Autoxidation of the furan fatty acid ester, methyl 9,12-epoxyoctadeca-9,11-dienoate

Article abstract of DOI:10.1007/s11746-998-0177-6

The objective of this study was to identify autoxidation products of methyl 9,12-epoxyoctadeca-9,11-dienoate (F_9,12). Previous work has shown that F_9,12 is a product both of autoxidation and singlet oxygen oxidation of the methyl ester derivative of conjugated linoleic acid (CLA). F_9,12, 95% pure, was synthesized from methyl ricinoleate. The synthetic F_9,12 was heated at 50 °C in sealed tubes containing air. Each tube contained 6 mg F_9,12 and 1 mg methyl stearate as an internal standard. Samples were taken at 4.5, 7, 23, 46.5, 69.5, and 93 h. The oxidized F_9,12 was dissolved in isooctane and analyzed by gas chromatography (GC), GC-direct deposition-Fourier transform infrared spectroscopy, and GC-electron ionization mass spectrometry. CLA methyl ester was oxidized in a similar manner. Under these conditions, the half-lives of CLA and F_9,12 were 40 and 35 h, respectively. Oxidation products of F_9,12 that were identified included: 5-hexyl-2-furaldehyde (I), methyl 8-oxooctanoate (II), methyl 13-oxo-9,12-epoxytrideca-9,11-dienoate (III), methyl 8-oxo-9,12-epoxy-9,11-octadeca-dienoate (IV), and methyl 13-oxo-9,12-epoxy-9,11-octadeca-dienoate (V).

Full text of DOI:10.1007/s11746-998-0177-6

Autoxidation of the Furan Fatty Acid Ester,
Methyl 9,12-Epoxyoctadeca-9,11-dienoate
N. Sehat, M.P. Yurawecz*, J.A.G. Roach, M.M. Mossoba, K. Eulitz,
E.P. Mazzola, and Y. Ku
U.S. Food and Drug Administration, Center for Food Safety and Applied Nutrition, Washington, DC 20204
ABSTRACT: The objective of this study was to identify autoxi-
dation products of methyl 9,12-epoxyoctadeca-9,11-dienoate
(F9,12). Previous work has shown that F9,12 is a product both of
autoxidation and singlet oxygen oxidation of the methyl ester
derivative of conjugated linoleic acid (CLA). F9,12, 95% pure,
was synthesized from methyl ricinoleate. The synthetic F9,12
was heated at 50°C in sealed tubes containing air. Each tube
contained 6 mg F9,12 and 1 mg methyl stearate as an internal
standard. Samples were taken at 4.5, 7, 23, 46.5, 69.5, and
93 h. The oxidized F9,12 was dissolved in isooctane and ana-
lyzed by gas chromatography (GC), GC-direct deposition-
Fourier transform infrared spectroscopy, and GC-electron ion-
ization mass spectrometry. CLA methyl ester was oxidized in a
similar manner. Under these conditions, the half-lives of CLA
and F9,12 were 40 and 35 h, respectively. Oxidation products of
SCHEME 1
(8), and plants (9). In these F-acids, the 3- and 4-positions of
the furanyl moiety may also contain a methyl group.
Previous studies of oxidations of the 2,5-disubstituted
F-acid moiety showed that it does not have the strong antiox-
idant properties that are associated with the tri- or tetra-alkyl
substituted F-acids (10). Oxidation with lipid hydroperoxides
(11,12), m-chloroperoxybenzoic acid (13), or ultrasound (14)
yields products with a dioxoene functional group (e.g.,
methyl 9,12-dioxooctadec-10-enoate). Compounds with the
dioxoene function are reported to be strong inhibitors of
blood platelet aggregation (15). Oxidation of F9,12 on silica is
reported to produce several products including methyl 9,12-
dioxooctadeca-7,10-dienoate and methyl 9,12-dioxooc-
tadeca-10,13-dienoate (16).
F
9,12 that were identified included: 5-hexyl-2-furaldehyde (I),
methyl 8-oxooctanoate (II), methyl 13-oxo-9,12-epoxytrideca-
9,11-dienoate (III), methyl 8-oxo-9,12-epoxy-9,11-octadeca-
dienoate (IV), and methyl 13-oxo-9,12-epoxy-9,11-octadeca-
dienoate (V).
JAOCS 75, 1313–1319 (1998).
KEY WORDS: Autoxidation of furan fatty acids, conjugated
linoleic acid (CLA), furan fatty acids.
The results presented here differ from previous oxidations
both in the experimental design (exposure to ambient air) and
in the products identified. Data presented here are limited to
those compounds that elute in <1.5 h using a CP-Sil 88 capil-
lary column.
Methyl 9,12-epoxyoctadeca-9,11-dienoate (F9,12) is one of
several secondary autoxidation products of methyl esters of
conjugated linoleic acid (CLA) (1). CLA, a mixture of geo-
metric and positional isomers of octadecadienoic acid con-
taining conjugated double bonds, is reported to have a broad
range of anticarcinogenic and other physiological properties
in animals dosed near the levels that occur naturally in some
foods (2,3).
The furan moiety of F9,12 was first identified as a compo-
nent of Exocarpus seed oil (4). This was the first report of a
naturally occurring 2,5-disubstituted furan fatty acid moiety
(F-acid). The general formula for F-acids is shown as the
methyl ester in Scheme 1. (For a C18 F-acid m = 7 and n = 5.)
Since the first report of F9,12, many different C18 and C20
F-acids have been reported in fish (5,6), latex (7), cattle liver
MATERIALS AND METHODS
Oxidation. F9,12 (60 mg) and methyl stearate (10 mg) were
dissolved in 10 mL of petroleum ether (bp 30–60°C). This so-
lution (1 mL) was added to each of seven test tubes (110-mL
volume) and the petroleum ether was removed from the tubes
by a stream of argon. The tubes, containing air, were capped
and placed in an oil bath at 50°C. Individual tubes were re-
moved from the bath at 4.5, 7.0, 23.0, 46.5, 69.5, and 93 h and
stored for at least 1 h at −20°C prior to gas chromatography
(GC) analysis. The materials were dissolved in 2 mLof isooc-
tane for injection. A similar experiment was performed with
the methyl esters of CLA. The solution was injected into the
gas chromatograph without any further cleanup.
*To whom correspondence should be addrssed at Food and Drug Adminis-
tration, Center for Food Safety and Applied Nutrition, Office of Food Label-
ing (HFS-175), 200 C St., S.W.Washington, DC 20204.
E-mail: mpy@cfsan.fda.gov
Copyright © 1998 by AOCS Press
1313
JAOCS, Vol. 75, no. 10 (1998)

1314
N. SEHAT ET AL.
GC and GC–electron ionization mass spectrometry corresponding to acquisition times of ca. 4.1 and 0.82 s, re-
(GC–EIMS). GC was performed by using a Hewlett-Packard spectively. Both 1H and 13C chemical shifts were referenced
(Avondale, PA) 5890A instrument under the following condi- to CDCl3 and are reported relative to TMS. Homonuclear 2-
tions: column, 50 M × 0.25 mm i.d. CP Sil 88 (Chrompack, dimensional spectra (COSY, NOESY) were recorded with
Raritan, NJ) capillary; helium carrier gas; flame ionization spectral widths of 4164 Hz in each domain while heteronu-
detector (FIO); temperatures (°C): injector 220, detector 280, clear 2-D spectra (HETCOR, HMBC) had spectral widths of
1
column 75 for 2 min, raise 20°C/min to 185°C and hold for 4164 Hz in the H domain and 20K Hz in the 13C domain.
33 min, raise at 4°C/min to 225°C. Samples were run in both Both types of spectra were recorded with 1024 data points in
split and splitless modes.
the F2 dimension and with 256 incremented spectra of either
Low- and high-resolution GC–EIMS analyses were ob- 16 (COSY, NOESY), 64 (HMBC), or 256 scans (HETCOR)
tained with a Hewlett-Packard 5890 series II gas chromato- each. Free-induction decays were processed as 2048 × 2048
graph coupled to a Micromass (Manchester, United Kingdom) matrices with appropriate linear prediction, zero-filling, and
Autospec Q mass spectrometer and OPUS 4000 data system. sine bell (COSY, HETCOR), gaussian (NOESY), or shifted-
The GC–MS system utilized version 2.1 BX software.
The same capillary GC column used to obtain FID data
sine bell (F2)/gaussian (F1) weighting.
Attenuated total reflection (ATR) FTIR spectroscopy. Neat
was used to obtain GC–EIMS data. Adjusting the capillary 5-hexyl-2-furaldehyde (50 µL) was examined by using the Bio-
GC column head pressure to 10 psi gave chromatographic re- Rad FTS-60A system equipped with a ZnSe single bounce hor-
sults comparable to those used to obtain the GC FID data. The izontal ATR (18) cell (Spectra-Tech, Stamford, CT).
GC–EIMS conditions were: splitless injection with helium
Methyl 9,12-epoxyoctadeca-9,11-dienoate. F9,12 was syn-
sweep restored 1 min after injection; injector and transfer thesized from methyl ricinoleate and purified by silica chro-
lines 250°C; oven 75°C for 2 min after injection, then raise matography as previously described (16,19).
20°C/min to 185°C, hold at 185°C 15 min, raise 4°C/min to
225°C, hold at 225°C for 5 min.
Methyl 9,12-dioxo-octadeca-cis-10-enoate. Methyl 9,12-
dioxo-octadeca-cis-10-enoate was synthesized by the reac-
The mass spectrometer was tuned to a resolution of 1000 tion of F9,12 with m-chloroperoxybenzoic acid as reported for
(5% valley) by observing m/z 305 in the EI mass spectrum of methyl 8,11-dioxo-octadeca-cis-9-enoate (13).
perfluorokersene (PFK). The mass scale was calibrated with
5-Hexyl-2-furaldehyde (I). I was prepared as previously
PFK for magnet scans from 440 to 44 daltons at 1 s per described (20): 1.2 mL (1.974 g = 0.014 mol) phosphorus
decade. Filament emission was 200 µA at 70 eV. Ion source oxychloride was added dropwise to 4 mL dimethylformamide
temperature was 250°C. High-resolution MS data were ob- at <0°C. The solution was stirred for 1 h at <0°C and then a
tained by GC–EIMS with the mass spectrometer operated in solution of 1.52 g (0.01 mol) 2-hexylfuran in 4 mL dimethyl-
voltage scan mode at 11 K resolution.
formamide was added slowly. The reaction mixture was
GC–direct deposition–Fourier transform infrared (GC– stirred at <0°C for 1 h and maintained at room temperature
DD–FTIR). A Bio-Rad Digilab division (Cambridge, MA) di- for an additional 1.5 h. The reaction mixture was then poured
rect deposition Tracer FTS-60A mid-infrared transmission into ice water, and solid sodium carbonate (5 g) was added to
(4000–600 cm−1) system and a Hewlett- Packard model 5890 the aqueous mixture. The mixture was left at room tempera-
Series II gas chromatograph were used. This system, which ture for 16 h. The aqueous mixture was extracted three times
was used with a CP-Sil-88 capillary column, has been de- with petroleum ether. The petroleum ether was dried with
scribed in detail (17). Sixty-four scans were collected at 4 sodium sulfate and then removed, yielding 0.96 g of a dark
cm−1 resolution. The GC effluent was sprayed on a moving yellow liquid (53%). Purity was 95.3% as determined by GC.
60 × 30 mm ZnSe crystal; the crystal was mounted on a cop- The mass and FTIR and NMR spectra are presented in the Re-
per block cooled with liquid N2. The block was mounted on sults and Discussion section.
motorized X-Y stage in a vacuum chamber; the IR beam was
focused onto the frozen analytes that gave rise to GC peaks.
The transmitted IR beam was focused onto a 0.1 mm mercury
RESULTS AND DISCUSSION
cadmium telluride detector and the IR measurement took The GC data, based on a neat injection of 0.1 µL, showed a
place in real-time and post-GC run modes.
95+% purity for the synthesized F9,12. Figure 1 shows a chro-
1H nuclear magnetic resonance (NMR) and proton-decou- matogram obtained for an open-air oxidation of F9,12 for 3 d
pled 13C NMR spectra. The spectra, described by 32,768 data at 50°C (see the Material and Methods section for parameters
points (real part), were obtained at 400 and 100 MHz (with used with the CP-Sil 88 column). The numbered peaks were
broad-band irradiation at 400 MHz for the latter), respec- identified from information obtained by using GC-DD-FTIR
tively, by using a Varian NMR Instruments VXR-400S spec- and high- and low-resolution GC–EIMS.
trometer equipped with a Nalorac Z-SPEC MD-400-3 micro-
Peak 1, or (I), was confirmed by comparison with a syn-
1
probe. H and 13C pulse widths of 6.7 and 5 µs were em- thetic reference (Material and Methods section). The NMR
ployed at transmitter powers of 55 and 53, which correspond assignments are given in Scheme 2 (those for 13C are under-
to tip angles of 60 and 45°, respectively, with 3-mm sample lined) and the connectivities are indicated in Table 1. The
tubes. 1H and 13C spectral widths of 4 and 20 kHz were used, FTIR and GC–EIMS spectra obtained for the synthetic I were
JAOCS, Vol. 75, no. 10 (1998)

AUTOXIDATION OF FURAN FATTY ACID ESTER
1315
FIG. 2. Attenuated total reflection Fourier transform infrared (FTIR) spec-
trum of 5-hexyl-2-furaldehyde (I). Frequencies are listed for 4 cm−1 res-
olution. The three weak bands at 2709, 2750, and 2810 cm−1 are char-
acteristic of the aldehyde C–H stretch, and overtone or combination vi-
brations (Ref. 24).
FIG. 1. Gas chromatogram with electron ionization mass spectrometry
(GC–EIMS) total ion current detection for F9,12, oxidized for 3 d in open
air at 50°C. Numbered peaks are identified in the Results and Discus-
sion section. Peaks eluting from 29–34 min were present in the original
F9,12 reference material.
(1672 cm−1) with –RC=CH– (1583 and 1514 cm−1). The
same three bands, obtained by GC–DD–FTIR (Fig. 3), oc-
curred for peaks 3–5. An additional frequency at 1730 cm−1
was also present for these compounds, indicating the presence
of an additional ester carbonyl group. Peak 2 was identified
as methyl 8-oxooctanoate (II) by comparison with its library
mass spectrum and the presence of carbonyl IR bands (cm−1)
at 1733 (R′COOR) and 1720 (R′CHO) in its IR spectrum.
The GC–EIMS data show molecular weights (m/z) of 180
for I, 252 for peak 3, and 322 for both peaks 4 and 5. The EI
SCHEME 2
TABLE 1
1H and 13C Nuclear Magnetic Resonance Dataa
for 5-Hexyl-2-furaldehyde
Position
1H
13C
Indirect CH connectivitiesb
2
3
4
5
—
151.7
123.6
108.5
164.1
28.3
27.4
28.7
31.3
22.4
9.48, 7.15, 6.21
9.48, 6.21
7.15, 2.69
7.15, 6.21, 2.69, 1.67
1.67, 1.30
2.69, 1.30, 1.26
2.69, 1.67, 1.26
1.67, 1.30, 1.26, 0.85
1.30, 1.26, 0.85
1.26
7.15d (3.5)
6.21d (3.5)
—
2.69 t (7.6)
1.67 qu (7.6)
1.30 m
1.26 m
1.26 m
0.85 t (6.7)
9.48
1′
2′
3′
4′
5′
6′
CHO
13.9
176.8
7.15
aChemical shifts referenced to CDCl3 at 7.26 ppm 1H and 77 ppm 13C; J-
values in Hz in parentheses.
bProton multiple bond connectivities (HMBC).
FIG. 3. Gas chromatography–direct deposition–FTIR spectra for peaks
3–5 shown in Figure 1. Wavenumbers (cm−1) are indicated for ester car-
bonyl (1730) and for α,β-conjugated R2C=O (1665–1672). The
–RC=CH– structure exhibited bands at 1517 and 1584–1588 cm−1 (Ref.
useful for the identification of peaks 3–5 (Fig. 1). The FTIR
of I is illustrated in Figure 2. The band frequencies (cm−1) for
the specific functional groups are indicated in the figure. Of
special note are bands indicating the conjugation of –C=O 24). For abbreviation see Figure 2.
JAOCS, Vol. 75, no. 10 (1998)

1316
N. SEHAT ET AL.
FIG. 4. EI mass spectra of (A) synthetic 5-hexyl-2-furaldehyde (I) and (B) gas chromatographic peak 1 in Figure 1. For abbreviations see Figures 1
and 2.
FIG. 5. EI mass spectra of peak 4 identified as methyl 8-oxo-9,12-epoxy-9,11-octadecadienoate (IV) in Figure 5A
and peak 5 identified as methyl 13-oxo-9,12-epoxy-9,11-octadecadienoate (V) in Figure 5B. For abbreviation see
Figure 1.
JAOCS, Vol. 75, no. 10 (1998)

AUTOXIDATION OF FURAN FATTY ACID ESTER
1317
mass spectrum of I is shown in Figure 4A. Their mass spec-
tra are characterized by successive losses from the alkyl chain
(m/z 165, 151, 137, 123, 109) and a hydrogen rearrangement
with cleavage of the chain β to the furan ring (m/z 110). The
low-mass ion series (m/z 81, 95, 109) attributed to furylalkyl
compounds (21) is present in the mass spectrum of I and in
the mass spectra of peaks 1 and 3–5. The IR spectrum of peak
3, which is consistent with an α,β-conjugated carbonyl and
an additional ester carbonyl, was identified as methyl 13-oxo-
9,12-epoxytrideca-9,11-dienoate (III).
would result in radical-directed cleavage γ to the furan ring
(Scheme 4).
γ-Hydrogen rearrangement to a carbonyl oxygen with
cleavage β to the carbonyl (category iii, above) can occur at
the ester or keto function. This process is the origin of m/z 74
in the EI mass spectra of saturated fatty acids. γ-Hydrogen re-
arrangement at the keto function with cleavage β to the car-
bonyl will give rise to m/z 194 if the keto is at C8 (compound
IV), and m/z 266 if the keto is at C13 (compound V)
(Scheme 5).
The components eluting at retention times between 29 and
34 min are impurities in the synthesized F9,12 such as dihy-
drofuran fatty acid methyl esters (16).
γ-Hydrogen rearrangement in the molecular ion and neu-
tral losses of the elements of methanol and carbon monoxide
may account for the presence of m/z 290 and 262 in the spec-
tra. Corresponding losses from the ion at m/z 266 for com-
pound V would result in m/z 234 and 206 in Figure 4B.
Cleavage γ to the ketone (category iv, above) would give rise
to three possible ion structures corresponding to m/z 207 for
compound IV and m/z 279 for compound V (Scheme 6).
On the basis of the MS and IR data, the mass spectrum in
Figure 4A is identified as that of the 8-oxo isomer (IV) and
the mass spectrum in Figure 4B is identified as that of the 13-
oxo isomer (V). Thus, GC peak 4 is assigned the structure of
compound IV and GC peak 5 that of compound V.
The rate of oxidation for F9,12 was determined in sealed
tubes containing 110 cm3 air in order to avoid ambient singlet
oxygen, which is known to react with conjugated dienes (23).
The appearance (mg × 50) of oxidation products I, II, and IV
is shown in Figure 6. The appearance of V is quantitatively
equivalent to IV and is therefore not shown. The time at
which [I] and [III] maximize is ca. 28 h, whereas concentra-
tions of IV and V (not shown) continue to increase at 93 h.
These data are consistent with these sets of products arising
GC–FTIR data provided evidence (above) of the pres-
ence of a conjugated carbonyl in both peaks 4 and 5. The
GC–EIMS data for peaks 4 and 5 are shown in Figure 5. Loss
of the elements of water (m/z 304) from the molecular ion
(m/z 322) and the presence of the acylium ion (m/z 291) are
consistent with a methyl ester structure. The abundant ions in
the EI mass spectra of methyl 8-oxo-9,12-epoxy-9,11-oc-
tadecadienoate (IV) and methyl 13-oxo-9,12-epoxy-9,11-oc-
tadecadienoate (V) arise from fragmentation directed by the
keto and furan functions. The most readily understood
processes that may occur are (i) cleavage α to the carbonyl,
(ii) cleavage β to the furan ring, (iii) γ hydrogen rearrange-
ments with cleavage β to the carbonyl, and (iv) cleavage γ to
the carbonyl corresponding to allylic bond rupture of the enol
form or formation of a cyclic oxonium ion (22). The struc-
tures of IV and V are such that some of the postulated frag-
mentation pathways may produce isobaric products. For ex-
ample, cleavage of the carbon chain of the 8-oxo isomer (IV)
α to the carbonyl (i) and cleavage β to the furan ring (ii) of
the 13-oxo isomer (V) would give rise to m/z 179 (Scheme 3).
Cleavage of the carbon chain α to the carbonyl (i) would give
rise to m/z 251 for compound V, but cleavage β to the ring (ii)
of IV would also give rise to m/z 251. Thus, the task of dis-
tinguishing between the two molecules requires rationalizing
fragmentations that are unique for each structure. This task is
aided by elemental composition data for the ions of interest.
An elemental composition of C12H17O2 for m/z 193 in Fig-
ure 4B is consistent with cleavage γ to the furan ring for com-
pound V. γ-Hydrogen rearrangement to the furan oxygen
SCHEME 4
SCHEME 3
JAOCS, Vol. 75, no. 10 (1998)

1318
N. SEHAT ET AL.
SCHEME 5
SCHEME 6
SCHEME 7
JAOCS, Vol. 75, no. 10 (1998)

AUTOXIDATION OF FURAN FATTY ACID ESTER
1319
abe, The Composition of Furan Fatty Acids in the Crayfish,
Lipids 23:694–700 (1988).
6. Ishii, K., H. Okajima, Y. Okada, and H. Watanabe, Studies of
Furan Fatty Acids of Salmon Roe Phospholipids, J. Biochem.
103:836–839 (1988).
7. Hasma, H., and Subramaniam, A., The Occurrence of a Fura-
noid Fatty Acid in Hevia brasiliensis Latex, Lipids 13:905–907
(1978).
8. Schödel, R., and G. Spiteller, Über das Vorkommen von
F-Säuren in Rinderleber und deren enzymatischen Abbau bei
Gewebeverletzung (The Occurrence of F-Acids in Cattle Liver
and their Enzymatic Degradation During Tissue Damage),
Liebigs Ann. Chem. 5:459–462 (1986).
9. Hannemann, K., V. Puchta, E. Simon, H. Ziegler, G. Ziegler,
and G. Spiteller, The Common Occurrence of Furan Fatty Acids
in Plants, Lipids 24:296–298 (1989).
10. Okada, Y., H. Okajima, H. Konishi, M. Terauchi, K. Ishii, I.-M.
Liu, and H. Watanabe, Antioxidant Effect of Naturally Occur-
ring Furan Fatty Acids on Oxidation of Linoleic Acid in Aque-
ous Dispersion, J. Am. Oil Chem. Soc. 67:858–862 (1990).
11. Batna, A., and G. Spiteller, Effects of Soybean Lipoxygenase-1
on Phosphatidylcholines Containing Furan Fatty Acids, Lipids
29:397–403 (1994).
12. Batna, A., and G. Spiteller, Oxidation of Furan Fatty Acids by
Soybean Lipoxygenase-1 in the Presence of Linoleic Acid,
Chem. Phys. Lipids 70:179–185 (1994).
13. Boyer, R.F., D. Litts, J. Kostishak, R.C. Wijesundera, and F.D.
Gunstone, The Action of Lipoxygenase-1 on Furan Derivatives,
Ibid. 25:237–246 (1979).
14. Lie Ken Jie, M.S.F., M.K. Pasha, and C.K. Lam, Ultrasonically
Stimulated Oxidation Reactions of 2,5-Disubstituted C18 Fura-
noid Fatty Ester, Ibid. 85:101–106 (1997).
FIG. 6. The disappearance of F9,12 and the amounts (mg × 50) of peaks
15. Graff, G., J.L. Gellerman, D.M. Sand, and H. Schlenk, Inhibi-
tion of Blood Platelet Aggregation by Dioxo-ene Compounds,
Biochim. Biophys. Acta 799:143–150 (1984) .
1, 3, and 5 vs. oxidation time (sealed tubes at 50°C).
16. Rosenblat, G., M. Tabak, M.S.F. Lie Ken Jie, and I. Neeman,
Inhibition of Bacterial Urease by Autoxidation of Furan C-18
Fatty Acid Methyl Ester Products, J. Am. Oil Chem. Soc.
70:501–505 (1993).
17. Mossoba, M.M., Applications of capillary GC-FTIR, INFORM
4:854–859 (1993).
18. Mossoba, M.M., M.P. Yurawecz, and R.E. McDonald, Rapid
Determination of Total Trans Content of Neat Hydrogenated
Oils by Attenuated Total Reflection Spectroscopy, J. Am. Oil
Chem. Soc. 73:1003–1009 (1996).
19. Lie Ken Jie, M.S.F., H.W.M. Chan, J.S.M. Wai, and S. Sinha,
Fatty Acids: XXIII. A Rapid Method for the Preparation of C18
Furanoid Fatty Ester Involving Dry-Column Chromatography,
Ibid. 50:705–706 (1981).
from different intermediates (Scheme 7). One-half of the F9,12
is oxidized at ca. 35 h. Methyl esters of CLA are 50% oxi-
dized at ca. 40 h under similar conditions. The quantities of
oxidation products detected by GC is low compared to the
[F9,12] at time zero (e.g., the sum of IV and V present at 93 h
of oxidation amounts to 5% of the total mass balance). Of the
six compounds indicated in Figure 1, only F9,12 and II were
detected in the oxidation of methyl CLA. This work is lim-
ited to the identification of compounds eluting <1.5 h on the
CP-Sil 88 column. The compound methyl 9,12-dioxooc-
tadecadec-10-enoate did not give a response either for the
compound itself or for any artifact or breakdown products
using the parameters described here.
20. Lie Ken Jie, M.S.F., and C.H. Lam, Fatty Acids, Part XVI: The
Synthesis of All Isomeric C18 Furan-Containing Fatty Acids,
Chem. Phys. Lipids 21:275–287 (1978).
21. McLafferty, F.W., and F. Turecek, Interpretation of Mass Spec-
tra, 4th edn., University Science Books, Mill Valley, 1993, p.
352.
22. Budzikiewicz, H., C. Djerassi, and D. Williams, Mass Spectrom-
etry of Organic Compounds, Holden Day, San Francisco, CA,
1967, pp. 135–138.
23. Ogawa, S., S. Fukui, Y. Hanasaki, K. Asano, H. Uegaki, S.
Fujita, and R. Shimazaki, Determination Method of Singlet
Oxygen in the Atmosphere by Use of α-Terpinene, Chemo-
sphere 22:1211–1225 (1991).
24. Nakanishi, Koji, Infrared Absorption Spectroscopy, Holden-
Day, Inc., San Francisco, 1962, p. 43.
REFERENCES
1. Yurawecz, M.P., J.K. Hood, M.M. Mossoba, J.A.G. Roach, and
Y. Ku, Furan Fatty Acids Determined as Oxidation Products of
Conjugated Octadecadienoic Acid, Lipids 30:595–598 (1995).
2. Belury, M.A., Conjugated Dienoic Linoleate: A Polyunsaturated
Fatty Acid with Unique Chemoprotective Properties, Nutr. Rev.
53:83–89 (1995).
3. Parodi, P.W., Cow’s Milk Fat Components as Potential Anticar-
cinogenic Agents, J. Nutrition 127:1055–1060 (1997).
4. Morris, L.J., M.O. Marshall, and W. Kelly, A Unique Furanoid
Fatty Acid From Exocarpus Seed Oil, Tetrahedron Lett.
36:4249–4253 (1966).
5. Ishii, K., H. Okajima, T. Koyamatsu, Y. Okada, and H. Watan-
[Received November 11, 1998; accepted July 15, 1998]
JAOCS, Vol. 75, no. 10 (1998)