2-fatty acrylic acids: New highly derivatizable lipophilic platform molecules

Article abstract of DOI:10.1007/s11746-014-2459-5

This paper reports the incorporation of an alpha-methylene unit into fatty acid skeletons. Since the new olefin is conjugated with the carboxylate, it is susceptible to 1,4-(Michael) additions. We have used multifunctional thiols and amines for additions at the methylene. The resulting products possess clusters of functionality grouped at one end of a hydrophobic tail. These structural patterns will be of use in the design of new types of bio-based surfactants and polymers. One particularly promising pattern of functionality that can be obtained through oxidation and reduction chemistry is a 2-fatty 1,2,3-propanetriol, or a lipophilized glycerol moiety.

Full text of DOI:10.1007/s11746-014-2459-5

J Am Oil Chem Soc
DOI 10.1007/s11746-014-2459-5
ORIGINAL PAPER
2-Fatty Acrylic Acids: New Highly Derivatizable Lipophilic
Platform Molecules
•
Jonathan A. Zerkowski Daniel K. Y. Solaiman
Received: 18 September 2013 / Revised: 11 March 2014 / Accepted: 17 March 2014
Ó AOCS (outside the USA) 2014
Abstract This paper reports the incorporation of an
alpha-methylene unit into fatty acid skeletons. Since the
new olefin is conjugated with the carboxylate, it is sus-
ceptible to 1,4-(Michael) additions. We have used multi-
functional thiols and amines for additions at the methylene.
The resulting products possess clusters of functionality
grouped at one end of a hydrophobic tail. These structural
patterns will be of use in the design of new types of bio-
based surfactants and polymers. One particularly promising
pattern of functionality that can be obtained through oxi-
dation and reduction chemistry is a 2-fatty 1,2,3-propane-
triol, or a lipophilized glycerol moiety.
could react. If the definition of ‘‘reaction’’ is broadened to
include biological whole-cell routes, there may be hydro-
xylase/cytochrome-dependent oxidations that can occur at
the distal alkyl end [1]. Essentially, FA contain a few
reactive sites buried in a mostly unreactive medium. One
site that has been somewhat underused is the alpha carbon.
Classical chemical procedures can be performed there, but
these generally require harsher reagents such as a strong
base or ones that will also react with unsaturation [2]. The
targets that can emerge are appealing however. 2-Hydroxy
oleic acid, also known as Minerval, has promise as an
anti-cancer and hypotensive drug and is known to alter
membrane structure [3, 4]. Triacylglycerol derivatives of
2-amino oleic acid are inhibitors of human pancreatic
lipase [5].
Keywords Oleochemistry ꢀ Surfactants ꢀ Detergents ꢀ
Biobased products
It is clear that the range of applications where FA could
be employed would be enlarged by being able to add extra
functionality near to the carboxyl group. Another way to
state this is that there would be a benefit to being able to
attach FA to other functional groups through procedures
other than esterification or amidation. When viewed from
this perspective, the desired chemical changes become not
just derivatization of a preexisting FA, but a way to graft
the lipophilic tail of the FA into some other kind of
structure, that is, a generalized protocol for lipophilization
of other molecules [6]. Even more beneficial than adding a
single kind of new functional group in a one-at-a-time
approach would be a modification that could serve as a
gateway or springboard to increased molecular diversity.
Conceptually, an analogous situation is offered by intro-
duction of an azide unit, which subsequently allows a huge
number of derivatives to be made by click chemistry. We
report here a modification of the alpha-region of the FA
that enables this kind of extensive diversification, namely
insertion of a conjugated 1,1-disubstituted methylene [7].
Introduction
From the molecular standpoint, the options for derivatizing
a fatty acid (FA) are limited. There are transformations
of the carboxylic unit itself; if unsaturation is present,
reactions can of course be done there. Slightly more
complicated, if precise regiocontrol is desired, would be
modification at allylic sites, since either side of the olefin
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). USDA is an equal opportunity provider and
employer.
J. A. Zerkowski (&) ꢀ 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

J Am Oil Chem Soc
The new molecules that are produced, which could be
labeled 2-fatty acrylic acids, will serve as valuable plat-
forms not only for the medicinal chemistry applications
previously mentioned, but as research tools for probing
lipid membranes [8] and as building blocks for new kinds
of surfactants and polymeric materials.
Materials and Methods
General
Chemicals and reagents were obtained commercially from
Sigma-Aldrich (St. Louis, MO) and Alfa Aesar (Ward Hill,
MA). All solvents and reagents were used as received.
˚
Silica gel used for column chromatography (grade 60 A,
mesh 230–400, particle size 40–63 lm), was obtained from
Fisher Scientific (Fairlawn, NJ). NMR spectra were
recorded at room temperature in CDCl3 referenced to tet-
ramethylsilane on either a Varian Associates (Walnut
Creek, CA) Gemini 200 MHz or Inova 400 MHZ instru-
ment. LCMS data were recorded on a Waters 3100
instrument with positive-mode atmospheric pressure
chemical ionization (APCI), using direct injection (no
column) with 5:4:1 acetone/CH3CN/0.5 % HCO2H in
CH3CN as eluant. Masses, reported in Da, were determined
as sodiated ionic adducts of the products [M ? Na]? or
protonated adducts [M ? H]?.
Preparation of 2-Fatty Acrylic Acids (Fig. 1 [7])
A solution of oleic acid (1.0 g, 3.5 mmol), 4-dimethyl-
aminopyridine (DMAP, 0.68 g, 5.5 mmol), and Meldrum’s
acid (0.58 g, 4.0 mmol) was prepared in 25 mL CH2Cl2
and cooled in an ice bath. To this solution is added EDC
(N-(3-dimethylaminopropyl)-N0-ethylcarbodiimide hydro-
chloride, 0.80 g, 4.2 mmol) over the course of 1 h. The
reaction was allowed to warm to room temperature and
stirring was continued overnight. It was extracted with
10 % w/v citric acid solution (2 9 100 mL), and the
organic solvent was dried over MgSO4, cooled in an ice
bath, and stirred under N2. Acetic acid (2.4 mL) was added,
followed by NaBH4 (0.33 g, 8.7 mmol) piecewise over the
course of 2 h, then the mixture was stirred overnight.
Solvent was removed, the crude was taken up in ethyl
acetate, and washed with water (3 9 100 mL). The organic
layer was dried over MgSO4, and solvent was removed. It
was then dissolved in 25 mL MeOH and (N,N-
dimethyl)methyleneammonium iodide (1.6 g, 8.7 mmol)
was added and heated to reflux for 8 h, then stirred over-
night at room temperature. Solvent was removed on the
Fig. 1 Synthetic conditions for adding an alpha-methylene unit to a
FA chain. The open arrow indicates the carbon that was initially the
COOH of the FA. See text for abbreviations
123

J Am Oil Chem Soc
rotary evaporator, and the residue was taken up in 4:1
hexane/ethyl acetate and applied to a column of silica gel.
Yields ranged from 75 to 91 % over six separate prepara-
as mono-methyl esters at the 1-COOH). These compounds
were separated on silica gel using 4:1 hexanes/ethyl acetate
to elute the less polar ketone, and ethyl acetate to elute the
mono-acid (0.5 g), which was then subjected to the general
acrylation conditions above.
1
tions. 1a: H NMR: 0.87 (t, 7.2 Hz, 3H, CH3), 1.37 br s,
1.47–1.75, 1.98–2.18 (m, 4H, allylic H), 2.37 (t, 7.4 Hz,
2H, CH2C=CH2), 3.84 (s, 3H, Me ester), 5.33–5.54 (m, 2H,
olefin), 5.62 and 6.21 (each s, 2 9 1H, =CH2). 13C NMR:
14.0, 22.6, 27.1, 28.5, 29.2–29.7, 31.8, 51.7, 124.2, 127.7,
127.9, 140.8, 167.6. Calculated for C22H40O2ꢀNa? 359.3,
found 359.6.
General Conditions for Michael Additions (Fig. 2)
To a solution of the acrylate fatty ester (on the scale of
several 100 mg) in 5 mL methanol was added 20 mol %
each of 4-dimethylamino pyridine and 1,8-diaza-bicy-
clo[5.4.0]undec-7-ene. The nucleophile was then added in
twofold excess for thiols or threefold excess for amines and
stirred under nitrogen. For compound 7 the ratio was 2.2 eq
FA to 1 eq dithiothreitol. The solution was refluxed for
amines and kept at room temperature for thiols. Disap-
pearance of the starting acrylate was monitored by TLC.
Subsequent in situ derivatization was performed as follows.
For cysteine derivative 5, 2 eq Boc anhydride was added.
After 2 h, oleyl amine (2 eq) and EDC (3 eq) were added,
and the mixture stirred overnight. For thiosuccinate 6,
solvent was removed, then 4 eq p-MeO-benzyl alcohol and
pivalic anhydride (6 eq) [11] were added in 20 mL THF
and stirred overnight. For the double adduct 8, the product
of the first Michael addition was isolated from unreacted
starting material by chromatography on silica gel, then the
second Michael addition (with a different nucleophile) was
performed. For amine 9, the solution was cooled to room
temperature and Boc anhydride (3.5 eq) was added and
stirred for 2 h. Methanol solvent was removed, 25 mL THF
added, then octanoic acid (4 eq) and pivalic anhydride
(6 eq) were added and the reaction was stirred overnight.
Solvent was removed on the rotary evaporator, the crude
was taken up in ethyl acetate, and extracted with 10 % w/v
citric acid, then water, and the solvent again removed. The
crude was then taken up in 2:1 hexane/ethyl acetate and
passed through a short plug (20 g) basic alumina to remove
excess octanoic acid. Finally, the material was chromato-
1b: 1H NMR: 0.87 (t, 7.2 Hz, 3H, CH3), 1.37 br s,
1.47–1.79, 1.56 (s, 9H, t-Bu ester), 1.95–2.16 (m, 4H,
allylic H), 2.38 (t, 7.3 Hz, 2H, CH2C=CH2), 5.28–5.54 (m,
2H, olefin), 5.50 and 6.13 (each s, 2 9 1H, =CH2). 13C
NMR: 14.0, 22.5, 27.2, 28.0, 28.5, 29.2–29.6, 31.9, 80.3,
123.1, 129.7, 129.9, 142.6, 166.7. Yield = 72 %. Calcu-
lated for C25H46O2ꢀNa? 401.3, found 401.5.
2a: 1H NMR: 0.88 (t, 7.1 Hz, 3H, CH3), 1.37 br s,
1.47–1.75, 2.37 (t, 7.3 Hz, 2H, CH2C=CH2), 3.60–3.73 (m,
1H, CH–OH), 3.84 (s, 3H, Me ester), 5.62 and 6.21 (each s,
2 9 1H, =CH2). 13C NMR: 14.0, 22.6, 25.6, 28.3,
29.3–29.5, 31.8, 37.4, 51.7, 71.9, 124.3, 140.8, 167.8.
Yield = 83 %. The starting hydroxy FA was prepared
according to [9] by formic acid addition to oleic acid.
Calculated for C22H42O3ꢀNa? 377.4, found 377.5.
2b: 1H NMR: 0.86 (t, 7.2 Hz, 3H, CH3), 1.37 br s,
1.96–2.21 (m, 4H, allylic H), 2.36 (t, 7.4 Hz, 2H, CH2-
C=CH2), 2.85 (t, 6.9 Hz, 2H, C(11)H2), 3.83 (s, 3H, Me
ester), 5.28–5.57 (m, 4H, olefin), 5.62 and 6.20 (each s,
2 9 1H, =CH2). 13C NMR: 14.0, 22.5, 25.5, 27.1, 28.3,
29.1–29.6, 31.4, 31.8, 51.6, 124.3, 127.9 (2 peaks), 130.1
(2 peaks), 140.8, 167.7. Yield = 88 %. The starting acid
for this product was linoleic acid (Sigma-Aldrich). Calcu-
lated for C22H38O2ꢀH? 335.3, found 335.5.
3: 1H NMR: 1.37 br s, 1.56 (s, 9H, t-Bu ester),
1.59–1.83, 2.24–2.42 (m, 4H, CH2C(O) and CH2C=CH2),
3.73 (s, 3H, Me ester), 5.49 and 6.09 (each s, 2 9 1H,
=CH2). 13C NMR: 24.9, 27.9, 28.6, 29.1–29.6, 31.8, 51.2,
80.1, 123.0, 142.6, 166.7, 174.2. Yield = 71 %. Calculated
for C26H48O4ꢀNa? 447.4, found 447.5. The starting mono-
methyl ester of 1,18 octadecanedioic acid was prepared as
follows, starting with H2SO4-catalyzed methanolysis of
15 g sophorolipids as previously described [10] to afford a
mixture of 17-OH oleates and 18-OH oleates (*7 g; the
18-OH variant comprises about 10 % of the whole). These
esters were then hydrogenated in methanol (50 mL) with
H2 over 5 % Pd/C for 4 h. After removal of the catalyst and
removal of solvent on the rotary evaporator, the crude
mixture was dissolved in 15 mL dimethylformamide and
treated with 12 g pyridinium dichromate at rt overnight.
Water (200 mL) was added, and the mixture was extracted
with diethyl ether (3 9 50 mL), which after rotary evap-
oration gave the 17-keto and 18-COOH compounds (both
1
graphed on silica gel to yield 9. H and 13C NMR, yields,
and MS:
4a: 0.87 (t, 7.2 Hz, 6H, 29 CH3), 1.35 br s, 1.55–1.82,
1.97–2.18 (m, 4H, allylic H), 2.48–2.97 (m, 5H, CH–CH2–
S–CH2), 3.80 (s, 3H, Me ester), 5.33–5.54 (m, 2H, olefin);
14.0, 22.6, 27.1, 28.8, 29.1–29.7, 31.7, 31.8, 32.0, 32.5,
34.0, 46.1, 51.5, 129.7, 129.9, 175.5. Yield = 93 %. Cal-
culated for C30H58O2SꢀNa? 505.4, found 505.7.
4b: 0.86 (t, 7.3 Hz, CH3), 1.38 br s, 1.56–1.80,
1.95–2.18 (m, 4H, allylic), 2.49–2.92 (m, 5H, CH–CH2–S–
CH2), 3.73 (t, 6.3 Hz, 2H, CH2OH), 3.80 (s, 3H, Me ester),
5.31–5.52 (m, 2H, olefin); 14.0, 22.6, 25.5, 27.1, 28.5,
29.2–29.7, 31.8, 32.0, 32.4, 34.0, 46.2, 51.5, 62.7, 129.7,
129.9, 174.8. Yield = 90 %. Calculated for C28H54O3-
SꢀNa? 493.4, found 493.5.
123

J Am Oil Chem Soc
Fig. 2 Adducts formed by nucleophilic addition of thiols or amines to the alpha-methylene
5: 0.86 (t, 7.1 Hz, 6H, 29 CH3), 1.14–1.47 br s, 1.53 (s,
9H, Boc tBu), 1.94–2.17 (m, 8H, allylic H), 2.63–3.09 (m
5H, CH–CH2–S–CH2), 3.35 (q, 6.8 Hz, 2H, C(O)NHCH2),
3.79 (s, 3H, Me ester), 4.20–4.35 (m, 1H, cysteine alpha
CH), 5.33–5.54 (m, 5H, olefin ? Boc NH), 6.40–6.55 (m,
1H, amide NH); 14.0, 22.6, 26.8, 27.1, 28.2, 29.2–29.7, 31.8,
32.3, 32.5, 33.5, 34.7, 35.1, 39.6, 45.9, 46.3, 51.6, 54.3, 80.3,
127.7, 127.9, 155.0, 170.0, 175.1. Yield = 80 %. Calcu-
lated for C48H90N2O5SꢀH? 807.7, found 807.8.
7: 0.87 (t, 6H, CH3), 1.35 br s, 1.54–1.80, 1.96–2.18 (m,
8H, allylic), 2.59–2.98 (m, 10H, 29 CH–CH2–S–CH2),
3.71–3.84 (m, 2H, 29 CH–OH), 3.80 (s, 6H, 29 Me ester)
5.34–5.52 (m, 4H, olefin); 14.0, 22.6, 27.1, 29.2–29.7, 31.8,
32.1, 34.1, 36.6, 46.2, 51.7, 70.8, 129.7, 129.9, 175.2.
Yield = 89 % based on dithiothreitol. Calculated for C48-
H90O6S2ꢀNa? 849.6, found 849.7.
8: 0.96 (t, 7.2 Hz, 3H, CH3), 1.33 br s, 1.42–1.55,
1.55–1.77, 2.50–2.92 (m, 10H, 29 CH–CH2–S–CH2), 3.73
(t, 6.5 Hz, 2H, CH2–OH), 3.79 (s, 6H, 29 Me ester); 14.0,
22.6, 25.2, 27.1, 28.4, 28.8, 29.1–29.5, 31.7, 32.0, 32.4,
32.5, 34.0, 46.1, 51.5, 62.8, 175.6. Yield = 81 %. The
starting diacid for this product was azelaic acid (Sigma-
Aldrich); the di-acrylic precursor was prepared entirely
analogously to 1a as described above, except that the molar
ratios of the reagents (Meldrum’s acid etc.) were doubled to
allow for reaction with both of the terminal COOH groups.
Calculated for C31H60O5S2ꢀNa? 599.4, found 599.5.
6: 0.86 (t, 7.3 Hz, CH3), 1.14–1.55 br s, 1.56 (s, 9H, t-
Bu ester), 1.60–1.80, 1.95–2.18 (m, 4H, allylic), 2.53–3.02
(m, 5H, CH–CH2–S and succinate CH2), 3.84 (m, 1H, S–
CH–C(O)), 5.22 (br s, 4H, 29 CH2-Ar), 5.31–5.52 (m, 2H,
olefin) 6.92 (d, 6.6 Hz, 4H, Ar), 7.15 (d, 6.5 Hz, 4H, Ar);
14.0, 22.6, 27.1, 29.2–29.7, 31.8, 32.1, 33.7, 34.3, 43.8,
45.5, 55.2, 65.8, 66.0, 80.2, 113.9, 129.7, 129.9, 132.1,
133.1, 159.6, 170.4, 177.4. Yield = 69 %. Calculated for
C45H68O8SꢀNa? 791.4, found 791.5
123

J Am Oil Chem Soc
9: 0.86 (t, 7.4 Hz, 9H, 39 CH3), 1.35 br s, 1.52 (s, 9H,
Boc tBu), 1.60–1.80, 1.94–2.18 (m, 4H, allylic), 2.25–2.46
(m, 4H, 29 CH2C(O)), 2.78–2.98 (m, 1H, C(O)CH),
3.05–3.66 (m, 4H, CH2–N–CH2), 3.87 (s, 3H, Ar-OMe),
3.95–4.38 (m, 2H, aminopropyl CH2O), 5.12 (br s, 2H,
CH2-Ar), 5.22–5.39 (m, 1H, aminopropyl CH), 5.39–5.51
(m, 2H, olefin), 6.95 (d, 6.5 Hz, 2H, Ar), 7.38 (d, 6.5 Hz,
2H, Ar); 14.0, 22.6, 27.1, 28.2, 29.2–29.7, 30.2, 31.8, 45.2,
50.7, 51.7, 55.2, 63.4, 70.7, 71.0, 80.9, 113.9, 129.7, 129.8,
132.1, 133.1, 157.3, 159.4, 175.6, 176.2. Yield = 66 %.
Calculated for C53H91NO9ꢀNa? 908.7, found 908.8.
10: 0.85 (t, 7.2 Hz, 3H, CH3), 1.35 br s, 1.56 (s, 9H, Boc
t-Bu), 1.60–1.80, 1.96–2.20 (m, 4H, allylic), 2.53–2.88 (m,
7H, CH–CH2–N–(CH2)2), 3.20–3.44 (m, 4H, (CH2)2N-Boc),
3.80 (s, 3H, Me ester), 5.34–5.48 (m, 2H, olefin); 14.0, 22.6,
27.1, 27.4, 28.3, 29.2–29.6, 30.6, 31.8, 43.9, 51.3, 52.9, 60.6,
79.6, 129.7, 129.9, 154.8, 175.9. Yield = 72 %. Calculated
for C31H58N2O4ꢀH? 523.4, found 523.6.
Reduction and Dihydroxylation to Give Triol 11
(Fig. 3)
Fatty acrylic ester 1a (0.55 g, 1.6 mmol), polymethylhy-
drosiloxane (0.25 g, 4 mmol SiH units) and Ti(OiPr)4
(1.6 mmol) were placed in a round-bottom flask equipped
with a drierite tube and heated to 65–70 °C under nitrogen
with magnetic stirring for 4 h [12]. After the reaction was
cooled to room temperature, 20 mL THF was added, fol-
lowed by the slow addition (portion-wise over 10 min) of
20 mL 4 M aqueous NaOH. The mixture was stirred
overnight, then extracted with 100 mL diethyl ether. After
removal of solvent, the material was passed through a short
plug of silica gel with 3:1 hexane/ethyl acetate. The pri-
mary alcohol thus obtained (0.45 g, 1.5 mmol, 93 %) was
suspended in 10 mL 9:1 t-BuOH/water and stirred vigor-
ously in an ice bath, then AD-mix-a (1.5 g, 0.75 eq based
on the recommended 1.4 g/mmol of olefin [13]) was added.
The mixture was stirred overnight, warming to room
temperature. Water (100 mL) was added and the mixture
was then extracted with 100 mL ethyl acetate, organic
solvent was removed, and the crude purified by chroma-
tography on silica gel (0.34 g, 88 % based on limiting
reagent AD-mix). 1H NMR: 0.87 (t, 7.1 Hz, 3H, CH3), 1.27
br s, 1.38–1.45, 1.95–2.18 (m, 4H, allylic), 3.62 (dd, 38.9
and 11.1 Hz, 4H, 29 CH2OH), 5.30–5.45 (m, 2H, olefin).
13C NMR: 14.0, 22.6, 22.9, 27.1, 29.2–29.7, 30.2, 31.8,
34.7, 67.5, 73.7, 129.7, 129.9. Calculated for C21H42O3-
Na? 365.3, found 365.6.
Fig. 3 Oxidation and reduction reactions used to alter compound 1.
See text for abbreviations
metal (0.18 g turnings, 7.6 mmol, 4 eq) over the course of
2 h [14]. A few minutes after addition of the first piece, faint
bubbling was observed arising from the surface of the metal.
The metal gradually became frayed and ultimately dis-
solved, with the appearance of a white cloudiness to the
solution. The solution was stirred for 18 h, then an aliquot
was taken, quenched with acetic acid, and solvent was
removed under a stream of N2. NMR showed that
*20–25 % of the starting material remained, so another
4 eq of Mg was added as above and the mixture stirred
overnight. Acetic acid (5 mL) was added dropwise, solvent
was removed, and the mixture was extracted with 50 mL
ethyl acetate vs. water (3 9 100 mL). Organic solvent was
removed, and the residue was passed through a short column
of silica gel with 3:1 hexane/ethyl acetate to afford 12 (0.6 g,
94 %). 1H NMR: 0.87 (t, 7.3 Hz, 3H, terminal CH3), 1.23 (d,
7.2 Hz, 3H, alpha CH3), 1.28–1.46, 1.60–1.82, 1.92–2.21
(m, 4H, allylic), 2.52 (hex, 7.0 Hz, CHC(O)), 3.74 (s, 3H,
Me ester), 5.32–5.66 (m, 2H, olefin). 13C NMR: 14.0, 17.0,
22.6, 27.1, 29.2–29.7, 31.8, 33.7, 39.4, 51.3, 129.7, 129.9,
177.4. Calculated for C22H42O2ꢀNa? 361.3, found 361.5.
Reduction of Alpha-Methylene
To a solution of 1a (0.63 g, 1.9 mmol) in 10 mL MeOH,
stirred under nitrogen at room temperature, was added Mg
123

J Am Oil Chem Soc
Results and Discussion
important though is the similarity of 3 to another important
bio-derived building block, itaconic acid [15]. Molecule 3
is one member of a series of fatty analogues where chain
length could be varied to control materials properties (glass
transitions, etc.) while still permitting radical-mediated
crosslinking or co-polymerization through the reactive
alpha-methylene.
To prepare fatty acrylic acids, we employed a literature
sequence (Fig. 1) that pairs malonate chemistry with the
Mannich reaction [7]. We made the modification of
replacing dicyclohexylcarbodiimide, which can be difficult
to handle, with the water-soluble, easily removed EDC (N-
(3-dimethylaminopropyl)-N0-ethylcarbodiimide hydrochlo-
ride). This alteration appeared to cause no adverse effects,
since the yield of the addition of Meldrum’s acid (2,2-
dimethyl-1,3-dioxane-4,6-dione), a carbon nucleophile, to
the FA was virtually quantitative. The second step, addition
of the malonate carbon to the electrophilic methylene
group in ‘‘Eschenmoser’s salt’’ ((N,N-dimethyl)methyle-
neammonium iodide), with concomitant decarboxylation,
went in good but not excellent yield: several repetitions of
the preparation of 1a ranged from 75 to 91 % based on
starting oleic acid. While these yields were more than
adequate for these initial studies, further optimization
might include rigorous exclusion of water, since the salt is
hygroscopic. It is also expensive, so using it in large
excesses was impractical. Longer reaction times might also
help if incomplete decarboxylation of the malonate inter-
mediate is occurring.
Michael Additions
To start our exploration of Michael additions at the fatty
acrylate, we chose to use thiols on the expectation that they
would be good nucleophiles that were likely to work. For
the first test of the concept we used a simple alkyl thiol,
1-octanethiol. Literature suggested that a helper nucleo-
phile such as a tertiary amine would be required [16], so a
mixture of 4-dimethyl pyridine (DMAP) and 1,8-diaza-
bicyclo[5.4.0]undec-7-ene (DBU), both at 20 mol %, was
included for all Michael additions. We performed the
reaction under nitrogen to minimize the possibility of
disulfide formation. Progress of the reaction was easily
monitored by TLC; with S-nucleophiles a complete loss of
starting material 1a was always observed. A more inter-
esting thiol was 6-hydroxy-1-hexane thiol. To examine the
feasibility of using the introduced OH for further reactions,
we hydrolyzed the methyl ester of 4b and performed self-
condensation of the resulting hydroxy acid (using the
pivalic anhydride esterification route from [11]). MS and
NMR showed the presence of oligomeric species, namely
oligoesters with a sulfide group along the backbone and
pendant oleic units. It is worth pointing out the structural
similarity between molecules 4a and 4b and Guerbet acids.
The latter are useful in constructing branched molecules
which have good low-temperature fluidity properties [17],
but they are not available with extra functional groups.
Conjugate addition to fatty acrylates such as 1 by contrast
does afford functionalized branched acids. As long as a
specific application of Guerbet acids can tolerate the pre-
sence of a sulfur atom, the options for utilizing this kind of
FA-based structure should therefore be enhanced.
This sequence of reactions should be easily applicable to
essentially any FA to elongate it by two carbons in the
chain (three counting the branching a-methylene). The
only other functional group that might occasionally be
encountered in FA molecules, a hydroxy unit, was not
discussed in the original paper [7], so we were interested to
see how such a substrate would fare. We used the mid-
chain hydroxy FA obtained by hydration of oleic acid in its
unprotected form [9], and the reaction worked with no
complications to give 2a, although to avoid possible
intermolecular ester formation in the carbodiimide step we
used an excess of Meldrum’s acid. Another familiar motif
in FA chemistry, the 9,12 diene pattern, was also com-
patible with these reaction steps, giving 2b, the linoleic
analogue. No alteration to a conjugated diene was
observed. In the unlikely but possible event that there were
a nitrogen present in the FA, it would need to be protected,
e.g. as a Boc derivative. Borohydride reduction does not
affect the isolated mid-chain olefin, unsurprisingly.
According to the original literature, other alcohols besides
methanol can be used for esterification in the Eschenmoser
salt/decarboxylation step. The only other one we tried was
t-butanol, which gave somewhat lower yield. Carboxylic
esters at other sites in the substrate are unchanged by these
conditions, as demonstrated by the successful preparation
of the a,x-diester 3 possessing an acrylic group only at one
end. Since long-chain diacids have found use as monomers
for polyesters and polyamides, this route would provide
new building blocks for those materials. Perhaps more
Other thiol nucleophiles can be used to introduce a
denser set of functionality. Cysteine was an obvious choice
to employ since it is bio-based; addition to the methylene
was complete in 1 h, so the proximity of the carboxylic
group was apparently not a problem. Attempting to purify
the zwitterionic adduct would have been difficult, so
instead we derivatized it in situ (with the DMAP/DBU
combination still present) by first adding Boc anhydride to
protect the NH2, then amidating the cysteine COOH with
oleyl amine and EDC to give 5 (Fig. 2). A small thiol that
can introduce two carboxylic units is thiosuccinic acid. For
this example we used the tert-butyl ester of the fatty
acrylate 1b, and addition was slower, presumably because
123

J Am Oil Chem Soc
of steric hindrance: after 24 h, even with excess thiol,
unreacted starting material was still present. Again, deriv-
atization was done in situ to make isolation easier, in this
case by di-esterifying with para-methoxybenzyl alcohol.
This protecting group was chosen because like the tert-
butyl group in 6 it can be removed with acid, which in this
case would reveal a fatty tri-acid. Finally, a commonly
used reagent in biochemistry, dithiothreitol, allows con-
struction of some interesting dimeric FA structures. Mol-
ecule 7, where the two hydroxy groups remain available to
attach other polar headgroups if desired, may be a useful
starting point for design of new gemini surfactants.
we tried was mercapto salicylic acid (2-thio benzoic acid)
and this example was totally unsuccessful. Steric and
electronic effects of the COOH group (which unlike in the
cysteine or thiosuccinic cases is rigidly oriented next to the
SH) may be the cause, but thiophenols in general might just
be too poorly nucleophilic. We chose not to try plain
thiophenol, since no extra functional groups beyond the
aromatic unit would be introduced, but possibly other
examples such as the commercially available para-thio
aniline might be worth a try.
Interesting structural possibilities are created by apply-
ing the acrylation and Michael addition to a long chain di-
acid. Azelaic acid gives the symmetrical di-acrylic acid,
which can then be desymmetrized by sequential reaction
One class of thiols that we had hoped to be able to apply
to the Michael additions were aromatic ones. The only one
1
Fig. 4 Representative H-NMR
spectra for selected Michael
adducts. Top the mono-adduct
formed after addition of
6-mercapto 1-hexanol to the
diacrylic derivative of azelaic
acid, showing one acrylic unit
remaining, and the di-adduct
(compound 8) formed after
addition of 1-octanethiol.
Bottom compound 9
123

J Am Oil Chem Soc
with two different thiols. We used a substoichiometric
amount of the first thiol, isolated the mono-adduct with one
acrylic unit remaining (see Fig. 4), and then performed an
addition with an excess of the second thiol to give 8. This
sequence of reactions and the subsequent separations were
quite easy to accomplish, suggesting that differentially
substituted a,x-di-FA will be readily available for pre-
paring new kinds of polymers.
commercially available. Furthermore, if a fatty analogue
with no internal unsaturation is used, the reaction scheme
becomes even simpler. The 1,2,3-triol obtained (11) is an
interesting new merger of glycerol with a fatty tail, minus
the oxidatively unstable hydrogen at the 2-position. As such,
it is an analogue of trimethylol propane (TMP), which is in
wide use in ester lubricants. This combination of reac-
tions—FA to fatty acrylate, then reduce to alcohol and di-
hydroxylate—provides a straightforward way to make a
whole range of FA-derived TMP analogues with different
length chains so that properties can be fine-tuned.
By contrast, nitrogen nucleophiles were sluggish at
Michael addition. We used amino propanediol with the
intention that it could lead to structures analogous to tria-
cylglycerols. The first attempt was with the tert-butyl ester
1b. After 24 h of reaction at room temperature, TLC
showed much fatty acrylate remaining, while MS indicated
that none of the desired adduct had formed. Refluxing the
reaction for 16 h did not improve matters; sterics are again
presumably responsible. Since we desired a combination of
protecting groups on the nitrogen and the carboxylic unit
that would be removable with acid, we re-tried the reaction
using the 4-methoxy benzyl ester variant of 1. This time,
MS after 16 h of reflux did show the desired adduct. As
with the cysteine adduct, derivatization was performed
in situ (attaching the Boc group, then esterifying the two
hydroxy groups) to aid in purification, affording 9. The
piperazine adduct 10 likewise required reflux for its prep-
aration. These molecules are lipophilic b-amino acids.
They should find use not only in design of new surfactants
by themselves, but by incorporation into peptides that may
be surface active, e.g. analogues of surfactin. They would
also be building blocks for new types of poly-amides
possessing a hydrophobic side chain.
We also sought to differentiate between the two olefins
using reduction chemistry. According to the literature, the
conjugated olefin should respond to some kinds of reduc-
tion agents that do not affect the internal, non-conjugated
one. Sodium dithionite is known to reduce conjugated
enones [18]; for its operation on molecule 1, NMR indi-
cated that a small amount of the desired compound may
have been formed, but interpretation and purification were
complicated by the need for a phase-transfer reagent in this
procedure. By contrast, Mg metal in MeOH [14] gave the
desired 2-methyl FA in good yield, roughly 75 % judging
by NMR, after one treatment. A second treatment was
necessary to convert the last trace of the 2-methylene
group. FA branched near to the carboxylate are believed to
be more resistant to hydrolysis than unbranched FA [19],
so 12 or analogues of it could enhance the range of con-
ditions where FA esters are used. If there were no internal
olefin, or if its reduction were immaterial, more forcing
conditions (high pressure H2) could be used to reduce the
2-methylene group. This reduction can even be performed
enantioselectively.
Oxidations and Reductions
The 2-fatty acrylates described here, easily prepared in
very good yield through a few simple operations, introduce
fertile new opportunities for utilizing FA. The single con-
jugated methylene that is added represents a multiplier
group through which a wide variety of functionality-laden
thiol or amine nucleophiles can be lipophilized.
The proximity of the methylene group and the carboxylic
group permit the synthesis of some intriguing new FA
variants without relying on nucleophilic additions. The
pathway to one such compound is shown in Fig. 3. We first
reduced the methyl ester to a primary alcohol using a very
convenient method reported by Buchwald, with Ti(OiPr)4
and poly(methylhydrosiloxane) [12]. Once the methylene
group had been deconjugated from the carbonyl it could be
used for other kinds of reactions. In line with our goal of
adding several functional groups to one end of the fatty
chain, we reasoned that dihydroxylation would be a pro-
ductive scheme to try. The presence of the internal olefin of
1 complicates the reactivity, but 1,1-disubstituted olefins are
known to be more susceptible to osmium dihydroxylation
reagents than 1,2-disubstituted olefins [13]. To minimize
reaction at the internal olefin, we used a substoichiometric
amount, about 75 mol %, of the Sharpless dihydroxylation
reagent. Note that the asymmetric aspect of this reagent is
superfluous here, but the reagent is conveniently
References
1. Huf S, Kru¨gener S, Hirth T, Rupp S, Zibek S (2011) Biotech-
nological synthesis of long-chain dicarboxylic acids as building
blocks for polymers. Eur J Lipid Sci Technol 113:548–561
2. Hwang H-S, Erhan SZ (2001) Highly selective asymmetric syn-
thesis of 2-hydroxy fatty acid methyl esters through chiral oxa-
zolidinone carboximides. JAOCS 78:205–211
´
´
`
´
3. Teres S, Llado V, Higuera M, Barcelo-Coblijn G, Martin ML,
Noguera-Salva MA, Marcilla-Etxenike A, Garcıa-Verdugo JM,
´
´
Soriano-Navarro M, Saus C, Gomez-Pinedo U, Busquets X,
Escriba PV (2012) 2-Hydroxyoleate, a nontoxic membrane
´
binding anticancer drug, induces glioma cell differentiation and
autophagy. Proc Natl Acad Sci 109:8489–8494
123

J Am Oil Chem Soc
´
¨
4. Cordomı A, Prades J, Frau J, Vogler O, Funari SS, Perez JJ,
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
´
´
Escriba PV, Barcelo F (2010) Interactions of fatty acids with
phosphatidylethanolamine membranes: X-ray diffraction and
molecular dynamics studies. J Lipid Res 51:1113–1124
13. Sharpless KB, Amberg W, Bennani YL, Crispino GA, Hartung J,
Jeong KS, Kwong HL, Morikawa K, Wang ZM (1992) The
osmium-catalyzed asymmetric dihydroxylation: a new ligand
class and a process improvement. J Org Chem 57:2768–2771
14. Gabriels S, Van Haver D, Vandewalle M, De Clercq P, Viterbo D
(1999) On the unexpected stereochemical outcome of the mag-
nesium in methanol-conjugate reduction of an exocyclic a, b-
unsaturated ester. Eur J Org Chem 1999:1803–1809
15. Okuda T, Ishimoto K, Ohara H, Kobayashi S (2012) Renewable
biobased polymeric materials: facile synthesis of itaconic anhy-
dride-based copolymers with poly (L-lactic acid) grafts. Macro-
molecules 45:4166–4174
5. Magrioti V, Verger R, Constantinou-Kokotou V (2004) Triacyl-
glycerols based on 2-(N-tert-butoxycarbonylamino)oleic acid are
potent inhibitors of pancreatic lipase. J Med Chem 47:288–291
´
´
6. Torres CF, Martın D, Torrelo G, Casado V, Fernandez O, Tenl-
´
´
lado D, Vazquez L, Moran-Valero MI, Reglero G (2011) Lipids
as delivery systems to improve the biological activity of bioactive
ingredients. Curr Nutr Food Sci 7:160–169
7. Hin B, Majer P, Tsukamoto T (2002) Facile synthesis of a-
substituted acrylate esters. J Org Chem 67:7365–7368
8. Best MD, Rowland MM, Bostic HE (2011) Exploiting bioor-
thogonal chemistry to elucidate protein-lipid binding interactions
and other biological roles of phospholipids. Acc Chem Res
44:686–698
16. Hoyle CE, Bowman CN (2010) Thiol–ene click chemistry. An-
gew Chem Int Ed 49:1540–1573
9. Knight HB, Koos RE, Swern D (1953) Addition of formic acid to
olefinic compounds. I. Monoolefinic compounds. J Am Chem Soc
75:6212–6215
10. Zerkowski JA, Solaiman DKY (2006) Synthesis of polyfunc-
tional fatty amines from sophorolipid-derived 17-hydroxy oleic
acid. JAOCS 83:621–628
11. Sakakura A, Kawajiri K, Ohkubo T, Kosugi Y, Ishihara K (2007)
Widely useful DMAP-catalyzed esterification under auxiliary base-
and solvent-free conditions. J Am Chem Soc 129:14775–14779
17. O’Lenick AJ Jr (2001) Guerbet chemistry. J Surfactants Deterg
4:311–315
18. Camps F, Coll J, Guitart J (1986) Regiospecific reduction of
unsaturated conjugated ketones with sodium dithionite under
phase transfer catalysis. Tetrahedron 42:4603–4609
19. Tai A, Kawasaki D, Sasaki K, Gohda E, Yamamoto I (2003)
Synthesis and characterization of 6-O-acyl-2-O-a-D-glucopyran-
osyl-L-ascorbic acids with a branched-acyl chain. Chem Pharm
Bull 51:175–180
123