EPA-600/7-82-063
November 1982
EVALUATION OF POTENTIAL VOC
SCREENING INSTRUMENTS
by
Kenneth T. Menzies
Rose E. Fasano
Arthur D. Little, Inc.
Cambridge, MA 02140
Contract 68-02-3111
Task Number 121
EPA
Project Officer
Merrill D. Jackson
Technical Support Staff
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
PB83-139733
Evaluation of Potential VOC
Screening Instruments
Arthur D. Little, Inc.
Cambridge, MA
Prepared for
Industrial Environmental Research Lab,
Research Triangle Park, NC
Nov 82
U.S. Department of Commerce
National Technka! Information Sen*
NT1S
-------
. TECHNICAL REPORT DATA
fftetlr rtod luitnictioni on tht nvmt btfort completing)
It. REPORT NO. 2.
IEPA-600/7-82-063
14 TITLE AND SUBTITLE
(Evaluation of Potential VOC Screening Instruments
i
17. AUTMCRIS*.
1 Kenneth T. Menzies and R.E. Fasano
1 ' '
J9 PERFORMING ORGANIZATION NAME AND ADDRESS
Arthur D. Little, he.
I Acorn Park
1 Cambridge, Massachusetts 02140
117 SPONSORING AGENCY NAME AND ADDRESS
1 EPA Office of Research and Development
1 Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
is. SUPPLEMENTARY NOTES j£RL-RTP project officer is Merrill
919/541-2559.
3. RECIPIENT'S ACCESSION NO.
PB83 139733
V REPORT DATE 1
November 1982
ft. PERFORMING ORGANIZATION CODE 1
8. PERFORMING ORGANIZATION REPORT NO. 1
C-82-480-03
10. PROGRAM ELEMENT NO. I
11. CONTRACT/GRANT NO. 1
68-02-3111, Task 121
^fcTiSTjW'-W'H
1«. SPONSORING AGENCY CODE 1
EPA/600/13
L D. Jackson. Mail Droo 62. 1
Reproduced from grrel 1
best available copy, TJfflRJ^ ^\
• » -> «•
[screening instruments for analysis of volatile organic compounds (VOCs). An initial
I review of available portable VOC detection instruments indicated that detectors
I operating on several principles (i.e., flame ionization, catalytic combustion, photo-
I ionization, infrared absorption, and thermal conductivity) might be useful for VOC
analysis. However, flame ionization and catalytic combustion devices evaluated pre-
viously showed poor sensitivity for highly substituted aliphatic and aromatic organic
compounds. Instruments utilizing photo ionization and infrared may be able to meet
necessary criteria for practical and accurate VOC analysis of highly substituted or-
ganics. Therefore, three commercially available instruments (i.e., HNU PI-101,
AID 580, and Foxboro Miran 80) were modified and evaluated for 32 such compounds
in concentrations of 100-10,000 ppmv. Results show that photo ionization may be
suitable for general VOC screening, but a reliable instrument/dilution system does
not exist. Infrared absorption will apparently not provide suitable general VOC
screening, but may be useful for analyzing some classes of organic compounds.
117.
KEY WORDS AND DOCUMENT ANALYSIS
\i DESCRIPTORS
[Pollution Photochemical Reac-
1 Analyzing tions
Organic Compounds Infrared Radiation
Volatility Electromagnetic Ab-
Selection sorption
Photo ionization Processing
Leakage
[19 DISTRIBUTION STATEMENT
Release to Public
b IDENTIFIERS/OPEN ENDED TERMS
Pollution Control
Stationary Sources
Volatile Organic Com-
pounds (VOCs)
Screening Instruments
Fugitive Emissions
IB. SECURITY CLASS iThil Rtportl
Unclassified
JO SECURITY CLASS tThil pogt)
Unclassified
o. COS ATI 1 if Id Croup |
13B
14B 07E 1
07C 20F
20M
14G 20C
07B 13H
71 NO. Of PAGES 1
96
37. PRICE I
CPA Form 2210-1 (»-7J)
-------
ACKNOWLEDGMENT
This report Is submitted in partial fulfillment of the requirements of
EPA Contract No. 68-02-3111, Technical Directive 121. The authors wish
to acknowledge the support of she Project Officer for this work,
Merrill Jackson of the Technical Support Staff, IERL-RTP. Robert
O'Neil, Kenneth Alder and Koyong Cho also contributed to the
project.
ii
-------
TABLE OF CONTENTS
Acknowledgement ji
List of Tables and Figures iv
ABSTRACT 1
1. INTRODUCTION 2
2. INSTRUMENT SELECTION 6
A. General Rationale 6
B. Unit Selection 18
C. Unit Modification 21
3. COMPOUND SELECTION 25
4. EXPERIMENTAL PROCEDURES 30
A. Introduction 30
B. Instrument Operation 30
C. Preparation of Gas Standards/Samples 30
D. Calibration Protocol 33
E. Instrument Sampling 33
F. Data Analysis 34
5. TEST RESULTS 35
6. DISCUSSIONS AND CONCLUSIONS 63
A. Photoionization Detection 63
B. Infrared Detection 64
C. Conclusions 71
7. LITERATURE CITED 72
8. APPENDICES 72
A. HNU Systems, Inc.- Model PI-101 A-l
B. AID, Inc. -Model 580 B-l
C. Foxboro/Wilks, Inc. - Miran 80 C-l
ill
-------
LIST OF TABLES AND FIGURES
Table No. Page
1 Compounds with Response Factors
Equal to or Greater than Five 3
2 Portable VOC Detection Instrument
Certification 4
3 Portable lonization Detectors 7
4 Portable Infrared Instruments 9
5 Portable Combustibles Analyzers n
6 Potential VOC Detectors 19
7 Additional Compounds to Test 26
8 Synthetic Organic Compounds with Vapor
Pressure Greater than 0.3 kPa (20°C)
and Not Tested Previously 27
9 Compounds for Evaluation 29
10 Calibration Scheme for Miran-80 31
11 Response Factors on PI-101 36
12 Response Factors on Miran 1A/80 39
13 Calibration Data for Miran 80 , 66
«
14 Substituted Aliphatic Compounds with
Response Factors Less than Twenty at
3.3 ym 68
15 Chlorinated Compounds and Response
Factors at 13.5 ym 69
ivr
-------
LIST OF TABLES AND FIGURES (Continued)
Table No. Pa^e
Al Specifications for Model PI 101
Photoionization Analyzer A-2
A2 Brief Description of Instrument
Controls and Functions A-4
Cl Specifications C-2
C2 Systems Specifications C-4
Figure No.
1 Photoionization Efficiency Curves as
a Function of Photon Energy for an
Aromatic Hydrocarbon and an Alkane 14
2 Instrument Response vs lonization
Potential for Several Classes of
Compounds 15
3 HNU Systems, Inc. Dilution Probe « 22
4 AID, Inc. Dilution System 23
5 Calibration Curve for 1,2-Dichloro-
ethane for the HNU Instrument 65
Al Control Panel Functions A-5
-------
ABSTRACT
This report describes the evaluation of potential fugitive source
emission screening instruments for analysis of volatile organic
compounds (VOC). An initial review of available portable VOC detection
instruments indicated that detectors operating on several principles,
i.e., flame ionization, catalytic combustion, photoionization, infrared
absorption and thermal conductivity, might be useful for VOC analysis.
However, flame ionization and catalytic combustion devices evaluated
previously have shown poor sensitivity for highly substituted aliphatic
and aromatic organic compounds. Instruments operating on the photo-
ionization and infrared principles may be able to meet necessary
criteria for practical and accurate VOC analysis of highly substituted
organics. Therefore, three commercially available instruments were
selected, modified, and evaluated for 32 such compounds in the
concentration range of 100 ppmv to 10,000 ppmv. The results indicate
that the photoionization principle may be suitable for general VOC
screening but a reliable instrument/dilution system does not exist at
present. The infrared absorption principle will apparently not provide
a suitable general VOC screening device but may be useful for analysis
of some classes of organic compounds.
-------
1. INTRODUCTION
The U.S. Environmental Protection Agency has issued performance standards
and guidelines1 to limit emissions of volatile organic compounds (VOC)
from several stationary source categories such as surface coating oper-
ations. These guidelines apply to industries which emit significant
quantities of air pollutants. It has become apparent that sources other
than classical point sources may also emit large amounts of VOCs into
the workplace and surrounding atmosphere. The EPA's Office of Air
Quality Planning and Standards (OAQPS) is, therefore, evaluating the
need for the control of fugitive emissions of VOCs from such sources
as valves, pumps and drains. As described in EPA Method 21, Determina-
tion of Volatile Organic Compound Leaks,2 technically and economically
feasible devices suitable for monitoring such leaks include only a few
portable detectors. These devices can be placed near possible points
of emissions and will respond to releases of the organic compounds.
Specific instruments suitable for this purpose include, but are not
limited to, catalytic oxidation, flame ionization, infrared absorption
and photoionization detectors.
Unfortunately, due to the chemical complexity of many fugitive VOCs
and the lack of universal sensitivity of these detectors, the detectors
previously evaluated cannot adequately measure the concentration of all
chemicals likely to be released. This fact has been documented3 for
two commercially available detectors using flame ionization (FID) and
catalytic combustion principles. Among 168 compounds tested, 23 showed
sufficiently poor response that the actual and measured concentrations
differed by a factor of greater than five (Table 1). The classes of
compounds which show the poorest agreement with the actual concentration
generally incorporate functional groups such as halides, hydroxyl
(alcohols), carbonyl (aldehydes, ketones) and carboxylate (acid) and
include both substituted aromatic hydrocarbons and lew molecular weight,
highly substituted aliphatic compounds.
Additional portable devices which respond accurately to these compounds
are needed for VOC screening. Instruments other than flame ionization
or catalytic oxidation detectors which might meet this goal operate
on the principles of infrared absorption, photoionization and thermal
conductivity.1*
The first step in this task was to select and procure one or more units
of those detectors which meet the specifications of Method 21. The VOC
instrument must be rugged, reliable, relatively inexpensive, portable
and easy to operate. Of course, it must respond to the organic compounds
of interest and be able to measure the leak definition concentration
specified in the regulations. According to Method 21, the instrument
must be intrinsically safe for operation in explosive atmospheres as
defined by the applicable National Electric Code. At this time, there
are few detectors which are "approved" for such an environment (Table 2).
-------
TABLE 1
COMPOUNDS WITH RESPONSE FACTORS
EQUAL TO OR GREATER THAN FIVE
OCPDB3 FID
ID No. Compound Names Response Factor
120 Acetophenone 10.98
— Acetyl-1-propanol, 3- 10.87
490 Benzcyl Chloride 6.40
790 Carbon Disulfide 571.92
810 Carbon Tetrachloride ' 21.28
•830 Chloro-Acetaldehyde 13.40
— Dichloro-1-propanol, 2,3- 61.51
— Dichloro-2-propanol, 1,3- 29.34
— Diisopropyl Benzene, 1,3- ^ 9.43
— Dimethyl Styrene, 2,4- * •„ 37.09
2060 Formic Acid 34.87
1221 Freon 12 9.65
2073 Furfural 7.96
2105 Glycidol 8.42
— Hydroxyacetone 8.70
2500 Methanol 5.69
— Methyl-2,4-pentanediol, 2- 96.34
2690 Methylstyrene, a- 10.24
1660 Monoethanolanlne 28.04
2770 Nitrobenzene 29.77
2910 Phenol 11.75
— Phenyl-2-propanol,2- 89.56
3291 Tetrachloroethane,l,l,2,2 6.06
aOrganic Chemical Producers Data Base
'Response Factor - A^tual Concentration
Measured Concentration
Cource: Reference 3
-------
TABLE 2
PORTABLE VOC DETECTION INSTRUMENT CERTIFICATION
Manufacturer
Model No.
Certification
Bacharach Instrument Co.,
Santa Clara, California
Century Systems,
Arkansas City, Kansas
HNU Systems, Inc.
Newton Upper Falls,
Massachusetts
Mine Safety Appliance Co.,
Pittsburgh, Pennsylvania
Survey and Analysis, Inc.
Norchboro, Massachusetts
TLV Sniffer
OVA-128
OVA-108
PI-101
AO
OnMark
Model 5
Intrinsically safe, Class I, Division 1, Groups C & D
Intrinsically safe, Class I, Division 1, Groups C & D, and
Class I, Division 2, Groups A & B
Intrinsically safe, Class I, Division 1, Groups A, B, C & D
Intrinsically safe, Class I, Division 1, Groups A, B, C & D
Intrinsically safe, Class I, Division 2, Groups A, B, C & D
Intrinsically safe, Class I, Division 1, Group D, and
Class I, Division 2, Groups A, B, & C
Intrinsically safe, Class I, Division 1, Groups A, B, C & D
Source: Reference A
-------
The second step in this task was to set up a laboratory system
capable of mixing known volumes of vapors with air and delivering the
mixtures of known concentration to the detectors. Tedlar bags and
a volumetric mixing system were selected for sample preparation since
they provide adequate accuracy/precision and require little cost ov
time to set up.
The third step in this task was evaluation of the detectors for
response to the compounds of interest. The response factors were
determined at several concentrations over the range of 100 ppmv to
10,000 ppmv. Measurements were limited to concentrations approaching
about 90% of the saturation concentration or 75% of the lower explosive
limit (LEL). In order to permit statistically valid interpretation of
the measured response factors, five replicate measurements at three
concentrations were conducted. Data analysis included calculations
of mean response factors and confidence intervals.
-------
2. INSTRUMENT SELECTION
A. •General Rationale
A recent summary of available portable VOC detection devices'* lists a
number of instruments operating on the following principles:
Flame lonization (FID)
Photoionization (PID)
Infrared Absorption (IR)
Thermal Conductivity (TC)
Hot Wire/Catalyst Combustion (Combustion).
The majority of available instruments operate on one of three principles,
i.e., FID, IR or Combustion (Tables 3, 4, 5). As noted above and in
previous work,3 two specific FID and Combustion devices show poor
sensitivity to several substituted organic compounds. Due to this
observation and with the understanding that other FID or Combustion
detectors available from different manufacturers probably do not differ
significantly in construction or sensitivity, alternative VOC screening
devices were evaluated. These were selected from instruments operating
on other detection principles, including photoionization, IR and thermal
conductivity. Fhotoionization or thermal conductivity detectors are
highlighted in Tables 3 and 5, respectively, while infrared detectors
are listed in Table 4.
The selection of potential VOC detectors from this list depends in
several criteria which are outlined in EPA Method 21. That is, an
instrument suitable for screening should hav». the following characteristics:
(1) Fast response (<30 seconds);
(2) Measurement range 100 to 10,000 ppmv;
(3) Similar responsiveness to a variety of organic vapors;
(4) Portable;
(5) Rugged;
(6) Reliable;
(7) Inexpensive;
(8) Easy to operate; and
(9) Intrinsically safe (as per National Electric Code).
Each of the first tiu.'r»e characteristics is of primary importance in
providing a practical instrument for VOC screening. Fast response time
is necessary for rapid screening of a large number of fugitive sources.
The specified measurement range is required by the need to limit
significant leaks of volatile organic compounds. Equal molar sensitivity
-------
TABLE 3
PORTABLE IONIZATION DETECTORS
Manufacturer
Ar^lvtUnl
IntftTuavnC
Inc.
.\votu1alr.
i
I
i
'
Sendtx.
Environmental
ar.d ?rccess
Instruaents
Otvlsioa
West Virginia
C«ntury
Svitcis, Inc.
Arkansas City,
Kansas
General Elec-
tric Instro-
nent Products
tynn.
Massachusetts
Heath Consul-
tants. Inc.
Jtotighlon,
Hassachasett
Hods I
Mo.
J)0b
end
JM
jjjC
511-12*
8401*
OVA-U8
OVA-128*
OVA-98
OVA-108*
TVH-1-
-
Detecto
PAR II1
Pollutant (•)
Detected
Nonncthnnc
total hydro-
carbons
Total hydro-
carbons
Total hydro-
carbons and
Individual
troop ounds
with CC
Total hydro-
carbons
Total hydro-
carbons
Total hydro-
carbons
Total hydro-
carbons
Total hydro-
carbons
Ualogcnated
ccapounds
Total hydro-
carbons
Principle
o(
Operation
FID
rio
FID/CC
no
FID
FID/CC
FID
FID/CC
capture0
FID
Cost
»
3711
3987
4968
3193
3300
4200
3500
420C
4060
2930
Weight
Ib
16.5
20.3
41
40
12
12
12
12
23
8
ppa
0-200 and
O-ZOOO far
Model 550;
n-rao nnj
o-lu.ooo lor
H.'iU-l iil
0- lo.uon.
0-2UOU, and
0-100
1-1000
0-10 and 0-100
0-1000
0-10,000
0-10,000
9 ranges:
0-1 through
0-10,000
1
0-10
0-100
0-1000
Acourscy
+3
+1
±-
+2
+2
±-
+2
+10
J
+4
1
Mensttlvltjr
0.1 ppa on
n Bviilt- of
0-200 ppa
1 pim on •
Mcala of
0-2000 ppa
0.03 ppa ea
propane
0.01 ppa
0.2 ppa
•ethane
0.2 ppa
•ethane
0.5 ppa
•ethane
0.5 ppa
nt thane
0.1 ppa
2 opa
2 ppa
3 PC"
Precision
X
+3
+3
±2
+2
+2
Raoponnc
Tine
3
3
8
2
2
120
13
Hois*
U.--S than
0.1 t-pa on
0-200 ppa
1 ppa on •
(IClllli Of
O-20OO ppu
±-s
Anbimt
Teop«rature
•c
0-/.0
1
I
0-40
5-43
3-40
-20 to 40
-20 to 40
-20 to 40
-20 to 40
0-5S
0-30
Drift*
s - +U «4>
• a *1£ 124)
» ^ _^»» V*— /
S - ±11 (1}
*- +11 (1)
s " +1!! (?)
s - +;z (i>
Be.ll.iil.
s - ^41 (7)
Reproduced from
bett available copy.
-------
TABLE 3 (Continued)
Manufacturer
HXV Syatema.
Inc.
N.-vton
Vrrtt r»ll».
X.iMi«4>-liuii»)tts
M*lre>v lab*.
SrrliUflcld,
VI r 'laid
!
Mlo* Safety
Appliance* Go.
Pittsburgh.
Pennsylvania
Survey and
Analysis. lac.
Scrthhcrought
Massachuietts
Model
He.
FI-1011
UC-500
Total BC
analyser
Snlfty
Model
A-SOO*
Follut«nt>l iiv light
rcnults In
lOlllailtiull
Total hydro-
cnrbons
Total hydro—
carboos
Total hydro-
carbon*
t
Principle
of
Operation
Fbotolon-
fttation
FID
FID
FID
Cost
»
3395
3S50
1695 for
baste
unit.
2295 for
entire
porta-
bility
package
Weight
Ib.
<9
40
35
17
Range
ppm
0-20.
0-200 and
0-2000
0-10.
0-50.
0-100. and
0-1000
0-4 and
0-12.000
0-10
0-100
0-1000
0-10.000
Ace iracy
I
-7
-7
11 on low
seal*
11
-55
-35
-3
+20
Sensitivity
1 ftm
0.1 ppa
CH,
2 rr*
Freelaion
X
*
±1
teeponsa
Tins
a
5
45
1
4
Kol.e
tO. 05 ppsi
aI4
±0.5
Aaiblent
Teaperatur*
•c
-IS to 50
10-40
4-45
0-50
Drift*
« • IX (7)
s - 32 (6)
for a, rAnga ef
0-20 ftm; no drift
for other rtnrjf*
S • 0.2 fpm (2*)
s - 0.2 ppa (24)
0-10 H-o
• - 0.51 (24)
• *20I (7)
- »12X (7)
- 0
- «; (7)
• 40.71 (7)
• 0
- 10.71 (7)
• 0 »
00
The letters "x" and "s" indicate zero drift and span drift. The nuabera of hours over which drift occur* 1* liven in pareotheali.
A charcoal tube 1* used to adsorb organic*, except methane, and • rang* of 0-10 ppn 1* available with the recorder. The instrument can be used
as alara by selling in 0-1000 range. Thia 1* a screening and leak detection device.
cThe following features are available! a range of 0-100 ppa with recorder only; Internal power, oxygen, and hyi'roge" •uppl.lea; a heated probe;
'and a battery-operated recorder with a range of 0-100 oV d.c.
Ca;abilities equal or exceed those of Models 550 and 555.
'Capabilities to detect higher concentrations require further investigation. This aay be suitable for ambient fir measurement only.
'Optional CC.
•optional CC.
A heated pletinua wire embedded In rubidium combusts incoming gases. Combustion of halogenated material* cause* electron* to flow from the
rubidlua. The electric*! flow measured 1* proportional to the amount of htlogenated material* present.
^erforaence characteristic* •• measured by KIOSB. Reference 6. Section 3.
^Inaccurate.
Source: lafarenc* 4
-------
TABLE 4
PORTABLE INFRARED INSTRUMENTS
>Uuuf«cturar
An.itv.u. Inc..
California
'
•
Ajtro
Resources
Corp. .
KC-JI ton.
Texas
Chrysler
HunttvlUe
Electronics
Division.
Hvn-. jvllle.
Al.ibio.1
foxboro
Analytical.
MU« Infra-
red Center,
S. Xorvalk,
Connecticut
Us Tech. Inc.
Mountain View,
California
Mndel
Mo.
A«-400b
S000e
III-C'1
HDpAe?
it
Atlasd
Hlran
-104*
Hlraa-lA
Dallde
Detector
Pollutant (e)
Detected
Individual
species
abnorblns. XR
Individual
hydrocarbons
Total hydro-
carbons
Total hydro-
carbons
Total hydro-
carbons
Any species
absorbing IK •
between 2.5
and 14.5 vm
in wavelength
Any species
absorbing I*
between 2.5
and 14.5 la
in wavelength
Ualogenated
hydrocarbons
Principle
of
Operation
IR
XX
IK
IX
IK
IK
IK .
Inhancement .
>f radiation
from a spark
>y halogen*
Coet
S
2393
or an
nalyier
•ensuring
nliiMlu
gas
»7*i
far an
analylur
Measuring
hrce
gate*
Model
403)
Not sold
in U.S.
but price
osy be
siallar to
Uran-lA
6600
112'
Weight
Ib.
25
20
20
30
24
32
13
Range
PPM
Specified by
customer, up
to 100Z
0-300 and
0-2000
0-300 and
0-2000
0-300 and
0-2000
ppa to
percent
-------
TABLE 4 (Continued)
Manufacturer
Infrared
InJuiit rtf».
Inc.
*alUtf«ii*r"'
1
1
|
Xlae Safety
Arpllacce Co.
Pittsburgh.
PeansTlvanla
Model
No.
Ut-711
1R-702*
1R-703*
f
U-703
LIRA 301*
P<
-------
TABLE 5
PORTABLE COMBUSTIBLES ANALYZERS
Manufacturer
M.ri«r«h
Ln»truovnt Co.
Sjnu Clara.
California
3U-:urlne
Industries,
Inc.
Malvern.
Pennsylvania
Control
Instru>e3ts
Cor;.
Fiilrflcld,
New Jersey
las Tech. lac.
Mcuntaln View.
California
International
Sensor
Technology
Santa Ana,
California
tine Safety
Appliance Co.
Pittsburgh,
Pennsylvania
HoJel
Ha.
C
L
l|b
TLV
Snlfferc
922
900. 900t
and 300RS
rrAPd
1177
. 1238
AC5100
20*
30*
40«
Pollutant (a)
Detected
Co&bustlbl*
gases
Coabustibla
tases and
vapors
Coabustibla
gases and
vapors
Fl unable
gases and
vapors
Combustible.
gases
Coabustibla
gases
Coabustlbl*
gasea and
vapora
Combustible
gases
Combuatlbla
gases
Coobustlbla
gases
rrlnclpla
of
Operation
Catalytic
combuatlon
Catalytic
combustion
Catalytic
coabustlon
Catalytic
coabuatlon
Thermal
coobustlon
Catalytic
combustion
Catalytic
combustion
Change in
resistance*
within
detector
Catalytic
combustion
Catalytic
conbustlon
Catalytic
combustion
Cost
233
160
279
896
495
685.
695 and
715
523
695
1200
for ppm
scale;
825
for LEL
acala
374
374
374
Volght
It
4
4
3
5
1.3
3
28
6
7
6
Range
O-IOOX LEL
o-ioo: LEL
0-100X LEL
0-100. 0-1000
t 0-10.000 ppm
0-100X LEL
0-100X LEL
0-100X LEL
0-1001 LEL
0-1001 LEL
4 0-500 ppm
LEL and ppa
0-100X LEL
0-1001 LEL
0-lOX and
0-100Z LEL
Accuracy
X
±3
J5t LEL
15X LEL
0
13
t2
t3
ganaltlvlty
2 ppm.
Precision
X
13
±1
12
±2
15
Rcaponaa
Time
a
3
5
<10
4
10
10 on LEL
scale and
60 on ppm
*Cfll4
Ambient
Tmmperatura
•c
O-50
-15 to 40
-15 to 40
0-52
0-40
-20 to 50
Drift*
c • <±5X (1 yr)
a - <15X (1 yr)
negligible in a
3 -month period
Reproduced from
available rnpy
-------
TABLE 5 (Continued)
Manufacturer
Survcv and
Analyrl*. Inc.
Xortliboro.
MlSfJchutettS
Tcicdvnc
Analytical
Instruments,
San Gabriel.
California
He-del
No.
OnKark
Model Sh
980
pollutant (a)
Detected
Combustible
f,as«a and
vapors
Total com-
bustibles
and oxygen
Principle
o(
Operation
thermal con-
luctlvlty
Catalytic
combustion
Coat
»
285
Height
Ib
<3
17
Range
fP*
0-5 and
0-100X
Scale: O-5X
'nethane; othera
are available
Accuracy
X
13
12
Senaltlvtty
0.5X of
Cull «cal.
fraclalon
X
a
lUfiponae
Tla«
a
<10
20
Ambient
T»l>uratur«
•c
0-50
0-50
Drift*
« - 1*1 (2 mo)
*lhe lettera "t" and "a" Indicate sero drift and apan drift given a* percent of scale over the time specified la par on the «l.
Catalytic coobuallon (hotwire) In low range and thernal conductivity la Ugh range.
cTie raage* of TLV caa be nultlplied by 10 with a dilution probe.
ihe FFAP uses a propane flaae to coabust aanple gas and la fully portable.
*Can be factory calibrated for five gases, such as pentane.
Calibrated to oeasure natural gas and petroleum vapors In air nlxturea.
^Factory calibrated for pentane. %
Nllacent for the 0-1001 scale are thereal aensors heated to 300* to 400*F. and fllaMntS) for the 0-51 seal* are
catalytic sectors heated to 1200* to 1300'F.
Source: Reference
-------
to compounds of widely differing functional character is not achievable
with currently evaluated instruments but is a desirable goal. The
other characteristics such as portability and instrlnsic safety are
also important but none should be considered individually critical to
the acceptance of a potential detector.
Assuming that characteristics of fast response and appropriate measurement
range are available in potential VOC detectors, the ability of the devices
to meet the criterion of similar responsiveness needs to be reviewed prior
to final instrument selection. Thus, the efficacy of various operating
principles to meet this criterion is discussed below.
(1) Fhotoionization
Fhotolonization detectors utilize ultraviolet radiation to ionize a. small
fraction of molecules introduced into an ionization chamber. The ioniza-
tion process is initiated by absorption of a photon of sufficient energy,
i.e., greater than the ionization potential, to remove an electron from
its ground state to infinity. A free electron and positively charged
ion are thus formed:
R + hv •*• R+ + e~
If the energy of the UV lamp is less than the ionization potential of
the compound, no ionization takes place. Ions formed in the detector/
ionization chamber may reach the electrodes under the influence of an
electric field and produce a small current. The number of ions which
reach the electrode is proportional to the concentration, although
only a very small fraction (M3.012) of Che molecules in the ionization
chamber are ionized by incident radiation. Depending on the character
of the electrons, e.g., sigma ys pi electrons, the yield of ions
(photoionization efficiency) may vary as a function of the energy of
incident photons (Figure 1). At present, UV sources are-available for
commercial instruments which emit photons of approximately 9 eV, 10 eV
or 12 eV. Based on the ionization potentials of organic compounds,5
it is apparent that certain classes of compounds, e.g., aromatics and
aliphatics greater than carbon number C7, can be ionized by a 10 eV lamp
while many substituted aliphatics require photons of at least 11 eV
(Figure 2). This observation leads to the conclusion that with sufficient
. energy most organic compounds can be ionized and detected. A practical
upper energy limit for VOC analysis is about 12 eV since the major
components of air such as nitrogen, carbon monoxide, carbon dioxide
and water have ionization potentials above this level. As well as the
ionization potential, the photoionization efficiency is Important since
this parameter determines sensitivity of the technique to different
compounds. A recent report6 indicates that the molar sensitivity of
aliphatic and oxygenated aliphatic compounds is several times less than
that of aromatic compounds if incident radiation is about 10.2 eV. In
fact, for aliphatic hydrocarbons of carbon number less than C8, the
relative sensitivity is less than one-tenth that for benzene. If
incident radiation is about 11.7 eV, the relative sensitivity of aliphatic
13
-------
RELATIVE
PHOTO-
IONIZATION
AROMATIC'
ENERGY OF
UV LAMP
ALKANE
J I
I I I
I I
95 10p
PHOTON ENERGY («V)
I I
; Figure 1. Photo ionizat ion efficiency! curves as a function of photon energy
f~~" for an aromatic hydrocarbon and an alkane (source: Ref. 5).
14
TYPING GUIDi: NHfltIT
-------
NITROGEN
HELIUM
CARBON MONOXIDE
CARSON DIOXIDE
-lYOROGEN
SULFUR OIOXIOE
ACETONITRILE
WATER
HEXANE
CYCLOHEXANE
METHYLBROMIOE
HYDROGEN SULFIOE
ETHYLENE
ETHYLENE OXIDE
ACETIC ACID
METHANE
/CHLOROFORM
/ CARSON TETHACHLORIOE
/ METHYLENE CHLORIDE
/ METHYL CHLORIDE
/ FORMALDEHYDE
/ FORMIC ACID
11.7—L ACRYLONITRILE
I METHANOC
/ METHYL FORMATE
IB.^^* NITROMETHANE
FREONS
PHOSGENE
ETHANE PROPANE BUTANES
I _L
CARBON OISULFIOE
AMMONIA
PMOSPHIN6
AASINf
ETHANOL
TRICHLOROETHYLENE
BUTYL ACETATE
OlETHYL ETHER
ACETONE
ACETALOEHYOE
VINYL CHLORIDE
VINYL BROMIDE
BROMINE
NITROBENZENE
/ BENZENE AND AROMATICS COMPOUNDS
/ AMINES
/ ORGANIC SULFUR
/ METHYL iOOIDE
/ PROPYL ETHE^
/ MIBK
1. OMF
A VINYL ACETATE
/ / PHENOL
' I ANILINE
9.5
CHLOROBENZENE
OCTANES
I
OECANES
I
10
IONIZATION POTENTIAL
H!
Figure 2. iistrument response vs. ionization potential for several classes of compounds (Ref. 5).
-------
and aromatic compounds is similar7 and perhaps within a factor of two.
Based on this assumption, a commercially available photoionization
instrument with a lamp of about 12 eV may provide a generally applicable
VOC detection technique.
(2) Infrared Detection
Typical nondispersive infrared devices operate by passing infrared
radiation through two separate absorption cells: ?. reference cell and
a sample cell. The sealed reference cell is filled with nonabsorbing
gas, such as nitrogen or argon. The sample cell is physically identical
to the reference cell and receives a continuous stream of gas b3ing
analyzed. Subsequently, the net radiation in the two beams are passed
into and absorbed in matched selective detectors (e.g., Luft detector)
containing the vapor to be detected. When organic vapors are present
in the sample cell, energy is absorbed, and the temperature and
pressure in the corresponding detector is reduced relative to that
in the detector, on the reference side of the analyzer. A diaphragm
between the two detectors is displaced and the amount of displacement
is detected, electronically amplified, and an output signal proportional
to concentration produced; In other NDIR systems, narrow bandwidth
filters which pass energy which corresponds to that absorbed by the
compound of interest are used along with simple solid state IR detectors.
In both cases, interference from compounds with overlapping absorption
bands is possible. More importantly, the maximum absorbing wavelength
for different organic species in the sample gas may not correspond to
the maximum absorbing wavelength of the calibration compound used in
the detector. Within reason, several different calibration compounds
could be used in the detector to improve responsiveness for several
compounds. Alternatively, by selection of a single narrow bandwidth
filter with a wavelength corresponding to a general alir' atic C-H
stretch, many aliphatic hydrocarbons might be detected _te uniformly.
Based on the maximum absorption wavelength of aromatic hydrocarbons, a
separate filter or cell would be needed for this class of compounds.
In practice, the specificity of the detection principle has precluded
the manufacture of an NDIR device suitable as a general (i.e., both
aliphatic and aromatic) organ/.c vapor detector.
An alternative IR detection scheme involves dispersive infrared analysis
in which the specific wavelength absorbed by the organic vapor of interest
is passed through a single sample cell. In this case, selectivity is
provided by a monochromatic light source rather than a. selective detector.
Such a device is inherently more selective than an NDIR and thus may be
less appropriate as a VOC screening device. However, by successive,
rapid monitoring of IS. absorption at several selected wavelengths
corresponding to the mavimim absorption wavelengths for several organic
functional groups, e.g., aliphatic CH, aromatic CH, C-C1, OO, it may
be possible to identify and quantify a wide variety of organic vapors
in a fugitive emission source.
16
-------
(3) Thermal Conductivity
Thermal conductivity (TC) of gases and vapors provides a physical
method used extensively in gas chromatography where mixtures of compounds
are resolved into individual components and quantified. The detector
responds proportionally to a change in thermal conductivity of the trace
gas in a background (helium, for example). The physical property of
thermal conductivity is not specific to any class of compounds. In fact,
the conductivity of many gases with the exception of hydrogen and helium
is quite similar as shown below:
Thermal Conductivity
Gas " flO"5 Cal/SeC"Cm2/(°C/cm)1
Air 7.5
Hydrogen 53.4
Helium 41.6
Nitrogen 7.5
Oxygen 7.6
Carbon Dioxide 5.3
Methane 10.9
Ethane 7.3
Propane 6.3
As molecular weight increases, the thermal conductivity decreases
somewhat, but for organic compounds of interest in fugitive sources
the thermal conductivity does not differ from that of air by more
than a factor of two.8 Thus, the concentrations of organic contaminants
in air must be very large (e.g., 1-100%) in order for the vapor to be
detected against the air matrix which has a similar thermal conductivity.
This severely limits the usefulness of TC detectors for VOC screening.
Two additional problem.?, exist for TC detectors in this application.
First, inorganic gases (e.g., C02, H20, HC1) which may be present in
fugitive sources will be detected along with any organic vapors. Second,
most thermal conductivity detectors consist of heated wires which are
subject to degradation by oxidizing or humid atmospheres or by the
presence of chlorinated hydrocarbons. Thus, the practical application
of TC detectors to VOC screening is in doubt.
(4) Other Detection Schemes
Two other detection principles, implicitly limited in response only
to chlorinated organic compounds, may be appropriate for this compound
class. They include electron capture, which is used commonly in gas
chromatography, and enhancement of radiation from halogens by a spark
source. Each scheme depends on a unique property of halogens:
electronegativity and electron energy levels, respectively. Their
17
-------
application for organic detection is very limited and their practical
use for VOC screening is in doubt.
B. .Unit Selection
On the basis of the factors discussed above, both the IR and photoion-
ization principles might be suitable for general VOC screening. However,
a comparison of the specifications 'of commercially available instruments
(Tables 3, 4) operating by these principles and the criteria of Method 21
lead to a rather unfortunate conclusion: no commercial instruments of
these types are available for VOC screening. In terms of a desire to
expand the list of potential detectors, such a finding is unsatisfactory.
What criteria led to this finding? The most obvious answer is the
requirement for an intrinsically safe device. No IR or FID devices are
certified for use in Class I, Division 1 environments (Table 2). One
PID device is certified for use in Class I, Division 2. It should be
noted, however, that in the future other devices may be modified so
as to meet Class I, Division 1 certification. Alternatively, the use
of an instrument only in less hazardous environments may not be
considered as particularly restrictive. For these reasons, the criterion
of intrinsic safety was given lesser significance and not used to rule
out potential devices for screening in this program.
The other criteria listed previously were ranked in approximately
descending order of importance. Thus, a response time of less than
30 seconds was given the highest importance. In fact, fast response
time is very important to practical measurement of VOC leaks and several
instruments with faster than 30-second response time are available.
Thus, it was decided that this criterion must be met by any instrument
to be evaluated. The criteria of portability, ruggedness and ease of
operation are also important but were not chosen as absolute selection
criteria. Portability can be evaluated subjectively and devices
operated with automobile batteries placed on a small cart may be con-
sidered to have adquate portability. Devices operated with AC power
are less practical in many industrial environments.
Tables 3 and A and additional manufacturers' literature were reviewed
using the modified criteria described above. A list of potential VOC
detectors was developed (Table 6). This list includes instruments
which operate on photoionization and IR principles and which meet most
of the Method 21 criteria. It is obvious that not all of the IR
instruments that meet most of the criteria are included. Since the
goal of this study is to evaluate the usefulness of the IR principle
rather than all available IR devices, only selected IR devices suitable
on the basis of the criteria for VOC screening were included.
The list, therefore, includes a dispersive IR device and NDIR devices
with or without solid state detectors. The NDIR devices may be useful
for a specific group of organics, i.e., aliphatic hydrocarbons. Also
included are two devices operating on other detection principles, i.e.,
ion capture and UV spark. These instruments are only useful for a
18
-------
TABLE 6
PO
Instrument Principle Range (ppra)
HNU Systems, Inc. Photo- 0-20, 0-200,
(PI-101) ionization 0-2000
CENTIAL VOC DETECTORS1*
Wgt
Response (a) (Ibs) Pump Power
5 9 Yes Battery
General Electric Ion Capture 0-1 •*• 0-10,000 120 23 Yes No
(TVM 1) (Halogenated'
Gas Tedh, Inc. UV Spark 0-100, 0-10,
(Halid'e) (Halogenated)
Foxboro, Inc. IR ppra •> %
Miran 1A/80
103 IR ppm -*• %
000 5 13 Yes AC
1-40 32 Yes Car Battery
(37 Ibs)
1-40
ANARAD, Inc. IR 100-10,000 ppm 5 24 Yes AC
AR-400 <5»J
Infrared IR 1000 ppra,
Industries Solid State 100% LEL
IR-711 <5n
AID, Inc. Photo- 0-200
(580) ionization 0-2000
5-120 9 Yes Battery
2 8 Yes Battery
Intrinsically
Safe
Yes
No
No
No
No
Yes
Pending
-------
specific class of organics, i.e., halogenated hydrocarbons, but were
included due to the widespread industrial use of these solvents.
The final instruments selected for evaluation were as follows:
Instrument Manufacturer Principle of Operation
Model 580 AID, Inc. Photoionization
PI 101 HNU Systems, Inc. Photoionization
Miran 80 Foxboro/Wilks, Inc. Dispersive Infrared
The rationale for the selection of these instruments is based on both
suitability and availability. As noted previously, photoionization
may be a particularly suitable VOC detector for aliphatic, aromatic,
and substituted organic vapors. Either of the two commercially
available, portable instruments meet most Method 21 criteria and may
be suitable for evaluation. Both photoionization instruments were
selected on the basis of the following considerations:
1. Each was supplied free of charge; ' „
«
2. Each was modified by the manufacturer with a dilution
probe (see discussion in Section 2c);
3. The location of HNU Systems, Inc. provided the potential
for rapid modification/repair; and
4. Negligible additional time would be expended in
evaluating both devices due to similarity of design.
I
The Miran instrument is the one available infrared device which pennies
a selection of wavelength as opposed to selection of a test compound in
a reference cell. This option permits the rapid (a few seconds)
assessment of the suitability of several wavelengths for the measurement
of the substituted organics of interest. Specific examples are an
aliphatic C-H stretch, arociatic C-H stretch or a C=0 stretch. Other
portable IR devices utilize a filter at one specific wavelength band
corresponding, for example, to an aliphatic C-H stretch. Thus, they
have an inherent selectivity against aromatic or substituted species.
Since all three classes of compounds are of interest, the latter devices
are not preferred as VOC screening devices. It is possible that one
wavelength may be suitable for analysis of a wide variety of organic
vapors. If that is the case, other IR instruments could potentially
be used for VOC leak detection. In summary, the Miran 80 (with
associated microprocessor) permits the most rapid and cost-effective
assessment of the IR principle as a general VOC detector. Other IR
devices were, therefore, not evaluated.•
The halocarbon specific detectors, i.e., General Electric TVM-1 and
Gas Tech Halide Detector, were not selected for evaluation despite
20
-------
the potential usefulness of such a device in environments subject to
halocarbon solvent combination. Neither the GE TVM-1 nor an equivalent
model is now sold. The Gas Tech device could not be modified to meet
the intrinsic safety requirements of Method 21 due to the presence of
a spark in the detector section.
No other instruments appeared to have a reasonable expectation of
meeting the Method 21 criteria and of providing a significantly
different performance than those devices evaluated previously or
selected for evaluation in this study.
C. Unit Modification
Both photoionization devices operate with a maximum quoted linear range
of 0-2000 ppmv. In fact, the linear range is frequently reported to be
only about 1500 ppmv. Since the ma yttrium concentration of concern in
VOC screening is 10,000 ppmv, dilution of sample air is necessary for
both instruments to operate in the linear range. Both HNU Systems, Inc.
and AID, Inc. provided their instruments with dilution systems designed
in their respective laboratories. The HNU Systems, Inc.'design (Figure 3)
consisted of (1) a fine bore restrictor which limited the flow of sample
air, and (2) a charcoal tube which passed an excess (lOx) of hydrocarbon-
free air (methane is not removed but does not respond in the detector).
The sample stream is thus diluted about 1 to 10. The AID, Inc. design
(Figure 4) consisted of a pump and needle valve which diverted 90% of
the incoming sample air through a charcoal tube and 10% to the norm*' .
exhaust point. The hydrocarbon-free (except for methane) sample air ij
combined with the incoming sample stream and thus a continuous tenfold
dilution is provided.
Problems were observed with these dilution systems and the UV lamps
provided with both instruments. The absolute accuracy of the dilution
ratios is in some doubt since independent flow rates were difficult to
measure. The UV lamps provided with both instruments were subject to
degradation during the life of the study. In fact, the 11.8 eV lamp
supplied with the AID, Inc. device failed during the study and,
unfortunately, a replacement could not be obtained in time to collect
useful data with this instrument. The 11.7 eV lamp supplied with the
HNU Systems, Inc. device failed during the study and a replacement was
provided. The difference in energy output from the two HNU Systems, Inc.
lamps was large (i.e., a factor of three to ten depending on the age
of the lamp). This variation affected the linear range of the instrument
and created problems in obtaining consistent results. In some cases
with the new lamp, saturation of the detector occurred even with the
dilution probe attached to the instruments. The test results reported
in Section 5 must, therefore, be carefully interpreted and conclusions
narrowly drawn.
The Miran 80 operates over the concentration range from ppm to percent.
The wide dynamic range is provided by a cell in which the pathlength of
IR radiation can be changed by optical folding of the incident beam.
21
-------
FigureS. HNU Systems, Inc. dilution probe.
22
-------
Reproduced from
best available copy.
Figure 4. AID, Inc. dilution system.
23
-------
Ac the concentration range of interest, i.e., 100 ppnv to 10,000 ppmv,
the incident beam traversed a distance of about 0.75 m. At this
pathlength, the full-scale absorbance for vapors of interest at a
concentration of 10,000 ppmv was about one absorbance unit. Once the
cell pathlength was set, no other modifications of operating conditions
were required.
24
-------
3. COMPOUND SELECTION
As noted in the Introduction, 168 compounds had previously been tested
for response factor on two commercially available VOC detectors.3
Twenty-three showed sufficiently poor response that the actual and
measured concentrations differed by a factor of greater than five
(Table 1). The classes of compounds showing poor agreement were
generally highly substituted aliphatic and aromatic compounds and those
compounds incorporating functional groups such as carbonyl and hydroxyl
groups.
These 23 compounds were selected for testing on the alternative VOC
screening devices to be evaluated in this study. Several other compounds
(Table 7), which were not evaluated in the previous work, were added to
the list.
These compounds (Tables 1 and 7) include only a portion of those
commonly used in chemical production. At the request of OAQPS, other
industrial compounds which have a vapor pressure greater than 0.3 kPa
but which were not considered previously have been tabulated (Table 8).
This extensive list of 76 compounds includes many species for which an
FID or catalytic combustion detector would respond well. However,
others are highly substituted compounds which will probably not give
adequate response on these two detectors. Selected substituted
compounds froa Table 8 were included in the detector evaluation.
The selection criteria required a response to several questions:
1. Are substituent groups present or absent? If absent, don't test,
2. Are the compounds similar (functionally and/or isomerically)
to others previously evaluated? If they are, don't test,
3. Are response factors on an FID instrument likely to exceed
five? If not, don't test, and
4. Do the compounds pose a serious health hazard to laboratory
personnel? If they do, cautiously consider evaluation.
As a result of the responses, the compounds were separated into two
groups: compounds that should and those that need not be analyzed.
Within the first group, the compounds were prioritized on the basis
of (1) their similarity to other vapors to be analyzed (for example,
positional isomers of compounds selected for testing were given lower
priority), and (2) their health hazard (extremely toxic compounds with
little commercial application or likelihood of release &nd which require
complex/expensive Katun-ing were given lover priority). • -
The compounds selected for evaluation in this program are listed in
Table 9. They are listed in the approximate order of testing.
25
-------
TABLE 7
ADDITIONAL COMPOUNDS TO TEST
OCPDBa
ID No. Compound Name
1660 Ethanol
— Formaldehyde
1235, 1236 Ethylene Dichloride (Dichloroethylene)
— Chlorinated Ethanes (C2H.C1, etc.)
Chlorinated Methanes CCH-Cl, etc.)
aOrganic Chemical Producers Data Base
26
-------
TABLE 8
SYNTHETIC ORGANIC COMPOUNDS WITH VAPOR PRESSURE
GREATER THAN 0.3 kPa (20°C) AND NOT TESTED PREVIOUSLY
Compound. Vapor Pressure (kPa)
Acctal • 2.7
Acetaldehyde 98.6
Acrolein 29.7
Acrylic esters: methyl aerylace 8.7
Allyl chloride . A0.8
Arayl acetate 0.5
Amyl amine 2.0
Amyl chloride . 2.2
Amyl mercaptans: 1-pentanethiol 1.5
2-methyl-2-butanethiol 5.0
2-methyl-l-butanethiol . 2.0
3-mechyl-l-butanethiol - 2.1
Aniline hydrochloride ' - 4.0
Benzyl benzoate 0.5
Chlorobenzoyl chloride 2.6
Chlorodifluoromethane G
Chlorodifluoromethane G
Chloroprene 23.1
Chlorotrifluproraethane G
Cyanoacetic acid- 2.9
Cyanogen chloride G
Cyclooctadiene (1,5-) 3.3
Dichlorodifluorotne thane G
Dichloropropene (isomers) 70.3
Diethylamine 24.0
Difluoroethane G
Diketene 1.6
Dimethylamine G
Dimethyl ether G
Dimethyl sulfide 48.8
Dioxolane ' 9.3
Ethyl bromide 51.4
Ethyl chloride G
Ethylene ehlorohydrin 0.7
Ethylene dibroraide 1.5
Ethylene glycoldimethyl ether 7.8
.' monoethyl ether 0.6
monoethyl ether acetate 0.3
monomethyl ether 1.2
monomethyl ether acetate 0.3
monopropylether 0.5
Ethyl orthoformate 0.4
Glycerol dichlorohydrin 0.9
SOURCE: REFERENCE 3 27
-------
TABLE 8 (continued)
Compound Vapor Pressure (kPa)
Hydrogen cyanide 66.6
Isoamylene G
Isobutanol 1.0
Isobutyl acetate 1.9
Isobutyraldehyde 18.3
Isopentane 76.5
Isopropylamine 29.3
Ketene G
Methallyl chloride (isoraers) 70.3
Methylamine G
Methyl bromide G
Methylene chloride 46.1
Methyl isobutyl carbinol ,0.6
Methyl isobutyl carbinol 2.1
Neopentanoic acid 1.1
Nonene 0.5
Paraldehyde 3.4
Pentene 70.3
Perchloroethylene 1.8
Perchloromethyl mercaptan 0.6
o-Phenylene diaraine 1.18
Phosgene t G
Propylamine 33.0
Propylchloride 37.3
Propylene chlorohydrin 0.7
Propylene dichloride 5.3
Quinone 5.0
Tetramethyllead 3.0
Toluene sulfonic acids 70.3
Toluene sulfonylchlorides 70.3
Trichlorofluoromethane 79.5
Trichlorotrifluoroethane 31.8
Trimethylamine G
SOURCE: Vapor pressure distribution of synthetic organic chemicals,
Weber, R.C.; P. Parker and M. Bowser. IERL, Cincinnati,
Draft Report, November 1980.
G - Gas; VP >101.3 kPa.
28
-------
TABLE 9
CCMPOCHDS FOR EVAtGATXOS
2*3-
1,3-
1,3-
Carbon Bisulfide
Carbon Tetrachloride
Chloro-Acetaldehyde
Dichloro-1-propanol,
Dichloro-2-propanol,
Diisopropyl Benzene,
Dimethyl Styrene, 2,4-
Formic Acid
Freon 12
Methanol
Methylstyrene, o-
Tetrachloroe thane, 1,1,2,2
Ethanol
Formaldehyde
Ethylene Bichloride (Dichloroethylene)
Chlorinated Ethanes ^^Cl, etc.)
Chlorinated Methanes (C^Cl, etc.)
Acetophenone
Benzoyl Chloride
Furfural
Monoe thanolamine
Nitrobenzene
Phenol
Acetyl-1-propanol, 3-
Glycidol
Hydroxyacetone
Methyl-2,4-pentanediol, 2-
Phenyl-2-propanol, 2-
An^^•fw» Hydrochloride
DiHnoroetbaae
Diketene
Dimethylsulfide
Glyceroldichlorohydrin
Paraldehyde
Perchloromethylmercaptan
Propylene Chlorohydrin
Toluenesulfonic Acid
Toluene Sulfonylchloride
Ethyleneglycoldimethyl Ether
Ethyleneglycolmonoethyl Ether Acetate
1-Pentanethiol
Acetal .
Chlorobenzoylchlorl.de
Chlorodifluoromethane
Chlorotrifluoromethane
Trichlorofluoromethane
Trichlorotrifluoroethane
Cyanoacetic Acid
Neopentanoic Acid
Any Inercap cans
2-methyl-2-butanethiol
2-methyl-l-butanethiol
3-me thy1-1-butanethio1
Glycols
Ethylene Glycolmonoethyl Ether
Ethylene Glycolmonomethyl Ether
Ethylene Glycolmonomethyl
Ether Acetate
Ethylene Glycolsonopropyl Ether
Glycolaethyl Ether (Dioxolane)
29
-------
4. EXPERIMENTAL PROCEDURES
A. Introduction •
Determination of response factors required initial calibration of the
VOC detectors with a gas or gases of known concentration. Methane
had been used previously for calibration of FID and combustion analyzers.3
However, the photoionization instruments do not respond to this compound
and the multiple wavelength analysis by dispersive infrared spectrometry
cannot be carried out by the use of methane as a single calibration gas.
Therefore, 1,2-dichloroethane was selected as the calibration gas for
the photoionization detectors. This compound can be detected by the
instruments and it has a response factor of about one (compared to
methane) when analyzed on an FID instrument. As a result, data collected
in this study may be comparable to data collected in a previous EPA
study.3
Several calibration compounds were used for the IR evaluation•since
several wavelengths were scanned in the dispersive infrared•instrument
to determine if any wavelength gave similar response factors for all
the compounds of interest. The wavelengths were selected to correspond
to key functional groups of the test compounds to be analyzed. The
•aveleagtas, functional groups and calibration cospouods are listed
in Table 10.
Once calibrated, the instruments were used to analyze the test compounds
at three concentrations over the range of 100-10,000 ppmv. The response
factor was determined by ralculgting the ratio of the actual concentration
to the concentration indicated by the instrument. The following sections
describe the procedures involved in calibration and operation of the
instruments, preparation of test gas samples, and calculation of response
factors.
B. Instrument Operation
Three instruments were selected for evaluation as VOC screening devices.
Two iastmaeats operated on the photoionization principle, i.e., AID, lac.
Model 580 and BSD Sysceas, Inc. Model PI-101. The other in instruaeat
was the Foxboro/lrilks, Inc. Miran 80 which operates on the principle
of dispersive infrared spectrophotometry. Details of the operation of
each instrument are given in the appendices.
During the. tests, the lamp in the Model 580 failed and a replacement
could not be obtained. Therefore, only the PI-101 and Miran 80 were
evaluated.
C. Preparation of Gas Standards/Saaples
Gas mixtures tested in this study were prepared in Tedlar gas sampling
bags of a nominal 25 liter volume. These bags provide a relatively
30
-------
TABLE 10
CALIBRATION SCHEME FOR MIRAS-80
Wavelength (urn) Functional Groups(3) Calibration Compound
3.3 Aromatic & Unsaturated C-H Toluene
3.4 Saturated C-H Pentane
3.6 Aldehyde C-H Butyraldehyde
4.0 Reference Wavelength Air
5.7 Carhonyl C»0 Acetone
6.35 Aromatic C-C, .Toluene
conj C«C (also N-H, C-S)
8.8 Ether C-O-C Diisopropyl Ether
9.5 Alcohol C-O-H Isopropanol
13.5 C-C1 1,2-dirhl oroethane
-------
inert surface to preclude adsorption, reaction, or permeation. They
also perait visual inspection of the bag interior to provide as.
indication of sample condensation or reaction. The bags are equipped
with two valves to facilitate flushing of sample gas and a septum
to permit injection of sample liquid with a syringe.
Gas samples were prepared by the following procedure:
1. Flush and evacuate bag three tizes with hydrocarbon-free
air (i.e., until no hydrocarbons are detected on each
instrument).
2. Fill bag with 20.0 L of hydrocarbon-free air.
3. Inject a known voluae of test coopcuod into the bag.
4. Permit at least one hour equilibration to insure adequate
evaporation and mixing. •*
5. Draw gas saaple froa bag with each instrument.
The hydrocarbon-free air was prepared by passing house compressed air
through silica gel, charcoal and a high efficiency filter. A known
volume (20.0 L) of air was introduced into the hose fitting of the
Tedlar bags through a calibrated rotameter. The volume was calculated
on the basis of rotaaeter flow rate (L/xdn) and duration of flow (zin)
a"3 y?>rm
©3SSTT3C.
Known volumes of the test compounds (all liquids) were injected through
the septa of the bags with Hamilton microliter syringes. The volumes
injected were in the range of 10 to 100 uL. The mass of material
injected was calculated from density data. Manual manipulation of
the bag, visual observation and at least one hour equilibration period
were used to ensure complete mixing.
target concentrations prepared for each cospound were 500, 1000,
5000 and 10,000 ppav. In several cases, it was not possible to prepare
the higher concentrations due to the low vapor pressure of the coapound
or due to safety reasons, that is, such a concentration would exceed
the lower explosive limit. In these cases, a concentration of 100 ppn
was often prepared. For each target concentration, the required volume
of liquid was calculated and measured in a microliter syringe. The
volume required to produce the test concentration was calculated
according to the following equation:
VL - (C) (VA) flay (BP)/ (62.36 x 106) (OL) (T)
where V Is the voluae of liquid ccapound in the syringe in oilliliters,
\f
C is the target concentration of compound "L" in parts per aillion
by volume (ppmv),~ -
32
-------
VA is the sample bag volume in liters (i.e., 20.0 L),
MW is the molecular weight of compound "L,"
Id
BF is the barometric pressure in mm Hg,
62.36 x 10 is a combined constant with the unit (liters) (mm Hg)
(g-mols) (°K). This constant incorporates the ideal gas
volume of 22.4 liters per g-mol, standard temperature and
pressure, and a factor of 10& to go from volume fraction
to ppmv,
PT is the liquid density in g/ml, and
Ij
T is the laboratory temperature in °K.
D. Calibration Protocol
Each instrument was initially calibrated (spanned) with a gas sample
prepared in triplicate at a concentration of 10,000 ppmv. Calibration
curves vere then prepared by introducing samples of the calibration
gas, prepared in triplicate at five concentrations over the range of
100 to 10,000 ppmv, into the instruments and recording the response.
The PI-101 was calibrated with 1,2-dichloroe thane while the Miran 80
was calibrated at individual analytical wavelengths with the compounds
listed in Table 10.
During subsequent analysis of each test compound, the HNU PI-101
instrument was spanned with an 8040 ppmv 1,2-dichloroethane certified
standard provided by Scott Specialty Gas, Inc. of Plumsteadville,
Pennsylvania. This span was carried out just prior to analysis of
each set of sample bags for each test compound.
The Foxboro/Wilks Miran instrument was electronically zeroed and
spanned according to the manufacturer's instructions. This zero and
span check was carried out prior to analysis of each set of sample bags
for each test compound.
Z. Instrument Sampling
The procedure used to obtain response data involved the following
steps for each of five replicate sample bags for each of three target
concentrations of each test compound:
1. Span and zero instruments.
2. Connect bag to photoionization instrument (PI-101).
3. Observe instrument response and record three instrument
readings at equilibrium point.
33
-------
4. Remove bag and permit instrument response to return to
zero,
5. Repeat steps 1-4 with each sample bag.
6. Connect bag to infrared instrument (Miran 80); start pump.
7. Empty bag to approximately 10% of original volume; stop pump.
8. Record instrument response at each of eight analytical
wavelengths and one reference wavelength.
9. Repeat steps 6-8 with each replicate sample bag at first
target concentration.
10. Remove bag, start pump and rezero instrument on .zero air
(room air was adequate).
11. Repeat steps 6-10 for each target concentration.
F. Data Analysis
The response factor reported in the following test results section is
the number that, when multiplied by the apparent concentration based on
instrument response, yields the actual concentration as calculated to
exist in the gas bag sample. That is:
_ _ ,__N Actual Bag Concentration (C)
Response Factor (RF) - concentration Calculated from
Instrument Response
Response factors were determined at three actual concentrations,
i.e., generally 100, 500, 1000, 5000 or 10,000 ppmv. No attempt was
made to fit the three response factors for each compound to a
particular function. For some compounds, the response factor is
nearly identical for each concentration whereas for others it differs
dramatically and in a complex manner. The response factor for
individual compounds is, therefore, not reported for an observed instru-
ment response of 10,000 ppmv. Instead, the mean response factors
calculated from up to five replicate data points at each of the three
actual bag concentrations are reported along with the standard deviation.
Also reported is the 95% confidence intervals for the response factors
as calculated from Student's t-test.
34
-------
5. TEST RESULTS
The response factors for the compounds tested in this instrument
evaluation program are listed on the following tables. The response
factors for all 16 compounds tested on the HNU Systems, Inc. Model PI-101
are reported in Table 11. The data are limited due to several factors
including:
1. The low vapor pressure of many compounds of interest and thus
the low concentration prepared in the sample bag and found
in the dilutor outlet;
2. .The failure of the dilution system to operate suitably;
»
3. The declining intensity of the UV lamp; and • .
4. Saturation of the instrument's detector.
The response factors for 32 individual compounds tested on the Miran
are reported in Table 12. The response factors are reported only for
those wavelengths where the detector was sufficiently sensitive to
yield an absorbance value above the background noise. Where this is
not the case, the response factor would be much greater than ten.
The remainder of the 57 compounds listed in Table 9 were not tested
for several reasons including:.
1. Similarity with compounds tested, e.g., amylmercaptans
with pentanethiols and glycols with ethyleneglycol-
monoethylethe.r acetate,
2. Low vapor pressure, e.g., phenyl-2-propanol,
3. Solid state, e.g., phenol,
4. Reactivity in Tedlar bags, e.g., toluene sulfonylchloride,
5. Poor availability of gaseous fluorinated methanes, and
6. Difficulties with operation of analyzer.
35
-------
TABLE 11
RESPONSE FACTORS ON PI-101
Compound
Actunl
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard Response
Deviation (n)* Factor
957.
Confidence
Interval
Acetal
Carbon Disulfidu
! '
Carbon Tetrachloride
Chloroform
Diketene
»
1 1000
5000
10000
1000
10000
500
1000
10000
1000
5000
10000
1000
5000
10000
925
7200
13200
1990
12900
784
1070
6070
756
2550
5250
148
318
460
13 (5)
10 (5)
mo (5)
71 (5)
921 (5)
59 (5)
72 (5)
475 (5)
«• * (5)
!0 (5)
10 (A)
7.6 (5)
13 (4)
7.1 (5)
1.1
0.69
.0.76
0.50
0.78
0.64
0.94
1.6
1.3
2.0
1.9
6.8
16
22
1.1 - 1.0
(0.69)
(0.76)
0.57 - 0.45
0.97 - 0.65
0.94 - 0.48
1.2 - 0.77
2.1 - 1.3
,1.4 - 1.3
(2.0)
(1.9)
7.9 - 5.9
18 - 14
23 - 21
to
o>
*Numbcr of replicates analyzed
-------
TABLE 11 (Cont.)
RESPONSE FACTORS ON PI-101
Compound
Actual
Concentration
(pprov)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
957.
Confidence
Interval
Dime thy laulfida.
Ethanol
I
•
EthyleneglycoldAmethyl
Ethvlene DichlorAde (tr«
»
Methanol
Pentanethiol. l-
Ef.her
r,B)
*
1000
5000
_IOOOO
1000
5000
10000
1000
IPOJO
1000
5000
10000
1000
1180
9200
11000
360
1330
3630
7620
1040
125
798
1060
1260
10 (4)
10 (4)
100 (5)
20 (5)
45 (5)
250 (5)
250 (5)
55 (4)
•
12 (5)
4.5 (5)
54 (4)
69 (4)
0.85
0.54
0.9l
2.8
3.8
2.8
1.3
0.96
8.0
6.3
9.4
0.79
(0.85)
(0.54)
(0.91)
3.4 - 2.4
4.9 - 3.4
3.4 - 2.3
1.4 - 1.2
1.1 - 6.84
11 - 6.3
6.4 - 6.2
11 - 8.3
0.96 - 0.68
-------
TABLE 11 (Cent.)
Compound
RESPONSE FACTORS ON PI-101
Actual
Concentration
(ppmv)
Instrument
Concentration Standard Response
(ppmv) Deviation (n) Factor
95%
Confidence
..Interval
Perchloromethyl Mercapt
Toluene
Tet rnchloroethane, 1 , 1 , 2 ,
i
Trichloroe thane, 1,1,
Trlclilorotrifluoroethane
,
n
-
•
1,1,2-
5000
1000
1000
5000
10000
1000
5000
10000
5000
10000
103
1180
736
1170
1880
1020
6170
9430
155
430
2.9 (5.)
120 (5)
5.5 (4)
10 (4)
10 (4)
77 (5)
66 (5)
200 (5)
1 (5)
1 (5)
,
'
48
0.85
1.4
4.3
5.3
0.98
0.81
1.1
32
23
55 - 43
1.2 - 0.67
1.4 - 1.3
(4.3)
(5.3)
1.4 - 0.74
0.84 - 0.79
1.1 - 1.0
(32)
(23)
-------
TABLE 12
RESPONSE FACTORS ON MIRAN 1A/80
Compound
Wavelength
(um)
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard Response
Deviation (n)* Factor
95X
Confidence
Interval
Acetal
»
i
i
»
•
'
3.3
3.4
3.6
»
5.7
8.8
1000
5000
10000
1000
5000
10000
1000
5000
10000
1000
5000
10000
1000
5000
10000
9670
28000
37600
1500
8480
22600
226
1420
2980
698
1150
1840
1890
9640
15800
149 (4)
150 (4)
386 (4)
49.3 (4)
64.5 (4)
50 (4)
119 (4)
79.4 (4)
164 (4)
115 (3)
129 (4)
257 (4)
75.7 0)
540 <5)
141 <«>
0.103
0.179
0.266
0.667
0.590
0.442
4.42
3.52
3.36
1.43
4.35
. 5.34
•
0.529
0.519
0.633
0.109 - 0.0986
0.182 - 0.176
0.275 - 0.258
0.744 - 0.604
0.604 - 0.576
(0.442)
7.82 - 1.65
4.28 - 2.99
4.07 - 2.86
4.91 - 0.839
6.76 - 3.20
9.78 - 3.76
•
0.639 - 0.451
0.614 - 0.449
0.651 - 0.615
*Number of replicates analyzed
-------
TABLE 12 (Cont.)
Compound
Wavelength
(UP)
RESPONSE FACTORS ON MIRAN 1A/80
Actual
Concentration
(ppmv)
Instrument 95%
Concentration Standard Response Confidence
(ppmv) Deviation (n) Factor Interval
Acetal
Acetyl-l^propanol, 3-
1
i
i
Benzoyl Chlori'de
j
Carbon Tetrachlorido
9.5
3.3
9.5
6.35
5.7
1000
5000
10000
500
1000
100
500
1000
100
500
1000
500
1000
10000
6690
23400
27200
247
813
39.2
217
406
•<70
5080
5420
115
232
390
672 (^
699 (5)
54.8 (5)
33.7 (5)
66.7 (5)
12.9 (4)
11.4 (5)
41.5 (5)
80.1 (5)
50.3 (5)
270 (5)
5.13 (3)
42.9 (4)
45.7 (5)
0.149
0.214
0.368
2.02
1.23
2.55
2.30
2.46
0.0209
0.0984
0.185
•
4.35
4.31
25.6
0.220 - O.lp
0.233 - 0.197
0.370 - 0.366
3.26 - 1.47
1.59 - 0.501
3.78 - 1.25
2.70 - 2.01
1.44 - 1.92
'
.0215 - 0.0196
.101 - 0.0192
.214 - 0.162
.38 - 3.65
0.5 - 2.71
8.0 - 19.3
-------
TABLE 12 (Cont.)
Compound
Wavelength
(um)
RESPOND
E FACTORS
Actual
Concentration
(ppmv)
ON MIRAN 1A/80
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Fnctor
95%
Confidence
Interval
Carbon Tetrachlorlde
i
Chloro-Acetaldehyde
.
-
6.35
9.5
13.5
3.3
3.4
3.6
5.7
10000
10000
500
1000
10000
1000
10000
10000
1000
10000
500
1000
10000
7920
64
1810
4300
33100
116
3660
580
64.0
2020
1870
2350
7620
330 (5)
4.28 (3)
143 (5)
309 (5)
572 (5)
92.3 (5)
384 (4)
40.9 (4)
3.80 (5)
127 (4)
13.0 (5)
28.8 (5)
356 (4)
1.26
156
0.276
0.233
0.302
8.62
2.73
17.2
15.6
4.95
0.267
0.426
1.31
.1.43 - 1.13
219 121
0.354 - 0.226
0.291 - 0.194
0.317 - 0.288
14.71 - 2.68
3.00 - 0.205
22.2 - 14.1
18.7 - 13.4
6.00 - 4.21
0.273 - 0.262
0.441 - 0.412
1.54 - 1.14
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIRAN 1A/80
Compound
Wavelengtn
(urn)
ts>
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95*
Confidence
Interval
Chloro-Acetaldohyde
)
!
I
Chloroform
.
'
Dlchloro-1-proponol, 2, 3-
6.35
9.5
13.5
%
13.5
3.3
500
1000
10000
500
1000
10000
500
1000
10000
1000
5000
10000
1200
4840
5680
6760
76
228
1880
709
2300
21800
6680
22200
34200
64.9
55.9 . (5)
151 (5)
813 (4)
12.9 (5)
5.89 (5)
64.3 (3)
29.9 (4)
82.9 (5)
802 (3)j
747 (5)
1260 (5)
2430 (4)
22.6 (3)
0.103
0.176
1.48
6.58
4.39
5.32
0.705
0.435
0.459
0.150
0.225
0.292
18.5
'
2.40 - 1 07
12.5 - 4.47
4 71 - L HO
6.27 - 4.64
0.814 - 0.621
0.483 - 0 V)s
0.545 - 0.396
0,217 - 0.114
0.267 - 0 1QS
0.378 - 0.239
29.7 - 7.40
-------
TABLE 12 (Cont.)
Compound
RESPONSE FACTORS ON MIRAN 1A/80
Wavelengtn
(um)
Actual
Concentr
(pprav)
Instrument
atlon Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95%
Confidence
Interval
Dlchloro-1-propanol, 2, 3
i
i - -~- -
Dichloro-2-propanol, 1, 3-
* •
t
>
Dlisopropy 1 Benzene ,1,3
9.5
13.5
3.3
9.5
t
13.5
3.3
100
500
1200
500
1200
1200
100
500
1200
500
1200
100
500
1225
85.2
447
. 747
1160
2230
227
65.4
304
653
1070
2300
133
703
1270
12.2 (5)
6.81 (3)
20.6 (4)
58.0 (4)
154 (4)
10.1 (3)
6.95 (5)
22.3 (3)
35.2 (4)
70.9 (3)
177 (3)
30.8 (4)
60.1 (5)
108 (3)
1.17
1.12
1.61
0.431
0.538
5.29
1.53
1.64
1.84
0.467
0-.522
' '
0.774
0.716
0.965
.95 - 0.840
.20 - 1.05
.76 - 1.48
.513 - 0.372
.690 - 0.441
6.54 - 4.44
•
M7 - 1.18
2.40 - 1.25
2.22 - 1.57
0.653 - 0.364
0.780 - 0.392
•
2.94 - 0.446
0.938 - 0.578
1.52 - 0.706
j J^i"
OJ
-------
TAJJL.E.
RESPONSE FACTORS ON MIRAN 1A/80
Compound
Wavelength
Actual
Concentration
(ppmv)
Instrument 95^
Concentration Standard Response Confidence
(ppmv) Deviation (n) Factor Interval
Dii sop ropy 1 Benzene , 1 , 3-
i
Diketene
»
,
'
6.35
5.7
3.3
5.7
»
9.5
500
1225
100
500
1225
5000
10000
1000
5000
10000
1000
5000
10000
•
134
507
311
343
380
354
1240
2280
6390
8600
69.4
377
580
34 (3)
77.9 (3)
8.04 (5)
12.5 (4)
14.2 (3)
13.9 (3)
197 (3)
226 (5)
171 (4)
487 (4)
15.7 (4)
23.6 (5)
65.9 (4)
3.75
2.42
0.331
1.47
3.22
14.1
8.06
0.439
0.782
1.16.
14.4
13.4
17.2
5.65 - 1.80
7.12 - 1.45
0.359 - 0.309
1.66 - 1.31
3.84 - 2.78
17.0 - 12.1
25.5 - 4.79
•
0.605 - 0.344
0.855 - 0.721
. 1.42 - 0.985
51.4 - 8.38
16.1 - 7.41
27.0 - 12.7
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIRAN 1A/80
Compound
Wavelength
(um)
Actual
Concentration
(ppmv)
Instrument
Concentration
(pptnv)
Standard
Deviation (n)
Response
Factor
95Z
Confidence
^Interval
Dimethyl Styrene,2,4-
i
Dlmethylsulflde »
•
'
3.3
5.7
6.35
3.3
3.4
500
1170
100
500
1170
100
500
1170
1000
5000
10000
1000
5000
10000
146
567
956
964
978
2540
2710
3010
2030
10100
20500
66.0
1510
3250
23.1 (4)
27.0 (5)
8.81 (5)
4.55 (5)
7.56 (5)
73.0 (5)
50.7 (5)
33.6 (5)
49.9 (4)
394 (4)
462 (5)
5.96 (4)
55.1 (3)
48.3 (5)
3.43
2.06
0.105
0.520
1.20
0.394
0.185
0.389
0.493
0.495
0.488
15.2
3.31
3.08
6.91 - 2.28
2.38 - 1.82
0.107 - 0.102
0.527 - 0.513
1.22 - 1.17
0.0428 - 0.0365
0.195 - 0.184
0.401 - 0.377
fr.534 - 0.457
0.565 - 0.440
0.520 - 0.459
•
21.3 - 11.8
3.93 - 2.86
3.21 - 2.95
en
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIRAN 1A/80
Compound
Wavelength
(urn)
Actual
Concentration
(pprov)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95%
Confidence
Interval .
Dimathvlaulflde
r '-
1
1
;
»
Ethanol
»
S.7
6.35
•
9.5
•
*
3.3
3.4
1000
5000
10000
1000
5000
10000
' i
1000
5000
10000
1000
5000
10000
1000
5000
10000
S22
1010
1180
2A80
4590
6540
15.3
120
270
3830
18500
34300
430
3420
7530
33. s (4)
52.4 (4>
95.4 m
90.7 (4)
112 (4)
190 (4)
2.90 (4)
3.51 (3)
27.4 (4)
181 (5)
432 (5)
217 (5)
16.8 (4)
47.2 (5)
72.1 (5)
1.92
4.95
8.47
0.403
1.09
1.53
65.4
41.7
37.0
0.261
0.270
0.292 ,
2.33
1.46
1.33
1.40 - 1L08
5.93 - 4,25
13.0 - 6.29
0.456 - 0.361
1.18 - 1.01
1.68 - 1.40
165 - 40.8
47.7 - 37.0
54,7 - 28.0
*
0.301 - 0.231
0.289 - 0.254
0.297 - 0.287
•
2.66 - 2.07
1.52 - 1.41
1.36 - 1.29.
-------
'I'AHLE 12 (Cont.)
Wavelength
(uro) __
RESPON8IJ FACTORS ON MIRAN 1A/80
Actual
Concentration
(ppniv)
Instrument
Concentration
(ppmv)
Standard
Deviation (11)
Response
Fnctor
95%
Confidence
Interval
Ethunol
i
i - - . . . .
i
Etlmnolamine
»r» *l
Etliylene Dicliloridc (tr
j
3.6
8 8
9.5
9.5
»
13.5
ns) 5.7
5000
10000
10000
1000
5000
10000
100
500
500
1000
5000
10000
386
941
668
2210
8440
16800
25.9
135
5620
684
815
940
4.24 (4)
10.7 (5)
20 8 (51
53.2 (4)
290 (4)
255 (5)
3.67 (4)
8.38 (4)
1050 (4)
14.5 (3)
8.02 (4)
19.1 (4)
13.0
10.6
15.0
0.452
0.592
0.595
3.86
3.70
0.089'
1.46
6.13
10.6
13.4 - 12.6
11.0 - 10.3
1 ft A _ 1 i H
J.0 .1 — JLJ. o
0.490 - 0.420
0.665 - 0.534
0.621 - 0.571
.
7.03 - 2.66
4.61 - 3.09
»
0.219 - 0.075
Is 61 - 1.34
6.33 - 5.95
11.4 - 9.99
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON HIRAN 1A/80
Compound
Wavelength
(um)
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
957.
Confidence
Interval
Ethylene Dichloride (tra
.
i
|
£ | Ethyleneglycoldlmethyl
Er.her
w
*
i
B) 6.35
8.8
3.3
3.4
3.6
1000
5000
10000
5000
10000
1000
5000
10000
1000
5000
10000
1000
5000
10000
1160
1760
2270
658
1540
5110
21100
33800
2310
11700
20600
284
1870
3920
34.1 (4)
30.0 (3)
136 <5)
23.1 (3)
41.2 (4)
86.2 (5)
460 (5)
351 (6)
42.8 (5)
358 (5)
501 (6)
7.09 (5)
74.6 (5)
93.5 (6)
0.862
2.84
4.41
7.60
6.49
.
0.196
0.237
0.296
0.433
0.427
0.485
3.52
2.67
2.55
0.951 - 0.78fl
3.07 - 2.65
5.29 - 3.78
8.95 - 6.60
7.10 - 5.98
0.205 - 0.187
0.252 - 0.223
0.304 - 0.288
0.456 - 0.412
0.431 - 0.394
0.518 - 0.457
•
3.78 - 3.29
3.01 - 2.41
2.72 - 2.40
-------
TABLE 12 (Cont.)
Compound
Wavelength
(u«n)
RESPONSE FACTORS ON MIRAN 1A/80
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95%
Confidence
Interval
Ethyleneglycoldimethyl
Ether
t
1 --
!
Ethyleneglycolmonoethyl
Ether Acetate
6.35
8.8
9.5
*
»
3.3
»
i
3.4
•
1000
5000
10000
1000
5000
10000
1000
5000
10000
200
1000
2GOO
1000
2000
1040
2280
4110
1570
9160
16000
1230
5130
9620
410
3570
4890
457
817
150 (5)
169 (5)
324 (6)
82.0 (5)
342 (5)
268 (6)
72.8 (5)
202 (5)
195 • (5)
51.2 (4)
122 (4)
50.3 (3)
67.0 (5)
80.0 (3)
0.962
2.19
2.43
0.637
0.546
0.625
0.813
0.975
1.04
0.488
0.280
0.409
2.19
2.45
1.61 - 0.686
2.76 - 1.82
3.0$ ~ 2.02
0.745 - 0.556
0.609 - 0.435
0.653 - 0.599
0.973 - 0.698
1-Q9_- 0.87?
1.10 - 0.934
*
0.809 - 0.349
0.314 - 0.253
0.428 - 0,392
•
3.69 - 1.55
4.23 - 1.72
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIRAN 1A/80
Compound
Wavelength
(um)
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95%
Confidence
Interval
Ethyleneglycolmonoethyl
Ether Acetate
i
!
1
i
Formaldehyde
3.6
5.7
8.8
9.5
V
•
3.3
i
3.4
1000
2000
200
1000
2000
1000
2000
200
1000
2000
500
1000
1000
50.8
158
2590
5110
6960
261
808
472
2190
3470
266
916
72.4
5.79 (3)
6.93 (3)
75.5 (3)
177 (4)
230 (4)
32.7 (3)
58.3 (3)
9.63 (4)
165 (4)
129 (3)
36.4 (6)
27.7 (6)
11.2 (6)
19.7
12.7
0.0772
0.196
0.287
3.83
2.48
0.424
0.457
0.576
1.88
1.09
13.8
8.6 - 13.2
15.6 - 10.6
0.0883 -.0.0686
0.220 - 0.176
0.321 - 0.260
8.31 - 2.49
3.59 - 1.89
0.453 - 0.398
0.600 - 0.368
0.686 - 0.497
2.90 - 0.991
1.18 - 1.01
22.9 - 9.88
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIRAN 1A/80
Compound
Wavelength
(um)
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
952
Confidence
Interval
Formaldehyde
t
i
i i
i
i
Formic Acid
3.6
5.7
9.5
3.3
x
V
3.4
»
3.6
500
1000
500
1000
500
1000
5000
10000
5000
10000
5000
10000
234
626
2490
3290
1
180
347
6930
18900
906
2860
1410
4510
47.1 (7)
34.2 (6)
147 (8)
99.4 (7)
19.8 (6)
44.0 (6)
79.7 (5)
335 (5)
11.5 (5)
51.3 (5)
22.8 (5)
101 (5)
2.14
1.60
0.201
0.304
2.78
2.88
0.722
0.529
5.52
3.50
3.55
2.22
4.22 - 1.43
1.86 - 1.40
0.233'- 0.176
0.328 - 0.283
3.87 - 2.17
4.27 - 2.17
•
0.745 - 0.699
0.557 - 0.504
5.72 - 5.33
3.61 - 3.38
3.95 - 3.22
2.36 - 2.09
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIRAN IA/80
Compound
Wavelength
(urn)
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95%
Confidence
Interval
Formic Acid
,
1
Freon 12
*
5.7
8.8
9.5
6.35
8.8
500
5000
10000
5000
10000
500
5000
10000
- 1212.5
2425
4850
•
1212.5
2425
4850
4990
23600
31300
1000
2920
1190
9120
14100
5940
6470
7490
1714
3130
4680
136 (5)
89.4 (5)
182 (5)
26.0 (5)
60.6 (5)
66.7 (5)
111 (5)
130 (5)
270 (5)
140 (5)
92.6 (5)
263 (5)
74.3 (5)
49.3 (5)
0.100
0.212
0.319
5.00
3.42
0.420
0.548
0.709
0.204
0.375
0.648
0.707
0.775
1.04
0.108 - 0.0931
0.214 - 0.210
0.325 - 0.314
5.39 - 4.66
3.63 - 3.24
0.498 - 0.364
0.567 - 0.530
0.728 - 0.691
0.234 - 0.181
0.399 - 0.354
0.671 - 0.626
1.23 - 0.496
0.830 - 0.727
1.07 - 1.01
to
-------
TABLE 12 (Cont.)
Compound
RESPONSE FACTORS ON Mlran 1A/80
Wavelength
(ura)
Actual
Concentn
(pprav)
Instrument
ition Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95%
Confidence
Interval
Freon 12
Furfural
i
i
i
9.5
3.6
5.7
6.35
ft
9.5
1212.5
2425
4850
1200
100
500
1200
100
. 500
1200
100
500
1200
6280
55000
72600
53.4
1230
1310
1420
4240
8040
13400
32.3
138
321
730 (5)
2130 (5)
1290 (5)
6.74 (5)
9.57 m
12.6 <4)
26.3 U)
113 (4)
120 (4)
420 (4)
3.50 (4)
8.66 (4)
15.2 (4)
0.193
0.0441
0.0668
22.5
0.0813
0.382
0.845
0.0236
0.0622
0.0896
3.10
3.26
3.74
.285 - 0.14b
0.0494 - 0.0398
0.0703 - 0.0637
34.8 - 16.6
0.0837 - 0.0793
0.394 - 0.370
0.898 - 0.798
0.0258 - 0.0217
0.0652 - 0.0594
0.0995 - 0.0814
4.72 - 2.30
4.53 - 3.02
4.40 - 3.25
Ul
to
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIRAN 1A/80
Wavelength
Compound (ym)
Actual
Concentration
(pprov)
Instrument
Concentration
(ppmv)
Standard
'Deviation (n)
Response
Factor
95%
Confidence
Interval
Furfural
Glycidol ;
i
o, I
*- 1
Hydroxyacetone
13.5
3.3
•
3.6
5.7
6.35
%
9.5
5.7
6.35
9.5
100
500
1200
100
100
100
100
»
100
100
100
100
656
5470
12200
262
572
3100
6540
132
1950
6870
24.6
24.9 (3)
260 (4)
532 (4)
24.6 (3)
27.0 (4)
52.6 (A)
99.3 (4)
10.4 (4)
24.3 (8)
72.8 (7)
3.58 (8)
0.152
0.0914
0.0984
0.382
0.175
0.0323
0.0153
0.758
0.0513
0.0146
4.07
0.182 - 0.131
0.108 - 0.0794
0.114 - 0,864
0.640 - 0.272
0.206 - 0.152
0.0341 - 0.0306
0.0161 - 0.0146
1.01 - 0.606
0.0528 - 0.0498
• •
0.0149 - 0.0142
6.21 - 3.02
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIRAN 1A/80
Wavelength
Compound (pm)
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppjnv)
Standard
Deviation (ri)
Response
Factor
95X
Confidence
Interval
Methanol
i
i
•
3.3
3.4
3.6
6.35
9.5
1000
5000
10000
1000
5000
10000
5000
10000
1000
5000
10000
1000
5000
10000
2520
17000
24400
101
1810
3840
10.3
181
2920
4230
5540
•
'438
1940
2870
102 (5)
7800 (5)
340 (4)
9.56 (5)
20.5 (5)
45.1 (4)
•
1.63 (5)
3.40 (4)
226 (5)
785 (4)
90.9 (4)
303 (4)
57.4 (4)
116 (4)
0.397
0.294
0.410
9.90
2.76
2.60
485
55.2
0.342
1.18
1.81
2.28
2.58
3.48
0.447 - 0.357
0.458 - 0.129
0.429 - 0.392
13.4 - 7.84
2.85 - 2.6B
2.71 - 2.51
867 - 337
58.8 - 52.1
0.436 - 0.282
1.26 - 1.12
1.90 - 1.72
2.93 - 1.07
2.81 - 2.3H
4.00 - 3.09
\J\
-------
TABLE 12 (Cont.)
Compound
RESPONSE FACTORS ON MIRAN 1A/80
Wavelength
(um)
i Actual
Concentra
(pprov)
Instrument
tion Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95X
Confidence
Interval
Methyl Styrene, o-
'
i -
!
!
!
Methylene Chloride
3.3
.
5.7
6.35
9.5
»
13.5
3.3
1030
5000
103
1030
5000
1030
5000
1030
5000
1030
5000
5000
10000
976
2830
330
1230
1570
4490
6960
73.6
178
167
948
1740
3740
38.6 (4)
229 (5)
9.20 (4)
10.0 (5)
48.3 (5)
128 (5)
322 (5)
3.67 (4)
» 5.57 (3)
49.2 (5)
' 92.2 (4)
62.7 (5)
144 (5)
1.06
1.77
0.312
0.837
3.18
0.229
0.718
14.0
28.1
6.17
5.27
2.87
2.67
1.21 - 0.937
2.28 - 1.44
0.342 - 0.287
0.858 - 0.819
3.48 - 2.93
0.249 - 0.213
0.824 - 0.637
16.6 - 12.1
32.5 - 24.8
34.1 - 3.39
7.64 - 4.03
3.19 - 2.61
2.99 - 2.42
Ul
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIRAN 1A/80
Compound
Wavelength
(lim)
Actual
Concentration
(pprov)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95%
Confidence
Interval
•fethylene Chloride
Pentanethiol, 1-
- i
13.5
3.3
3.4
%
3.6
9.5
1000
5000
10000
1000
5000
10000
1000
5000
10000
5000
10000
1000
5000
10000
30050
98400
119000
3180
11300
15800
648
4590
8650
267
634
89.4
483
886
933 (4)
4760 (5)
2880 (5)
15.3 (3)
370 (5)
265 (3)
14.8 (A)
115 (4)
97.1 (3)
«
23.4 (5)
18.5 (4)
i
18.6 (3)
51.6 (A)
50.7 (3)
0.0333
0.0508
0.0840
0.314
0.442
0.633
L.54
1.09
1.16
18.7
15.8
11.2
10.4
11.3
0.0369 - 0.0303
0.0587 - 0.0448
0.0901 - 0.0787
0.321 - 0.308
0.487 - 0.406
0.682 - 0.590
.1.66 - 1.44
1.18 - 1.01
1.21 - 1.10
24.8 - 15.1
17.4 - 14.4
106 - 5.90
15.7 - 7.73
12.7 - 10.1
O)
-J
-------
TABLE 12 (Cent.)
RESPONSE FACTORS ON MI RAN 1A/80
Wavelength
Compound (uro)
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95X
Confidence
Interval
Pentunethiol. 1-
Perchloromethvlmercaptan
!
;
i
13.5
3.3
3.6
5.7
8.8
9.5
5000
10000
5000
5000
500
1000
5000
5000
500
1000
5000
,- 5300
10«iOO
612
64.0
1730
3410
7660
426
36.7
132
303
217 (4)
289 f3)
28,2 (5)
4.99 (5)
•570 (3)
112 (4)
306 (5)
31.4 m
2,65 (1)
13.3 m
20,7 f4)
0.941
0.9S2
ft. 17
78.1
0.289
0.293
0.653
11.7
13.6
7.58
16.5
-1, OH_- 0 fl35
-1.08 - O.H52
_9J7 - 7,24
99t7 - fiA.2
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIHAN 1A/80
Compound
Wavelength
(um)
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
95X
Confidence
Interval
Propylene Chlorohydrln
*
1
i
•
i
3.3
3. 4
3.6
•
6.35
%
8.8
9.5
500
1000
5000
1000
5000
5000
500
1000
5000
5000
500
1000
5000
1240
2990
14200
36.6
1490
51.6
3040
3630
3960
•
795
462
966
4470
31.1 (4)
92.0 (5)
81.6 (4)
3.39 (4)
39.6 (5)
2.30 (3)
151 (4)
69.4 (5)
37.7. (5)
53.3 (5)
>
ll.'A (4
45.4 (5
157 (5
0.403
0.334
0.352
27.3
3.36
96.9
0.164
0.275
1.26
6.29
1.08
1.04
1.12
0.438 - 0.373
0.366 - 0.308
0.359 - 0.346
38.7 - 21.1
3.62 - 3.12
120 - 81.3
•
0.195 - 0.142
0.291 - 0.262
1.30 - 1.23
7.73 - 5.30
1.17 - 1.00
1.19 - 0.916
1.24 - 1.02
in
-------
TABLE 12 (Cont.)
RESPONSE FACTORS ON MIRAM 1A/80
Compound
Wave Length
,m\
Actual
Concentration
(pprov)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Response
Factor
953!
Confidence
Interval
Propylene Chlorohydrin
Tetrachloroe thane, 1, 1, 2,
i
i
i
»
Trichloroethane, 1, 1, 1-
j
13.5
- 3.3
8.8
13.5
%
3.3
3.4
500
1000
5000
5000
10000
10000
t
1000
5000
10000
1000
5000
10000
5000
10000
3800
8510
38600
582
1010
404
20000
73000
101000
266
2910
5920
38.8
421
128 (4)
194 (4)
904 (5)
26.1 (4)
21.6 (4)
25.1 (4)
616 (4)
2750 (4)
1670 ^4)
17.5 (4)
8.16 ,(4)
28.9 ' (4)
3.25 (4)
2.59 (5)
0.132
0.118
0.130
8.59
9.90
24.8
0.0500
0.0685
0.0990
3.76
1.72
1.69
129
23.8
0.147 - 0.119
0.127 - 0.110
0.139 - 0.122
10.0 - 7.52
10.6 - 9.27
30.8 - 20.7
•
0.0554 - 0.0455
0.0778 - 0.0612
0.105 - 0.0941
4.75 - 3.11
1.73 - 1.70
1.72 - 1.66
175 - 102
24.2 - 23.4
-------
TABLE 12 (Cent,)
RESPONSE FACTORS ON MIRAN IA/80
Wavelength
Compound (urn)
Actual
Concentration
(ppmv)
Instrument
Concentration
(ppmv)
Standard
Deviation (n)
Rcsponne
Factor
95X
Confidence
Interval
Trichloroethaner 1,1,1- 1 6.35
9'5
,'
,
: 13.5
.
1 »
Trichlorotrlfluoro- j 5.7
ethane, 1,1, 2- ]
-
'
6.35
1000
5000
10000
1000
5000
10000
1000
5000
10000
1000
5000
10000
1000
5000
10000
3010
14500
27400
4330
13900
19900
4120
16400
29100
640
856
1060
1390
2390
3520
87*2 (3)
114 (5)
95.7 (4)
238 (5)
283 (5)
1520 (4)
431 (4)
171 (4)
340 (4)
46.5 (4)
26.7 (5)
55.1 (3>
-
181 (3)
97.8 (5)
189 (4)
0.332
0.345
0.365
0.231
0.360
0.503
0.243
0.305
0.344
1.56
5.84
9.43
0.719
2.09
2.84
.380. - 0.295
0.353 - 0.337
0.369 - 0.361
0.273 - 0.200
0.381 - 0.340
0.664 - 0.404
0.364 - 0.182
0.315 - 0.295
0.357 •• 0.331
2.03 - 0.127
6.40 - 5.38
12.1 - 7.71
1.69 - 0.544
2.36 - 1.8U
3.43 - 2.43
-------
TABLE 12 (Cont.)
Compound
Wavelength
(um)
RESPONSE FACTORS ON MIRAN 1A/80
Actual
Concentration
(ppray)
Instrument 95%
Concentration Standard Response Confidence
v) Deviation (n) Factor Interval
Trichlorotrlfluoro-
ethane,l,l,2-
.'
•
*
j
8.8
9.5
13.5
\
1000
5000
10000
1000
5000
10000
5000
10000
5840
16100
18500
977
3690
6280
1100
2270
76. A (5)
50.0 (4)
134 (5)
55.0 (5)
6.93 (3)
58.0 (4)
28.9 (4)
132 (4)
0.171
0.311
0.541
1.02
1.36
1.59
•
4.55
4.41
• »
>
j
0.178 - 0.165
0.314 - 0.308
0.552 - 0.530
1.21 - 0.885
1.37 - 1.34
1.64 - 1.55
4.96' - 4.19
5.40 - 3.72
•
ro
-------
6. DISCUSSIONS AND CONCLUSIONS
A. Photoionization Detection
As noted previously, the photoionization technique vas evaluated for a
limited number of compounds due to both chemical and, more significantly,
equipment problems. The PI-101 was calibrated with dichloromethane so
as to permit direct comparison with response factors reported in
Reference 3. The response factors observed for the 16 compounds tested
on the photoionization detector, PI-101, range from 0.50 to 48. Seventy-
five percent (12) of the compounds have response factors of less than
five and greater than 0.2. There appears to be no obvious trend of
response factor with molecular weight (carbon number) or functionality
within this group. On the other hand, it is interesting to note that for
both alcohols tested, i.e., methanol and ithanol, the response factors are
inversely proportional to carbon number. Thus, it appears that non-
binding electrons on the oxygen atom of the alcohols do not provide a
much greater photoionization yield than other sigma-bonded electrons in
compounds with similar carbon numbers. The high response factor for
trichlorotrifluorethane is consistent with its high ionization poten-
tial (11.78 eV). In fact, this ionization potential is slightly higher
than the quoted energy of the UV lamp used in the study. This may
indicate that thermal energy provides sufficient additional energy to
permit some ionization when coupled to the energy provided by the UV
light.
Although the specific response factors for the limited number of com-
pounds tested do not unequivocally confirm the suitability of photo-
ionization as a general VOC screening technique, an important but cautious
observation can be made. That is, based on this small sample of com-
pounds tested, which includes an aromatic compound (i.e., toluene), an
ether (i.e., acetal), an alcohol (i.e., ethanol) and chlorinated alkanes
(i.e., trichloroethane and chloroform), the response factor over a con-
centration range of 500 ppmv to 10,000 ppmv may be within a factor of
five. This result is consistent with an expectation (Figure 1) of more
similar photoionization yield from sigma and pi electrons when the com-
pound is influenced by UV radiation of approximately 12 eV rather than
10 eV. The expectation that photoionization yield for aliphatic and
aromatic compounds may be similar indicates the potential usefulness
of photionization as a VOC screening tool.
In terms of current availability as a potential VOC detector, the most
significant result with respect to the photoionization detector (HNU
Systems, -Inc. PI-101 and AID, Inc. 580) is probably the difficulty
observed in operating the prototype dilution system. Both dilution
probes were designed and fabricated by the respective manufacturers
under severe time limitations. Neither probe was designed in a manner
which permitted reliable independent measurement of dilution ratio or
reproducible adjustment. Thus, the absolute dilution ratio is in some
doubt. The ability to adjust the dilution ratios was practically non-
existent. As noted previously, the fixed dilution ratio's were inappropriate
63
-------
for analysis of vapor concentrations which yielded instrument responses
much above 10,000 ppnrv or much below 1000 ppmv. Detector saturation
was observed somewhat above an instrument response of 10,000 ppmv. At
the span settings required for adequate operation, the background
instrument response to zero air was quite high.
*
Whenever the intensity of the UV lamps began to decrease (note that the
AID, Inc. lamp failed early in the program), the instrument span had to
be increased regularly. Some alteration to the span potentiometer
setting could be made to correct for this decrease in response. However,
for some tests the correction was not sufficient to yield an identical
calibration. Under these conditions, response factors were calculated
at a different absolute instrument response. The data included in
Table 11 reflects this variation in span point. However, since the
calibration curve is linear over the range of 0-10,000 ppmv (with
dilution, that is about 0-1000 ppmv) (Figure 5), no systematic error
should occur due to the change in absolute response.
Due'to declining instrument response and low vapor pressure of many •
compounds, one-half of the compounds tested did not yield reliable
response factors. The problems noted above and limited data obtained
indicate :hat, at the present time, a reliable photoionization system
does not exist to operate over a VOC concentration range of 100 ppmv to
10,000 ppmv. More accurately, a reliable dilution/photoionization system
is not available.
B. Infrared Detection
The results of the evaluation of the Miran 80 are much more complete.
A total of 32 compounds were analyzed. As noted previously, other
compounds were not tested for several reasons, including (1) low vapor
pressure; (2) reactivity; (3) lack of availability; and (4) close
chemical similarity to compounds previously tested. Prior to testing,
the instrument was calibrated with individual span gases at eight
analytical wavelengths which correspond to individual functional groups,
e.g., C-H; C-C1; C-OH. The calibration curve data (Table 13) indicate
that the absorbance values observed over the concentration range of
100 to 10,000 ppmv are linear. Test compounds were then run and the
instrument response calculated on the basis of the response indicated
by the specific span gas used at individual analytical wavelengths.
An analysis of the data in Table 12 indicates that the response factors
for most compounds with a particular functional group, determined at an
analytical wavelength which corresponds to that functional group
(Table 10), are generally less than a value of twenty. This is consistent
with the general observation that the functional group is more important
than the remainder of the molecule in determining the IR extinction
coefficient of the compound at the wavelength of interest.
•
For example, three of the four aromatic compounds tested have -
reasonable response factors (<5)_at 6.35 urn as shown below. This
64
-------
SLOPE: 0.711
IBTZBCEPT: 6£4
S- 15
COU. OOEFF: 0.995
4 6 8
ACTUAL CONCDtTRAXXOH a
10
5. Calibration curve for 1,2-dichloroethaneJor the HNU instrument.
... ~I~.~~ ~ >
l5 ' f^.9^
-------
TABLE 13
CALIBRATION DATA FOR MIRAN 80
Wavelength
(um)
3.3
3.4
3.6
5.7
6.35
8.8
9.5
13.5
Functional
Group(a)
Aromatic and
Unsaturated C-H
Saturated C-H
Aldehyde C-H
Carbonyl C-0
Aromatic C-C,
conj C-C. N-H,
C-S
Ether C-O-C
Alcohol C-O-H
C-C1
Calibration
Compound
Toluene
Pentane
Butyraldehyde
Acetone
Toluene
Dllsopropyl
ether
Isopropanol
1,2-dichloro-
ethane
Slope
(xlO-4 AU/ppm)
.255
Y-Intercept
(AU)
.0114
Correlation
Coefficient
.999
1.35
.436
.307
.059
.640
.587
.105
.0495
.00675 •
.00449
.0153
.0552
.00227
.00670
.994
.999
.999
.999
.992
.999
.992
AU - Absorbance Units
-------
wavelength is within a broad aromatic ring stretch area.
Compound Response Factor Range
A
Diisopropyl Benzene 2.42 - 3.75
Dimethyl Styrene,2,4- 0.185 - 0.934
Methyl Styrene 0.229 - 0.718
Within this group, the addition of the large aliphatic group (isopropyl)
on the benzene ring appears to reduce the sensitivity (larger response
factor) at the aromatic C ~ C stretch wavelength as compared to less
alkylated aromatics.
In the case of aliphatic and substituted aliphatic compounds, the OH
stretch wavelength of 3.3 ym yields suitable response factors (<5) for
about 52% of those tested. The classical aliphatic C-H stretch is
observed at 3.4 ym, but some overlap of 3.3 and 3.4 ym IR bands may
occur in the Miran due to incomplete resolution. Also, some shift of
the CH stretch wavelength probably occurs due to nearby oxygen or
halogens. A list of aliphatic compounds and corresponding response
factor ranges at this wavelength are shown in Table 14. If one includes
alkylated aromatic compounds (4) in the list of compounds with response
factors less than five at 3.3 ym - 3.4 ym, the percentage of compounds
tested with suitable response factors increases to 62%.
Ten chlorinated hydrocarbons tested in this program yielded measurable
response factors of 13.5 ym. Seventy percent were observed to yield
response factors less than five at this wavelength. The compounds
and respective response factors are given in Table 15.
Since the ultimate goal of this instrument evaluation is to assess the
suitability of IR as a general VOC screening technique, an assessment of
the usefulness of a single wavelength for measurement of orgauic compounds
of varied molecular weight and functionality is in order. A review of the
data in Table 12 indicates that the number of test compounds (total of 32)
which yield response factors of less than 20 or greater than 0.05 at each
analytical wavelength are as follows:
Wavelength (ym) Number of Compounds
3.3 23
3.4 12
3.6 13
5.7 17*
6.35 17
8.8 11
9.5 25
13.5 1A
67
-------
TABLE 14
SUBSTITUTED ALIPHATIC COMPOUNDS WITH RESPONSE FACTORS
LESS THAN TWENTY AT 3.3 ym (Mlran 80)
Compound Response Factor Range
Acetyl-l-propanol,3- 1.23 - 2.02
Chloro-acetaldehyde 2.73 - 8.62
Dichloro-l-propanol,2,3- 18.5
Dichloro-2-propanol,l,3- 5.29
Diketene 8.06 - 14.1
DimethyIsulfide . 0.488-0.495
Ethanol 0.261 - 0.292
Ethyleneglycoldimethyl Ether 0.196 - 0.296
Ethyleneglycolmonoethyl Ether Acetate 0.280 - 0.488
Formaldehyde 1.09 - 1.88
Formic Acid 0.529 - 0.722
Glycidol 0.382
Methanol 0.294-0.410
Methylene Chloride 2.67 - 2.87
Pentanethiol,!- 0.314 - 0.633
Propylene Chlorohydrin 0.334 - 0.403
Tetrachloroethane,l,l,2,2- 8.59-9.90
Trichloroethane,1,1,1- 1.69 - 3.76
68
-------
TABLE 15
CHLORINATED COMPOUNDS AND RESPONSE FACTORS AT 13.5 ym
(Miran 80)
Compound
Carbon Tecrachloride
Chloro-acetaldehyde
Chloroform
Dichloro-1-propanol,2,3-
Dichloro-1-propanol,lf3-
Methylene Chloride
Propylene Chlorohydrln
Tecrachloroethane,1,1,2,2-
Trichloroethane,1,1,1-
Trichlorotrifluoroethane,1,1,2-
Response Factor Range
0.233 - 0.302
0.435 - 0.705
0.150 - 0.292
0.431 - 0.538
0.467 - 0.522
0.0333 - 0.0840
0.118 - 0.132
0.0500 - 0.0990
0.243 - 0.344
4.41 - 4.55
69
-------
In some cases, the response factors at a particular wavelength
(e.g., 5.7 urn) are strongly a function of concentration. It appears
that this may be due to a concentration broadening phenomenon which is
frequently observed in gas phase infrared spectrometry. If only those
compounds which show a response factor between 5 and 0.2, and those
which show no strong variation in response factor with concentration
(i.e., < a factor of two from 1000 to 10,000 ppmv) are summarized as
above, fewer compounds yield suitable response factors:
Wavelength (urn) Number of Compounds
3.3 12
3.4 4
3.6 3
5.7 1
6.35 3
8.8 4
9.5 15
13.5 7
The results indicate that only 3.3, 9.5 and 13.5 ym analytical wavelengths
respond acceptably for a large number of compounds (i.e., greater than
10% of the total number of compounds). However, in any case, fewer than
50% of the compounds are reliably detected. The aliphatic and aromatic
compounds do not overlap at 6.35 ym but do overlap at 3.3 ym. However,
note that only alkylated aromatics have good response at 3.3 ym. Thus,
there is apparently no useful agreement in response factors between,
for example, a large number of aromatic compounds and aliphatic compounds
(e.g., 50% of those tested) at analytical wavelengths specific to each
compound class. It is apparent that the overlap of IR absorbance bands
of different functional groups is not sufficient to yield one analytical
wavelength which might be used to quantify both compound classes with
the expectation of agreement within a factor of five. This observation
indicates that infrared spectrophotometry is not particularly suitable
for general VOC screening.
On the other hand, the fact that the response factors do not vary by
large values (i.e., greater than five) for some classes of compounds,
e.g., halogensted aliphatics at 13.5 ym and aliphatic and alkylated
aromatics at 3.3 - 3.4 urn, corroborates the suitability of infrared
spectrophotometry for VOC screening of compounds belonging to one
functional group. Even in this case, only 30 to 80% of the compounds
in a given class may yield response factors less than five at a single
specific IR wavelength.
70
-------
C. Conclusions
In summary, based on the results of this evaluation, it appears that:
1. Infrared (IR) spectrophotometry may not be suitable for
general VOC screening, with the exception of analysis
of VOC emissions of a single organic functional group
character.
2. IR screening of organic compounds of a single functional
class, e.g., C-C1, may be suitable for as much as 80%
of compounds in the class.
3. IR screening at a wavelength corresponding to both
aliphatic and aromatic CH stretches may be suitable
for as much as 30-50% of organic compounds.
4. A portable photoionization device is not currently
available for VOC screening in the concentration range
of 100 ppmv to 10,000 ppmv.
5. The development of a reliable dilution probe for use on
a photoionization device is close at hand.
6. With such a dilution probe, it appears that a photoionization
device with an 11.7 or 11.8 eV UV lamp may be used for
reliable analysis of VOC fugitive emissions.
71
-------
7. LITERATURE CITED
tf
1. U.S. Environmental Protection Agency, "Measurement of Volatile
Organic Compounds," Report No. PB80-221674 (September 1979).
2. U.S. Environmental Protection Agency, "Method 21. Determination
of Volatile Organic Compound Leaks," proposed regulations, Federal
Register. .46 (Monday, January 5, 1981).
3. Brown, G.E., D.A. DuBose, W.R. Phillips and 6.E. Harris, "Response
Factors of VOC Analyzers Calibrated with Methane for Selected
Organic Chemicals," Task Report, U.S. Environmental Protection
Agency, EPA 600/2-81-002 (September 1980).
4. Anastas, M.Y. and H.J. Belknap, "Summary of Available Portable
VOC Detection Instruments," U.S. Environmental Protection Agency,
Report No. EPA-340/1-80-010 (March 1980).
5. Spain, D., J.J. Decorpo, J.R. Holtzclaw, "Use of a Photoionization
Detector as a Hydrocarbon Trace Gas Analyzer," Naval Research
Laboratory, Memo Report No. 4239 (August 1980).
6. Langhorst, M.L., "Photoionization Detector Sensitivity of Organic
Compounds," J. Chromat. Sci., 19. 98-103 (February 1981).
7. Driscoll, J., HNU Systems, Inc., personal communication (June 1981)
8. Dietz, W., J. Gas Chromatog.. j>, 68 (1967).
8. APPENDICES
The attached appendices include instrument operating procedures which
were abstracted from instruction manuals provided by the manufacturers.
72
-------
A. HNU Systems, Inc. - Model PI- 101
(1) Introduction
The Model PI-101 has been designed to measure the concentration of
trace gases in many industrial or plant atmospheres. The analyzer
employs the principle of photoionization for detection. This process
is teraed photoionization since the absorption of ultraviolet light
(a photon) by a molecule leads to ionization via:
RH + hv -*• RH+ + e~
where RH - trace gas, and
hv - a photon with an energy J> ionization potential of RH.
sensor consists of a sealed ultraviolet light source that emits
photons which are energetic enough to ionize oany trace species
(particularly organics) but dc not ionize the major components of air
such as 02, N2, CO, C02 or H20. A chamber adjacent to the ultraviolet
source ocntains a pair of electrodes. When a positive potential is
applied to one electrode, the field created drives any ions, formed by
absorption of UV light, to the collector electrode where the current
(proportional to concentration) is measured.
To minimize absorption of various sample gases, the ion chamber is
made of an inert fluorocarbon material, is located at the sampling
point, and a rapid flow of sample gas is maintained through the small
ion chamber volume.
The analyzer will operate either froa a rechargeable battery for acre
than ten hours or continuously from the AC battery charger. A solid
state amplifier board in the probe and a removable power supply board
in the readout module enable rapid servicing of the unit in the field.
The useful range of the instrument is from a fraction of a ppm to about
2,000 ppm. For measurement at levels above 2,000 ppm, dilution of the
sample stream with clean air is recommended. Some typical specifications
for the Model PI-101 Photoionization Analyzer are given in Table Al.
(2) Operation
Turn the function switch to the battery check position. The needle on
the meter should read within or above the green battery arc on the scale-
plate. If the needle is in the lower portion of the battery arc, the
instrument should be recharged prior to naV-ing any measurements. If red
LH> coaes on, the battery should be recharged.
Sext, turn the function switch to the on position. In this position, the
UV light source should be on. Look into the end of the probe to see the
purple glow of the lamp.
A brief description, Table A2^ of the instrument controls and functions
is shown in Figure Al.
A-l
-------
TABLE Al
SPECIFICATIONS FOR MODEL PI 101 PHOTOIONIZATION ANALYZER
Performance (benzene referred)
Range: 0.1 to 2000 ppm
Detection limit: 0.1 ppm
Sensitivity (max): 0-2 ppm FSD over 100 division meter scale
Repeatability: ±1% of FSD
Linear Range: 0.1 to 600 ppm
Useful Range: 0.1 to 2000 ppm
Response Time: <3 sec to 90% full scale
Ambient humidity to 95% RH
Operating temperature ambient to 40°C*
Physical
Size: Probe 6.3 DIA x 28.5 L (cm) (2-1/2 x 11-1/4")
Readout 21W x 13D x 16.5H (cm) (8-1/4 x 5-3/16 x 6-1/2")
Stowed 21W x 13D x 24H (cm) (8-1/4 x 5-3/16 x 9-1/2")
Cable 80 cm long (32")
Weight: Probe .55 kg (20 ounces)
Readout 3.2 kg (7 pounds)
Total (shipping) 5.4 kg (12 pounds)
Controls and functions
Mode switch OFF, Battery Check, Standby (zero), 0-2000, 0-200,
0-*20 ppm
Low battery indicator light
Zero (10 turn £ 300% FSD max)
Span (10 turn counting dial 1.0 to 10 times nominal sensitivity)
Readout 4-1/2" (11.3 cai) meter Taut Band movement graduated 0-5-10-
15-20 divisions
Signal output for recorder 0-(-5V) FSD
Power output for recorder 12 VDC - jack on side of instrument
Instrument is temperature compensated so that a 20°C change in
temperature corresponds to a change in reading of < ±2% full-
scale at ;r.n?rlmm sensitivity.
A-2
-------
TABLE Al (Continued)
Power requirements of operating times
Continuous use, battery > 10 hours
Continuous use with HNU recorder reduces instrument battery
operating time to 1/2 normal time
Recharge time, max < 14 hours, 3 hours to 902 of full charge
Recharge current, max .4 amps @ 15 VDC
Construction
Designed to withstand the shock and abuse to which portable instru-
ments are often subjected. The readout is housed in a two-piece
aluminum case, and finished with a solvent resistant baked
acrylic textured paint.
The probe is fabricated from extruded aluminum sections and
machined plastic.
Serviceability
The probe and readout are of a modular design allowing rapid servicing
and/or replacement of mechanical and electrical components. All
module interwiring includes quick disconnects.
Maintenance
The instrument contains only one moving part, and consumes no gases
or reagents. The only routine maintenance procedure is cleaning the
light source window every several weeks.
Calibration check
Check instrument calibration at least once per week with HNU
calibration standard to ensure that the high sensitivity of the
instrument is
A-3
-------
TABLE A2
BRIEF lESCRIPTION OF INSTRUMENT CONTROLS AND FUNCTIONS*
Control Function
Six Position Switch OFF - Shuts off all power and removes DC
voltages.
CN - In any other function position or
measuring mode, the electronics are en.
BATTERY CHECK - Indicated the condition of
the battery. If needle position is in
lower portion of green battery arc, the
instrument should be recharged.
STANDBY - UV lanp is off but electronics are
on. This position will conserve power
and extend the useful operating time
between recharges of the battery. This
position is also utilized to adjust the
electronic zero.
RANGES.- 0-20, 0-200, 0-2000 direct reading
ranges available at minimum gain for
benzene. More sensitivity is available
by adjusting the span potentiometer.
Zero Potentioiteter A ten-turn potentiometer is employed to adjust
the zero electronically when the instrument is
placed in the standby position with the probe
attached. This eliminates the need for a hydro-
carbon-free gas.
Span Potentiometer A ten-tuii counting potentiometer is utilized
for upscale setting of the meter on calibration
gas. Counter-clockwise rotation increases the
sensitivity (M.O times}. This pot can increase
the sensitivity to make the instrument direct
reading for nearly any gas which the instrument
responds to.
For position of layout controls see Figure Al.
A-4
-------
Battery Cfieck
Low Battery Indicator
Light (LED)
Power Off
Sensitivity
Aoji'Stment
Hi-Voltage
Interlock
12 Pin Interface Connector
between readout unit and
seosor.
Ranges (ppm)
Function
Switch
Zero Adjustment
Recorder Output
1-5VDC)
Figure Al. Control panel functions.
A-5
-------
To zero the instrument, turn the function switch to the standby position
and rotate the zero potentiometer until the meter reads zero. Clockwise
rotation of the zero potentiometer produces an upscale deflection while
counterclockwise rotation yields a downscale deflection. Note: No zero
gas is needed, since this is an electronic zero adjustment (see below).
If the span adjustment setting is changed after the zero is set, the
zero should be rechecked and adjusted, if necessary. Wait 15 or 20
seconds to ensure that the zero reading is stable. If necessary,
readjust the zero.
The instrument is now ready for calibration or measurement by switching
the function switch to the proper measurement range. The instrument is
supplied calibrated to read directly in ppm (v/v) 0-20, 0-200, 0-2000
of benzene with the span position set at 9.8. For additional sensitivity,
the span potentiometer is turned counterclockwise (smaller numbers) to
increase the gain. By changing thci span setting from 10.0 to 1.0 the
sensitivity is increased approximately tenfold. Then, the 0-20, 0-200,
and 0-2000 ppm scales become 0-2, 0-20, and 0-200 ppm full scale,
respectively. This span control is also utilized to make the instrument
scale read directly in ppm of the compound being measured, e.g., it is
adjusted to match the value of a calibration gas to that same reading
on the instrument scale. The span control can be utilized to calibrate
nearly any compound, measured by photoionization, to be direct reading
on the 0-20 ppm range.
A small DC-operated fan is used to pull air through the photoionization
sensor at a flow rate of three to seven hundred centimeters per minute
(ca. 0.5 1pm). The fan provides nearly instantaneous response times
while consuming little power. The characteristics of a fan are such
that it cannot tolerate a significant pressure drop without affecting
the flow rate and, therefore, either the instrument reading or response
time. Since photoionization is essentially a nondestructive technique,
changes in flow rate do not affect the signal but if a large pressure
drop is imposed at the inlet of the probe, the sample may not reach
the sensor.
The instrument was designed ta measure trace gases over a concentration
range from less than 1 ppm to 2000 ppm. Higher levels of various gases
(to percentage range) can be measured but the recommended procedure is
to dilute the sample with clean air to a concentration of less than
500 ppm. This is generally within the linear range of the instrument
and if the measured concentration is mult ID He'd by the dilution ratio
the correct concentration in the stream cats be determined. A calibration
curve for diluted dichloroethane was included in the basic report.
If the probe is held close to AC power lines or powar transformers, an
error may be observed. For measurements made in close proximity to
such items, their effect on measurements can tr. determined by the
following procedure. Zero the instrument in aa electrically Quiet.
area, in the standby pocition, then move the instrument: to the
A-6
-------
questionable area involved. If AC pickup is going to be a problem,
the meter (in the standby position) will indicate the magnitude of
the error.
The instrument is equipped with an automatic solid state battery
protection circuit. When the battery voltage drops below A* 11 volts,
this circuit will automatically turn off the power to the instrument.
This prevents deep discharging of the battery and considerably extends
the battery life. If the instrument is unintentionally left on over-
night, the battery will be unharmed because of the battery protection
circuit. If the instrument battery check reads low and the lamp doesn't
fire, plug the charger into the instrument. The power to the analyzer
should then be returned.
To charge the battery, place the mini-phone plug into the jack on left
side of the bezel prior to plugging charger into 120 VAC. When
disconnecting charger, remove from 102 VAC before removing mini-phone
plug. The battery is completely recharged overnight (ca. 14 hours).
To ensure that the charger is functioning, turn the function switch
to the battery check position, place phone plug into jack and plug
charger into AC outlet. The meter should go upscale if the charger is
working and is correctly inserted into the jack.
The instrument can be operated during the recharge cycle. This will
lengthen the time required to completely recharge the instrument battery.
A-7
-------
B. AID. Inc. Model 580
The Model 580 Organic Vapor meter has been designed for measurement of
low levels of most organic vapors in an air sample. The sample Is
contlnously pulled Into the Model 580 by Its own Internal sampling pump
at a rate of approximately 600 mL/mln. The sample Is pulled directly
into the Photionization Detector where high energy from a UV source
causes ionlzation of organic materials in the sample. The amount of
ionlzation that occurs is determined by measuring the ion current in
an electrical field. This current is amplified through an electrometer-
type amplifier and presented on a digital display on the front panel as
well as being made available on the rear of the unit for attachement to
a 100 millivolt recorder. The Model 580 used in this study was equipped
with a lamp providing an ionizing energy of 11.8 eV. In general, anything
with an ionization potential below 11.8 will be sensed by the Model 580.
The ionlzation potentials do not necessarily give an indication of the
sensitivity of the ionlzation detector for that particular material.
Some materials that have an ionlzation potential as high as 12.2 will
also be ionized using ths 11.8 lamp.
The Model 580 is a completely portable instrument, meaning that it
can operate independent of any power requirements on its own internal
battery system for a period of at least 8 hours. It is also capable
of being operated from a line voltage using the plug-in charger circuit
that accompanies the instrument. With the charger plugged into the
system, by means of the 3-position selector switch, the line voltage
can be used to either recharge the batteries in the Model 850 or to
operate the Model 580 from line voltage. The third position on the
selector switch is the operation of the instrument from its own internal
batteries.
The instrument may be set up on a bench with the charging circuit
plugged to a wall outlet and into the rear of the Model 580. The power
switch on the rear of the instrument should be set for AC operation.
The switch on the front panel should then be turned to the ON position.
When this occurs, the LCD display should be activated allowing digits
to appear, the pump should be activated, and the lamp in the detector
should come on. The lamp can be observed by the small hole in the end
of the detector housing. The lamp gives off a pale blue light when it
is in operation.
Initial calibration of the 580 requires a supply of zero air. This
means simply it has to have zero concentration of the components to
which the 580 will respond. Thus, it is not necessary to clean methane
from ambient air for this zeroing. In many cases, ambient air itself
will be sufficient for zeroing of the instrument. When basically zero
air is supplied to the instrument, the zero can be set through the ZERO
adjust hole in the rear panel. There is a small pot immediately inside
this hole that will allow adjustment of this zero. The Model 580 is
then presented with a sample of known concentration of a span gas to be
B-l
-------
measured. The SPAN Pot, again located on the rear panel, is then
adjusted via a screwdriver for proper reading of the concentration of
this particular span gas.
The Model 580, as it leaves the plant, has been calibrated on a mixture
of butadiene in air. The Model 580 has two ranges of operation, 0 to
200 and 0 to 2000 ppm. In general, the span gas used should be relatively
close in concentration to the expected levels to be determined on the
actual air being sampled. This would also imply that the particular
range of the unit should be on the same range for calibration as well
as for measuring.
The standard Model 580 is designed to obtain organic vapor concentrations
in the ppm region with an upper limit of approximately 2000 ppm. There
are two basic reasons for this upper limit. The first reason is that of
the electronics in the system. If the concentration or, literally,
the electrical signal arriving at the detector amplifier gets much
greater than that required to provide a readout of 2000 ppm, the
amplifier will begin moving into a nonlinear response region; thus, not
provide adequate output for the organics present in the sample.
The second reason is the detector design itself. This particular
detector was designed for industrial hygiene-type analysis in which the
concentration ranges of Interest are certainly covered by the 2000 ppa
upper lisdt. As one exceeds the 2000 ppa, the detector response itself
becomes nonlinear; such that doubling the amount of organic concentra-
tion in the sample does not double the output signal from the system.
When it becomes necessary, such as in the area of fugitive emissions,
to measure concentrations at the 10,000 ppm level using the Photo-
ionization Detector, it is necessary to dilute the sample such that
the concentration actually presented to the detector is below 2000 ppm.
Dilution of the sample will lower the reading by a known factor, e.g.,
1:10, on the instrument and raise the minimum discernible amount of
organic vapor from, for example, .1 ppm on a standard 580 to 1 ppm on
the modified version for high concentrations.
«
In the standard Model 580, the sample is pulled in through'the probe
to the probe fitting and from there into the detector. The detector
exit then connects directly to the pump inlet. The pump exit exhausts
through the rear panel of the instrument. In order to provide a 1:10
dilution of the Incoming sample, it is necessary to add, to that stream
prior to the detector, clean air at the flow ratio of 9-to-l. Thus,
directly behind the sample probe fitting on the front panel, a Teflon T
arrangement was installed whereby diluent air is placed into the sample
line. This diluted sample then continues in the normal fashion through
the detector out of the detector into the pump inl'et and exits at the
pump exit. Just prior to exiting the instrument, another T was placed
in the sample line. The pump exit connects to one part of the T. The
second part of the T goes to a restrictor valve and then to the exit
connection on the rarea panel. The third part of this T passes through
B-2
-------
tubing in the back panel to a charcoal filter. The exit from the
charcoal filter passes back into the SCO and connects to an inlet of
the Teflon T placed upstream of the detector. It is desired to maintain
500 mL/min through the detector. Thus, the incoming sample should be
50 mL/min and the makeup air coming from the charcoal filter to the
inlet T should be 450 mL/min. This ratio is adjusted by using the
restrictor between the exit T and the exit port on the rear panel
of the 580. This restrictor is closed until the inlet sample sampling
rate is 50 mL/min. The integrity of the system is checked by measuring
the exit flow at 50 mL/min.
Under this type of system, th^re is 450 mL/min of air circulating
from the pump outlet through the charcoal filter into the inlet line to
dilute the sample. The charcoal in the filter removes interfering
materials that would give a response on the Photoionization Detector.
Normally, the charcoal will remove,organic material, other than methane
and possibly ethane, from the sample. These materials will not be
ionized in the Photoionization Detector. The charcoal filter is a
disposable cartridge and is held on the outside rear panel of the
Model 580. This provides easy replacement of this charcoal filter.
With the 1:10 dilution introduced into the Model 580 by the modification
system described above, the inlet sample flow is of necessity reduced
to 50 mL/min. With the standard probe assemblies in the Model 580,
this would Introduce a serious time constant before the sample is
actually presented to the detector for analysis. For this reason,
the probe has been altered to a 1/16" ID tube to provide approximately
the same time constant on the sampling tube, as one would obtain with
a larger probe.
B-3
-------
C. Foxboro/Wilks, Inc. - Miran 80
(1) General Description
The Miran 1A Portable Gas Analyzer-is a single-beam, variable filter
spectrometer, scanning the infrared spectral range between 2.5 and
14.5 ym. The instrument is equipped with a gas cell having pathlength
variable between 0.75 and 20.25 meters. Other basic equipment are a
pump and 3 m (10 ft) air sampling hose with a particulate filter, a
zero gas filter, and a carrying case.
The Gas Analyzer System consists of two components, the gas cell and
the analyzer (Table Cl). The variable gathlength gas cell has a 5.6
liter capacity body, vacuum-tight to 10 ^ torr and pressurizable to
1000 kPa (10 atmospheres), an internal optical path variable in 1.50
meter increments between 0.75 and 20.25 rosters, a pair of windows
transparent to infrared energy between 2.5 and 14.5 ym, inlet and outlet
ports and a safety valve. The internal optics are gold plated and the
inside of the cell is ptfe coated to resist sample absorption and
corrosion.
The analyzer consists of a radiation source, mirror systea, mechanical
chopper, circular filter (variable in three segments between 2.5 and
14.5 ym), a scanning motor, pyroelectric detector, a signal preamplifier,
logarithmic range compensating circuitry, regulated power supplies, a
deter providing absorbance and percent-transmission scales and a 0-1
volt output for a strip chart recorder.
The Miran 1A System operates from either 110 or 220 V ac, 50-60 Hz
power supply. By means of an inverter, portable oper:r'on from a 12
volt battery is readily accomplished. This expands the instrument's
use to monitoring beyond the confines of the laboratory and greatly
facilitates the determination of environmental pollutants and meeting
OSHA (Occupational Safety and Health Administration) industrial require-
ments.
The Wilks Model 80 is a quantitative analysis system combining a high
performance single beam infrared spectrometer (Model 1A) with a
programmable microcomputer system (Table C2). This results in an
instrument of speed and versatility. The Model 80 accepts multicomponent
liquids, solids and/or gaseous samples directly without the necessity
of vaporizing or dissolving them, separating them into their individual
components.
The Model 80 has the following features:
• The measurement of up to 18 separate wavelengths in less than
two minutes. The interval between sets of measurements Is
variable from 7 seconds to 30 minutes.
C-l
-------
TABLE Cl
SPECIFICATIONS
SPECTROMETER
Type:
Wavelength Range:
Wavelength Range Control:
Resolution (Approximate):
Noise Level:
Drift:
Slit Settings:
Response Time Settings:
Absorbance Ranges:
Wavelength Drive Speed:
Infrared Source:
Infrared Detector:
Power Requirements:
Weight:
Single-beam infrared spectrometer.
2.5 to 14.5 ym in three steps:
2.5 to 4.5; 4.5 to 8; 8 to 14.5 with
small overlaps.
Manual or motor driven.
0.05 ym at 3 ym wavelength
0.12 ym at 6 ym wavelength
0.25 ym at 11 ym wavelength
Maximum of 0.003 absorbance units/8 hours
under the following conditions: 20.25
meter pathlength, 1 mm slit, 1 second
time constant, 12.0 ym wavelength, 23°C
oper Ing temperature.
Maxu um of 0.006 absorbance units/8 hours
using instrument conditions as specified
above for noise level.
Closed, 0.5, 1.0, 2.0 millimeters.
1, 4, 10 and 40 seconds.
*
0 to 0.025, 0 to 0.1, 0 to 0.25,
0 to 1 absorbance units and 0 to
100Z transmission.
2.5 minutes per segment.
Regulated nichrome wire heating element.
Pyroelectric type, lithium tantalate element.
25 watts at either 115 or 230 VAC, 50-60 Hz.
5.8 Kg (12.5 Ib) without cell.
11.6 Kg (30.0 Ib) with gas cell.
C-2
-------
Dimensions:
Temperature Range:
TABLE Cl (Continued)
245 x 155 x 155 mm without cell.
190 x 280 x 720 mm with cell.
0° to +40"C (32° to 104°F) Operating.
-20° to +60°C (-4° to 140°F) Storage.
VARIABLE PATH TWENTY METER GAS CELL
Pathlength: 0.75 to 20.25 meters in steps of 1.5 meter,
externally set.
7oluae: 5.6 liter.
Pressure Range, Operating: 10~ torr vacmsm to 1000 k?a (10 acaospheres).
Valves:
Internal Finish:
Windows:
Inlet and exhaust valves designed for both
vacuum and pressure, ptfe sealed.
ptfe lined, with mirrors and other components
gold plated.
NaCl or AgBr normally furnished. Others
on special order.
C-3
-------
TABLE C2
SYSTEM SPECIFICATIONS
Type:
Wavelength Range:
Resolution of CVF:
(full width at half height)
Minimum (Wavelength Step Size):
Wavelength Repeatability:
Noise Level:
Drift:
Photometric Accuracy:
Slit Settings:
Time Constant:
Full Scale Range:
Wavelength Drive Speed:
Power Requirements:
Ambient Temperature Range:
Dimensions:
Single-beam spectrometer.
2.5 to 14.5 ym.
0.05 y at 3.5 ym, 0.08 y at 6 urn,
0.25 y at 12 ym.
0.0005 pm at 3.5 urn* 0.0008 ym at 6 ym.
0.0016 ym at 12 ym.
0.0007 ym at 3.5 ym, 0.0003 ym at 12 ym,
48 hrs at 23°C.
-4
Max. of 1 x 10 absorbance units, without
cell, 1 mm slit; 3.5 microns, 1 x 10~3
absorbance units at 12 microns; 23°C.
Maximum of 0.002 absorbance units at 23°C,
3.5 microns, 24 hours.
Better than 0.1%.
0.5, 1, 2 mm, and closed.
0.25 seconds for recorder output.
1.6 absorbance units (useful range).
•
Selectable from 40 seconds to 1 hour.
100, 120, 220, 240 VAC + 152, -122, 50-60 3z,
75 watts.
0* to +40°C (32° to +104°F).
35 x 26 x 18 cm.
C-4
-------
H I I
• Signal averaging, 256 measurements are made at each wavelength
over a 1-180 second interval, enhancing the signal to noise
and improving precision.
• Quantitates up to 11 components with one reference wavelength.
Compensates for interferences. Absorbance repeatability within
the noise level.
• Keyboard entry of instrument settings and factors for data
manipulation. Changing instrumental conditions for a new
analysis is rapid, i.e., less than 10 minutes to set up a
5-component liquid analysis.
• Printout of memory parameters such as wavelengths, standard
factors in data matrix, gain settings, etc.
• Scans a short section of the spectrum (automatical!y) and
prints out absorbance as a function of wavelength.
• Data printout presentation includes absorbance for each
wavelength and concentration for each component.
• Digital display shows keyboard entry, and displays wavelength
during the analysis routine.
• Compatible with all Wilks standard liquid, gas and solid
sampling accessories.
(2) Operation
(a) Precautions and Preliminary Steps
Precautions
Using the particulate filter whenever sampling to prevent dust and dirt
from damaging optical components.
Avoid water condensation in the cell when the instrument is cold. Before
use, purge the cell with dry air or inert gas, such as nitrogen or
helium. Allow sufficient warm-up time to prevent water condensation on
cold internal surfaces.
Before storage or transporting the instrument in a cold environment,
purge the cell and close cell ports.
C-5
-------
Preliminary Steps
The operator should first perform Initial Checkout. The following
quick tests verify proper instrument operation and detect long-term
changes in performance.
1. Switch on POWER and allow 5 to 15 minutes warm up. Adequate
warm-up is indicated by no detectable down scale drift on
the 0.1A scale. Some random fluctuation due to noise is
normal and should not be confused with drift. Drift is
indicated by uniform meter deflection in one direction over
a period of a minute or two.
2. Set the following conditions:
RANGE ZT
PATH 0.75 aeters
SLIT 1 mm
WAVELENGTH 3.5 urn
ZERO CONTROLS XI, minimum (0.0)
RESPONSE TIME 1.0 second
Record the reading. This record will show long-term changes
in instrument performance. Some degradation in infrared
optical components should be expected. ' After several years
of normal service, cell windows, for example, may have to be
replaced. A continous record of this test is useful in
deciding whether instrument service is required. It will also
show sudden performance changes as occur with exposure of
NaCl windows to very wet samples or exposure of AgBr windows
to asmonia or pyridlne.
(b) Atmospheric Sampling Techniques
Atmospheric sampling requires only that samples be flushed directly
through the instrument via the sampling hose and particulate filter.
The analyzer's built-in pump is used. To minitor a single material
or contaminant, the analyzer is set to the appropriate wavelength for
the substance and changes in concentration are recorded. Periodic
flushings of the cell with zero gas or clean air is necessary to recheck
the zero absorbance setting. Where there is high level local contamination
of certain orgr.r^ic compounds, bottled air may be necessary for cell
flushing purposes. Such situations may arise during analysis within
closed environments such as r-jlvent tanks or storage areas or when
checking for leaks in pipes or duct work.
C-6
-------
(c) Monitoring Concentration of a Test Gas
Where several substances are of interest, such as in VOC screening*
select wavelengths to be used based on likely compounds (functional
groups) and proceed as follows:
1. The instrument should have first passed the Initial Checkout.
2. Set up instrument for wavelengths to be monitored.
3. Select a RESPONSE TIME setting that will-give a smooth
meter response without being unduly sluggish.
4. With clean air or "zero gas" in the cell, adjust for zero
absorbance reading with the ZERO CONTROL.
5. With the FUNCTION SWITCH on the 0-1 absorbance scale,
the meter should read zero.
6. Select the desired absorbance RANGE for aonitoring-
7. Connect the sampling hose and particulate filter to the
sample inlet port.
'8. Open VALVES and switch on the ambient air pump.
9. Read or record absorbance values.
«
•
(d) Calibration for Quantitative Analysis
For optisnm accuracy, it is necessary to calibrate the analyzer at the
wavelengths used for each sample. This is a one-time procedure unless
the wavelength filter ±s replaced or a new cell is installed. Absorbance
is recorded at one-to-five concentrations and a calibration curve
prepared. An inversion matrix may be used with the Model 80 micro-
processor to convert directly to ppmv.
------- |