EPA-650/2-75-069
June 1975 Environmental Protection Technology Series
REACTIVE HYDROCARBON ANALYZER
FOR MOBILE SOURCES
U.S. Environmental Protection Agency
Office of Research and Development
Washington, 0. C. 20460
-------
EPA-650/2-75-069
CHEMILUMINESCENT
REACTIVE HYDROCARBON ANALYZER
FOR MOBILE SOURCES
by
A. Fontijn, H. N. Volltrauer, R. Ellison
Aero Chem Research Labs, Inc.
P. O. Box 12
Princeton, New Jersey
Contract No. 68-02-1224
ROAPNo. 26ACV
Program Element No. 1A1010
EPA Project Officer: P.M. Black
Chemistry and Physics Laboratory
National Environmental Research Center
Research Triangle Park. North Carolina 27711
Prepared for
U. S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, D. C. 20460
June 1975
-------
EPA REVIEW NOTICE
This report has been reviewed by the National Environmental Research
Center - Research Triangle Park, Office of Research and Development.
EPA, and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environ-
mental Protection Agency, have been grouped into series. These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These 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
9. MISCELLANEOUS
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to
develop and demonstrate instrumentation, equipment and methodology
to repair or prevent environmental 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 for sale through the National
Technical Information Service, Springfield, Virginia 22161.
Publication No. EPA-650/2-75-069
ii
-------
TP-319a
ABSTRACT
A chemiluminescence method for measuring total reactivity of hydro-
carbon (HC) mixtures has been developed and a prototype analyzer based on
this method has been built. The difference between the OH(A2S2-X2n) emis-
sion intensities at 308.9 and 312.2 nm from O-atom/hydrocarbon reactions
near 1 Torr is measured. For C^H* 1308-9 >:> *3iz.z> *or C2H2 *308.9 * *3iz.2« The
other hydrocarbons tested yield the same spectral distributions as C2H4; CH4
yields no emission. Two PMTs are used for 308.9 and 312.2 nm measure-
ment respectively. When the apparatus is zeroed, the difference in signal
from the two PMTs is insensitive to C2H2. The relative response to the
individual reactive HC species can be set to give good agreement with reac-
tivity ratings. The response to HC mixtures is additive. A limit of sensi-
tivity of * 0.05 ppm C2H4-equivalent HC and a linear response to individual
HCs to 2500 ppm is obtained; greater sensitivity appears feasible. CO, CO2,
SO2, CH4, C2H2 and NOX do not interfere with instrument response. A 1%
Change in [ O2] causes < 1% change in signal; 3% H2O causes & 12%decrease.
This report was submitted in fulfillment of Contract 68-02-1224 by
AeroChem Research Laboratories, Inc. under the sponsorship of the
Environmental Protection Agency. Work was completed as of 12 June 1975.
-------
TP-319a
CONTENTS
Page
Sections
I Conclusions *•
II Recommendations 2
III Specifications of Prototype Instrument 3
IV Introduction *
Part I: Feasibility Study
by Arthur Fontijn and Roy Ellison 7
V Survey Study 8
A. Approach 8
B. Experimental 1®
C. Investigation of Emissions Other than OH(Aa£-Xan) 13
1. CN(BaE-X*2) 13
2. N0(0,y) and CO Cameron Emissions 1*
D. OH(A2E-XaJl) from 0-Atom Reactions 15
1. Spectral Distribution and Intensities 15
2. Reaction Kinetics and Kinetic Spectroscopy 19
VI The Chemiluminescence Hydrocarbon Analyzer Test Apparatus
Experiments 22
A. Experimental 22
1. Gas Handling System and Flow Conditions 22
2. Radiation Measurement 24
3. Electronics 2*
B. Results and Discussion 26
1. Acetylene Zero 26
2. Response to Individual Hydrocarbon Species 26
3. Interference by Other Species Present in Mobile
Engine Exhausts; Influence of Variable % Oa 30
4. Measurements with Other Reagent Gas Compositions 34
5. Hydrocarbon Mixtures 35
Part II: Manual for Prototype Analyzer
by Hermann N. Volltrauer and Arthur Pontijn
40
VII Technical Description
IV
-------
TP-3l9a
Page
VII A. Gas Handling System 40
1. Sample Gas System 40
2. Reagent Gas System 40
3. Reactor and Vacuum Pump 46
B. Light Detection System 47
C. Electronics 47
1. Measuring Electronics 47
2. Microwave Power Supply 49
VIII Operating Instructions 52
A. Descriptions of Panels and Controls 52
1. Front Panel 52
2. Rear Panel 53
B. Installation and Set-Up 53
C. Operation ^4
D. Calibration 55
IX Maintenance and Troubleshooting 57
A. Planned Maintenance 57
B. Disassembly and Reassembly 57
1. Cabinet 57
2. PMT Housing 58
3. Sample Inlet Nozzle 59
4. Reactor 59
5. Vycor Discharge Tube 60
6. Valves 60
7. Microwave Power Supply 60
C. Troubleshooting **•
D. Signal Processor Board Adjustments 63
E. Shipping 6A
X Tests 65
A. Reagent Gas Purification Traps 65
B. Collimators ^
C. Flow Rates ~
D. Nozzle Distance and Microwave Power oo
E. Linearity and Limit-of-Sensitivity 66
F. Response Ratios 66
XI References 69
XII Inventions and Publications 72
APPENDIX A: List of Manufacturers of Parts Used in Prototype
Instrument
73
-------
FIGURES
No. Page
1 Schematic of Conventional Fast Flow Reactor. 11
2 OH(A2E-X2n) Emission from the 0/C2H<, and 0/C2H2 Reaction. 18
3 1308.9 versus C2Hi, Flow. 18
4 Influence of 02 on 308.9 nm Intensities from 0/C2Hi, and 0/C2H3. 21
5 The Chemiluminescence Hydrocarbon Analyzer Test Apparatus. 23
6 Diagram of Electronics for the Hydrocarbon Analyzer Test Apparatus. 25
7 Exponential Dilution Plot for Propadiene. 27
8 Response versus Concentration of Ethylene and n-Butane with He/9%
02 Reagent Gas. 28
9 Absence of Interference by Methane. 3!
10 Absence of Interference of NOX with n-Butane Measurement. 32
11 Effect of 02 in Sample Gas on AI30e.9 Response for 12.5 ppm C2H<,. 33
12 Response versus Concentration of Ethylene and n-Butane with
Undiluted 02 Reagent Gas. 34
13 Comparison of Response of Ethylene and n-Butane to that of Their
Mixture at the Same Individual Concentrations. 37
14 Front View. 41
15 Rear View. 42
16 View with Side Panel Folded Down. *3
17 Top Inside View. 44
18 Side View with Cover Off. 45
19 Signal Processor Circuit *°
20 Microwave Power Supply Wiring. 50
fi9
21 Chassis Wiring. °*
22 Response of Ethylene, Iso-Octane and Benzene as a Function
of Nozzle Distance and Microwave Power. 67
vv
-------
No.
TABLES
Page
Relative Chemiluminescence Signal Intensities of Class V, IV,
III, and I Hydrocarbons, at High Hydrocarbon Concentration in
Survey Apparatus 17
Effect of Nozzle Distance on AI308t» for He/92! 02 Discharge
Gas and 1250 ppm Hydrocarbon 27
Relative Chemiluminescence Signal Intensities of Class V, IV,
III, and I Hydrocarbons Obtained at 1.5 * 1(T2 sec Reaction
Time for He/9% Oa Discharge Gas 29
Concentration Ranges of Hydrocarbons (in ppm) for which Inter-
ference was Investigated at the Indicated Interference Gas
Concentration 31
Changes in &I30e .9 from 125 ppm C2H<, and n-C«H10 as a function
of 0-Atom Concentration for He/9% 02 Reagent 36
Prototype Analyzer Relative Chemiluminescence Signal Intensities
of Class V, IV, III, and I Hydrocarbons at Microwave Power Levels
of 20% (A) and 30% (B) of Full Scale 68
vn
-------
TP-319a
ACKNOWLEDGMENTS
We are grateful to the contract monitor, F.M.. Black, and to J. Sigsby
of EPA for many helpful suggestions in the course of this work. We also had
an illuminating discussion with A.P. Altshuller and B. Dimitriades of EPA
on the concept of hydrocarbon reactivity.
We thank P. J. Howard for assembling the prototype instrument.
J. Rose assisted with some of the survey experiments.
Vlll
-------
TP-319a
SECTION I
CONCLUSIONS
This work demonstrates that instruments based on the AIaos.9 method ,
here developed give readings of total hydrocarbon reactivity closely approxi-
mating those which can be obtained by measuring each hydrocarbon individu-
ally with a gas chromatograph and multiplying its concentration with the
appropriate .reactivity factor. Since a ^Iao8.9 instrument, at continous flow,
gives a continuous direct reading, far less time and effort is involved in using
apparatus based on this method.
In this work emphasis was placed on approaching the Dimitriades reac-
tivity scale. However, the response to the individual reactive hydrocarbons
can be adjusted by varying (l) the relative sample and reagent gas flow rates,
(2) the O-atom concentration (by varying the microwave discharge power) or
(3) the reaction time (by adjusting the hydrocarbon nozzle to PMT distance
and/or changing the average gas velocity). Such adjustment to the desired
reactivity scale is possible due to the differences in the rate coefficients of
the reactions of O atoms with the different hydrocarbons as well as the hydro-
carbon fragments formed in the initial O atom attack. Thus a response where
all reactive hydrocarbons are weighted roughly equal would seem possible.
Such an instrument setting becomes of interest when rural smog situations
have to be considered.
By making the 1303.9/^312.2 ratio unity for reactive hydrocarbons, acetylene
can in principle also be measured. This would require minor modifications
of the prototype instrument. However, the O -atom/acetylene reaction has
many strong emission features other than those used in the present instrument.
Some of those emissions appear much more suitable for an instrument built
specifically for acetylene monitoringt
-------
TP-319a
SECTION II
REG OMMENDATIONS
The prototype instrument delivered is suitable for measurement of ethyl -
ene equivalent hydrocarbon concentrations from 0.05 ltd 2500 ppm. This
range is adequate for mobile source work. In the future a monitor for ambient
reactive hydrocarbons, i.e. a monitor of greater sensitivity may be required.
The same basic principle as used for the present instrument can very likely
be employed to yield an instrument with a sensitivity on the order of 10 ppb or
better by making the following modifications:
1. Using lenses to increase the light-gather ing capability of the photo-
multiplier tubes (PMTs) monitoring the reaction zone.
2. Using quartz (instead of Pyrex) viewing ports, and PMT envelopes.
3. Using a larger reactor to cover the entire PMT face.
4. Optimizing the reactor design using information obtained from the
present instrument and the literature on O-atom/hydrocarbon
reactions.
5. Optimizing filter and collimator selection by a careful analysis of
the O/hydrocarbon and background spectra near 310 nm.
Item 5 may decrease the background signal relative to the sample signal
while the others are expected to increase both signals proportionally. Items
1, 2 and 3 can result in a factor of 10 to 15 more light reaching the PMTs.
Since the noise in the background (the sensitivity limiting factor) is mainly
due to the statistical nature of the signal, it will increase approximately as
the square root of the signal. A factor of 10 to 15 more signal can therefore
result in an increase of 3 to 4 in sensitivity. Items 4 and 5 will undoubtedly
result in some improvement in sensitivity but the amount cannot be quantita-
tively assessed.
-------
TP-319a
SECTION III
SPECIFICATIONS OF PROTOTYPE INSTRUMENT
SENSITIVITY (ethylene equivalent) 0. 05 ppm (S/N s 2)
USEFUL, CONCENTRATION RANGE 0. 05 to 2500 ppm
FLOW RATES: SAMPLE Iml.atmsec'1
REAGENT 3 ml.atm. sec'1
PRESSURE 1.6 Torr
OPERATING RANGES Eight full scale ranges: 1, 2.5, 10,
25, 100, 250, 1000, 2500 ppm
TIME CONSTANTS 0.1, 0.3. 1, 3. 10 sec.-
LINEARITY Linearized to within 3 % of full scale
TEMPERATURE Operates at room temperature
READ OUT 15 cm analog meter on front panel with
adjustable recorder output of 0-1 V
LINE VOLTAGE RANGE 105-125 V
POWER REQUIREMENTS (exclusive 115 V, 60 Hz, 400 W
of vacuum pump)
EXTERNAL CONNECTIONS A source of O2 (10% in Ar) and a
vacuum pump
VACUUM PUMP 150 1/min capacity, Welch Duo-Seal
1402B-01 or equivalent
SIZE 51 X 56 X 67 cm
WEIGHT 70 kg (150 Ibs)
-------
TP-319a
SECTION IV
INTRODUCTION
The formation of photochemical smogs requires both NOX and organic
pollutants, particularly hydrocarbons. However the ability of hydrocarbons
to produce smogs varies greatly depending upon the rate coefficients of the
many reactions involved (particularly attack by OH, O3, and O).1'2 For ex-
ample, methane and acetylene are considered to be 'unreactive, ' while eth-
ylene and n-butane are 'reactive' hydrocarbons in terms of photochemical
smog. However, severalfold higher n-butane than ethylene concentrations
are required to produce similar atmospheric effects. Current U.S. hydro-
carbon standards are based on total non-methane hydro'carbons, the meas-
urement of which requires several steps, e.g. a flame ionization total hy-
drocarbon analyzer in combination with an infrared methane monitor or a
gas chromatographic methane separator .3 Alternatively a complete gas
chromatographic analysis gives detailed information but is very time con-
suming. There is, accordingly a need, especially in mobile'engine exhaust
monitoring, for a measurement method that gives an indication of total hy-
drocarbon reactivity, i.e., gives the sum of the concentrations of each
hydrocarbon species multiplied by an individual or group reactivity factor.
Since the reactivity of hydrocarbons differs with respect to several para-
meters, e.g. butane contributes more heavily in late or downwind than in
early smog situations, any hydrocarbon reactivity scale is of necessity
somewhat arbitrary (compare Ref. 4); nonetheless a meaningful measure-
ment technique should be approachable. The present work was undertaken
to establish such a technique.
In Part I an initial survey study (Section V) of chemiluminescence in
reactions of hydrocarbons with O and N atoms and mixtures thereof suggested
one suitable method, i.e. the difference in OH(A2 S -X2n) intensities, from O-
atom reactions, at 308. 9 nm and 312.2 nm. This work was followed byanexten-
1. Niki, H., Daby, F..E. and Weinsfock, B., "Mechanisms of Smog Reactions,"
in Photochemical Smog and Ozone Reactions. Advances in Chemistry Series
113, American Chemical Society, Washington, DC, 1972, p. 16-57.
2. Leighton, P.A., Photochemistry of Air Pollution, Academic Press, New York,
1961.
3. Coloff, S.G., Cooke, M., Drago, R.J. and Sleva, S.F., "Ambient Air
Monitoring of Gaseous Pollutants," American Laboratory, July 1973, 10-22.
4. Proceedings of the Solvent Reactivity Conference. EPA-650/3-74-010,
November 1974.
-------
TP-319a
sive feasibility study of this "AI308.9" technique (Section VI). Since this
study showed that this method gives OH intensities in reasonable propor-
tion to the desired hydrocarbon reactivity ratings5 and has the sensitivity
required for mobile exhaust monitoring, the construction of a prototype
instrument for delivery to EPA was undertaken. This instrument
is discussed in Part 2.
5. Dimltriadas, B., "The Concept of Reactivity and Its Possible Applications
in Control," Ref. 4, p. 13.
-------
TP-3l9a
(THIS PAGE IS BLANK)
6
-------
TP-319a
PART I
FEASIBILITY STUDY
by
Arthur Fontijn
and
Roy Ellison
-------
TP-319a
SECTION V
SURVEY STUDY
A. Approach
The objective of our initial effort was to find a prominent emission
feature resulting from a reaction of C2H4 that would not occur to a measurable
degree with C2H2 under the same reaction and observation conditions. The
sllO-900 nm spectral region was investigated since it lends itself to study
with common photomultiplier tubes (PMT), which is the most suitable type
of detector for measurement of low light levels, i.e. measurement of emis-
sion from trace concentrations. Once such a feature was found we checked
whether other reactive, i.e. photochemical smog reactive, hydrocarbons
would also yield the same emission as C2H4.
In view of the available time and funds it was decided to concen-
trate on reactions of O and N atoms which, on the basis of our past experi-
ence and the literature, offered the best prospects for yielding a suitable
emission feature, t The other alternatives considered are:
(i) Q3 Reactions. These can readily be handled at near atmos-
pheric pressure which is an advantage in a practical in-
strument. The successful operation of the Nederbragt
type detector (C2H4/O3) for O3 suggests that this would
be a suitable approach for C2H4 measurement. In view
of the fact that other olefins can be used to replace
C2H46"8 this route might be a promising one for an olefin-
specific detector. However, it is commonly thought
that saturated hydrocarbons would not yield any emission
with O3. Since this has apparently not been explicitly
investigated it may remain worthwhile to further explore
O3 reactions.
^ C2H2 and C2H4 are the only two hydrocarbons whose chemiluminescence
had extensively been studied in the past.
6. Finlayson, B.J., Pitts, J.N. and Akimoto, H., "Production of Vibra-
tionally Excited OH in Chemiiuminescent Ozone-Olefin Reactions," Chem.
Phys. Lett. .12, 495-498 (1972).
7. Rummer, W.A., Pitts, J.N. and Steer, R.P., "Chemiiuminescent Reactions
of Ozone with Olefins and Sulfides," Env. Sci. Techn. 5_, 1045-1047 (1971).
8. Hodgeson, J.A., McClenny, W.A. and Martin, B.E., "Environmental Protec-
tion Agency, Private communications 1973 and 1974.
-------
TP-319a
(ii) OH Reactions... Little is known about chemLluminescence
in OH reactions. Since OH is the major reactive com-
pound in photochemical smog, its chemiluminescent re-
actions might well correlate with hydrocarbon reactivity.
The study of such reactions remains of interest, but
could be complicated by emission from O-atom reactions
since OH radicals react rapidly -with each other to produce
O-atoms via9"10
OH + OH — H20 + O (1)
C2H2 is known not to produce chemiluminescence with
OH.11
(iii) Halogen Atom Reactions. Very few chemilumine scent
reactions of Cl, Br and I have ever been observed. F
atoms, which are more difficult to handle than O or N
atoms (but not prohibitively so), produce a number of
chemiluminescent reactions.12
(Lv) H-Atom Reactions^. H-atom/hydrocarbon reactions are
apparently not chemiluminescent. '
9. Wllaon, W.E. Jr., "A Critical Review of the Gas-Phase ^f^ion Kinetics
- of the Hydroxyl Radical," J. Phys. Chem. Ref. Data 1, 535-573 (1972).
10. Del Greco, F.P. and Kaufman, F., "Lifetime and Reactions of OH Radicals
in Discharge Flow Systems," Disc. Faraday Soc. 33, 128-138 (1962).
11. Bayes, K.D. and Jansson, R.E.W "The Origin of ™&™"™ **.***.
Atomic Hydrogen-Acetylene Flame," Proc. Roy. Soc. A282, 275-282 (1964).
12. Schatz, G. and Kaufnan, M. , "Chemiluminescence Excited by Atomic Fluorine,"
J. Phys. Chenu 76, 3586-3590 (1972).
13. Gaydon, A.G., The Speccroscoov of Flames. John Wiley, New York, 1957,
p. 252.
-------
TP-319a
B. Experimental
A Pyrex fast-flow reactor, Fig. 1, available from previous
studies of hydrocarbon chemiluminescent reactions, e.g. Refs. 14, 15
was used. With a mechanical vacuum pump such reactors are suitable for
O- and N-atom reaction studies involving pressures from about 0.5 to 20
Torr (1 Torr = 133.3 Pa). The reaction tube is 2.2 cm i.d. Carbon com-
pounds are introduced via a traversable nozzle into a gas stream containing
either O atoms, N atoms, or a mixture of the two. The resulting chemilu-
mine scence was observed in a direction perpendicular to the flow, through
a LiF window downstream from the nozzle. The nozzle is adjustable to
allow variation of reaction time, t; under otherwise steady flow conditions
t is proportional to the nozzle to (observation) window distance. The effect
of O2 addition was studied by adding O2 through an 'additive gas1 inlet.
Atomic O was produced either by passing O2 or mixtures of O2
and Ar or He through a microwave discharge (2450 MHz) or, as shown in
Fig. 1, but only occasionally used in this study, by passing N2 (or N2 + He
or Ar) through the discharge, followed by titration of the resulting N atoms
via
N + NO — O + N2 (2>
The advantages of the latter method are easy measurement of the O-atom
concentration, [O], and production of O atoms free from Oz. The former
method however yields larger quantities of O atoms. [O] in the case of
the O2 or O2/Ar,He discharges was measured by comparison to the O/NO
chemilumine scence intensity from over-titrated (excess NO) N/NO mixtures,
as described by Fontijn and Lee.16 In the N-atom studies no NO was added.
To produce N/O mixtures, flows of NO less than those of N were added.
14. Fontijn, A., Ellison, R. , Smith, W.H. and Hesser, J.E., "Chemiluminescent
Emission of CO Fourth Positive Bands in Nitrogen Atom/Oxygen Atom/Reactive
Carbon Compound Systems. Relation to Chemi-Ionization," J. Chem. Phys.
£3, 2680-2687 (1970).
15. Fontijn, A. and Johnson, S.E. , "Mechanism of CO Fourth Positive VUV Chemi-
luminescence in the Atomic Oxygen Reaction with Acetylene. Production of
'D)," J. Chem. Phys. 59., 6193-6200 (1973).
16. Fontijn, A. and Lee, J., "Comparison of the Absolute Quantum Yields of
the Gas-Phase O/NO Reaction and the Liquid-Phase Luminol Oxidation Chemi-
luminescence Standards," J. Opt. Soc. Am. 62, 1095-1098 (1972).
10
-------
TP-319*
He+N,
CARBON »»
COMPOUND
-MICROWAVE
CAVITY
( ADDITIVE
GAS
ACTION
TUBE
TO
MANOMETER
TO
PUMP
M)P
ENING
FIGURE 1. Schematic of conventional fast flow reactor.
All permanent gases used were c.p. grade or better. N2, O2, He
and Ar, were passed through activated alumina (drying agent). The gaseous
hydrocarbons were dried by passage through silica gel. NO was passed
through silica gel and "Ascarite" NO2 absorbent. The normally liquid hy-
drocarbons (Fisher certified grade toluene, n-heptane, iso-octane and
benzene) were placed inside a flask submerged in a constant temperature
bath. N2 introduced through a fish tank bubbler passed through the flask:
the hydrocarbon content of the saturated N2 thus obtained was calculated
using the data of Stull.17 Flows were regulated with fine control needle
17. Stull, D.R., "Vapor Pressure of Pure Substances. Organic Compounds,"
Ind. Eng. Chem. 39, 517-540 (1947).
11
-------
TP-319a
valves and measured via an equivalent oil flow in a partially submerged
buret.18 The lowest hydrocarbon flow rates measurable in this manner are
alO~3 ml.atm sec"1, while the practical total flow rates ranged from si to
10 ml.atm sec"1. It follows that the lowest hydrocarbon concentrations
that could be measured with this apparatus were in the range 100 to 1000
ppm, corresponding to the approximate upper limit of interest. Hence the
apparatus used in the Section V experiments could not as such have been
used for demonstrating feasibility, but only to indicate which features might
be worth exploring further.t Pressures were measured with a Barocel
electronic manometer calibrated against a U-tube oil manometer equipped
with a micrometer depth gauge.
A Czerny-Turner type 0.5m Minute man Model 305-MHA mono-
chromator with an f/6.9 aperture ratio, equipped with a Centronic Q4242BA
bialkali PMT mounted in a thermo-electrically cooled housing was used. In
the studies below «190 nm a magnetic (photo) electron multiplier (MEM)
detector with a Cu I photocathode was used instead of the 4242 tube and a
pressure of alO"5 Torr was maintained within the monochromator by a me-
chanical vacuum pump backing a 10 cm Edwards Speedivac oil diffusion
pump with a thermo-electrically cooled chevron baffle (Speedivac DCB2B)
and a liquid N2 trap between the diffusion pump and the monochromator.15
Gratings blazed at 200 and 500 nm having 600 grooves mm"1 were used to
cover the spectrum; the monochromator linear dispersion with these grat-
ings is 1.6 nm mm"1. Signals from the PMT and the MEM were amplified
with a Keithley 417 picoammeter.
All concentrations and flow rates in this report are on a volume (number
density) basis not on a mass or mass-per-volume basis. For the type
of experiments discussed in Section V, where all gases are individually
flow metered, the concentration in ppm of a given species, X, mixed
with a flow of a carrier gas, Y, then is :
(How rate of X) X 106/ (flow rate of X + flow rate of Y).
18. Daniels, F., Williacs, J.W., Bender, P., Alberty, R.A. and Cornwall,
C.D., Experimental Physical Chemistry. 6th ed., McGraw-Hill, New York,
1962, p. 439.
12
-------
TP-319a
C. Investigation of Emissions Other than OH(A2S-X2Il)
A number of spectra covering the 150-600 nm range were taken.
Higher wavelengths were not investigated since (i) it is known that in that
region O atoms produce the OH Meinel bands with the same spectral dis-
tribution and with similar intensities with both C2H4 and C2H219 and (ii) CN
emission from N and N/O reactions inside the 600-850 nm region belongs
to the same band system (A-X) which emits at shorter wavelengths2 a and
behaves similarly to the (uv/visible) B-X system. 14» 21> 22
Most emissions in the 150-600 nm range are far more intense for
C2H2 than for C2H4 and appear unattractive for that reason. The exceptions
are the OH(A-X)emission from O-atom reactions (see Section V.D.), the
CN(B-X)emissions from N and N/O mixtures and the NO(P,"Y) and CO Cam-
eron emissions in the region 200 to 250 nm from N/O mixtures.
1. CN(B2S-X22)
From our earlier work14'21'23 it was known that the intensities of
the CN(B.A-X) system in N-atom/carbon compound reactions can be en-
hanced by partially replacing N atoms with O atoms. Moreover, the [O] /
([N] + [O]) ratio which gives the maximum CN intensity differs for different
compounds. The latter fact made it attractive to study not only N-atom but
N/O reactions as well. The bulk of the experiments were made in the
presence of 6 to 12% O2, an O2 range representative of that encountered
in bag samples if such samples were to be mixed inside a monitor in roughly
equal amounts with a N/O/N2 reagent flow. The spectral distributions ob-
tained with C2H4 and C2H2 are very similar. The most intense emissions
19. Krieger, 3., Maiki, M. and Kummler, R., "Chemiluminescent Reactions of
Oxygen Atoms with Reactive Hydrocarbons. I. 7000-9000 A," Env. Sci.
Techn. j6, 742-744 (1972).
20. Pearse, R.W.B. and Gaydon, A.G., The Identification of Molecular Spectra.
Chapman and Hall, London, 1963, Third ed., (a) p. 111-113. (b) p. 241-242.
21. Fontijn, A., "Mechanism of CN and NH Chemiluminescence in the N-0-CaHa
and 0-NO-CaHa Reactions," J. Chem. Phys. 43, 1829-1330 (1965).
22. Kiess, N.H. and Broida, H.P., "Emission Spectra "from Mixtures of Atomic
Nitrogen and Organic Substances," Seventh Symposium (International) on
Combustion, Butterworths, London, 1959, p. 207-214.
23. Fontijn, A. and Ellison, R., "Formation of Electronically Excited Species
in Nitrogen Atom-Oxygen Atom Reactions Catalyzed by Carbon Compounds. NO
(AaE, Ban) and OC'S)," J. Phys. Chem. 12. 3701-3702 (1968).
13
-------
TP-319a
are from the B-X,£v = o transitions; these were studied near 388.3 nm.
Under no condition (P = 1 to 7 Torr; bath gas 90%Ar/10% N2 to 100% N2;
[O]/([N] + [O]} ratio variation from 0 to 1) could we observe ratios
*G H /*C H - ^ or tren<*s that suggested that larger ratios are obtainable.
It thus appears that on the basis of the selection criteria used here
there is little prospect for CN(B.A-X) emission as the basis for reactive hy-
drocarbon measurements. However, Baity, McClenny and Bell24 have shown that
reasonable agreement can be obtained between CN(B-X) emission measured
at 388.3 nm and hydrocarbon reactivity with the exception of C2H2. The latter
could then be subtracted if an independent C2H2 measurement could be devised.
Such an approach is less attractive than the OH Alaos.? tnethod (Sections V.D
and VI) which is not subject to C2H2 interference. The CN 388.3 nm emission
intensity, moreover, varies strongly with small variations in O2 content of
the sample24 (factor of 2.4 decrease for an increase of the O2 content of the
sample from 10 to 20%) and NOx could also be a major interference because
of the rapidity with which it destroys N atoms.23 In* view of the advantages
on all these points of the OH Alaos.9 method we have not further investigated
the CN emission.
2. NO(P,"Y ) and CO Cameron Emissions
Our previous chemiluminescence studies of N/O mixtures had
shown that the intensity of the NO P and V bands, which are characteristic
for such mixtures, is slightly enhanced upon addition of C2H4, but apparently
not upon addition of C2H2.14'23 This phenomenon could best be observed in the
200-250 nm region. More recently we found.that C2H4, but evidently not C2H2,
leads to CO Cameron band emission in this wavelength region.26 These earlier
observations were made in an O2-free system in a study in which C2F4 was
normally used and C2H4 and C2H2 were used, only incidentally; the observed
NOV light intensities using C2F4were two orders of magnitude higher than
with CjH^ the Cameron band Intensities from C2F4 were at least one order
of magnitude higher than those from C2H4.
24. Baity, P.W., McClenny, W.A. and Bell, J.P., "Detection of Hydrocarbons
by Chemiluminescence with Active Nitrogen," American Chemical Society,
Division of Environmental Chemistry, Preprints of Papers 167th National
Meeting, Los Angeles, CA, April 1974, p. 310-312.
25. Brocklehurst, B. and Jennings, K.R., "Reactions of Nitrogen Atoms in the
Gas Phase," Progress in Reaction Kinetics .4, 1-36 (1967).
26. Johnson, S.E., Fontijn, A. and Miller, W.J., "Kinetics of Vacuum Ultra-
violet Chemiluminescence," AeroChem TP-289, AFRPL-TR-73-17, April 1973.
14
-------
TP-319a
By optimizing conditions (N2 bath at 1.6 Torr; |[O]/([N] + [O])| =
0.3; observations near the C2H4 inlet nozzle) we succeeded in getting con-
siderable NOY intensity increases (factor of 10) with C2H4 and, moreover,
found the intensities to be proportional to [ C2H4] from about 100 to 1000 ppm.
O2 has a definite, but not necessarily prohibitive, quenching effect. However,
under these conditions C2H2 led to similar increases in intensity. Thus since
our goal was to find a method not responsive to C2H2 we did not continue this
study.
The Gartner on bands emission is weakly interspersed between the
MOV bands and does not appear practical for hydrocarbon analysis. Since it
can be made the dominant spectral feature in the 190-230 nm region when
C2F4 is used, this emission, as well as the NOY emission, may be useful for
chemiluminescence analysis of halo carbons.
D. OH(A22>X2II) from O-Atom Reactions
1. Spectral Distribution and Intensities
Figure 2 shows a comparison of the pertinent portion of the
OH(A2S-X2II) system from the O/C2H4 and O/C2H2 reactions. For both re-
actants the (O,O)Q2 head at 308.9 nm is a_ dominant feature; however in the
C2H4 case it is the dominant feature while for C2H2 it is just a_ dominant
feature. Specifically, for C2H2 the 312.2 nm (1,1)RX head has almost the
same intensity as the 308.9 nm head, while for C2H4 the (1, I)R! head is
very weak. These observations suggest that a measurement of Al3os.9 =
1308.9 " *3iz.2 could provide a direct measurement of C2H4 in the presence
of C2H2.T To further investigate this point we measured the spectral dis-
tribution of a number of other hydrocarbon species specified by the EPA
technical monitor, which are given in Table 1. All of these hydrocarbons
(except CH4 from which no emission could be detected) were found to yield
the same spectral distribution as C2H4, giving further indications of the
An alternate choice which might be suitable is to subtract IM6.4 rather
than Iai2.2 from 1308.9*
15
-------
TP-319a
suitability of the Alsos.9 measurement as a measurement method for reactive
hydrocarbons free from interference by C2H2 and CH4. The spectra of Fig. 2
were taken at relatively high hydrocarbon concentrations, [HC] , to obtain a
high light intensity allowing good quality spectral distribution measurements.
In further experiments the flow rates of the hydrocarbons of Table 1 were
varied over large ranges (by as much as a factor of 40)* and reaction time
was varied from 2 x 10'3 to 5 x 10"3 sec. None of these changes produced
noticeable changes in spectral distribution, suggesting that the AljoB.9 method
would be valid at all hydrocarbon concentrations of interest. The spectral
distribution from O/C2H4 also did not change when the pump was throttled to
increase the pressure from «0.8 Torr at the standard flow conditions (Table 1
and Fig. 3) to 2.0 Torr. This change did however result in a decrease in
intensity and further experiments were therefore done at «0.8 Torr. The
O/C2H4 spectral distribution also was found to be the same whether a 6% or
12% O2 in He bath or a 100% O2 bath was used. The O/C2H2 spectral distri-
bution showed some change with pressure; the intensity of the~ "subtract1
wavelengths (306.4,,and,312. 2 nm) decreased somewhat with respect to that
at 308.9 nm. Thus we have several reasons for favoring »0.8 Torr (near
the practical lower limit for a mechanical vacuum pump) to higher pressures.
In Section Vtp.2 the physico-chemical reasons for some of these observa-
tions are discussed.
The next point to establish is the relative response of the indivi-
dual hydrocarbons other than C2HZ,^ at 308.9 nm. Since all hydrocarbons
give the same spectral distribution this measurement is the equivalent of
a Al308.9 measurement. To this end spectra were taken over a range of
concentration within the linear part of the response plots (at very high [HC]
deviations from linearity occur, cf. e.g. Fig. 3, which can be attributed to
In the experiments with the test apparatus, Section VI below, the hydro-
carbons were introduced with a flow of air (the counterpart of hydrocar-
bon diluted in the N2/O2 mixture present in bag samples) and the number
of ppm (v/v) of hydrocarbon in the sample flow is a meaningful figure,
the same as in most air pollutant analyzers. In the experiments of the
present section the hydrocarbons flowed directly in the main gas flow
and no such definition is possible. The total flow under the conditions
of Table 1 contained (before reaction) 1600 ppm of hydrocarbon, cor-
responding to 4800 ppm in the sample flow under the conditions of
Section VI where the reagent flow was twice as large as the sample
flow.
The 308.9 nm intensity of C2H2 is 0.25 that of C2H4> the ^I308.s from C2H2
is zero by definition (and measurement, cf. Section VI).
16
-------
TP-319&
TABLE 1. RELATIVE CHEMILUMINESCENCE SIGNAL
INTENSITIES OF CLASS V, IV, III, AND I
HYDROCARBONS,* AT HIGH HYDROCARBON
CONCENTRATION IN SURVEY APPARATUS
Class V, Reactivity = 14.3
Ethylene
Propylene
Butene-1
Butene-2
Isobutene
Propadiene
Butadiene
"Class IV. Reactivity = 9.7
Toluene
Class III. Reactivity = 6.5
n-Butane
n-Heptane
Iso-octane
Class I. Reactivity =1.0
Methane
Ethane
Propane
Benzene
b,c
100
104
91
94
78
48
113
6
10
10
<10'2
3
7
3
4.
2.
2.
1.
9 X 10*
2 X 109
3 X 10'
2 X 1010
1.2 X 1010
unavailable
1.2 Xl010
1.4 X 108
1.6 X 107
7.7 X-107
5.5 X 107
*1 x 104
5.5 X 105
9 XlO6
2.4 X 107
a Reactivity classes and numbers as suggested by B. Dimitriades. 5
b Ethylene is taken as 100.
c The observation conditions were the following: Reaction time = 2 X 10'3
sec; P = 0.8 Torr; O2 flow = 3 ml.atm sec'1; O-atom flow = 4.7 X 10'2
ml.atm sec'1; Hydrocarbon flow = 5 X 10"3 ml.atm sec"1.
d Rate coefficients, in 1 mole'^ec'1, at w25°C for the attack of O atoms
on the specific hydrocarbon as recommended by Herron and Huie from
the availa-ble measurements.
e No detectable signals from methane were obtained; the highest methane
flow tested was 5 X 10"1 ml.atm sec~'.
27. Herron, J.T. and Huie, R.E., "Rate Constants for the Reactions of Atomic
Oxygen (09P) with Organic Compounds in the Gas Phase," J. Pays. Chem. Ref,
Data!, 467-518 (1973).
17
-------
TP-319a
0,0 0.0
m n
-3O6.4
-307.8
-3089
0/C2H2
310 320 330 310 320
WAVELENGTH, nm
330
340
FIGURE 2. OH(A2S-X2n) emission from the O/C2H4 and O/C2H2 reaction.
The O/C2H4 spectrum was obtained at a 3-times more sensi-
tive scale (lower intensity) than the O/C2H2 spectrum.
Spectrometer bandpass: 0.3nm; P = 0.8 Torr; O2 flow 3 ml.
atm sec"1; O-atom flow 1. 1 X 10"1 ml.atm sec'1; C2H4 flow
6.7X 1(T3 ml.atm sec'1; C2H2 flow 2.3X 10'2 ml.atm sec'1.
I
UJ
or
FIGURE 3.
6 8 10 12 14
FLOW, 10-5 ml otm sec''
16
^os.g versus C2H4 flow. Reaction time = 2 X 10"3sec; P = 0. 7
Torr; He flow 2.2 ml.atm sec'1; O2 flow 2.4 X 10'1 ml.atm
sec"1; O-atom flow not measured, may be estimated as
=sl X 10~2 ml.atm sec'1.
18
-------
TP-319a
the fact that the O-atom consumption is no longer negligible); as can be seen
from Fig. 3 such deviations occur at concentrations much higher than are of
practical interest, i.e. larger than those corresponding to 1000 ppm. The
results of the comparison are given in Table 1. The hydrocarbon flow rate
of 5 X 10~3 ml.atm sec"1 used fell within the linear range for all the hydro-
carbons. The agreement between the intensities obtained and the Dimitrlades
reactivity ratings5 is gratifyLngly good. On the basis of these observations it
was decided to continue with the AI308.9 method and to build a test apparatus '
suitable for a thorough testing of the method at realistic hydrocarbon and po-
tential Interference compounds concentrations (see Section VI).
It was further observed, in some additional experiments with the
conventional flow tube in which the hydrocarbon nozzle-to-observation window
distance was varied, that the strong signals from the Class V compounds de-
creased with reaction time, while weaker signals from the Class IV, III, and
I compounds did not decay within the observed reaction time interval (2 X 10"3
to 5 X 10'3 sec) J It follows that the AI308.9 response to the various hydrocar-
bons can be varied by changing reaction time; this is further confirmed in
Section VI.
2, Reaction Kinetics and Kinetic Spectroscopy
The difference in spectral distribution resulting from the C2H2 and
C2H4 reactions with O atoms had been observed previously and is indicative
of two different OH(A22) formation reactions leading to formation of rotation-
ally and vibrationally 'hot' and 'cold1 OH(A2S), respectively." The fact that
the other hydrocarbons produce the same spectral distribution as C2H4 sug-
gests that the mechanism of OH(A2Z) formation is the same in all those cases;
however this reaction mechanism has not yet been established. The O/C2H2
mechanism has been shown to involve O2 and can probably be attributed to29
CH + 02 — OH(A2Z) 4- CO (3)
t The signals from toluene and n-heptane actually showed a slight increase
and those from benzene a strong (factor of 2.5) increase with increasing
reaction time.
28. Becker, K.H., Kley, D. and Norstrom, R.J., "OH Chemiluminescence in
Hydrocarbon Atom Flames," Twelfth Symposium (International) on Combustion.
The Combustion Institute, Pittsburgh, 1969, p. 405-413.
29. Krishnamachari, S.L.N.G. and Broida, H.P., "Effect of Molecular Oxygen
on the Emission Spectra of Atomic Oxygen-Acetylene Flames," J. Chem.
Phys. 34, 1709-1711 (1961).
19
-------
TF-319*
While in accord with Reaction (3) the OH(A2S-Xzn) emission from O/C2H2
increases initially with O2 addition, that from O/C2H4, etc. only decreases
in intensity upon such addition,30 indicative of OH(A2S) quenching by O2 and
of free radical scavenging of an emission precursor by O2. This difference
in O2 influence on I3os.9 is illustrated in Fig. 4 for which an N2/O mixture
was used to provide a comparison to an essentially O2--free situation. ' It may
be seen that the O/C2H2 intensity reaches a maximum and then decreases;
this is again indicative of quenching and free radical scavenging effects.
Qualitatively similar results to those of Fig. 4a were obtained when propy-
lene, butene-1 or butene-2 was used instead of C2H4 and when longer reaction
times, higher pressures and Ar/3% O2 discharge gas were investigated. For
the test/prototype instrument design, these observations indicate [O2] should
be kept relatively low since,at constant [O], Aljos 9 decreases with increasing
[Oa].
The observation (Section V.D.I) that with increasing pressure
the O/C2H2 spectral distribution becomes more like the O/C2H4 distribution
is merely evidence for rotational/vibrational relaxation of the relatively
long-lived (Trad = 8 X 10'7 sec)31'32 emitter OH(A2Z) with increasing pres-
sure. Part of the OH{A2Z) molecules formed will be electronically quenched,
i.e. do not contribute to the light emission. Other factors being equal, the
fraction of OH(A2S) formed that radiates should be kept as high as possible
for high sensitivity. To give an idea of this fraction, consider the rate co-
efficients for electronic quenching, kq. These are32 for He., Ar, N2, and O2,
respectively, < 6x108, 6x 10a, 6x 109 to 3X 1010, and «3x 10ie 1 mole'1 sec'1.
The rate coefficient for light emission krad equals T rad~l and the fraction of
OH(A22} molecules formed which radiate is krad/(krad + kq[N2] ). For 0.5
In the present measurements no special O2 removal traps in the N2
line were used. Previous observations29 suggest that in radically
O2-free N2/O/C2H2 mixtures the 308.9 nm intensity at 0% added O2
is much lower than in the case of Fig. 4.
30. Fontijn, A., "Mechanism of Chemiluminescence of Atonic Oxygen-Hydrocarbon
Reactions. Formation of the Vaidya Hydrocarbon Flame Band Emitter," J.
Chem. Phys. 44, 1702-1707 (1966).
31. Sutherland, R.A. and Anderson, R.A., "Radiative and Predissociative Life-
times of the A2!:"1" State of OK," J. Chem. Phys. 58, 1226-1234 (1973).
32. Becker, K.H. and Haaks, D., "Measurement of the Natural Lifetimes and
Quenching Rate Constants of OH(aS+, v=0,l) and OD(aE+, v=0,l)," Z.
Naturforsch. 28a, 249-256 (1973).
ZO
-------
TP-319a
Torr of N2 (2.7 X 10'5 mole N2 per liter) and taking the lower limit to kq this
fraction is (8 X 10'7)"1 / [(8 X 10'7)'1 + 6 X 109 x 2.7 X 10'5 ] = 0.89.
Taking the upper limit to kq [N2] or taking 0.5 Torr O2 we obtain 0.61.
Electronic quenching thus is far from negligible even at low pressures.
The rate coefficients for O-atom attack on the hydrocarbons are
shown in the last column of Table 1. In general the rate coefficients de-
creased in the same direction as the photochemical smog reactivity ratings.5
This appears to be a principal reason that the AI308.9 measurements give a
reasonably good indication of photochemical smog reactivity (the reactions
following the initial O-atom attack are of course also important in determin-
ing the light intensity). Because the less reactive hydrocarbons are consumed
less rapidly than the more reactive hydrocarbons their intensity will decrease
less rapidly with reaction time (distance), which is the reason that temporal/
spatial adjustment of the relative AJ308.9 from the various hydrocarbons is
possible (cf. Section V.D.I). The increases in intensity with reaction time
observed in e.g. the benzene case, are indicative of the buildup in concentra-
tion of a reaction intermediate. The fact that at 2 X 10"3 sec (the observation
condition of Table 1) the butenes give a lower reading than C2H4 notwithstanding
their larger rate coefficient is probably indicative of their faster consumption.
74-149
6 8 10 0 2 4
[02 ], % OF TOTAL FLOW
8
10
FIGURE 4.
Influence of O2 on 308.9 nm intensities from O/C2H4 and
O/C2H2. Reaction time is 1 X 10"3 sec; P = 1.6 Torr; N2
flow 11.7 ml.atm sec'1; O-atom flow 9.4X 10"2ml.atm
sec'1; C2H4and C2H2 flows 1.1 X 10'2 ml.atm sec'1.
21
-------
TP-319a
SECTION VI
THE CHEMILUMINESCENCE HYDROCARBON
ANALYZER' TEST APPARATUS EXPERIMENTS
The basis for the AI308.9 method has been discussed in Section V.D.
In order to test the method under realistic bag sample conditions hydrocar-
bon concentrations of «0.1 to alOO ppm should be used. This could not be
done with the survey apparatus since (i) a spectrometer is not a sufficiently
sensitive tool and (ii) regular continuous flow metering devices are not ade-
quate for providing such low concentrations. Another flow tube apparatus
(the AI308.9 test apparatus) was therefore built, in which (i) interference fil-
ters replaced the monochromator, (ii) an exponential dilution flask was used
to supply low concentrations of the hydrocarbons, and (iii) a differential
electrometer was used to automatically subtract 1312.2 from 1303.9-
A. Expe rimental
1. Gas Handling System and Flow Conditions
The test apparatus is shown in Fig. 5. The O atoms are produced
by a 2450 MHz microwave discharge in a 13 mm o.d. Vycor tube. The re-
mainder of the apparatus is made of Pyrex. A 22 mm i.d. Pyrex reaction
tube is used. We will refer to the gas passing through the discharge into the
reaction tube as the reagent (= second reactant) gas. This gas for most of
the work was He/9% O2 (1 ml.atm sec"1 He/1 X lO^ml. atm sec"1 O2). In
some experiments 100% O2 was used to provide a comparison. The He/9%
O2 was chosen for the majority of the experiments since it gives a reasonably
high [O] , comparable to that available from 100% O2, yet keeps O2-quench-
ing at a low level.* The hydrocarbons were introduced through a 2 liter ex-
ponential dilution flask using a 0.5 ml.atm sec'1 flow of Scientific Grade
Air? as the carrier gas; the air/hydrocarbon mixture thus simulates the bag
sample gas in automotive exhaust monitoring. The average gas velocity
down the flow tube under these conditions was 270 cm sec"1. In a few ex-
With the test apparatus routine [O] measurements could not readily be
made. In Section VLB.4 some relative [o] measurements are discus-
sed. For the work of Sections VLB. 1-3, the flow conditions and micro-
wave discharge power input were kept constant which resulted in a con-
stant [O] , as testified by the reproducibility (±5%) of the £1303.9 response
with ethylene over the period of these experiments, cf, Section VLB. 2.
This grade air obtained from MG Scientific has the following stated im-
purities: NOX < 5 X 10"3 ppm; CO2 < 2ppm; CO w5 ppm; total hydro-
carbons (mainly methane) < 0.1 ppm.
22
-------
TP-319a
74-39B
AIR
HYDRO-
CARBON
INJECTION PORT
•3-WAY VALVE
if
EXPONENTIAL-
DILUTION FLASK
MAGNETIC STIRRER
REACTION TUBE
306.9 nm FILTER
HV SUPPLY-:
DIFFERENTIAL
ELECTROMETER
MOVABLE
-NOZZLE TUBE
-COLLIMATOR
•312.2 nm FILTER
MANOMETER
CHEMILUMINESCENCE
ZONE
TO PUMP
FIGURE 5. The chemilumlnescence hydrocarbon analyzer test apparatus.
periments at higher concentrations (=;500-5000 ppm) the HC was also flow
metered In directly, bypassing the exponential dilution flask; these experi-
ments served to calibrate the dilution flask output. Gas-tight syringes were
used to introduce the hydrocarbons into the dilution flask. Liquid hydro-
carbons were injected first into a 5 liter predilution flask and after evapor-
ation and mixing were withdrawn from this flask with a gas syringe for in-
jection into the apparatus dilution flask. Potential interference gases (see
Section VI.B. 3) were co-injected with the hydrocarbons into the dilution
flask. In the case of H2O it was noticed that the presence of H2O interfered
with the delivery of n-butane. Therefore, in all H2O interference tests,
H2O was introduced with 0.5 ml.atm sec"1 Scientific Grade Air, as shown in
Fig. 5, while a constant flow of hydrocarbon, corresponding to 850 ppm in
the sample, was used. The pressure in the reaction tube was maintained at
1.2 Torr using a throttling valve in the pump line and a 5 cfm mechanical
vacuum pump. All materials were purified as discussed in Section V.B. The
exponential dilution flask was kept at 38 Torr, which pressure gave a conven-
ient time decay (x7 min for a factor 10 decrease in concentration).
The hydrocarbons were dried by passage through silica gel, the
remaining reagent and sample gases by passage through activated alumina.
Upstream from the microwave discharge and the exponential dilution flasks
23
-------
TP-319
the gases passed through liquid N2 to remove most carbonaceous impurities
(except CO, CH4) that might have been present. To prevent interference
from hydrocarbons which might have been present in the flow system, the
apparatus was washed with dilute HF before it was put to use. Nonetheless,
after several weeks running noticeable increases in the background emission
were observed; these were traced to the hydrocarbon inlet nozzle, which was
therefore frequently removed and washed with dilute HF. No such repeat
cleaning of the rest of the apparatus was found necessary. At the end of the
day, the liquid N2 traps were allowed to warm up and were backpumped; in
this way any accumulated impurities were prevented from entering the re-
action tube.
2. Radiation Measurement
The chemiluminescence was measured with two matched (approxi-
mately equal anode sensitivity, A/lm) Centronic 4242 bialkali PMTs, viewing
the reaction zone through a 308.9 and 312.2 nm filter respectively, cf. Fig.
5. The filters were 2.5 cm diam and had a 1 nnvhalf-width with >20% peak
transmission. Matched PMTs were used to obtain approximately equal in-
tensities at both wavelengths for C2H2; electronic gain adjustment allowed
the intensities to be made exactly equal. Since the stated wavelength for in-
terference filters is that for perpendicular incidence collimation is required
when, as in the present case, a spatially extended glow is observed; the col-
limators between the reaction tube and the PMTs were 2.5 cm diam, 2.5 cm
long (radiator honeycomb) tubes. Since the observation wavelengths are close
to the cut-off of Pyrex the transmission of a tube similar to the reaction tube
was tested. A clear Pyrex or Vycor tube wall (half cylinder) placed between
our present reactor and the monochromator resulted in a 30% decrease in
308.9 nm intensity. (A 50% decrease in intensity was observed when either
tube was covered with phosphoric acid.) These losses were considered
sufficiently small to retain Pyrex without special windows as the reaction
tube material.
3. Electronics
The electronics for the hydrocarbon analyzer basically consist of
two current to voltage converters, one for each PMT, and two difference
amplifiers whose inputs are the amplified PMT signals. Overload protec-
tion for the PMT is also provided.
The light intensity of the chemiluminescent reaction taking place
inside the reaction tube, R, Fig. 6 is monitored by PMT1 and PMT2, re-
spectively. The method of amplification of the two PMT signals is the same
and consists (for PMT1) of a current to voltage converter, CVC1 (CVC2 for
PMT2, etc.), with variable gain selectable on the front panel and two voltage
24
-------
TP-3l9a
T4-I4S
no v
60 Hz
RANGE
ONSTANT
o
FIGURE 6. Diagram of electronics for the hydrocarbon analyzer teat
apparatus.
amplifiers, AAl and ABl, with a gain adjust pot between them. AB1 provides
for selection of time constants of 0.3, 1, 3, and 10 seconds with a front panel
switch. Also included on each amplifier board (containing CVC, AA and AB)
is a current sensor which, when a predetermined anode current on its PMT
is exceeded, will shut off the voltage to both PMTs by relay RL. A reset
button on the front panel will re-apply power to the high voltage supply when
depressed.
The two amplified PMT signals (S^ S2) are applied to two differ-
ence amplifiers, Dj and D2, in such a way as to give KSi-S2 and S2-Slt re-
spectively, as outputs where K ranges between 0 and 1 and is set on the
front panel. K is adjusted to null out the response to C2H2 taking into ac-
count both the differences in PMT sensitivity and the light intensities at
308.9 and 31Z.2 nm.
25
-------
TP-3l9a
A front panel meter and one set of banana jacks display the am-
plified outputs of either PMT or the output of Dl (3 position switch SA).
Another set of banana jacks provides the output D2, while two more sets of
jacks (not shown) make the outputs of ABl and AB2 available for measure-
ment of 1308.9 and I3i2.2 individually and for zeroing out dark current and
background signal of the respective PMTs. In normal operation the Dl
output corresponding to AI^., was used. The D2 output available from S2-
Sj could be used to measure C2H2, which was not done in this work.
B. Results and Discussion
1. Acetylene Zero
A constant flow of C2H2 at 8000 ppm was introduced. The relative
gain of the two PMT signals was adjusted to give a zero reading of AI308>9
under these conditions. If the spectral distribution of O/C2H2 does not
change when smaller concentrations of C2H2 are used, no reading at those
concentrations should be obtained either. C2H2 from the exponential dilu-
tion flask was therefore passed through the zeroed system at concentrations
varying from 100 to 1 ppm. No AI308<9
instrument is zeroed with C2H2 at one concentration it is zeroed at all con-
centration ranges, which also implies that the spectral distribution from
O/C2H2 is invariant with [C2H2] in this range.
One inherent aspect of the electro-optical arrangement used
should be stressed here. The AI308.9 output of the instrument is zero for
acetylene, once the instrument is zeroed with acetylene. However, the in-
dividual PMTs still measure a signal from C2H2 at 308.9 and 312. Z nm.
As a result one cannot measure concentrations of reactive hydrocarbons in
the presence of much larger concentrations of acetylene. Since this situa-
tion will not arise in practice, we have not modified the electronics to allow
such measurements. The ultimate limitation of course would have been
that no reactive HC measurements could be made when their absolute 308.9
nm intensity is comparable to, or less than, the noise in the corresponding
C2H2 signal.
2. Response to Individual Hydrocarbon Species
To obtain an average Class III reading in roughly the proportion
to ethylene indicated by the Dim.itrlades reactivity table5 a number of ex-
periments were made in which the nozzle-to-center of observation port
distance, i.e. the reaction time, was varied for ethylene and n-butane; the
results are shown in Table Z. A distance of 4 cm was selected, corres-
ponding to 1.5 X 10"2 sec, for measurement of all hydrocarbons.
-------
TP-319a
TABLE 2. EFFECT OF NOZZLE DISTANCE ON
AI308>9 FOR He/9% O2 DISCHARGE
GAS AND 1250 PPM HYDROCARBON
Nozzle-to-Center of
Observation Port Distance
cm
2
3
4
5
C2H4
nA
1850
1100
1030
700
nA
180
160
270
320
Ratio
10
7
4
2
[C3H4], ppm (v/v)
125 12.5 1.25
74-99
OJ25
OAO INITIAL CONCENTRATIONS
•A*FROM EXPONENTIAL DILUTION
10 15 20
TIME, minutes
30
FIGURE 7. Exponential dilution plot for propadiene. Reaction time
1.5X 10~2 sec; Reaction tube P = 1.2 Torr; Reagent gas
flow: He flow 1 ml.attn sec~1/O2 flow 0.1 ml.atm sec"1;
Sample (cylinder air) flow from exponential dilution flask
0.5 ml.atm sec'1 with variable hydrocarbon concentration.;
Exponential dilution flask pressure: 38 Torr.
27
-------
TP-319a
The relation between signal response and hydrocarbon concentra-
tions had to be established for each hydrocarbon species. This was done by
making a number of injections over a wide (typically factor of 100) range of
concentrations, cf. e.g. Fig. 7. The points from the exponential dilution
plots giving the same response were then assumed to correspond to the same
concentrations. Plots of response vs. concentration, e.g. Fig. 8, were
prepared from these plots. The response to the individual hydrocarbon
species was found to be linear (first power) in their concentration, cf. Figs.
7 and 8 to within a factor of 2 over the full concentration range.
Table 3 compares the AI308.9 measurements thus obtained for the
individual hydrocarbons. Since the plots are parallel (first order in con-
centration) any concentration point along the line can be used for this com-
parison (we used 125 ppm). Because of this linear response the limit of
sensitivity, which is aO.2 ppm for ethylene, is inversely proportional to the
signal response given in Table 3, i.e. al ppm for n-butane, cf. also Fig. 8.
74-6TA
io-
10
i-T
I I T
OD INITIAL CONCENTRATIONS
FROM EXPONENTIAL DILUTION
C2H4
10"
n-C4H|0
ID'1
10° 10' 10*
CONCENTRATION, ppm (v/v)
FIGURE 8.
Response versus concentration of ethylene and n-butane with
He/9% O2 reagent gas. Conditions as in Figure 7.
28
-------
TP-319a
TABLE 3. RELATIVE CHEMILUMINESCENCE SIGNAL INTEN-
SITIES OF CLASS V, IV, III, AND I HYDROCARBONS*
OBTAINED AT 1.5X 10'2 SEC REACTION TIME FOR
He/9% O2 DISCHARGE GAS
Class V, Reactivity =14.3
Ethylene 100
Propylene 42
Butene-1 79
Butene-2 53
Isobutene 38
Propadiene 133
Butadiene 56
Class IV, Reactivity = 9.7
Toluene 55
Class ni, Reactivity = 6.5
n-Butane 23
n-Heptane 57
Iso-octane 43
Class I, Reactivity =1.0
Ethane 1 . 3
Propane 7.9
Benzene 20
Q
Acetylene 0
Methaned < 10~2
a
Reactivity classes and numbers as suggested by B. Dimitrlades.5
Ethylene is taken as 100.
The acetylene signals are zero since the Instrument is zeroed
using acetylene.
No detectable signals from methane were obtained at concentrations
up to 1250 ppm, the highest concentration investigated.
29
-------
TP-319&
In general the AI308.9 measurements scale quite satisfactorLly with
the reactivity factors. Benzene appears to give too high a response and
some compounds from Class V too low a response. This should be improv-
able by decreasing the observation time, cf. Table 2 and Section V.D, but
this presumably would lead to a decreased reading for Class HI compounds
as well. Further control over the relative A 1303.9 response is possible by
varying the [O] , cf. Section VLB.4.
Over a period of several months the response to 125 ppm C2H4
was frequently checked. No systematic trends with time were observed
and the signal was constant to within
-------
TP-3l9a
TABLE 4. CONCENTRATION RANGES OF HYDROCARBONS (IN PPM)
FOR WHICH INTERFERENCE WAS INVESTIGATED AT THE
INDICATED INTERFERENCE GAS CONCENTRATIONS
CO
1250
Ethylene
n-Butane
125-2
CO 2 H20
20.000 30.000 12.5
C2H2 CH4 NOX
1250-10 1250 125
850
125-2
Propadiene 1000-0.5 1000-0.5
Propane 1000- 10
b
tf
0.25
0 -
850
125-1
125-0.5
1250-10 6-1 125 -5
1000 -5
12.5-0.5
74-90*
•5 ppm C2H4
(a)
—INJECTION
125 ppm CH4
FIGURE 9.
01234
TIME, minutes
Absence of interference by methane. Recorder trace (a)
co-injection with C2H4, and (b) CH4 alone. Conditions as
in Figure 7.
31
-------
TP-319a
[n-C4H|o], ppm (w/v)
12.5
I
7«-9«
1.25
I
INJECTION
125 ppm NOX
O NOX ABSENT
• WITH NOX
10 15 20
TIME, minutes
FIGURE 10.
Absence of interference of NOX with n-butane measure-
ment. Exponential dilution plots with and without added
NOX. Conditions as in Figure 7.
No evidence for positive interference was found. The only nega-
tive interference occurred with H2O which caused a 12% decrease in inten-
sity for 850 ppm C2H4 (because the interference is only a factor two larger
than the variation of consecutive exponential dilution runs, these H2O mea-
surements were made at a steady C2H4 flow after exponential dilution runs
suggested that a quenching effect might occur). The interference of the
other gases thus was < 5% at the concentrations tested.
32
-------
TP-3l9a
74-144
16
S 14
ii 12
10
8
10
20 30 40
% 02IN SAMPLE GAS
50
FIGURE 11.
Effect of O2 in sample gas onAl308.9 response for 12.5 ppm
C2H4. Reaction time 1.5x 10~2 sec; P = 1.2 Torr; Reagent
gas (He + O2) flow 1.1 ml.atm sec'1; Sample gas flow from
exponential dilution flask =0.5 ml.atm sec'1.
Bag samples also will contain a somewhat variable amount of O2
within the range 15-20%. To investigate the influence of variable O2, the
O2 content of the sample gas was varied from 14% to 40% (remainder N2)
for 12.5 ppm C2H4 injections. This [O2] range is of course much larger
than that of practical interest. The results are shown in Fig. 11. They
suggest that a 1% increase or decrease in O2 content of the sample gases
causes somewhat less than about a 1% decrease or increase in the response
to ethylene. Thus the effect of normal variations of the O2 in bag samples
is negligible.
33
-------
TP-319a
4. Measurements with Other .Reagent Gas Compositions
Experiments have been made using undiluted O2 as the reagent
gas with ethylene and n-butane sample gas; two different nozzle distances
were used under flow conditions otherwise very similar to those used for
the He/9% O2 experiments. The results are shown in Fig. 12. At both
distances the response is again first power in [HC]. The C2H4/n-C4H1o
response ratios at 4 and 2 cm are 14 and 2. 5, respectively, similar to
those obtained in the He/9% O2 experiments; however the absolute inten-
sities are a factor of 2 to 3 lower for the undiluted O2 reagent flow. (Com-
pare Fig. 12 to Fig. 8 and Table 2.) The experiments are not strictly com-
parable to each other since no attempt was made to keep [o] the same as
in the He/Oz experiments. Experiments in which [o] rel was measured at
10-6
KT1
Ul
CO
o
a.
(o
iu
oc
10"
•8
100
10' 10*
CONCENTRATION, ppm (v/v)
FIGURE 12. Response versus concentration of ethylene and n-butane with
undiluted O2 reagent gas. P = 1.2 Torr; O2 flow 1.0 ml.atm
sec"1; Sample air flow 0.5 ml.atm sec"1.
34
-------
TF-319*
a number of discharge power levels for the two reagent gas compositions
suggest however that [O] was similar for the discussed conditions. Thus
this comparison again suggests a quenching effect by O2, similar to the ex-
periments of Fig. 11 and of Section V.D.
Experiments were next made in He/9% O2 in which [O] was delib-
erately varied by a factor of 2 by varying discharge power input. The re-
sults are shown in Table 5. It may be seen that decreasing [O] increases
the ratio Al308.9(G2H4)/Al308.9(n-c4Hio)' At 6 cm nozzle distance the abso-
lute response to ethylene also increased. The explanation of these obser-
vations is that at lower [O] the ethylene (which has a much higher reaction
rate coefficient than n-butane for O-atom attack, cf. Table 1) consumption
is decreased and more of it remains to lead to a light producing reaction
after longer reaction times. Not enough n-butane consumption occurs on
this time scale to observe a similar effect on the absolute A 1309.9 signal
from n-butane. The response under all these conditions is again propor-
tional to [HC].
The conclusion from these experiments is that both nozzle dis-
tance and discharge power, i.e. [O], can be used to vary relative AI308.9
response of the hydrocarbons in any desired direction.
5. Hydrocarbon Mixtures
Up to now only samples containing one reactive hydrocarbon have
been discussed (in Section VLB. 3 it was shown that unreactive hydrocar-
bons- -C2H2 and CH4--do not interfere with the response of reactive hydro-
carbons). Bag samples of course contain a mixture of hydrocarbons. To
determine whether synergistic effects between reactive hydrocarbons occur,
we investigated mixtures of ethylene/n-butane (in a 1:10 concentration ratio)
and ethylene/propadiene (in a 1:1 concentration ratio). The results obtained
for the ethylene/n-butane mixture are shown in Fig. 13 (those for the ethylene/
propadiene mixture spanned a concentration range of 0.5 to 500 ppm). The
signals obtained from these mixtures are, within experimental error, equal
to the sum of those obtained from the same quantities of the individual com-
pounds. Hence the AlaoB.? method is apparently not subject to synergistic
effects.
35
-------
TP-3l9a
TABLES. CHANGES IN AI308<9 FROM 125 PPM CEH4
AND n-C4Hlo AS A'FUNCTION OF O-ATOM
CONCENTRATION FOR He/9% O2 REAGENT
Nozzle Distance , , a Al3o8.9 (C2H4) AI308.9 (C^io) AIa08'9 *CzHlL
cm L J rel . nA nA AI308.9 (C4H10)
6 2 23 100 0.23
6 1 34 11 3.1
2 2 200 10 20.0
2 1 135 5 27.0
a Absolute [O] was not measured. The relative [O] was obtained in a
separate experiment in which NO rather than hydrocarbon was passed
through the nozzle and the intensity of the emission from the reaction
O + NO —• NO2 + hv was measured. Since33 I « [o] [NO] and [NO]
was kept constant, a factor of 2 change in light intensity corresponded
to a factor of 2 change in [O].
33. Fontijn, A., Meyer, C.B. and Schiff, H.I., "Absolute Quantum Yield
Measurements of the NO-0 Reaction and Its Use as a Standard for Cheni-
lueinescent Reactions," J. Chem. Phys. 40. 64-70 (1964).
36
-------
TP-319a
10-6
10°
n-C4H,0 CONCENTRATION, ppm (v/v)
I01 I02 10s
74-I4T
10-7
in
£
a.
E
CO
o
a.
to
Ul
a:
10-8
10-
•9
IOT'0
Cg H4 * n- €4 HIQ
I
.10° 10'
C2H4 CONCENTRATION, ppm (v/v)
102
FIGURE 13. Comparison of response of ethylene and n-butane to that of
their mixture at the same individual concentrations. The
individual points on the C2H4 + n-G^xo line are from the ex-
ponential dilution trace of the mixture, the line itself repre-
sents the sum of the responses to the individual hydrocarbons
as shown in the lower two lines. Conditions as in Figure 7,
except for Oatom flow which was not measured.
37
-------
TP-319a
(THIS PAGE IS BLANK)
38
-------
TP-319a
PART II
MANUAL FOR PROTOTYPE ANALYZER
by
Hermann N. Volltrauer
and
Arthur Fontijn
39
-------
TP-319a
SECTION VII
TECHNICAL DESCRIPTION
A. Gas Handling System
1. Sample Gas System
The sample gas is brought into the instrument through a 1/4 in.
Gyrolock connector SAMPLE* located on the rear panel (see Fig. 15). From
there it flows at a rate of about 1 ml. atm sec"1 through a metering valve
embedded in a heated aluminum block (to prevent its condensation) to an
adjustable inlet nozzle on the reactor. The aluminum block is heated by a
25W cartridge heater (also embedded in the block) which contains a 0. 6 cm
tube through which the sample flows prior to entering the valve. The block
temperature, nominally 120°C, can be adjusted by changing the value of a
75 Ohm resistor wired in series with the heating element and located on a
terminal board, TB1, on the side panel above the microwave power supply
(see Figs. 16 and 17). Resistors of zero and 100 Ohms result in tempera-
tures of approximately 150 and 110°C, respectively.
2. Reagent Gas System
As discussed in Section X (Tests) below, the 10% O2 in Ar
reagent gas can be used directly from the cylinder, or it can be purified prior
to use by a series of traps. In either case the gas is brought into the instru-
ment through a rear panel 1/4 in. Gyrolock connector 02 at a pressure of 5
to 20 psig. After passing through a normally closed solenoid valve activated
by the MAIN POWER switch (which can be overridden by a miniature switch
located on-terminal boad TB1, Fig. 18), the pressure is reduced to about
1 psig by an internal pressure regulator. The gas then flows either through
a toggle valve and metering valve directly into the discharge or first passes
through the following series of traps before flowing through these valves and
into the discharge:
1. Silica Gel - for removal of water and some hydrocarbons
2. Heated 'Catalyst F1 (Engelhard Industries) - operated at about
375°C to oxidize the remaining hydrocarbons and CO to CO2 and
H20
The terms in capitals refer to the markings on the instrument panels
(see Figs. 14 and 15).
-------
75-57
FIGURE 14. Front view.
41
-------
75-58
p
c©
p
FIGURE 15. Rear view.
42
-------
MEASURING ELECTRONICS-i
PMT SUPPLY
PRESSURE
GAUGE
PMT HOUSING
REACTOR
75- 43
MICROWAVE POWER SUPPLY
TERMINAL BOARD TBI
PRESSURE REGULATOR
SAMPLE
METERING VALVE
CATALYST TRAP
TRAPS 3 and 4
CONSTANT VOLTAGE
TRANSFORMER
BALLAST
VACUUM PUMP
CONNECTION
PMT HOUSING
MICROWAVE CAVITY
MOVABLE SAMPLE NOZZLE
(NOZZLE DISTANCE INDICATOR UNDERNEATH)
PYREX CONNECTING TUBE
MICROWAVE CAVITY TUNING KNOB
FIGURE 16. View with side panel folded down.
43
-------
75-41
ADJUSTMENT POINTS
FOR MEASURING
ELECTRONICS
PMT HIGH
VOLTAGE CABLE
TERMINAL BOARD TBI-
PMT SUPPLY
MICROWAVE POWER SUPPLY
VARIABLE TRANSFORMER
TRAP I
CATALYST TRAP
SAMPLE HEATER BLOCK
SAMPLE METERING VALVE
TRAPS 3 and 4
MICROWAVE CAVITY
TUNING KNOB
02 METERING VALVE
MOVABLE SAMPLE NOZZLE
REACTOR-
-02/Ar INLET
FIGURE 17. Top inside view.
-------
75-42
MICROWAVE CAVITY TURNING KNOB
MICROWAVE
CAVITY
TRAP I
MICROWAVE
POWER SUPPLY
VARIABLE
TRANSFORMER
REAGENT FLOW OVERRIDE SWITCH
MEASURING ELECTRONICS
JT*
PMT SUPPLY
FIGURE 18. Side view with cover off.
-------
TP-319a
3. Silica Gel - for removal of the water produced by 2
4. Ascarite - to remove CO2
5. Filter - to protect the metering valve and discharge tube
The traps, regulator and filter are located on the rear panel. The
first, third and fourth traps are 3. 2 cm o.d. Pyrex tubing containing the
active materials, while the heated catalyst is inside a 2.5 cm i.d. stainless
steel tube wrapped with a 150 W heating element and surrounded by about
2 cm of insulation in an aluminum box. Proper voltage for the heating ele-
ment for the catalyst trap to operate at about 375°C is obtained by half-wave
rectification of the line voltage, but provision for adding a resistor to allow
variation of the catalyst temperature is available on terminal board TB1
which also contains the diode used for rectification. Placing a resistor in
series or parallel with the diode lowers or raises the temperature of the
catalyst, respectively.
With the flow set by the metering valve at about 3 ml.atm sec"1, the
gas proceeds through the microwave discharge cavity via a short length of
phosphoric acid coated 1. 3 cm o. d. Vycor tubing, and then through about
30 cm of 1. 2 cm o.d. Pyrex tubing to the reactor, to be mixed with the
hydrocarbon-containing sample being analyzed. The Vycor discharge tube,
the connecting tube and the reactor are connected with 1/2 in. stainless steel
Cajon fittings with Teflon sleeve inserts to prevent destruction of O atoms by
the stainless steel.
A 50 cm length of light pipe with one end near the discharge and the
other on the front panel DISCHARGE INDICATOR allows monitoring the condi-
tion of the discharge. The discharge may also be viewed directly through the
fan opening in the rear panel.
3. Reactor and Vacuum Pump
The reactor consists of 2. 5 cm o. d. Pyrex tubing 22. 9 cm long with
a 5 cm long 1. 2 cm o. d. side arm 5.1 cm from its inlet end. Cajon fittings
at both ends and the side arm facilitate removal of the reactor for cleaning
or exchange. The O-atom containing gas enters the reactor through the side
arm while the sample gas is introduced through the top of the reactor via a
movable 0.6 cm o.d. Pyrex tube. The end of the tube inside the reactor has
8 small holes through which the sample enters the reactor. A 1 to 1/4 in.
Cajon adaptor allows easy adjustment of the nozzle distance over a 10 cm range.
This distance is indicated by a pointer attached to the tube and can be read
from the scale attached to the side panel.
46
-------
TF-319a
The reactor pressure (which can be used as a quantitative indication
of flows) is monitored by a thermocouple pressure gauge (Hastings model
VT-4) located at the bottom (downstream end) of the reaction tube in a stain-
less steel elbow into which the connecting tube (serving as a half liter ballast)
for the vacuum pump is fitted.
B. Light Detection System
The reaction is viewed through the Pyrex reactor walls by two matched
bialkali PMTs (Centronic type 4242) through narrowband (1 nm) interference
filters centered at 308. 9 and 312. 2 nm. Since the bandpass frequency of the
interference filters is dependent on the angle of incidence of the chemilumin-
escence radiation, collimators (consisting of about 0. 3 cm cell size honey-
comb) are used between the reactor and filters to insure well-defined
transmitted wavelength characteristics of the filters.
C. Electronics
1. Measuring Electronics
The function of the measuring electronics is to (1) amplify the two
PMT signals to useable voltage levels and (2) obtain the weighted difference of
these voltages and display them on the panel meter and recorder output.
A schematic of the signal processor circuit boards (two are used) is
shown in Fig. 19. With the exception of the subtracting circuit which has been
added to the 312 board, the two circuits are identical and consist of a current
to voltage converter (Ul) whose gain is determined by resistors on the ganged
(with the corresponding resistors for the 308 circuit to vary both gains simul-
taneously) range switch. The resistors were chosen to correct from one range
to the next for the slight inherent non-linearity of the response of the instru-
ment (see Section X. E) with changes in concentration. This makes the maxi-
mum error on any scale due to the non-linearity ~ 3% of full scale. * Offset
current is also fed into Ul to permit zeroing out any background present when
zero gas is sampled.
Two ganged potentiometers acting as voltage dividers and accessible
on the front panel CALIBRATION are wired between a second (U2) and third
(U3) voltage amplifier to allow simultaneous variation of gain of the two cir-
cuits for calibration purposes. Front panel selectable TIME CONSTANT, SEC
capacitors are included in the feedback circuits of the U3s on a ganged switch
A possibly more elegant solution for future instruments might be the use
of an operational amplifier programmed to linearize the response.
47
-------
00
312
f*r
AttUf
*>
4
on
ZEflt
31*
Owe.
j« o i
ra n
6£4'5
^ A/N A/
1 , ^ ^
ci 1 'W «**
^X :K "
^ II 't
Cz'IfOm-
'*«
HHI-
«to»
_l>^
Ttf/^'v.
umrrrM .£>J
Itm
I
UttT OMfVtUr BM*nMS
C/6 , Ol , RK, 111
ommeo •
HTF^
.^Sb-y,
FIGURE 19. Signal processor circuit.
-------
TP-319a
for selection of time constants of 0.1, 0.3, 1, 3 and 10 seconds. The largest
range resistor, 50 megohms (which is the one for the 1 ppm range), on each
circuit is paralleled with a 0.47 p.F capacitor to make the time constant on
that range 25 sec.
The 308 output is connected through a 6 V Zener diode to a light
emitting diode located on the front panel OVERRANGE which is turned on
whenever the voltage across it reaches about 1.5V. When hydrocarbon con-
centrations are displayed on the meter this overrange light informs the
operator that the signal due to acetylene is greater than the electronics can
handle and a higher range should be used. This situation occurs when the
308 signal exceeds about twice full scale.
Subtraction of the 312 from the 308 signal, to obtain a reading corre-
sponding to the hydrocarbon concentration, is performed by U4 on the 312
board. The inverting input of U4 receives the 312 signal to be subtracted
directly from U3 through a 50 kilo-Ohm resistor. The gain of U4 to the 308
signal at the non-inverting input is adjusted with the ratio pot on the front
panel ZERO C2H2 to give zero output with an acetylene sample. The output,
HC, of U4 is related to its inputs, (308) and (312) by
HC = 2R(308) - (312)
where R is the fraction showing on the ZERO C2H2 dial (0-1).
Power for driving the operational amplifiers on the two signal processor
boards is obtained from a * 15 V, 220 mA power supply mounted on a third
printed circuit board in the electronics enclosure. To prevent high humidity
from affecting the performance of the circuits, a higher temperature is main-
tained inside the electronic enclosure by means of a 1300 Ohm resistor wired
to 110 V.
Both PMTs draw their current from a common high voltage power
supply located under the electronics enclosure.
2. Microwave Power Supply
Power for the magnetron (Raytheon RK 5609) is supplied by a high
voltage (~ 1500 V) d.c. supply (see Fig. 20) employing a pi section filter to
reduce ripple. A variable transformer in the primary circuit is used to con-
trol the microwave power, an indication (in % of full scale) of which is
obtained on the front panel meter with the MODE switch in the MW PWR posi-
tion. Access to the Variac transformer dial is obtained on the left side of
the instrument with the cover off (see Fig. 18). Regulation for the supply is
provided by a 500 VA constant voltage transformer.
49
-------
HX
01
o
/77
FIGURE 20. Microwave power supply wiring.
-------
TP-319a
Ignition of the microwave discharge is performed by discharging a.
0.1 uF capacitor, through the primary of a high voltage transformer (a 6 V
automobile ignition coil) by pressing a push button switch DISCHARGE
RESTART on the front panel. The secondary of the transformer is placed
near the cavity to cause the arc from the transformer to ignite the microwave
discharge. The capacitor is charged by a voltage divider connected to the
high voltage d. c. of the power supply.
51
-------
TP-319a
SECTION VIII
OPERATING INSTRUCTIONS
A. Description of Panels and Controls
1. Front Panel (see Fig. 14)
METER
RANGE
TIME CONSTANT SEC
ZERO 308, ZERO 312
CALIBRATION
ZERO C2H2
MODE
PRESSURE
DISCHARGE INDICATOR
DISCHARGE RESTART
OVERRANGE
Continuously displays the concentration
of the sample being analyzed. The scale
is calibrated for .0-1.0 and 0-2.5 parts
per million (ppm).
Selects the measuring range. There are
8 ranges:!.0, 2.5, 10, 'ZSriOO, 250,
1000, and 2500 ppm full scale.
Selects the electronic time constant
from 0.1, 0.3, 1, 3, and 10 sec.
Provide manual means for balancing out
background currents.
Provides means to adjust meter reading
to a known gas sample concentration.
This ratio potentiometer is used to null
the instrument response to acetylene.
Permits displaying one of the following
on the front panel meter: 308 signal,
312 signal, hydrocarbon concentration
(HC), and relative microwave power in
% (MW PWR).
Monitors pressure in the reactor.
Gives indication if discharge is lit
(light) or not (dark).
Permits activating the discharge by
pressing button.
When on indicates that the signal due to
acetylene has saturated the 308
electronics
52
-------
TP-319a
MAIN POWER
DISCHARGE POWER
Turns instrument on and off.
Turns microwave power supply on
(only if main power is on) and off.
2. Rear Panel (see Fig. 15)
POWER PLUG
RECORDER OUTPUTS
SAMPLE
°2
VACUUM
Provides power (110 V a.c. 60 Hz) to
the instrument. The main fuse MAIN
(8A) and the microwave power supply
fuse MW (5A) are located next to the
power plug.
Three double banana plug output jacks
supply an adjusted (via the three
potentiometers) 0-1 V d. c. output for
the 308, 312, and HC signals.
Provides means for connecting the
sample to the instrument.
Provides means for introducing oxygen
to the instrument.
A 2. 6 cm o.d. port for connection to
an external vacuum pump.
B. Installation and Set-Up
The Pyrex tube connecting the discharge tube to the reactor and the
sample inlet nozzle are packed separately for shipment to prevent breakage.
They must be re-installed before the instrument is turned on. In addition it
is suggested that the following procedure be followed when initially setting up
the instrument*:
Step 1. Remove the cover (see Section IX. B» 1).
Step 2. Unpack the Pyrex connecting tube with fittings and Teflon inserts and
install (cf. Section IX. B. 4).
If other relative reactivity ratings are required see Section X.
53
-------
TP-319a
Step 3. Unpack and install the sample inlet nozzle (cf. Section IX. B. 3).
Step 4. Connect a 150 1/min vacuum pump to the VACUUM port on rear
panel.
Step 5. Connect a source of- 10% O2 in Ar to the 62 inlet on rear panel.
Pressure should be between 5 and 20 psig.
Step 6. Cap SAMPLE inlet and close toggle valve in Q2 line located near
traps on rear panel.
Step 7. Plug instrument power cord and pump power co.rd into a grounded
110 V, 60 Hz outlet. Turn both on. Press DISCHARGE POWER
switch if not lit. The high voltage supply for the magnetron has an
automatic three-minute delay from the time the DISCHARGE POWER
light has come on.
Step 8. If pressure does not rapidly drop to about 0.1 Torr, find,leak and
repair.
Step 9. When pressure is ~ 0. 1 Torr uncap SAMPLE inlet and if necessary
adjust sample metering valve to obtain a pressure of 0.85 Torr.
Step 10. Open toggle valve in O2 line and let lines fill with the Oz/inert gas
mixture, \djust pressure to 1.6 Torr with O2 metering valve. This
pressure is correct when Ar is used as inert gas and corresponds to
a flow of 3 ml. atm sec'1 second reactant.
Step 11. With the MODE switch at MW PWR press DISCHARGE RESTART
button. After a few seconds adjust variable transformer (accessible
through left side of instrument with cover off, see Fig. 18) for a reading
of 20% of full scale.
Step 12. With a hydrocarbon-containing sample connected to the sample inlet
and the MODE switch at 308, maximize the meter reading by adjusting
the plastic knob on the microwave cavity.
C. Operation
When the installation and set-up have been completed, the instrument
should be ready to operate.
If lower concentration samples (50 ppm or less) are to be measured,
the background must be zeroed out. It will take from thirty minutes to one
hour for the background to reach its minimum value of about 2 ppm equivalent
54
-------
TP-319a
in the HC mode. To zero out the background connect a zero gas sample to
the SAMPLE input and with the MODE switch and the appropriate ZERO knob,
alternately display and null the meter for both 308 and 312.
D. Calibration
Complete calibration will involve the measurement of a number of
individual hydrocarbons. After the instrument (including the discharge) has
been operating for several hours proceed (to obtain a constant background)
as follows:*
Step 1. With the TIME CONSTANT at 3 sec and the RANGE switch at 2. 5 ppm,
or higher if necessary, turn the MODE switch to 308 and, using the
ZERO 308 knob, null the meter. Turn the MODE switch to 312 and
null the meter with the ZERO 312 knob.
Step 2. With the MODE switch at HC and the RANGE at 2500 ppm, introduce
an acetylene sample through the rear panel SAMPLE input having a
concentration between 10 and 200 ppm. Turn the RANGE switch
counterclockwise until the OVERRANGE light comes on, then back
off one stop. Null, the meter with the ZERO C2H2 knob.
Step 3. Introduce an ethylene sample of a known concentration between 5 and
500 ppm (with the RANGE switch set accordingly). Adjust the CALI-
BRATION dial to get agreement between the sample and the meter.
If more adjustment than possible with the CALIBRATION dial is
needed, turn the CALIBRATION dial to 2 and calibrate using the PMT
high voltage adjustment pot, see Fig. 17.
Step 4. Introduce a sample of hydrocarbon (HCX) from a reactivity class
other than V (ethylene),e. g. iso-octane, in a known concentration
between 50 and 1000 ppm. From the meter reading calculate the
ratio of ethylene to HCX response normalized to the same concentra-
tion . Using Section X as a guide determine approximate changes in
nozzle distance and microwave power needed to obtain the desired
ratio for the ethylene to HCX response.
Step 5. With the cover removed (Section IX.B.l) change the nozzle distance by
loosening the 1/4 in. Cajon fitting on top of the reactor and gently
twist the 0. 6 cm Pyrex hydrocarbon inlet tube; bring it to its new
location as read on the scale on the side panel. With the MODE
switch at the MW PWR position adjust the variable transformer
(accessible through the left side of the instrument, see Figs. 17 and
18) until the value determined in Step 4 is obtained.
55
-------
TP-319a
Step 6. Repeat Steps 3 and 4, and if necessary Step 5, until the desired ratio
is obtained.
Step 7. Repeat Steps 1 and 2. This "completes the calibration.
An abbreviated (and less accurate) calibration could be carried out by
performing only Steps 1, 2, and 3 above or performing Steps 1 and 3 only.
NOTE: Other hydrocarbons can be used for the calibration once enough data
is available to relate their responses under varying conditions.
Desirable ratios of responses may differ from those suggested here
and will depend on the application.
56
-------
TF-319a
SECTION IX
MAINTENANCE AND TROUBLESHOOTING
A. Planned Maintenance
It is anticipated that only the sample metering valve and some of the
Pyrex and Vycor parts will need occasional cleaning. The pressure gauge will
give a good indication of the condition of the sample valve, while the sample
inlet nozzle is the prime suspect if the background increases.
The only item in the instrument that has a known limited life is the
magnetron*; since it is operated well below its maximum current rating, it
should have a useful life well in excess of one year. It is important, however,
to keep the grid directly above the magnetron clean so as not to obstruct the
cooling air circulation.
B. Disassembly and Reassembly
CAUTION: Be sure the MAIN POWER switch is off and the power cord is
disconnected before attempting any servicing of the instrument
unless explicitly stated otherwise.
1. Cabinet
To gain access to the inside of the instrument, the U-shaped
cover must be removed by the following procedure:
Step 1. Remove the 6 screws on the handles.
Step 2. Remove the 7 screws on each side of the cover.
Step 3. Remove the 4 screws holding the fan, lift it out and disconnect it.
Step 4. Lift up the cover.
Due to the limited life and long delivery times, we recommend that a
spare magnetron (Raytheon RK 5609) be kept on hand.
57
-------
2. PMT Housing
To replace or service a filter or collimator or to remove a PMT
housing for servicing, proceed as follows:
Step 1. With instrument and pump unplugged, remove the cover
(Section IX. B. 1).
Step 2. Loosen all four screws visible through the two short horizontal slots
in the right (hinged) side panel.
Step 3. Remove the 14 screws around the perimeter of-the right side panel
but none at the hinge.
Step 4. Disconnect:
(a) the I/ 4 in. Cajon fitting at the top of the sample inlet nozzle on
the reactor,
(b) the 1/8 in. Gyrclock connector (from traps) at the O2 metering
valve,
(c) the microwave cable at the cavity. It may help to tilt the side
panel down slightly.
Step 5. Fold down the side panel, after providing support for it.
Step 6. Loosen the 4 screws .on the plate holding the two halves of the
aluminum reactor block together.
Step 7. Remove the 4 screws for each PMT housing that is to be removed
a-nd carefully back up the housing away from the reactor. When free,
lift it out. If both housings are to be removed, support for the
reactor will be required.
NOTE: Each PMT housing is fitted to its half of the reactor block and to its
location on the side panel.
Step 8. After removal place the housing on a clean surface. Remove the 4
plastic screws and pull off the PMT to reactor coupler with collima-
tor. The collimator is press-fitted into the coupler and is easily
removed.
Step 9. Remove the plastic spacer and 4 flat head screws. Turn one of the
flat head screws partially into a threaded hole of the filter holder and
gently pull the filter holder out. This last step exposes the PMT and
should be carried out in subdued light.
58
-------
TP-319a
To reassemble follow above steps in the reverse order'being care-
ful in Steps 9 and 8 to properly fit the plastic spacer in the groove. Some
twisting of the coupler relative to the housing may be needed in Step 7 to
align the PMT housings.
3. Sample Inlet Nozzle
Step 1. With instrument and pump unplugged, remove top panel (see
Section IX. B. 1).
Step 2. Disconnect the 1/4 in. Cajon fitting at the top of the 0. 6 cm sample
inlet nozzle tube.
Step 3. Loosen the 1/4 in. Cajon connector at the 1/4 to 1 in. adaptor on
top of the reactor and carefully pull out the 0.6 cm tube, twisting
it if necessary.
Step 4. Clean the tube by soaking it in dilute (1-5% in H2O) HF for a few
minutes.
CAUTION: HF should not be allowed to come in contact with the skin.
Step 5. Rinse in distilled water and dry.
Step 6. Reassemble.
4. Reactor
Step 1. Follow Steps 1 through 5 of Section IX. B. 2 then perform Steps 2
and 3 of IX. B. 3 above.
Step 2. Loos^en the two 1 in. Cajon fittings at the top and bottom of the
reactor. Remove the top one.
Step 3. Loosen both ends of the 1/2 in. Cajon fitting on the side arm.
Step 4. Loosen the 1/2 in. Cajon fitting at the Pyrex elbow end of the dis-
charge cavity.
Step 5. Very carefully lift out the 1. 2 cm Pyrex tube containing the bends.
Step 6. Remove the 1/2 in. Cajon fitting at the reactor side arm. Do not
lose the Teflon insert.
Step 7. Lift out the reactor and clean as in Steps 4 and 5 of Section IX. B. 3,
then reassemble.
59
-------
TP-319a
5. Vycor Discharge Tube
Step 1. Follow Steps 1 through 5 of Section IX. B. 2.
Step 2. Loosen the 1/2 in. Cajon fitting on the side of the bent tube at the
side arm of the reactor.
Step 3. Loosen all three connections at the two Cajon fittings at the cavity.
Step 4. Carefully lift out the 1. 2 cm bent Pyrex tube.
Step 5. Remove the Vycor tube and clean as in Steps 4.and 5 of Section IX. B. 3
above.
Step 6. Before re-installing this tube coat the inside with phosphoric acid.
The recommended method for doing this is to expose some powdery
P2O5 to the atmosphere and apply the resulting syrupy liquid to the
inside of the tube, with e. g. Pyrex wool.
CAUTION: Do not allow the phosphoric acid to come in contact with an
organic material; this could result in high background readings.
NOTE: The various parts of the 1/2 in. Cajon fittings are not interchangeable
and must therefore be replaced in exactly the same position.
6. Valves
Removal of the O2 metering valve is straightforward. The only pre-
caution necessary is to avoid excessive pressure on the Vycor tube passing
through the cavity.
Access to the heated sample valve is achieved by removing the two
screws in the split aluminum heating block and lifting off the top half. Re-
moving the 1/8 in. Gyrolock fitting on the sample inlet line frees the valve
and heater assembly which can then be separated by undoing the 1/8 in. Gyro-
lock connection at the valve. Removing the other 1/8 in. Gyrolock connection
frees the valve which can then be opened and cleaned.
7. Microwave Power Supply
Step 1. With instrument unplugged remove top cover (see Section IX. B. 1).
Step 2. Remove microwave cable at supply.
60
-------
TP-319a
Step 3. Unscrew the wires on the terminal strip on top of the power supply
leading away from it. Connections on the terminal strip are from
left to right: black, white, orange, yellow green, black, and blue.
Step 4. Remove the 4 screws on the left side of the instrument holding the
supply.
Step 5. Remove the 2 support bars for the fan on top of the instrument and
carefully lift out the supply.
C. Troubleshooting
After some initial construction faults were corrected, no problems
with the instrument have been experienced at AeroChem, except for the
microwave interference which can be easily corrected (see Section IX. D.
below). Some possible problems that may arise are the following:
1. Measuring Electronics - if any problems occur that seem to be due
to the electronics check the following:
(a) Is the heater in the electronics enclosure operating? (It should
be hot to the touch.)
(b) Do the circuit boards fit snugly into their sockets?
(c) Are all cables plugged into PMT bases and into the PMT supply?
(d) Are the fuses good?
If none of the above locates the problem, check for broken wires
with the aid of Fig. 21.
2. Meter - If the meter does not read zero with zero air input on the
2500 ppm scale, see Section IX. D.
3. Microwave Power Supply - Operation of the microwave supply can
be checked by connecting a UHF diode (IN23) to a milliammeter and holding
the diode near the cavity. Zero current indicates the absence of microwave
radiation. The most likely source of this problem would be the magnetron,
followed by the high voltage power supply (see Fig. 20).
4. Light Leaks - These should be no problem because of the small
bandwidth of the filters used. If tests indicate the contrary (more than 1 ppm
equivalent signal at 308 or 312 in a normally lit room with the cover off and
the microwave supply off) it is likely that a filter has become displaced (see
Section IX. B. 2).
61
-------
li -19
— a
B«
»
•
(^
r
i
i
0
;
&
c
i
"
"i
. —
Ml
T
1
1
I
*»
I
T
3
4
51
6
7
30*
wr
HOOVH&
/tor
C'/JltR
SlZ
«y
W
LJ
I/fVOC.
*irni
i
HO AC
WMSM6
FIGURE 21. Chassis wiring.
-------
TP-319a
5. Air Leaks - With both metering valves off, the pressure gauge
should read less than 0. 1 Torr. Higher readings indicate a leak, a faulty
valve or a faulty pump. Leaks can be easily pinpointed by spraying acetone
on suspected joints and watching for pressure increases.
D. Signal Processor Board Adjustments
Due to a small microwave radiation induced electrical interference,
the signal may be affected by changes in the microwave power level. This
interference can easily be zeroed out with the voltage offsets of the opera-
tional amplifiers. When changes in microwave power are made the output
of the instrument should be checked on the 2500 ppm scMe by sampling
ambient air. If the output is not zero, the following alignment procedure
should be followed.
Step 1. Note reading on CALIBRATION dial (it will be used in Step 15).
Step 2. Unplug analyzer from the 110 V supply.
Step 3. Remove the cover (see Section IX. B. 1).
Step 4. Unplug PMT high Voltage cable at the power supply, see Fig. 17.
Step 5. Plug in unit and depress MAIN POWER switch. (Depress DISCHARGE
POWER also if this button is not lit.) Wait until meter reads magne-
tron anode current ( i. e. with the MODE switch at MW PWR position
a positive reading should be obtained) and depress DISCHARGE RE-
START button a few times until discharge lights.
Step 6. Turn both (308 and 312) ZERO knobs fully clockwise.
Step 7. With MODE switch at 312 and CALIBRAT ION knob at 10 (fully clock-
wise) adjust R7 (R7, R13 and R22 are the adjustment points for the
measuring electronics, see Fig. 17) of the 312 card so that the
same positive meter reading is obtained on the 1 and 2500 ppm
range. R13 may have to be adjusted to get a positive reading.
Step 8. Reduce reading to « 5% of full scale with R13 and note exact
reading.
Step 9. Turn CALIBRATION to zero and note reading. If meter is below
zero bring up with R22, then obtain readings with CALIBRATION
at 10 and 0.
63
-------
TP-319a
Step 10. Adjust R13 to make reading with CALIBRATION at zero 10% less
than the difference in readings between CALIBRATION at 10 and
zero, e.g. if reading is 17 and 5 for CALIBRATION settings of 10
and 0 respectively adjust, while at 0, for 5-0. 1(17-5) s 3.8.
Step 11. Repeat Step 10 until the same reading is obtained at both extremes
of the CALIBRATION pot (e.g. 10 and 0).
Step 12. Null the meter with R22.
Step 13 Repeat Steps 7 through 11 with MODE switch at 308 and operating
on 308 card.
Step 14. Turn MODE switch to HC and zero meter using R22 of 308 card.
Step 15. Return CALIBRATION dial to original setting obtained in Step 1.
For any one CALIBRATION setting the adjustment can be simplified
as follows. Perform Steps 2, 3 and 5 of above. Then while sampling ambient
air on the 2500 RANGE of 312 adjust R22 (also R13 if necessary) of the 312
card to null the meter. Turn the MODE switch to HC and adjust R22 (also
R13 if necessary) of the 308 card to again null the meter. If the meter read-
ing can be made zero with the ZERO knobs at both 308 and 312 MODE settings
on the 1 ppm scale, the instrument is zeroed on all ranges.
E. Shipping
Should it become necessary to ship the instrument, please observe
the following:
Step 1. Remove all external connections.
Step 2. Remove and pack separately the Pyrex connecting tube and the
0. 6 cm hydrocarbon inlet tube (see Sections VIII. B and IX. B).
Step 3. Use the original shipping container or another suitable container
and surround the instrument by at least 10 cm of packing material.
64
-------
TP-319a
SECTION X
TESTS
The instrument was subjected to a number of tests to (i) check its
general performance, (ii) select conditions that result in a satisfactory agree-
ment with the Dimitriades scale and at the same time lead to high sensitivity
and (iii) determine the linearity and sensitivity of the instrument at those
conditions.
A. Reagent Gas Purification Traps
The instrument was found to operate satisfactorily when the traps
in the reagent gas line were bypassed; we therefore performed all further
tests in that mode. However the traps have been included with the instrument
for possible use under special operating conditions e.g. when O2/He reagent
gas, which may contain hydrocarbons,is used. They may also provide zero air.
B. Collimators
A number of lengths of 0. 3 cm cell-size honeycomb were used in
front of each PM.T and the signal-to-noise ratio was measured for ethylene
(after zeroing for acetylene). The differences in signal-to-noise ratios were
small; the best ratio was achieved when the 308 collimator was 2.5 cm long
and the 312 collimator was 1.3 cm long*.
C. Flow Rates
A reagent flow rate of a 3 ml. atm sec'1 with a sample flow of
' * 1 ml. atm sec"1 gave satisfactory agreement with the Dimitriades ratings3
(Sections X.D and F) and resulted in a sensitivity of 0.05 ppm ethylene equiva-
lent. These flow rates result in a reading of 1. 6 Torr on the front panel pres-
sure gauge and were used for all further tests. Increased sample flow may
result in an increased sensitivity; however as a result of such increases, the
linearity will suffer at the upper limit (near 2500 ppm) and response ratios
will change. Any choice of flow rates is to a large extent arbitrary and other
considerations may dictate flow rates different from those chosen.
65
-------
TP-319a
D. Nozzle Distance and Microwave Power
Under the flow conditions of Section X. C the response of the instru-
ment to ethylene, iso-octane and benzene* was determined as a function of
nozzle-to-center of observation port distance and microwave power level (as
% of full scale; 100% » 125 mA anode current), see Fig. 22.
E. Linearity and Limit-of-Sensitivity
At 20% microwave power detailed data were obtained for ethylene
and limited (over 1 or 2 orders of magnitude) data for some other hydrocarbons.
The lowest detectable ethylene concentration (S/N = 2)"was found to be ** 0.05
ppm. In these tests known amounts of hydrocarbons were introduced with air
into a 5 liter exponential dilution flask. To calibrate the dilution flask concen-
tration of NO versus time was measured with an AeroChem Model AA-5
commercial chemiluminescence NO monitor which has a known linear response.
The signal of the hydrocarbon analyzer was initially found to be proportional
to [hydrocarbon concentration]0*96. This dependence would result in a 7%
deviation from linearity per decade. A similar slight deviation had been
suggested by the test apparatus results. In the instrument delivered to EPA
this deviation is corrected electronically, cf. Section VII. C. 1.
F. Response Ratios
The response of some other hydrocarbons relative to that of ethylene
for a given concentration of each is presented in Table 6 for a nozzle-to-center
of observation port distance of 1 cm and 20% and 30% microwave power levels.
The 5 liter exponential dilution flask (Section X. E) was again used in these tests.
Benzene was used since it is the Class I hydrocarbon most likely, cf.
Table 3, to give too high a reading and therefore required closest
scrutiny.
66
-------
12
10
- 6
1
ui 4
i r
MICROWAVE POWER
' 20%
— — — — - 30%
BENZENE X5
ISO OCTANE
I 2 3
NOZZLE DISTANCE, cm
FIGURE 22.
Response for ethylene, benzene and iso-octane
as a function of nozzle-to "Center of observation
port distance and microwave power. Sample
flow 1 ml.atm sec'1! Reagent (10% O2/Ar) flow
3 ml. atm sec"; P e 1.6 Torr.
67
-------
TP-319a
TABLE 6. PROTOTYPE ANALYZER RELATIVE CHEMILUMINESCENCE
SIGNAL INTENSITIES OF CLASS V, IV, III AND I HYDRO-
CARBONS* AT MICROWAVE POWER LEVELS OF 20% (A)
AND 30% (B) OF FULL SCALE
Class V. Reactivity** = 14.3 A B
Ethylene 100 100
Propylene 82 77
Butene-1 96 91
Isobutene 70 71
Class IV. Reactivity = 9.7
Toluene 40 42
Class III. Reactivity = 6.5
n-Butane 14 18
Is o-octane 47 54
Class I. Reactivity = 1.0
Propane 7.3 8.0
Benzene 8.6 9.8
Acetylene 0 0
Reactivity classes and numbers as suggested by Dimitriades. 5
Operating conditions: Nozzle-to-center of observation port
distance, 1 cm; Sample flow, 1 ml. atm sec'1; Reagent
(Oz/10% Ar) How 3 ml.atm sec'1; P = 1.6 Torr.
68
-------
TP-319a
SECTION XI
REFERENCES
1. Niki, H., Daby, E.E. and Weinstock, B., "Mechanisms of Smog Reactions,"
in Photochemical Smog and Ozone Reactions. Advances in Chemistry Series
113, American Chemical Society, Washington, DC, 1972, p. 16-57.
2. Leighton, P.A., Photochemistry of Air Pollution. Academic Press, New York.
1961.
3. Coloff, S.G., Cooke, M., Drago, R.J. and Sleva, S.F., "Ambient Air
Monitoring of Gaseous Pollutants," American Laboratory, July 1973, 10-22.
4. Proceedings of the Solvent Reactivity Conference. EPA-650/3-74-010,
November 1974.
5. Dimitriades, B., "The Concept of Reactivity and Its Possible Applications
in Control," Ref. 4, p. 13.
6. Finlayson, B.J., Pitts, J.N. . and Akimoto, H., "Production of Vibra-
tionally Excited OH in Chemiluminescent Ozone-Olefin Reactions," Chem.
Phys. Lett. 1£, 495-498 (1972).
7. Rummer, W.A., Pitts, J.N. and Steer, R.P., "Chemiluminescent Reactions
of Ozone with Olefins and Sulfides," Env. Sci. Techn. 5_, 1045-1047 (1971).
8. Hodgeson, J.A., McClenny, W.A. and Martin, B.E., "Environmental Protec-
tion Agency, Private communications' 1973 and 1974.
9. Wilson, W.E. Jr., "A Critical Review of the Gas-Phase Reaction Kinetics
of the Hydroxyl Radical," J. Phys. Chem. Ref. Data .1, 535-573 (1972).
10. Del Greco, P.P. and Kaufman, F., "Lifetime and Reactions of OH Radicals
in Discharge Flow Systems," Disc. Faraday Soc. 33, 128-138 (1962).
11. Bayes, K.D. and Jansson, R.E.W., "The Origin of Light Emission in the
Atomic Hydrogen-Acetylene Flame," Proc. Roy. Soc. A282. 275-282 (1964).
12. Schatz, G. and Kaufman, M., "Chemiluminescence Excited by Atomic Fluorine,"
J. Phys. Chem. 26, 3586-3590 (1972).
13. Gaydon, A.G., The Spectroscopy of Flames. John Wiley, New York, 1957,
p. 252.
14. Fontijn, A., Ellison, R., Smith, W.H. and Hesser, J.E., "Chemiluminescent
Emission of CO Fourth Positive Bands in Nitrogen Atom/Oxygen Atom/Reactive
Carbon Compound Systems. Relation to Chemi-Ionization," J. Chem. Phys.
53, 2680-2687 (1970).
15. Fontijn, A. and Johnson, S.E., "Mechanism of CO Fourth Positive VUV Chemi-
luminescence in the Atomic Oxygen Reaction with Acetylene. Production of
C(3P,1D)," J. Chem. Phys. _59, 6193-6200 (1973).
69
-------
TP-319a
16. Fontijn, A. and Lee, J., "Comparison of the Absolute Quantum Yields of
the Gas-Phase 0/NO Reaction and the Liquid-Phase Luminol Oxidation Chemi-
luminescence Standards," J. Opt. Soc. Am. 6£, 1095-1098 (1972).
17. Stull, D.R., "Vapor Pressure of Pure Substances. Organic Compounds,"
Ind. Eng. Chem. _39, 517-540 (1947).
18. Daniels, F., Williams, J.W., Bender, P., Alberty, R.A. and Cornwell,
C.D., Experimental Physical Chemistry. 6th ed., McGraw-Hill, New York,
1962, p. 439.
19. Krieger, B., Malki, M. and Kummler, R., "Chemiliiminescent Reactions of
Oxygen Atoms with Reactive Hydrocarbons. I. 7000-90DO A," Env. Sci.
Xechn. 6, 742-744 (1972).
20. Pearse, R.W.B. and Gaydon, A.G., The Identification of Molecular Spectra,
Chapman and Hall, London, 1963, Third ed., (a) p. 111-113. (b) p. 241-242.
21. Fontijn, A., "Mechanism of CN and NH Chemiluminescence in the N-0-C2H2
and 0-NO-C2H2 Reactions," J. Chem. Phys. 43_, 1829-1830 (1965).
22. Kiess, N.H. and Broida, H.P., "Emission Spectra from Mixtures of Atomic
Nitrogen and Organic Substances," Seventh Symposium (International) on
Combustion, Butterworths, London, 1959, p. 207-214.
23. Fontijn, A. and Ellison, R., "Formation of Electronically Excited Species
in Nitrogen Atom-Oxygen Atom Reactions Catalyzed by Carbon Compounds. NO
(AaZ, B2n) and O^S)," J. Phys. Chem. 72. 3701-3702 (1968).
24. Baity, P.W., McClenny, W.A. and Bell, J.P., "Detection of Hydrocarbons
by Chemiluminescence with Active Nitrogen," American Chemical Society,
Division of Environmental Chemistry, Preprints of Papers 167th National
Meeting, Los Angeles, CA, April 1974, p. 310-312.
25. Brocklehurst, B. and Jennings, K.R., "Reactions of Nitrogen Atoms in the
Gas Phase," Progress in Reaction Kinetics 4_, 1-36 (1967).
26. Johnson, S.E., Fontijn, A. and Miller, W.J., "Kinetics of Vacuum Ultra-
violet Chemiluminescence," AeroChem TP-289, AFRPL-TR-73-17, April 1973.
27. Herron, J.T. and Huie, R.E., "Rate Constants for the Reactions of Atomic
Oxygen (09P) with Organic Compounds in the Gas Phase," J. Phys. Chem. Ref.
Data 2f 467-518 (1973).
28. Becker, K.H., Kley, D. and Norstrom, R.J., "OH Chemiluminescence in
Hydrocarbon Atom Flames," Twelfth Symposium (International) on Combustion,
The Combustion Institute, Pittsburgh, 1969, p. 405-413.
29. Krishnamachari, S.L.N.G. and Broida, H.P., "Effect of Molecular Oxygen
on the Emission Spectra of Atomic Oxygen-Acetylene Flames," J. Chem.
Phys. 34, 1709-1711 (1961).
70
-------
TP-319a
30. Fontijn, A., "Mechanism of Chemiluminescence of Atomic Oxygen-Hydrocarbon
Reactions. Formation of the Vaidya Hydrocarbon Flame Band Emitter," J.
Chem. Phys. 44, 1702-1707 (1966).
31. Sutherland, R.A, and Anderson, R.A., "Radiative and Predissociative Life-
times of the Aar>- State of OH," J. Chem. Phys. 58, 1226-1234 (1973).
32. Becker, K.H. and Haaks, D., "Measurement of the Natural Lifetimes and
Quenching Rate Constants of OH(aE+, v=0,l) and OD(2£+, v*=0,l)," Z.
Naturforsch. 28a, 249-256 (1973).
33. Fontijn, A., Meyer, C.B, and Schiff, H.I., "Absolute Quantum Yield
Measurements of the NO-0 Reaction and Its Use as a-Standard for Cheml-
luminescent Reactions," J. Chem. Phys. 40, 64-70 (1964).
71
-------
TP-319a
SECTION XII
INVENTIONS AND PUBLICATIONS
A patent application entitled "Chemiluminescent Method and Appara-
tus for Determining the Photochemical Reactivity of Organic Pollutants in a
Gaseous Mixtures" by Arthur Fontijn was filed by EPA on February 10, 1975;
serial number 548471.
A paper -was submitted to Environmental Science and Technology on
April 14, 1975. It is entitled "Homogeneous Gas-Phase Chemiluminescence
Measurement of Reactive Hydrocarbon Air Pollutants By Reaction with Oxygen
Atoms" by Arthur Fontijn and Roy Ellison.
72
-------
TP-319a
APPENDIX A: LIST OF MANUFACTURERS OF PARTS
USED IN PROTOTYPE INSTRUMENT
AeroChem Research Laboratories, Inc.
Princeton, NJ 08540
Signal Processor Boards
AeroChem D-28
PMT Housing
AeroChem G-l
API Instruments Co.
Chesterland, OH 44026
Automatic Switch Co. (ASCO)
Florham Park, NJ 07932
Bailey Instruments Co.
Saddle Brook, NJ 07662
Bertan Associates, Inc.
Hicksville, NY 11801
Conoflow Corp.
Blackwood, NJ 08012
Corion Instrument Corp.
Waltham, MA 02154
Hastings-Raydist Co.
Hampton, VA 23361
Raytheon Service Company
Newton Upper Falls, MA 02164
Scintillonics, Inc.
Fort Collins, CO 80521
Panel Meter
API603 0-1 mA DC
Solenoid Valves
Asco 8262C 35 VM
PMTs
Centronic 4242
PMT Power Supply
Bertan Model 602
Low Pressure Regulators
Conoflow H 10XT 1014 05
Interference Filters
308. 9 nm and 312. 2 nm
(both 1 nm FWHM, S 20% peak
transmission and 2. 5 cm diam)
Vacuum Gauge
Hastings VT4 with DV40 tube
Magnetron
Raytheon RK 5609
Microwave Supply Components
Scintillonics Model HV-15A
73
-------
BIBLIOGRAPHIC DATA
SHEET
1. Report No.
EPA-650/2-75-069
3.N^ecipient's Accession No.
4. Title and Subtitle
A CHEMILUM1NESCENCE REACTIVE HYDROCARBON ANALYZER FOR MOBILE
SOURCES
3. Report Date
June 1975
6.
7. Author(s)
Arthur Fontijn, Hermann N. Volltrauer and Roy Elliaon
8' Performing Qrganization'Rept.
No. xp_3iga
1. Performing Organization Name and Address
AeroChem Research Laboratories, Inc.
P.O. Box 12
Princeton, New Jersey 08540
10. Project/Task/Work Unit No.
68-02-1224 (J113)
II. Contract/Grant No.
12. Sponsoring Organization Name and Address
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. Type of Report & Period
Covered Final Report
13 June 1973 to
12 June 1975
14.
15. Supplementary Notes
16. Abstracts A chemilumlnescence method for measuring total reactivity of hydrocarbon (HC)
mixtures has been developed and a prototype analyzer based on this method has been built
The difference between the OH(Aa£a-XaJI) emission intensities at 308.9 and 312.2 run from
0-atorn/hydrocarbon reactions near 1 Torr is measured. For CaH«, I90e.9 » laia.a, for
CaH2 ISOB.S a lau.s. The other hydrocarbons tested yield the same spectral distribu-
tions as CaHi,; CM* yields no emission. Two PMTa are used for 308.9 and 312.2 nm measure-
ment respectively. When the apparatus is zeroed, the difference in signal from the two
PMTs is insensitive to C2Ha. The relative response to the individual reactive HC species
can be set to give good agreement with reactivity ratings. The response to HC mixtures
is additive. A limit of sensitivity of * 0.05. ppm CaH<.-equivalent HC and a linear
response to individual HCs to 2500 ppm Is obtained} greater sensitivity appears feasible
CO, C02, S02, Cm, CaHa and NOX do not interfere with instrument response. A 1% change
in [02] causes < 1% change in signal; 3% H30 causes a 12% decrease.
17. Key Words and Document Analysis. 17o. Descriptors
Hydrocarbons
Reactivity Measurement
17b. Idennfiers/Open-Ended Terras
Chemilumlnescence Monitor
Technical Manual
I7e. COSATI Field/Group
18. Availability Statement
19. Security Class (This
Report)
UNCLASSIFIEE
ASSIFIED
Ilass (This
20. Security Class (This
"IjNCLASSIFIF.D
21. No. of Pages
80
22. Price
FORM NTIS-39 IREV. 3-72)
THIS FORM MAY BE REPRODUCED
USCOMM-OC I49S2-P72
-------
INSTRUCTIONS FOR COMPLETING FORM NTIS-35 (10-70) (Bibliographic Data Sheet based on COSATI
Guidelines to Format Standards for Scientific and Technical Reports Prepared by or for the Federal Government,
PB-180 600).
1. Report Number. Each individually bound report shall carry a unique alphanumeric designation selected by the performing
organization or provided by the sponsoring organization. Use uppercase letters and Arabic numerals only. Examples
FASEB-NS-87 and FAA-RD-68-09.
2. Leave blank.
1 Recipient's Accession Number. Reserved (or use by each report recipient.
4- Title and Subtitle. Title should indicate clearly and briefly the subject coverage of the report, and be displayed promi-
nently. Sec subtitle, if used, in smaller type or otherwise subordinate it to main title. When a report is prepared in more
than one volume, repeat the primary title, add volume number and include subtitle for the specific volume.
5. Report Dote. F.ach report shall carry a date indicating at least month and year. Indicate the basis on which it was selected
(e.g., date of issue, dace of approval, date of preparation.
6. Performing Organisation Code* Leave blank.
7. Authors). Give name(s) in conventional order (e.g., John R. Doe, or J.Robert Doe). List author's affiliation if it differs
from the performing organization.
8. Performing Organization Report Number. Insert if performing organization wishes to assign this number.
9. Performing Organization Name and Address. Give name, street, city, state, and zip code. List no more than two levels of
an organizational hierarchy. Display the name of the organization exactly as it should appear in Government indexes such
as USGROR-I.
10. Project/Task/Work Unit Number. Use the project, task and work unit numbers under which the report was prepared.
11. Contract/Grant Number. Insert contract or grant.number under which report was prepared.
12. Sponsoring Agency Name and Address. Include zip code.
13- Type of Report and Period Covered. Indicate interim, final, etc., and, if applicable, dates covered.
14. Sponsoring Agency Code. Leave blank.
15. Supplementary Notes. Enter information not included elsewhere but useful, such as Prepared in cooperation with . . .
Translation of ... Presented at conference of ... To be published in ... Supersedes . . . Supplements .
16. Abstract. Include a brief (200 words or less) factual summary of the most significant information contained in the report.
If the report contains a significant bibliography or literature survey, mention it here.
17. Key Words and Document Analysis, (a). Descriptors. Select from the Thesaurus of Engineering and Scientific Terms the
proper authorized terms that identify the major concept of the research and are sufficiently specific and precise to be used
as index entries for cataloging.
(b). Identifiers and Open-Ended Terms. Use identifiers for project names, code names, equipment designators, etc. Use
open-ended terms written in descriptor form for those subjects for which no descriptor exists.
(c). COSATI Field/Group. Field and Group assignments are to be taken from the 196? COSATI Subject Category List.
Since the majority of documents are multidisciplinary in nature, the primary Field/Group assignment(s) will be the specific
discipline, area of human endeavor, or type of physical object. The application(s) will be cross-referenced with secondary
Field/Group assignments that will follow the primary postmg(s).
18. Distribution Statement. Denote releasability to the public or limitation for reasons other than security for example "Re-
lease unlimited". Cite any availability to the public, with address and price.
19 & 20. Security Classification. Do not submit classified reports to the National Technical
21. Number of Pages. Insert the total number of pages, including this one and unnumbered pages, but excluding distribution
list, if any.
22. Price. Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
FORM NTIS-39 (REV. 3-721 USCOMM-OC MgS2-P72
------- |