Journal Pre-Proofs: Food Chemistry
Journal Pre-Proofs: Food Chemistry
Journal Pre-Proofs: Food Chemistry
PII: S0308-8146(20)32710-2
DOI: https://doi.org/10.1016/j.foodchem.2020.128848
Reference: FOCH 128848
Please cite this article as: Simoes, S., Lelaj, E., Rousseau, D., The presence of crystalline sugar limits the
influence of emulsifiers on cocoa butter crystallization, Food Chemistry (2020), doi: https://doi.org/10.1016/
j.foodchem.2020.128848
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
crystallization
*Corresponding author
Address:
Ryerson University
350 Victoria St
Fax: +1-416-979-5044
Email: rousseau@ryerson.ca
1
ABSTRACT
The effects of 2 wt% emulsifier and crystalline sugar on the isothermal solidification and
polymorphic behavior of cocoa butter were evaluated. The emulsifiers investigated were
commercial soy lecithin, polyglycerol polyricinoleate (PGPR), citric acid esters of mono- and
cocoa butter nucleation and growth from the melt, with PGPR showing the smallest
enhancement. Lecithin and AP minimally affected the polymorphic form IV-to-V transition
contrary to PGPR and CITREM, which both promoted the formation of form V crystals. The
presence of sugar alone accelerated cocoa butter solidification while limiting the ability of the
emulsifiers to do so. Sugar alone, and in the presence of emulsifier, hindered the polymorphic
form IV-to-V transition. This study shows that the effects of emulsifiers on the isothermal
crystallization of cocoa butter can be muted in the presence of crystalline sugar, suggesting a
2
1. INTRODUCTION
Emulsifiers are a key component in confectionery manufacturing as they help to control the
viscosity and yield stress of molten chocolate formulations. It is well-known that emulsifiers
will also affect triacylglycerol (TAG) crystallization (Smith, Bhaggan, Talbot, & Van Malssen,
2011). When mixed with bulk fats, they may enhance crystal growth by crystallizing or self-
organizing prior to the TAGs themselves, thereby acting as a template for nucleation and
growth (Davis & Dimick, 1989; Ishibashi, Hondoh, & Ueno, 2017). Conversely, they may
polyricinoleate (PGPR), ammonium phosphatides (AP) and citric acid esters of mono- and di-
glycerides (CITREM) also see usage in commercial products (Ačkar et al., 2015; Beckett,
2008). There is no current consensus regarding the effects of commercial soy lecithin on
confectionery fat crystallization as, in some cases, it may enhance cocoa butter fat crystal
growth rate, particularly at lower concentrations (0.2 - 0.3 wt%), but inhibit it at higher
concentrations (0.5 – 2 wt%) (Man, Rahman, Yusoff, Aini, & Miskandar, 2007; Miyasaki,
Luccas, & Kieckbusch, 2016; Rigolle et al., 2015). PGPR, an oligomer of polyricinoleic acid
chains esterified onto a polyglycerol backbone, is generally thought to have little effect on the
crystallization of cocoa butter or hydrogenated fats (Miyasaki et al., 2016; Wang, Liu, Jin,
Huang, Meng, & Wang, 2011). AP is an alternative to lecithin based on the same molecular
rather than a zwitterion. Reports to-date have presented conflicting conclusions stating that it
can either accelerate or slow fat crystallization (Rossi, 1998; MacMillan, 2000). Finally,
CITREMs are synthetic emulsifiers where fatty acids are esterified onto a glycerol moiety with
subsequent attachment of a polar citric acid group. CITREMs have shown improved
3
rheological and textural properties of compound chocolates with a reduced fat content, but little
complex systems while overlooking the important role of dispersed particles. Other than
emulsifiers, exploring how dispersed particles affect cocoa butter crystallization is important
of TAG crystals off the surface of added particles was observed with the addition of 1 % foreign
particulate (Yoshikawa, Kida, & Sato, 2014). Comparatively little is known about fat
The present study explores the effects of added sugar and emulsifiers relevant to the
confectionery industry on cocoa butter isothermal crystallization and its form IV-to-V
polymorphic transition. The governing hypothesis was that emulsifiers would alter cocoa butter
crystallization, with their effect(s) dependent on molecular structure. Dispersed sugar, other
than enhancing cocoa butter crystallization by acting as a site for heterogeneous nucleation,
would interfere with the emulsifier’s capacity to affect crystallization. Assessing such effects
is highly relevant for the confectionery industry given the dominant role of dispersed particles
in confectionery formulations.
2.1 Materials
Commercial cocoa butter, canola oil and lecithin were provided by Mondelēz International
(East Hanover, New Jersey, USA), with the composition of cocoa butter provided in
Supplemental Table 1. Canola oil was purified on silica (60Å, 70-230 mesh, 63-200 µm)
purchased from Sigma Aldrich (St. Louis, Missouri, USA) to remove any partial acylglycerols
4
and free fatty acids. PGPR 4150, CITREM 4205 and AMP 4455 were acquired from Palsgaard
(Juelsminde, Denmark) and had average molecular weights of 1500, 465 and 700 g/mol,
respectively. CITREM 4205 and AMP 4455 are produced using sunflower oil and have a high
oleic content (75 – 90 %). All emulsifiers were used as received. Confectioner’s sugar (10x,
average particle size ~ 60 µm) was purchased from Domino Foods (Iselin, New Jersey, USA)
and used as received. Cocoa butter-emulsifier blends were prepared by adding 2 wt%
emulsifier to molten cocoa butter followed by magnetic stirring at 80 °C for 1 hour. Cocoa
butter sugar blends were prepared as follows: a 1:1 mixture of 10x confectioner’s sugar and
molten cocoa butter was magnetically stirred at 40 °C for 1 hour. If present, emulsifier was
premixed with molten cocoa butter as above, before addition of sugar. In samples containing
sugar, the mixture consisted of 1 wt% emulsifier (i.e., 2 wt% relative to the fat phase).
SFC was determined using a pulsed NMR (Minispec mq20, Bruker, Karlsruhe, Germany). A
2.0 mL aliquot of sample was added to an NMR tube (10 mm O.D.). Samples were heated in
waterbath for 1 minute to cool. The tube was immediately transferred to the NMR sample port
with the sample chamber set at 20 °C. Isothermal crystallization measurements were taken
every 60 seconds over 12 hours using an adapted protocol from the AOCS Official Method
Cd16b-93 for direct SFC measurements (American Oil Chemists' Society, 1998).
The direct method, when used to determine the SFC of sugar-containing blends, will result in
inflated SFC values as sugar crystals cannot be differentiated from solid fat. To circumvent
this problem, AOCS Official Method Cd 16-81 was used to determine the SFC of sugar-
containing samples (American Oil Chemists' Society, 1998). This indirect method compares
the liquid and solid signal from a sample and a control oil at the temperature of interest (20 °C)
and a temperature at which the fat is fully molten (60 °C). As a result, four separate
5
measurements are performed at two temperatures to determine SFC. This approach isolates the
contribution of sugar to the overall signal, which is then removed revealing the SFC of the fat
phase only. Here, after the melting and cooling treatments were applied to the direct method
samples above, the SFC of all cocoa butter-sugar-emulsifier blends were determined after 15
minutes and 1, 3 and 5 hours of isothermal crystallization, and compared to the results of the
direct method. The Avrami rate and onset parameters of the bulk SFC vs. time data obtained
via the direct and indirect methods were fitted to the linearized Avrami equation:
SFCt SFCeq 1 e
K ( t Tlag ) n
(1)
where SFCt and SFCeq represent the SFC at time t and at equilibrium and Tlag corresponds to
the onset time prior to the primary and secondary crystallization onsets. The value of n, the
Avrami exponent, gives an indication of the dimensionality of crystal growth and is sensitive
to the time dependence of nucleation. The parameter K, the Avrami constant, is an indicator of
crystal growth and nucleation rate. A program was written in Microsoft Excel v.14 (Redmond,
WA, USA) that allows for determination of relevant parameters (n, SFCeq, K and Tlag).
Cocoa butter and cocoa butter-sugar blend melting behavior in the absence and presence of the
Instruments, New Castle, DE, USA). Molten fat samples (~ 5 mg, corrected for the presence
of sugar) were placed in an aluminum DSC pan, hermetically sealed, and heated in an 80 °C
oven for 10 minutes, with subsequent ageing in a 20 °C incubator. Measurements were taken
at 24 hour intervals over a period of 96 hours. Thermal profiles were evaluated from 20 °C to
80 °C at a heating rate of 5 °C/min. Peak fitting and integration were performed using
6
illustrates the general approach, where curve fitting is used to differentiate the respective
enthalpic contributions of forms IV and V. The percentage enthalpy associated with form V
was determined from the total enthalpy and was used to demonstrate the relative polymorphic
Wide-angle x-ray diffraction (WAXD) was investigated using a powder X-ray diffractometer
(Miniflex 600, Rigaku, Tokyo, Japan), where molten samples (2.5 mL) were placed in 5 cm
diameter, circular aluminum pans, heated in an 80 °C oven for 10 minutes, and subsequently
sample holder. Only cocoa butter-emulsifier blends were characterized using X-ray diffraction
Fat crystal morphology following storage at 20 °C for 24 or 96 hours was observed by confocal
laser scanning microscopy (CLSM) (LSM800 Zeiss Inc., Toronto, ON, Canada) with a 63 ×
oil immersion objective. Each 1 mL sample was melted at 80 °C for 10 min and stored at 20
°C. Morphology was evaluated after 24 and 96 hours by gently scraping ~ 5 mg onto a glass
slide along with ~ 0.1 mL of a 0.1 wt% solution of Fluorol Yellow (Sigma Aldrich, St. Louis,
Missouri, USA) in canola oil which was then gently covered with a glass cover slip, and imaged
All reported results are the arithmetic mean (± standard deviation) of triplicate experiments.
Analyses were performed using one-way analysis of variance (ANOVA) with a post-hoc Tukey
HSD test to compare multiple groups. A pairwise comparison was performed using the Student
7
3. RESULTS AND DISCUSSION
Rosen and Dahanayake proposed that emulsifiers should serve specific functions. One function
is that, when added to a system, they should adsorb to a surface or interface and change its
properties (Rosen & Dahanayake, 2000). The other function is that they should aggregate in
the solvent in which they are dissolved and change the properties of the solution. The ability
of emulsifiers to influence fat crystallization has, as one of its fundamental tenets, molecular
In chocolate, ~ 0.6 wt% emulsifier is added to the final mass of chocolate to facilitate flow
However, that quantity of emulsifier is relative to the entire chocolate mass, which takes into
account sugar crystals, cocoa particles, milk powder or other particles, if present. Using a
model system with only sugar and emulsifier, this relative ratio must be taken into
consideration. Here, 2 wt% emulsifier in cocoa butter was chosen relative to cocoa butter which
Characteristic two-stage cocoa butter crystallization was observed in all blends, with a primary
stage where higher-melting TAGs crystallized and a secondary stage, where mass
(Dewettinck, Foubert, Basiura, & Goderis, 2004). The addition of either emulsifier or oil
shortened nucleation times and enhanced crystal growth rates, with the largest effects conferred
by the low molecular weight emulsifiers. Figure 1 reports the primary and secondary onsets
and rates of crystallization of cocoa butter in the presence/absence of sugar and emulsifier.
Addition of lecithin alone shortened the onset of nucleation from 25 minutes to 4 minutes.
Compared to the control cocoa butter, secondary growth occurred ~ 40 minutes earlier and its
rate increased from 2.2 × 10-6 s-1 to 2.2 × 10-4 s-1 - an increase of 100 ×. This was similar to
other efforts that have shown that lecithin accelerates cocoa butter crystallization (Ashkezary
8
et al., 2018; Miyasaki et al., 2016; Svanberg et al., 2011). In the high-melting fraction of cocoa
butter that crystallizes during primary nucleation, there tends to be a greater fraction of
phospholipids that are native to the cocoa butter, and it has been suggested that they facilitate
nuclei formation (Davis & Dimick, 1989; Rigolle et al., 2015). However, this depends on the
& Dimick 1995). Given the observed effects, it was phosphatidylcholine that dominated the
behavior of lecithin.
CITREM and AP both decreased the onset of nucleation of cocoa butter from 25 minutes to ~3
minutes and decreased the onset of secondary growth by ~ 40 minutes, similar to lecithin
(Figures 1A and 1C). One postulated mechanism that describes enhancement of fat nucleation
incorporate into the crystal lattice due to unsaturation of their constituent fatty acids (Rigolle
et al., 2015). As lecithin, CITREM and AP all displayed similar effects on nucleation, it is
likely that their polar headgroups were responsible for the shortened onset times. Other liquid-
state emulsifiers with similar unsaturation in the TAG moieties such as sorbitan monooleate
and sorbitan trioleate have shown minimal effects on cocoa butter nucleation despite bearing
the same fatty acid moieties, which further supports that headgroup polarity influences
The presence of PGPR reduced the primary onset of crystallization to ~ 11 minutes rather than
~ 25 minutes in cocoa butter alone (Figure 1A). Secondary growth also occurred earlier, after
~ 86 minutes instead of ~ 111 minutes, with the secondary rate increasing to a lesser extent
than with any other emulsifier (Figure 1D). It was surmised that PGPR’s oligomeric nature
would result in dissimilar behavior to the other emulsifiers, given its inability to form ordered
organized assemblies in solution. For comparison, the effects of 2 wt% canola oil on cocoa
9
butter crystallization were investigated and closely resembled that of PGPR. Canola oil is
primarily comprised of triolein, with a cis configuration in the olefin at the n-9 position. Due
to its kinked structure, it is also unlikely that canola oil would integrate itself into cocoa butter
crystal nuclei (Rizzo, Norton, & Norton, 2015). Given their structural dissimilarity to the
crystallizing TAGs, any changes in cocoa butter crystallization were attributed to a dilution
effect as both canola oil and PGPR increased the liquid fraction of cocoa butter, mimicking
softer cocoa butters, which also have faster nucleation rates (Foubert, Vanrolleghem, Thas,&
Dewettinck, 2004). In contrast to our results, PGPR was previously shown to have no tangible
effect on cocoa butter crystallization rate (Miyasaki et al., 2016). This discrepancy may have
arisen from differences in concentration, where up to 0.8 % PGPR was used by Miyasaki et al.
compared to 2 wt% in the present study. As well, the present samples were isothermally
driving force for crystallization is greater which would potentially diminish the effect of an
additive (Smith et al., 2011), however this would greatly depend on cooling rate (Bowser,
2006).
The Avrami constant n describes the dimensions that crystal growth occurs, where lower values
of n indicate growth in fewer directions, i.e., rod-like crystal growth, whereas higher values of
n indicate disc-like or spherulitic growth (Avrami, 1939, 1940; Wright, Hartel, Narine, &
Marangoni, 2000). The addition of emulsifier and oil did not have a significant effect on the
primary growth mode with all values remaining at n ~ 1, indicating rod-like growth from
instantaneous nuclei. Secondary growth n values in the presence of lecithin, AP and CITREM
were also ~ 1.0 whereas cocoa butter alone and in the presence of PGPR or canola oil yielded
n ~ 1.4.
Early-stage SFCeq showed a statistically insignificant increase from 4.1 to 4.5% in the presence
of emulsifiers, with no increase upon addition of canola oil (p ≥ 0.05). Second-stage SFCeq
10
showed similar trends, as canola oil showed no increase in SFC whereas the emulsifiers
no crystallizable species, the increase in SFC was strictly due to the solidification of the TAGs
the primary growth phase) after 1 and 3 hours (representative of the secondary growth phase)
and after 5 hours (representative of SFCeq). Emulsifier presence modestly increased the SFC
of the cocoa butter after 15 min and 1 hour. After 3 hours, the emulsifiers greatly increased
cocoa butter’s SFC of ~ 17 % to values of 50 - 53% for the low molecular weight emulsifiers
whereas PGPR and canola oil enhanced it to 32 – 36 %. At 5 hours, cocoa butter SFC was ~
and PGPR as well as canola oil SFC values of 47 – 51 %. These results showed that all
The presence of sugar lowered the cocoa butter primary and secondary nucleation onsets from
25 to 9 minutes and from 111 to 64 minutes, respectively (‘CB + Sugar’ results in Figures 1A
and 1C). Secondary growth rate increased over 280 ×, demonstrating that sugar greatly
nucleation (Figures 1B and 1D) (Dhonsi & Stapley, 2006; Svanberg et al., 2011). Interestingly,
sugar did not impact cocoa butter’s primary rate of nucleation, which suggested that the higher-
melting TAGs in cocoa butter were not affected by the presence of sugar.
It was expected that the presence of sugar would limit the ability of emulsifier molecules to
associate with each other and template fat crystallization, owing to their preferential interaction
with the sugar crystal surface. Across all emulsifiers tested, sugar muted any effect added
emulsifier had on crystallization kinetics (using the direct method), resulting in similar Avrami
11
K values regardless of the emulsifier added (p ≥ 0.05). The onset of primary and secondary
growth across all samples occurred within 3 - 8 minutes and 69 - 81 minutes, respectively. The
rate constant, K, showed similar values across samples, 9 × 10-4 s-1 for primary growth and ~ 4
– 10 × 10-4 s-1 for secondary growth. These results suggested that the surface effects of sugar
were dominating crystallization events, and the ability of any emulsifier to affect crystallization
was diminished either through their immobilization at the surface of sugar or simply through
the magnitude of the effect of sugar. Presence of sugar did not change the calculated mode of
growth in the primary stage where, similar to samples without sugar, values suggested rod-like
Table 1 contrasts the differences in SFC of the emulsifier-containing cocoa butter-sugar blends
at 15 min and after 1, 3 and 5 hours. The presence of sugar alone enhanced cocoa butter
crystallization most notably at 1, 3, and 5 hours, with increases of 12, 44, and 25 %,
respectively. The combination of sugar and emulsifier did yield greater SFCs in the early stages
Addition of lecithin to the cocoa butter-sugar blend increased SFC to a lesser extent than either
AP and CITREM, but all three showed a more substantial effect than canola oil and PGPR.
The largest difference was seen at the 1 hour mark, with the emulsifiers increasing SFC from
increase. By 3 hours, all SFC values were 57 – 62 % and by 5 hours, all samples were at 64 –
66 % SFC, demonstrating that the emulsifier effects were more prevalent during the initial
period of stage 2 crystallization. Due to the speed of primary crystallization and the use of
standard solutions involved in indirect measurements, it was difficult to obtain primary rate
12
Figure 2 contrasts the fat crystal morphology of bulk cocoa butter isothermally crystallized for
96 hours at 20 °C with that observed in a 1:1 cocoa butter:sugar blend solidified under the same
conditions. In the bulk, cocoa butter crystallized in an unimpeded fashion based on the
spherulitic morphology and presence of fine needles jutting into surrounding molten fat (Figure
2A). Addition of any emulsifier to the cocoa butter did not appreciably alter the resulting fat
crystal morphology or size, with spherulitic crystals dominating all samples (Supplemental
Figure 3).
The typical spherulite size of the cocoa butter fat crystals in Figure 2A (D ~ 50 μm) was
significantly reduced in the presence of sugar, where the average crystal size was well-below
~ 50 μm (Figure 2B). After storage for 96 hours, the network of the cocoa butter-sugar blends
were composed of haphazardly organized sugar and needle-shaped fat crystals (Figure 2B).
Restricted fat crystal growth was apparent, and presumably resulted from the limited physical
space available, as newly-formed fat crystals were now forced to grow within a physically-
confined environment, given the significant volume fraction occupied by the dispersed sugar
Cocoa butter alone showed a majority of form IV crystals with d-spacings of 4.32 and 4.13 Å
after 24 hours of isothermal storage at 20 °C, similar to Chapman et al. (Chapman, Akehurst,
& Wright, 1971) (Supplemental Figure 4). Storage for 96 hours revealed a gradual transition
from form IV to V, given the presence of a d-spacing at 4.53 Å. The form V polymorph is
associated with a strong peak at 4.5 – 4.6 Å accompanied by fingerprint peaks in the 3.7 – 4.0
Å region. The melting points for all fat-emulsifier blends ranged from 26.4 – 27.2°C for form
Emulsifiers that share structural similarities with the solidifying fat may integrate themselves
into the fat crystal lattice, allowing them to modify the properties of the crystal rather than that
13
of the oil phase. Conversely, dissimilar emulsifiers will not be able to change the properties of
fat crystals directly and may instead alter crystallization by increasing the liquid fraction and
molecular motion via dilution, or by changing the polarity of the mother liquor (Garti,
Schlichter, & Sarig, 1986). The present results strongly suggested that all emulsifiers were not
integrated within the cocoa butter crystal lattices, which was expected given their molecular
dissimilarity and/or molecular weight difference (Norton, Lee-Tuffnell, Ablett, & Bociek,
1985). Elsewhere, Ueno and coworkers showed that mixing palm mid-fraction with 5%
sorbitan tripalmitate, which share the same palmitic acid fatty acid moiety, resulted in mixed
melting behavior given the emergence of a melting peak distinct from either palm mid-fraction
or sorbitan tripalmitate (Ishibashi, Hondoh, & Ueno, 2018). By contrast, sorbitan tristearate
and sorbitan tribehenate, showed no influence on the melting of palm mid-fraction given their
molecular dissimilarity.
Figure 3 shows the melting profile of cocoa butter isothermally aged at 20 °C for 96 hours.
Under these conditions, cocoa butter will initially crystallize in the form II polymorph and
transition to form IV within a few hours (Dewettinck et al., 2004). Here, a combination of
deconvolution, curve fitting, and integration applied to the thermograms was used to determine
the relative enthalpies of the form IV and V polymorphs (Figure 4). The enthalpies were then
converted into a ratio of polymorphs, as they relate to ageing and composition. By following
the transition from form IV to form V over a period of 96 hours, we determined the ability of
the emulsifier and sugar, and combinations thereof, to alter this polymorphic transition.
Figure 4A compares the evolution in the proportion of form V in the cocoa butter-emulsifier
blends as a function of storage time following isothermal storage at 20 °C over 96 hours. Cocoa
butter alone showed the lowest proportion of form V after 96 hours as all emulsifiers, as well
as the canola oil, promoted the form IV to V transition. Lecithin and AP promoted form V
formation in a non-linear manner, given the increase from ~ 5-10 % at 24 hours to ~ 65-70 %
14
after 96 hours, with a large increase between 24 and 48 hours. The CITREM and PGPR
promoted the formation of form V in a similar manner (from 25-30 % after 24 hours to ~ 80 %
after 96 hours).
Canola oil also promoted the form VI to V transition likely through a dilution mechanism, with
only slight changes in form V content over 96 hours, as after 24 hours ~ 59 % of the cocoa
butter had transition to form V. This result was expected as fat crystallization from a solvent
will generally lead to the formation of more stable polymorphs, typically the β polymorph
(Arishima & Sato, 1989). By considering the canola oil as a solvent, albeit at only 2 wt% of
the cocoa butter, it diluted the cocoa butter enough to provide greater space for the TAGs to
Neither the addition of emulsifier or oil affected the total melting enthalpy of cocoa butter (p
< 0.05). When encompassing the contributions of both the form IV and V polymorphs, cocoa
butter melting enthalpy rose from 102 to 131 J/g over 96 hours. The addition of all emulsifiers
resulted in similar trends in enthalpy, ranging from 99 – 102 J/g after 24 hours to 124 - 140 J/g
after 96 hours. The only outlier was AP, which yielded melting enthalpies of 124 J/g after 3
and 4 days, suggesting a hindrance of the form IV to V transition. With canola oil, the overall
melting enthalpy after 24 hours was ~ 116 J/g and 130 J/g after 96 hours.
Presence of sugar alone hindered the form IV to V polymorphic transition, both in rate and
extent (Figure 4B). After 24 hours, the cocoa butter-sugar blend solely consisted of form IV,
which gradually increased to ~ 49 % after 96 hours. Similar to the nucleation and growth results
above, sugar generally minimized the effect of added emulsifier, with most emulsifiers yielding
proportions of 65 - 75 % form V after 96 hours. The same effect was seen with canola oil,
which was surprising, as it was expected that the oil would remain free in the fat phase as it is
15
environment at the surface of the sugar crystals, which would reduce molecular motion
(Acevedo, Block, & Marangoni, 2012; Arishima & Sato, 1989; Lishchuk & Halliday, 2009).
The only exception to the rule was the mixture of AP + sugar, which completely blocked the
form IV to V transition over 96 hours. In the absence of sugar, AP behaved similarly to lecithin
with a minimal effect on the cocoa butter form IV-to-V polymorphic transition. Yet, when
combined sugar, there was a synergistic effect between the sugar and AP. Figure 4B shows that
AP prevented the transition from form IV to form V, perhaps by limiting the fluidity of the
continuous cocoa butter phase (Garti & Sato, 2001). Previously, AP was shown to limit the
absorption of water to sugar crystals in chocolates submerged in alcohol syrups (Holdgaard &
Wikman, 2003). This was attributed to a tight adsorption of AP to the surface of the sugar
crystals, but also supports the concept of a less plastic continuous phase, hindering the transfer
of water to the surface of the sugar crystals. Why AP caused this only in the presence of sugar
is unknown. Amongst the emulsifiers tested, AP had the smallest headgroup and lowest
compositional variability, which both likely contributed to its ability to efficiently pack on the
available sugar crystal surface and organize in the continuous phase around these. This ability
would increase the local micro-viscosity around the sugar crystals unlike the other emulsifiers,
thereby slowing the rate of cocoa butter form IV towards form V. Clearly, this topic warrants
further scrutiny.
The combined effect of sugar and emulsifier on the total enthalpy of melting of cocoa butter
crystals was limited. As noted, cocoa butter alone showed melting enthalpies of 102 - 131 J/g
over a period of four days. In the presence of all emulsifiers, the cocoa butter-sugar blend
showed similar enthalpies ranging from 90 – 120 J/g after 24 hours, and 101 – 139 J/g after 96
hours. Canola oil did reduce the overall enthalpy slightly, but this was not surprising as it
increased the liquid TAG fraction in the samples, thereby reducing the amount of fat that could
crystallize. Finally, the presence of sugar had no effect on the melting point of the form IV or
16
V polymorphs either alone or in the presence of emulsifier, as melting points ranged from 26.8
– 27.2 °C for form IV and from 30.3 – 31.0 °C for form V, respectively (p > 0.05).
Figure 5 depicts possible mechanisms relevant to the role of emulsifiers and sugar in the
systems studied. Emulsifier clustering in solution forming templates for crystallization to occur
has previously been reported for phospholipids in cocoa butter (Figure 5A) (Davis & Dimick,
1989). Rizzo, Norton and Norton also described the formation of mono and di-acylglycerol
micelles in a coconut - sunflower oil blend, which enhanced crystallization (Rizzo et al., 2015).
The ability for lecithin, CITREM and AP to enhance cocoa butter crystallization may be
attributed to this clustering and templating phenomenon. The PGPR and canola oil were
unlikely to participate in this behavior due to their structures and polarity, respectively, as they
are not conducive to the formation of an ordered region that would promote fat crystallization
(Figure 5B). Norton and coworkers described the exclusion of triolein during the crystallization
of tristearin and tripalmitin, with the presence of triolein also promoting the formation of
higher-melting polymorphs, similar to the present effects of PGPR and canola oil (Norton et
al., 1985). Sonwai et al. saw similar enhancement of cocoa butter crystallization and the
promotion of higher melting polymorphs in the presence of canola oil, and attributed the result
The presence of sugar introduced a surface that may induce TAG heterogeneous nucleation
(Figure 5C). Yoshikawa and coworkers saw crystallization accelerated with the introduction
of 1% dispersed solid particles in cocoa butter (Yoshikawa et al., 2014). Sugar surfaces also
provide sites for emulsifiers to adsorb to, which changes the lipophilicity of the coated sugar
crystals (Rousset, Sellappan, & Daoud, 2002). Though the Avrami-based values for the
solidification in the presence of sugar, SFC values in the presence of sugar (with or without
17
emulsifier) were similar to, or greater, than that of cocoa butter and emulsifier alone at
intermediate times. However the form IV to V polymorphic transition was retarded, suggesting
the existence of a synergistic effect between the sugar and emulsifier that merits further
investigation. West and Rousseau saw similar trends in palm oil whereby sugar enhanced
crystallization and hindered polymorphic evolution (West & Rousseau, 2017). Similarly,
Acevedo, Block and Marangoni observed that increased viscosity decreased molecular
mobility during crystallization and promoted the formation of less stable polymorphs (Acevedo
et al., 2012).
Finally, canola oil’s ability to influence cocoa butter crystallization was likely due to a dilution
effect rather than surface activity. Thus, the presence of sugar would presumably not impact
its ability to accelerate polymorphic transitions. Since sugar hindered polymorphic evolution
even in the presence of canola oil, its effects were likely due to an increase in viscosity and a
decrease in molecular motion, rather than the association of additives to the surface of sugar
(Figure 5D).
4. CONCLUSION
We have observed the effects of 2 wt% emulsifier (on a fat basis) on cocoa butter nucleation,
crystal growth and polymorphism both in the presence and absence of sugar. We demonstrated
that all low molecular weight emulsifiers promoted crystallization of bulk cocoa butter during
the early stages of nucleation and growth to a similar extent despite the variety in polar
headgroup. Both PGPR and canola oil modestly enhanced nucleation and growth possibly
suggesting dilution as a consequence of their presence rather than an integration into the cocoa
butter crystal lattice. The low molecular weight emulsifiers demonstrated a minimal effect on
the form IV to V transition, given their liquid state and lack of molecular complementarity
whereas PGPR behaved similarly to added oil, accelerating the form IV to V transition, and
18
behaved as a solvent. The presence of sugar accelerated cocoa butter nucleation and growth
but muted the effect of emulsifiers, which was surmised as resulting from their adsorption to
the sugar crystal surface. The sugar retarded the form IV to V transition both in the presence
and absence of emulsifiers, potentially through a form IV stabilization mechanism with this
phenomenon enhanced in the presence of AP. Despite the variety of emulsifiers explored, there
was no significant impact on the sub-cell packing due to the inability of an emulsifier to become
integrated within the crystal lattice of the cocoa butter crystals as corroborated by the thermal
analysis. We demonstrated that, though the presence of sugar does indeed limit the ability of
emulsifiers to impact cocoa butter crystallization, there does exist and unknown cooperative
mechanism between AP and sugar that will retard the kinetics of polymorphic transition in
cocoa butter, from form IV to form V. Future work will continue to explore the mechanistic
relationship between emulsifier structure, presence of sugar and effects on fat crystallization
in greater depth.
ACKNOWLEDGMENTS
The Natural Sciences and Engineering Research Council of Canada (NSERC) is gratefully
19
REFERENCES
modification of the mechanical strength and oil binding capacity of a model edible fat
https://doi.org/10.1021/la303365d
Ačkar, D., Škrabal, S., Šubarić, D., Babić, J., Miličević, B., & Jozinović, A. (2015).
Arishima, T., & Sato, K. (1989). Polymorphism of POP and SOS III. Solvent crystallization
of β2 and β1 polymorphs. Journal of the American Oil Chemists Society, 66(11), 1614–
1617. https://doi.org/10.1007/BF02636187
Ashkezary, M. R., Yeganehzad, S., Vatankhah, H., Todaro, A., & Maghsoudlou, Y. (2018).
Avrami, M. (1939). Kinetics of phase change. I: General theory. The Journal of Chemical
https://doi.org/10.1089/jam.2007.0571
124–127.
Chapman, G. M., Akehurst, E. E., & Wright, W. B. (1971). Cocoa butter and confectionery
20
fats. Studies using programmed temperature X-ray diffraction and differential scanning
https://doi.org/10.1007/BF02609292
Davis, T. R., & Dimick, P. S. (1989). Lipid composition of high-melting seed crystals formed
during cocoa butter solidification. Journal of the American Oil Chemists Society, 66(10),
1494–1498. https://doi.org/10.1007/BF02661979
Dewettinck, K., Foubert, I., Basiura, M., & Goderis, B. (2004). Phase behavior of cocoa
butter in a two-step isothermal crystallization. Crystal Growth and Design, 4(6), 1295–
1302. https://doi.org/10.1021/cg049772n
Dhonsi, D., & Stapley, A. G. F. (2006). The effect of shear rate, temperature, sugar and
emulsifier on the tempering of cocoa butter. Journal of Food Engineering, 77(4), 936–
942. https://doi.org/10.1016/j.jfoodeng.2005.08.022
Foubert, A.I.; Vanrolleghem, P.A.; Thas, O.; Dewettinck, K. (2004). Influence of Chemical
69(9), 478–487.
Garti, N., & Sato, K. (2001). Crystallization rocesses in fats and lipid systems. Marcel
Dekker. https://doi.org/10.1017/CBO9781107415324.004
Garti, N., Schlichter, J., & Sarig, S. (1986). Effect of food emulsifiers on polymorphic
transitions of cocoa butter. Journal of the American Oil Chemists’ Society, 63(2), 230–
236. https://doi.org/10.1007/BF02546144
Holdgaard, J., & Wikman, H. H. (2003). Alcohol resistant chocolate. Palsgaard A/S.
Ishibashi, C., Hondoh, H., & Ueno, S. (2017). Epitaxial growth of fat crystals on emulsifier
crystals with different fatty acid moieties. Crystal Growth and Design, 17(12), 6363–
6371. https://doi.org/10.1021/acs.cgd.7b01039
Ishibashi, C., Hondoh, H., & Ueno, S. (2018). Influence of Fatty Acid Moieties of Sorbitan
21
Esters on Polymorphic Occurrence of the Palm Mid-Fraction. Journal of the American
Lishchuk, S. V., & Halliday, I. (2009). Effective surface viscosities of a particle-laden fluid
interface. Physical Review E - Statistical, Nonlinear, and Soft Matter Physics, 80(1), 1–
7. https://doi.org/10.1103/PhysRevE.80.016306
http://ezproxy.lib.ryerson.ca/login?url=https://www.proquest.com/docview/301586980?
accountid=13631
Man, Y. B. C., Rahman, R. A., Yusoff, M. S. A., Aini, I. N., & Miskandar, M. S. (2007).
Miyasaki, E. K., Luccas, V., & Kieckbusch, T. G. (2016). Modified soybean lecithins as
Norton, I. T., Lee-Tuffnell, C. D., Ablett, S., & Bociek, S. M. (1985). A Calorimetric, NMR
and X-Ray Diffraction Study of the Melting Behavior of Tripalmitin and Tristearin and
their Mixing Behavior with Triolein. Journal of the American Oil Chemists’ Society,
Rigolle, A., Gheysen, L., Depypere, F., Landuyt, A., Van Den Abeele, K., & Foubert, I.
https://doi.org/10.1002/ejlt.201400555
22
Rizzo, G., Norton, J. E., & Norton, I. T. (2015). Emulsifier effects on fat crystallisation. Food
Rosen, M. J., & Dahanayake, M. (2000). Industrial Utilization of Surfactants: Principles and
Practice.
University of Strathclyde.
Rousset, P., Sellappan, P., & Daoud, P. (2002). Effect of emulsifiers on surface properties of
101. https://doi.org/10.1016/S0021-9673(02)00370-9
http://www.gomc.com/firstpage/199505127.pdf
Smith, K. W., Bhaggan, K., Talbot, G., & Van Malssen, K. F. (2011). Crystallization of fats:
Influence of minor components and additives. Journal of the American Oil Chemists’
Sonwai, S., Podchong, P., & Rousseau, D. (2017). Crystallization kinetics of cocoa butter in
https://doi.org/10.1016/j.foodchem.2016.07.092
Svanberg, L., Ahrné, L., Lorén, N., & Windhab, E. (2011). Effect of sugar, cocoa particles
and lecithin on cocoa butter crystallisation in seeded and non-seeded chocolate model
https://doi.org/10.1016/j.jfoodeng.2010.09.023
West, R., & Rousseau, D. (2017). Modelling sugar, processing, and storage effects on palm
oil crystallization and rheology. LWT - Food Science and Technology, 83, 201–212.
23
https://doi.org/10.1016/j.lwt.2017.05.021
Wright, A. J., Hartel, R. W., Narine, S. S., & Marangoni, A. G. (2000). Effect of minor
components on milk fat crystallization. Journal of the American Oil Chemists’ Society,
Yoshikawa, S., Kida, H., & Sato, K. (2014). Promotional Effects of New Types of Additives
https://doi.org/10.5650/jos.ess13155
24
FIGURE LEGEND
Figure 1 Avrami model results of the effect of 2 wt% emulsifier on the isothermal
crystallization of cocoa butter. (A) Primary onset; (B) Primary rate; (C) Secondary
Figure 2 Morphology of cocoa butter alone (A) and mixed with sugar (B) following 96
hours of isothermal crystallization at 20 °C. Dark areas in (B) are sugar crystals.
for up to 96 hours. For each time point, the left endotherm is associated with form
IV and the right endotherm with form V. Peak integration was used to determine
Figure 4 Evolution in the proportion of cocoa butter that has transitioned from form IV to V
(A), dilution (B), heterogeneous nucleation (C) or viscosity (D) effects. Yellow
micelles; B) Canola oil and PGPR are triskelion-like graphics; C) and D) Sugar
25
Figure 1:
30 14
A B
12
25
Primary Onset (minutes)
8
15
6
10
4
5
2
0 0
CB CITREM AP Lecithin PGPR Canola oil CB + Sugar CB CITREM AP Lecithin PGPR Canola oil CB + Sugar
125 800
C D
700
Secondary Onset (minutes)
100
500
75
400
50
300
200
25
100
0 0
CB CITREM AP Lecithin PGPR Canola oil CB + Sugar CB CITREM AP Lecithin PGPR Canola oil CB + Sugar
26
Figure 2:
A B
27
Figure 3
24 Hours
2.5
48 Hours
72 Hours
2.0 96 Hours
Heat Flow (W/g)
1.5
1.0
0.5
0.0
20 25 30 35 40
Temperature °C
28
Figure 4
29
Figure 5
30
Table 1: SFC values at set time points showing the combined effect of sugar and emulsifier on
cocoa butter crystallization. SFC values were measured with either the direct method (a) or the
31
Supplemental Figure 1:
Heat Flow (W/g)
Form IV Form V
Temperature (°C)
Figure 1 Illustration of the curve fitting used to separate the contributions of the form IV
32
Supplemental Figure 2:
80
70
60
Solid Fat Content (%)
50
40
CB
CB+Lec
30
CB+PGPR
CB+CITREM
20 CB+AP
CB+canola
10
0
0 2 4 6 8 10 12
Time (hours)
Figure 2 Cocoa butter isothermal crystallization at 20 °C for 12 h using pulsed NMR in the
presence of 2 wt% emulsifier or canola oil. Error bars removed for clarity (n=3).
33
Supplemental Figure 3
Figure 3 Morphology of cocoa butter alone (top) and mixed with sugar (bottom) following 96
hours of isothermal crystallization at 20 °C. Dark, geometric areas are sugar crystals.
Micrographs taken at 100x magnification, and are part of, at least, triplicate experiments that
34
Supplemental Figure 4
35
Figure 3 Wide-angle X-ray diffraction of isothermally-aged cocoa butter at 20 °C after 24
36
Supplemental Table 1
Declaration of interests
☒ The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.
☐The authors declare the following financial interests/personal relationships which may be
considered as potential competing interests:
37
HIGHLIGHTS
Low molecular weight emulsifiers promote cocoa butter crystallization
Emulsifiers do not alter cocoa butter polymorphic behavior
Sugar accelerates cocoa butter crystallization, but slows its form IV to V transition
The presence of sugar mutes the effects of emulsifiers on crystallization
38