The LipidWeb blank

Introduction to Mass Spectrometry of Fatty Acids



What You Will Find in These Web Pages

This series of documents is primarily a practical guide to structure determination of natural fatty acids by mass spectrometry with electron-impact ionization, as opposed to a mechanistic review, and it must be considered a personal account. It is illustrated with many more mass spectra than would be possible in a conventional review - approximately 700 at the last count - and these do not include our Archive pages where well over 2000 mass spectra are displayed (without interpretation). To view them, visit our Archive of methyl esters --- 3‑pyridylcarbinol ('picolinyl') esters --- DMOX derivatives --- pyrrolidides --- other esters and miscellaneous lipids. I hope the old adage that "a picture is worth a thousand words" applies here.

I have tried to illustrate spectra of C18 fatty acids wherever possible to simplify comparisons, but fragmentations can usually be extrapolated easily to other chain-lengths. These pages have been divided into six main sections, starting with basic information on selection and preparation of derivatives both at carboxyl groups and double bonds, together with some useful ancillary techniques. Then there are sections dealing with mass spectra of the main derivative types, methyl esters, 3-pyridylcarbinol ('picolinyl') esters, pyrrolidides and 4,4-dimethyloxazolines (DMOX). In the final section, spectra of fatty acids in the form of esters other than methyl are described together with those of some miscellaneous lipids encountered in our research, e.g. acetals, alcohols and sterols. Some contaminants that may be encountered in fatty acid preparations are also discussed.

flaskMy approach to these web pages was made possible by the wide range of mass spectra available to me and gathered over 40 years of research into lipid chemistry. For example, we have spectra of methyl esters of more than 500 different fatty acids on file, as well as 480 3‑pyridylcarbinol ester derivatives, and 320 DMOX derivatives not to mention pyrrolidides (over 250) and many other fatty acid derivatives (and related lipids). For example, we are unique in having spectra of all the possible cis-18:1 isomers, from 2‑18:1 to 17‑18:1, as 3‑pyridylcarbinol esters, DMOX derivatives and pyrrolidides, and also for all the methylene-interrupted 18:2 isomers (2,5‑18:2 to 14,17‑18:2), most of which are illustrated here. To this can be added, branched-chain, cyclic, oxygenated, sulfur-containing, halogenated, allenic, non-methylene-interrupted dienes and polyenes, and so forth.

I must gratefully acknowledge that all of these spectra have been obtained as part of the lipid chemistry research effort at the James Hutton Institute (Dundee, Scotland), and many have not been published elsewhere. Only a few of these are available in the commercial mass spectrometry libraries (and then I have often been the source). I am also grateful to a number of scientists who have provided samples for analysis or mass spectrometry files obtained on Agilent instruments (for which I have appropriate software), and they are acknowledged in the relevant web pages as well as in our credits page. An extensive bibliography is provided in separate documents in the various sections of this website, listing more than a thousand relevant references, but I have kept citations in the text to an essential minimum, as it is not always possible to establish priority of publication of individual spectra. These pages are revised and improved as new information and spectra become available.


Why Mass Spectrometry is Invaluable

The common fatty acids of animal and plant origin have even-numbered chains of 16 to 22 carbon atoms with zero to six double bonds of the cis configuration with methylene-interrupted double bond systems predominating. Nature provides countless exceptions, however, and odd- and even-numbered fatty acids with up to nearly a hundred carbon atoms exist. In addition, double bonds can be of the trans configuration, acetylenic and allenic bonds occur, and there can be innumerable other structural features, including branch points, rings, oxygenated functions, and many more. My best estimate is that 2000 different fatty acids of natural origin must exist, as well as others produced as artefacts when fats are used in commerce and in cooking for example.

Scottish thistleIn many research fields, it is essential to have simple rapid methods for determination of fatty acid structures and for isolation of pure components of mixtures for further analysis. In particular, new methods involving gas chromatography-mass spectrometry (GC-MS), GC linked to Fourier-transform infrared spectroscopy (FTIR), and silver ion and reversed-phase high-performance liquid chromatography (HPLC) are available, amongst others. I have described methodology in published reviews cited below, some of which may be aging but are still relevant. Of these methods, GC-MS with electron-impact ionization is especially useful, and the simplest commercial bench-top instruments are well suited to the purpose. Straightforward derivatization procedures are required that utilize readily available reagents and have simple glassware requirements. A feature of particular importance with GC-MS is that it is rarely necessary to isolate components in a pure form, as may be required for other spectroscopic methods (e.g. NMR spectroscopy) or for chemical degradative procedures. The instrumentation is much less costly and simpler to use and maintain than modern LC-MS equipment, for example that with electrospray ionization, which is making such an impact on so many fields of lipid research.

Fatty acids are usually analysed by GC as methyl ester derivatives, but their mass spectra may not always contain ions indicative of key structural features; the positions of double bonds in the aliphatic chain, for example, can only rarely be determined unequivocally. That said, useful 'fingerprint' spectra are often obtainable with polyunsaturated derivatives, enabling identification by comparison with standard spectra. Indeed, there are many occasions when it is convenient to analyse fatty acid methyl esters by mass spectrometry, for example for confirmatory purposes or as a guide to what further work may be required. Molecular weight and retention times are useful analytical data from such analyses, some limited structural information may be available, and indeed definitive spectra can be obtained often from branched-chain fatty acids or from those with additional oxygenated functional groups. Methyl esters may also give better resolution on GC than other derivatives, for example of components that otherwise might overlap.

In the most useful approach to structure determination, the carboxyl group is derivatized with a reagent containing a nitrogen atom. When the molecule is ionized in the mass spectrometer, the nitrogen atom not the alkyl chain carries the charge, and double bond ionization and migration is minimized. Radical-induced cleavage occurs evenly along the chain and gives a series of relatively abundant ions of high mass from the cleavage of each C-C bond. When a double bond or other functional group is reached, diagnostic ions usually occur. The first useful nitrogen-containing derivatives, i.e. pyrrolidides, were described over forty years ago; they give useful spectra and should not be discounted (indeed I believe they have been greatly under-valued, especially for labile fatty acids, such as those with epoxide rings, or with terminal functional moieties). However, most analysts now prefer either 3‑pyridylcarbinol ester or 4,4-dimethyloxazoline (DMOX) derivatives.

Derivatives for mass spectrometry

Both 3‑pyridylcarbinol esters and DMOX derivatives have their merits in mass spectrometry terms, and neither should be neglected. Each has advantages for particular types of fatty acid, and they are best considered as providing complementary information rather than simply as alternatives. With difficult samples, I have prepared both types of derivative, and often pyrrolidides also. As methyl esters are usually available for other purposes, it is often convenient to analyse these, simply for confirmation or as a record. 3‑Pyridylcarbinol esters and pyrrolidides tend to give spectra that are easier to interpret than those of DMOX derivatives when functional groups are near the terminal end of the fatty acyl chain (see Hamilton, J.T.G. and Christie, W.W. Chem. Phys. Lipids, 105, 93-104 (2000); DOI). On the other hand, DMOX derivatives may have advantages for functional groups in positions 4 to 6.

flaskIn choosing a derivative for mass spectrometry, good chromatographic properties are also important. One advantage of DMOX derivatives is that they are only slightly less volatile than methyl esters; they can be subjected to GC analysis on polar stationary phases under similar conditions and can often give comparable resolution. Pyrrolidides also have reasonable chromatographic properties, and can give some unexpected separations. 3‑Pyridylcarbinol esters, on the other hand, require column temperatures about 50°C higher than for methyl esters, and that meant initially that they had to be separated on non-polar phases, such as DB-5TM, which gave relatively poor resolution. From time to time, we still find a DB-5TM column to be of value, for example for fatty acids of high molecular weight. With the introduction of new polar phases, which are stable to high column temperatures and have low-bleed characteristics for MS analysis, such as BPX-70TM or even some of those of the Carbowax type, such as Supelcowax 10TM, the problem of GC resolution of 3‑pyridylcarbinol esters is greatly lessened and only very-long-chain fatty acids (>C24) tend to cause problems. We have a separate web page that discusses the merits of different types of GC columns.

It is worth noting that methyl esters, 3‑pyridylcarbinol esters and pyrrolidide derivatives are prepared under relatively mild conditions, and they are stable chemically so can be stored for long periods at –20°C. DMOX derivatives may require harsh conditions for preparation, and precautions are necessary to prevent hydrolysis on storage, but at least it is a simple one-pot method that can be applied to most lipid types. However, milder methods are now available when required. Preparation of the nitrogen-containing derivatives and of methyl esters are described in detail on separate web pages. Of course, there are times when it is necessary to prepare additional derivatives of functional groups other than the carboxyl, and for example, trimethylsilyl ethers can aid the analysis of hydroxy fatty acids.

Other chromatographic and spectroscopic methods are available as an aid to characterization. In particular, we have found high-performance liquid chromatography in the reversed-phase and silver ion modes to be especially useful for simplifying mixtures prior to GC-MS (see our web page on this aspect of the topic).


Alternative Mass Spectrometric Techniques

Many alternative types of fatty acid derivatives to those discussed in these pages have been described in the scientific literature, but as we have no experience of them, we cannot discuss them here. Any new derivatives described should have a better combination of chromatographic and mass spectrometric fragmentation properties than the existing ones if they are to be taken seriously. Thus, while simple oxazoline derivatives, as opposed to the dimethyloxazolines, look excellent on paper, there is no body of published spectra for comparison purposes (see Kuklev, D.V. and Smith, W.L. J. Lipid Res., 44, 1060-1066 (2003); DOI).

In the last few years, some very interesting papers have appeared dealing with acetonitrile-chemical-ionization tandem mass spectrometry in the gas phase for locating double bonds in fatty acid methyl esters mainly from the laboratory of Professor J. Thomas Brenna, and this technique has the virtue of being able to distinguish cis/trans isomers. A bibliography of relevant published papers is available on this website. Mass spectrometry with electrospray ionization is being used increasingly for structural analysis of fatty acids, but I have no personal experience of this technique and cannot comment further at present. Some interesting results have been obtained by direct infusion or ‘shotgun’ methods, but they seem only suitable for rapid screening in my estimation at least. That said, derivatization of fatty acids with N-(4-aminomethylphenyl) pyridinium (AMPP) in conjunction with these techniques looks promising. Similarly, I have no experience of remote-site fragmentation methods involving tandem mass spectrometry or collisional activation of carboxylate anions or alkali metal-cationized fatty acids, although they appear to be useful for locating double bonds especially. These methods and the other mild ionization techniques require more sophisticated and expensive instrumentation than the electron-impact ionization methods described here, and they rely heavily on computerized analysis of the data.


Quantitative Analysis

I have NOT described the use of mass spectrometry for quantitative analysis of fatty acids in these web pages. Then quite different problems arise, especially a need for careful calibration with appropriate standards, which may not be available; methyl ester derivatives may be as good as any other for this purpose. I will leave that topic for someone else to discuss, but readers should be aware that gas chromatography (GC) with flame-ionization detection is by far the simplest and arguably the most accurate approach to quantitative analysis when a sample contains the mainstream range of fatty acids.


What to Read

For general information on fatty acid analysis or lipid analysis in general, I recommend that readers consult my book - Lipid Analysis (4th Edition) - listed below. The first two reviews in the list, though now somewhat dated, give more detailed information on our GC-MS methodology -

The last of these is most useful for those interested in mechanistic aspects of lipid mass spectrometry. Of course, there are many specialist books available that deal with the principles of mass spectrometry in general, but I cannot offer specific guidance. Many more review articles are listed in our Bibliography - review articles document. On the other hand, readers will find that these web pages contain much more practical information and vastly more illustrations of mass spectra than are available in any more formal publication.


Credits/disclaimer Updated: March 23rd, 2017 Author: William W. Christie LipidWeb icon

Return to the top of the page