EPA/600/4-85/043
June 1985
DEVELOPMENT OF AN OPTICAL MONITOR
FOR TOXIC ORGANIC COMPOUNDS
IN AIR
by
T. Hadeishi, M. Pollard, R. McLaughlin and M. Koga
Lawrence Berkeley Laboratory
University of California
Berkeley, CA 94720
Interagency Agreement No. DW 930479-01
Project Officer
Donald R. Scott
Methods Development and Analysis Division
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina 27 711
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, N.C. 27711
-------
DISCLAIMER
The information- in this document was funded wholly or in part by
the U. S. Environmental Protection Agency under Interagency Agreement
DW-930479-01 to Lawrence Berkeley Laboratory. It has been subject to
the Agency's peer and administrative review, and it has been approved
for publication as an EPA document. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
ii
-------
FOREWORD
Measurement and monitoring research efforts are designed to antici-
pate environmental problems, to support regulatory actions by developing
an in-depth understanding of the nature and processes that impact health
and the ecology, to provide innovative means of monitoring compliance
with regulations, and to evaluate the effectiveness of health and
environmental protection efforts through the monitoring of long-term
trends. The Environmental Monitoring Systems Laboratory, Research Tri-
angle Park, North Carolina, has responsibility for: assessment of
environmental monitoring technology and systems for air, implementation
of agency-wide quality assurance programs for air pollution measurement
systems, and supplying technical support to other groups in the Agency
including the Office of Air and Radiation, the Office of Toxic Sub-
stances, and the Office of Solid Waste.
Analysis of organic compounds in the atmosphere is a difficult task
requiring a variety of analytical techniques to ensure that identifica-
tion and quantification are valid. The environmental analyst requires
methods that are precise, accurate and easy to implement. This study is
concerned with the development of a new instrument which utilizes the
information available in the high resolution absorption spectra of
organic compounds in the ultraviolet and visible regions. The use of
this instrument in the analysis of organic compounds may greatly improve
the ability of the analyst to produce valid results.
Thomas R. Hauser, Ph.D.
Director
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina 27 711
iii
-------
ABSTRACT
The objectives of this study were: a) to design, construct, and
deliver a prototype atomic line molecular spectrometer (ALMS) benzene
monitor and b) to locate matches of atomic lines and sharp molecular
absorption features in other toxic organic compounds for possible use
with the ALMS or TALMS technique. ALMS and TALMS are newly developed
high resolution molecular absorption techniques that are used in the
vacuum ultraviolet and ultraviolet regions of the spectrum to detect
organic molecules in the gas phase. The dual beam prototype ALMS
instrument was designed constructed, tested, and delivered to the
Environmental Monitoring Systems Laboratory, U. S. EPA, Research Trian-
gle Park, North Carolina, in December, 1984. It was designed for moni-
toring benzene and other organics with the 1S4.9 and 253.7 nm mercury
lines. The instrument consisted of three units: The optical unit
(weight: 28 lbs, dimensions: 28x10x12"); the electronics unit (weight: 6
lbs, dimensions 19x7x5.25") and a lamp driver (weight: 24 lbs, dimen-
sions 14.5x14x6.5"). The total weight was 58 lbs, which is less than
the TALMS benzene monitor previously developed (82 lbs). Tests of the
performance of the ALMS benzene monitor showed an approximate detection
limit of 250 ppbv at 184.9 nm.
The process of searching for signals in organic compounds has been
simplified by the development of a computer accessible data base of
atomic line location and relative intensities. This data base was used
to select lines for ALMS detection of o- and m-xylenes. Line matches
and TALMS signals were found for three new compounds: p-difluorobenzene
(Pt: 265.9 nm), m-dichlorobenzene (Ge: 269.1 nm), and p—
chlorofluorobenzene (Fe: 275.6 nm). The high resolution absorption
spectrum of p—difluorobenzene was determined near the 265.9 nn platinum
line.
This report was submitted in fulfillment of Interagency Agreement
DW 930479-01 by the Lawrence Berkeley Laboratory under the sponsorship
of the U, S. Environmental Protection Agency. This report covers the
period from October, 1983, to November, 1984, and the work was completed
as of November, 1984.
iv
-------
CONTENTS
Foreword ' iii
Abstract.... iv
Figures vi
1. Introduction 1
2. Conclusions and Recommendations 3
3. ALMS Instrument Construction and Testing 5
Benzene Detection by ALMS 5
Prototype Construction..... 8
4. TALMS and ALMS Detection of Substituted Benzenes 13
Atomic Line Data Base 13
ALMS Detection of Xylenes 14
TALMS Detection of m-Dichlorobenzene 14
TALMS Detection of p-Chlorofluorobenzene 15
TALMS Detection of p-Difluorobenzene 19
Line Shape Measurements of p-Difluorobenzene 20
Summary of TALMS Signals 23
References 26
Appendices
A. ALMS Instruction Manual 27
B. Electronic Schematics 30
v
-------
FIGURES
Number Page
1 Vacuum ultraviolet absorption spectra of
benzene derivatives and superimposed atomic lines 6
2 Diagram of the ALMS prototype instrument 7
3 Extinction coefficients of benzene
at various Iodine lines 9
4 Photograph of prototype ALMS optical unit 10
5 Photograph of front and back
of ALMS electronics module 11
6 TALMS signal for m-dichlorobenzene
at the Ge 269. 1 nm line 16
7 Iron emission line superimposed on
the p-chlorofluorobenzene absorption maximum 17
8 TALMS signal from p-chlorofluorobenzene
at the Fe 275.6 nm line IS
9 TALMS analytical curves for
p-difluorobenzene with the Pt 265.9 ran line 21
10 Percentage difference between o1'" and
-------
SECTION 1
INTRODUCTION
There is a great need for instruments to monitor specific organic
pollutants and classes of pollutants in ambient air and near sources
such as waste disposal and industrial production sites. Direct monitors
are needed for specific chlorinated hydrocarbons and various aromatic
hydrocarbons such as benzene, toluene, and other substituted benzenes.
Since substituted benzenes and other organic compounds absorb light in
the vacuum ultraviolet and the ultraviolet spectral regions, one possi-
ble detection method is the use of optical absorption techniques in the
gas phase. The use of high resolution absorption methods would increase
the selectivity of the technique by exploiting roLational-vibrational -
fine structure in the absorption spectra of the compounds.
Tunable atomic line molecular spectroscopy (TALMS) is a high reso-
lution, molecular absorption approach to monitoring organic vapors that
is different from most present analytical techniques [1] [2]. Whereas
present methods depend on some form of chromatographic separation for
compound identification, the TALMS technique depends only on ultraviolet
absorption properties. There is no separation procedure involved. It
is highly specific because it responds only to very sharp rotational-
vibrational molecular absorption features. However, TALMS has not been
shown to be very sensitive due to difficulties in locating atomic probe
lines near high intensity molecular absorption maxima. The lowest
detection limit found for benzene is approximately 10 ppmv [3], which is
too high for direct ambient air measurement.
A related technique, atomic line molecular spectroscopy (ALMS),
resulted from extending the ideas behind the TALMS technique in an
attempt to improve the detection limits. Differential absorption at two
wavelengths is the basis of both methods. In the TALMS case the
wavelength difference is determined by the very small Zeeman splitting
of the atomic line chosen for measurement. The small splitting insures
that the background correction will eliminate most interferences. In
the ALMS case, the wavelength positions are determined by choosing two
different atomic lines that are necessarily separated by a much greater
distance than in the TALMS technique. The inherent detection limit in
both techniques depend upon the differences in intensities in the molec-
ular absorption spectra at the two wavelength positions. Because the
wavelength separation is much greater with ALMS, the absorption differ-
ence can be much larger yielding a great improvement in detection lim-
its. For example a TALMS detection limit of 10 ppmv has been found for
benzene at 253.6 nm [3]. Judging from the much higher extinction coef-
ficients in the vacuum ultraviolet region for benzene, detection limits
1
-------
should be at least one thousand times better with the ALMS technique.
This would result in detection at 10 ppbv levels, a useful limit for
ambient air monitoring.
The major disadvantage of the ALMS technique is the possibility of
interferences from compounds other than those sought. The use of multi-
ple analysis lines will reduce this problem. However, the essential
assumption for background correction with additional analysis lines is
that absorption by interfering compounds is constant over the wavelength
interval used. If several different lines are used to monitor the same
molecule, the presence of an interferent can be detected because dif-
ferent apparent concentration values will be obtained at different
wavelengths. Although this may not provide a background correction, it
will alert the analyst to the presence of a problem. With the ALMS
technique the use of multiple analysis lines at properly chosen
wavelengths in the vacuum ultraviolet and in the ultraviolet regions may
also allow the determination of classes of compounds.
Previous studies of TA.LMS [4] [5] have resulted in construction and
evaluation of prototype instruments for general laboratory use and for
monitoring benzene. The goals of the present study are to: a) design,
construct, and deliver to the Environmental Monitoring Systems Labora-
tory, Research Triangle Park, North Carolina, an ALMS monitor and b) to
determine the spectral location of absorption features and matching
atomic lines in toxic compounds other than benzene. An ALMS instrument
would be useful in laboratory detection and field monitoring for benzene
and other toxic compounds. Determination of spectral locations of
molecular absorption maxima is necessary to optimize instrument perfor-
mance and to extend the technique to other organic compounds.
2
-------
SECTION7 2
CONCLUSIONS AND RECOMMENDATIONS
Design and construction of a prototype ALMS instrument for the
detection of benzene and other compounds was completed on schedule and
delivered to the Environmental Monitoring Systems Laboratory at Research
Triangle Park, North Carolina, in December 1984. It consisted of three
parts: an optical module, an electronics module, and a lamp driver. The
optical module weighs 28 lbs, the electronics module 6 lbs, and the lamp
driver 24 lbs. An instruction manual and schematic diagrams of the
electronics were also supplied. The instrument was constructed to
operate in the vacuum ultraviolet and ultraviolet spectral regions and
was equipped with a mercury lamp. Benzene and other organic compounds
can be detected with the mercury 184.9 and 253.7 nm lines. Tests with
benzene at 184.9 nm gave approximate detection limits of 250 ppbv. This
is an improvement by a factor of forty over the best TALMS detection
limits. Other atomic lamps can be used to obtain different analysis
wavelengths as required.
Searches for TALMS signals in other organic compounds were contin-
ued. This tedious process was greatly improved by development of a com-
puter search technique using National Bureau of Standards atomic line
information. After considerable experimentation with a variety of lamps
and medium resolution absorption spectra, TALMS signals were found for
p-difluorobenzene (Pt: 265.9 nm) m-dichlorobenzene (Ge: 269.1 nm) and
p-chlorofluorobenzene (Fe: 275.6 nm). TALMS signals have now been found
for benzene, bromobenzene, chlorobenzene, toluene, p-xylene, aniline,
phenol, pyridine, formaldehyde, m-dichlorobenzene, p—
chlorofluorobenzene, and p—difluorobenzene. Wavelengths for the ALMS
detection of ortho and meta xylenes also were selected. The high reso-
lution absorption spectrum of p-difluorobenzene was determined near the
platinum 265.9 nm line.
It was recommended that several modifications be made to improve
ALMS instrument performance. An arrangement could be devised to alter-
nately send two different lines from the same light source through the
instrument. Electronic subtraction of the signals from these two lines
will greatly reduce light source noise and electronic noise with a
corresponding increase in sensitivity. For simultaneous detection of
several compounds a device that automatically positions different
wavelength regions on the exit slit of the monochromator should be con-
structed. If it is necessary to make measurements in the vacuum ultra-
violet region or use weak emission lines, an arrangement for purging the
instrument and optical path will be important in order to increase
transmission. Addition of commercially available intense lamps of other
3
-------
elements is also important. The goal of these modifications is to
improve sensitivity, reliability and portability
Since major decreases in detection limits and extension of this
technique to other compounds are dependent on the location of proper
analysis lines, it is recommended that more studies be carried out to
locate such lines for compounds of interest.
A
-------
SECTION 3
ALMS INSTRUMENT CONSTRUCTION AND TESTING
BENZENE DETECTION BY ALMS
The gain in sensitivity that can be realized by using the ALMS
technique can be seen by comparing the extinction coefficients of com-
pounds to be monitored with the extinction coefficients now being used
for the TALMS technique. For benzene, at the mercury 253.7 nm line,
the extinction coefficient is approximately 100. Figure 1 shows the
absorption spectrum of a number of benzene derivatives [6] with the
position of several emission lines of iodine and arsenic superimposed.
The peak of the benzene absorption in the vacuum ultraviolet region
corresponds to an extinction coefficient of 126,000 [7]. This leads to
the expectation that a gain in sensitivity of 1000 is possible. Figure
1 indicates that with the proper choice of atomic lines in the vacuum
ultraviolet region, comparable gains can be achieved in the detection of
other benzene derivatives.
A block diagram of a double beam ALMS instrument is given in Figure
2. Light from an atomic lamp, e. g. mercury, passes through a purgable
monochromator to a beam splitter. Part of the light passes through the
sample cell to the photonm.ltiplier tube (PMT 1). The remaining light is
reflected to another photomultiplier tube (PMT 2) which is used to com-
pensate for lamp fluctuations. The signals from the photomultipliers
are processed through amplifiers and filters, and an absorbance signal
is produced on a strip chart recorder.
Before design and construction of the ALMS prototype instrument
began, tests for benzene detection were carried out on a laboratory
instrument. The instrument was like the one shown in Figure 2 with a
GCA McPherson Model 218 monochromator. The light path in the monochro-
mator was one meter and was purged with flowing argon during the experi-
ments to prevent absorption by oxygen in the vacuum ultraviolet region.
The light path outside the the monochromator was also purged. The 13 cm
cell had a volume of 124 m_L. With high purity quartz lenses and windows
it was possible to extend the usable spectral region to 165 nm by gas
purging. In these experiments the wavelengths were selected manually.
Specially constructed iodine lamps were used. The iodine lines may
be obtained at high intensity by the use of an electrodeless discharge
lamp. The iodine lamp used in this experiment was excited using a one
quarter wave stub cavity powered by a 200 watt microwave diathermy unit
produced by The Burdick Corp., Milton, WI (Model MW/200). The tip of
the lamp extended beyond the cavity and was immersed in a beaker of
water to lessen the effect of temperature changes on the vapor pressure
5
-------
LO CM CTD CM
ni'j uj Ln tt
If) CO O)
px. CO (J)
00 00 00 Is*-
\ \ \ i7 f//
§ CM
rrt
Benzene
Toluene
Ethyl Benzene
c
CC
JO
Isopropyl Benzene
o
w
JO
<
O-xylene
M-xylene
P-xylene
¦a
1600
2200
2000
1800
1400
1200
Wavelength A
XBL 355-11124
Figure 1. Vacuum ultraviolet absorption spectra of benzene derivatives
and superimposed atomic lines [7].
6
-------
Light source Monochromator
L1
Half mirror
L2
Sample cell ^ pmt 1 voltage follower
L4
I
Outlet
t
Inlet
Ar gas inlet
A)PMT 2
Voltage follower
AC amplifier
L-D^A—O
Band pass filter
DC amplifier
Log
amplifier
Strip chart
recorder
XBL 844-9310
Figure 2. Diagrnm of ALMS prototype instrument. L designates lens and
PMT photomultiplier tube.
-------
of iodine in the lamp. The tube was operated continuously for a week at
80 % power and produced intense, neutral lines at a low noise level.
Because this was a continuous, rather than a pulsed, light source; the
ac amplifier and filtering electronics were bypassed in the signal pro-
cessing circuit.
Molar extinction coefficients for benzene were determined at five
iodine wavelengths: 179.91, 183.OA, 184.44, 187.64, and 206.16 nm by
injecting known amounts of benzene vapor in air into the sample cell.
The results are shown in Figure 3. The largest extinction coefficient
was approximately 72,000 at 179.91 nm. This is considerably larger than
the extinction coefficient at the mercury 253.6 nm line of about 100.
The approximate detection limit of benzene at 179.9 nm was found to be
100 ng in a 124 mL cell. This corresponds to a detection limit of 250
ppbv.
PROTOTYPE CONSTRUCTION
A portable TALMS monitor that had been in use at the EPA Environ-
mental Monitoring Systems Laboratory at Research Triangle Park, North
Carolina, was returned to the Lawrence Berkeley Laboratory and converted
to an ALMS instrument. This involved removal of the magnet and squeezer
assemblies and modification of the electronics. A new mercury light
source was installed, and the output was checked to verify that a steady
baseline was produced. With new cell windows and lenses the instrument
can be operated in the vacuum ultraviolet and ultraviolet regions. Pho-
tographs of the optical and electronic units are shown in Figures 4 and
5 respectively.
The prototype ALMS that resulted had a total weight of 58 lbs coin-
pared to a weight of 82 lbs for the benzene TALMS monitor developed pre-
viously. The optical unit weighed 28 lbs and had dimensions:28" x 10" x
12". The electronics unit weighed 6 lbs and had dimensions of 19" x 7"
x 5.25". The lamp driver weighed 24 lbs and had dimensions of 14.5" x
14" x 6.5". Brief descriptions of the design and construction of the
instrument are given below. The operating instructions and circuit
diagrams for the prototype unit are given in Appendices A and B.
Optical Unit
A photograph of the 28 inch long optical unit is shown in Figure 4.
The basic arrangement of the optical unit is shown in Figure 2. A mer-
cury Penray lamp was used with a JY H—10 monochromator for line isola-
tion. A 15 cm sample cell equipped with high quality quartz windows was
mounted on the optical bench. Photoraultipliers were model R128 from
Haraamatsu Photonics.
Electronic Components
A photograph of the electronics module is shown in Figure 5. The
lamp power supply will differ depending upon the lamp used but might
typically provide a discharge that is pulsed at 10 kHz. The photomulti-
plier tube voltage output is buffered with a voltage follower and ampli-
fied. A band pass filter is used to reject noise of frequencies
8
-------
— 8
180
190
Wavelength (nm)
205
XBL 855-11122
Figure 3. Extinction coefficients of benzene at various iodine lines.
9
-------
Figure 4. Photograph of prototype ALMS optical unit
-------
Figure 5. Photograph of front (top) and back of ALMS electronic
module.
11
-------
differing from that of the lamp power supply* The filter provides one
dB attenuation at 4 kHz and at 20 kHz. The filtered ac signal is then
converted to dc, amplified, and fed into a log amplifier (Analog Dev-
ices, 755P), whose output is a measure of absorbance. This output
(absorbance) is continuously displayed on a strip chart recorder.
Schematic diagrams are given in Appendix B.
The prototype instrument was tested for the detection limit of ben-
zene at 184.9 nm before delivery. An approximate detection limit of 250
ppbv was found.
12
-------
SECTION A
TALMS AND ALMS DETECTION OF SUBSTITUTED BENZENES
The effectiveness of TALMS or ALMS detection is largely dependent
on the atonic lines chosen for monitoring. The goal is to maximize the
difference in molecular absorption at two lines while, at the same time,
minimizing the wavelength separation between them. If the wavelength
separation is small, background correction will be more accurate.
Literature searches indicate that many organic molecules possess sharp
absorption features. However, data in the literature are incomplete for
most compounds, nonexistent for others, and sometimes inaccurate.
Atomic line compilations show that lines can be found in any wavelength
region, although these lines are often very weak. Finding matches is a
tedious process because in many literature citations molecular absorp-
tion is not accurately measured, and because quiet, intense light
sources that operate in a magnetic field are not easy to produce. How-
ever, advances in the application of the ALMS and TALMS techniques to
multicomponent determinations at low concentrations depend on the loca-
tion of additional regions of potential TALMS and ALMS signals. There-
fore, searches for TALMS signals in xylenes, m-dichlorobenzene, p-
chlorofluorobenzene, and p-difluorobenzene were carried out. A double
beam TALMS instrument with a 15 cm cell was used.
ATOMIC LINE DATA BASE
One year ago a magnetic tape was received from the National Bureau
of Standards (NBS) containing information on atomic emission lines
listed in the CRC Handbook of Chemistry and Physics. This information
consists of wavelengths and intensities corresponding to a specified
element in a certain ionization state. Sometimes a code referring to
some descriptive feature of the emission line also is included. This
element, wavelength, intensity information is very useful in attempts to
find line matches for both ALMS and TALMS techniques. If these tables
could be entered into a computer data base, rapid searches could be made
for emission lines within designated wavelength intervals. Unwanted
elements and low intensity lines could be screened out in the process.
There is also the possibility of obtaining useful information about the
statistical distribution of emission lines over particular wavelength
regions that would be helpful in deciding what light sources to build.
The NBS data were stored in one hundred files on eight 5.25
inch floppy discs. In general there was only one element per file.
Irrelevant material was eliminated, and the remainder of the data were
processed on an HP 9826 computer with a storage capacity of 4.886 Mbytes
using an HP 9135A Winchester hard disc drive. Only wavelengths in the
13
-------
region from 180 to 500 nm were retained.
Sorting the complete data base in numerical order by wavelength was
accomplished with internal sorting routines written in the extended HP
BASIC language (MAT SORT and MAT REORDER). The radioactive elements
were omitted from the data base since lamp production with these ele-
ments would be difficult.
After the NBS atomic line information had been transferred and
arranged, a search was made for nonrandom wavelength distributions of
atomic lines. To conduct this search, a program was written to count
the number of lines over a specified wavelength interval with a greater
than specified intensity. After a trial and error process, it was found
that between wavelengths of 180 and 500 nm an interval of 0.5 nm and a
minimum intensity of 500 would most clearly reveal any tendency for
atomic lines to cluster in particular regions. However, such clustering
of lines was not found. The number of lines available for matching did
seem to decrease between 410 and 500 nm and between 205 and 180 nm.
ALMS DETECTION OF XYLENES
TALMS detection of p-xylene at the cobalt 252.9run' line was success-
ful in previous studies [5]. However, attempts to detect o-xylene and
m-xylene were not. Hence, methods of using the ALMS technique to detect
these latter components were considered. Our approach to this problem
is to use two wavelengths that are close together (separated by 0.2 or
0.3 nm) and assume that differences in absorption caused by substances
other than the analyte are negligible over this wavelength range. The
ideal situation would be to have two close wavelengths from a single
element whose spectrum is easy to excite.
Spectral data obtained from our medium resolution spectrograph
indicate that o—xylene has a number of intense absorption features
between 267.9 and 269.5 nm, but there are few usable atomic lines in
this region. However, germanium (265.16 and 265.12 nm), lead (261.42
and 261.37 nm), or manganese (257.61 and 257.55 nm) lamps could be used
for this compound.
Absorption spectra of o-xylene reported by Humby et al. [8] also
indicate that platinum (270.24 and 270.59 nm) or iron (271.90 and 272.09
nm) lamps might be used.
The m-xylene absorption features are much broader than those of o-
and p—xylene. In fact, it may be impossible to monitor this molecule
with the TALMS technique. The best possibility for ALMS detection is to
use a platinum lamp which has lines at 270.59 and 270.24 nm.
TALMS DETECTION OF m-DICHLOROBENZENE
Meenakshi and Ghosh have recently published a paper on the high
resolution spectrum of m-dichlorobenzene [9]. The wavelengths of
absorption features are reported, and absorption intensities are divided
into seven categories that range from "very very strong" to "very weak".
Data from this publication were converted to wavelengths in air, and
wavelengths of features with intensities that were "moderately strong"
-------
or greater were compared with those of atomic emission lines. This was
done using the computer program that contained atomic line information
from the NBS magnetic tapes. The results indicated that molecular
absorption-atomic line matches might occur at the platinum (273.396,
262.803 nm); iron (275.014, 273.358, 271.902 nra); iridium (266.479 nm);
or germanium (265.117 nm) lines. Less probable matches were indicated
for. the ruthenium (273.572 nm); zirconium (271.026 nm); or rhenium
(271.547 nm) lines.
Meenakshi and Ghosh determined the intensities of the absorption
features from photographic plates using absorption cells that varied in
path length from 0.5 to 3 meters. This information was of little value
in predicting the change in intensity that would be obtained when this
compound was placed in the cell of the TALMS instrument. In fact, when
the iron and platinum lamps were used, it was found that only a very
small change in intensity resulted from absorption features that were
reported as "moderately strong". For this reason the effort was devoted
to lines listed as "very very strong", "very strong", and "strong".
Platinum, Iron, lead, and iridium light sources were constructed5
but no TALMS signals for m-dichlorobenzene were found with them. A ger-
manium lamp was assembled}and a definite signal was obtained at 269.134
nm with a 15 kG (1.5 T) magnetic field in the perpendicular Zeeman
direction.Under the same operating conditions no TALMS signals were
observed for benzene, p-xylene, toluene, or bromobenzene. Figure 6
displays the TALMS signal obtained with a 15 cm cell.
TALMS DETECTION OF p-CHLOROFLUOROBENZENE
The high resolution spectrum of p—chlorofluorobenzene has been,
reported by Cvitas and Hollas [10]. The wavelengths of sharp absorption
features from this work were compared with those of atomic emission
lines. The comparison indicated a TALMS signal might be obtained with
the iron 275.573 nm line and this was confirmed experimentally. Convert-
ing this wavelength to wavenumbers in vacuum yields the value 36,277.36
cm-*. A representation of the high resolution absorption spectrum of
p-chlorofluorobenzene is presented in Figure 7 with the iron line super-
imposed .
The TALMS signal was obtained using a McPherson 0.3m spectrograph
with a 150 pm slit to isolate the iron line. The TALMS instrument was
used in the dual channel mode with automatic gain control. The TALMS
signal is displayed in Figure 8. It was obtained in the perpendicular
Zeeman direction with a field strength of 15 kG (1.5 T); the intensity
of the signal increased with increasing magnetic field up to 17 kG (1.7
T). The experimental arrangement was tested by inserting an ultraviolet
neutral density filter (~50 % transmission) Into the optical path. No
TALMS signal occurred. Five mL of pyridine vapor at room temperature
were injected into the cell. Again no TALMS signal occurred although
this vapor resulted in 50 % absorption.
Vapor pressure data for chlorobenzene were used to estimate that of
p-chlorofluorobenzene [11]. This will only be approximate,but vapor
pressure data are not available for the chlorofluorobenzenes. The vapor
15
-------
A
M-dichlorobenzene
c
o
P-xylene
TALMS
signal
Time
¦<-
XBL 855-11119
Figure 6. TALMS signal for m-dich.loroberizerie at the Ge 269.134 nm line.
-------
75
I
I 1 1 1
1 Fe 36,277 cm-1 ~
ion
en
o
— /
1 —
Q_
O
w
< 25
n
1
III,
10 0 -10
Wavenurnber (crrT1)
XBL 855-11117
Figure 7. Iron emission line superimposed on the p-chlorofluorobenzene
absorption maximum at 36,277 cm ^ [10].
17
-------
P-chlorofluorobenzene
c
o
TALMS
signal
Optical
filter
4—'
Q.
1_
o
w
.Q
<
Pyridine
Time
XBL 855-11120
Figure 8, TALMS signal from p-chlorofluorobenzene at the Fe 275.6 nm line.
-------
pressure of chlorobenzene as a function of temperature was fit by least
squares to the relationship:
1/T = A + B InP
where A = 3.8596 x 10"3 and B = 2.0695 x 10~4 if T is expressed in
degrees K and P is in torr. At 25 °C, 0.1 mL of a p-chlorofluorobenzene
air mixture produced a signal equal to two times the noise level. This
corresponds to an approximate detection limit of 30 ppmv.
TALMS DETECTION OF p-DIFLUOROBENZENE
A TALMS signal was found for p-difluorobenzene at the platinum
265.945 nm emission line. Aside from the boiling point, no vapor pres-
sure information for this compound could be found in the literature. In
order to construct an analytical curve and to determine the lower limit
of detection, the vapor pressure must be known. Therefore, the experi-
ments described below were carried out to determine a relationship
between vapor pressure and temperature.
The p—difluorobenzene was purchased from the Aldrich Chemical Co.
and was represented as being 99+% pure with a boiling point range of 88
- 89 C. The vapor pressure-temperature relationship was determined by
placing the material in a glass tube capable of being evacuated while at
liquid nitrogen temperature. After evacuation, the liquid nitrogen bath
was replaced with an ice bath, and the pressure was measured using a
Balzers pressure gauge that read directly to 0.1 millibar and could be
estimated to 0.05 millibar. The sample was immersed in a liquid nitro-
gen bath, and evacuation started after the first signs of crystalliza-
tion (pressure = 20 millibars). This was repeated four times to insure
that all of the oxygen and other gases were pumped out of the system
before a vapor pressure measurement was made. The following pressures
were measured at the listed temperatures:
Temperature Pressure
(C) mbar (torr)
19.5 (14.6)
59.75 (44.81)
(760)
The 89 c point was taken from the literature [12]. These data fit the
relat ionship:
log P = a + b (1/T)
where P = pressure (mm of Kg), T ¦ temperature (K), a = 8.185 and b =
-1928.4, with the square of the coefficient of determination = 0.996.
Data for an analytical curve were obtained by drawing off known
volumes of an air and vapor mixture in equilibrium at 0 °C. A glass
syringe was used because previous experiments had shown that plastic
syringes could be a source of contamination. The signal and background
0
24.6
89
19
-------
Levels were averaged for 90 seconds at each point. A ten second time
constant was used. The platinum 265.945 nm line was used in the paral-
lel Zeeman direction with magnetic fields of 12.2 kG (1.22 T) and 16.4
kG (1.64 T). The results are plotted in Figure 9.
When as much as two mL (200 pg) of the vapor were injected into the
cell; there was almost no transmitted light and the signals were dom-
inated by shot noise. The analytical curve was linear up to injections
of about one nL. From the baseline noise of 0.04 v and the analytical
curve, two times the-noise level corresponds to a detection limit of 4
pg or 50 ppmv.
To measure the extinction coefficient of p-difluorobenzene the
ratio of the transmitted light intensity to the incident intensity was
determined using the 265.945 nm platinum line. Defining the extinction
coefficient (<) by the relationship:
I(transmitted) = I(incident) x 10—^ x concentration x pathlength
the value of < was found to be 2200 [cm(mole/L)]—1 at this wavelength.
LINK SHAPE MEASUREMENT OF p-DIFLUOROBENZENE
High resolution line shape measurements of molecular features
responsible for TALMS signals have several important applications.
Because these shapes are characteristic of the molecule, they can be used
to avoid interferences. If the molecular line shape is known, the
optimum field strength and field direction for maximum sensitivity can
be readily chosen. In addition, line shape measurements reveal a funda-
mental property of the molecule that may have future uses that are not
readily predictable.
In previous studies [1] line shape measurements have been made of
benzene and chlorobenzene absorption features. These measurements were
made In the 253.7 nm region using a relatively noise—free mercury
discharge light source. However, line shape measurements made with the
magnetically confined lamps of this study are more difficult because
such discharges tend to be associated with a higher level of noise.
Measurement of the line shape of p—difluorobenzene reported here is
more difficult because the platinum 265.9 nm line must be used and it
can only be generated with magnetically confined lamps.
When transmission of the cF" and cf components of the 265.9 nm line
was compared, the absorption difference was found to be less than two
percent. This provides sufficient information for a TALMS signal
because the automatic gain control can be used to smooth out noise from
the light source. However, when line shapes are determined, the time
elapsed between measurement of the two components precludes use of the
automatic gain control.
A measurement system was devised that circumvents this type of
noise. It takes advantage of the fact that the difference in absorbance
of the o-+ and components is greater the higher the concentration in
the cell. This larger absorbance difference more than compensates for
the fact that a percentage difference of a lower total light intensity
20
-------
2.0
1.8
1.6
12.2 kG
16.4 kG
0 20 40 60 80 100 120 140 160 180 200 220
Amount injected (^g)
XBL 855-11115
Figure 9. TALMS analytical curves for p-difluorobenzene with the Pt
265.945 run line.
21
-------
(70 — 80% absorbed) must be determined.
A double beam arrangement was used that monitored intensity before
and after passing through the cell. Signals from the photomultiplier
tube responding to the reference signal and the one responding to the
sample signal were fed into a log ratio module. Ratio recording at high
absprbance values greatly reduces the influence of light source noise.
The log ratio module includes two integrated circuits, AD 755N and
AD 755P, which according to the manufacturer's specifications are accu-
rate within 1% when they are used for voltage logging between 10 mv and
10 v. The error was found to be much greater than this. The ICs were
checked by imposing known voltages from an Analogic voltage source. It
was found that in the range below 100 mv the 755P was in error by 1 —
1.5% and the 755N was in error by as much as 8%. The 755N was within
specifications in the 0.1 - 1.0 v range. The sample signals were
expected to be less than 100 mv,so they were processed using the circuit
containing 755P. The reference signals were expected to be greater than
100 mv so they were processed by the circuit containing 755N.
A combination of two linear polarizers on either side of a retarda-
tion plate was used to isolate the o3*" and the
-------
indicate the optimum current is 4.0 amperes (14.4 kG). A plot of the
extinction coefficient versus magnet current for the and
-------
2.0 —
-V-'
c
o
o
1—
0
Q.
0.5
0
o
£Z
0
0
0.0
H—
Q
-0.5
-1.0
0
5
2
6
3
1
4
(4.5) (9.5) (12.2) (14.4) (16.4)
Amps (kG)
XBL 855-11125
Figure 10. Percentage difference between a+ and a sbsorbance of
p-difluorobenzene at Pt 265.9 nm vs. magnetic field current.
24
-------
2000
1900
JJ3
o
E
E
o
1800
1700
_ l_ i J ,_l _I _.i l_ I I L_j
543210 12345
(16.4) (14.4) 12.2) (9.5) (4.5) (4.5) (9.5) (12.2) (14.4) (16.4)
Amps (kG)
XSL85S-111I3
Figure 11, Molar extinction coefficient of p-difluorobenzene vs. magnetic
field strength for the o+ and cr Zeeman components at 265.9 ran.
25
-------
REFERENCES
T. Hadeishi, H. Koizumi, R. D. McLaughlin, and J. E. Millaud, "Tun-
able Atomic Line Molecular Spectrometer", Spectrochim. Acta 37B, 501
( 1982).
2. T. Hadeishi, R. D. McLaughlin, J. G. Conway, and D. R. Scott,
"Selection of Atomic Emission Lines for Tunable Atomic Line Molecular
Spectroscopy of Benzene", Anal. Chem. _54_, 1517 (1983).
3. R. Berkley, D. R. Scott, and R. Hedgecoke, "Performance Optimization
of TALMS", American Chemical Society Meeting, Miami, Florida, April,
1984.
4. T. Hadeishi, R. McLaughlin, and J. Millaud, "Development of a Tun-
able Zeeman Spectrometer for Analysis of Toxic Organic Compounds", EPA—
600/S4-82-067, U. S. Environmental Protection Agency, Research Triangle
Park, North Carolina, 1982, 55 pp., NTIS # PB-83-139535.
5. T. Hadeishi, R. McLaughlin, J. Millaud, and M. Pollard, "Development
of a Continuous Monitor for the Detection of Organic Compounds", EPA.
600/4-83-034, U. S. Environmental Protection Agency, Research Triangle
Park, North Carolina, 79 pp. ,NTIS # PB-83-234922.
b. M. B. Robin Higher Excited States of Polyatomic Molecules Vol.
II, Academic Press, New xork, m, 1975. ' ' ~
7. L. W. Pickett, M. Muntz, and E. M. McPearson, "Vacuum Ultraviolet
Absorption Spectra of Cyclic Compounds. I. Cyclohexane, Cyclohexene,
Cyclopentane, Cyclopentene, and Benzene", J.Am.Chem.Soc .7.3,4862 ( 1951).
g. C.M. Humby, G.P. Semeluk and R. D. S. Stevens, 'The Vapor Phase UV
Spectra of Ortho Fused Benzocycloalkenes", Spectros. Lett., _3,99
(1970).
g_ A. Meenakshi and D. K. Ghosh, "276 nm Absorption System of m-
Dichlorobenzene: Vibrational Fine Structure Analysis", J. Molec. Spec-
tros., 103, 208 (1984).
1q. T. Cvitas and J. M. Hollas, "Rotational Band Contour Analysis in
the 2760 & Svstem of p-Chlorofluorobenzene", Molec. Phys, _18_ (2)» 261
( 1970).
11. R. C. Weast, Ed, CRC Handbook of Chemistry and Physics 57th Ed.,
CRC Press, Cleveland, GK;—1975"
12. Editorial Board , Dictionary of Organic. Compounds , Oxford Univer-
sity Press , 19 65 .
26
-------
APPENDIX A
ALMS INSTRUCTION MANUAL
ASSEMBLY INSTRUCTIONS
1. Mount the monochromator on the optical bench. Remove the tape
holding the 0.25 mm slits in the monochromator. The 0.05 mm slits
are in a plastic bag with this manual. Slits for the H-20 will
also fit in this monochromator.
2. The mercury Pen-ray lamp is taped to the optical bench. Remove the
tape.
3. Insert the mercury lamp into the Cajon fitting in the lamp housing.
Push it in as far as it will go. Mount the lamp on the monochroma-
tor at the end labeled "lamp". The top of the lamp housing screws
off so the lamp can be oriented properly.
4. Remove the tape from PMT1. Do not lose the 0-ring.
5. Mount the beam splitter, the sample cell, and the PMT1 housing to
the optical bench as shown.
6. Connect the beam splitter to the monochromator.
7. Slide the sample cell into the beam splitter and the PtCTl housing
over the other end of the sample cell. Slide them together as far
as possible. Light leaks can occur at these connections.
8. Remove the tape from PMT2. Do not lose the 0—ring.
9. Mount the PMT2 housing on the beam splitter.
10. There are five wires from each PMT tube buffer. The wires from
PMT2 are numbered 1, 2, 3, 4, and 5. The wires from PKT1 are num-
bered 6, 7, 8, 9, and 0. Attach these wires to the corresponding
numbers on the bus bar mounted on the wall of the optical bench.
The optical bench is now assembled.
ELECTRONICS MODULE
Each card of the electronics module must be inserted into the
proper slot. The connectors at the back of the bin are not the same.
Look to see where each card is plugged into the bin before removing it.
To remove the power supply the screws on the front and back must be
removed. This module does not slide out freely. (The wires must be
27
-------
disconnected before the module will completely slide out of the bin.)
The other three modules, (the PMT power, signal processing, and time
constant modules) will slide out of the bin freely. An extender card
has also been provided to assist in repairs of the modules if needed.
OPERATING MANUAL FOR ALMS
The prototype ALMS instrument consists of an optical bench, a lamp
driver, and an electronics unit. The electronics unit contains a DC
power supply, photomultiplier tube power supply, signal processing
module, and a time constant module. Circuit diagrams for all four of
these modules are provided in Appendix B.
1. Power to the ALMS instrument unit is turned on by plugging the unit
into a wall socket. This power module runs the electronics unit
and the output buffer amplifiers on the phototubes on the optical
bench. Power is supplied from the electronics unit to the photo-
tube output buffers through the gray three conductor cable that is
provided with the instrument. This cable plugs into the back of
the DC power supply in the electronics unit and into the three pin
connector on the optical bench.
2. The PMT power supply module supplies power to the photomultiplier
tubes. The outputs are at the back of the electronics unit. Chan-
nel 1 and channel 2 refer to PMT 1 (signal) and PMT 2 (reference)
and are marked on the module and the optical bench. Connect chan-
nel 1 and channel 2 on the electronics unit to the optical bench
using the two BNTC cables provided. The PMT power supply signal is
a 0 — 7.4 volt DC voltage. The high voltage DC to DC converter is
on the phototubes on the optical bench so there is no high voltage
on the two PMT power supply cables.
3. The signal processing module takes the PMT output buffer signals
from both phototubes and puts them into a log ratio amplifier to
get an absorbance signal. The outputs from the PMT buffer amplif-
iers on the optical bench should be connected to the PMT 1 and PMT
2 inputs on the back panel of the signal processing module using
the two BNC cables provided. The absorbance signal can be moni-
tored on the log output on the rear of the signal processing module
or the absorbance output on the front panel of the module. The
phototube signals can be monitored at the PMT 1 and PMT 2 output
connectors at the rear of the electronics unit.
When the front panel switch on the signal processing module is in
the up position,the absorbance output is the log ratio of the chan-
nel 1 and channel 2 inputs.
When the front panel switch is in the down position the input from
channel 2 is cut off and replaced by a one volt internal reference
signal. The ALMS instrument should be operated with the front
panel switch in the up position.
There is a log adjust screw on the front panel of the signal pro-
cessing module. This is used to set an offset in the log ratio
amplifier. Normally this should not be touched. The proper way to
28
-------
set the log adjust is to put the same signal into the PMI i and PMI
2 inputs. The absorbance output should be zero. If it is not
zero, then turn the log adjust screw until the absorbance output is
zero.
4. The lamp for the prototype ALMS is a standard mercury pen ray lamp.
It is mounted in the housing at the front of the monochromator.
There is access to the lamp at the top of the housing for visually
orienting the lamp for proper operation. Connect the separate lamp
driver into the blue Pomona box mounted on the optical bench using
the high voltage BNC cable provided.
5. The ALMS signal is simply an absorbance measurement from the two
photomultiplier tubes. With no sample in the cell, set both POT
outputs to the same value so that there is a zero absorbance out-
put. Set the photomultiplier tube outputs to the 1-2 volt range.
6. The output of the signal processing module is the input to the time
constant module. Turn the knob on the front panel of the time con-
stant module counterclockwise to its final position. The output
now has a one second time constant. The next positions, in a
clockwise direction, are 10, 20, 30, and 50 second time constants.
The zero button zeros the output by discharging the capacitors in
the time constant amplifier.
The ALMS instrument is now ready to operate.
29
-------
APPENDIX B
ELECTRONIC SCHEMATICS
This appendix contains schematics of the ALMS electronic modules
that were sent to the U. S. Environmental Protection Agency at Research
Triangle Park, NC, during the month of December, 1984.
30
-------
3 A
21.1 V
7 ABC
4000
33 K
1 ABC
4000
33
4 ABC
+ 15V
5 ABC
-15V
6 ABC
7915
7815
XBL 855-11126
Figure 12. ALMS power supply.
-------
© 1
^ V
0 + 15 V
0 15 V
10 K
PMT
#1
10 M
10 K
j Wv
PMT
10 M
PMT #2 "on
FP switch
PMT #2 "off" and I V Ref. "on
RP Absorbance
—(14.
Figure 13
10 nf
PMT
Mon #1
AD 757 N
Log Amp
• 10 K
PP
[_ PMT
Mon #2
+ 15 V-
10 K
51 K
—vw-
+ 15 V
->£20 K
--15 V
2N2219
£_20kJ
-rFP
Labsorbance
ALMS logarithmic amplifier.
XBL 855-11127
-------
Ch 1
out
7.4 V
max
Ch 2
out -
d>
G>
-©—
1 ^ 4000
10 Q
———
10 Q
—vw—
f 15 V
5.6
K
F.P.
10 K
2N5192
l/
V
2N5192
5.6 k:
Fp
10 K
[/
k
3
V
I
\7
:iMf
/
2N2219
f
1 uf
2N2219
XBL 855 11121
figure 14, ALMS photomultiplier tube power supply.
33
-------
10^F
5mF
3 /iF
0.1 mf /
10K
10K
.en 10K
nput gj—mv
n output
741
10 K
503
10 K
Button cj |
Electro!
RA30382121
XBL 855-11116
Figure 15. ALMS time constant module.
34
-------
' TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completingj
1.REFOFTNO. 2.
EPA/600/4-85/043
3. RECIPIENT'S ACCESSION NO.
?3S 5 2 2362r^B
4. TITLE AND SUBTITLE
Development of an Optical Monitor
for Toxic Organic Compounds in Air
5. REPORT DATE
June 1985
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
T. Hadeishi, M. Pollard, R. McLaughlin, M. Koga
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Lawrence Berkeley Laboratories
University of California
Berkeley, CA 94720
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
IAG No. -DW 930479-01
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. Environmental Protection Agency
Office of Research & Development/EMSL/MDAD
Research Triangle Park, N. C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final 10/ft.Vm/M
14. SPONSORING AGENCY CODE
EPA/600/Q8
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The objectives of this study were: a) to design, construct, and deliver a proto-
type atomic line molecular spectrometer (ALMS) benzene monitor and b) to locate
matches of atomic lines and sharp molecular absorption features in other toxic organic
compounds for possible use in ALMS or TALMS techniques. ALMS and TALMS are newly
developed, high resolution molecular absorption techniques which are used in the
vacuum untraviolet and ultraviolet regions of the optical spectrum to detect organic
molecules in the gas phase. The dual beam prototype ALMS instrument was designed,
constructed, tested and delivered to the Environmental Monitoring Systems Laboratory,
U. S. EPA, Research Triangle Park, North Carolina, in December, 1984. It was designed
for monitoring benzene and other organic compounds with the 184.9 and 253.7 nm mercury
lines. The instrument consisted of three units: the optical unit (weight: 28 lbs,
dimensions: 28x10x12"); the electronics unit (weight: 6 lbs, dimensions: 19x7x5.25");
and a lamp driver (weight: 24 lbs, dimensions: 14.5x14x6.5"). The total weight was
58 lbs. which is less than that of the TALMS benzene monitor previously developed
(82 lbs). Tests of the performance of the benzene monitor showed an approximate
detection limit of 250 ppbv at 184.9 nm. Line matches and TALMS signals were found
for three new compounds: p-difluorobenzene (Pt:265.9 nm); m-dichlorobenzene (Ge:
269.1 nm) and p-chlorofluorobenzene (Fe:275.6 nm).
i7. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.lDLNTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
21. NO. OF PAGES
41
20. SECURITY CLASS /This page)
22. PRj.CE
EPA Form 2220-1 (Rev. 4—77) frevious edition is obsolete
------- |