Crystallization, polymorphism, and binary phase behavior of model enantiopure and racemic 1,3-Diacylglycerols

Article abstract of DOI:10.1021/cg101536q

1,3-Diacylglycerols (1,3-DAG) are components in many natural, commercial, and food systems. These compounds are always asymmetric, and diacid forms are chiral. To understand what effect this has on their crystallization behavior, model enantiopure (1-decanoyl-3-palmitoyl-sn-glycerol) and racemic (1,3-decanoyl-palmitoyl-rac-glycerol) 1,3-DAG were prepared and characterized. In addition, binary phase diagrams were prepared to investigate their phase behavior and the racemate's crystalline tendency. The major finding for this work is eutectic phase behavior was seen for blends of opposite enantiomers indicating racemic mixtures form conglomerates (mechanical mixtures of enantiopure crystals) in the solid phase. Differential scanning calorimetry melting curves of the racemic mixture display marked polymorphism, whereas, the pure enantiomer did not. This can be understood from a structural perspective since chain-end matching and hydrogen-bond optimization (via orientation of glycerol) are simultaneous for enantiopure, but are multistage for racemic DAG. Thus, there are critical differences between the crystallization behavior of enantiopure and racemic 1,3-DAG, and future physical analysis of these compounds should reflect this finding.

Full text of DOI:10.1021/cg101536q

ARTICLE
pubs.acs.org/crystal
Crystallization, Polymorphism, and Binary Phase Behavior of Model
Enantiopure and Racemic 1,3-Diacylglycerols
R. John Craven and Robert W. Lencki*
Department of Food Science, University of Guelph, Guelph, Ontario, Canada
ABSTRACT: 1,3-Diacylglycerols (1,3-DAG) are components
in many natural, commercial, and food systems. These com-
pounds are always asymmetric, and diacid forms are chiral. To
understand what effect this has on their crystallization behavior,
model enantiopure (1-decanoyl-3-palmitoyl-sn-glycerol) and
racemic (1,3-decanoyl-palmitoyl-rac-glycerol) 1,3-DAG were
prepared and characterized. In addition, binary phase diagrams
were prepared to investigate their phase behavior and the
racemate’s crystalline tendency. The major finding for this work
is eutectic phase behavior was seen for blends of opposite enantiomers indicating racemic mixtures form conglomerates (mechanical
mixtures of enantiopure crystals) in the solid phase. Differential scanning calorimetry melting curves of the racemic mixture display
marked polymorphism, whereas, the pure enantiomer did not. This can be understood from a structural perspective since chain-end
matching and hydrogen-bond optimization (via orientation of glycerol) are simultaneous for enantiopure, but are multistage for
racemic DAG. Thus, there are critical differences between the crystallization behavior of enantiopure and racemic 1,3-DAG, and
future physical analysis of these compounds should reflect this finding.
’
INTRODUCTION
crystallization. In addition, it is also possible to obtain the β form
1
5,6
by storing β crystals close to their melting point.
2
Acylglycerols are the main components in fats, oils, and cell
β1-form 1,3-DAG adopt a V-shaped (herringbone) conforma-
membranes. In nature, acylglycerols are often chiral due to
the positional specificity of biosynthetic pathways. For instance,
cell membranes typically only contain the (R)-stereoisomer of
phospholipid (phosphate moiety in the sn-3 position of
glycerol). Likewise, lipid components from natural sources
incorporated into commercial products may also be chiral. For
example, at least 15 commonly occurring triacylglycerols in palm
tion in which the two hydrocarbon chains extend out from the
7
polar center. This structure was determined using data from
both single-crystal X-ray (for 1,3-di-(3)-thiadodecanoyl-
8
1
glycerol) and infrared spectroscopy, and has since been con-
firmed for a different monoacid 1,3-DAG (viz. 1,3-di-(11)-
bromoundecanoyl-glycerol) and for a racemic diacid 1,3-DAG
9,10
2
(viz. 1,3-stearoyl-oleoyl-rac-glycerol).
Recently, it has been
oil have been identified as chiral. On the other hand, racemic
demonstrated that this same structure holds for 1,3-acyl-palmi-
mixtures (of chiral acylglycerols), monoacid acylglycerols and
other nonchiral acylglycerols also occur in nature and are
common in industry and research. It is well-known that there
are significant differences between modes of crystallization for
pure enantiomers and mixtures of opposite enantiomers, but
these stereochemical aspects have been largely overlooked in
most lipid crystallization studies to date.
toyl-rac-glycerols with acyl groups of varying chain length (from
11
butyric through to oleic; 4:0 to 18:1).
All diacid 1,3-DAG are chiral about the sn-2 position of
glycerol. Nevertheless, there has been a tendency to examine
the physical properties of 1,3-DAG without taking this into
account. Consequently, the crystallization behavior of pure
enantiomers has not been fully distinguished from that of mixed
enantiomers or racemic mixtures in spite of critical differences.
The role of stereochemistry in crystallization behavior is deter-
mined using phase diagrams derived from binary mixtures of
pure enantiomer and the corresponding racemic mixture. In the
resulting phase diagrams, a characteristic liquidus profile is
associated with each particular crystalline tendency—racemic
3
1,3-Diacylglycerols (1,3-DAG) play a key role in a number of
industrially important systems and show promise as food in-
gredients in a new class of weight-reducing products. When
1
,3-DAG are substituted for dietary triacylglycerol (TAG), weight
4
loss and improved blood cholesterol levels result. To develop
,3-DAG-based products, however, a detailed understanding of
1
their physical chemistry (in particular crystallization behavior) is
required. For the most part, 1,3-DAG crystallize in one of two
monotropic triclinic parallel β polymorphs which are, in order of
increasing thermodynamic stability, β and β . In broad terms, β
1
12
compound, conglomerate, or pseudoracemate.
2
1
is obtained by crystallizing 1,3-DAG from less polar solvents at
Received: November 19, 2010
higher temperatures and β is obtained by crystallizing 1,3-DAG
Revised:
March 29, 2011
2
from more polar solvents at lower temperatures or through melt
Published: April 01, 2011
r 2011 American Chemical Society
1566
dx.doi.org/10.1021/cg101536q
|Cryst. Growth Des. 2011, 11, 1566–1572

Crystal Growth & Design
ARTICLE
Figure 1. Stereochemistry of (R)-1-decanoyl-3-palmitoyl-sn-glycerol.
Hydrogen atom displays stereochemistry at sn-2 position.
The main purpose of this work is to compare the physical
chemistry of an enantiopure 1,3-DAG with its racemic counter-
part. This will improve understanding of the role stereochemistry
plays in crystallization and perhaps lend some insight into the
polymorphism of 1,3-DAG. 1-Decanoyl-3-palmitoyl-sn-glycerol
and 1,3-decanoyl-palmitoyl-rac-glycerol were chosen as models
for study because the racemic form has relatively simple poly-
1
1
morphic behavior, and the difference in chain-length was
regarded as sufficient for the potential effects of chirality to be
detectable.
’
MATERIALS AND METHODS
Unless noted otherwise materials, methods and procedures were
12
identical to those described previously. Prepared samples were stored
in a freezer (ꢀ30 °C) to minimize acyl migration. Samples analyzed by
infrared and X-ray spectroscopy were stored at ꢀ30 °C for at least one
month prior to testing. Samples analyzed by differential scanning
calorimetry (DSC) were stored at ꢀ30 °C for at least 3 days prior to
Figure 2. Gas chromatography traces for (a) 1-decanoyl-3-palmitoyl-
sn-glycerol and (b) 1,3-decanoyl-palmitoyl-rac-glycerol. Peaks prior to 5
min are solvent and derivatizing agent.
1
testing. DSC analysis indicated that the high-melting form (β ) was
obtained within this time frame. The purity of prepared compounds was
determined by GC/FID analysis of trimethylsilyl- derivatives prepared
using N-trimethylsilylimidazole and pyridine.
Preparation of Enantiopure 1-Decanoyl-3-palmitoyl-sn-
glycerol. Decanoyl chloride (2.84 g) dissolved in 30 mL of methylene
chloride was added dropwise to a solution of 3-palmitoyl-sn-glycerol
where all enthalpy and temperature values are for fusion,
temperature is in Kelvin, and the entropy of mixing for the solid
is assumed to be zero. This calculated value was compared with
13
3
the theoretical molar entropy of mixing calculated using
m
l
ΔS ¼ ꢀ RðxE ln xE þ x
0
ln xE
E
0
Þ
ð3Þ
(
5.0 g), triethylamine (2.73 g) and N,N-dimethylaminopyridine (0.18 g)
12
in 70 mL of methylene chloride stirred in an ice bath. After addition,
the solution was stirred at room temperature (∼22 °C) for 3 h. Solvent
was removed under vacuum, the residue was taken back up in hexane and
filtered, then the raw product was recrystallized from the filtrate.
Enantiopure 1-decanoyl-3-palmitoyl-sn-glycerol (Figure 1) was further
purified by flash chromatography with hexane/ethyl acetate (7:2, v/v) to
yield 4.70 g of >99% (by GC) pure product.
’ RESULTS AND DISCUSSION
For simplicity, enantiopure 1-decanoyl-3-palmitoyl-sn-glycer-
ol (Figure 1) will occasionally be identified by the letter E or the
short form E-DAG, and racemic 1,3-decanoyl-palmitoyl-rac-
glycerol by the letter R or the short form R-DAG. It is worth
noting that the racemic mixture (1,3-decanoyl-palmitoyl-rac-
glycerol) is composed of 1-decanoyl-3-palmitoyl-sn-glycerol
Preparation of Racemic 1,3-Decanoyl-palmitoyl-rac-gly-
cerol. This compound was produced as for the enantiopure 1-decanoyl-
0
0
and 1-palmitoyl-3-decanoyl-sn-glycerol (E or E -DAG) in a 1:1
ratio. Consequently, R-DAG could also be identified as 50%
3-palmitoyl-sn-glycerol but using racemic 1(3)-palmitoyl-rac-glycerol as
a starting material. This produced 4.10 g of >99% (by GC) pure product.
0
E-DAG and 50% E -DAG in the accompanying figures, tables and
text. For consistency, we will use terminology as defined in
Jacques, Collet, and Wilen’s book: Enantiomers, Racemates and
Resolutions.
Chemical esterification with decanoyl chloride was used to
produce both enantiopure and racemic DAG from the appro-
priate monopalmitin. Subsequent flash chromatography yielded
product (E: 65% and R: 56% yield) in very high purity (>99% by
GC) (Figure 2). As expected, NMR spectra were identical for
’
CALCULATIONS
The theoretical melting point depression was calculated using
3
a form of the Hildebrand equation,
!
2
RT
E
ΔT ꢁ
ln γx
ð1Þ
ΔHf
where ΔT is the melting point depression, T is assumed to be
both enantiomer and racemic mixture. δ = 0.88 (6H, t), 1,25
(36H, m), 1.63 (4H, m), 2.05 (1H, s), 2.35 (4H, t), 4.14 (5H, m)
E
H
14
approximately equal to T , and γ is assumed to be one.
R
The molar entropy of mixing for the liquid was calculated
(Figure 3).
using
Infrared spectra for the highest-melting forms of both the
enantiopure compound and racemic mixture are similar, and any
differences can be attributed to variations in the strength and
orientation of hydrogen bonding between hydroxyl and carboxyl
ΔHE ΔHR ΔHE ꢀ ΔHR
TE
TR
m
ꢀ
ΔS ¼
ꢀ
ꢀ
ln
ð2Þ
l
TE
TR
TE ꢀ TR
1
567
dx.doi.org/10.1021/cg101536q |Cryst. Growth Des. 2011, 11, 1566–1572

Crystal Growth & Design
ARTICLE
groups (Figure 4, Table 1). As a result, absorbances around 3491,
Substantial differences between 1-decanoyl-3-palmitoyl-sn-
glycerol and 1,3-decanoyl-palmitoyl-rac-glycerol are apparent
from an inspection of their respective DSC crystallization and
melting data (Table 2). The initial melt of tempered solids (high-
melting β -form) yields a melting point (T ) of 60.69 °C for E
ꢀ1
1
730, 1710, and 1194 cm are greater for the enantiopure
compound. In addition, differences in location for the carbonyl
ꢀ1
stretching frequencies also occur (1707 and 1730 cm for E and
1
ꢀ
1
712 and 1730 cm for R). All told, this indicates differences in
15
1
e
hydrogen-bonding environment between E and R. Both the
Table 1. Main Infrared Absorbances for 1-Decanoyl-3-pal-
ꢀ
1
mitoyl-sn-glycerol (E) and 1,3-Decanoyl-palmitoyl-rac-gly-
16
cerol (R)
ꢀ1
wavenumber (cm
)
description
Similarly, X-ray powder diffraction indicates that both com-
pounds are in the β polymorph (Figure 5). Main peaks in the
WAXS region for the enantiopure compound (3.69 vs, 3.81 s, and
3
491
OꢀH stretch for ꢀOH H-bonded to
carbonyl O (E > R)
2
849 and 2912
CꢀH stretch for acyl chain and
glycerol backbone
4
3
.56 s with a shoulder at 4.50 Å) and racemic mixture (3.72 vs,
.80 s, and 4.60 s with a shoulder at 4.53 Å) are similar to those
E: 1707 and 1730
R: 1712 and 1730
1412
CdO stretch (E > R); two peaks due to
differences in H-bonding
reported previously for the β polymorph of 1,3-lauroyl-palmi-
1
17
toyl-rac-glycerol (3.76, 3.80, and 4.58 Å). Some additional
peaks occur at lower angles in the X-ray powder diffraction
spectra (both E and R: 2θ = 2.22, 4.42, 6.6, 8.82° and R only:
strongest CꢀH deformation and wagging band
CꢀH deformation and wagging bands
CꢀOH bending and CꢀO stretching (E > R)
CꢀOH bending and CꢀO stretching bands
CꢀC stretch and CꢀH rocking vibrations
methylene rocking vibrations
1300ꢀ1400
1194
2
θ = 3.32, and 5.52°). Similarly, SAXS (1.2 to 4.5°) of both
enantiomer and racemic mixture were identical (2θ = 2.20°)
indicating their lamellae are the same size (39.75 Å). Several
minor peaks were also observed in the SAXS measurement
1
8
7
142 ꢀ 1186
70ꢀ1150
16
(E: 2θ = 4.43° and R: 2θ = 1.10, 3.31 and 4.44°).
single peak indicates triclinic parallel
arrangement (β polymorph)
Figure 3. NMR spectrum of 1,3-decanoyl palmitoyl-rac-glycerol. Peak
assignments are provided in the inset. Spectrum for 1-decanoyl-3-
palmitoyl-sn-glycerol was identical.
Figure 5. X-ray powder diffraction patterns for (a) 1-decanoyl-3-
palmitoyl-sn-glycerol and (b) 1,3-decanoyl palmitoyl-rac-glycerol.
ꢀ
1
Figure 4. Infrared spectra for (a) 1-decanoyl-3-palmitoyl-sn-glycerol with inset emphasizing two intense absorbance bands at 1707 and 1730 cm and
ꢀ1
(b) 1,3-decanoyl palmitoyl-rac-glycerol with inset emphasizing two absorbance bands at 1712 and 1730 cm .
1
568
dx.doi.org/10.1021/cg101536q |Cryst. Growth Des. 2011, 11, 1566–1572

Crystal Growth & Design
ARTICLE
Table 2. Crystallization and Melting Data for Pure Enantio-
mer and Racemic Mixture
and 54.60 °C for R with very similar enthalpies of fusion (95.69
and 100.62 kJ/mol, respectively). In cooling/heating cycles
conducted at 2, 5, and 10 °C/min, lower-melting solids
(β -form) are obtained, and the difference (ΔT ) between the
1
-decanoyl-3-palmitoyl-sn-glycerol
2
e
higher-melting (β ) and lower-melting (β ) forms is less for E
1
2
crystallization melting
than it is for R (on average 2.27 and 3.71°, respectively). There
are marked differences between the DSC crystallization and
melting curves for 1-decanoyl-3-palmitoyl-sn-glycerol and 1,3-
decanoyl-palmitoyl-rac-glycerol (Figure 6). Most notably, the
crystallization of metastable forms is visible in distorted exother-
mic peaks (5 and 2 °C/min) and the endothermic double peak
ΔH ΔH
c
f
T
e
(°C)
T
p
e p
(°C) (kJ/mol) T (°C) T (°C) (kJ/mol)
a
5
2
5
1
°C/min
°C/min
°C/min
60.69
58.57
58.33
58.37
61.76
60.86
61.26
61.98
95.69
94.77
95.69
96.17
49.18
48.79
49.52
48.04
47.65
96.95
96.80
97.24
(
2 °C/min) for R, whereas the polymorphism of E appears to be
0 °C/min 47.78
limited to the occurrence of β and β forms with slightly
1
2
different melting points and, perhaps, to the crystallization of a
metastable form leading to the distorted exothermic peak in the
sample cooled at 2 °C/min.
The DSC melting curves for binary mixtures of 1-decanoyl-
3-palmitoyl-sn-glycerol and 1,3-decanoyl palmitoyl-rac-glycerol
in 10% increments were plotted together to facilitate comparison
1
,3-decanoyl-palmitoyl-rac-glycerol
crystallization melting
ΔH ΔH
c
f
T
e
(°C)
T
p
e p
(°C) (kJ/mol) T (°C) T (°C) (kJ/mol)
(
Figure 7). These melting curves were analyzed, and T for each
a
p
5
2
°C/min
54.60
50.87
56.79
100.62
binary mixture was used to display liquidus data (Figure 8). The
observed trend is also evident from casual observation of the
compiled melting curves (Figure 7). The line in Figure 8
represents a linear regression on data obtained from complete
cooling/heating cycles (see below). The enantiopure compound
b
b
b
b
°C/min 45.18
44.47
97.53
52.7
47.92
b
b
5
2.55
54.49
53.43
54.03
49.56
5
°C/min 44.03
0 °C/min 42.7
High-melting form. First and second peak; total ΔH
43.51
41.91
97.28
96.85
51.04
50.76
97.14
95.54
1
a
b
f
= 97.48 kJ/mol.
has the highest melting point (T ). The melting point decreases
p
Figure 6. DSC curves for the (a) crystallization and (b) melting of enantiopure 1-decanoyl-3-palmitoyl-sn-glycerol, and the (c) crystallization and (d)
melting of 1,3-decanoyl-palmitoyl-rac-glycerol. Dashed line distinguishes the high-melting form and solid lines delineate cooling/heating cycles starting
with completely melted samples.
1
569
dx.doi.org/10.1021/cg101536q |Cryst. Growth Des. 2011, 11, 1566–1572

Crystal Growth & Design
ARTICLE
Figure 7. DSC melting curves for high-melting form of binary blends of
Figure 9. DSC melting curves for binary blends of 1-decanoyl-3-
palmitoyl-sn-glycerol (E) and 1,3-decanoyl-palmitoyl-rac-glycerol (R).
Samples were cooled from 100 °C to ꢀ20 at 2 °C/min and then melted
at 2 °C/min. Labels refer to relative proportion of enantiopure
compound and racemic mixture.
1-decanoyl-3-palmitoyl-sn-glycerol (E) and 1,3-decanoyl palmitoyl-rac-
glycerol (R). Labels refer to relative proportion of enantiopure com-
pound and racemic mixture.
Figure 10. DSC melting curves for binary blends of 1-decanoyl-3-
palmitoyl-sn-glycerol (E) and 1,3-decanoyl palmitoyl-rac-glycerol (R).
Samples were cooled from 100 °C to ꢀ20 at 5 °C/min and then melted
at 5 °C/min. Labels refer to relative proportion of enantiopure
compound and racemic mixture.
Figure 8. Liquidus data (T
p
) for high-melting form of binary blends of
1
-decanoyl-3-palmitoyl-sn-glycerol (E) and 1,3-decanoyl palmitoyl-rac-
glycerol (R). Line represents linear regression on data collected in
complete cooling/heating cycles (Figure 12).
the binary mixture. Likewise, the calculated entropy of mixing for
liquid E- and E -DAG (5.5 J/mol by eq 2) is close to the
0
theoretical value (5.76 J/mol) predicted using eq 3.
with each successive addition of racemic mixture, finally reaching
the lowest melting point, that of the racemic mixture. In this case,
the racemic mixture of 1-decanoyl-3-palmitoyl-sn-glycerol and
In addition to those for the high melting (β ) form (Figure 7),
1
DSC melting curves for cooling/heating cycles of binary mix-
tures of 1-decanoyl-3-palmitoyl-sn-glycerol and 1,3-decanoyl-
palmitoyl-rac-glycerol crystallized and melted at 2, 5, and
10 °C are presented in Figures 9, 10, and 11, respectively. These
cooling/heating curves provide information on the low-melting
1
-palmitoyl-3-decanoyl-sn-glycerol has a lower melting point
than either of the enantiopure compounds. This eutectic phase
behavior, where the pure enantiomers’ melting points are
mutually depressed, is typical for systems crystallizing as
form’s (β ) phase behavior including any additional polymorph-
2
3
conglomerates. The actual melting point depression (6.1°) is
ism. It is interesting to note that samples rich in the enantiopure
compound (E7R3 to pure E) tend to produce one main peak
whereas samples rich in racemic compound (pure R to E6R4)
often display distorted or double peaks. This echoes the differing
extent of polymorphism found for pure enantiomer and racemic
mixture above (Figure 6). From a structural perspective this
makes sense, since chain-end matching and hydrogen-bond
optimization (orientation of glycerol) would occur
close to the predicted value (6.7°) determined using the Hildeb-
rand equation (eq 1) and DSC data for pure E. Thus, reduction in
the melting point of 1-decanoyl-3-palmitoyl-sn-glycerol by addi-
tion of 1-palmitoyl-3-decanoyl-sn-glycerol (in the form of a
racemic mixture) is a colligative property and similar melting
point reduction would be expected for addition of a completely
dissimilar compound (in place of the opposite enantiomer) in
1
570
dx.doi.org/10.1021/cg101536q |Cryst. Growth Des. 2011, 11, 1566–1572

Crystal Growth & Design
ARTICLE
Table 4. Melting Data (T ) for Binary Mixtures of 1-Decan-
e
oyl-3-palmitoyl-sn-glycerol (E) and 1,3-Decanoyl-palmitoyl-
rac-glycerol (R) at Three Different Cooling/Heating Rates
e
T (°C)
samples
2 °C/min
5 °C/min
10 °C/min
mean
SD
pure E
E9R1
E8R2
E7R3
E6R4
E5R5
E4R6
E3R7
E2R8
E1R9
pure R
58.56
57.24
55.62
53.97
51.42
51.07
51.87
52.10
52.17
58.33
56.98
55.44
53.45
51.38
50.45
51.56
51.82
49.43
50.62
51.04
58.37
57.00
55.63
53.64
51.31
50.21
51.55
49.60
50.46
50.79
50.76
58.42
57.07
55.56
53.69
51.37
50.58
51.66
51.17
50.69
50.71
50.89
0.12
0.14
0.11
0.26
0.06
0.44
0.18
1.37
1.38
0.12
0.14
Figure 11. DSC melting curves for binary blends of 1-decanoyl-3-
palmitoyl-sn-glycerol (E) and 1,3-decanoyl palmitoyl-rac-glycerol (R).
Samples were cooled from 100 °C to ꢀ20 at 10 °C/min and then melted
at 10 °C/min. Labels refer to relative proportion of enantiopure
compound and racemic mixture.
50.86
Table 3. Melting Data (T ) for Binary Mixtures of 1-Decan-
p
oyl-3-palmitoyl-sn-glycerol (E) and 1,3-Decanoyl-palmitoyl-
rac-glycerol (R) at Three Different Cooling/Heating Rates
p
T (°C)
samples
2 °C/min
5 °C/min
10 °C/min
mean
SD
pure E
E9R1
E8R2
E7R3
E6R4
E5R5
E4R6
E3R7
E2R8
E1R9
pure R
60.86
60.28
59.77
58.80
58.37
57.44
54.76
54.63
55.15
54.76
54.49
61.26
60.82
60.29
59.43
58.75
57.79
57.80
55.70
56.38
54.78
54.43
61.98
61.00
60.72
60.26
59.37
58.33
59.16
57.01
56.70
55.23
54.03
61.37
60.70
60.26
59.50
58.83
57.85
57.24
55.78
56.08
54.92
53.98
0.57
0.37
0.48
0.73
0.50
0.45
2.25
1.19
0.82
0.27
0.53
simultaneously for the enantiopure compound. In contrast, these
two processes would be separate and independent for racemic
mixtures and, ultimately, both processes would be necessary for
the formation of the most-stable crystal form. At pre-
sent, however, it is difficult to know the extent to which
polymorphism and phase behavior (i.e., formation of addition
compounds) contribute to the observed melting behavior of
binary mixtures.
Figure 12. Melting data for cooling/heating cycles on binary blends of
1-decanoyl-3-palmitoyl-sn-glycerol (E) and 1,3-decanoyl-palmitoyl-rac-
glycerol (R). Triangles are for data collected at 2 °C/min, diamonds for
5
°C/min and squares for 10 °C/min. Extrapolated onset of melting
(
T
e
) values are represented by open symbols, maximum peak tempera-
ture (T ) values are represented by filled symbols. Solid lines were
p
derived by linear regression on mean data and the dashed line is the
suggested eutectic solidus.
Meltingdataforcooling/heating cyclesillustratedinFigures9,
1
0, and 11 is summarized in Tables 3 and 4 and plotted in
Figure 12. In spite of differences in cooling/heating rate the data
especially for X e 0.5) shows good reproducibility (Tables 3
behavior is evident for mixtures rich in enantiopure compound
(
(X e 0.5). Thus, the phase behavior can be summarized as
R
R
0
and 4). Linear regression for the mean values of T (peak
eutectic (between E- and E -DAG) having limited solid solution
p
maximum) yields an equation (T (°C) = ꢀ7.418x þ 61.57)
formation (for enantiomer-rich mixtures), and the racemic mix-
ture crystallizes as a mechanical mixture (conglomerate).
Values for T and T are far less regular for racemic and near
p
2
with R = 0.988. This line is shown in Figures 8 and 12. Linear
regression for the mean values of T (extrapolated onset) for X
e
R
p
e
e 0.5 yields an equation (T (°C) = ꢀ16.63x þ 58.61) with
racemic mixtures (X > 0.5) than for enantiomer-rich mixtures
e
R
2
R = 0.989. Thisline is shown in Figures 12. Phase behavior on the
(X e 0.5) (Figure 12). The melting data suggests that either
R
basis of T values is eutectic indicating the solid is a conglomerate
for 1:1 mixtures of E- and E -DAG. In addition, solid solution
polymorphs or addition compounds are formed for E- and
E -DAG at a number of ratios in particular for E4R6 cooled
p
0
0
1
571
dx.doi.org/10.1021/cg101536q |Cryst. Growth Des. 2011, 11, 1566–1572

Crystal Growth & Design
ARTICLE
and heated at 5 and 10 °C/min (X = 0.6). To a lesser
(16) Chapman, D. The Structure of Lipids by Spectroscopic and X-ray
Techniques: With a Chapter on Separation Techniques including thin layer
and gas liquid chromatography; Methuen and Co Ltd: London, 1965.
R
degree, similar behavior is seen for E3R7 and E2R8 (X =
R
0
.7 and 0.8). As a consequence, the current data implies, but
(17) Sidhu, S. S.; Daubert, B. F. J. Am. Chem. Soc. 1946, 68 (12),
does not explicitly delineate, the eutectic solidus location.
A dashed line represents the likely location for the eutectic
solidus (Figure 12).
2603–2605.
In conclusion, investigation of phase behavior through binary
phase diagrams indicates 1,3-decanoyl-palmitoyl-rac-glycerol
crystallizes as a conglomerate. In the solid phase, opposite
enantiomers of 1,3-DAG are mutually immiscible and thus form
a mechanical mixture. Future crystal structure determinations for
1
,3-DAG should reflect this finding. Furthermore, more complex
melting behavior was observed for the racemic mixture and
systems rich in racemic mixture. This may be due to greater
polymorphism, the formation of addition compounds, or both.
In sum, this demonstrates the importance of stereochemistry in
determinations of crystal structure and understanding the crystal-
lization behavior of 1,3-DAG.
’
AUTHOR INFORMATION
Corresponding Author
*Phone: 519-824-4120, x54327. E-mail: rlencki@uoguelph.ca.
’
ACKNOWLEDGMENT
Financial support for this project was provided by National
Sciences and Engineering Research Council of Canada; we are
grateful for their generosity. Thanks to ACD (Toronto) for their
free chemical structure drawing and NMR analysis software
(Chemsketch, 3D Viewer and 1D NMR Processor). Thanks
also to Japan VAM & POVAL Co., Ltd. for their donation of vinyl
esters of fatty acids.
’
REFERENCES
(
(
(
1) Weiss, R. M.; McConnell, H. M. Nature 1984, 310, 47–49.
2) Loh, S. K.; Choo, Y.-M. J. Oil Palm Res. 2003, 15 (1), 6–11.
3) Jacques, J.; Collet, A.; Wilen, S. H. Enantiomers, Racemates and
Resolutions; John Wiley and Sons: New York, 1981.
4) Christophe, A. B. Structural Effects on Adsorption, Metabolism,
and Health Effects of Lipids. In Bailey’s Industrial Oil and Fat Products,
(
6
2
th ed. ed.; Shahidi, F., Ed.; John Wiley & Sons, Inc.: Hoboken, NJ,
005; Vol. 3, pp 535ꢀ553.
(5) Larsson, K. Physical Properties - Structural and Physical Char-
acteristics. In The Lipid Handbook; Gunstone, F. D., Harwood, J. L.,
Padley, F. B., Eds.; Chapman and Hall Ltd: London, 1986; pp 321ꢀ384.
(6) Shannon, R. J.; Fenerty, J.; Hamilton, R. J.; Padley, F. B. J. Sci.
Food Agric. 1992, 60, 405–417.
(
(
(
(
7) Larsson, K. Acta Crystallogr. 1963, 16, 741–748.
8) Chapman, D. J. Chem. Soc. 1956, 2522–2528.
9) Hybl, A.; Dorset, D. Acta Crystallogr. 1971, B27, 977–986.
10) Goto, M.; Honda, K.; Di, L.; Small, D. M. J. Lipid Res. 1995,
3
1
1
1
6, 2185–2190.
11) Craven, R. J.; Lencki, R. W., J. Am. Oil Chem. Soc. 2011, DOI:
0.1007/s11746-011-1769-0.
12) Craven, R. J. Lencki, R. W. Cryst. Growth Des., 2011, DOI:
0.1021/cg101654c.
13) Craven, R. J.; Lencki, R. W. J. Am. Oil Chem. Soc. 2010, 87 (11),
281–1291.
14) Atkins, P. W. Physical Chemistry, 2nd ed.; Oxford University
Press: Oxford, 1982.
15) Chapman, D. Chem. Rev. 1962, 62, 433–456.
(
(
(
(
(
1
572
dx.doi.org/10.1021/cg101536q |Cryst. Growth Des. 2011, 11, 1566–1572