Omega-functionalized fatty acids, alcohols, and ethers via olefin metathesis

Article abstract of DOI:10.1007/s11746-012-2015-0

Methyl 17-hydroxy stearate was converted to methyl octadec-16-enoate using copper sulfate adsorbed on silica gel. This compound served as a useful substrate for the olefin metathesis reaction. As a result, several fatty acids with novel functional groups at the x-end were prepared: a glyceryl ether attached at the 18-carbon, an aromatic fatty acid from eugenol, and a ferrocenyl fatty acid. By employing the unsaturated fatty alcohol, other groups were introduced, namely the terminal fluoride, bromide, and iodide were prepared, as was a thiol derivative. The penultimate and omega olefins reported here should serve as building blocks that allow fatty acids to make a greater contribution to a range of emerging technological

Full text of DOI:10.1007/s11746-012-2015-0

J Am Oil Chem Soc (2012) 89:1325–1332
DOI 10.1007/s11746-012-2015-0
ORIGINAL PAPER
Omega-Functionalized Fatty Acids, Alcohols, and Ethers
via Olefin Metathesis
•
Jonathan A. Zerkowski Daniel K. Y. Solaiman
Received: 23 September 2011 / Revised: 22 December 2011 / Accepted: 5 January 2012 / Published online: 3 February 2012
Ó AOCS (outside the USA) 2012
Abstract Methyl 17-hydroxy stearate was converted to
methyl octadec-16-enoate using copper sulfate adsorbed on
silica gel. This compound served as a useful substrate for
the olefin metathesis reaction. As a result, several fatty
acids with novel functional groups at the x-end were
prepared: a glyceryl ether attached at the 18-carbon, an
aromatic fatty acid from eugenol, and a ferrocenyl fatty
acid. By employing the unsaturated fatty alcohol, other
groups were introduced, namely the terminal fluoride,
bromide, and iodide were prepared, as was a thiol deriva-
tive. The penultimate and omega olefins reported here
should serve as building blocks that allow fatty acids to
make a greater contribution to a range of emerging tech-
nological areas.
transformations of the carboxylic group) since that is
where nature overwhelmingly provides the functionality,
usually olefinic and more rarely an alcohol. From the
standpoint of molecular design of lipidic materials, this
limitation is regrettable, since there are fewer options for
controlling the shapes of the chemical building blocks.
While being able to tailor and make use of groups at the
a and x ends of a hydrophobic fatty acid chain would
have an impact on more familiar materials such as
polymers, films and coatings, and bilayers and micelles,
there are also newer areas of technology where an
expanded palette of fatty acid structures could be put to
use, if it were available. These include matrices for
protein crystallization [1], delivery vehicles for drugs or
nucleic acids [2–4], and design of structures ranging
from carbon nanotube arrays [5] to microfluidic reactors
[6] to membrane-permeable prodrugs [7]. Fortunately,
there are some readily available fatty acids that offer
different substitution patterns. In particular, we have
been exploring transformations of 17-OH or 18-OH oleic
and stearic acid, which are readily obtained from gly-
colipids, namely sophorolipids. The sophorolipids in
turn are obtained in good amounts by fermentation of
low-value agricultural byproducts [8]. Our previous
work involved replacing the OH with an azide or prop-
argyl group and performing click reactions [9]. While
this approach proved to be a useful way to link together
disparate types of molecules, since the click reaction is
widely tolerant of many functional groups, there may
be situations where the resulting polar 1,2,3-triazole
might not be desirable, e.g. in a predominantly hydro-
phobic environment. This paper reports the dehydration
of 17-OH stearic acid and subsequent products con-
taining enhanced functionality that are made using olefin
metathesis and other reactions.
Keywords Fats and oils Á Co-products
(waste) \ biobased products Á Polymers/coatings \
biobased products
Introduction
Chemical derivatization of fatty acids is almost exclusively
performed toward the middle of the chain (not counting
Mention of trade names or commercial products in this publication is
solely for the purpose of providing specific information and does not
imply recommendation or endorsement by the US Department of
Agriculture. USDA is an equal opportunity provider and employer
J. A. Zerkowski (&) Á D. K. Y. Solaiman
USDA, ARS, Eastern Regional Research Center,
600 E. Mermaid Lane, Wyndmoor, PA 19038, USA
e-mail: jonathan.zerkowski@ars.usda.gov
123

1326
J Am Oil Chem Soc (2012) 89:1325–1332
Experimental
General
Chemicals and reagents were obtained commercially from
Sigma-Aldrich (St. Louis, MO) and Lancaster Synthesis
(Alfa Aesar, Ward Hill, MA). All solvents and reagents
were used as received. Silica gel used for column chro-
˚
matography (Grade 60 A, mesh 230–400, particle size
40–63 lm), was obtained from Fisher Scientific (Fair-
lawn, NJ). NMR spectra were recorded at room temper-
ature in CDCl3 referenced to tetramethylsilane on either a
Varian Associates (Walnut Creek, CA) Gemini 200 MHz
or an Inova 400 MHZ instrument. LCMS data were
recorded on a Waters 3100 instrument with positive-mode
Electrospray Ionization (ESI), using direct injection (no
column) with 9:1 CH3CN:0.5% HCO2H in CH3CN as
eluent. Masses reported in Da, were determined as sodi-
ated ionic adducts of the products [M?Na]? or protonated
adducts [M?H]?.
Fig. 1 Conversion of a hydroxy fatty acid to a monounsaturated fatty
acid
General Conditions for Cross-Metathesis [11]
The reacting components were dissolved in 10 mL of
methylene chloride in the ratios listed below, Grubbs’
second-generation metathesis catalyst was added (5 mol %
based on limiting metathesis partner) and the mixture was
heated under nitrogen in a water bath at 40–45 °C over-
night. After 18 h, another portion of catalyst was added and
reaction continued for another 8 h. Solvent was removed
and the crude mixture was applied to a silica gel column.
Preparation of x-2 Fatty Acid 2 [10]
The copper sulfate/silica reagent was prepared by slowly
adding 35 g of silica gel to a warm aqueous solution
(approx. 50 °C) of copper sulfate pentahydrate (13.6 g
brought up to 50 mL) with stirring in a crystallization dish.
The slurry that formed remained fluid enough to be mag-
netically stirrable until the last few grams of silica were
added, at which point it was mixed by hand. This method
avoided the presence of a separate aqueous layer, which
would lead to the appearance of discrete copper sulfate
crystals. The wet silica was heated at 50 °C overnight, then
in an oven at 200 °C for 2 h, following which, it was stored
in a vacuum desiccator and used without any further
changes.
Compound 4
2 (0.5 g, 1.8 mmol) and 3 (0.93 g, 5.4 mmol), yield =
1
0.32 g (45%). H NMR 1.20–1.42 m, 1.38 and 1.44 (each
s, 3H, acetonide methyls), 1.55–1.72 m, 1.98–2.12 (m, 2H,
CH2–CH=), 2.32 (t, 7.3 Hz, 2H, CH2C(O)), 3.35–3.59 (m,
2H, CH2OCH2CH=), 3.68 (s, 3H, Me ester), 3.69–3.85 and
3.96–4.16 (dd, 6.4 and 8.2 Hz, 2H, acetonide ring CH2),
3.83–3.94 (m, 2H, CH2OCH2CH=), 4.22–4.35 (m, 1H,
glyceryl CH), 5.45–5.80 (m, 2H, olefin). 13C NMR 25.2,
27.1, 29.4–29.9, 32.8, 34.3, 51.6, 67.2, 71.1, 72.5, 74.9,
109.6, 126.0, 135.5, 174.6. MS calc. for C25H46O5ÁNa?
449.3 found 449.3.
To a solution of methyl 17-hydroxy stearate 1a (1 g,
3.2 mmol) in 25 mL of toluene was added approximately
2 g of the copper sulfate/silica reagent. The toluene mix-
ture was refluxed overnight, after which another portion of
the reagent was added. Total reflux time was 24 h. The
silica was filtered off, toluene removed under vacuo, and
the product 2 was separated from unreacted starting
material by passage through a short column of silica gel
(Fig. 1). Three repetitions gave yields of 53, 57, and 66%.
Carbon tetrachloride could be used as solvent instead of
Compound 7
6 (0.41 g, 1.3 mmol) and 3 (0.68 g, 3.9 mmol), yield =
1
0.28 g (51%). H NMR 1.21–1.42 m, 1.38 and 1.44 (each
1
s, 3H, acetonide methyls), 1.55–1.72 m, 1.98–2.11 (m, 2H,
CH2CH=), 2.06 (s, 3H, acetate CH3), 3.35–3.59 (m, 2H,
CH2OCH2CH=), 3.69–3.81 and 3.96–4.16 (dd, 6.5 and
8.2 Hz, 2H, acetonide ring CH2), 3.81–3.94 (m, 2H,
CH2OCH2CH=), 4.07 (t, 7.5 Hz, 2H, CH2OC(O)),
4.22–4.37 (m, 1H, glyceryl CH), 5.45–5.80 (m, 2H, olefin).
13C NMR 21.2, 25.6, 26.1, 27.0, 28.8, 29.3–29.8, 32.5,
toluene with no apparent differences in yield. H NMR
1.20–1.44 br s, 1.49–1.75 (m 5H, C(O)CH2CH2 and
=CHCH3), 1.90–2.11 (m, 2H, CH2–CH=), 2.32 (t, 7.2 Hz,
2H, CH2C(O)), 3.67 (s, 3H, Me ester), 5.30–5.45 (m, 2H,
olefin). 13C NMR 25.1, 27.1, 29.4–29.9, 32.8, 34.3, 51.7,
124.7, 132.0, 174.6. MS calc. for C19H36O2ÁH? 297.3
found 297.4.
123

J Am Oil Chem Soc (2012) 89:1325–1332
1327
64.9, 67.2, 71.0, 72.6, 75.0, 109.6, 126.1, 135.6, 171.5. MS
calc. for C26H48O5ÁNa? 463.3 found 463.3.
the crude material was almost pure. After removing the
solvent, the material was dissolved in 15 mL CH2Cl2 and
acetic anhydride (0.48 mL, 5 mmol), pyridine (0.5 mL,
6 mmol), and 4-dimethylaminopyridine (DMAP, 20 mg)
were added. The mixture was stirred overnight, extracted
with water, and organic solvent was evaporated. The ace-
tate 6 obtained (0.89 g, 77%) was subsequently used
without further purification. 1H NMR 1.20–1.42, 1.57–1.70
(m, 5H, –OCH2CH2 and =CHCH3), 1.91–2.09 (m, 2H,
CH2–CH=), 2.07 (s, 3H, acetate CH3), 4.07 (t, 7.4 Hz, 2H,
CH2–O) 5.38–5.48 (m, 2H, olefin). 13C NMR 18.0, 22.7,
25.8, 26.9, 29.3–29.8, 32.7, 32.8, 63.5, 124.6, 131.9, 171.7.
MS Calc. for C20H38O2ÁH? 311.3 found 311.4.
Compound 16
2 (0.5 g, 1.8 mmol), eugenol (0.89 g, 5. 4 mmol), yield =
0.42 g (59%). 1H NMR 1.16–1.41 br s, 1.50–1.71 m,
1.94–2.08 (m, 2H, CH2CH=), 2.30 (t, 7.4 Hz, 2H,
CH2C(O)), 3.25 (d, 5.9 Hz, 2H, ArCH2), 3.67 (s, 3H, Me
ester), 3.86 (s, 3H, eugenol OMe), 5.46–5.57 (m, 2H,
olefin), 6.63–6.73 (m, 2H, Ar), 6.84 (d, 6.2 Hz, 1H, Ar).
13C NMR 25.1, 29.2–29.7, 32.6, 34.2, 38.8, 51.5, 55.9,
111.1, 114.2, 121.1, 129.1, 132.1, 133.2, 143.8, 146.5,
174.5. MS calc. for C26H42O4ÁNa? 441.3 found 441.4.
Dihydroxylation to 5 and Epoxidation to 12
Compound 17
Compound 4 (0.15 g, 0.35 mmol) was dissolved in 10 mL
1:1 t-BuOH/H2O, and to the solution was added 0.7 g
commercially available ‘‘AD-b’’ mix (Sharpless’ reagent)
along with 100 mg methanesulfonamide. The mixture was
stirred overnight at room temperature, then sodium meta-
bisulfite was added. The reaction was poured into 200 mL
H2O and 100 mL ethyl acetate and extracted. Removal of
solvent followed by passage through a short plug of silica
gel with 1:1 hexane/ethyl acetate gave analytically pure
material, 94%. 1H NMR 1.13–1.34 br s, 1.36 and 1.44
(each s, 3H, acetonide methyls), 1.50–1.70 m, 2.31
(t, 7.2 Hz, 2H, CH2C(O)), 3.44–3.52 (m, 1H), 3.52–3.71
(m, 4H), 3.66 (s, 3H, methyl ester), 3.69–3.78 (m, 2H),
4.02–4.10 (dd, 6.5 and 8.1 Hz, 1H, one of the acetonide
ring CH2), 4.24–4.33 (m, 1H, glyceryl CH). 13C NMR 25.2,
27.1, 29.4–29.9, 32.8, 34.3, 51.6, 63.3, 66.5, 72.4, 72.8,
74.6, 74.8, 109.7, 174.4. MS Calc. for C25H48O7ÁNa? 483.3
found 483.7.
Allyl ethyl ferrocene ether (0.33 g, 1.2 mmol), 2 (0.15 g,
0.53 mmol), yield = 108 mg (41%). 1H NMR 1.56
(d, 7.5 Hz, 3H, OCHCH3), 1.58–1.73 m, 1.91–2.07 (m, 2H,
CH2CH=), 2.32 (t, 7.3 Hz, 2H, CH2C(O)), 4.16 (br s, 9H,
ferrocene), 4.11–4.20 (m, 2H, OCH2), 4.20–4.30 (m, 1H,
OCH), 5.33–5.43 (m, 2H, olefin). 13C NMR: 20.1, 25.1,
27.3, 29.2–29.9, 32.8, 34.2, 51.5, 65.9, 67.9, 68.2, 68.8 (br),
75.0, 130.0, 130.5, 174.4. MS calc. for C31H47FeO3ÁH?
524.3 found 524.6.
Compound 3 was prepared by dissolving commer-
cially available 3-allyloxy-1,2-propane diol in CH2Cl2,
adding 2,2-dimethoxy propane (3 equiv) and pyridinium
p-toluenesulfonate (PPTS, 5 mol%), and stirring at room
temperature overnight. The solvent was removed and the
crude was passed through a short plug of silica gel to
remove PPTS, affording 3. The allyl ether of ferrocenyl
ethanol was prepared by combining ferrocenyl ethanol
with allyl bromide (3 equiv) in DMSO and adding
powdered KOH (1.5 equiv), then pouring the mixture
into water after 2 h and extracting with 1:1 hexane/ethyl
ether.
Compound 7 (0.15 g, 0.34 mmol) was dissolved in
10 mL CH2Cl2 and 100 mg meta-chloroperbenzoic acid
was added. The reaction was stirred at room temperature
overnight. Sodium metabisulfite and 10 mL H2O were
added and stirred for 30 min, then the mixture was
extracted with saturated NaHCO3 (3 9 20 mL). Solvent
was removed and the material was passed through a short
plug of silica gel with 3:1 hexane/ethyl acetate to obtain 12
(0.13 g, 85%). 1H NMR 1.15–1.45 br s, 1.34 and 1.41
(each s, 3H, acetonide methyls), 1.50–1.65 m, 2.03 (s, 3H,
acetate methyl), 2.76–2.82 and 2.87–2.94 (each m, 2H,
epoxide ring), 3.39–3.61 (m, 4H, CH2OCH2), 3.67–3.78
(m, 1H, one of the CH2 acetonide ring protons), 3.96–4.08
(m, 3H, CH2OC(O) and one of the CH2 acetonide ring
protons), 4.21–4.32 (m, 1H, glyceryl CH). 13C NMR 21.3,
25.6, 26.1, 26.9, 28.8, 29.4–29.9, 31.9, 56.2, 57.1, 64.9,
66.8, 72.1, 72.5, 74.9, 109.8, 171.6. MS calc. for
C26H48O6ÁNa? 479.3 found 479.6.
Reduction of Methyl Ester to Give 6 [12]
Compound
2
(1.1 g, 3.9 mmol) and poly(meth-
ylhydrosiloxane) (0.58 g, 10 mmol Si–H equivalents) were
placed in a flask, and then titanium tetraisopropoxide
(1.1 g, 4.5 mmol) was added. The mixture was stirred
magnetically under nitrogen and heated to 60 °C for 3 h.
After cooling to room temperature, 30 mL THF was added,
followed by cautious dropwise addition of 30 mL 4 M
NaOH. This mixture was stirred overnight. To remove
some of the THF, a stream of nitrogen was directed over it
for a few hours, then it was poured into 300 mL H2O and
extracted with 1:1 hexane/ethyl ether. TLC indicated that
123

1328
J Am Oil Chem Soc (2012) 89:1325–1332
Eliminations to Give Terminal Alkenes [13]
and Substitution with Thioacetate
was added. The mixture was stirred overnight, then poured
into 300 mL of H2O and extracted with ether (50 mL 9 2).
Column chromatography in 9:1 hexane/diethyl ether
afforded terminal alkene 13 (154 mg, 71%) and fluoride 14
Compound 7 (0.25 g, 0.6 mmol) was dissolved in 30 mL
of MeOH and hydrogenated over 5% Pd/C for 4 h. The
solid was filtered off, the solution was concentrated in
vacuo to about 5 mL of methanol, and a solution of LiOH
(84 mg, 2 mmol) in 2 mL of water was added. THF was
added until the mixture was homogenous (approx. 3 mL),
then it was heated at 45 °C for 3 h. Solvent was removed,
citric acid was added to lower the pH to approximately 3,
and the crude was extracted with 1:1 hexane/ethyl ether.
The free alcohol thus obtained was dissolved in 5 mL
of 1,2-dichloroethane, then to this solution were added
NEt3 (0.34 mL, 2.5 mmol), triphenylphosphine (0.34 g,
1.3 mmol), and 1,2-dibromotetrachloroethane (0.42 g, 1.3
mmol). The mixture was stirred at room temperature for
1 h, after which time tlc indicated complete conversion of
the starting material to a less polar compound. Solvent was
removed, the material was applied to a silica gel column,
and the product 18-bromo compound was isolated. This
compound was then dissolved in 20 mL of acetone to
which NaI (0.9 g, 6 mmol) was added and heated at 45 °C
overnight. Solvent was removed in vacuo, and the material
was extracted with ether. The resulting iodide 8 (249 mg,
86% from 7) was split into two portions. One portion
(150 mg, 0.3 mmol) was dissolved in 3 mL DMSO and
tetrabutylammonium fluoride hydrate (0.34 g, 0.8 mmol)
was added. The mixture was stirred overnight, then poured
into 300 mL of H2O and extracted with ether (50 mL 9 2).
Column chromatography in 9:1 hexane/ether gave the ter-
minal alkene 9 (70 mg, 63%) and the fluoride 10 (29 mg,
25%). The other portion (80 mg, 0.16 mmol) was dissolved
in acetone along with potassium thioacetate (114 mg,
1 mmol) and heated to 45 °C overnight. Solvent was
removed, the crude applied to silica gel, and eluted with
1
(45 mg, 20%). 13: H NMR 1.20–1.45 br s, 1.53–1.71 m,
1.97–2.12 (m, 2H, CH2CH=), 2.32 (t, 7.2 Hz, 2H,
CH2C(O)), 3.68 (s, 3H, Me ester), 4.90–5.08 (m, 2H,
CH2=), 5.72–5.95 (m, 1H, =CH). 13C NMR 25.1, 29.0–29.7
br, 33.9, 34.2, 51.7, 114.2, 139.3, 174.5. MS calc. for
1
C19H36O2ÁH? 297.3, found 297.2. 14: H NMR 1.20–1.49
br s, 1.54–1.73 m, 1.73–1.87 m, 2.32 (t, 7.3 Hz, 2H,
CH2C(O)), 3.68 (s, 3H, Me ester), 4.34 and 4.57 (two
triplets, 6.6 Hz, 1J H/F = 47.3 Hz, 2H, CH2–F). 13C NMR:
25.1, 27.0, 28.9–27.9, 30.3, 30.7, 32.8, 34.7, 51.5, 82.7,
85.9, 174.2. MS calc. for C19H37FO2ÁNa? 339.4 found
339.3.
Non-Metathesis Route to Glyceryl Ether
Acetate 6 (0.65 g, 2.1 mmol) was dissolved in 10 mL
MeOH, then LiOH hydrate (0.13 g, 3 mmol) in 3 mL H2O
was added, followed by THF until the solution became
homogeneous with only slight haziness, about 7 mL total
THF. The reaction was heated at 45 °C in a waterbath for
3 h, then at RT overnight. Solvents were removed, satu-
rated citric acid solution was added and the mixture was
extracted with ether (2 9 50 mL). After removing the
ether, the crude material was dissolved in methylene
chloride, and pyridine (0.4 mL, 5 mmol) and methanesul-
fonyl chloride (0.35 g, 3 mmol) were added. The reaction
was stirred at RT overnight, then solvent was removed and
the crude was purified by passage through a short silica gel
plug. This mesylate 6a and solketal (0.66 g, 5 mmol) were
dissolved in 2 mL of DMSO, then powdered KOH (0.17 g,
3 mmol) was added and stirred vigorously for 2 h, after
which time the mixture was poured into 500 mL of H2O
and extracted with 100 mL of ether. Column chromatog-
raphy afforded 15 in 67% yield (0.49 g) based on the
1
4:1 hexane/ether to give 11 (69 mg, 93%). H NMR (only
the chemically relevant new peaks differing from 7 are
listed) 9: 4.90–5.08 (m, 2H, =CH2), 5.72–5.95 (m, 1H,
=CH). 13C NMR 114.2 and 139.5 (terminal olefin). MS
1
starting acetate 6. H NMR 1.20–1.35 br s, 1.38 and 1.44
(each s, 3H, acetonide methyls), 1.50–1.72 (m 5H,
OCH2CH2 and =CHCH3), 1.89–2.11 (m, 2H, CH2–CH=),
3.36–3.59 (m, 4H, –CH2OCH2–), 3.68–3.75 and 4.01–4.06
(dd, 6.4 and 8.2 Hz, 2H, acetonide ring CH2), 4.19–4.35
(m, 1H, glyceryl CH), 5.32–5.48 (m, 2H, olefin). 13C NMR
18.2, 25.6, 26.3, 27.0, 29.4–29.9, 32.8, 67.1, 72.1, 72.2,
75.0, 109.6, 123.8 (cis), 124.7, 131.1 (cis), 131.9. MS calc.
for C24H46O3ÁNa? 405.3 found 405.6.
Calc. for C24H46O3ÁNa? 405.3 found 405.6. 10: H NMR
1
4.34 and 4.57 (each t, 6.5 Hz, 2H, 1J H/F = 47.3 Hz, CH2–
F). 13C 83.6, 85.3 (CH2–F). MS Calc. for C24H47FO3ÁNa?
425.3 found 425.5. 11: 2.87 (t, 6.8 Hz, 2H, CH2–S). 13C
NMR 30.8 (S–CH2), 196.4 (S–C=O). MS Calc. for
C26H50O4SÁNa? 481.3 found 481.5.
Compound 1b (0.27 g, 0.86 mmol) was treated similarly
as 7 in the preceding paragraph to make the iodide,
using NEt3 (0.56 mL, 4 mmol), triphenylphosphine (0.45 g,
1.7 mmol), and 1,2-dibromotetrachloroethane (0.55 g,
1.7 mmol), then NaI (1.35 g, 9 mmol). The resulting
iodide (0.32 g, 0.77 mmol) was dissolved in 5 mL DMSO,
and tetrabutylammonium fluoride hydrate (0.84 g, 2 mmol)
Alkylation of Phenolic Fatty Acid 16
Compounds 16 (200 mg, 0.48 mmol), 1a (220 mg, 0.7
mmol) and triphenylphosphine (158 mg, 0.6 mmol) were
dissolved in 10 mL THF. To this solution was added a
123

J Am Oil Chem Soc (2012) 89:1325–1332
1329
Cross-Metathesis Reactions
solution of diisopropyl azodicarboxylate (121 mg, 0.6
mmol, in 3 mL THF) dropwise. The reaction was stirred
magnetically under N2 overnight, then solvent was
removed at reduced pressure and the crude was applied to a
silica gel column. Elution with 3:1 hexane/ethyl acetate
afforded 16a (235 mg, 69%). 1H NMR 1.22 (d, 6.8 Hz, 3H,
OCHCH3), 1.25–1.53 br s, 1.53–1.75 m, 1.90–2.06 (m, 2H,
=CHCH2), 2.31 (t, 7.4 Hz, 4H, 2 9 CH2C(O)), 3.23 (d,
5.9 Hz, 2H, ArCH2), 3.67 (s, 6H, 2 9 methyl ester), 3.81
(s, 3H, Ar–OCH3), 4.07–4.15 (m, 1H, ArOCH), 5.41–5.53
(m, 2H, olefin), 6.63–6.73 (m, 2H, Ar), 6.79 (d, 6.4 Hz, 1H,
Ar). 13C NMR: 19.3, 22.9, 25.2, 25.6, 29.3–29.9, 32.1,
32.8, 34.3, 38.9, 51.6, 56.1, 80.1, 112.7, 116.1, 120.4,
129.1, 132.2, 134.1, 146.5, 150.5, 174.6. MS calc. for
C45H78O6ÁNa? 737.8 found 737.8.
After installing the olefin at the C16–C17 position of the
chain, we set out to demonstrate the variety of functional
groups that could be added at this site of a fatty acid by
employing olefin cross-metathesis with Grubbs’ ‘‘second-
generation’’ catalyst [11]. This reaction has been applied to
mid-chain olefinic fatty acids before [15–18], but not, to
our knowledge, closer to the hydrophobic terminus of the
C18 chain. The obvious difference is that a larger pro-
portion of the carbon chain is retained, up to the C16 point
instead of C9 with oleates, which opens up a new range of
longer lipidic building blocks. A quite different but fasci-
nating reaction that affords long-chain a,x-FA using a Pd
catalyst has been dubbed ‘‘isomerizing carbonylation’’
[19]. Cross-metathesis is of course an inherently compli-
cated proposition, since the system is at equilibrium, and a
number of different products that can themselves partici-
pate in secondary equilibria are the result. Fortunately,
when the olefins are terminal or near the end of a chain, as
they are for all the examples here, low-molecular weight
byproducts (here ethylene, propene, or 2-butene) are lost to
evaporation, simplifying the resulting product mix.
Another way to drive the reaction to the desired mixed
product is to use one partner in excess [11]. In our work,
we used 3 eq of the non-FA component.
Results and Discussion
Preparation of x-2 Alkene
The necessary first step is to perform dehydration of the
penultimate alcohol to afford an alkene. Our standard
starting material was methyl 17-hydroxystearate, pro-
duced by hydrogenation of the oleic hydroxy fatty acid
that is the most abundant component in sophorolipids. In
principle, there should be many methods to accomplish
the dehydration, most of which involve converting the OH
to a leaving group (halide, sulfonate ester) and elimina-
tion. Even the most straightforward of these routes,
however, requires the cost and time of extra reagents
which could furthermore complicate cleanup. While such
routes can work well, and we have employed some in the
past [14], we sought a more direct route that could be
performed without an intermediate transformation. The
direct elimination of alcohols to alkenes using copper
sulfate adsorbed on silica gel has been reported [10]. It
worked adequately in our hands; several repetitions gave
yields in the 53–66% range after 24 h. While these yields
are not high, the simplicity of the reaction and the ease of
separation of product from unreacted starting material,
which can then easily be recycled for another round, more
than make up for moderate yields. A side benefit is that
this dehydration works tolerably on secondary alcohols
but not at all on primary ones, so that if the 18-OH
stearate is present, it remains unchanged. This terminal
alcohol is a minor component isolated from the sophor-
olipids, and normally we separated it away from the
17-hydroxy version using silica gel chromatography prior
to the dehydration, but in one test reaction where we did
not remove it, it was recovered intact, although there did
appear to be a little of an unknown side product that was
formed.
The first, fairly obvious moiety to include was a modi-
fied glycerol unit, 3. Cross-metathesis with this partner
would, after removal of the acetonide protecting group and
diesterification, give a structure analogous to a triacyl-
glycerol, in particular an ether version that is ‘‘inverted’’ in
the sense that one carboxyl group is at the opposite end
from the glycerol (Fig. 2). Glycerol ethers are of interest
for their biological activity, and have been identified as
crucial for inducing adipogenesis [20]. We obtained 4 in
45% yield. Yields reported here are unoptimized; see [18]
for an example of a thorough investigation of reaction
parameters. Interestingly, that work found that phenol-
derived Schiff base variants of the catalyst were particu-
larly active, while use of the same second-generation cat-
alyst as here gave 48% conversion of methyl oleate and
30% yield of desired cross-metathesis products even when
the cross-metathesis partner was present in a fivefold
excess. One modification that can be made to 4 is to use the
olefin near the glycerol unit as the site of further attach-
ment of fatty acids, namely by dihydroxylation to give 5.
This molecule possesses two distinct groups of hydroxy
units, two tied up as the acetonide and two free, that could
be used independently to attach different groups.
Glyceryl ether 4 also opens the door to novel chemistry
that can be done at the carboxylic acid so that func-
tional groups can be positioned at, for example, the inter-
layer region of a bilayer. Making esters would be a
123

1330
J Am Oil Chem Soc (2012) 89:1325–1332
straightforward way to proceed, but we chose to not retain
the carboxylic unit and instead explored options for dif-
ferent kinds of chemistry. (Although it is important to note
that molecule 4 is an AB2 monomer which could be used to
build hyperbranched or dendritic polyesters). Reduction of
the methyl ester to an alcohol would be the starting point
for several new lipidic structures. We were concerned,
however, that reduction protocols might also adversely
affect the acetonide protecting group, so instead we per-
formed the reduction on 2 to give an x-2 unsaturated
alcohol first, prior to cross-metathesis. Reduction of the
methyl ester using Buchwald’s poly(methylhydrosiloxane)/
Ti(OiPr)4 method worked smoothly [12]. This protocol
appears to be a good, general, mild method for transfor-
mation of FAs to fatty alcohols. The alcohol was protected
as an acetate (see [21] for a related approach using oleyl
alcohol), and yields of cross-metathesis with 6 were similar
to those for the methyl ester.
When the fluoride elimination conditions were applied
to 18-OH stearic variant 1b, subsequent to bromination and
iodination, the terminal alkene/fluoride 13 and 14 mixture
were again obtained and were separable. It is interesting to
note that overall this reaction sequence, using either 7 or
1b, affords three terminal halides, F, Br, and I; the fluoro
and iodo versions could be useful as radiolabeled com-
pounds. In our opinion, molecule 6 has the potential to be a
powerful synthon for constructing new fatty acid/alcohol
derivatives. Our aim was to use the acetate as a masked
alkene precursor, through the OH to halide to terminal
alkene sequence. In this way, iterative cross-metatheses for
example between 9 and 6 could be used to build up very
long chain molecules.
Utilizing the a,x groups of 6 in a different sequence
would form the basis of an alternate route to a close rela-
tive of 9 that did not rely on cross-metathesis. This alter-
native route consists of deprotecting the acetate, converting
the OH to a mesylate or other leaving group, and then
forming an ether linkage between this molecule and
We employed acetate 7 for three different sets of
transformations. First was elimination to give a new ter-
minal alkene. A mild, efficient procedure was recently
reported for this conversion, using the terminal iodide and
tetrabutylammonium fluoride hydrate [13]. Elimination
was accomplished by the route shown in Fig. 3. The ace-
tate was hydrolyzed, the OH converted to Br then to I, and
fluoride ion was used as base. To our surprise, the fluoride
by-product 10 was separable from the terminal alkene 9
using silica gel chromatography. Conveniently, this route
also was the set-up for a second transformation of 7, via 8,
namely substitution with thioacetate to give 11 (Fig. 4).
Positioning a sulfur atom at the end of a hydrophobic chain
will contribute to the longer-term goal of constructing tri-
acylglycerol-like molecules that could be adhered to metal
surfaces as self-assembled monolayers. Thirdly, by way of
analogy to the transformation of 4 to 5, we also used the
olefin of 7 in an oxygenating reaction, this time
epoxidation.
Fig. 3 Reagents and conditions
a
poly(methylhydrosiloxane),
Ti(OiPr)4, then NaOH; b acetic anhydride, pyridine, dmap; c Grubbs’
second-generation metathesis catalyst, 3; d H2 over Pd/C; e LiOH in
MeOH/water/THF; f Ph3P, BrCl2CCCl2Br in ClCH2CH2Cl; g NaI in
acetone; h Bu4NF hydrate, DMSO
Fig. 2 Conditions a Grubbs’ second-generation Ru olefin metathesis
catalyst, CH2Cl2, 40 °C, 24 h. b Sharpless’ asymmetric dihydroxy-
lation reagent, t-BuOH/H2O, 16 h
123

J Am Oil Chem Soc (2012) 89:1325–1332
1331
Fig. 5 Reagents and conditions a Grubbs’ second-generation metath-
esis catalyst, eugenol; b 1a, PPh3, diisopropyl-azodicarboxylate;
c Grubbs’ second-generation metathesis catalyst, allyl ferrocenyl
ethyl ether
adding aromatic units to fatty acids, on the theory that
improvement to properties such as lubrication or oxidation
resistance may result [22]. Olefin cross-metathesis is an
excellent way to install an aromatic unit. We used the bio-
derived compound eugenol to obtain 16. The phenolic OH
is still available for further chemistry. We employed it in
the Mitsunobu reaction with another unit of 1a to give the
pseudo-dimeric fatty acid ether 16a (Fig. 5). This series of
steps should be a potent way to create aromatic ethers of
fatty acids. Finally, this methodology presents an oppor-
tunity to include redox-active units in lipid bilayers [23].
Ferrocenyl fatty acid 17 is the first example of an entirely
new class of fatty acid derivative.
Fig. 4 Reagents and conditions a potassium thioacetate in ace-
tone; b m-ClC6H4CO3H in CH2Cl2; c Ph3P, BrCl2CCCl2Br in
ClCH2CH2Cl; d NaI in acetone; e Bu4NF hydrate, DMSO; f LiOH in
MeOH/water/THF; g methanesulfonyl chloride, pyridine, CH2Cl2;
h solketal, DMSO, KOH
solketal to give 15. Molecules 9 and 15 represent starting
points for experiments that would test subtle differences in
the sizes of bilayer-spanning linkers, since self-metathesis
of 15 would give a C32 chain connecting two glyceryl units
and self-metathesis of 9 would give a C34 chain. If the
small difference in carbon chain length were unimportant
to a given application, molecule 15 would probably be a
preferred building block, since it is prepared by a better
synthetic route.
Conclusions
Using a cheap and easily prepared heterogeneous reagent,
bio-derived hydroxy fatty acid 1a is conveniently con-
verted into an x-2 monounsaturated fatty acid. This com-
pound and its alcohol analogue that is simple to prepare are
useful substrates for transition metal-catalyzed metathesis,
which is a wide-ranging technique for introducing func-
tional groups. Taken collectively, these x-functionalized
fatty acids should promote investigations into new
advanced lipidic materials. The added functional groups
provide a springboard to continued derivatization, e.g.
through subsequent metathesis reactions for 9, 13, or 15,
disulfide formation for 11, alkylation of the phenolic OH
for 16, and transition metal catalyzed couplings for 8.
Other functional groups besides glycerol can of course
be x-appended to the fatty acid chain using metathesis. To
limit the range of choices, we investigated only non-polar
groups that would be of interest in a membrane or bilayer
environment. (Although as an aside, since allylation of
carbohydrates is readily performed, this method may prove
to be a powerful way to append fatty acids to sugars
through non-hydrolyzable linkages. Such structures would
be alternatives to fatty acid esters of sugars, which are
under investigation as bio-surfactants, and even to the
native sophorolipid structure, where the glycosidic linkage
to the fatty acid is also hydrolyzable.) There is interest in
123

1332
J Am Oil Chem Soc (2012) 89:1325–1332
12. Reding MT, Buchwald SL (1995) An inexpensive air-stable
titanium-based system for the conversion of esters to primary
alcohols. J Org Chem 60:7884–7890
References
1. Li D, Lee J, Caffrey M (2011) Crystallizing membrane proteins
in lipidic mesophases. A host lipid screen. Cryst Growth Des
11:530–537
2. Carmona-Ribeiro AM (2010) Biomimetic nanoparticles: prepa-
ration, characterization and biomedical applications. Int J Nan-
omed 5:249–259
3. Kusumoto K-I (2009) Structural features of glutamate-based
lipidic materials for small interfering RNA delivery system.
J Biomed Mater Res Part A 89:739–750
4. Belletti D, Riva G, Tosi G, Forni F, Barozzi P, Luppi M, Vandelli
MA, Ruozi B (2011) Novel polymeric/lipidic hybrid systems
(PLHs) for effective Cidofovir delivery: preparation, character-
ization and comparative in vitro study with polymeric particles
and liposomes. Int J Pharm 413:220–228
´
´
13. Berube M, Kamal F, Roy J, Poirier D (2006) A dehydrohalo-
genation methodology for synthesizing terminal olefins under
mild conditions. Synthesis 2006:3085–3091
14. Zerkowski JA, Solaiman DKY (2007) Polyhydroxy fatty acids
derived from sophorolipids. J Am Oil Chem Soc 84:463–471
15. Behr A, Krema S (2011) Metathesis applied to unsaturated lipid
compounds. Lipid Technol 23:156–157
16. Behr A, Westfechtel A, Gomes JP (2008) Catalytic processes for
the technical use of natural fats and oils. Chem Eng Technol
31:700–714
17. Miao X, Dixneuf PH, Fischmeister C, Bruneau C (2011) A green
route to nitrogen-containing groups: the acrylonitrile cross-
metathesis and applications to plant oil derivatives. Green Chem
13:2258–2271
18. Behr A, Gomes JP (2011) The cross-metathesis of methyl oleate
with cis-2-butene-1,4-diyl diacetate and the influence of pro-
tecting groups. Beilstein J Org Chem 7:1–8
5. Thauvin C, Perino A, Contal E, Morin E, Schultz P, Meunier S,
Wagner A (2011) Programmed dispersions of MWNTs in aque-
ous media by coating with photopolymerizable synthetic am-
phiphiles. J Phys Chem C 115:7319–7322
19. Stempfle F, Quinzler D, Heckler I, Mecking S (2011) Long-chain
linear C19 and C23 monomers and polycondensates from
unsaturated fatty acid esters. Macromolecules 44:4159–4166
20. Homan EA, Kim Y-G, Cardia JP, Saghatelian A (2011) Mono-
alkylglycerol ether lipids promote adipogenesis. J Am Chem Soc
133:5178–5181
21. Rybak A, Meier MAR (2008) Cross-metathesis of oleyl alcohol
with methyl acrylate: optimization of reaction conditions and
comparison of their environmental impact. Green Chem
10:1099–1104
6. Jesorka A, Stepanyants N, Zhang H, Ortmen B, Hakonen B,
Orwar O (2011) Generation of phospholipid vesicle-nanotube
networks and transport of molecules therein. Nat Protoc
6:791–805
7. Kaki SS, Adlercreutz P (2011) Lipase-catalyzed synthesis and
characterization of 1-butanoyl-2-palmitoyl phosphatidylcholine,
a potential lipidic prodrug of butyric acid. Chem Phys Lipids
164:246–250
8. Ashby RD, Solaiman DKY, Foglia TA (2008) Property control of
sophorolipids: influence of fatty acid substrate and blending.
Biotechnol Lett 30:1093–1100
9. Zerkowski JA, Nun˜ez A, Strahan GD, Solaiman DKY (2009)
Clickable lipids: azido and alkynyl fatty acids and triacylglyce-
rols. J Am Oil Chem Soc 86:1115–1121
22. Black KD, Gunstone FD (1996) The Friedel-Crafts adducts of
methyl oleate with benzene and toluene. Chem Phys Lipids
79:87–94
´
ˆ
23. Mertins O, Messina P, Labbe E, Vivier V, Arbault S, Lemaıtre F,
Buriez O, Amatore C (2011) Electrochemical analysis of the
interactions and reactivity of ferrocene-based drugs with a lipid
environment: a qualitative overview. Inorg Chim Acta 374:59–68
10. Nishiguchi T, Machida N, Yamamoto E (1987) Dehydration of
alcohols catalyzed by copper(II) sulfate adsorbed on silica gel.
Tetrahedron Lett 28:4565–4568
11. Chatterjee AK, Choi T-L, Sanders DP, Grubbs RH (2003) A
general model for selectivity in olefin cross metathesis. J Am
Chem Soc 125:11360–11370
123