ANALYST, JANUARY 1987, VOL. 112 45 Gas Chromatographic Determination of Acrolein in Rain Water Using Bromination of 0-Methyloxime Harumitsu Nishikawa and Tomokuni Hayakawa Gifu Prefectural Research Institute for Environmental Pollution, 58-8, Yabuta, Gifu-shi, 500 Japan and Tadao Sakai Department of Chemistry, Asahi University, 185 7 Hozumi, Hozumi-cho, Gifu, 501-02 Japan A gas chromatographic method using an electron-capture detector was developed for the determination of acrolein based on the bromination of 0-methyloxime. Acrolein was determined by gas chromatography with a 3% silicone Ge XE-60 packed column and the calibration graph showed good linearity in the range 0-0.06 yg ml-1 of acrolein in aqueous solution. The detection limit was 0.4 ng ml-1 of acrolein (signal t o noise ratio = 2) and the relative standard deviation from five determinations of 0.04 yg ml-1 of acrolein in aqueous solution was 4.5%.This method is satisfactory for the selective and reproducible determination of trace amounts of acrolein in rain water. Keywords: Acrolein determination; gas chromatography; acrolein bromination; 0 -methyloxime; rain water Aldehydes are present in vehicle exhaust gases and are formed by photochemical reactions with hydrocarbons in air. As aldehydes are therefore related to photochemical oxidant concentrations, it is very important to be able to determine them in air, as an indication of air pollution levels. Acrolein is a particularly important aldehyde and spectrophotometric and fluorimetric methods'-5 are generally used for its deter- mination.Several methods using gas and liquid chromato- graphy also have been reported for the determination of aldehydes, including acrolein,6-s but these methods are usually limited in either sensitivity or selectivity. Oxime derivatives used in the gas chromatographic deter- mination of carbonyls are rnethoximes,9JO benzyloximes,llJ2 p-nitrobenzyloximesll and pentafluorobenzyloximes. 13-15 For these methods, except for pentafluorobenzyloximes, flame- ionisation or nitrogen-specific detector systems are used. Although a gas chromatographic method based on the bro- mination of acrolein has been reported in earlier papers,l6,17 the brominated product of acrolein is unstable and the reproducibility of the determination is poor. This paper reports a method for the sensitive and selective determination of acrolein involving the bromination of acrolein O-methylox- ime and gas chromatography with electron-capture detection. Experimental Reagents and Materials Methoxylamine hydrochloride (MOA.HC1) (Wako Pure Chemical Industries, Osaka, Japan) was dried under reduced pressure. Other reagents used were of analytical-reagent grade.A standard solution of acrolein was prepared by dissolving 100 mg of the purest grade of acrolein available in distilled water and diluting to 100 ml. The brominated derivative of acrolein 0-methyloxime was supplied by Tokyo Kasei Kogyo (Tokyo, Japan), and the Sep-PAK CI8 (SP) cartridge was from Waters Associates (Milford, MA, USA). Apparatus and Conditions A Hitachi 073 gas chromatograph (GC) with a 63Ni electron- capture detector (ECD) and a Hitachi 663 GC with a flame- ionisation detector (FID) and a flame thermionic detector (FTD) were used.The following conditions were used for the GC with ECD: a 2 m glass column packed with 3% silicone GE XE-60 on 60-80 mesh Chromosorb W AW DMCS; column temperature, 90 "C; injection and detector tempera- ture, 170 "C; and carrier gas (nitrogen) flow-rate, 40 ml min-1. The GC with FID conditions were as follows: 1, a 2 m glass column packed with 20% TCP on 60-80 mesh Chromosorb W AW DMCS; a column temperature of 80 "C; an injection and detector temperature of 120 "C; and a carrier gas (nitrogen) flow-rate of 40 ml min-1; 2, a 2 m glass column packed with 10% DEGS on 60-80 mesh Chromosorb W AW DMCS; a column temperature of 140 "C; an injection and detector temperature of 170 "C; and a carrier gas (nitrogen) flow-rate of 40 ml min-1.The GC with FTD conditions were the same as the GC-FID conditions except for the carrier gas, which was helium, also with a flow-rate of 40 ml min-1. A Hitachi M52S GC - MS was used with a 10-20 eV ionisation energy. General Procedure A 1 ml aliquot of 2 M sodium acetate and 1 ml of MOA.HC1 (5 mg ml-1) are added to 5 ml of sample solution containing about 0.3 pg of acrolein. The mixture is allowed to stand for 10 min at room temperature and then 1 ml of 1.5 M sulphuric acid, 0.2 ml of 0.2 M potassium bromate and 2 g of potassium bromide are added and dissolved with stirring. After standing for 15 min at room temperature, the excess of bromine is reduced with 0.05 M sodium thiosulphate.The solution is forced through an SP cartridge and the derivative in the cartridge is eluted with 1.5 ml of diethyl ether. A 4 pl portion of the eluate is measured by GC with ECD and the peak-height method is used for the determination of acrolein. Results and Discussion Derivatisation Reaction Aldehydes are known to react with MOA to form O-methyl- oxime.10 In this work, we investigated a method based on the bromination of acrolein 0-methyloxime, followed by GC with ECD to determine acrolein with a high sensitivity and selectivity. It is assumed that the reaction proceeded as follows: CH,--CH-CHO + CH3-O-NH2 ---* CHFCH-CH=N-O-CH~ + H20 CH&H-CH=N-O-CH3 + Br2 + CH2-CH-CH=N-O-CH3 I 1 Br Br46 ANALYST, JANUARY 1987, VOL.112 !L 1 0 5 1 0 0 5 1 0 tR/min Fig. 1. Chromatograms of acrolein and its derivatives by GC-FID. 20% TCP: column temperature, 80 "C; carrier gas (N2), 40 ml min-1. 10% DEGS: column temperature, 140 "C; carrier gas (NJ, 40 ml min-1. (a Pre-reaction, A = acrolein; (b) 0-methyloxime derivative, A = acetic acid, B = brominated derivative of acrolein - MO A = acro r' ein - MO; and (c) brominated derivative of 0-methyloxime, r L DEGS L I I I I 0 5 0 5 rR/min Fig. 2. Chromatograms of acrolein and its derivatives by GC-FTD. 20% TCP: column temperature, 80 "C; carrier gas (He), 40 ml min-l. 10% DEGS: column tem erature, 140 "C; carrier gas (He), 40 ml min-l. (a) Pre-reaction;hby 0-methyloxime derivative, A = acrolein - MO; and (c) brominate derivative of 0-methyloxime, A = bromi- nated derivative of acrolein - MO In order to identify the reaction procedure described above, the solutions extracted with diethyl ether ( i e ., the pre- reaction solution, the solution containing 0-methyloxime and that containing the brominated derivative of 0-methyloxime) were measured by GC-FID and GC-FTD (Figs. 1 and 2). (b) 164 50 100 150 m/z 177 245 200 250 Fig. 3. Mass spectrum of acrolein - MO.Br derivative. (a) Bromi- nated derivative of acrolein 0-methyloxime and ( b ) synthesised 2,3-dibromopropionaldehyde 0-methyloxime I 1 I I 0 2 4 6 Amount of MOA.HCl/mg Fig. 4. Effect of amount of MOA on the formation of acrolein - MO. Acrolein, 1.0 pg; sodium acetate (2 M), 1.0 ml 0.5 1.0 1.5 2.0 Amount of 1.5 M H2SOdjml Fig. 5. Effect of volume of 1.5 M H2S04 on bromination of acrolein - MO. Acrolein, 1.0 pg; MOA.HC1, 5 mg On the chromatograms obtained by GC-FID, the acrolein peak [peak A in Fig.1 (a)] disappeared and two new peaks [Fig. 1 (b)] appeared with the formation of 0-methyloxime. Furthermore, in Fig. 1 (c), two of the peaks decreased and a new peak (peak B) appeared. This new peak is assumed to be caused by the bromination of O-methyloxime. Peak A in Fig. 1 (c) was identified as acetic acid by GC - MS measurement. On the chromatograms obtained by GC-FTD (Fig. 2), the acrolein peak did not appear, but two large peaks appeared as the O-methyloxime derivative was formed. Then, as can be seen in Fig. 2 (c), these peaks disappeared and a new large peak appeared as bromination took place.From these results, it is assumed that the nitrogenous compound was formed and converted to different nitrogenous compounds by bromination. The two peaks of 0-methyloxime may be due to syn- and anti-isomers. The peaks resulting from the brominated derivative of 0-methyloxime could not be separated under the proposed conditions. The extent of the reaction, determined by GC-FID from those compounds remaining after each reaction period, was about 92% for 1 mg of acrolein in 5 ml of aqueous solution. The mass of the brominated derivative of acrolein O-methyl- oxirne is shown in Fig. 3. The highest peak (mlz 245) of theANALYST, JANUARY 1987, VOL. 112 47 molecular ion peaks (mlz 243,245 and 247) appeared with an ionisation energy of 10 eV. The fragment peaks are assumed to be (M - Br) at mlr 164 and 166, and (M - 2Br) at mlz 85. The mass spectrum agreed with that of the standard 2,3-dibromo- propionaldehyde 0-methyloxime obtained from Tokyo Kasei Kogyo, Tokyo, Japan.As a result, the brominated derivative of acrolein O-methyl- oxime is assumed to be 2,3-dibromopropionaldehyde 0-methyloxime. 0-Me thy loxime For ma tion Levine et al. have reported the formation of 0-methyloxime in methanol solution.10 In this work, the formation of acrolein 0-methyloxime in aqueous solution was studied in order to be able to dissolve potassium bromide adequately in the second step. Fig. 4 shows the effect of MOA.HC1 in sodium acetate buffer solution. The yield of the product was maximum and constant in the range 4-6 mg of MOA.HC1. The temperature of the reaction had no influence, even at room temperature (15-20 "C), 40 and 60 "C.The minimum time for the completion of the reaction was found to be 10 min. Bromination of 0-Methyloxime The effect of acidity on the bromination of acrolein O-methyl- oxime is shown in Fig. 5. The yield was maximum in the range 0.8-1.0 ml of 1.5 M sulphuric acid, corresponding to pH 1.3-1.7. The peak height was constant in the range 2 4 g of KBr, and the yield was constant in the range 0.1-2.0 ml of 0.2 M potassium bromate. The minimum reaction time for complete bromination was found to be 10 min after the addition of the reagents. The derivative was stable for at least 5 d. Use of Sep-PAK CI8 Cartridge The brominated derivative of O-methyloxime was completely extracted with one 5 ml aliquot of diethyl ether.An SP 0 10 tRlmin Fig. 6. Typical chromatogram of rain water by GC-ECD. A = Acrolein derivative Table 1. Retention time of derivatives Retention time/ Relative Parent compound min retention time Acrolein . . . . 7.39 1 .oo Methacrolein . , 5.39* 0.73 6.951. 0.94 Crotonaldehyde . . 9.07 1.23 * Peak of brominated derivative of methacrolein. t Peak of brominated derivative of methacrolein 0-methyloxime. cartridge was used to enrich and clean up the acrolein derivative, and the following procedure was established. Pump the sample solution after reaction through the cartridge with a 10 ml syringe. Remove the cartridge from one syringe to another and pump 1.5 ml of diethyl ether through to elute the acrolein derivative. The recovery of the derivative obtained from the cartridge using 0.2 pg of acrolein was 98%.Selection of GC Column In order to separate the acrolein peak from those compounds with ethylenic bonds, such as methacrolein and crotonal- dehyde, five different columns were tested, namely, 1% PEG-HT, 10% TCEP, 10% DEGS, 3% silicone GE XE-60 and 5% silicone OV-225. As a result, it was found that the 3% silicone GE XE-60 column was the most efficient and effective for the separation of the peaks. The retention times are shown in Table 1. It was found by GC - MS that two peaks (5.39 and 6.95 min) of methacrolein were of the brominated derivative of methacrolein and the brominated derivative of meth- acrolein 0-methyloxime, respectively. This result shows that the 0-methyloxime formation from methacrolein had not proceeded to completion.Calibration Graph and Precision A calibration graph was prepared with an acrolein standard solution under the optimum experimental conditions. The relationship between peak height and the concentration of acrolein in aqueous solution was linear over the range 0-0.06 yg ml-1. The relative standard deviation from five replicates was estimated to be 4.5% for 0.04 yg ml-1 of acrolein. The detection limit (signal to noise ratio = 2) was 0.4 ng ml-1 of acrolein. Recovery of Acrolein from Rain Water In order to assess the proposed method, recovery experiments were carried out on mixtures of rain water and acrolein standard solutions. The results obtained are shown in Table 2. The proposed method can be satisfactorily applied to the determination of acrolein in rain water, as the recovery obtained was in the range 90-101%.Table 2. Recovery of acrolein from rain water Acrolein Rain water sample A . . . . . . . . B . . . . . . . . * ND = Not detected. Added PH Pg - 4.5 0.050 0.200 0.050 0.200 - 4.8 Found Recovery, Pg Yo ND* - 0.045 90 0.202 101 ND - 0.047 94 0.192 96 Table 3. Results obtained for the determination of acrolein in rain water Acrolein*/ng ml-1 Range Rain water sample Mean A-1 . . . . . . 1.8 1.5-2.0 A-2 . . . . . . NDT A-3 . . . . . . 2.8 2.5-3.1 A-4 . . . . . . 2.2 2.1-2.2 B-1 . . . . . . ND * Mean of three determinations. t ND = Not detected.48 ANALYST, JANUARY 1987, VOL. 112 Determination of Acrolein in Rain Water Acrolein was determined in rain water samples by the proposed method. Table 3 shows the results obtained, and Fig.6 shows a typical chromatogram of acrolein in rain water. Conclusion An improved method for the determination of acrolein in aqueous solutions, based on the bromination of 0-methylox- ime followed by GC with ECD, has been established. This method is highly sensitive and selective for the determination of acrolein in rain water compared with previous methods. We thank Dr. H. Tsukube, Okayama University, for his instructive advice and are grateful to Dr. K. Hojo, Tokyo Kasei Kogyo, for supply of standard materials. 1. 2. 3. 4. References Cohen, I. R., and Altshuller, A. P., Anal. Chem., 1961, 33, 726. Japanese Industrial Standard, JIS K 0089, 1983. Alarcon, R. A., Anal. Chem., 1968, 40, 1704. Suzuki, Y., Imai, S . , and Hamaguchi, A., Bunseki Kagaku, 1979, 28,445. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. Suzuki, Y., and Imai, S . , Anal. Chirn. Acta, 1982, 136, 155. Saito, T., Takashina, T., Yanagisawa, S . , and Shirai, T., Bunseki Kagaku, 1983, 32, 33. Swarin, S. J., and Kipari, F., J. Liq. Chromatogr., 1983,6,425. Kuwata, K., Uebori, M., Yamasaki, H., Kuge, Y., and Kiso, Y., Anal. Chem., 1983, 55, 2013. Fales, H. M., and Luukkainen, T., Anal. Chem., 1965,37,955. Levine, S . P., Harvey, T. M., Waeghe, T. J., and Shapiro, R. H., Anal. Chem., 1981, 53, 805. Magin, D. F., J. Chrornatogr., 1979, 178, 219. Magin, D. F., J. Chromatogr., 1980, 202, 255. Nambara, T., Kigasawa, K., Iwata, T., and Ibuki, M., J. Chromatogr., 1975, 114, 81. Kobayashi, K., Tanaka, M., and Kawai, S . , J . Chrornatogr., 1980, 187, 413. Nishikawa, H., Takahara, Y., Mori, H., and Hayakawa, T., J. Jpn. SOC. Air Pollut., 1984, 19, 387. Nishikawa, H., Hayakawa, T., and Ikeda, S., J. Chromatogr., 1986, 351,566. Nishikawa, H. , and Hayakawa, T., Bunseki Kagaku, 1985,34, 729. Paper A61185 Received June 6th, 1986 Accepted August 5th, 1986