Evaluation of aluminum canisters for the collection and storage of air

The use of a GC con- figuration coupled with the resolving power of capillary col- umns provides the potential for the rapid screening of chem- ically...
0 downloads 0 Views 557KB Size
Anal. Chem. 1990, 62, 1899-1902

transient adsorption studies to be performed on both bare and chemically modified quartz surfaces. The use of a GC configuration coupled with the resolving power of capillary columns provides the potential for the rapid screening of chemically modified surfaces for molecular recognition. Further work in this direction is currently in progress. Registry No. IPA, 67-63-0; silica, 60676-86-0.

LITERATURE CITED Blosensors: Fundemenfals and Appllcafions; Turner, A. P. F., Karube, I., Wilson, G. S., Eds.; Oxford University Press: Oxford, U. K., 1987. ChemiCel Sensor Technology; Seiyama, T., Ed.; Elsevier: New York, 1989; Vol. 1 and 2. Thompson, M.; Frank, M. D.; Heckl, W. M.; Marassi, F. M.; Vlgmond, S. J. I n Chemical Sensor Techno&y; Seiyama T., Ed.; Elsevier: New Yark. 2. rr DO _ 237-254. . -..., 1888: . _ _ _ Val. ,. -, _ .. Hecki, W. M.; Marassi, F. M.; Kallury, K. M. R.; Stone, D. C.; Thompson. M. Anal. Chem. 1990. 62. 32-37. Auld, B. A. Acoustic Fieus and Waves in Sol&; Wiley-Interscience: New York, 1973; Vol. 1 and 2. Dana, S. Surface Acoustic Wave Devices ; Prentlce-Hall: Enalewood Cliffs, NJ, 1986. Ristic, V. M. Rinclples of Acoustic Devlces; Why-Interscience: New Yark. . -. .., 1883. .- - -. D'Amico, A.; Verona, E. Sens. Actuators 1989. 77, 155-166. Bailantine, D. S.; Wohltjen, H. Anal. Chem. 1989, 67, 704A-715A. Fox, C. G.; Alder, J. F. Anelyst 1989, 774, 997-1004. Nieuwenhulzen, M. S.; Venema, A. Sensors Materials 1989, 5 , 261-300.

__

1899

(12) Wohltjen, H.; Dessy, R. Anai. Chem. 1979, 57, 1458-1464. (13) Wohltjen, H.; Dessy, R. Anal. C h m . 1979. 57, 1465-1470. (14) Hauden, D.; Michel, M.; Bardeche, G.; Gagnepain, J.J. Appl. Phys. Lett. 1977, 37,315-317. (15) Aider, J. F.; Fox, C. G.; Przybylko, A. R. M.; Rezgui, N.-D. D.; Snook, R. D. Analyst 1989, 714, 1183-1185. (16) Lewis, M, IEEE s,,,,,p, (Roc,) ,979, 612-622, (17) Wohltjen, H. Sens. Acfuators 1984, 5 , 307-325. (18) Wohltjen, H.; Snow, A.; Ballantine, D. I€€€ Unrason. Symp. (Roc .) 1985. 66-72. (19) Airokk C.; Santos, L. S., Jr. Thermochlm. Acta 1988, 704, 111-119. (20) Bernstein, T.; Michel, D.; Pfeifer, H.; Fink, P. J. Collokl Interface Sci. 1981, 84, 310-317.

Michael Thompson* David C. Stone Department of Chemistry University of Toronto 80 St. George Street Toronto, Ontario M5S 1Al Canada RECEIVEDfor review January 25,1990. Accepted May 17,1990. Support from the Institute of Chemical Science and Technology, Canada, and the Natural Sciences and Engineering Research Council of Canada for this work is gratefully acknowledged.

TECHNICAL NOTES Evaluation of Aluminum Canisters for the Collection and Storage of Air Toxics Alex R. Gholson, R. K. M. Jayanty,* and Julia F. Storm' Research Triangle Institute, P.O. Box 12194, Research Triangle Park, North Carolina 27709 Whole air collection techniques for determining volatile organic compounds (VOCs) in air have been widely used to study the effects of VOCs on atmospheric photochemistry and to monitor for toxic compounds in indoor environments, in the vicinity of point and area sources, and in ambient air. The passivated stainless steel canister has become the container of choice for whole air sampling because of the low background, ruggedness, and storage stabilities for most organic compounds (1-3). Limitations in the use of these containers include the surface reactivity with many oxygen-, nitrogen-, and sulfur-containing compounds, which results in significant wall losses, and the relatively high cost of the containers. As part of an audit material development program for hazardous waste incineration, the stability of 25 toxic organic compounds in high-pressure aluminum cylinders has been documented for up to 5 years a t a concentration as low as 5 ppbv. Of the organic compounds in the aluminum cylinder, five contained either nitrogen or oxygen (acetone, 1,4-dioxane, 2-butanone, acetonitrile, and acrylonitrile) and were found to be stable. Two oxygen-containing compounds, ethylene oxide and propylene oxide, were found to be unstable in the aluminum cylinder (4). A recent study showed that aluminum gas sampling loops provided better results for sampling oxygenated organics at the parts per billion by volume (ppbv) concentration than stainless steel loops (5). The stability of a compound in a high-pressure cylinder and in a flowing stream of gas in a sampling loop are substantially different than in a static, low-pressure (100-200 kPa) air sample. Current address: State of North Carolina, Division of Environmental Management, P.O. Box 27687, Raleigh, NC 27611. 0003-2700/90/0362-1899$02.50/0

Nevertheless, the findings suggest that aluminum may be a good material for sampling organic compounds at the part per billion level. This work presents the results of stability studies for 23 organic compounds in aluminum canisters and passivated aluminum canisters. Stability in the aluminum canister is a function of (1) surface reactions between the analytes and the canister walls, (2) reactions between analytes, and (3) reactions between analytes and other compounds present in the sample. To evaluate the aluminum canisters, the effects of the second and third mechanisms were minimized by using standard gas mixtures containing relatively nonreactive analytes prepared in either nitrogen or air. Because of the effect seen with water on passivated stainless steel canisters, water was added to some of the canisters. Water is believed to compete for active sites on the walls of the canisters, helping to passivate the surface. The amounts of water added was not great enough to result in condensation loss of the polar organics.

EXPERIMENTAL SECTION Canisters. Four spherical, 6-L aluminum canisters and one spherical, 6-L stainless steel canister provided by Andersen Samplers, Inc., were used for this evaluation. Two of the aluminum canisters and the stainless steel canister were passivated with the Summa process (Molectrics Corp.). A stainless steel bellows valve was fitted at the opening in the top of each canister. All canisters were cleaned with a series of evacuations and Nz pressurization at 150 "C followed by a pressurization with humidified N2 The canisters were then evacuated, refilled with N,, and blank checked by analyzing approximately 300 mL of N2from the canister. Sample Preparation. Stability studies were performed by using two gaseous mixtures of organic compounds. Samples were 0 1990 American Chemical Society

1000

ANALYTICAL CHEMISTRY, VOL. 62, NO. 17, SEPTEMBER 1, 1990

Table I. Results of t h e Stability Study i n Dry Aluminum Canisters" day 1

compound

expected concn, ppbv

day 0

A

B

A

vinyl chloride 1,3-butadiene bromomethane (ECD) trichlorofluoromethane (ECD) methylene chloride (ECD) chloroform (ECD) 1,2-dichloroethane l,l,l-trichloroethane (ECD) benzene carbon tetrachloride (ECD) 1,2-dichloropropane trichloroethylene toluene 1,l-dibromoethane (ECD) tetrachloroethylene chlorobenzene ethylbenzene o-xylene

3.91 9.30 4.10 3.49 4.32 4.38 4.03 4.06 4.33 3.96 4.07 4.29 4.30 4.26 4.63 4.15 3.89 4.11

0.97 0.84 0.86 0.96 0.71 0.56 0.05 0.74 0.33 0.96 0.06 0.45 0.08 0.02 0.36 tr tr tr

0.96 0.91 1.14 1.02 1.00 0.74 0.30 0.76 0.58 0.18 0.20 0.68 0.20 0.22 0.47 0.16 0.09 0.09

0.76 1.02 ND 0.85 0.67 0.56 tr 0.72 0.31 0.93 0.03 0.40 0.05 ND 0.35 0.02 tr tr

stability ratio (exutl/control) day 4 day 7 A B A B B

1.05 1.42 0.85 0.84 0.92 0.74 0.43 0.72 0.62 ND 0.18 0.66 0.18 0.15 0.50 0.13 0.04 0.04

1.23 0.30 ND 1.07 0.81 0.49 tr 0.76 0.40 0.87 0.04 0.46 0.07 tr 0.42 0.04 tr NA

1.05 0.83 0.51 1.00 0.97 0.60 0.44 0.78 0.69 ND 0.24 0.74 0.22 0.06 0.59 0.16 0.04 0.02

1.23 0.08 ND 0.78 0.75 0.58 tr 0.67 0.40 0.58 tr 0.41 0.07 ND 0.39 0.04 tr 0.01

1.13 0.79 0.23 0.72 0.98 0.69 0.44 0.74 0.71 ND 0.24 0.70 0.24 0.04 0.57 0.19 0.05 0.02

day 10

A

B

1.71 ND ND 0.78 0.72 0.64 0.05 0.72 0.40 0.52 tr 0.40 0.06 ND 0.42 0.04 0.01 0.01

1.77 1.47 0.18 0.72 0.94 0.71 0.49 0.73 0.79 ND 0.30 0.77 0.28 0.02 0.70 0.24 0.06 tr

" A = results for unpassivated canisters; B = results for passivated canisters; ND = not detected; tr = detected but below the limit of quantification; NA = results not available because of interference. Table 11. Results of t h e Stability Study with 2000 ppmv W a t e P

compound

A

B

vinyl chloride 1,3-butadiene bromomethane (ECD) trichlorofluoromethane (ECD) methylene chloride (ECD) chloroform (ECD) 1,2-dichloroethane l,l,l-trichloroethane (ECD) benzene carbon tetrachloride (ECD) 1,2-dichloropropane trichloroethylene toluene 1,2-dibromoethane (ECD) tetrachloroethylene chlorobenzene ethylbenzene o-xylene

3.91 9.30 4.10 3.49 4.32 4.38 4.03 4.06 4.33 3.96 4.07 4.29 4.30 4.26 4.63 4.15 3.89 4.11

1.00 0.98 1.12 1.06 1.10 1.06 1.30 1.13 1.11 1.02 1.00 1.09 1.03 1.35 1.03 1.04 0.98 1.59

1.27 0.92 1.18 1.02 1.05 1.02 1.03 0.98 1.03 0.91 1.00 1.10 1.02 1.26 1.15 1.10 1.04 1.05

0.98 NA 1.12 0.96 1.08 0.99 1.12 1.02 1.28 1.05 1.02 1.03 1.05 1.11 1.10 1.06 1.06 1.27

1.10 1.38 1.13 0.94 1.09 1.24 1.05 0.99 0.99 1.02 0.99 1.00 0.97 1.14 1.06 0.98 1.00 1.15

NA 1.50 1.09 1.09 1.16 0.85 1.34 1.16 1.25 1.13 1.14 1.13 1.10 1.24 1.16 1.14 1.10 1.18

11.1 15.7 9.43 17.1 19.2 10.1 10.2

0.97 0.55 1.00 1.00 0.04 0.71 0.57

0.96 0.42 1.01 1.02 0.09 0.68 0.58

1.05 0.43 1.03 0.98 0.05 0.61 0.43

1.05 0.47 1.02 0.96 0.02 0.72 0.62

CFC 114b(ECD) acetoneb 1,l-dichloroethyleneb CFC 113b(ECD) 1,4-dioxaneb tolueneb chlorobenzeneb

day 0

stability ratio (exptl/control) day1 day 4 day 7 A B A B A B

expected concn, ppbv

day 10

A

B

1.38 1.27 1.24 1.10 1.22 0.91 1.38 1.12 1.15 0.93 1.14 1.17 1.08 1.30 1.18 1.16 1.12 NA

1.43 1.37 0.98 0.78 1.11 1.03 1.20 1.04 1.18 0.98 1.00 1.02 1.02 0.97 1.05 1.12 1.01 NA

1.38 1.25 1.03 0.80 1.00 1.00 1.01 0.97 1.02 0.64 0.99 0.98 0.98 1.06 1.00 1.06 1.09 0.60

NA NA 1.15 0.85 NA 1.07 1.11 1.13 1.16 1.03 1.04 1.00 1.02 1.04 1.20 1.07 1.00 1.03

1.83 0.91 1.04 1.13 1.55 1.05 0.98 0.99 1.01 0.52 1.00 0.99 0.95 1.20 1.06 1.14 1.50 1.59

l.lOC

l.llC

0.4OC 1.02c 0.97c trc 0.56c 0.38'

0.5OC 1.02c 0.93c trc 0.72c 0.62c

1.01 0.38 0.99 0.99 tr 0.54 0.35

1.00 0.54 1.01 0.88 tr 0.76 0.67

1.0P 0.36c 1.05c 0.95c trc 0.43c 0.35c

1.08c 0.58c 1.06c 0.7F trc 0.67c 0.76c

" A = results for unpassivated canisters; B = results for passivated canisters; NA = not available because of interference. Concentration of water was 170 ppmv for this compound. CResultswere obtained for days 3 and 14, respectively; tr = detected but below the limit of quantification. prepared by adding certified gas standards (Scott Specialty Gases) to evacuated canisters. For samples containing water, a heated tee with a septum was placed between the standard and the canister. Deionized water that had been boiled t o remove any residual VOCs was injected into the canister through the septum. Table I includes the compounds and concentrations in the standards used t o prepare the evaluation canisters. For the stability study, a standard containing the 18compounds shown in Table I (which were in one cylinder) was added to all four aluminum canisters. Approximately 2000 ppmv of water was added to one passivated and one unpassivated canister. The remaining seven compounds in Table I1 (which were in a different

cylinder) were added to one passivated aluminum canister, one unpassivated aluminum canister, and the one passivated stainless steel canister. Approximately 170 ppmv of water was added to each canister. Analysis. For the 18-component study, the canister samples were analyzed on days 0, 1,4,7, and 10. For the seven-component study, the canister samples were analyzed on days 0, 1, 3, 7, and 14. All analyses were performed by gas chromatography (GC) with a flame ionization detector (FID) and an electron-capture detector (ECD). The air samples were concentrated by cryogenic trapping in a liquid argon cooled loop packed with glass beads. The volume

ANALYTICAL CHEMISTRY, VOL. 62, NO. 17, SEPTEMBER 1, 1990

concentrated was measured with a calibrated ballast system. The concentrated sample was flash-desorbed with boiling water onto a 60 m X 0.325 mm i.d. DB-1 fused-silica capillary column (J & W Scientific) in a carrier gas flow of helium at a linear velocity of 28 cm/s. The oven temperature was held initially at 10 "C for 2 min before it was increased to a final temperature of 150 "C at a rate of 8 OC/min. The column effluent was split by a ratio of 1:5 between the ECD and the FID with a fused-silica outlet splitter (SGE, Inc.). The analog signal from each detector was processed with a dual-channel integrator. The GC-FID-ECD system was calibrated with certified standard gas cylinders at the same concentration as that of the standard used to prepare the samples. A second standard used in the same concentration range was analyzed at the end of each day as a quality control check. An analytical system blank check was performed each day by analyzing approximately 300 mL of hydrocarbon-free N2 to ensure that no interference due to contamination or sample carryover was occurring.

RESULTS AND DISCUSSION The compounds in the 18-component standard represent a collection of nonpolar organics that have been found to be stable for up to 2 weeks in humidified passivated stainless steel canisters (3). The stability in aluminum canisters appears to be dependent on the amount of moisture in the sample and to some extent the passivation treatment applied to the canister. Tables I and I1 show the results of the dry and wet unpassivated aluminum canisters. The ratio between the measured canister concentration versus the control for each day is shown. Compound stability in the dry unpassivated canisters was extremely poor. Only vinyl chloride, trichlorofluoroethane, methylene chloride, and l,l,l-trichloroethane were consistently greater than 70% of the control over the 10-day period. Less than 10% of 1,2-dichloroethane, 1,2-dichloropropane7 toluene, 1,2-dibromoethane, chlorobenzene, ethylbenzene, and o-xylene added was detected in the sample within an hour after it was prepared. 1,3-Butadiene, bromomethane, and carbon tetrachloride had sample concentrations between 80 and 100% of the control's concentration, but 1,3-butadiene and bromomethane rapidly went to zero with time, whereas carbon tetrachloride slowly dropped to a value of 50% of the added concentration after 10 days. Concentrations of the other compounds remained constant and were between 10 and 70% of their control values. In contrast to the dry unpassivated sample, the wet unpassivated sample's stability was excellent. The concentration ratios for all compounds were between 0.98 and 1.60 initially, and no significant loss was found after 10 days. Some analytical difficulties, probably due to water, resulted in interferences with some of the early eluting compounds. The drastic effect water has on the stabilities of the 18component standard in aluminum canisters is also evident in the comparison of the passivated canisters. Tables I and I1 also show the results of the dry and wet passivated samples, respectively. In the dry passivated canisters, concentration ratios consistently greater than 0.70 were found for vinyl chloride, 1,3-butadiene, trichlorofluoromethane, methylene chloride, chloroform, and l,l,l-trichloroethane. Ratios consistently less than 0.20 were found for carbon tetrachloride, chlorobenzene, ethylbenzene, and o-xylene. Bromomethane and 1,2-dibromoethane showed a steady drop in their ratios from 1.14 to 0.18 and 0.22 to 0.02, respectively. All the remaining compounds had stable ratios between 0.20 and 0.70. In contrast, the wet passivated sample had ratios between 0.91 and 1.27 initially, and no significant loss was found for any of these compounds over the 10-day study except carbon tetrachloride. Carbon tetrachloride started at a ratio between 0.90 and 1.02 for the first 4 days but dropped to 0.52 after 10 days. Carbon tetrachloride was found to be more stable in the unpassivated canisters than the passivated canisters, both wet and dry.

1901

Table 111. Results of the Control Cylinder Analyses

compound vinyl chloride 1,3-butadiene bromomethane (ECD) trichlorofluoromethane (ECD) methylene chloride (ECD) chloroform (ECD) 1,2-dichloroethane l,l,l-trichloroethane (ECD) benzene carbon tetrachloride (ECD) 1,2-dichloropropane trichloroethylene toluene l,2-dibromomethane (ECD) tetrachloroethylene chlorobenzene ethylbenzene o-xylene fluorocarbon 114 (ECD) acetone 1,l-dichloroethylene fluorocarbon 113 (ECD) 1,4-dioxane

toluene chlorobenzene

av concn, ppbv f SD 3.5 f 0.7 7.2' f l.gb 4.1 f 0.3 4.6' f 0.5 4.4 f 0.3 4.4 f 0.3 4.6" f 0.2 4.1 f 0.2 4.1 f 0.2 3.9 f 0.4 4.1 f 0.2 4.3 f 0.2 4.3 f 0.3 4.3 f 0.2 4.3 f 0.4 3.7 f 0.Bb 4.4 f 0.5 4.8' f 0.6 11.1f 0.5 15.7 f 0.6 9.36 f 0.1 17.1 f 0.2 18.8 f 0.8 9.75 f 0.3 10.0 f 0.2

cert concn, PPbV 3.91 9.30 4.10 3.49

4.32 4.38 4.03 4.06 4.33

3.96 4.07 4.29

4.30 4.26 4.63

4.15 3.89 4.11 11.1

15.7 9.43 17.1 19.2 10.1

10.2

'Concentration is greater than 1standard deviation from certified value. Standard deviation is greater than 20%. Table I11 shows the 5-day average concentration for the control standard. The values obtained were compared with those certified by comparison with NBS primary standards, and all but four compound values were within 1 standard deviation of the certified value. Only two compounds, 1,3butadiene and chlorobenzene, had a relative standard deviation greater than 20%. Most standard deviations were less than 10% of the concentration value. This indicates that the analytical accuracy and precision for the standard analysis was within a range normally considered acceptable at these low concentrations. The day-to-day variability of the canister sample analyses was larger than the variability of the standard analyses for compounds that were found to be stable. Sample variability could be due either to small changes in the surface activity of the canister that were caused by fluctuations in room temperature or to analytical interference with water present in some of the samples or with trace impurities present in the canister. The second stability study involved seven compounds known to be more reactive with metal surfaces because of their polarity. Water was added to each sample, but the amount added was approximately 25 times less than the wet samples of the 18-component stability study. The stability results for the unpassivated and passivated aluminum samples in the presence of 170 ppm water vapor are also shown in Table 11. Little difference in stability can be seen between the two canisters for these seven compounds. There was no loss of the two chlorofluorocarbons and 1,l-dichloroethylene. Acetone, toluene, and chlorobenzene had initial stability ratios of only 0.4-0.7. The passivated aluminum canister showed no loss with time for these three compounds, whereas the unpassivated canister showed a slight loss with time. Less than 10% of the 1Q-dioxane added was found initially in both aluminum canisters, and it rapidly dropped below the detection limit after 3 days of storage. The stability ratios for toluene and chlorobenzene in the seven-component stability study at 170 ppmv H20 are shown in Table IV. These ratios fell between the 2000 ppmv H,O sample and dry sample

Anal. Chem. 1990, 62,1902-1904

1902

Table IV. Results of the Seven-Component Stability Study in Passivated Stainless Steel Canister with 170 ppmv of Water stability ratio (exptl/control) compound

day 0 day 1 day 4 day 7 day 14

1,2-dichloro-1,1,2,2-tetrafluoroethane (ECD)

0.98

1.05

1.09

1.01

1.05

acetone 1,l-dichloroethylene

0.99

0.98

0.98

0.88

1.00 1,1,2-trichloro-1,2,2-trifluor- 1.02

1.06 0.88

1.03 0.94

1.02 1.00

0.76 1.07 0.86

oethane (ECD) 1,4-dioxane

0.53

0.36 1.00 1.00

0.31 0.98 0.99

0.29 0.98 1.00

0.23 0.84 1.05

toluene

chlorobenzene

0.96 0.99

stability for toluene and chlorobenzene in the 18-component stability studies. This suggests that the stability of these two compounds may be a function of the water vapor concentration. Table IV shows the compound stabilities of the sevencomponent standard in the passivated stainless steel canister. Initial stability ratios for six of the seven compounds were between 0.96 and 1.02. 1,4-Dioxane has an initial stability ratio of 0.53, and a steady decrease to 0.23 was found over 14 days of storage. A small but steady decrease in the stability ratio with time was found for acetone. A linear correlation between stability ratio and time was found for acetone at 1.7% loss per day with B correlation coefficient of 0.99 and for 1,4-dioxaneat a 1.6% loss per day with a correlation coefficient of 0.79. The similar slopes found for both of these compounds suggest a common mechanism for the loss. Both 1P-dioxane and acetone are proton acceptors and electron pair donors and would form hydrogen bonds with available protons or Lewis acid-base adducts with metals. Aluminum, being an unusually strong Lewis acid, is an especially good electron acceptor and would have a strong affinity for acetone and l,4-dioxane, which is confirmed by the poor recovery of these compounds in the aluminum canister.

CONCLUSIONS The applicability of using aluminum canisters to collect ambient-level air samples depends on the reactivity of the compounds of interest and the concentration of water in the sample. Compounds with only slightly polar properties and lower volatiles (bp > 60 "C) are unstable in both dry un-

passivated and passivated aluminum canisters. To prevent a large initial loss of these compounds, it is estimated that the concentration of water in the canister sample must be greater than approximately 500 ppmv (1.6% relative humidity a t 25 "C). Summa passivation of the aluminum surface slightly lowers the amount of compound lost initially in dry samples and requires shghtly less water to completely passivate the surface. For polar oxygenated compounds such as acetone and l,Cdioxane, aluminum canisters are more reactive than the passivated stainless steel and should therefore not be used to sample for polar organic compounds. From the results of this study, the selected organic compounds showed no increase in stability when collected and stored in aluminum as compared to those collected and stored in stainless steel canisters. The hydrocarbons and halogenated hydrocarbons tested were found to be stable for more than a week if levels of water were greater than 500 ppmv. Further work is needed to find an effective passivation process that would both eliminate the need for water in the sample and provide a stable surface for collecting the polar organic compounds. Registry NO.CFC114, 76-14-2; CFC113,76-13-1;aluminum, 7429-90-5; stainless steel, 12597-68-1; vinyl chloride, 75-01-4; 1,3-butadiene, 106-99-0;bromoethane, 74-83-9; trichlorofluoromethane, 75-69-4;methylene chloride, 75-09-2;chloroform, 67-66-3; 1,2-dichloroethane, 107-06-2; l,l,l-trichloroethane, 71-55-6; benzene, 71-43-2; carbon tetrachloride, 56-23-5; 1,2-dichloropropane, 78-87-5; trichloroethylene, 79-01-6; toluene, 108-88-3; 1,2-dibromoethane, 106-93-4; tetrachloroethylene, 127-18-4; chlorobenzene, 10890-7;ethylbenzene, 100-41-4;o-xylene, 9547-6; acetone, 67-64-1; 1,l-dichloroethylene, 75-35-4; 1,4-dioxane, 123-91-1.

LITERATURE CITED (1) Harsch, D. E. Atmos. Environ. 1980, 14, 1105. (2) Westberg, H. H.; Holdren, M. W.; HIII, H. H. Analytical Methodology for the Identification and Quantlflcatbn of Vapor Phase Organic PoHutants. Final report the Coordinated Research Council on CRC-APRAC Project No. CAPA-11-71, Report No. 82-13-5, January, 1982. (3) Oliver, K. D.; Pleil. J. D.; McClenny, W. A. A t m s . Envlron. 1988, 20, 1403. (4) Jayanty, R. K. M.; Cooper. S. W.; Decker, C. E.; Von Lehmden, D. J. J. Air Poiiut. ControiAssoc. lS85, 35, 1195. (5) Allen, J. M.; Jayanty, R. K. M.; von Lehmden, D. Anal. Chem. 1087, 59. 1002-1084.

RECEIVED for review December 27,1989. Accepted April 23, 1990.

Separation of Ethyl Methacrylate-Butyl Methacrylate Copolymers by Liquid Adsorptlon Chromatography with an Ultraviolet Absorptlon Detector Sadao Mori

Department of Industrial Chemistry, Faculty of Engineering, Mie University, Tsu, Mie 514, J a p a n

INTRODUCTION Although molecular weight averages and a molecular weight distribution (MWD) of a homopolymer can be measured by size exclusion chromatography (SEC), those of a copolymer cannot be obtained accurately without knowing the chemical composition distribution (CCD) of the copolymer ( I ) . Therefore, the measurement of a CCD is required before calculating molecular weight averages and a MWD of the copolymer in order to obtain accurate data. The separation of styrene (SI-methyl methacrylate (MMA) random copolymers by high-performance liquid adsorption

chromatography (LAC) has been reported in our previous papers (1-6), and the technique was applied to the separation of styrene-alkyl methacrylate and styrene-alkyl acrylate copolymers (7) and S-MMA block copolymers (8), respectively. Fractionation by LAC gave a CCD, and a real MWD has been obtained by SEC followed by LAC of SEC fractions (5). Besides our work, several attempts have been reported for the separation of copolymers according to composition by high-performance liquid chromatography (HPLC): e.g., SMMA (9,lO);S-methyl acrylate (11);S-acrylonitrile (12);and S-butadiene (13). Similar to our work, these reports utilized

0003-2700/90/0362-1902$02.50/0 0 1990 American Chemical Society