Mass Spectra of Halogenated Fatty Acids

As with my other documents on mass spectrometry, this is a subjective account that details only those fatty acids relevant to this topic which have been encountered during our research activities here and for which we have spectra available for illustration purposes. Spectra of methyl esters, 3-pyridylcarbinol ('picolinyl') esters, DMOX derivatives and pyrrolidides are all described here (when available), but I will only describe key diagnostic ions as general features of each type of derivative are described elsewhere on this website. Only a few of the spectra have been published elsewhere, and references to prior publications are cited if known.

The fluoro fatty acids were perfectly stable and were easily derivatized. The others can lose the halogen atom readily unless care is taken during derivatization.

 

Fluoro Fatty Acids

One natural seed oil contains several fatty acids with a fluorine atom on the terminal carbon, i.e. Dichapetalum toxicarum. We reinvestigated this oil using 3-pyridylcarbinol ester and DMOX derivatives and, in addition to obtaining novel mass spectrometric information, we found several hitherto unknown minor fatty acid components (Christie et al., 1998). We also have some mass spectra from several synthetic fluoro-fatty acids kindly donated by Dr David O'Hagan of St Andrews University, and details of some of these have been published (Hamilton and Christie, 2000). They provided important evidence for some mechanistic concepts, but the aim of these pages is simply to act as a practical guide).

Methyl esters certainly give distinctive spectra, but information on a wider range of isomers would be required to give definitive interpretations. The mass spectrum of methyl 18-fluoro-octadecanoate follows -

Mass spectrum of methyl 18-fluoro-octadecanoate

In the high mass range, there is an ion at m/z = 296, which presumably represents loss of HF, and ions for loss of the methanol moiety from the carboxyl group (m/z = 285), and for the McLafferty ion (m/z = 74) can be recognized. Without further information, there is no feature that locates the fluorine atom on its specific position. Further comment would be merely speculation.

Similarly, in the mass spectrum of methyl 18-fluoro-octadec-9-enoate, there is an ion at m/z = 294, which presumably represents loss of HF, together with the expected ions for the loss of methanol ([M-31/32]+, m/z = 283, and [M-HF-32]+, m/z = 262).

Mass spectrum of methyl 18-fluoro-octadec-9-enoate

Mass spectrum of methyl 2-fluoro-octadecanoate -

Mass spectrum of methyl 2-fluoro-octadecanoate

The base peak is the McLafferty ion at m/z = 92, which confirms that the fluorine atom is on position 2. Abundant ions at m/z = 147 and 161 indicate preferential cleavages between C-6 and C-7, and C-7 and C-8, of each of the derivatives. Furthermore, losses of 47 (m/z = 269) and 61 (m/z = 255) amu from the molecular ion indicate the expulsion of fluorinated ethyl and propyl radicals, while other losses of 75, 89, 103 and 117 amu represent the expulsion of 4-, 5-, 6- and 7-carbon fragments that must contain a CHF group together with an additional hydrogen atom.

3-Pyridylcarbinol esters: As expected, these give much more informative mass spectra, locating the fluorine atom and other structural features, and that of 3-pyridylcarbinyl 18-fluoro-octadecanoate follows.

Mass spectrum of 3-pyridylcarbinyl 18-fluoro-octadecanoate

After the molecular ion, there are gaps of 20 amu to m/z = 373 (loss of HF) and one of 33 amu to 360 (loss of the terminal -CH2F). Thereafter, there are regular gaps of 14 amu for loss of successive methylene groups.

Mass spectrum of 3-pyridylcarbinyl 18-fluoro-oleate -

Mass spectrum of 3-pyridylcarbinyl 18-fluoro-oleate

This is very similar to the spectrum of 3-pyridylcarbinyl oleate, and the position of the double bond is determined as described in the section of this website on 3-pyridylcarbinol esters of monoenes. However, in the high mass range, there are again diagnostic gaps of 20 amu to m/z = 371 (loss of HF - just discernable) and one of 33 amu to 358 (loss of the terminal -CH2F).

Mass spectrum of 3-pyridylcarbinyl 18-fluoro-linoleate -

Mass spectrum of 3-pyridylcarbinyl 18-fluoro-linoleate

As anticipated, this is similar to the spectrum of 3-pyridylcarbinyl linoleate, and the position of the double bonds are determined as described in the section of this website on 3-pyridylcarbinol esters of dienes. However, in the high mass range, there is again a diagnostic gap of 20 amu to m/z = 369 (loss of HF). The expected gap of 33 amu to 356 (loss of the terminal -CH2F is not clearly seen, but successive gaps of 14 amu are apparent so the fluorine atom must be on the terminal carbon.

The mass spectrum of synthetic 3-pyridylcarbinyl 2-fluoro-octadecanoate -

Mass spectrum of 3-pyridylcarbinyl 2-fluoro-octadecanoate

The McLafferty ion is shifted from m/z = 151 to 169 as expected and indeed all ions from here upwards are increased by 18 amu in comparison to the those in the spectrum of 3-pyridylcarbinyl stearate.

DMOX derivatives of fatty acids with terminal fluorine atoms are much less useful (this might not be true of more centrally-placed fluorine atoms). As an example, the mass spectrum of the DMOX derivative of 18-fluoro-octadecanoate follows -

Mass spectrum of the DMOX derivative of 18-fluoro-octadecanoate

From the molecular ion, there is a gap of 15 amu to m/z = 340, representing loss of a methyl group, not of -CH2F. We now know that the explanation is a facile loss of a methyl group from the ring structure that confounds the picture (seeHamilton and Christie, 2000).

The mass spectrum of the DMOX derivative of 18-fluoro-oleate -

Mass spectrum of the DMOX derivative of 18-fluoro-oleate

The pattern is repeated; a loss of 15 amu from the molecular ion to m/z = 338 confuses the picture at the terminal carbon. The position of the double bond is determined as described in the section of this website on DMOX derivatives of monoenes.

Mass spectrum of the DMOX derivative of synthetic 2-fluoro-octadecanoate -

Mass spectrum of the DMOX derivative of 2-fluoro-octadecanoate

The McLafferty ion at m/z = 131 has been shifted upwards by 18 amu (from 113) as expected, and most other ions are 18 amu higher in comparison to the non-fluorinated analogue.

Pyrrolidine derivatives of fatty acids with terminal fluorine atoms are best compared with DMOX derivatives, and the mass spectrum of the pyrrolidine derivative of 18-fluoro-linoleate is illustrated next.

Mass spectrum of the pyrrolidine derivative of 18-fluoro-linoleate

Although the ions in the high mass range are less abundant than in the spectrum of the DMOX derivative (at least relative to the base peak), the diagnostic ions for the double bonds are clearly delineated (in the same place as with DMOX derivatives). Following the molecular ion, there is a gap of 20 amu to m/z = 331 (loss of HF) and of 33 amu to m/z = 318 (very small) for loss of -CH2F. There is no confusing ion at [M-15]+ as with the DMOX derivative, because there are no methyl groups on the ring.

 

Chloro Fatty Acids

We only have spectra available of 3-pyridylcarbinol esters of a few omega-chloro fatty acids, produced by microbial fermentation of chlorohydrocarbons. Other chlorinated fatty acids have been found in marine samples, but we do not have access to these. Our single attempt to prepare DMOX derivatives by the usual method of heating the free acid with 2-amino-2-methyl-propan-1-ol at 190°C caused elimination of the chlorine atom. The 3-pyridylcarbinol esters were prepared by hydrolyzing to the free acid, and proceeding via the imidazolide route (see the section on Preparation of derivatives for mass spectrometry).

Mass spectrum of 3-pyridylcarbinyl 16-chloro-hexadecanoate -

Mass spectrum of 3-pyridylcarbinyl 16-chloro-hexadecanoate

Following the molecular ion at m/z = 381, there is a gap of 35 amu to m/z = 346 for loss of the terminal chlorine atom. Thereafter, there is a regular series of ions 14 amu apart for cleavage at successive methylene groups.

Mass spectrum of 3-pyridylcarbinyl 16-chloro-hexadec-9-enoate -

Mass spectrum of 3-pyridylcarbinyl 16-chloro-hexadec-9-enoate

Again, there is a gap of 35 amu to m/z = 344 for loss of the terminal chlorine atom. Thereafter, there is at first a regular series of ions 14 amu apart for cleavage at successive methylene groups. The position of the double bond is determined as described in the section of this website on 3-pyridylcarbinol esters of monoenes.

 

Bromo Fatty Acids

Our first attempts to prepare derivatives of bromo fatty acids (produced by microbial fermentation of bromohydrocarbons) failed when our transesterification and hydrolysis procedures caused elimination of bromine with formation of ethoxyl or methoxy fatty acids (see the section of this website on Mass spectra of epoxy, furanoid and alkoxy acids). However, we obtained some synthetic 2-bromo-octadecanoic acid and successfully derivatized this as described below. Other brominated fatty acids have been found in marine samples, but we do not have access to these.

Methyl 2-bromo-octadecanoate was prepared by reacting the free acid with trimethylsilyldiazomethane. Its mass spectrum is -

Mass spectrum of methyl 2-bromo-octadecanoate

The spectrum is complex, but the expected double molecular ion at m/z = 348 and 350 can be seen (bromine exists as two isotopes in approximately equal amounts of atomic mass 79 and 81). However, the key diagnostic feature is the fact that the McLafferty ion forms a doublet of ions at m/z = 152 and 154, instead of at 74. The ion at m/z = 269 is for the loss of the bromine atom.

3-Pyridylcarbinyl 2-bromo-octadecanoate was prepared from the free acid via the imidazolide route. Its mass spectrum, which follows, is surprisingly unhelpful.

Mass spectrum of 3-pyridylcarbinyl 2-bromo-octadecanoate

The small molecular ion is a doublet (m/z = 425/7), but the base peak at m/z = 346 reflects loss of a bromine atom. The doublet of ions at m/z = 242/244 presumably results from cleavage between carbons 3 and 4, and there are a few small ions in the higher mass region that are doublets so presumably contain bromine.

DMOX derivative of 2-bromo-octadecanoate. This was prepared successfully by our relatively mild two-step method (see the section of this website on Preparation of derivatives for mass spectrometry). Its spectrum is -

The molecular ion is a small doublet at m/z = 387/9 and there is also a small doublet at m/z = 372/4 for loss of a methyl group (probably from the ring system). Again the key diagnostic feature that locates the bromine on carbon 2 is the fact that the McLafferty ion forms a doublet at m/z = 191/3. The base peak at m/z = 308 reflects loss of bromine from the molecule.

Spectra of further halogenated fatty acids are available, but without interpretation, in the Archive Sections of these web pages, i.e. for methyl esters -- 3-pyridylcarbinol ('picolinyl') esters -- DMOX derivatives -- pyrrolidides.

 

References

  • Christie, W.W., Hamilton, J.T.G. and Harper, D.B.Mass spectrometry of fluorinated fatty acids in the seed oil of Dichapetalum toxicarium. Chem. Phys. Lipids, 97, 41-47 (1998) (DOI: 10.1016/S0009-3084(98)00090-5).
  • Hamilton, J.T.G. and Christie, W.W.Mechanisms for ion formation during the electron impact-mass spectrometry of picolinyl ester and 4,4-dimethyloxazoline derivatives of fatty acids. Chem. Phys. Lipids, 105, 93-104 (2000) (DOI: 10.1016/S0009-3084(99)00133-4).

Updated April 5, 2012

pdf