Why I dislike boron trifluoride-methanol
Abstract: Boron trifluoride-methanol is very popular as a methylating reagent, but it has many disadvantages. Newcomers to lipid analysis should critically assess its use, and consider alternatives. Established analysts should be aware of the reasons for inconsistent results obtained with this reagent.
It has become apparent to me that many of my readers swear by boron trifluoride-methanol as a methylating reagent and do not understand why I take every opportunity to run it down. In preparing a substantial review of esterification methods  (now available online at this website), I had to re-evaluate the evidence - but did not change my conclusions.
The reagent has only one advantage - it can be bought off the shelf from a number of suppliers.
The remainder of this article will deal with the disadvantages.
Origins of the Methodology
The first important paper on the use of the boron trifluoride-methanol was by Metcalfe and Schmitz , but a later one published by Morrison and Smith in 1964 has had much more influence . The latter is a classic study for a number of reasons, and at one time was high on the list of the most highly cited papers of all time on biochemical topics. To my knowledge, it was the first truly comprehensive study of an esterification method. For example, they examined the rates of methylation of a wide variety of lipids, including triacylglycerols, cholesterol esters, phosphoglycerides and sphingolipids. In addition, the reaction with plasmalogens was studied, as was the effect of addition of an inert solvent to solubilize non-polar lipids, and a side reaction in which cholesterol was converted to cholestadiene was identified. This last effect is common to all acidic transesterification reagents, and has been "rediscovered" in at least six later publications .
I earnestly recommend young scientists actually to read Morrison and Smith's paper - not simply to take as gospel what scientist C took from scientist B who got his information from scientist A, who may have skimmed through the paper 20 years ago. It is available as a free download from the Journal of Lipid Research. In another of my articles on this website, I made a similar point with regard to the classic papers on extraction methodology. Not all the good science has been done in the last few years. But now I am onto another hobbyhorse.
No criticism of these first papers is intended here, and there are good reasons why no problems were encountered with the reagent at first. Lough was the first to raise doubts, when he noted that high concentrations of boron trifluoride in methanol (50% as opposed to the more usual 14-15%) catalysed addition of methanol across the double bonds of unsaturated fatty acids; loss of polyunsaturated components of mixtures was observed and new artefactual peaks appeared late in GC chromatograms as methanol was added across double bonds . Although he subsequently withdrew the criticism of the reagent, admitting that he had used excessive amounts of the catalyst, many others were soon to report similar problems. I was a post-doctoral fellow at the Hormel Institute in Minnesota at this time, and I recall that many of us had the same difficulties.
The Nature of the Problems
The reason for inconsistent results, some laboratories having massive problems while others had none, was difficult to track down. Indeed, definitive experiments to determine the mechanism do not appear to have been carried out even yet. It is, however, evident that the problems develop as the reagent ages, probably through formation of further fluoroboron compounds by reaction with atmospheric oxygen and methanol. Boron trifluoride-methanol should always be stored at refrigerator temperature, and the excess should be discarded after a few months (remember to write the date of purchase on the bottle!). Certainly, I know of regular users who get through a bottle in a week or two who do not seem to have problems. While there should not be any difficulties with fresh reagent, purchasers have no information on how long a bottle has been standing on the shelves of the supplier's warehouse.
Apart from the reaction with polyunsaturated fatty acids, it is well established that boron trifluoride-methanol will cleave the rings in cyclopropane fatty acids (commonly encountered in microorganisms), and it causes cis-trans isomerization of double bonds in conjugated fatty acids. In addition, it reacts with the antioxidant BHT to produce spurious peaks in chromatograms (the latter reaction has been "rediscovered" only three times ). In the period since I wrote my review, at least three new papers on formation of artefacts of fatty acids with the reagent have appeared [5-7].
My other major argument against the use of boron trifluoride-methanol is that it is simply no better than other acidic reagents. A primary reason for using it is that it is often claimed to be much more rapid than the alternatives. It may be a little faster than other acidic reagents, but only because it contains 15% catalyst instead of 1 to 5% in the alternatives. I prefer to use methanolic sulphuric acid (1%) or hydrogen chloride (5%), prepared by adding sulphuric acid or acetyl chloride, respectively, slowly to methanol. These procedures are so straightforward that there is no need to prepare more reagent than is required for a week or two ahead. If the reaction is carried out on a small scale in a fume cupboard, there is no significant hazard. Of course, when it is appropriate to use them, base-catalysed transesterification methods are much faster and safer in every respect .
Modern chromatographic techniques are designed to be used with very small samples, certainly less than the microgram scale, and it is standard practice for research scientists in medicine and many other fields to operate with samples of this size. It is then especially important to minimize the opportunities for artefact formation. In contrast, it is not uncommon in say quality control labs to operate on a 'bucket' scale (I consider this to be from 10 mg upwards), where problems in esterification are likely to be much less evident. Standard methods, such as those approved by the AOCS, are written with the latter in mind and can be needlessly complicated .
My concern is not for the established analyst in a large laboratory, but for newcomers to the subject who scan the literature, discover that boron trifluoride is widely used for methylation of fatty acids and proceed to base their methodology upon it. We may know of the history of nearly fifty years of problems and how they can be minimized. He or she may not. Perhaps, there should be a “health” warning - “BF3 can damage your chromatograms” on every bottle?
- Christie, W.W. Preparation of ester derivatives of fatty acids for chromatographic analysis. In: Advances in Lipid Methodology - Two, pp. 69-111 (ed. W.W. Christie, The Oily Press, Dundee) (1993).
- Metcalfe, L.D. and Schmitz, A.A. The rapid preparation of fatty acid esters for gas chromatographic analysis. Anal. Chem., 33, 363-364 (1961).
- Morrison, W.R. and Smith, L.M. Preparation of fatty acid methyl esters and dimethyl acetals from lipids with boron trifluoride-methanol. J. Lipid Res., 5, 600-608 (1964).
- Lough, A.K. The production of methoxy-substituted fatty acids as artifacts during the esterification of unsaturated fatty acids with methanol containing boron trifluoride. Biochem. J., 90, 4C-5C (1964).
- Orgambide, G,G., Reusch, R.N. and Dazzo, F.B. Methoxylated fatty acids reported in Rhizobium isolates arise from chemical alterations of common fatty acids upon acid-catalyzed transesterification procedures. J. Bact., 175, 4922-4926 (1993).
- Yurawecz, M.P., Molina, A.M., Mossoba, M. and Ku, Y. Estimation of conjugated octadecatrienoates in edible oils and fats. J. Am. Oil Chem. Soc., 70, 1093-1099 (1993).
- Stavarache, C., Vinatoru, M. and Maeda, Y. Ultrasonic versus silent methylation of vegetable oils. Ultrasonics Sonochem., 13, 401-407 (2006).
- Ackman, R.G. Remarks on official methods employing boron trifluoride in the preparation of methyl esters of the fatty acids of fish oils. J. Am. Oil Chem. Soc., 75, 541-545 (1998).
This article has been updated appreciably from one by the author that first appeared in Lipid Technology, 6, 66-68 (1994) (by kind permission of P.J. Barnes & Associates (The Oily Press Ltd)), who retain the copyright to the original article.
James Hutton Institute (and Mylnefield Lipid Analysis), Invergowrie, Dundee (DD2 5DA), Scotland.
Updated: July 14th, 2011