J Am Oil Chem Soc (2010) 87:937–945

DOI 10.1007/s11746-010-1574-1

ORIGINAL PAPER

Synthesis and Physical Properties of Potential Biolubricants

based on Ricinoleic Acid
Linxing Yao • Earl G. Hammond • Tong Wang •

Satyam Bhuyan • Sriram Sundararajan

Received: 31 December 2009 / Revised: 3 March 2010 / Accepted: 10 March 2010 / Published online: 31 March 2010
Ó AOCS 2010

Abstract The desire to replace petroleum-based materi- recently because of environmental concerns and the highly
als with environmentally friendly and sustainable alterna- variable price of crude petroleum. More than 60% of the
tives has stimulated the development of vegetable oil-based lubricants currently used in the United States are reported
materials as biolubricants. Our studies have focused on to become pollutants through spillage, evaporation, and
molecules that might be produced by biosynthesis of loss to the environment [1]. Lubricants that are produced
genetically altered oilseed plants with limited post-harvest from natural fats and oils can be biodegradable and non-
modification. Various ricinoleate and 12-hydroxystearate toxic as well as reducing America’s dependence on foreign
esters and estolides were synthesized and their melting oil. But some properties of conventional vegetable oils
points and viscosities were documented. The antifriction limit their use as biolubricants even though they can have
and antiwear properties of some esters were evaluated with excellent lubricity and viscosity indices [2]. Long-chain
a microtribometer. The purities of all the products were fatty acids have high melting points if saturated or poor
[98–99% by gas chromatography. Some of these com- oxidative stability if polyunsaturated, and these traits are
pounds showed melting points, viscosities, and lubricity particularly undesirable in lubricant applications. Branched
suitable for uses as biolubricants. Various ricinoleate esters and monounsaturated esters, such as oleates, can exhibit
acylated at the 12 positions with short-chain acids were acceptable oxidative stability and have melting points
particularly promising. acceptable for many lubricant applications [3–5].
Although chemical modifications of conventional veg-
Keywords 12-Acetylricinoleate esters  Friction  etable oils can improve their performance in lubrication
12-Hydroxystearate  Melting point  Ricinoleate  application [1, 2], fatty acid derivatives with desirable
Viscosity  Wear lubrication properties might be produced in oilseeds that
have been modified by genetic engineering techniques
[6, 7]. Potentially, the latter approach could bring about
Introduction major savings [Dowling A, Zakharenko O (2005) Assessing
biodiesel’s cold flow limitation and gains from re-engineered
The use of vegetable oil-based lubricants to replace those soybeans, Unpublished data].
based on petroleum has received significant attention The development of biolubricants based on estolides
derivatives from castor, lesquerella, and meadowfoam oils
through chemical modifications has been studied thor-
L. Yao  E. G. Hammond (&)  T. Wang
oughly [8–10]. While only a limited number of plant
Department of Food Science and Human Nutrition,
Center for Crops Utilization Research, Iowa State University, species synthesize oxygenated fatty acids, such as hydroxy
2312 Food Science Building, Ames, IA 50011, USA and epoxy acids, in their seed oils, all higher plants pro-
e-mail: hammond@iastate.edu duce a variety of oxygenated fatty acids and fatty acid
derivatives during the synthesis of surface waxes. These
S. Bhuyan  S. Sundararajan
Department of Mechanical Engineering, Iowa State University, plant surface waxes contain a variety of hydroxy and
2025 Black Engineering, Ames, IA 50011, USA epoxy fatty acids. Li and Beisson [11] have reviewed the

123
938 J Am Oil Chem Soc (2010) 87:937–945

use of plant surface waxes as possible sources of high 200 mL of hexane and washed with 100 mL of methanol–
value-added lipids. water–acetic acid (90:10:1, by vol). The methanol-rich
The general goal of our work has been to design and phase was transferred to a second funnel and washed with
identify molecules that might be biosynthesized in oilseed 200 mL of fresh hexane. After being washed with another
crops and that would possess properties suitable for lubri- 200 mL of fresh hexane in a third funnel, the methanol-rich
cation, including excellent low temperature properties, phase was collected as fraction 1. The MeR was equili-
suitable viscosity, great stability to oxidation, and good brated successively with the same hexane washes until 12
lubricity. In our work, we have measured the physical fractions of the methanol–water–acetic acid phase had been
properties of quite pure examples of these compounds to collected. Then 1 mL of each of the 12 methanol fractions
illustrate the best properties that could be achieved, was partitioned between hexane (2 mL) and water (5 mL),
although it is unlikely that such high purity can be expected and the hexane layer was analyzed by gas chromatography
in a biosynthetic product without further fractionation. This (GC). Fractions with [98.5 GC % of MeR were combined
comparison of pure targets also facilitates comparisons and 1 part of the combined fractions was mixed with 1 part
among various candidates for the usually long genetic and of hexane and 3 parts of 2% aqueous acetic acid by vol-
biological process of molecular alteration. In this paper, we ume. The [98.5% pure MeR was now in the hexane layer.
explore the synthesis, melting, viscosity, and lubricity The hexane was removed and the MeR (about 16 g) was
properties of several pure ricinoleate and 12-hydroxystea- stored at -5 °C.
rate derivatives. Alternatively, about 3 g of crude castor oil FAMEs was
purified using a 30-g silica column so that non-hydroxy
FAMEs and MeR were eluted separately with hexane-
Materials and Methods diethyl ether at ratios of 95:5 (v/v) and 80:20 (v/v),
respectively. The purity of the resulting MeR was [99%
Materials (by GC). Less than 0.5% of methyl dihydroxystearate was
present in the MeR. MeR obtained from silica gel purifi-
Castor oil was purchased from a local pharmacy (Wal- cation was used in melting point determinations, whereas
green, Ames, IA). Potassium isopropylate (19% in iso- that obtained from countercurrent-solvent fractionations
propanol) and potassium propylate (20% in propanol) were was used in the synthesis of other derivatives and viscosity
purchased from BASF Corp. (Evans City, PA). Silica gel and lubricity measurements.
(40–140 Mesh) was from J. T. Baker (Phillipsburg, NJ). Ethyl, propyl, butyl, pentyl, and isopropyl ricinoleates
Light mineral oil was from Fisher Scientific (Pittsburgh, were synthesized via the same procedure as MeR in 5–8 g
PA). Other chemicals including 12-hydroxystearate, were amounts using the metal alkoxide corresponding to the
purchased from Sigma-Aldrich (St Louis, MO). esterifying alcohol. The catalysts for producing esters with
alcohols longer than propyl were generated by reacting
Synthesis and Purification of Ricinoleate Esters sodium metal with the appropriate alcohols [4]. The
with Free Hydroxy Groups resulting esters were purified with silica gel using the same
procedure used for MeR.
About 20 g of castor oil was dissolved in 300 mL of Enzymatic methods were used for the synthesis of
methanol containing 6 mL of 5 N sodium methoxide. The methyl and isopropyl 12-hydroxystearate. The reaction was
mixture was vigorously stirred overnight at room temper- carried out with a molar ratio of 12-hydroxystearic acid to
ature. The methyl esters were extracted with hexane and the alcohol of 50:1 and Novozyme 435 at 5% of the weight
washed with 2% aqueous acetic acid and with water, and of the alcohol for 24 h.
weighed about 18–20 g after solvent removal. The GC
response factor of methyl ricinoleate (MeR, see Table 1 for Synthesis and Purification of 12-Acylated Ricinoleate
a key to abbreviations) relative to methyl stearate was and 12-Acylated Hydroxystearate Esters with Alcohols
found to be 0.86. Applying this factor and assuming the of Various Chain Lengths
other fatty acids had the same response as methyl stearate,
GC showed that the resulting methyl esters contained Acetyl esters of various alcohols were used to place acetyl
88.0% MeR, 4.6% methyl linoleate, 3.2% methyl oleate, groups on the 12-position of ricinoleate in instances where
0.6% methyl palmitate, 0.6% dihydroxystearate, 0.3% the alcohol group on ricinoleate was also available as an ester
methyl eicosanoate and additional minor impurities. of acetic acid we wished to place on the 12 position of
MeR was freed of non-hydroxy FAMEs by a counter- ricinoleate. Thus, MeR (6 mM) was reacted with methyl
current fractionation method adapted from Berdeaux et al. acetate (300 mM) catalyzed by sodium methoxide (2% by
[12]. The crude castor oil FAMEs (30 g) was dissolved in weight of methyl acetate) for 3 h at 25 °C with stirring. The

123
J Am Oil Chem Soc (2010) 87:937–945 939

Table 1 Melting points and reaction conversion rate of various ricinoleate and 12-hydroxystearate esters
Compounds Melting point of unstable Tbcom (°C) Reaction
polymorphsa (°C) conversionc (%)

Methyl ricinoleate (MeR) -3.5 98

Ethyl ricinoleate (EtR) -15.7 98
Propyl ricinoleate (PrR) -19.4 98
Butyl ricinoleate (BuR) -35.5 -29.4 98
Pentyl ricinoleate (PeR) -22.2 98
Isopropyl ricinoleate (i-PrR) -50.6, -42.6 -20.3 98
Methyl 12-hydroxystearate (MeHS) 35.3, 52.6 59.4 95
Isopropyl 12-hydroxystearate (i-PrHS) 43.8 49.0 95
Methyl 12-acetylricinoleate (MeR-Ac) -50.0e 85(95)d
e
Ethyl 12-acetylricinoleate (EtR-Ac) -50.0 85
Propyl 12-acetylricinoleate (PrR-Ac) -50.0e 85
e
Butyl 12-acetylricinoleate (BuR-Ac) -50.0 85
Isopropyl 12-acetylricinoleate (i-PrR-Ac) -50.0e 85
Methyl 12-butyrylricinoleate (MeR-Bu) -50.0e 34
Isopropyl 12-butyrylricinoleate (i-PrR-Bu) -50.0e 25
Methyl 12-hexanoylricinoleate (MeR-Hx) -50.0e 95
Methyl 12-octanoylricinoleate (MeR-Oc) -50.0e 95
Methyl 12-decanoylricinoleate (MeR-10) -45.0, -33.5 -12.1 81(86)d
Methyl 12-undec-10-enylricinoleate (MeR-11u) -23.8 -18.7 80
Methyl 10-undecylenatef (C11:1 FAME) -19.8 99
Methyl 12-acetylstearate (MeHS-Ac) -24.0 -5.7 45
Methyl 12-butyryl stearate (MeHS-Bu) -14.1, -3.7 4.9 7
Isopropyl 12-acetylstearate (i-PrHS-Ac) -31.2 -9.8 71
Isopropyl 12-butyryl stearate (i-PrHS-Bu) -15.0 24
Abbreviations for each molecule are in parentheses
a
Melting points of polymorphic peak were the peak temperatures of the DSC curves
b
Tcom is the temperature at completion of melting
c
The conversion standard deviation were about 2
d
The value in the parenthesis is from the reaction catalyzed by DMAP and DCC; the value without parenthesis is from the reaction catalyzed by
sodium methoxide
e
Approximate gelling temperatures
f
Methyl 10-undecylenate (C11:1 FAME) was prepared by methylating a commercial sample of 10-undecenoic acid

product, methyl 12-acetylricinoleate (MeR-Ac), was washed reactions between MeR and acetic anhydride at a molar ratio of
with 2% aqueous acetic acid and water and extracted with 1:1.2:2.4 of MeR:acetic anhydride:pyridine according to
hexane. Excess methyl acetate was removed by distillation Sharma et al. [13]. The excess acid anhydride was removed
and the product was purified by chromatography on silica under vacuum after the reaction was complete. The residual
gel. The non-polar MeR-Ac was eluted with 5% diethyl ether mixture of MeR-Ac and unreacted MeR was separated by silica
in hexane, and the yield after purification was *67%. Other gel chromatography. The desired product was eluted with 5%
12-acetylricinoleate esters with longer alcohol chains, were diethyl ether in hexane. Similar synthesis strategies were
synthesized by the same procedure using the corresponding applied for methyl 12-butyrylricinoleate (MeR-Bu), isopropyl
acetate esters, appropriate ricinoleate ester and sodium alk- 12-butyrylricinoleate (i-PrR-Bu), methyl 12-acetylstearate
oxide. Ethyl 12-acetylricinoleate, for example, was obtained (MeHS-Ac), methyl 12-butyryl stearate (MeHS-Bu), isopropyl
from the reaction of ethyl acetate and ethyl ricinoleate and 12-acetylstearate (i-PrHS-Ac) and isopropyl 12-butyryl stea-
catalyzed by sodium ethoxide. rate (i-PrHS-Bu).
Traditional methods also were used to synthesize MeR- Methyl 12-hexanoylricinoleate (MeR-Hx), and methyl
Ac in order to avoid some of the side products that occurred 12-octanoylricinoleate (MeR-Oc) were synthesized by
with alkoxide catalysts. Pyridine was used to catalyze the esterification of MeR with hexanoic acid and octanoic acid,

123
940 J Am Oil Chem Soc (2010) 87:937–945

respectively, catalyzed by 4-dimethylaminopyridine and n-decane were described previously [4]. Visual
(DMAP) and 1,1-dicyclohexylcarbodiimide (DCC), fol- observations of samples cooled with dry ice-acetone were
lowing the procedure of Kodali et al. [14]. used to determine the approximate gelation temperature for
Methyl 12-decanoylricinoleate (MeR-10) was synthe- compounds that did not crystallize in the DSC.
sized by reacting methyl ricinoleate with methyl decanoate The dynamic viscosities of the various ricinoleate esters
at a molar ratio of 1:5, and catalyzed by 2% sodium in the temperature range -5 to 80 °C were determined with
methoxide. The excess methyl decanoate was removed by a Brookfield DV II ? viscometer (Brookfield Inc.,
distillation. GC analysis showed that the product contained Stoughton, MA) using a CP42 cone spindle. The instru-
*80% MeR-10 and small amounts of MeR (10%), methyl ment condition, temperature calibration, and procedure for
ricinoleate estolide dimer (MeR-Estolide, 2%) and methyl measurement were described previously [4]. The viscom-
12-decanoylricinoleate estolide dimer (MeR-Estolide-10, eter has an error of ±1% under the conditions used for our
8%). The polar constituents, MeR and MeR-Estolide were measurements.
removed by silica gel chromatography with a gradient The viscosities of 12-acetylricinoleate esters in the
elution of hexane/diethyl ether (95:5 to 80:20, v/v). The temperature range of -34 to 10 °C were determined with
non-polar MeR-Estolide-10 was removed by silver ion- the same viscometer and spindle. The temperature was
silica gel chromatography. controlled with a Brinkmann RK 20 refrigerated circulator
The preparation of silica gel containing 20–30% silver (Westbury, NY) and was monitored with a total immersion
nitrate for use in chromatography was adapted from a Fisherbrand ASTM thermometer (Fisher Scientific, Pitts-
procedure of Ghebreyessus et al. [15]. About 140 g silica burgh, PA) with the temperature of -38 to 2 °C, which
gel was added to a 300 mL of 22% aqueous solution of was inserted in a glass tube on the return line of the cir-
silver nitrate. The mixture was well stirred for 10 min and culating water bath. Data were collected between -34 and
dried at 110 °C for *5 h. The powder was protected from 0 °C at 5 °C intervals. Viscosities were recorded after
light. A gradient elution of 45–60% diethyl ether in hexane equilibrating at the desired temperature for 5 min and at the
washed out MeR-10 that had one double bond, and left maximum attainable torque.
MeR-Estolide-10, which had two double bonds, on the The density of MeR was measured by a pycnometer
column. with a 5-mL capacity (Ace Glass, Vineland, NJ) and water
The synthesis and purification of methyl 12-undec-10- was used to verify the exact volume of the pycnometer.
enoylricinoleate (MeR-11u) were similar to those for MeR- The oxidative stability index (OSI) of MeR-Ac was
10. A gradient elution of 85–90% ether/hexane collected measured using an ADM Oxidative Stability Instrument
MeR-11u from the silver ion-silica gel column. (Omnion, Rockland, MA) at 110 °C, according to AOCS
The reaction conversion percentages of various esters Official Method Cd 12b-92 [16].
are given in Table 1.
Friction and Wear Test
Instrumental Analysis
Friction and wear tests were performed on a custom-built
GC analysis was done on an HP 5890 Series II instrument ball-on-flat microtribometer, which has been previously
(Hewlett-Packard, PA) with a SPB-1 fused silica column described [17]. In both tests, a stainless steel ball rubbed
(15 m 9 0.25 mm 9 0.25 lm) (Supelco, Bellefonte, PA). against a polished stainless steel disk at specified speed for
The carrier gas (helium) flow rate was 1.9 mL/min, and the a specified time. The esters to be tested were mixed with
split ratio was 2.64. The injector and flame ionization light mineral oil at 10% (by wt).
detector were at 300 °C. The oven temperature was pro- Preliminary test had shown that the stainless steel disks
grammed from 100 to 300 °C at a rate of 10 °C/min, and were the major source of variance, so a balanced incom-
then held at 300 °C for 10 min. plete randomized block design was used in the friction
The structures of various esters in CDCl3 solution were tests. Twelve disks were treated as blocks, and each block
confirmed by 1H NMR with a Varian VXR-400 NMR was assigned randomly to three esters. Four repeated
spectrometer (Varian Inc., Palo Alto, CA) and compared measurements were performed for each ester on different
with chemical shifts predicted by ACD/ChemSketch Pre- disks and averaged. A ramped load test was used to obtain
dictor software (Advance Chemistry Development, Inc., the coefficient of friction (COF) in which the normal force
Toronto, Ontario, Canada). was increased from 0.2 to 1,000 mN while the probe was
Melting points of the synthesized esters were measured sliding against the disk for 40 mm at 1 mm/s. The repe-
with a DSC7 differential scanning calorimeter equipped ated measurements of each ester were performed in a
with an Intracooler System I (Perkin Elmer, Norwalk, CT). parallel fashion and 0.9 mm apart. After each test, the
The temperature programs and calibrations with indium stainless steel disk was cleaned with soapy water, and then

123
J Am Oil Chem Soc (2010) 87:937–945 941

sonicated for 10 min in acetone before the next ester was NMR Results
tested.
A completely randomized block design was used for the The 1H-NMR results for all the purified compounds were
wear tests. Three disks were treated as blocks, and each compared with the NMR spectra predicted by ACD/
ester was tested once on each disk. Reciprocal sliding of ChemSketch Predictor software and in all instances they
the probe against the disk under 800 mN normal load over were in agreement with the relative size and location of the
a 30-mm distance for 500 cycles at 5 mm/s was used. Ten predicted peaks. There was no sign of peaks indicating
random locations on each wear track were selected, and extraneous impurities. Our results also agree with the
1
wear depths were measured with a contact profilmeter H-NMR spectra of MeR and MeR-Ac previously reported
Dektak IIA (Veeco Instruments, Santa Barbara, CA). [18, 19].
Data were fitted to a mixed linear model with SAS
(Version 9.1, SAS Institute Inc. Cary, NC, USA) to test Effects of Chemical Structures on Melting Points
significant difference at p = 0.05.
The melting points (completion of melting) of ricinoleate
and 12-hydroxystearate esters and acylated derivatives are
Results and Discussion shown in Table 1. Those who work with lubricants tradi-
tionally use parameters such as cloud point and pour point
Synthesis and Purification of Various Ricinoleate Esters to characterize low temperature properties. For relatively
pure compounds, such as those we are characterizing, the
During the counter-current purification of MeR, the yield cloud and pour points are approximately the same as their
of the MeR in the collected fractions increased from frac- melting point [20]. Among ricinoleate esters with a free
tions 1 through 3 and then slowly declined with successive hydroxy group, MeR had the highest melting point at
fraction to low values. The purity of the MeR markedly -3.5 °C, which is slightly higher than the literature value
increased from fraction 1 to 3 and remained above 99% [20]. BuR melted at the lowest melting temperature of
through fraction 11, declining to 98% in fraction 12. The -29.4 °C. For longer-chain esters of ricinoleate, the
use of small amount of acetic acid prevents the production melting point increased with chain length. Such V-shaped
of free fatty acids. This proved to be a convenient method plot of melting points versus chain length has been found
for obtaining pure MeR. The reaction conversions of for esters of oleate, stearate and for saturated free fatty
12-acetylricinoleate esters using the acetate esters transe- acids [4, 21]. References 4 and 21 provide data and spec-
sterification were about 85%. A small amount (\2% by ulation about the causes of this phenomenon. The melting
GC) of ricinoleate estolide dimer was inevitably produced point of i-PrR, which melts at -20.3 °C, is very similar to
during the synthesis of 12-acetylricinoleate esters by this that of the corresponding straight chain, PrR, which melted
method even though a large excess of the acetate esters at -19.4 °C. In the oleate ester series, isopropyl oleate
were added to drive the reaction to completion and sup- melted 6.2 °C lower than propyl oleate [4]. Presumably,
press the formation of ricinoleate estolide dimer. The MeR this difference is because the isopropyl group interferes
estolide dimer further reacted with the acetate esters to with the close packing of alkyl chains. The similar melting
form 1–2% of non-polar acylated MeR estolide dimer that points of i-PrR and PrR suggest that the hydroxy group of
could not be removed from the desired product by silica gel the ricinoleate dominates its crystal packing pattern and
chromatography. In spite of the presence of small amounts makes the effect of the isopropyl branch negligible.
of impurities, this method provided an effective way to Figure 1 shows that the melting points of the ricinoleate
obtain acetylated ricinoleates with satisfactory reaction esters of various chain lengths are slightly higher than
conversions. The acylation reactions catalyzed by DMAP those of the corresponding oleate esters. Similarly, the
generally gave higher conversion rate than those catalyzed reported melting point for 12-hydroxystearic acid, 76 °C
by pyridine. Acylation with short chains acids tended to [22], is slightly higher than that of stearic acid, 70 °C.
have higher conversion rate than long chain compounds These observations suggest that a polar hydroxy group
when using the same methods. Reactions between MeR positioned in the middle of a carbon chain slightly
and a medium chain length fatty esters, such as methyl increases the melting point of a fatty acid or its esters.
decanoate or methyl 10-undecylenate, had high conversion The ricinoleate esters whose 12-hydroxy groups were
([80%) when catalyzed by sodium methoxide, but silver acylated with short-chain acids having carbon number 2 to
ion-silica gel chromatography was not very efficient in 8, did not crystallize in the temperature range we measured
separating these compounds from small amounts of (-60 to 0 °C). Instead, they formed transparent gel-like
accompanying acylated ricinoleate estolide dimer. solids at about -50 to -55 °C.

123
942 J Am Oil Chem Soc (2010) 87:937–945

The melting points of ricinoleate esters with free

hydroxy groups and those acylated at the C-12 position by
medium-chain fatty acids, i.e. MeR-10 and MeR-11u,
ranged from -3.5 to -29.4 °C. The compounds in Table 1
that did not crystallize upon cooling to -50 °C included all
the 12-acylated ricinoleate esters with an acyl side chain of
eight carbons or less. These are promising candidates for
biolubricants; however, Knothe [20] has suggested that
ricinoleate ester may be considerably less stable to oxida-
tion than oleate esters. A preliminary study in our labora-
tory showed that for samples that had been stripped of
tocopherols, MeR-Ac had an OSI time longer than MeR.
Esters derived from 12-hydroxystearic acid had fairly high
melting points except for i-PrHS-Bu, which melted at
Fig. 1 Melting points of ricinoleate and oleate esters. The melting -15.0 °C, but they should have high stability because of
points of the oleate esters are from Ref. 4 the absence of double bonds.

In contrast to the shorter 12-acyl branches, the decanoyl Effect of Chemical Structures on Viscosities
and 10-undecenoyl derivatives of ricinoleates were able to
crystallize at low temperatures. Seemingly, an acyl group The dynamic viscosities between -5 and 80 °C of most of
that is a little longer than eight or nine carbons allows the products synthesized in this study are shown in Table 2.
crystallization. The melting points of MeR-10 and MeR- As usual, the viscosities of the compounds we examined
11u were -12.1 and -18.7 °C, respectively. Compared increased with carbon numbers and decreased with tem-
with MeR, which possesses a free hydroxy group, the perature. At temperatures below 40 °C, ricinoleate esters
melting points were lowered considerably by adding these with a free hydroxy group had higher viscosities than
longer branches on the 12-position of MeR. The double 12-acylated ricinoleate esters. Indeed, a free 12-hydroxy
bond on the 10-undecylenate moiety further depressed the group on MeR gave a higher viscosity than MeR acylated with
melting point. Possibly the isopropyl esters of these com- a 10 or 11-carbon chain. Above 40 °C, the differences in
pounds would have even lower melting points than their viscosity among these compounds became relatively small.
methyl ester counterparts. The viscosity at 40 °C of compounds in Table 2 are
Considering that there were small amounts of acylated considerably lower than biolubricants made from triacyl-
estolide esters presented in these 12-acylated ricinoleates, glycerol estolides, or oils with ether-linked side chains
we compared the melting profiles of MeR-Ac and MeR-10 chiefly because of the difference in molecular sizes [8].
of which the same compounds were alternatively synthe- Their viscosities were slightly lower than ricinoleate
sized with pyridine or DMAP. No differences were estolide dimers esterified with monohydroxy alcohols [9];
detected between the pure samples and those contaminated but comparable to other straight or branched-chain mono-
with acylated MeR estolide dimer. esters [2].
The low temperature properties of 12-hydroxystearate The viscosities of some 12-acetylricinoleate esters also
esters also were studied. As shown in Table 1, MeHS and were measured in the range of -10 to -34 °C, and the
i-PrHS melt at 59.4 and 49.0 °C, respectively. Esterifying results are shown in Table 3. Their viscosities at -34 °C
the 12-OH position with acetate or butyrate lowered the were not particularly high, which indicated that the
melting points by more than 50 °C for both the methyl and 12-acetylricinoleates still flow freely at -34 °C. Our
isopropyl esters. equipment was unable to achieve temperatures below
During DSC measurements, most of the compounds -34 °C. As expected, the longer the carbon chain of the
showed polymorphism. Changing the temperature program esterifying alcohol, the greater the viscosity of the ester.
and letting the compounds temper for 7–10 min at the The viscosity of isopropyl esters showed some inter-
temperature where the highest melting polymorph begins esting changes with temperature relative to other ricinole-
to melt, reduced or deleted the peaks of the low tempera- ate derivatives. In our previous work on branched esters
ture polymorphs. All the melting points reported in Table 1 [4], we found that the isopropyl ester gave viscosities
are from the polymorphic form with the highest melting similar to n-propyl esters. Of the ricinoleate esters shown in
temperature, which is the most stable form. The melting Table 2, the isopropyl ester had a higher viscosity than the
temperatures of other polymorphic forms that we encoun- n-propyl ester and butyl ester at -5 °C. The isopropyl ester
tered for each compound also are recorded. continued to have slightly higher viscosity than the

123
J Am Oil Chem Soc (2010) 87:937–945 943

Table 2 Dynamic viscosities of various ricinoleate esters

Temperature (°C) Viscosity (cp)
MeRa EtR PrR BuR i-PrR MeR-Ac EtR-Ac PrR-Ac BuR-Ac i-PrR-Ac MeR-10 MeR-11u

-5 130.8 145.6 162.3 171.6 182.6 55.0 59.2 66.4 68.9 73.7 65.4 96.8
0 95.1 105.1 117.1 124.1 130.2 40.9 45.4 50.0 52.4 54.8 50.4 76.7
5 70.2 77.8 84.9 90.8 94.2 31.5 35.1 38.3 40.3 41.6 39.5 56.6
10 52.7 58.1 63.1 67.6 69.1 24.7 27.4 29.9 31.4 32.0 31.0 45.3
15 40.0 43.8 47.9 51.3 51.8 20.2 22.1 23.9 25.1 25.2 25.4 35.9
20 31.2 34.0 37.0 39.6 39.8 16.9 18.1 19.6 20.3 20.2 20.4 28.0
25 25.9 27.0 29.0 31.4 31.3 14.0 14.8 16.1 16.8 16.7 17.3 22.8
30 20.0 21.5 23.1 24.9 24.7 12.0 12.3 13.3 14.0 13.8 14.5 19.0
35 16.8 17.3 18.6 20.1 19.7 10.1 10.4 11.2 11.8 11.6 12.4 15.9
40 13.8 14.2 15.2 16.5 16.1 8.84 8.89 9.54 10.0 9.79 10.7 13.5
45 11.5 11.8 12.8 13.6 13.2 7.57 7.65 8.20 8.64 8.38 9.27 11.6
50 9.70 9.87 10.8 11.4 11.0 6.59 6.65 7.12 7.48 7.23 8.06 9.97
55 8.25 8.40 9.10 9.71 9.28 5.75 5.85 6.25 6.55 6.31 7.10 8.71
60 7.05 7.15 7.76 8.30 7.92 5.06 5.15 5.48 5.77 5.54 6.27 7.63
65 6.09 6.19 6.70 7.13 6.81 4.50 4.58 4.86 5.12 4.91 5.59 6.74
70 5.30 5.39 5.81 6.21 5.89 4.01 4.09 4.35 4.57 4.38 5.03 5.98
75 4.67 4.72 5.11 5.43 5.15 3.61 3.68 3.90 4.11 3.91 4.54 5.34
80 4.11 4.19 4.51 4.80 4.55 3.27 3.33 3.54 3.71 3.54 4.11 4.79
100b 2.77 2.80 3.06 3.21 3.06 2.37 2.44 2.59 2.72 2.55 3.04 3.46
c
VI 22 19 53 53 29 100 115 123 134 105 171 157
a
See Table 1 for the key to abbreviated names
b
Calculated from Eq. 1 with Ea in Table 4
c
VI was calculated by assuming the density of these esters were 0.9

Table 3 Dynamic viscosities of 12-acetylricinoleates at -10 to well below, not only that of the corresponding butyl ester,
-34 °C
but also that of the methyl ester. Seemingly, the branched
Temperature (°C) Viscosity (cP) isopropyl group and the branch on the 12 position of
MeR-Aca BuR-Ac i-PrR-Ac ricinoleate caused complex changes in the relative vis-
cosities of these esters.
-10.0 88.0 91.5 81.1 The viscosities of the esters examined in this paper
-15.2 124.0 126.1 112.3 were fitted with the Eq. 1 developed by Yao and Ham-
-20.4 177.8 186.0 165.0 mond [4] for oleate esters,
-24.6 257.6 263.9 234.2    
-30.8 398.9 413.5 355.5 R B þ CT þ DT 2
ln g ¼ ln þ ln T  ð1Þ
-34.9 620.9 635.9 556.1 Nh RT
a
See Table 1 for the key to abbreviated names where g is the dynamic viscosity, N is Avogadro’s num-
ber, h is Planck’s constant, T is the temperature in degrees
n-propyl ester throughout the temperature range studied, Kelvin, R is the gas constant, and B, C, and D are con-
but the difference in the viscosity of the isopropyl and butyl stants such that B ? CT ? DT2 = Ea, the energy of acti-
esters became smaller as the temperature increased, and at vation or the energy required for sliding one molecule past
25 °C and higher, the butyl esters had a greater viscosity. another.
Table 2 shows a similar relationship among the propyl, The results gave good fits of the data, which generally
butyl and isopropyl esters of the 12-acetylricinoleates. The are within the experimental error of the measurements
isopropyl ester had a greater viscosity at -5 °C but was (B±1%) when temperature is above 5 °C. However, larger
surpassed by the butyl ester at 20 °C. This observation is variations between calculated and observed viscosities
further complicated by data in Table 3, which shows that at were observed at low temperature measurements where the
-10 °C, the viscosity of isopropyl 12-acetylricinoleate was viscosity increased greatly with temperature decreases. The

123
944 J Am Oil Chem Soc (2010) 87:937–945

Table 4 Equations for calculating Ea with Eq. 1 significant difference in the COFs among mineral oil and
Compounds a
Ea calculations b mineral oil containing 10% MeR, BuR, or i-PrR. The COFs
ranged from 0.20 to 0.24. The COFs of unlubricated disks
MeR Ea = -0.2609T2 ? 531.95T – 58996 were significantly different (p = 0.0194) under our
EtR Ea = -0.2899T2 ? 552.85T – 62790 microscale conditions, which is not surprising [26] but is
PrR Ea = -0.3101T2 ? 565.30T – 64908 frequently overlooked in the statistical design of tribolog-
BuR Ea = -0.2920T2 ? 553.58T – 63202 ical tests. We overcame this problem with a statistical
i-PrR Ea = -0.3231T2 ? 576.31T – 67199 block design.
MeR-Ac Ea = -0.2053T2 ? 481.71T – 47511 The wear depths for various lubrication conditions are
EtR-Ac Ea = -0.2489T2 ? 511.33T – 52584 shown in Fig. 2. A low value indicates better wear pro-
PrR-Ac Ea = -0.2597T2 ? 518.69T – 54016 tection. Disks lubricated with MeR-Ac had significantly
BuR-Ac Ea = -0.2565T2 ? 516.09T – 53644 less wear depth than those of an unlubricated surface or a
i-PrR-Ac Ea = -0.2474T2 ? 512.28T – 53284 surface lubricated with mineral oil. Disks lubricated with
MeR-10 Ea = -0.2286T2 ? 493.75T – 49538 MeR-10, BuR, i-PrR and MeR showed significant less wear
MeR-11u Ea = -0.2582T2 ? 517.92T – 54829 depths than an unlubricated surface, but were not different
a from surfaces lubricated with mineral oil. Apparently,
See Table 1 for the key to abbreviated names
b adding 10% of MeR-Ac to mineral oil enhanced the pro-
T is in Kelvin
tection against wear damage. Again a significant effect for
equations for Ea and values for B, C, and D are given for disk blocks (p \ 0.0001) suggested that variation in the
each compound in Table 4. surface of the unlubricated disks should not be ignored in
In lubricants studies, it is customary to determine the the statistical design.
kinematic viscosity of a fluid at 40 and 100 °C and use this We confirmed that the locations of each wear track on a
data to calculate the viscosity index (VI) [23]. The kinematic particular disk, which were randomly selected to test for
viscosity can be obtained by dividing the dynamic viscosity wear depths, had no significant difference.
of a fluid by its density. The density of fatty acid esters are In conclusion, 12-acylated ricinoleate esters, and some
typically about 0.9 g/mL. For example, we found MeR’s acylated 12-hydroxystearate esters reported in this study
density to be 0.928 g/mL at 25 °C and 0.913 g/mL at 40 °C. have shown considerable potential for use in lubrication
We have calculated the approximate VI of the compounds applications because of their low melting temperatures,
listed in Table 2 according to ASTM D2270 [23] by viscosities and lubricity. Isopropyl 12-hydroxystearate
assuming the density of all these compounds was 0.9 and acylated with acetate and butyrate melted, respectively, at
calculating the viscosity at 100 °C of each compound with -9.8 and -15 °C. They should be quite stable to oxidation.
our Eq. 1 using the appropriate Ea from Table 4. As shown in The 12-acylated ricinoleate esters with acyl chain lengths
Table 2, the 12-acylated ricinoleate esters gave VIs in the of 8 or less all melted below -50 °C. In addition, MeR-Ac
range of 100 to 171 whereas ricinoleate esters with a free promises to reduce the wear of metal surfaces compared
hydroxy group had VIs below 100. The ricinoleate esters with petroleum-based lubricants. The oxidative stability of
with long acyl groups, i.e. MeR-10 and MeR-11u, had a MeR seems to be less than that of methyl oleate [20], but
greater VIs than 12-acetylricinoleates. MeR-11u had a lower
VI than that of MeR-10 possibly because of MeR-11u’s
terminal double bond. In general, the VI increased with
molecular weight but decreased with branching, i.e. the
isopropyl ester. The VI of simple fatty esters usually was
about 200 [2]. Having a branch in the carbon chain (such as
12-acyl ricinoleates) decreased the VI in agreement with
Hwang et al. [24]. But the VI reduction could be reduced
by increasing the chain length of the branch. The VI of
12-acylated ricinoleate esters are comparable to those of
synthetic oil and superior to those of mineral oil with similar
kinematic viscosities at 100 °C [25].

Friction and Wear Test

Fig. 2 The wear depths observed for unlubricated surfaces and
surfaces lubricated with mineral oil or mineral oil with 10% of
The COF was calculated as the slope of the linear regres- various esters. Treatments with the same letter had no significant
sion line of normal force versus lateral force. There was no difference (p \ 0.05)

123
J Am Oil Chem Soc (2010) 87:937–945 945

preliminary data from our laboratory suggests that MeR-Ac 7. Napier JA (2007) The production of unusual fatty acids in
is considerably more stable to oxidation than MeR. transgenic plants. Annu Rev Plant Biol 58:295–319
8. Isbell TA, Lowery BA, DeKeyser SS, Winchell ML, Cermak SC
Kortenska and Yanishlieva [27] have found that long-chain (2006) Physical properties of triglyceride estolides from lesque-
alcohols accelerate oxidation presumably by hydrogen rella and castor oils. Ind Crops Prod 23:256–263
bonding antioxidants as well as a direct effect on the oxi- 9. Cermak SC, Brandon KB, Isbell TA (2006) Synthesis and
dation reaction. Most of their work has been done with physical properties of estolides from lesquerella and castor fatty
acid esters. Ind Crops Prod 23:54–64
primary alcohols, and they found that the longer the chain, 10. Erhan SZ, Kleiman R, Isbell TA (1993) Estolides from mead-
the greater the prooxidant effect. Methyl ricinoleate has owfoam oil fatty acids and other monounsaturated fatty acids. J
been reported to exhibit a similar prooxidant effect [28]. If Am Oil Chem Soc 70:461–465
this mechanism is correct, then MeR-Ac might have a 11. Li Y, Beisson F (2009) The biosynthesis of cutin and suberin as
an alternative source of enzymes for the production of bio-based
stability comparable to oleate because its 12-hydroxy chemicals and materials. Biochimie 91:685–691
group is acylated. Further work is needed on the oxidative 12. Berdeaux O, Christie WW, Gunstone FD, Sebedio JL (1997)
stability of these compounds. Large-scale synthesis of methyl cis-9, trans-11-octadecadienoate
The isolation and acylation of pure ricinoleate esters can from methyl ricinoleate. J Am Oil Chem Soc 74:1011–1015
13. Sharma BK, Adhvaryu A, Liu Z, Erhan SZ (2006) Chemical
be accomplished by fairly simple chemistry. The ricinole- modification of vegetable oils for lubricant applications. J Am Oil
ate esters can be isolated in a pure form by the counter- Chem Soc 83:129–136
current distribution process described in this report. This 14. Kodali DR, Atkinson D, Redgrave TG, Small DM (1987)
step is critical to the removal of long-chain saturated and Structure and polymorphism of 18-carbon fatty acyl triacylgly-
cerols: effect of unsaturation and substitution in the 2-position. J
monounsaturated esters that would elevate the melting Lipid Res 28:403–413
point of the product. The acylation can be accomplished by 15. Ghebreyessus KY, Schiltz H, Angelici RJ (2006) Partial separa-
the sodium alkoxide process we used with suitable recy- tion of polyunsaturated fatty acid esters from FAMEs mixtures by
cling and drying of the large amount of acetate ester adsorption on silver nitrate-impregnated silica gel. J Am Oil
Chem Soc 83:645–652
required. The unacylated product of the reaction can be 16. AOCS (1994) Official methods and recommended practices of
recovered and recycled by the counter-current method used the American Oil Chemists’ Society. AOCS Press, Champaign
to isolate the original ricinoleate esters. The formation of 17. Bhuyan S, Sundararajan S, Yao L, Hammond EG, Wang T (2006)
esters of ricinoleate or other hydroxy alkyl esters in oilseed Boundary lubrication properties of lipid-based compounds eval-
uated using microtribological methods. Tribol Lett 22:167–172
plants that had probably been modified by molecular 18. Lie Ken Jie MSF, Cheng AKL (1993) Confirmation of the carbon
genetics would probably reduce the expense of this sort of chemical shifts of ethylenic carbon atoms in methyl ricinoleate
lubricant significantly. Smith [6] and Li and Beisson [11] and methyl ricinelaidate. Nat Prod Lett 3:65–69
have discussed ways in which this could be accomplished, 19. Spectral database for organic compounds (Accessed Feb 2010)
National Institute of Advanced Industrial Science and Technol-
although such a process has yet to be achieved. ogy, Japan. http://riodb01.ibase.aist.go.jp/sdbs/
20. Knothe G (2008) ‘‘Designer’’ biodiesel: optimizing fatty ester
Acknowledgments We thank the Plant Science Institute and Center composition to improve fuel properties. Energy Fuels 22:1358–
for Crop Utilization Research at Iowa State University for funding, 1364
and A. Pintar for statistics advice. 21. Bailey AE (1950) Melting and solidification of fats. Interscience
Publishers, New York, pp 146–147
22. Zakharkin LI, Guseva VV, Churilova IM (1983) Synthesis of
References 12-hydroxycarboxylic acids from 11-cyanoundecanoic acid.
Zhurnal Organicheskoi Khimii 19:1866–1869
1. Erhan SZ, Sharma BK, Perez JM (2006) Oxidation and low 23. ASTM (2007) Standard test method for standard practice for
temperature stability of vegetable oil-based lubricants. Ind Crops calculating viscosity index from kinematic viscosity at 40 and
Prod 24:292–299 100 °C. Designated D 2270
2. Erhan SZ, Adhvaryu A, Sharma BK (2006) Chemically func- 24. Hwang H, Adhvaryu A, Erhan EZ (2003) Preparation and prop-
tionalized vegetable oils. In: Rudnick LR (ed) Synthetics, mineral erties of lubricant basestocks from epoxidized soybean oil and
oils and bio-based lubricants. CRC Press, Taylor & Francis 2-ethylhexanol. J Am Oil Chem Soc 80:811–815
Group, Boca Raton, pp 361–389 25. Rudnick LR (2006) Polyalphaolefins. In: Rudnick LR (ed) Syn-
3. Yao L, Hammond EG (2006) Isolation and melting properties of thetics, mineral oils and bio-based lubricants. CRC press, Taylor
branched-chain esters from lanolin. J Am Oil Chem Soc 83:547– & Francis Group, Boca Raton, pp 1–33
552 26. Bushan B (2002) Introduction to Tribology. Wiley, New York,
4. Yao L, Hammond E, Wang T (2008) Melting points and vis- p 533
cosities of fatty acid esters that are potential targets for engi- 27. Kortenska VD, Yanishlieva NV (1995) Effect of the phenol
neered oilseed. J Am Oil Chem Soc 85:77–82 antioxidant type on the kinetics and mechanism of inhibited lipid
5. Fatemi SH, Hammond EG (1980) Analysis of oleate, linoleate oxidation in the presence of fatty alcohols. J Sci Food Agric
and linolenate hydroperoxides in oxidized ester mixtures. Lipids 68:117–126
15:379–385 28. Popov A, Yanishlieva N (1967) Influence of some components of
6. Smith MA (2007) Engineering oilseed for industry. In: Janick J, a nonglyceride complex of natural fats on accumulation and
Whipkey A (eds) Issues in New Crops and New Uses. ASHS decomposition of hydroperoxides. Grasas y Aceites (Sevilla,
Press, Alexandria, p 90 Spain) 11:645–655

123