Temperature-Dependent Electron Capture Detector Response to Common Alternative Fluorocarbons

Article abstract of DOI:10.1021/ac9703921

The relative electron capture detector (ECD) response to alternative fluorocarbons (AFCs) using gas chromatography are found to be at least 1 order of magnitude lower than that for CFC-12. Detection limits for the chlorofluorocarbons CFC-12, HCFC-22, HCFC-123, and HCFC-124 are found to be 2.5, 90, 30, and 90 pg, respectively. Those for the hydrofluorocarbons are significantly poorer; 14 and 45 ng for HFC-125 and HFC-134a, respectively. HFC-152a was not detected using ECD. Since atmospheric concentrations of these compounds are in the low part-per-trillion level, GC-ECD is apparently not sensitive enough to be used for AFC analysis without substantial preconcentration. Two columns are evaluated for the AFC separation. The Poraplot Q WPLOT column showed good separation ability, though column bleed limits detection performance. A Carboxen 1004 packed column exhibits much lower interference. But separations are time consuming and peak broadening adversely affects limits of detection. Mechanisms for the ECD response are proposed based on thermodynamics and temperature-dependent ECD responses. CFC-12, HCFC-123, and HFC-125 apparently undergo ion-forming dissociative electron capture. The electron capture process for HCFC-22 and HFC-134a appear to form molecular ions. Both mechanisms appear to be operative for HCFC-124 electron capture. Dissociative electron capture rate constants for HCFC-123, HCFC-124, and HFC-125 are estimated to be 3.5 × 10^-10, 1.0 × 10^-10, and 5.6 × 10^-13 cm³ s^-1, respectively at 300 °C.

Full text of DOI:10.1021/ac9703921

Anal. Chem. 1997, 69, 3871-3878
Te m p e ra t u re -De p e n d e n t Ele c t ro n Ca p t u re
De t e c t o r Re s p o n s e t o Co m m o n Alt e rn a t ive
Flu o ro c a rb o n s
Sonia R. Sous a a nd Ste phe n E. Bia lkow s ki*
Department of Chemistry and Biochemistry, Utah State University, Logan, Utah 84322-0300
The relative electron capture detector (ECD) response to
alternative fluorocarbons (AFCs) using gas chromatogra-
phy are found to be at least 1 order of magnitude lower
than that for CFC-1 2 . Detection limits for the chlorofluo-
rocarbons CFC-12, HCFC-22, HCFC-123, and HCFC-124
are found to be 2 .5 , 9 0 , 3 0 , and 9 0 pg, respectively.
Those for the hydrofluorocarbons are significantly poorer;
hydrogen allows for relatively rapid atmospheric reaction with OH
and O ( D). Faster tropospheric reactions predictably result in
1
lower atmospheric lifetime and correspondingly lower fractions
entering the stratosphere.6 Production of HFC compounds is
unregulated. HCFC compounds are classified as class II sub-
stances in the Clean Air Act since they contain chlorine. Produc-
tion of the later is scheduled to halt by the year 2030.
-8
1
4 and 4 5 ng for HFC-1 2 5 and HFC-1 3 4 a, respectively.
Predictions regarding the environmental fate of the AFC
compounds were summarized by the Alternative Fluorocarbon
Environmental Acceptability Study (AFEAS).8,9 Progress toward
understanding the chemistry and toxicology of the degradation
products was reviewed.10 Products thought to occur with the
largest yields are the carbonyl and acetyl halides.11,12 Acid,
hydroxide, peroxide, and nitrate degradation products will prob-
ably not prevail in high concentration since they are soluble and
HFC-1 5 2 a was not detected using ECD. Since atmo-
spheric concentrations of these compounds are in the low
part-per-trillion level, GC-ECD is apparently not sensitive
enough to be used for AFC analysis without substantial
preconcentration. Two columns are evaluated for the AFC
separation. The P oraplot Q WP LOT column showed good
separation ability, though column bleed limits detection
performance. A Carboxen 1 0 0 4 packed column exhibits
much lower interference. But separations are time con-
suming and peak broadening adversely affects limits of
detection. Mechanisms for the ECD response are pro-
posed based on thermodynamics and temperature-de-
pendent ECD responses. CFC-1 2 , HCFC-1 2 3 , and HFC-
2 2
hydrolyze. Carbonyl halides, CFClO, CF O, and CCl O, have been
observed in the stratosphere.13,14 Dissolution is thought to be the
main loss mechanism for carbonyl compounds.8 However, labora-
tory measurements indicate that the carbonyl halides are not very
soluble.15 Atmospheric lifetimes are estimated to be from 5 to 30
2 3 3
years for CF O, CFClO, CF C(O)F, and CF C(O)Cl and up to 1500
years for HFCO.10 It has been suggested that some fluorocarbon
radicals, considered as benign to the ozone layer, may actually
destroy it through catalytic processes.16 AFC compounds are
effective absorbers of infrared radiation, and increased use may
contribute to global warming.7
The underlying assumption to the fates of the AFC compounds
is that atmospheric chemistry models are accurate enough to
predict atmospheric fate based on laboratory measurements. A
1
2 5 apparently undergo ion-forming dissociative electron
capture. The electron capture process for HCFC-2 2 and
HFC-1 3 4 a appear to form molecular ions. Both mecha-
nisms appear to be operative for HCFC-1 2 4 electron
capture. Dissociative electron capture rate constants for
HCFC-1 2 3 , HCFC-1 2 4 , and HFC-1 2 5 are estimated to
-
1 0
-1 0
, and 5 .6 × 1 0 -1 3 cm3
-1
be 3 .5 × 1 0
, 1 .0 × 1 0
s
,
respectively at 3 0 0 °C.
(
6) Fisher, D. A.; Hales, C. H.; Gilkin, D. L.; Ko, M. K. W.; Sze, N. D.; Connell,
P. S.; Wuebbles, D. J.; Isaksen, I. S. A.; Stordal, F. Nature 1 9 9 0 , 344, 508-
The production and use of most volatile chlorocarbons is
restricted by international agreement because of their contribution
512.
to stratospheric ozone depletion.1-4 The program for the phase-
(7) Fisher, D. A.; Hales, C. H.; Wang, W.; Ko, M. K. W.; Sze, N. D. Nature
1
9 9 0 , 344, 513-516.
out of use of these compounds in the United States is Title VI of
the Clean Air Act Amendments of 1990.5 To circumvent problems
associated with banning these substances, hydrofluorocarbons
(
(
8) Scientific Assessment of Stratospheric Ozone: 1989, Volume II Appendix:
AFEAS Report World Meteorological Organization, Global Ozone Research
and Monitoring Project, Report 20, 1990.
9) Watson, R. T.; Kurylo, M. J.; Prather, M. J.; Ormond, F. M. Present State of
Knowledge of the Upper Atmosphere 1990: An Assessment Report, NASA
Reference Publication 1242, 1990.
(
HFCs) and hydrochlorofluorocarbons (HCFCs) have been in-
troduced as substitutes. These are collectively referred to as the
alternative fluorocarbons (AFCs). The AFC compounds have
physical properties similar to CFCs, but the incorporation of
(
(
10) Wallington, T. J.; Schneider, W. F.; Worsnop, D. R.; Nielsen, O. J.; Sehested,
J.; Debruyn, W. J.; Shorter, J. A. Environ. Sci. Technol. 1 9 9 4 , 28, 320A-
326A.
11) Tuazon, E. C.; Atkinson, R. J. Atmos. Chem. 1 9 9 3 , 16, 301-312.
(
(
(
1) Rowland, F. S. Environ. Sci. Technol. 1 9 9 1 , 25, 622-628.
2) Hamill, P.; Toon, O. B. Phys. Today 1 9 9 1 , 44, 34-42.
3) Simmonds, P. G.; Cunnold, D. M.; Dollard, G. J.; Davies, T. J.; McCulloch,
A.; Derwent, R. G. Atmos. Environ. 1 9 9 3 , 27A, 1397-1407.
4) Cunnold, D. M.; Fraser, P. J.; Weiss, R. F.; Prinn, R. G.; Simmonds, P. G.;
Miller, B. R.; Alyea, F. N.; Crawford, A. J. J. Geophys. Phys. 1 9 9 4 , 99, 1107-
(12) Tuazon, E. C.; Atkinson, R. J. Atmos. Chem. 1 9 9 3 , 17, 179-199.
(13) Rinsland, C. P.; Zander, R.; Brown, L. R.; Farmer, C. B.; Park, J. H.; Norton,
R. H.; Russel, J. M.; Raper, O. F. Geophys. Res. Lett. 1 9 8 6 , 13, 769-772.
(14) Wilson, S. R.; Crutzen, P. J.; Schuster, G.; Griffith, D. W.; Helas, G. Nature
1 9 8 8 , 334, 689-691.
(
(15) De Bruyn, W. J.; Duan, S. X.; Shi, X. Q.; Davidovits, P. Geophys. Res. Lett.
1
126.
1 9 9 2 , 19, 1939-1942.
(
5) Reilly, W. K. Fed. Regist. 1 9 9 1 , 56, 2420.
(16) Zhuangjie, Li; Francisco, J. S. Chem. Phys. Lett. 1 9 9 3 , 209, 151-159.
S0003-2700(97)00392-2 CCC: $14.00 © 1997 American Chemical Society
Analytical Chemistry, Vol. 69, No. 19, October 1, 1997 3871

more direct approach is to measure concentrations in situ. There
have been relatively few reports of measured atmospheric
abundance of AFCs given the potential impact.17-23 HCFC-22
the solid stationary phase columns tested were apparently capable
of performing adequate separation of volatile AFC compounds.
The purpose of this study is to evaluate the potential for using
the ECD to detect AFCs. The ECD response to the common
AFCs are measured relative to that of CFC-12, which is used as
an internal standard. We also examine the utility of a Poraplot-Q
WPLOT and Carboxen 1004 packed column in the separation of
the AFCs when ECD is used. No attempt was made to enhance
the ECD signals using oxygen doping.28,29 This technique has
been shown to increase detection limits for some AFC com-
pounds.24 On the hand, ECD response is known to be affected
by detector temperature. We perform a temperature-dependent
response study. The temperature-dependent response results
allows deduction of the electron capture mechanism and molecular
ion thermodynamics.
(
CHF
2
Cl), a refrigerant, has been in use the longest and has
received the most attention.1
7-20
3 2
HCFC-142b (CH CFCl ), used
since 1988 as a foam blowing agent and a solvent, has also been
measured.17,21,22 Atmospheric concentration trends for the popular
refrigerant HFC-134a (CF CH
3 2
F) have recently been reported.23
For the most part, these reports indicate higher than expected
concentrations.
With the exception of a single passive infrared absorption
report,19 AFC compounds have been detected using gas chroma-
tography with mass spectrometry (GC/ MS) operating in mass-
selective detection mode. Air samples are collected in either
stainless steel or aluminum cylinders and stored for later labora-
tory analysis. Analysis samples are concentrated prior to GC/
MS analysis. Detection requires preconcentration of nearly 1 L
of whole-air sample. Preconcentration has been performed using
cryogenic trapping techniques. There are problems with this
method such as supplying the liquid nitrogen, the cost, and the
band broadening that occurs when large volumes of sample are
desorbed and introduced into the column. Several researchers
EXPERIMENTAL SECTION
Apparatus. A Hewlett-Packard Model 5890a gas chromato-
62
graph equipped with both a Model G1223A Ni â source, electron
capture detector (ECD) and a Model 19231D/ E flame ionization
detector (FID) is used in this study. The ECD is a potentially
hazardous radioactive source and a semiannual wipe test is
required to assess its stability. The ECD and FID are used at
temperatures of 300 and 250 °C, respectively, except for the ECD
temperature-dependent response study described below. The
ECD makeup flow rate is 20 mL min- nitrogen. Gas samples
are introduced either by direct injection using the standard
splitless injector or by volumetric injection using a Valco, T series,
six-port valve with a 10 µL volume injection loop. The direct
injector is maintained at 200 °C for all experiments. The six-port
injection valve is mounted in the gas chromatograph oven.
Ultrapure nitrogen (Whitmore Oxygen) is used as the carrier and
make-up gas. Oxygen trap (Oxy-trap Alltech) and a molecular
sieve 40/ 50 mesh (Hewlett-Packard) are used to purify the gas
supply. Carrier and make-up gas flow rate are measured with
the Hewlett-Packard electronic flow sensing.
have reported the use of “microtraps” with carbonaceous adsor-
bents to address these problems.2
4-26
Simmonds et al.26 reported
on an automated GC/ MS instrument for field measurements using
both glass bead and carbonaceous adsorbent “microtraps” to
concentrate the AFC compounds. Since selective-ion monitoring
was not used, several liters of air are required for concentration
prior to GC/ MS analysis. Problems with this method are that
large numbers of potential interferents have to accounted for in
the MS data analysis and that the large sample volumes are subject
to problems associated with the use of adsorbent traps. The large
size of the instrument and the requirement for liquid nitrogen
prevent use of this method for true field measurements.
Sturges and Elkins24 reported on the use of electron capture
detection (ECD) with ambient temperature traps for preconcen-
tration. This method has potential for field monitoring due to
the relatively small size, the cost, and no need for cryogenics.
Speciation would have to be totally accomplished with chroma-
tography, though some selectivity may be attained with the ECD.
Sturrock et al.27 have evaluated several wide-bore capillary columns
for use in AFC analysis. Although no single column was perfect,
1
Separations are performed using two columns. The first is a
3
0 m, 0.53 mm internal diameter WPLOT Poraplot Q (J&W
Scientific, CA) column. The stationary phase of the Poraplot Q
is a styrene-divinylbenzene copolymer similar in characteristics
to the Porapak phase used in packed columns. The second
column is a Carboxen 1004 packed column (Supelco). This
stainless steel column has a length of 2 m, with 0.75 mm internal
diameter. Carrier flow rates are experimentally derived using a
Van Deemter plot of data obtained under isothermal conditions.
(
(
(
17) Pollock, W. H.; Heidt, L. E.; Lueb, R. A.; Vedder, J. F.; Mills, M. J.; Solomon,
S. J. Geophys. Res. 1 9 9 2 , 97, 12993-12999.
18) Montzka, S. A.; Myers, R. C.; Butler, J. H.; Elkins, J. W.; Cummings, S. O.
Geophys. Res. Lett. 1 9 9 3 , 20, 703-706.
19) Risland, C. P.; Goldman, A.; Murcray, F. J.; Blatherwick, R. D.; Kosters, J.
J.; Murcray, D. G.; Sze, N. D.; Massie, S. T. J. Geophys. Res. 1 9 9 3 , 95,
The carrier flow rates used in this study are 5.6 mL min- for the
1
Poraplot Q column and 9.6 mL min-1 for the Carboxen 1004
column. These carrier flow rates are used for isothermal condi-
tions and for the starting temperature with temperature program
conditions. Separation of the compounds tested could be achieved
isothermally and with temperature programming on the Poraplot
Q column. In order to achieve the separation of HFC-125, HCFC-
1
6477-16490.
20) Irion, F. W.; Brown, M.; Toon, G. C.; Gunson, M. R. Geophys. Res. Lett. 1994,
1, 1723-1726.
(
(
2
21) Schauffer, S. M.; Heidt, L. E.; Pollock, W. H.; Gilpin, T. M.; Vedder, J. F.;
Solomon, S.; Lueb, R. A.; Atlas, E. L. Geophys. Res. Lett. 1 9 9 3 , 20, 2567-
2
570.
22) Montzka, S. A.; Myers, R. C.; Butler, J. H.; Elkins, J. W. Geophys. Res. Lett.
9 9 4 , 21, 2483-2486.
2
2, and HCFC-124, on the Carboxen 1004 column, a temperature
(
(
program is used: 80 (1 min) and then 20 °C min- ramp to the
final 220 °C. Resolution was still poor in this case. Analysis time
was ∼12 min with the temperature program. Quantitative detector
response and retention time data are obtained using 200 °C
isothermal conditions by injecting each AFC compound in a
mixture with CFC-12.
1
1
23) Montzka, S. A.; Myers, R. C.; Butler, J. H.; Elkins, J. W.; Lock, L. T.; Clarke,
A. D.; Goldstein, A. H. Geophys. Res. Lett. 1 9 9 6 , 31, 169-172.
(
(
24) Sturges, W. T.; Elkins, J. W. J. Chromatogr. 1 9 9 3 , 642, 123-134.
25) O’Doherty, S. J.; Simmonds, P. G.; Nickless, G. J. Chromatogr. A 1 9 9 3 ,
6
57, 123-129.
(
(
26) Simmonds, P. G.; O’Doherty, S.; Nickless, G.; Swaby, R. A.; Knight, P.;
Ricketts, J.; Woffendin, G.; Smith, R. Anal. Chem. 1 9 9 5 , 67, 717-723.
27) Sturrock, G. A.; Simmonds, P. G.; Nickless, G. J. Chromatogr. A 1 9 9 3 , 648,
(28) Grimsrud, E. P.; Miller, D. A. Anal. Chem. 1 9 7 8 , 50, 1141-1145.
(29) Miller, D. A.; Grimsrud, E. P. Anal. Chem. 1 9 7 9 , 51, 851-859.
4
23-431.
3872 Analytical Chemistry, Vol. 69, No. 19, October 1, 1997

Ta b le 1 . P h ys ic a l P ro p e rt ie s o f AFC a n d CFC S p e c ie s
Us e d in Th is S t u d y
trade name
formula
bp (°C)
molecular mass
CFC-12
CCl2F2
CHClF2
-29.8
-40.8
27.6
-11.0
-48.5
-26.2
-25.0
120.91
86.7
HCFC-22
HCFC-123
HCFC-124
HFC-125
HFC-134a
HFC-152a
CF3CHCl2
CF3CHClF
CF3CHF2
CF3CH2F
CH3CHF2
152.90
136.48
119.98
102.0
66.03
Temperature-Dependent ECD Response. Data for the
temperature-dependent ECD response to the tested AFC and CFC
compounds is obtained by monitoring the relative response of
standardized gas mixtures while varying the ECD temperature
from 200 to 350 °C in increments of 30 °C. Separation is
performed with the Poraplot Q column under isothermal condi-
tions. Carrier and make-up gas flow rates are also the same as
those used in the separation and relative response studies.
Data Analysis. The data are acquired with either a Sergent-
Welch Model XKR chart recorder or a PC computer with a
Metrabyte Model DAS-HRES, 16-bit analog-to-digital converter
board. The data collection software is written in Turbo-C (Bor-
land). The data acquisition software records the chromatogram
and writes files that may be subsequently analyzed. Chromato-
gram data are processed using Quattro-Pro (Borland). A 30-point
moving average smoothing filter is often used to improve the
signal-to-noise ratio (SNR). Peak areas are estimated using
triangulation based on peak tangents. Three or more replicate
measurements were used to estimate the detector response of
each compound. A relative ECD response is defined as the ratio
of the peak area of the compound of interest relative to that of
CFC-12 at equal injected amounts.
Fig u re 1 . Chromatographic analysis of a mixture of AFC-22, -124,
and -125 and CFC-12 on the Poraplot Q with ECD detection and
temperature programming. (1) HFC-125; (2) HCFC-22; (3) CFC-12;
(4) HCFC-124.
Ta b le 2 . Re t e n t io n Tim e s (Min u t e s ) a n d Re la t ive ECD
Re s p o n s e (3 0 0 °C) o f t h e Flu o ro c a rb o n s Us e d in Th is
S t u d ya
retention time
compound
Poraplot Q
Carboxen 1004
rel ECD response
1
CFC-12
HCFC-22
HCFC-123
HCFC-124
HFC-125
5.8
4.7
14.3
8.3
4.0
4.4
10.0
3.5
123
14.5
2.1
-2
(2.0 ( 0.3) × 10
-
-
-
1
2
4
(1.1 ( 0.2) × 10
(3.2 ( 0.2) × 10
(1.8 ( 0.1) × 10
(4.7 ( 0.1) × 10
not detected
-5
HFC-134a
HFC-152a
2.8
3.4
4.6
a
-1
Initial Poraplot Q flow rate was 5.6 mL min . The temperature
program was used. Carboxen 1004 flow rate was 9.6 mL min-1. Column
temperature was 200 °C. An FID was used to obtain the retention time
for HFC-152a.
Reagents and Sample P reparations. HCFC-22, HCFC-123,
HCFC-124, HFC-125, HFC-134a, and HFC-152a were received as
a gift-in-kind from AFEAS, furnished by Allied, DuPont, and ICR.
The physical and chemical properties of the AFCs and CFCs
studied in this paper are given in Table 1. CFC-12 and HCFC-
peratures. A program with an initial temperature of 45 °C for 5
min followed by a 9 °C min-1 ramp to 180 °C final temperature
was found to produce satisfactory resolution of all components
within 16 min. The program also helped reduce detector noise
due to column bleed over the majority of this time. Retention
times for this temperature program are given in Table 2.
1
34a was purchased from PCR. A primary CFC-12 standard 10.5
ppm (v/ v) in Ar (Matheson) is used as an internal standard for
the mixtures. All these compounds are the highest purity available
Carboxen 1004 is a porous carbonaceous stationary phase with
a high affinity for smaller molecules.33 All compounds tested on
(
at least 99%) and used without further purification. The samples
are prepared by volumetric dilution of the standards with ultrapure
argon (99.999%, Matheson, CA) in two calibrated glass flasks. The
this column are strongly adsorbed, and elevated temperatures are
needed to get elution in reasonable times. The elution time order
is HFC-125 < HFC-152a < HFC-134a < HCFC-22 < CFC-12 <
HCFC-124 < HCFC-123. HCFC-123 is strongly retained even at
4
flask pressure is maintained at 9.2 × 10 Pa under Ar. Glass beads
are added in order to enhance mixing upon shaking.
2
2
25 °C. A temperature program found to resolve HFC-125, HCFC-
2, and HCFC-124 is 80 °C for 1 min with a 20 °C min-1 ramp to
RESULTS
Separation of the AFC on P oraplot Q and Carboxen 1004.
On Poraplot Q, separation of CFC-12, HCFC-22 and -124, and HFC-
a final temperature of 220 °C. Column bleed is significantly less
than found for the Poraplot Q WPLOT column even at the highest
temperature (220 °C). The Carboxen 1004 column is not expected
to perform as well as the Poraplot Q. The column geometry is
less favorable for separation being much shorter, wider, and
having a more heterogeneous stationary phase than the WPLOT
Poraplot Q.
125 is obtained with the isothermal temperature of 100 °C. Figure
1
shows a typical chromatogram obtained using the Poraplot Q
WPLOT column under isothermal conditions. The elution time
order of HFC-125 < HCFC-22 < CFC-12 < HCFC-124 < HCFC-
1
23 follows the boiling points of these substances. Separation of
these substance takes ∼10 min. However, HCFC-123 is more
strongly adsorbed and does not elute in reasonable times at 100
(30) Clemons, C. A.; Altshuller, A. P. Anal. Chem. 1 9 6 6 , 38, 133-136.
(
31) Pellizzari, E. D. J. Chromatogr. 1 9 7 4 , 98, 323-361.
°
C. Increasing the temperature decreases the HCFC-123 elution
(
32) Chen, E. C. M.; Albyn, K.; Dussack, L.; Wentworth, W. E. J. Phys. Chem.
time, but the lower molecular weight substances are not resolved.
In addition, significant column bleed occurs at the higher tem-
1
9 8 9 , 93, 6827-6832.
(33) Bruno, T. J.; Wertz, K. H.; Caciari, M. Anal. Chem. 1 9 9 6 , 68, 1347-1359.
Analytical Chemistry, Vol. 69, No. 19, October 1, 1997 3873

Fig u re 4 . Detector response as a function of inverse temperature
from 200 to 350 °C for HCFC-124. Injection of 1 µL of 1.3 parts-per-
thousand (v/v).
Fig u re 2 . Detector response as a function of inverse temperature
from 200 to 350 °C for CFC-12, HCFC-125, and HCFC-123. Injection
of 1 µL of 10.5 ppm (v/v).
As seen in Figure 2, CFC-12 and HFC 125 have negative slopes,
indicating that electron attachment occurs by the dissociative
mechanism. HCFC-22 and HFC-134a have positive slopes, shown
in Figure 3, indicating the electron attachment is due to a
reasonable molecular electron attachment mechanism. A complex
Arrhenius plot trend for HCFC-124 can be seen in Figure 4.
Assuming the trends at the extreme temperatures are lines, the
ECD response for this species may be a combination of both
molecular ion formation and dissociative mechanisms. In this
case, dissociative electron capture occurs at temperatures greater
than ∼280 °C, and electron attachment is a resonance mechanism
at lower temperatures.
Fig u re 3 . Detector response as a function of inverse temperature
from 200 to 350 °C for HCFC-22 and HCFC-134a. Injection of 1 µL
of 42 parts-per-thousand (v/v).
DISCUSSION
A. Separation of the AFCs on P oraplot Q and Carboxen
1
0 0 4 . Several types of columns have been evaluated for
resolving the AFCs, and no single column is entirely satisfactory.27
Alumina suffers from a major drawback of inducing a dehydro-
genation reaction of smaller HCFCs. Since the degree of the HCl
elimination reaction is a function of column temperature and
elution time, this stationary phase is not satisfactory for HCFC
analysis. It was also shown that the Poraplot Q WPLOT produces
good separation of the AFCs.27 A resolution greater than 1 is
observed for all of these compounds. Higher temperatures are
required for the timely elution of HCFC-123. Unfortunately, severe
column bleeding is observed at temperatures above 200 °C. A
temperature program is used to maintain separation between HFC-
125 and HCFC-22 while producing timely elution of HCFC-123.
The lower, beginning temperature produces band broadening.
Carbonaceous molecular sieves are used as adsorbent for their
suitability as concentrating traps for AFCs.25 Favorable Kovats
indices for AFC separation on hexafluoropropylene-modified
carbonaceous phase,33 and the reported high affinity of carbon-
aceous phases toward fluorocarbons, prompted us to test the
relatively new Carboxen 1004 packed column stationary phase as
a potential column for AFC separation.33 Another reason for
testing this column is that column bleed could be lower than that
observed for the polymer phases, thus lowering detection limits.
In fact, negligible column bleed was observed, even at the highest
operating temperatures. This suggests that temperature program-
ming could be used to separate all the components of interest:
AFC-125, -152a, -22, -12, -124, and -123. We observed poor
resolution from the attempt to separate the AFCs. The highest
Relative ECD Response. Table 2 shows the relative re-
sponses of ECD to the AFCs. The relative response is calculated
as the peak area ratio of the AFC to that obtained for CFC-12
with equal mole injections. CFC-12 was used as an internal
standard to avoid errors due to variation in injection volume. Listed
uncertainties are (σ obtained from replicate measurements.
It is easy to see that the ECD has a much lower response to
the AFCs than to CFC-12. The relative response apparently
decreases with substitution of hydrogen for chlorine on the
molecule, as in the response of HCFC-22 (CHF
2
Cl) relative to
CFC-12 (CF Cl
2
2
). This trend has been reported in the past.30 The
major trend gleaned from this table is that the relative ECD
response decreases with increasing number of hydrogen atoms
on the molecule.
Temperature-Dependent ECD Response. Data represent-
ing the temperature-dependent ECD response are illustrated in
Figures 2-4. Each datum represented in the plots is the peak
area of the species on the chromatograms. The data are plotted
in Arrhenius fashion, i.e., the natural logarithm of the relative peak
area as a function of the inverse temperature. Slopes obtained
by linear regression are proportional to the activation energy for
ion fragmentation or the enthalpy of molecular anion formation.
A negative Arrhenius plot slope indicates a dissociative ion
fragmentation electron attachment mechanism, whereas a positive
slope indicates electron attachment by the resonance mecha-
nism.31,32
3874 Analytical Chemistry, Vol. 69, No. 19, October 1, 1997

-
-
R value obtained was 0.56. HFC-123 is strongly retained on this
column, rendering quantitative analysis impractical. This has been
observed using Carboxen 1000 and 1003 adsorbents trap.25
Temperatures in excess of 260 °C are necessary to desorb HCFC-
CF Cl + e h CF Cl
∆H° ) -5.4 kcal/ mol (2)
∆H° ) +3.1 kcal/ mol (3)
∆H° ) -37 kcal/ mol (4)
∆H° ) +2.5 kcal/ mol (5)
2
2
2
2
-
2
-
CF Cl f CF Cl‚ + Cl
2
2
-
-
1
23. These results can be justified by the fact that all these
CF Cl‚ + e h CF Cl
2 2
2
stationary phases have a large surface area, 1200 and 750 m / g
for the Carboxen 1000 and the Carboxen 1004, respectively, and
strong interactions between the solute and the adsorbate have
been shown to occur.25 In the end, our attempt to obtain better
detection limits was not successful. Detection limits were not
improved because the peaks were much broader than those
obtained using the Poraplot WPLOT column.
-
-
CF Cl f CF + Cl
2
2
Overall, the heat of reaction to produce two chlorides is -34.5
kcal/ mol. However, the first dissociation, eq 3, is significantly
endothermic relative to kT (∼1.1 kcal/ mol at 300 °C). If energy
from electron capture in eq 2 is lost by collision faster than
chloride dissociation, this step may limit the overall rate for
efficiency electron capture. The negative slope obtained in the
Arrhenius plot shown in Figure 2 is consistent with a forward
reaction rate-limiting step. The slope indicates an Arrhenius
activation energy of E
range of this study. The activation energy is based on the
assumption that k . k and is thus the activation energy for eq
This value is in reasonable agreement with values of E
between 3.4 and 3.6 kcal/ mol reported previously.34
HCFC-123 (CF CHCl
) exhibits ∼1 order of magnitude lower
B. Mechanisms for AFC ECD Response. The instrumental
response of an ECD is directly related to the number of electrons
removed from the gas by capture processes.31 Although several
individual factors like the cross section for electron collision,
electron affinity, unimolecular ion reactions, and cation/ anion
reactions are all important in interpreting absolute response, it is
easier to interpret the response by simpler general mechanisms
prior to describing the details. The temperature-dependent
A
) 6 kcal/ mol over the limited temperature
2
1
3
.
A
response can yield definitive information in this regard.
A general mechanism for alkyl halide electron capture is32
3
2
k1
k2
ECD response than CFC-12. Both molecules possess the same
number of chlorine atoms. In fact, based on the number of
fluorine atoms, HCFC-123 might be predicted to have an ECD
response greater than that of CFC-12. The overall lower ECD
sensitivity, compared to CFC-12, may be explained by the fact
that the electron-withdrawing, and thus bond-weakening, effect
of fluorine at the labile C-Cl bond is reduced in HCFC-123
because the fluorines are on an adjacent carbon atom. This lowers
the relative ECD response compared to chlorofluorocarbons with
the same number of chlorine atoms by strengthening the C-Cl
bond energy. In fact, a higher activation energy is expected due
to the increased bond strength.
The Arrhenius plot for the ECD response is shown in Figure
2. As with CFC-12, the negative slope is indicative of dissociative
electron capture. If loss of the first chloride is the rate-limiting
step for electron capture, the lower response may be attributed
to the fact that the first C-Cl bond strength could be higher than
that in CFC-12 due to hydrogen and carbon substituents on the
carbon with chlorine. For example, the first C-Cl bond enthalpy
-
-
-
AX + e y z AX
98 A + X
(1)
k-1
-
where X is a halogen. Halide, X , production is not energetically
favorable for fluorine and may be unfavorable for chlorine as well.
2
When the rate of halide production, the k process, is slow on the
time scale of the ECD measurement, electron capture is essentially
an equilibrium between the molecule and the molecular anion,
-
AX . The ECD response is proportional to the ratio of forward,
1
k , to reverse k-1, rate constants comprising the equilibrium
-
-
constant K ) [AX ]/ [AX][e ]. In the linear operating range of
the detector, electron concentrations are much greater than that
of the ion or molecule. The ECD response is proportional to
-
-
[
AX ] ) K[AX][e ]. Thus, larger K result in greater relative
signals. Large ECD responses are often observed for species that
stabilize the captured electron in resonant orbitals, e.g., alkyl
fluorocarbon and aromatic compounds.
Equilibrium allows connection to the thermodynamics of
molecular ion formation. The enthalpy for electron attachment
increases with number of hydrogens in the CH
x
Cl4-x chlo-
may be estimated from the slope of Arrhenius plots of ln[ST3/ 2
vs inverse temperature,31,32 where S is the ECD signal. The
romethane series. The higher C-Cl bond strength would
increase the minimum activation energy for molecular ion dis-
sociation producing chloride ion, thereby decreasing the ECD
response at a given temperature.
]
electron attachment mechanism is indicated for a positive Arrhe-
nius plot slope. In this case d ln[ST3/ 2]/ d(1/ T) ) -∆H/ k, where
k is Boltzmann constant.
The regression slope indicates an activation energy of E
.3 kcal/ mol. Assuming that the activation energy is the energy
required for loss of the first chloride, the heat of formation of the
CF CHF radical may be estimated from the dissociative capture
mechanism
A
)
7
When the rate for halide production from the molecular anion
is fast, the free electron is efficiently captured by dissociative halide
production.31 In this case, the rate-limiting step is often the rate
3
2
at which halide ions are produced, essentially, k . An Arrhenius
plot of ln[S] versus inverse temperature is more appropriate.32
The dissociative electron attachment mechanism is indicated by
a negative Arrhenius plot slope. In this case d ln[S]/ d(1/ T) )
-
-
CF CHCl + e f CF CHCl‚ + Cl
3
2
3
∆H ) +7.3 kcal/ mol (6)
E
A
/ k, where E
A
is the activation energy for halide production.
ECD is particularly sensitive to chlorofluorocarbons. This may
where the apparent heat of reaction is that determined from the
be due to favorable thermodynamics for chloride ion formation.
In the case of CFC-12, a thermodynamically favorable mechanism
is the multiple-step process
(
34) Smith, D.; Adams, N. G.; Alge, E. J. Phys. B: At. Mol. Phys. 1 9 8 4 , 17, 461-
472.
Analytical Chemistry, Vol. 69, No. 19, October 1, 1997 3875

Arrhenius plot activation energy. Based on this, and the heats of
formation of HCFC-123 and chloride, the heat of formation of the
Ta b le 3 . He a t s o f Fo rm a t io n (k c a l/m o l) o f Ha lo c a rb o n s
a n d At o m s a t S t a n d a rd S t a t e (2 9 8 .1 5 K)a
CF
3
CHF radical is ∼-114 kcal/ mol.
The enthalpy for chloride ion production from HCFC-22
CHClF ) electron capture is slightly endothermic
∆
Hf°
∆Hf°
neutral
molecule
neutral
anion
fragment
anion
(
2
CF2Cl2
CHF2Cl
CF3CHCl2
-117.90 -123.33
-115.6
CF3
-111.7
-59.2
-7.8
-154.9
-27.7
CHF2
CH2F
CH3
CF2Cl
CF3CF2
CF3CHF
CF2
CHF
H
F
-
-
CHF Cl + e f CHF ‚ + Cl
∆H° ) 0.5 kcal/ mol (7)
2
2
-177.0
CF3CHFCl* -207.8
34.8
139
-103.0
-255.0
CF3CHF2
CF3CH2F
CH3CHF2
CF2CF2
CHFCF2
HF
-264.0
-214.1
-119.7
-157.4
-117.4
-65.13
-66.0
-213.0
-162.7
-44.6
39.0
An ECD response on the same order of magnitude as CFC-12
might be expected on the basis of reaction enthalpy alone.
However, as shown in Table 2, the relative ECD response to this
compound is a factor of 50 less than that for CFC-12. In addition,
the Arrhenius plot shown in Figure 3 exhibits a positive slope,
indicating that molecular electron attachment is not followed by
chloride ion formation. The activation energy for chloride ion
production is apparently greater than the electron affinity of HCFC-
-102.0
<27.0
33.2
-59.5
-55.9
52.1
18.36
29.0
Cl
a
Heats of formation are from ref 43, the NIST WWW CKMech
database, or the JANAF tables except for (*), which are estimated using
HyperChem with the semiempirical AM1 method.
2
2. In this case, the slope of the Arrhenius plot (ln[ST3/ 2] vs 1/ T)
is related to the enthalpy for molecule electron capture. The
positive slope indicates an enthalpy for electron capture of ∆H )
A four-center HF elimination step
-
-
6.1 kcal/ mol. This places the heat of formation for the CHClF
ion at ∆H ) ∼-121.7 kcal/ mol.
HCFC-124 (CF CH Cl) has a relative response similar to
2
-
-
2
CF CF H + e f CF CF + HF
(9)
3
2
2
3
2
HCFC-22 at 300 °C. However, the Arrhenius plot in Figure 4
shows a complex thermal behavior for electron capture. The slope
is negative slope at high temperatures and becomes positive at
low temperatures. This behavior is indicative of the electron
capture mechanism of eq 1, but with competition between the
back reaction producing the neutral and an electron, and stable
anion formation by dissociation of the molecular ion.31 At low
temperatures, the temperature-dependent response is that for
molecular ion formation. As the detector cell temperature is
increased, more energy is available; thus the production of a stable
may be possible. The neutral reaction
CF CF H f CF CF + HF ∆H° ) 41.5 kcal/ mol (10)
3
2
2
2
requires relatively less energy to proceed. If the electron affinity
of C is sufficiently high, then this reaction might be responsible
2 4
F
for the positive activation energy. Perfluoro-2-butene and 1-(tri-
fluoromethyl)-1-difluoroethane apparently form very stable mo-
lecular anions,36 indicating a high electron affinity. The electron
2 5 2 3
affinity for C F and C F are in the -41 to -55 kcal/ mol range.
-
ion, perhaps Cl , increases.
On the other hand, positive electron attachment activation ener-
gies have been observed for some compounds.34,37,38 It is possible
that the measured activation energy is that required for molecular
electron capture. In fact, positive activation energies are observed
for perfluoromethylcyclohexane molecular electron capture.38,39
ECD response to the HFCs should not be due to a dissociative
mechanism. Because of the relatively high C-F bond strength,
electron capture cannot lead to stabilization via fluoride ion
formation.35 As shown in Table 2, ECD response to HFC
compounds decreases in the order: HFC-125 (CF
34a (CF CFH ) . HFC-152a (CH CHF ). The response is clearly
related to the number of fluorides on the substituted ethane. HFC-
25 and HCFC-134a have trifluoromethyl groups, which may help
stabilize the captured electron relative to over HFC-152a.
ECD response to HFC-125 (CF CHF ) is nearly 4 orders of
magnitude lower than that for CFC-12. The Arrhenius plot shown
in Figure 2 indicates an activation energy, E ) 8.6 kcal/ mol for
3 2
CF H) > HFC-
(
See Table 3.)
1
3
2
3
2
HFC-134a has an ECD response that is nearly 5 orders of
magnitude less than that for CFC-12. In addition, the Arrhenius
plot exhibits a positive slope, indicating a direct electron attach-
ment mechanism
1
3
2
-
-
CF CH F + e h CF CH F
(11)
3
2
3
2
A
electron stabilization, even in the lower temperature range. It is
doubtful that electron stabilization occurs by a direct (two-center)
dissociative mechanism because the ion-forming reactions are
highly endothermic;
The slope of the Arrhenius plot indicates an enthalpy for electron
attachment of ∆H ) -10.3 kcal/ mol. The heat of formation of
the molecular anion is thus estimated to be ∆H ) -223 kcal/
f
mol at 300 K.
-
-
•
•
A question arises regarding why the ECD response is so much
lower for the HFC than for CFC-12. The -10.3 kcal/ mol electron
CF CF H + e f CF + CHF2
∆H° ) 49.9 kcal/ mol
∆H° ) 61.1 kcal/ mol
∆H° ) 124.6 kcal/ mol
3
2
3
-
-
CF CF H + e f CF CF + H
3
2
3
2
(
36) Ravishankara, A. R.; Solomon, S.; Turnipseed, A. A.; Warren, R. F. Science
9 9 3 , 259, 194-199.
1
CF CF H + e f CF + CHF2-
-
•
3
2
3
(37) Wentworth, W. E.; Chen, E. C. In Electron Capture; Zlatkis, A., Poole, C. F.,
Eds.; Elsevier: Amsterdam, 1981; Chapter 8, pp 151-191.
(
8)
(
38) Culbertson, J. A.; Grimsrud, E. P. Int. J. Mass Spectrom. Ion Processes 1 9 9 5 ,
49, 87-98.
1
(
35) Knighton, W. B.; Grimsrud, E. P. J. Am. Chem. Soc. 1 9 9 2 , 114, 2336-
342.
(39) Smith, D.; Herd, C. R.; Adams, N. G.; Paulson Int. J. Mass Spectrom. Ion
2
Processes 1 9 9 0 , 96, 431.
3876 Analytical Chemistry, Vol. 69, No. 19, October 1, 1997

attachment enthalpy for HFC-134a is significant compared to kT
at 300 °C. One plausible explanation for the low relative response
constants may be estimated in this case.
-
9
3
-1
Smith et al. reported a rate constant of 3.2 × 10 cm s for
is that the molecular anions react with N +
, present in the ECD
dissociative electron capture by CFC-12 (CCl
F
+ e f CClF
-
+
2
2
2
2
-
34
detector, producing free electrons. This type of reaction has been
observed for larger fluoroalkanes.38
Cl ) at 300 K. Based on this rate constant, and the relative
response of the AFCs found to capture electrons by a dissociative
process in Table 2, reaction rate constants for HCFC-123, HCFC-
C. Detection Limits. Detection limits are estimated based
on signals produced for the relatively high injection amounts and
the baseline noise of a chromatogram. The noise is primarily due
to column bleed. A SNR of 3 is used to estimate the detection
limit. The detector is assumed to respond linearly down to the
detection limit. The 10 µL sample loop was used to inject 9.9
ppm (v/ v) primary CFC-12 standards onto the Poraplot column
at 100 °C. The resulting ECD peak and baseline were analyzed
with the spread sheet program. SNRs based on both the peak
height to baseline noise standard deviation and the integrated area
to integrated baseline standard deviation were obtained. SNRs
of 510 and 80 were obtained for peak height and peak area
measurements, respectively. These correspond to SNR ) 3, CFC-
-
10
-10
-13
124, and HFC-125 are 3.5 × 10 , 1.0 × 10 , and 5.6 × 10
3
-1
cm
s
, respectively, at 300 °C. These rate constants can be
extended to other temperatures with the activation energies.
These rate constants probably describe the decomposition after
capturing a free electron since the temperature-dependent re-
sponse studies indicate dissociative electron capture for these
compounds.
The CFC and AFC compounds may react with free electrons
present in the atmosphere, thus initiating the destruction of these
compounds. In fact, the role of electron capture-initiated decom-
position has been reported to be a major decomposition route in
considerations regarding the fate of long-lived halogenated
compounds.36,41 Although ions are found throughout the atmo-
sphere, electron densities are significant only above 50-60 km.
To predict the destruction of the AFCs in the mesosphere, the
mechanism by which the molecule captures the electron needs
to be considered. Compounds such as CFC-12, HCFC-123, HCFC-
124, and HFC-125, which apparently undergo rapid dissociative
electron capture, would readily decompose in an electron-rich
environment.
The fate of compounds that apparently form stable molecular
anions is more complicated since electron capture does not alone
lead to decomposition. The reaction of the molecular ion with
cations, in this case N
2
+, produced by â collisions, complicates
the assessment of the degree to which electron capture will
decompose the molecule. Culbertson and Grimsrud have shown
that a large fraction of the perfluoromethylcyclohexane anion is
converted back to the neutral through reaction with cations
present in an ECD.38 They proposed that the atmospheric
chemistry of species that form stable molecular anions is deter-
mined by the recombination of the anions with positive ions by
1
2 mass detection limits of 2.5 pg for peak height and 16 pg for
peak area using the ECD. Although integrated peak area is
normally expected to yield higher SNR, excessing baseline drift
and noise due the column resulted in lower SNR, even for this
ideal case. If baseline drift were perfectly reproducible, the
minimum extrapolated detection obtained from the noise alone
would be 14 fg.
Based on the LOD for CFC-12, and the relative responses of
the AFC, mass detection limits for HCFC-22, -123, and -124 are
9
0, 30, and 90 pg, respectively. The HFC mass detection limits
are considerably higher, 14 and 45 ng for HFC-125 and -134a,
respectively. These results are worse than those of Vidal-Madjar
et al.40 Their ECD mass detection limits for CFC-11 are 75 fg in
pulsed detector mode, 250 ft in dc mode, corrected for a SNR of
3
. A mass detection limit for CFC-12 of 1.6 pg for dc detector
mode is estimated based on their data. However, these detection
limits were extrapolated by assuming constant baseline noise. Our
detection limits for HCFC-123 and HFC-134a are comparable to
those of Sturges and Elkins.24 Mass detection limits of 12 and
4
10 pg for HCFC-123 and HFC-134a, respectively, may be obtained
-
+
from the pulsed mode ECD concentration detection limits reported
in the later paper. Although it is possible to enhance ECD
sensitivity to halocarbon species with hydrogen, using oxygen in
the makeup gas of the ECD detector,29 the noise level is also
enhanced. Sturges and Elkins found that detection limits for
HCFC-123 were better without oxygen due to the increased noise
level associated with oxygen.
D. Electron Capture Rates and Atmospheric Fat. The
ECD can be used to measure the relative dissociative electron
capture rate constant using a calibration compound for which the
electron capture rate constant is known.32 The conditions have
to be such that the compound concentration is not significantly
depleted by the dissociative electron capture reaction itself. The
later is created by using a fast pulse rate. The measured response
should be proportional to the rate constants. Although the
Hewlett-Packard ECD detector operates in a variable pulse rate
mode, the relative AFC signals were adjusted by changing the
relative concentration such that maximum signals were about
the same as that for CFC-12. Thus, the detector pulse rates
were similar for the all compounds. Dissociative capture rate
M + p f M + p
(12)
(13)
-
+
M + p f P + neutrals
+
where p is a positive ion. Equation 12 indicates that the molecule
is regenerated to the original molecule. In eq 13, molecular ion
reaction produces new product compound and, thus, degradation.
For compounds such as HCFC-22, HCFC-124 (at low tempera-
tures) and HFC-134a, recombination with positive ions could
convert the molecular anion to the original molecule or result in
new species. The branching ratio to produce the parent neutral
can be high.38
Contributions to the atmospheric decomposition of species that
decompose upon capturing an electron may be extremely low.
The lifetime of SF
estimated as being greater than 600-800 years.42 The rate
constant for SF is much greater than that of the AFC compounds
studied here. The estimated rate constants for HCFC-123 and
6
due to dissociative electron capture has been
6
(
41) Morris, R. A.; Miller, T. M.; Viggiano, A. A.; Paulson, J. F.; Solomon, S.;
Reid, G. J. Geophys. Res. D 1 9 9 5 , 100, 1287-1294.
(
40) Vidal-Madjar, C.; Parey, F.; Excoffier, J. L.; Bekassy, S. J. Chromatogr. 1 9 8 1 ,
03, 247-261.
(42) Lais, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; Mallard,
2
W. G. J. Phys. Chem. Ref. Data 1 9 8 8 , 17, Suppl. 1.
Analytical Chemistry, Vol. 69, No. 19, October 1, 1997 3877

-
10
3
-1
HCFC-124 are on the order of 10 cm s while HFC-125 is on
Ta b le 4 . Ma s s De t e c t io n Lim it Es t im a t e s Ba s e d o n a
S ig n a l-t o -No is e Ra t io o f 3 fo r 3 0 0 °C ECD w it h P o ra p lo t
Q Co lu m n
the order of 10- cm s . By comparison, the electron attach-
13
3
-1
-7
ment rate constant for SF
6
is 3.1 × 10 cm3 s-1 at 300 K.
Assuming that the atmospheric lifetime is inversely proportional
to the rate constant, the lifetime of AFC compound would be from
compound
LOD (pg)
concn LOD for 1 L sample
CFC-12
2.5
90
30
90
14
0.2 ppt (v/ v)
9.7 ppt (v/ v)
1.8 ppt (v/ v)
6.1 ppt (v/ v)
1.1 ppb (v/ v)
4.1 ppb (v/ v)
3
to 6 orders of magnitude greater than that of SF
to 10 years if electron capture were the only decomposition route.
The tropospheric lifetime of the AFC compounds based on OH
and O( D) are estimated to be between ∼2 years, for HCFC-123,
6
, or from 106
HCFC-22
HCFC-123
HCFC-124
HFC-125
HFC-134a
HFC-152a
9
1
45
nd
6
,8
a
a
and ∼20 years, for HCFC-22.
Clearly, dissociative electron
nd
capture decomposition plays a minor role at best in the decom-
position of the AFC species.
a
nd, not detected.
given above.6 HCFC-22 is apparently vented at greater rates than
have the CFC compounds. Thus, our simple estimation calcula-
tion above is conservative in that real future concentrations will
probably be higher.
CONCLUSION
The maximum atmospheric concentration of the AFC com-
pounds may only on the order of a few parts-per-trillion (ppt).
Assuming that atmospheric concentrations of these species have
reached steady state due to vertical and horizontal mixing, that
steady-state concentrations are inversely proportional to atmo-
spheric lifetime, τ, and that AFC release rates will equal the past
CFC-11 rates
Our mass detection limits indicate that the ECD may be
adequate for quantitation of the HCFC compounds if preconcen-
tration is used. Shown in Table 4 are concentration detection
limits for the AFC compounds for a 1 L sample size, assuming
complete recovery from preconcentration. Concentration detec-
tion limits in the low ppt range should be adequate. These
indicate that from 10 to 100 L samples would have to be used to
reach the limit of quantitation (10σ) for the HCFCs. The present
atmosphere contains less of the newer HCFC compounds, so
preconcentration of over 100 L is needed for quantitation. There
is an apparent need for better selective detection if the AFC
concentrations are to be monitored to in our changing atmosphere.
CAFC ) CCFC-11(τAFC/ τCFC-11
)
The atmospheric lifetime of CFC-11 is ∼60 years. AFC concentra-
tions can be found from estimates of atmospheric lifetimes based
1
29
on the reaction with OH, O ( D), and vertical migration. Future
global concentrations for AFCs are as follows (ppt, v/ v): HCFC-
22, 90; HCFC-123 10; HCFC-124, 40; HFC-125, 170; HFC-134a, 110;
and HFC-152a, assuming that the AFCs will be used in the same
capacities and amounts of CFC-11. By comparison, Irion et al.20
ACKNOWLEDGMENT
We thank the Alternative Fluorocarbon Environmental Accept-
ability Study (AFEAS) for supplying the alternative fluorocarbons.
found 1993 CFC-22 concentrations to be ∼115 ppt (v/ v) and
-
1
increasing at ∼6.6% year . HCFC-22 has not been in use for as
long as the CFCs though its concentration is relatively high.
However, Montzka et al.18 found that global averaged concentra-
tions are consistent with an atmospheric lifetime of 13.6 years,
which is, in fact, significantly shorter than most model predictions
based on the reaction rate constants used to obtain the estimates
Received for review April 11, 1997. Accepted July 18,
X
1997.
AC9703921
X
Abstract published in Advance ACS Abstracts, September 1, 1997.
3878 Analytical Chemistry, Vol. 69, No. 19, October 1, 1997