic Laser System."
Appl. Opt. 15, 2080, 1976.
-------
pathlength and environment. This compares with an estimated
interference-limited sensitivity of about 2 ppm-m. Only the
known interferants, H20 and CO-, have been considered.
The principal advantage of using a tunable, narrow-line
source is the ability to select frequencies that lie in clear
spaces within the densely spaced interference lines of H20 and
hot C02. With a conventional infrared source, such as a Nernst
glower filtered by a grating spectrometer, energy consideration
dictates a spectral slitwidth of about 1-2 cm"1, which is wider
than interference line spacings. . Thus, a field instrument using
conventional spectrometer will fail to maintain accurate cali-
bration due to fluctuations in temperature and concentrations
of C02 and H20.
-------
SECTION 4
RECOMMENDATIONS
Based on the experimental results on atmospheric H2S04 ab-
sorption measurements and on S02/ H20 and C02 interference mea-
surements, the ultimate sensitivity of a monitoring instrument
is likely to be limited by interference. This is especially
true for the 1222 cm'1 region where the S02 lines are very dense.
Therefore, it is recommended that a laser spectrometer be used
for H2S04 vapor detection. The main advantage of this type of
instrument is that it is least subject to interference, since
the wavelength of operation can be tuned to a region away from
absorption lines.
A dual wavelength absorption method using two diode lasers
appears to be the best approach at the present. One laser should
be tuned to the absorption peak near 880 cm"1 and the other near
1000 cm"1, where H2S04 absorption is low. An absorption cali-
bration can be made by inserting a short cell containing H2S04
vapor in atmospheric nitrogen into the beam path. The cell de-
sign should use a permanent, stable, sealed-off feature similar
to that in the present study. For the highest stability in laser
frequency, the lasers would be frequency locked to absorption
line center. The system should use a retroreflector to increase
the pathlength. High frequency modulation techniques (>10 kHz)
should be employed to minimize turbulence effects. The results
of this study indicate that the existence of interferences will
severely limit the performance of a conventional IR instrument
for monitoring H2S04 in stacks, but that sufficient sensitivity
can be achieved with a tunable diode laser instrument.
It has been shown that the partial vapor pressures of H20,
803 and H2SO4 can be determined spectroscopically. The degree
of dissociation indicated by the equation H2S04 -»• H20 + 503 can
also be measured spectroscopically by monitoring the 803 and H20
absorptions for different cell wall temperatures at a given
H2S04 reservoir temperature.
The SO^ molecule is also of interest. The present high
resolution laser spectrometer should be useful in measuring all
the 803 lines in the V3 band. Such spectroscopic parameters as
line strength and pressure broadening can easily be determined.
The result could be helpful in further understanding catalytic
reactions such as 2S02 + O2 ^ 2303.
-------
It appears that the fine absorption structure near 1223 cm"-'-
might be that of gaseous H^SO^. It would seem worthwhile to
further pursue similar high resolution measurements at this and
other regions using a longer cell. Increasing the cell path-
length to 10 meters would permit an operating temperature reduc-
tion to 80°C or below. At this lower temperature, pressure
broadening is probably negligible and fewer hot vibrational
bands are present.
In future experiments with HoSO^ vapor, the source of S02 in
a low-pressure cell, as detected in the present experiment,
should be determined. If it is an impurity in ^SO^ it should
be purged with helium bubbling through the hot H2S04 liquid.
Since the presence of S02 can be monitored spectroscopically, the
effect of purging can be observed.
Finally, the "U" tube manometer used for the total pressure
measurement has been proven to be a rather simple instrument to
use. It is recommended that it be used in a system where only
moderately precise pressure measurements are required. An all
quartz spiral gauge should be employed for more precise pressure
measurements.
-------
SECTION 5
EXPERIMENTAL APPARATUS, METHODS AND RESULTS
In this section of the report, we shall first outline the
specific tasks required. For each task, we will discuss the ap-
paratus and methods used and the experimental results. The fol-
lowing is a list of required tasks:
High Resolution Absorption Measurements of Low Pressure
H2S04 Vapor. Spectral measurement is to be performed with
a suitable spectrometer system at the azeotropic point of
the H2S04-H20 vapor liquid phase diagram. Spectral scans
throughout the 870-895 cm"1 and 1210-1240 cm'1 regions are
required.
High Resolution Absorption Measurements of H2S04 Vapor in
Atmospheric Pressure of Nitrogen. Upon identification of
prominent H2S04 absorption lines in both spectral regions,
atmospheric H2S04 absorption measurements are to be per-
formed with continuous coverage of at least 2 cm""1 and
centered on the most prominent peak in each region.
High resolution Absorption Measurements of Interferants -
H20, S02 and CO2- Absorption spectra of H20, C02 and S02
shall be obtained in the identical spectral regions and
under similar experimental conditions.
Spectroscopic Determination of Partial Vapor Pressures above
Hot Azeotropic H2S04 Solution. The partial pressures of
H20, S03 and H2S04 vapors are to be determined spectroscopi-
cally for azeotropes of H2S04 at 100, 150, 200 and 250°C.
The above measurements shall be performed with a resolution equal
to or better than 0.05 cnr1 and all spectral line positions shall
be referenced with respect to known spectral lines.
HIGH RESOLUTION ABSORPTION MEASUREMENTS OF LOW PRESSURE H2SO4
VAPOR
Experimental Apparatus and Methods
Figure 1 is a schematic diagram of the experimental setup
which was employed to perform the low-pressure H2S04 absorption
measurements from 1210-1240 cm"1 and from 890-895 cm"1.
10
-------
A diode laser of either Pbg. 989Sn0.011Se (122° cm band) or
PbO. 955SnO. Q45Se (88° cm"1 band) 'was mounted on a vibrationally-
isoiated cold finger of a closed-cycle cryogenic cooler. The
required fractional value of tin depends on the composition-to-
wavelength tuning characteristics of the material. The tempera-
ture of the cold finger was accurately controlled to ±0.0003°K
anywhere in the range from 12°K to 80°K by the temperature con-
troller. In operation, the laser frequency was tuned grossly by
changing the cold finger temperature and finely by changing the
laser diode injection current, supplied by the diode current
power supply. The cryogenic cooler, the temperature controller
and the diode current power supply are major components of a
laser spectrometer system (Laser Analytics, Inc. Model LS-3) .
As shown in the diagram, the laser radiation is collected by a
2.5 cm f.l., f/1 lens and is steered into a 25 cm long pyrex ab-
sorption cell. The far end of the pyrex cell was gold coated for
efficient spectral reflection. After traversing the cell twice,
the beam is directed into a grating monochromator for spectral
mode filtering. The transmitted beam is then detected, syn-
chronously amplified and is displayed on an x-y recorder. The
drive for the x-axis of the x-y recorder is proportional to the
diode current. For absolute frequency calibration a reference
gas cell was used. For relative frequency calibration, a fixed
Ge etalon with a free spectral range of 0.04844 cm"1 at 1222 cm"1
(0.04856 cm"1 at 880 cm"1) was used.
The procedure used in absorption measurement is as follows:
1) perform a background scan with no I^SOg vapor present in
the cell, I^SO^ reservoir at 60°C or lower
2) perform a reference frequency scan with a reference gas
cell in the beam path
3) perform an etalon scan
4) perform an H2S04 absorption scan with hot H2S04 vapor
in the cell, H2S04 reservoir at 100°C or higher.
The cell wall temperature is to be maintained at a higher
(10 to 15°C) temperature than that of the reservoir for all of
the above steps.
The design and construction of the H^SCK absorption cell is
of the utmost importance in the success of tne experiment, since
hot H2SC>4 acid is corrosive to many vacuum and optical material
coming in contact with it.
For the low-pressure ^SO^ cell, pyrex glass tubing was
used for the cell body. The inside diameter of the tubing was
28 mm. One end of the cell has a gold-coated flat pyrex mirror.
11
-------
TEMPERATURE
CONTROLLER
COOLER
LASER
POWER
SUPPLY
X-Y
RECORDER
•*-
Y
GRATING
SPECTROMETER
ABSORPTION CELL
LOCK-IN
AMPLIFIER
DETECTOR
Figure 1. Low-pressure H-SO, absorption measurement setup
-------
The other end has a 1 mm thick, optical-grade silicon window with
a 2° widge. The cell length from the window to the gold mirror
is 25 cm. A tight vacuum seal was successfully made between the
Si window and the pyrex glass tube (pyres No» 332) by shrinking
the I.D. of the latter on the edge of the silicon window at high
temperature (~800°C). The seal could apparently withstand re-
peated cycling between 20°C and 275°C without deterioration as
witnessed at the end of the low-pressure absorption measurements
that the absorption linewidth of SO? (in minute presence in the
cell possibly as an impurity in I^SO^) remained quite narrow,,
This is a condition indicating a low background pressure in the
cell.
To conveniently control the vapor pressure of 112804 in the
low-pressure cell, a pyrex side arm was incorporated into the
cell. Figure 2 is a picture of the low-pressure cell with liquid
H2SC>4 in the side arm. In operation, the cell body was typically
at about 200°C while the side arm was at a lower temperature.
The side arm temperature was controlled independently by sur-
rounding the side arm with a heatable copper tube. A thermo-
couple attached to the tip of the side arm tube sensed the liquid
H2S04 temperature and was part of a feedback temperature control
system. In a background scan, the heater to the copper tube was
turned off and the side arm temperature was cooled down to below
60°C. For an absorption scan, the heater was automatically
turned off and on to maintain a preset temperature (-170° ± 1°C)
on the liquid I^SO*.
The cell was baked out at 250°C and back filled with dry
nitrogen. Then a few cm3 of a 37N H2S04 solution (98,,5 to 99.3%
H2S04 concentration) was placed on the side arm reservoir. The
azeotrope was prepared by evacuating about a third of the origi-
nal H2S04 acid in the reservoir at 160°C. Here we followed con-
servatively the recommendation by Kunzler (11) who obtained an
atmospheric azeotrope by boiling off 20 percent of an original
91.9 wt.% H2S04 solution. Since the HoSC^ partial pressure near
the azeotropic composition is insensitive to a change in the
azeotropic composition (8), a small change, if any, in the azeo-
tropic composition does not affect the ^SO. absorption coeffi-
cient measurement result. The low pressure cell was permanently
sealed off for absorption measurements.
Experimental Results
Since the vapors above azeotropic solution of H2SC>4 are I^O,
SC>3 and HjSO^, it is desirable to identify their presence. Scan-
ning the low-pressure cell at high resolution is one way of
searching for them. For strong HO lines, the positions are
11. Kunzler, J. E. Anal. Chem. 25, 93 (1953)
13
-------
II 1
II •'
I1 '
1 1
1 1
I' 1
. 1 i
SILICON
WINDOW
1 mm THICK
2° WEDGE
L A
GOLD
MIRROR
TOP VIEW
25 cm
28 mm i,D, PYREX
H2S04
RESERVOIR
FRONT VIEW
Figure 2. Low-pressure H-SO, absorption cell
-------
known to better than 0.05 cm"1. For 803, from the work of
Lovejoy (12), it is known that there is a strong absorption band
centered at 1391 cm""1., Using the experimental setup shown in
Fig. 1 and a diode laser tuned to the 1400 cm""! region, a spec-
trum of 803 lines and one 1^0 line was obtained. Figure 3 shows
a scan near the 1416.13 cm"-*- 1^0 line. These data were taken
with the H2S04 reservoir near 165°C and the cell wall temperature
about 180°C. Since an atmospheric path of about 2.2 m was in
series with the cell, absorption by atmospheric H^O vapor shows
up as a broad dip. The 1^0 vapor in the low-pressure cell shows
up as a narrow line located at the center of the atmospheric
water line. The rest of the lines are all SOg„ The S02 band
centered about 1375 cm"-*- does not extend far in this region even
at high temperature. A 30 cm cell filled with 10 Torr of S02 at
200°C shows no absorption lines when scanned in this region.
Further evidence that the observed lines in Fig. 3 are S03 is
that the width of the H20 line (in an expanded scale) is greater
than that of the 803 lines. At low pressure, the widths are due
to Doppler broadening. The measured linewidths are 0.0055 cm"1
for the H20 line and 0.0028 cm"1 for SO, lines. The calculated
linewidths are 0.005 cm"1 and 0.0024 cm !. The measured width
of 803 appears slightly too great in comparison with the calcu-
lated value. This is perhaps due to pressure broadening. These
are spectroscopic evidences that the vapors above hot H2S04 solu-
tion did contain H20 and 803 and they are detectable by tunable
diode laser spectroscopy.
After H2O and SO3 absorption lines were observed, the laser
was tuned to the 8.2 ym region in searching for I^SO^ absorption
lines. Figure 4 is a recorder trace of H2SC>4 vapor absorption
near 1223 cm"1. The cell body was operated at about 200°C while
the liquid H2SO4 reservoir was at 170°C and 155°C for the two
absorption scans. The reference lines are ^0 lines. The cali-
bration trace was obtained with a 16 cm cell filled with 10 Torr
of N20 gas. The frequency of one of the N20 line was identified
to be 1223.36 cm"1 (13).
The dotted trace near the bottom is the etalon scan with a
fringe spacing 0.0484 cm"1. By counting the number of fringes,
the width of this scan was calculated to be about 0.6 cm"1. This
width is typical for diode lasers in this spectral region. As
can be seen, the absorption traces show that there are fine
12. Lovejoy, R. W., J. H. Colwell, D. R» Eggers and G. D. Halsey.
"Infrared Spectrum of Gaseous Sulfur Trioxide." J. Chem.
Phys., 36, 612 (1962).
13. McClatchey, R. A., W. S. Benedict, S. A. Clough, D. E. Burch,
R. F. Calfee, K. Fox, L. S. Rothman and J. S. Caring. "AFCRL
Atmospheric Absorption Line parameters Compilation." AFCRL-
TR-73-0096, 1973.
15
-------
6
tu
o
<
LU
a:
1416.13 cm
H20 LINE
-1
FREQUENCY (cm"1)-^
Figure 3. Absorption spectra above hot I^SO. solution (T~165°C)
-------
FREQUENCY (cm'1)
Figure 4. H SO. absorption near 1223 can
-1
-------
structure absorption as well as smooth absorption. The smooth
part of the total absorption shows a large variation with tem-
perature. Except in this spectral region, there was no fine
structure observed anywhere else in the 1222 cm"1 and 880 cm'1
bands.
In the background scan with the cell body at about 60°C or
lower, several narrow lines of SC>2 are present. The SC>2 is be-
lieved to be present in H2SC>4 as an impurity. Apparently evacu-
ating about one-third of HoS04 solution was not sufficient to
completely remove the S02 from the I^SO^ solution. The amount
of S02 present was estimated to be less than 0.1 Torr, so that
its presence did not affect the absorption measurement accuracy.
Since each recorder scan is between 0.5 cm'1 and 0.75 cm"1
wide, any gradual absorption change vs. frequency is not imme-
diately obvious by visual inspection. However, the transmission
in the range covered in Fig. 5 does change quite rapidly from
left to right. As a matter of fact, this trace showing gradual
change in absorption is large in comparison with other recorder
traces.
Figure 6 is a similar recorder trace for the 11.3 ym band.
The experimental procedure was the same except that a different
diode laser was used. The reference lines were either those of
13C02 or NH3 for this region. As mentioned before, there was no
fine structure anywhere throughout this band.
The spectral slitwidth of the monochromator was about 2 cm"1
or less. It was sufficiently narrow to isolate one single laser
mode. Throughout the low pressure ^SO^ scans, both background
and absorption scans were repeated to determine how well the
curves could be repeated. If there was a small drift due to
electronics, laser source or the temperature of the absorption
cell, an average value is used. In particular, a strip chart
recorder was used to view this slight drift. Figure 7 is such
a strip chart recorder trace near 1222 cm"1. The transmitted
signal through the cell was monitored as a function of time as
the heater for the H2S04 reservoir was turned on near transmis-
sion peak and turned off near transmission minimum. The tempera-
ture at each peak transmission was about 90°C and at each minimum
transmission was 170°C.
Figure 8 is a plot of low pressure I^SO. absorption vs. fre-
quency from 1210-1240 cm"1 by combining a large number of scans
like that shown in Fig. 4 and Fig. 5. The absorption curve has
two peaks with the peak at 1222 cm"1 being much sharper. At this
peak, the absorption coefficient is 7.2 citi'Vatm. This compares
with a value of 6.0 cm~1/atm from Majkowski, which was measured
at one atmosphere pressure of nitrogen and a value of about
4.0 cm~1/atm from Burch, the I^SC^ pressure for which was not
accurately known.
18
-------
6
LJ
>
ACKGROUND (70°C)
FREQUENCY (cm'1)
Figure 5. Low-pressure absorption H2SC>4 absorption near 1224 cm
-1
-------
RESERVOIR
TEMPERATURE
FREQUENCY (cm"1)
Figure 6. Low-pressure H SO absorption near 880.5 cm
-------
to
ELAPSED TIME (HOURS)
Figure 7. Strip chart recording of low pressure
vs . time
absorption
-------
8
KJ
fO
4-1
H3
o
~ 4
o
0
a.
a:
o
co
RESERVOIR TEMPERATURE = 170°C
CELL BODY TEMPERATURE = 200°C
PH2S04 = °'67
I
1210
1220
1730
FREQUENCY (cm'1)
1240
1250
Figure 8.
Low pressure H_SO absorption (P~0.67 Torr, T=«1700C)
-------
A similar plot was constructed for the 870-895 cm"1 region
(see Fig. 9). There is only one peak. There was no sharp struc-
ture. The peak occurs at 880 cm"1 and is much broader in com-
parison with the peak at 1222 cm"1. The absorption coefficient
at this peak is about the same as that at 1222 cm'1 and is equal
to about 7.0 cm~1/atm.
HIGH RESOLUTION ABSORPTION MEASUREMENTS OF H?SO. VAPOR IN
ATMOSPHERIC PRESSURE OF NITROGEN
Experimental Apparatus and Methods
To perform the atmospheric scan, a different experimental
setup with a different H2S04 absorption cell was used. Figure
10 is a schematic diagram for the atmospheric pressure, flowing-
gas HoSO^ vapor absorption measurements. With the exception of
the absorption cell, the optical and electronic equipments are
the same as those shown in Fig. 1. The flowing-gas cell is a
transmission type. It has a window-to-window length of 64 cm.
Again, wedged silicon windows were used and they were similarly
sealed to the pyrex tube. Figure 11 is a detailed diagram of the
flowing-gas H2S04 absorption cell and the hot H2S04 vapor genera-
tor. The inlet and outlet of the cell are closed to the windows
to ascertain gas density uniformity throughout the cell. Thermo-
couples were attached to the cell wall at 10 cm intervals and to
the cell windows. High temperature heating tapes were wrapped
around the cell body with extra turns near the windows. The cell
was placed inside a wood box, the inside surface of which was
lined with heavy ceramic fiber cloth. Two salt windows were used
to cover the openings on the box in order to minimize air convec-
tion and maintain a stable temperature operation around 230°C.
The HoS04 vapor generator is similar to that used by
Majkowski (7). A 15 cm3 of 37N solution of reagent grade H2S04
acid was introduced into the reservoir, which was then immersed
in a heated oil bath. Dry nitrogen was used as a carrier gas at
a flow-rate of about 1 liter/hr. The heat exchanger coil is a
6 mm O.D. pyrex glass tubing closely wound on a 37 mm O.D.,
77 mm mandrel. The end of the pyrex coil extends into the
reservoir and terminates in a glass frit (40-60 vim). The glass
frit was necessary to obtain fine bubbling through the H2S04
liquid and to prevent large aerosol from getting into the H2S04
solution. At the outlet of the reservoir, quartz wool was used
as a gas filter which allows only vapor to pass but prevents
liquid aerosol from passing through. The oil bath used a high
temperature oil (Dow Corning 200 fluid) which has a rather high
flash point. However, it has a rather high viscosity even at
200°C so that constant stirring had to be applied to obtain a
uniform temperature distribution throughout the bath. A thermo-
couple immersed in the bath forms part of a temperature control
system.
23
-------
M
8
.u
-------
TEMPERATURE
CONTROLLER
COOLER
LASER
LASER
POWER
SUPPLY
CHOPPER
GRATING
SPECTROMETER
ABSORPTION CELL
X-Y
RECORDER
LOCK-IN
AMPLIFIER
DETECTOR
Figure 10. Experimental setup for H SO absorption at atmospheric
N_ pressure
-------
ro
SILICON
WINDOW
H2S04 ABSORPTION CELL
N-
PREHEATER
COIL
64 cm
28 mm I,D, PYREX
I
^--AZEOTROPIC
<^r" H2S04
OIL BATH
TO TRAP
••II
Figure 11. Absorption cell for H2S04 absorption at atmospheric nitrogen pressure
and H2S04 vapor generator
-------
In operation, the absorption cell body was kept at near
230°C and the oil bath at 180°C or at 200°C<, The temperature at
the connecting tube between the I^SO^ reservoir and the cell in-
let was independently controlled to be as close to the bath tem-
perature as possible. To achieve an azeotropic condition, the
liquid H2S04 was heated in the oil bath to 180°C and hot dry
nitrogen was bubbling through the liquid for about 15 hours* At
the end, only about 90% of the fluid remained in the reservoir.
Since the 37N solution has an H2S04 concentration of 98.9 ± 0.4%
a 10% removal is believed sufficient in attaining the azeotropic
condition (see page 13).
For a background scan, the reservoir temperature was lowered
to about 60°C by cooling it with room air. Because the rate of
dry nitrogen flow was about 1 liter/hr. and the volume of the
cell was about 0.5 liter, sufficient time (~1 hour) was allowed
after each reservoir temperature change before any data was
taken. In practice, the equilibrium condition was noted by ob-
serving that the transmitted laser signal through the cell
reached a steady-state value.
Experimental Results
Using the experimental setup for atmospheric pressure f^SO^
absorption measurements (see Fig. 10), the region from 1220 to
1222 cm"^- was scanned. The reasons that this particular 2 cm~l
region was selected are that the I^SO^ absorption peak occurs
near 1222 cm""-'- and that there is a very strong 1^0 line near
1225 cm"-*- which would present an interference problem. Figures
12 through 14 are the recorder traces. As in the low-pressure
scans discussed earlier, for each spectral region, there are
several traces vs. the same diode current, namely, background
trace, etalon trace, reference trace, and absorption trace. As
noted, there are two absorption traces, one at 180°C and the
other at 200°C. The absorption vs. frequency profile is smooth.
Since the recorder trace covers only about 0.5 to 0.7 cm~ , about
3 or 4 similar scans are necessary to cover a 2 cnT^ continuous
region. Figure 15 is a summary of the atmospheric scan data.
The cell temperature was maintained at about 230 ± 10°C for both
the 180°C and 200°C scans.
Since SC>2 have many lines in this 2 cm~^ region, it was used
conveniently as frequency markers. The absolute frequency was
determined by the 1^0 line at 1220.320 cm=^.
The absorption coefficients are given in units of
10"^ ppm~l-m""-*-. We have used the partial pressure spectroscopi-
cally determined in the present experiment (see section on par-
tial pressure measurements, p. 42).. At the peak absorption,
1222 cm" , points based on Majkowski's data and Burch's result
(2 Torr of H2S04 pressure) were also plotted. The 180°C absorp-
tion curve agrees within experimental error with Majkowski's
27
-------
CO
1219.80 TO 1220.50 en
Figure 12.
FREQUENCY (cnT1) -»•
absorption at atmospheric nitrogen pressure
-------
1220.5 TO 1221.22 cm
FREQUENCY (cm-1) -»-
Figure 13. H SO absorption at atmospheric nitrogen pressure ,
-------
CO
o
1221.22 TO 1221.98 cm
BACKGROUND
ATTENUATED BY 1.11
FREQUENCY (cm~)
Figure 14. H->sod absorption at atmospheric nitrogen pressure
-------
o
X
E
200°C
0 (MAJKOWSKI)
u>
z
UJ
li.
u.
01
o
o
Q.
OC.
O
a.
cs
8
6
4
180°C
O (MAJKOWSKI)
A (BURCH)
I
1219
1220
1221
FREQUENCY (cm'1)
1222
Figure 15. H2SO4 absorption at atmospheric nitrogen pressure
-------
result. Burch's data appears to be too low. Since only one scan
of Majkowski"s 25 scans was performed at 200°C and the majority
of his scans were performed at 180°C, his 180°C data point (in
good agreement with our 180°C data) is likely to be more accurate
than his 200°C data point, which is about 15% smaller than our
result. Figure 16 shows the H2S04 absorption for the spectral
region 1210-1240 cm"1 and the 2 cm region we picked for atmos-
pheric scan.
Similarly, we obtained absorption scans for the 880 cm"1
region. Figures 17 through 20 show the recorder traces covering
the spectral region from 878.75 to 881 cm"1. The reference lines
are l3C02 lines. Figure 21 is a summary for the 2 cm"1 region
around 880 cm"1.
Majkowski did not measure this region. Burch's data point
plotted on Fig. 21 was based on his 2-Torr H2S04 measurement at
1222 cm"1 and his 3-Torr H2S04 absorption ratio at 1220 and
879 cm"1. Although his data point is in good agreement with our
180°C data, a direct comparison cannot be made because our mea-
surements were performed at atmospheric pressure.
Figure 22 shows where the 2 cm atmospheric scan is located
with respect to the 880 cm"1 band.
HIGH RESOLUTION ABSORPTION MEASUREMENTS OF INTERFERANTS - H~0,
S02 AND C02
Experimental Apparatus and Methods
The experimental apparatus and methods used for the absorp-
tion measurements of interferants are exactly the same as those
used for the H2SO, vapor absorption measurements except that a
larger cell was used. The absorption cell was 1.2 m long, only
1.1 m of which was heated to 200°C. The same reference gases
were used, namely H20 and N20 in the 1222 cm"1 region and 1^C02
and NH3 in the 880 cm"1. Because the lines of S02 are very dense
in the 1222 cm"1, it was conveniently used as a frequency marker.
Experimental Results
— 1 —1
For the 2 cm region around 1221 cm , there are a large
number of hot S02 lines as observed in the 1.2 m interference
cell filled with 9 Torr of S02. Figure 23 shows one of several
recorder traces in this region. Figure 24 is a summary plot of
S02 absorbance vs. line position over a region about 3 cm"1 wide
near 1221 cm"1. It can be seen that the distribution of absorp-
tion lines are not uniform. Several strong peaks are present.
Also, there are about four low-absorption regions, centered at
about 1220.0 cm"1, 1221, 1221.5 and 1222.1 cm"1. For the same
frequency region, a continuous absorption scan was also made
32
-------
o
CQ
(£
o
1.6
3
o' 0.4-
in
CM
X
ATMOSPHERIC PRESSURE
L = 64 cm
— 200°C
— 180°C
Q3
0.2
01
co
CL
o
CO
o 0.1
fN
1220
1221
WAVENUMBER (cm
-1,
1222
LOW PRESSURE
L = 50 cm
T = 170°C
I
'1210
1220
WAVENUMBER
~
1230
1240
Figure 16. Location of the 2 cm atmospheric scan region with respect
to the 1210-1240 cm low-pressure scan
33
-------
LO
878.75 TO 879.39 cm
BACKGROUND
ATTENUATED BY 1.11
FREQUENCY (cm
Figure 17. Hoso>i absorption at atmospheric nitrogen pressure
-------
U)
U1
879.33 TO 879.98 cm
Figvire 18.
FREQUENCY (cm-1) •*-
absorption at atmospheric nitrogen pressure
-------
en
FREQUENCY RANGE
880.14 TO 380.64
FREQUENCY (cm
Figure 19. H SO absorption at atmospheric nitrogen pressure
-------
OJ
880.49 TO 881.01 cm
FREQUENCY (cm"1)
Figure 20. H2SO4 absorption at atmospheric nitrogen pressure
-------
o
^H
X
E
I
I
a
CO
00
LU
o
o
o
a.
cc
o
in
ca
a
6
878
(BURCH) A
879
880
180°C
881
FREQUENCY (cm'1)
Figure 21. H«SO absorption at atmospheric nitrogen pressure
-------
2.4*
ATMOSPHERIC PRESSURE
L = 64 cm
200°C
u
CQ
a:
o
to
CQ
0.8
O
CO
tN
X
•180°C
WAVENUMBER (cm )
LOW PRESSURE
L = 50 cm
T = 170°C
900
WAVENUMBER (cm )
Figure 22. Location of the 2 cm atmospheric scan region with respect
to the 870-895 cm~ low pressure scan
39
-------
*>.
o
L = 1.1 m
T = 200°C
1221.50 TO 1222.15 cm
FREQUENCY (cm
Figure 23. SO., absorption near 1222 cm
-1
-------
6
5
i 4
3
HI
o
^»
ABSORBAt
u>
(N
O
2
1
0
—
1219
I
i
i
1220
,
L =
T =
p
2
1221
1.1 m
200°C
= 9 TORR
i
1222
WAVENUMBER (cm )
Figure 24. SO absorption
-------
using 10 Torr of H^O in the same 1.2 m cell at 200°C. Figure 25
is a scan around 1220.320 cm'1 showing two H20 absorption lines.
A cell filled with S02 gas at room temperature was used for pro-
viding frequency markers. Figure 26 is a summary plot of H20
absorbance vs. H20 line positions. There are only three signi-
ficantly strong H20 lines between 1219 and 1222 cm"1. For stack
gas monitoring, any one of the four regions just mentioned in
which the S02 absorption is relatively low can be chosen. How-
ever, consideration of interference by nearby H20 lines reduces
the number of choices to two, namely, the 1221.5 and 1222.1 cm"1
regions. For example, for a typical stack condition, the absorp-
tion at 1220.0 cm"1, 1221.0 cm"1 and at 1221.5 cm"1 by the wing
of the 1220.3 cm"1 H20 line are estimated to be about 2.5 per-
cent, 0.56 percent and 0.2 percent respectively, and the absorp-
tion at 1222.1 cm"1 by the wing of the 1225 cm"1 H20 line is
estimated to be about 0.1 percent.
For the 880 cm"1 region, Figs. 27 and 28 show the absorp-
tion of H20 and C02 lines vs. frequency from 878 through 881 cm"1.
Only one H2Q line of small absorbance was observed near
878.56 cm"1. The CO^ lines were not observed using the 1.2 m
cell at 200°C with a t'02 pressure of 100 Torr. A good choice of
spectral position for H2S04 monitoring in this 3 cm"1 region is
880 cm"1 where the H2S04 absorption coefficient is near its peak.
SPECTROSCOPIC DETERMINATION OF PARTIAL VAPOR PRESSURES ABOVE
HOT AZEOTROPIC H2S04 SOLUTION
Experimental Apparatus, Methods and Results
Using tunable diode laser spectroscopic methods, the partial
pressures of H^O and SO., vapors above a hot aqueous sulfuric acid
solution were determined by measuring the absorption line
strength of H2O (803) vapor and comparing it against a calibrated
strength vs. H20 (803) pressure curve. The partial pressure of
H2S04 vapor is then equal to the difference between the total
pressure (also measured) and the sum of the H20 and SOi partial
pressures. The pressure measurements were performed with the
H2S04 reservoir at 107°C, 150°C and 200°C and the cell body at
20°C higher in each case.
The 64-cm long absorption cell used in the atmospheric sul-
furic acid absorption measurements was modified slightly for the
measurements. The modification included adding a side arm and
an all-teflon valve. With the exception of the absorption cell
the experimental setup is similar to that shown in Fig. 1. The
following paragraphs describe the methods and results of the mea-
surements.
42
-------
OJ
L = 1.1 m
T = 200°C
LEGEND: 0 H20 LINES
X S02 LINES
FREQUENCY ( cm )
Figure 25. HO absorption
-------
10-
0.8-
3 0.6h
OQ
CK
O
V)
DQ
o 0,4-
CM
X
}
LINES
L = 1.1 m
T = 200°C
P., n = 10 TORR
H2
-------
0.20 •
L = 1.1 m
T = 200°C
Pr_ = 100 TORR
C02
0.16
01
0.12 •
CQ
O£
O
CO
OQ
0.08
0.04-
WEAK C02 LINES (NOT OBSERVED)
878
879 -i860
WAVENUMBER (cm )
881
Figure 27. CO absorption
-------
020-
OJL6-
u
Q12-
m
ce.
o
CO
CQ
ao4 -
/Ha
0 LINE
L = 1.1 m
T = 200°C
PH20 = 15 T°RR
878
srs _x88o
WAVENUMBER (cm )
881
Figure 28. HO absorption
-------
H20 Partial Pressure Measurements--
The modified cell was used for both the absorption strength
measurements and the calibration measurements . In the calibra-
tion measurements, a few cm^ of distilled H20 was placed in the
side arm which was then immersed in saturated brine solution at
about -19 °C. At this temperature the water vapor pressure is
about one Torr. The absorption strength of an H20 line at low
pressure near 1330 cm~3- was measured at a number of cell body
temperatures by tuning a diode laser through the line. Figure
29 is a plot of the absorbance per Torr of H20 pressure vs. cell
body temperature. After obtaining the calibration curve, the
water in the side-arm was replaced by an H2SO^ azeotrope and an
absorption scan was obtained . Figure 30 is a recorder trace of
the same H20 absorption line above 150°C H~SCK azeotrope. The
azeotrope was prepared by distilling off about one third of a
37N H2S04 at the azeotrope temperature. By comparing the absorp
tion strength of the line in Fig. 30 with the calibration curve
in Fig. 29, the partial pressure of H20 was determined to be
about 0.23 Torr. That is, for a 150°C azeotrope (side arm
temperature)
aL (calculated from Fig. 30, T^. = 170°C)
PH0 = aL/PR Q(from Fig. 29 at T- = 170°C)
Azeotrope at other temperatures were prepared and the partial
pressures were determined similarly. Table 1 summarizes the H2O
partial pressure measurement results. The numbers in parenthe-
sis are from Luchinskii for the atmospheric azeotrope with a 98.3%
H2SC>4 concentration (9) .
S03 Partial Pressure Measurements —
For the SOj partial pressure measurements the same absorp-
tion cell used for the H20 partial vapor pressure measurements
was used. The SO^ was purchased from a commercial source with a
stated purity of 99%. Because of the much higher vapor pres-
sure (14), the side arm was immersed in a methanol-water mixture
at about -40°C (~1 Torr of S03 vapor pressure) for the calibra-
tion runs. Figure 31 is the plot of absorbance per unit pressure
vs. temperature for an S03 line near 1354 cm"1. Figure 32 is a
recorder trace of the same 803 line above a 150°C H2SO, azeo-
trope. Table 2 is a summary of the 803 partial pressure measure-
ment results. Again, the numbers in parenthesis are from
Luchinskii. It can be seen that the S03 pressure is much higher
than that of the 98.3% azeotrope.
14. Handbook of Chemistry and Physics, 46th Ed. 1964, p. D101,
The Chemical Rubber Col, Cleveland, Ohio.
47
-------
CO
1.0
a:
a:
a
Q.I
250
200
ISO
100
75
50
8
10
400/T (K"1)
Jl
14
Figure 29. Absorbance per unit pressure (aL/Torr) vs. temperature of cell
body for the H20 line at 1330 cm"1. (L = 64 cm)
-------
VO
Ul
u
z
ce.
LLJ
>
UJ
ee
0.0048 cm
-1
J_
FREQUENCY (cm'1)
Figure 30.
H2O line at 1330 cm
L=64 cm, T, , =*'.
body
-1
above a 150°C H_SO azeotrope
-------
T f ' C )
250 200 ISO 100 75 fiO 2,5
a:
cc
O
O.I
4000/T
12
13
14
Figure 31. Absorbance per unit pressure (aL/Torr) vs. temperature of
cell body for the SO line near 1354 cm (I/*64 cm)
50
-------
= 64 cm
Tbody ' 170°C
FREQUENCY (cm"1) •*-
Figure 32. SO absorption lines near 1354 cm
-1
-------
TABLE 1. H20 PARTIAL PRESSURE VS. TEMPERATURE OF
AZEOTROPE
Temperature (°C)
Partial Pressure (Torr)
107
150
200
0.024 ± 0.002 (0.01)
0.23 ± 0.02 (0.145)
2.3 ± 0.2 (1.7)
TABLE 2. S03 PARTIAL PRESSURE VS. TEMPERATURE OF
AZEOTROPE
Temperature (°C)
Partial Pressure (Torr)
107
150
180
200
0.022 ± 0.002 (0.005)
0.21 ± 0.02 (0.086)
0.90 ± 0.09 (0.43)
2.0 ± 0.2 (1.3)
52
-------
Total Pressure Measurements--
Since it was not possible to isolate any single H2S04 line
for absorption strength measurements and since ^SO/i vapor does
not exist by itself alone„ the partial pressure of §2^°4 cann°t
be measured independently? therefore, it was necessary to mea-
sure the total vapor pressure and then determine indirectly the
partial pressure of H2S04 by subtracting the H20 and SO^ partial
pressures from the total pressure. The apparatus used for mea-
suring the total pressure consists of a pyrex "U" tube manometer
connected to a glas bulb containing the azeotrope. Figure 33
is a sketch of the total pressure measuring apparatus« Azeotrope
prepared by distillation was placed in the bulb (the "U" tube at
this time had no azeotrope liquid). The space above the bulb and
the entire "U" tube section were exhausted to high vacuum. This
was done with the oil bath near room temperature. Valve A was
then closed. The "U" tube was then filled to level as shown
with liquid from the bulb by tilting the apparatus momentarily.
The temperature of the bath and the section of the "U" tube
covered by the heating tape was then brought up to the azeotrope
temperature (100°C, 150°C or 200°C) with the glass tubing being
~15°C hotter than the bath. As the total pressure above H2S04
solution increased, a sufficient amount of dry nitrogen was let
into the right hand side of the "U" tube to keep the liquid
levels on both sides equal. When this occurred,, the total vapor
pressure would be exactly equal to the pressure gauge reading0
As a check, the apparatus was used to measure the water
vapor pressure at about 40 Torr. The result indicated that the
measurement was accurate to less than one Torr. Since the H^SO^
azeotrope has a specific gravity of about 1.83 and the liquid
level can be read to about 1/2 mm, the reading error is about
0.07 Torr. Table 3 is a summary of the total vapor pressure
measurements. The overall error includes the liquid level read-
ing error, pressure gauge inaccuracy, leveling of the "U" tube,
an estimated error due to possible condensation of the vapor
above the "U" tube liquid and oil bath temperature uncertainty.
Partial Pressure of I^SO^-—
Table 4 lists the partial pressure of H2SC>4 based on the
total pressure and partial pressure measurements discussed in
the preceding paragraphs. As mentioned in the beginning of ex-
perimental methods for partial pressure determination, the par-
tial pressure of H2SC>4 was equal to the difference between the
total pressure and the sum of the 803 and I^O pressures. It can
be seen that the partial pressure is considerably lower than
those from Luchinskii based on the 98.3% azeotrope. Therefore,
as a result, the absorption coeffijient of ^SCD, vapor is corres-
pondingly higher. For example, previously we obtained an atmos-
pheric H2S04 absorption coefficient of about 4.9 x 10"^ ppm-l-m"^-
based on Luchinskii"s pressure data. Based on our new pressure
data, the coefficient should now be 6.4 x 10=^
53
-------
PRESSURE
GAUGE
HEATING TAPE
en
TO PUMP
H2S04
AZEOTROPE
OIL BATH
Figure 33. Total pressure measurement apparatus
-------
TABLE 3. TOTAL PRESSURE VS. TEMPERATURE OF H-SO.
AZEOTROPE
Temperature (°C)
Total Pressure (Torr)
107
150
200
232
0.08 ± 0.02 (0.055)
0.75 ± 0.08 (0.68)
7.8 ± 0.8 (7.46)
23 ± 2 (26.2)
TABLE 4. H2S04 PARTIAL PRESSURE VS. TEMPERATURE
OF H2SO4 AZEOTROPE
Temperature (°C)
Partial Pressure (Torr)
107
150
200
0.034 ± 0.02 (0.04)
0.32 ± 0.08 (0.45)
3.5 ± 0.8 (4.4)
55
-------
Based on our experimental results, we computed the azeotrppe
composition and dissociation constant Kp. Table 5 is a list of
the azeotropic composition vs. temperature and Fig. 34 is the Kp
constant vs. temperature. The dissociation constant of Kp was
calculated from the partial pressures given in Tables 1, 2 and 4
for three azeotrope reservoir temperatures, namely 107°C, 150°C
and 200°C. Since the cell body temperature was 20°C higher in
each case, the calculated Kg was assigned to the corresponding
cell body temperature at 127°C, 170°C or 220CC. Figure 34 shows
that the dissociation constant agrees well with the extrapolated
value from Gmitro's plot. The spectroscopic technique has
greatly extended the measurable range of the dissociation con-
stant to lower temperature. Although the partial vapor pressure
of H20 above H^SC^ azeotrope near room temperature is very low,
the present spectroscopic method is capable of determining the
dissociation constant of H2S04 near room temperature.
Finally, Fig. 35 is the plot of vapor pressure vs. tempera-
ture for HjO, S03 and HoS04. For completeness, the atmospheric
azeotrope data previously measured are also included (8) .
As mentioned in the preceding paragraphs, our measured 112804
partial pressures are about 30% lower than those from Luchinskii
and our measured H^O and 803 partial pressures are considerably
higher than those from Luchinskii. The effect of further disso-
ciation of 119804 on going from the reservoir to the hotter cell
body (20°C higher in temperature than the reservoir) could ac-
count for the discrepancy.
The following paragraphs show that the measured partial
pressures are in good agreement with the calculated values tak-
ing dissociation into consideration.
In the atmospheric scan, in which the reservoir temperature
was 200°C and the cell body was at 230°C, a value of absorbance
equal to 2.028 (aL = Jin IO/I = Jin 15.2/2.0) was obtained (see
Fig. 17) . Since the cell length was 64 cm and the measured 1^804
partial pressure at 200°C reservoir temperature (Table 4) was
3.5 Torr, the absorption coefficient for the 880 cm"1 band was
calculated to be
_ 2.028 760 _ 2.028 (760) _ ^T1/**-™
Ot — - = - T, — p — — ;rj = = — b. bo Cltl /atltl
L J.b 54 3 • 5
which was plotted in Fig. 21.
We will now show that the observed value of H2SO4 partial
pressure, 3.5 Torr (which we determined using the spectroscopic
techniques to measure partial pressures) was the result of fur-
ther dissociation of 1*2804 vapor from the 200°C reservoir on
going to the 220°C cell body. The other measured pressures were
Pso = 2.1 Torr, PH,,O = 2'3 Torr and Ptotai = 7.8 Torr. From
••) ~
56
-------
TABLE 5. AZEOTROPIC COMPOSITION VS. TEMPERATURE OF
HS0 AZEOTROPE
Temperature °C
Azeotropic Composition (%
107
150
200
326
99.35
99.31
99.01
98.48
57
-------
LEGEND
" PRESENT DATA
SOLID CURVE
BLACK DOTS
PREVIOUS DATA
AND VERMEULEN)
1.3
Figure 34. Dissociation constant K^ of H SO gas vs. temperature
58
-------
lOOOr
T(°C
250 * a
(50
100
-PREVIOUS MEASUREMENT (REF, 8)
100
LEGEND:
A = H2so4
e = H2o
x = so3
10
ce.
a:
O
(X
(ft
CO
LLJ
Of.
c.
0.1
0.01
-t
•t
Figure 35.
4000/r (
Partial pressure of H-SO.
H_SO solution vs. temperature
I0
i
II
H0 and S0_ above azeotropic
J
59
-------
these partial pressures, the dissociation constant Kp = P
is equal to 1.38 at 220°C, which is only abou
11% smaller than the value of 1.541 calculated by extrapolating
Luchinskii's partial pressure data to 220°C. We can now compare
our observed partial pressures with the partial pressures calcu-
lated using the following set of equations (from Luchinskii).
PH20 + PS03 + PH2S04 - Ptotal - 7'65 Torr (1)
98.082 (PH SQ + PSQ )
24 3 = = W - 99.144% (2)
P • P
H90 SO,
—== i = Kp (220°C) = 1.541 (3)
PH2S04
where W is the azeotropic composition by weight.
Table 6 shows the results for two values of W, 99.144% and
98.402%. The last row in Table 6 contains the observed values.
Two values of W were used because of the lack of experimental
data for W vs. total pressure azeotrope in the low pressure range.
Based on Kunzler's data on total pressure vs. azeotropic compo-
sition (see Fig. 36), an extrapolated value of W is estimated to
be about 99.0 ± 0.2% for Ptotal =7.65 Torr. It turns out that
the calculated partial pressures are very insensitive to this
small change in the azeotropic composition (see Table 6). The
agreement between observed and calculated partial pressures are
indeed very good.
The justification for the use of Equations (1),(2) and (3)
are
1. total pressure in the cell body is controlled by the
pressure at the reservoir which is 7.65 Torr at 200°C
(Luchinskii)
2. the dissociation took place in the cell body and Kp at
220°C must be used
3. the azeotropic composition is that of the reservoir
liquid at 200°C, since heating the gaseous constituents
above an azeotrope does not significantly change its
azeotropic composition.
It appears from the above discussion on the partial pressure
measurements and independently calculated results that the H2S04
partial pressure was valid as measured, and Equations (1), (2)
60
-------
TABLE 6. CALCULATED PARTIAL PRESSURES INCLUDING THE EFFECTS
OF DISSOCIATION
W(%)
99.144
98.402
99.01
Pressures (Torr )
H20
2.349
2.459
2.3
so3
2.1
2.0
2.0
H2S04
3.201
3.191
3.5
Total
7.65
7.65
7.8
Method
calc.
calc.
observed
61
-------
0?-
GC
<£
O
K
e/5
co
UJ
K
a
10
98.5
98.7
% H2S04
98.9
Figure 36.
Total vapor pressure above H_SO solution vs. azeotropic
composition
62
-------
and (3) can be used with confidence to determine the partial
pressure of H2S04.
A careful comparison of the experimental conditions between
the atmospheric scan and the low pressure scan (partial pressure
measurements) shows that one cannot use these three equations to
calculate the H2SC>4 pressure for the atmospheric scan. The rea-
son is that the gas in the atmospheric scan was flowing into the
cell body from the reservoir and the total pressure of the con-
stituent gases ( H20, 803 and 112804) was not constrained to the
reservoir pressure. If the cell body were at the same tempera-
ture as the reservoir (200°C), the total pressure of the consti-
tuent gases would be 7.65 Torr. As much as the cell body was
maintained at 230°C more 1^804 vapor would be dissociated accord-
ing to the following relationship
(PH Q + AP)(PSQ + AP)
Kp(230°C) = 1.99 = ^-(p r-^lj (4)
where AP is the change in pressure due to dissociation. The
value of 1.99 was again based on extrapolated partial pressure
values from Luchinskii's data. PH?O = 1«81 Torr, PgOo = !••* Torr
and PH->SO>I = 4.54 Torr to be used in Equation (4) are the pres-
sures at the reservoir. Solving for AP, we obtained AP =
1.081 Torr. The final pressures in the cell body are then
PH 0 = 2.891 Torr, Pso = 2.38 Torr, PH2S04 = 3«459 Torr and
ptotal = 8.731 Torr. The buffer gas nitrogen at 1 atm .Is assumed
to have no influence at the dissociation coefficient Kp. Based
on this value of 112804 partial pressure, the peak absorption
coefficient at atmospheric pressure is 6.87 cnTVatm in the
880 cm"1 and about 6.46 cm~Vatm in the 1220 cm~l region. These
values are valid for a cell wall temperature of 230°C. The ab-
sorption coefficient should remain relatively unchanged in the
200-250°C range if it is assumed that the individual 1^804 lines
making up the absorption profiles are relatively independent of
temperature.
By considering the effects of dissociation we find that our
results for the partial pressures are in better agreement with
previous determinations and that our absorption coefficient re-
mains essentially unchanged from that determined from the re-
sults of transmission measurement and partial pressure measure-
ment.
63
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
' M-R60N0°/2-78-019
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
SPECTRAL MEASUREMENTS ON GASEOUS SULFURIC ACID
USING TUNABLE DIODE LASERS
5. REPORT DATE
February 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHO
Richard S. Eng, Kenneth W. Nil! and Jack F. Butler
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFOl-
•JG ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Laser Analytics, Inc.
38 Hartwell Avenue
Lexington, MA 02173
IAD 712 BA-22 (FY77)
11. CONTRACT/GRANT NOT
68-02-2482
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory - RIP,. NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park. North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final 11/76 - 7/77
14. SPONSORING AGENCY CODE
FPA/fifin/nq
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Using a tunable diode laser spectrometer with a spectral resolution of about
-4 -1
10 cm , the important central portions of the two infrared absorption bands
of H2S04 at 8.2 ym (1222 cm"1) and 11.3 urn (880 cm"1) have been scanned at low
pressure H.2 Torr of H2S04) and at atmospheric nitrogen pressure. Maximum
absorption coefficients have been measured to be 6.5 cm'Vatm and 6.9 cm'Vatm
at the 8.2 pm (1222 cm" ) and 11.3 ym (880 cm"1) bands, respectively.
Interference
,~ I » „, j r%r\r\ .*** I
spectra of S02> C02 and H20 near the H2S04 absorption peaks at 1222 cm"1 and 880 cm
were scanned using a 1.1 m cell at 200°C to determine interference-free regions.
A spectroscopic method was used to measure the partial pressures of H2SO., SO-
and H20 vapors above azeotropes of H2S04 at 107°C, 150°C and 200°C. The expected
performance characteristics of an H2S04 tunable diode laser stack monitor are
considered on the basis of the above results.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
* Air pollution
* Sulfuric acid
* Spectral determination
* Spectrometers
* Lasers
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
13B
07B
14B
20E
i:, rmuu i I
STATEMENT
19. SECURITY CLASS (ThisReport)
RELEASE TO PUBLIC
EPA FO.III 2320-1 (9-73)
UNCLASSIFIED
21. NO. OF PAGES
72
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
64
------- |