The LipidWeb blank

Mass Spectrometry of Methyl Esters

Saturated Branched-Chain Fatty Acids

As cautioned in the 'Introduction' to these documents, the mass spectra of methyl esters sometimes afford limited information only concerning the structures of fatty acids. Molecular weights are usually obtainable and are important for identification, as are gas chromatographic retention data. However, many saturated methyl-branched fatty acids can be identified from the mass spectra of their methyl ester derivatives, especially when spectra of model compounds are available for comparison purposes. Only those spectra encountered in my own research can be described in detail here, leaving some unfortunate gaps in my account, but the references listed provide further information. There is a document on the chemistry and occurrence of branched-chain fatty acids on this site here... Note that conflicting nomenclatures can be used in the literature, depending on whether the total number of carbons in the fatty acid or that in the aliphatic chain is listed.

Although they are arguably not ideal derivatives for the purpose, many natural branched-chain fatty acids have been successfully characterized by means of mass spectrometry of their methyl esters. 3-Pyridylcarbinol (‘picolinyl’) esters and pyrrolidides are the best derivatives in this instance (see the appropriate sections of these web pages), but dimethyloxazoline derivatives are less suitable, especially for the iso- and anteiso-methyl forms, which are most often encountered in nature. The definitive paper on the mass spectra of methyl esters of mono-methyl-branched fatty acids has data for the complete series of (2- to 170-methyl-octadecanoates (Apon and Nicolaides,1975), although much of the data for the more important acids from a biological standpoint were published earlier (Ryhage and Stenhagen, 1959). Some representative spectra are illustrated below with brief details of interpretation only.

Iso- and anteiso-Methyl-Branched Fatty Acids

The two most common type of branched-chain fatty acids are those with methyl branches in the iso or anteiso positions, and examples of the mass spectra are shown below. Unfortunately, the former has the least distinctive spectra of all the isomers. Note that mass spectra of branched isomers often appear busier or noisier than those of the analogous straight-chain esters.

The mass spectrum of methyl iso-methylhexadecanoate (or 15-methylhexadecanoate) -

Mass spectrum of methyl iso-methylhexadecanoate

In this instance, the features that may distinguish the spectrum from that of the straight-chain analogue and other branched isomers are not always easy to see. Though small, the [M-15]+ ion (m/z = 269) may be larger in the spectrum of an iso-isomer than in that of the straight-chain analogue, and there is usually a doublet of ions for [M‑31/2]+ (see the papers cited above). The other ions of diagnostic value are rather small (<1% of the base peak), and equivalent to [M‑65]+, [M‑55]+ and [M‑56]+, and here at m/z = 219, 229 and 228, respectively, but these may not apply uniformly across the chain-length range. As with normal saturated fatty acids, the McLafferty rearrangement ion at m/z = 74 is the base peak.

As a further example, the mass spectrum of methyl anteiso-methyl-hexadecanoate (or 14-methyl-hexadecanoate) is shown next -

Mass spectrum of methyl anteiso-methyl-hexadecanoate

The molecular ion is clearly seen and the important distinguishing feature from the spectrum of the straight-chain analogue and that of the iso-isomer is that an ion at [M-29]+ (m/z = 255) is more abundant than that equivalent to [M‑31]+. Also, an ion at [M‑61]+ (m/z = 223) is small but distinctive (sometimes with ions at [M‑60]+ and [M‑79]+) (again see the papers cited above). We have found these ions in the C14, C16 and C18 analogues, but they are less evident with the C20 and C24 analogues (number of carbons in the straight chain).

Certain other ions that are found in the spectra of all methyl-branched fatty acids are mentioned briefly in the next section. It can be helpful to be aware that it is more common to find iso-methyl fatty acids with an even number of carbons in total in natural samples, although the chain-length is odd-numbered, while the opposite is true for anteiso-isomers. It is also useful to know that these two isomers elute after the corresponding fatty acids with the same number of carbons in the linear chain (one fewer in total) from all GC phases with the iso-isomer eluting before the anteiso-isomer. Thus, as an example, the order of elution is pentadecanoate (15:0) < iso-methyl-pentadecanoate < anteiso-methyl-pentadecanoate < hexadecanoate (16:0). If in doubt, hydrogenation is a useful supplementary technique to remove unsaturated components that might confuse identification.

Fatty Acids with Centrally-Located Methyl-Branches

Fatty acids with methyl branches in more central positions are found in bacterial lipids and can also be produced by animal tissues, e.g. in ruminants or in human sebaceous secretions and vernix caseosa (the greasy secretion on the newborn baby), when methylmalonate is utilized instead of malonate in fatty acid biosynthesis. The methyl esters may then give spectra with distinctive fragments, ions 'a' and 'b' (together with 'a+1' and 'a+2'), on either side of the branch as shown -

Fragmentation of a branched ester

In addition, there may be diagnostic ions for 'b' after the loss of methanol, or strictly speaking a methoxyl group plus a hydrogen atom ([b‑32]+), and then for a further loss of a water molecule ([b‑50]+). These ions are present in the two spectra above, but are very small. For example, in the spectrum of the anteiso-isomer, the 'a' ion is barely detectable, but the 'b' ion at m/z = 255, and those for 'b‑32' and 'b‑50' at m/z = 223 and 205, respectively, are clearly seen. Analogous ions are less easily distinguished in the spectrum of the iso-isomer.

As mentioned earlier, the examples illustrated below were selected from the limited range available to us from our past research efforts. Most were found in sponges, where they were probably derived from bacteria ingested as part of their diet.

Thus, in the mass spectrum of methyl 15-methyl-docosanoate, ion 'a' is at m/z = 241, and ion 'b' at m/z = 269 (b‑50, m/z = 219).

Mass spectrum of methyl 15-methyl-docosanoate

In the mass spectrum of methyl 14-methyl-eicosanoate, ion 'a' is at m/z = 227, and ion 'b' at m/z = 255 (b‑50, m/z = 205).

Mass spectrum of methyl 14-methyl-eicosanoate

In the spectrum of methyl 13-methyl-eicosanoate, ion 'a' is at m/z = 213 and ion 'b' at m/z = 241 (b‑50, m/z = 191).

Mass spectrum of methyl 13-methyl-eicosanoate

In the spectrum of methyl 11-methyl-octadecanoate, ion 'a' is at m/z = 185 and ion 'b' at m/z = 213.

Mass spectrum of methyl 11-methyl-octadecanoate

This spectrum has been selected to demonstrate the presence of both of the [b‑32]+ and [b‑50]+ ions, which are at m/z = 181 and 163, respectively, simply because they appear in a relatively noise-free part of the spectrum in this instance. I will leave readers to find the first of these ions in most of the other spectra.

In the spectrum of methyl 10-methyl-octadecanoate, ion 'a' is at m/z = 171 and ion 'b' at m/z = 199 (b-50 is at m/z = 149) -

Mass spectrum of methyl 10-methyl-octadecanoate

This fatty acid has the trivial name 'tuberculostearic acid' as it was first detected in the pathogen Mycobacterium tuberculosis. In this instance, the spectrum was obtained from the lipids of the bacterium Rhodococcus ruber courtesy of Gary Dobson. The spectrum of methyl 10-methyl-hexadecanoate has the same key diagnostic ions.

In the spectrum of methyl 9-methyl-tetradecanoate, ion 'a' is at m/z = 157 and ion 'b' at m/z = 185. The b‑32 and b‑50 ions are clearly seen.

Mass spectrum of methyl 9-methyl-tetradecanoate

By extrapolation, for an 8-methyl-branch, we would expect the 'a' and 'b' ions to be at m/z = 143 and 171, respectively, and we can predict the diagnostic ions for other isomers in the same way. However, it is evident that many of the same ions occur in these various mass spectra, differing only in relative abundance. It is important to note that it might occasionally be difficult to identify an unknown unequivocally from such data without access to spectra of model compounds.

The spectrum of methyl 3-methylpentadecanoate is rather different. Ion 'b' at m/z = 101 (replacing that normally found at m/z = 87) is especially distinctive, together with 'b‑32' at m/z = 69, while ion 'a' is the McLafferty ion at m/z = 74. In the spectrum of a 2-methyl equivalent, the McLafferty ion is at m/z = 88 (see the spectrum of methyl pristanate in the next section).

Mass spectrum of methyl 3-methylpentadecanoate

In most of the spectra above, ions equivalent to [M-43]+ and [M-29]+, which are presumed to be due to the loss of fragments consisting of carbons 2 to 4 and 2 to 3, respectively, are prominent. In the spectrum of the 3-isomer, these characteristic ions are shifted to at [M‑57]+ and [M‑53]+, presumably because these fragments now incorporate the methyl branch (see the web pages on methyl esters of normal saturated fatty acids for a discussion of the mechanism).

Multi-Methyl Branched Fatty Acids

In the mass spectra of methyl esters of multi-methyl-branched fatty acids, ions reflecting the first branch point are much the most abundant. I have encountered one non-isoprenoid dimethyl-branched fatty acid in a sponge, i.e. 8,10-dimethyl-hexadecanoic acid and its methyl ester has the spectrum -

Mass spectrum of methyl 8,10-dimethylhexadecanoate

The 'b' ion for the methyl branch on C10 at m/z = 213, together with the ions representing [b‑32]+ and [b‑50]+ at m/z = 181 and 163, respectively, are clearly seen, but no 'a' ion is apparent. There is a complete complement of the expected ions for the first methyl group (on C8) as indicated on the spectrum, while the 'a*' ion at m/z = 143 is especially distinctive.

Isoprenoid fatty acids, derived from metabolism of the phytol of chlorophyll, are significant components of marine oils and of ruminant fats. Phytanic acid can accumulate in plasma of patients suffering from Refsum's syndrome in which there is a defect in one or other steps in the α-oxidation system. A number of different isomers have been found in nature, and the mass spectra of the methyl esters of the three most common of these are illustrated below.

The mass spectrum of methyl 4,8,12-tridecanoate (from a fish oil) -

Mass spectrum of methyl 4,8,12-tridecanoate

The McLafferty ion at m/z = 74 is no longer the base peak in the spectrum, because of the presence of the methyl branch on carbon 4; there is thus only one hydrogen atom available for abstraction (see the web pages on mass spectra of normal saturated fatty acids for a discussion of the mechanism). The 'a' ion for the first methyl group at m/z = 87 is now the most abundant ion. I will leave the reader to sort out most of the remaining 'a' and 'b' ions for this and the next two spectra - I am content to consider them as "fingerprints". A review by Lough (1975) discusses these spectra in some detail.

The mass spectrum of methyl pristanate or 2,6,10,14-tetramethylpentadecanoate is -

Mass spectrum of methyl pristanate or 2,6,10,14-tetramethylpentadecanoate

The mass spectrum of methyl phytanate or 3,7,11,15-tetramethylhexadecanoate is -

Mass spectrum of methyl phytanate or 3,7,11,15-tetramethylhexadecanoate

Note that the McLafferty ion at m/z = 88 is the base peak in the spectrum of methyl pristanate as it now contains the methyl group on carbon 2, while the ion at m/z = 101 for a fragment containing the methyl branch in position 3 dominates the spectrum of methyl phytanate.

Spectra of many more methyl esters of branched-chain fatty acids can be accessed from our Archive pages (without interpretation). Spectra of monoenoic fatty acids with methyl branches are discussed in the web page dealing with monoenes.

Note. Those with access to instruments with a facility for collisional dissociation of molecular ions generated by electron ionization (tandem mass spectrometry) will be able to obtain additional data that should remove any dubiety regarding interpretation (Ran-Ressler et al., 2012).


I also recommend - Christie, W.W. and Han, X. Lipid Analysis - Isolation, Separation, Identification and Lipidomic Analysis (4th edition), 446 pages (Oily Press, Woodhead Publishing and now Elsevier) (2010) - from Science Direct.

Credits/disclaimer Updated: April 16th, 2018 Author: William W. Christie LipidWeb icon