Anal. Chem. 2009, 81, 1646–1651
Method for Determining Double-Bond Positions in Mixtures of Linear Olefins Jeffrey C. Gee* and Daniel S. Prampin Chevron Phillips Chemical Company, 1862 Kingwood Drive, Kingwood, Texas 77339 We describe a new method for estimating the positions of double bonds in mixtures of linear olefins longer than octene. The method requires about a gram of olefin and a 5-10 min reaction catalyzed by a homogeneous metathesis catalyst. The carbon number distribution of the product mixture, as determined by gas chromatography, depends upon the distribution of double-bond isomers in the starting mixture. We describe a computer algorithm for using the final carbon number distribution to determine the locations of double bonds in the starting mixture. Examples are based on tetradecenes, hexadecenes, and octadecenes. Quantifying the levels of different positional isomers in mixtures of linear olefins poses an analytical challenge.1 Direct analysis by gas chromatography is often not entirely satisfactory, and investigators have employed a number of clever techniques to solve the problem. Most approaches appear to have involved some type of derivatization, followed by detailed gas chromatography and mass spectrometry to determine the positions of the double bonds in the original mixture. At least one direct gas chromatographic method provided good resolution of most pentadecene isomers, but this analysis required many hours to complete.2 In this paper, we propose a rather simple and relatively fast method for estimating double-bond positions in mixtures of medium to long olefins (i.e., > ∼C10). The method requires about a 5-10 min reaction, followed by a single gas chromatographic analysis lasting less than an hour. We have generally heated about 1-3 g of olefin to 50-60 °C with about 60-200 ppmw (parts per million by weight) of a homogeneous metathesis catalyst for 5-10 min. Metathesis scrambles the alkyl groups on the double bonds to give new olefins.3 The carbon number distribution of the product mixture is indicative of the locations of the double bonds in the starting olefin mixture. We developed a computer program to calculate the starting isomer distribution from the observed carbon number distribution in the metathesis product mixture. * To whom correspondence should be addressed. E-mail:
[email protected]. (1) (a) Marques, F. A.; Millar, J. G.; McElfresh, J. S. J. Chromatogr., A 2004, 1048, 59–65. (b) Fordham, P. J.; Chamot-Rooke, J.; Giudice, E.; Tortajada, J.; Morizur, J.-P. J. Mass Spectrom. 1999, 34, 1007–1017. (c) Schmitz, B.; Klein, R. A. Chem. Phys. Lipids 1986, 39, 285–311. (d) Francis, G. W.; Veland, K. J. Chromatogr. 1981, 219, 379–384. (e) Soja˘k, L.; Krupc˘ˇık, K.; Jana˘k, J. J. Chromatogr. 1980, 191, 199–206. (f) Francis, G. W.; Tande, T. J. Chromatogr. 1978, 150, 139–145. (g) Hsu, C. S.; Drinkwater, D. In Current Practice of Gas Chromatography-Mass Spectrometry; Niessen, W. M. A., Ed.; Chromatographic Science Series 86; Marcel Dekker:: New York, 2001; pp 55-94. (2) See ref 1e.
1646
Analytical Chemistry, Vol. 81, No. 4, February 15, 2009
Table 1. Best Distribution Determined for Dodecene Example 1 % 1-alk
% 2-alk
% 3-alk
% 4-alk
% 5-alk
% 6-alk
96.5 ± 0.72 1.45 ± 1.0 0.85 ± 0.53 0.40 ± 0.35 0.40 ± 0.23 0.40 ± 0.25
EXPERIMENTAL SECTION Metathesis Reactions. We heated and stirred about 1.0-2.5 g of olefin to 60 °C under nitrogen in a small sample bottle. Then we combined the sample with
