Investigation into the Microstructure, Texture

and Rheological Properties of Chocolate Ganache

Jade McGill and Rich W. Hartel

Abstract: Ganache is a mixture of chocolate and dairy. Although a popular confection, little is known about how it
functions as a system. Objectives were to (1) determine if dairy fats and cocoa butter mix in ganache, (2) characterize
ganache microstructure, and how structure affects texture and rheology, and (3) identify how changes in chocolate
composition alter ganache. Textural analysis, differential scanning calorimetry, stress sweep tests, and microscopy were
used to examine ganache formulations that varied in dairy source (cream or butter) or in solid fat content (SFC),
composition or type of chocolate. Melting temperatures for all ganache formulations were lower than for chocolate,
indicating that cream milk fat globules rupture during processing, and mix with cocoa butter. Altering the SFC of
chocolate affected ganache hardness, spreadability, melting enthalpy, and resistance to deformation. Chocolate systems
made with constant fat content and greater amounts of defatted cocoa powder relative to sugar or nonfat milk powder
yielded ganache that was harder, less spreadable, and more resistant to deformation. Ganache made with commercially
produced dark, milk, and white chocolates behaved similarly to model chocolate systems. Ganache attributes are affected
by chocolate crystalline fat content in addition to particle phase volume—greater levels of cocoa powder, which is mostly
insoluble, strengthens ganache structure, producing a firmer product, whereas greater levels of milk powder and sugar,
which dissolve in the aqueous cream component, produce a softer ganache.

Food Engineering, Materials

Science, & Nanotechnology
Keywords: chocolate, ganache, microstructure, rheology, texture

Practical Application: Understanding how ganache functions as a system and how differences in chocolate composition
affect its textural and rheological properties may allow for greater control over the desired characteristics of the final
product. For example, this research shows how changing cocoa content of the chocolate affects ganache, which is useful
when developing formulations involving chocolates with different cocoa percentages. There may also be cost saving
implications; for example, using a chocolate with a harder cocoa butter may allow less total chocolate to be used in a
formulation, while still achieving an appropriate texture.

Introduction then stirring the mixture until a uniform consistency is attained.

Ganache is a mixture of chocolate and dairy, usually cream. Most
commonly used as a filling for truffles and molded chocolates, it is
also frequently used as a component of pastries; for example, as a
tart filling, cake glaze, or sandwiched between cookies. Although
it is a popular confection that has been produced for over 150 y
(Robuchon, 2009), very little scientific research has been pub-
lished about ganache. The existing literature touches on fungal
growth (Bensoussan, Alcaraz, & Tourtel, 1997), calorie reduc-
tion through hydrocolloid use (Dias, Alvarenga, & Sousa, 2015),
and the effect of chocolate temper and production methodology
on texture (Greweling, 2007). However, none focuses on under-
standing ganache at a fundamental level—its microstructure, how
various components interact, or how changing certain ingredients
affects physical properties such as texture.
According to Greweling (2007, 2013), ganache is an oil-in-
water emulsion formed by the addition of cream to chocolate.
There are several methods of making ganache. One involves pour-
ing cream just off the boil over solid chocolate at room temper-
ature, allowing the heat of the cream to melt the chocolate, and
JFDS-2017-1639 Submitted 10/3/2017, Accepted 12/26/2017. Authors are with
Dept. of Food Science, Univ. of Wisconsin- Madison, 1605 Linden Dr., Madison,
Wis. 53706, U.S.A. Direct inquiries to Hartel (E-mail: rwhartel@wisc.edu).
Alternatively, cream heated to 40 °C can be added to liquid, but
tempered, chocolate followed by mixing. Traditionally, a ratio
of 2:1 chocolate to cream is used for dark/semi-sweet chocolate
ganache, whereas 2.5 or even 3:1 is required to make a good
ganache with milk or white chocolate.
Chocolate is a dispersion of particles (sugar crystals, cocoa par-
ticles, and potentially milk powder) in a cocoa butter continuous
phase whereas cream is an oil-in-water emulsion with fine milk
fat globules dispersed in a continuous phase of skim milk. Ag-
itation during mixing of warm cream and chocolate to create
ganache provides sufficient energy to create globules of fat, most
likely emulsified by the lecithin in the chocolate and/or the pro-
teins within the cream, within the aqueous phase provided by the
cream. The fate of the cream fat globules during mixing is un-
known. Further, if butter is added to the ganache, as called for
in some formulations, the milk fat from the butter must also be
incorporated into an emulsion as well. It is unknown whether or
not the milk fat from cream and butter mix well with the cocoa
butter or retain their own identity as separate, distinct emulsion
droplets. The cocoa butter, while mostly melted, must contain
sufficient β-V crystals to seed the cocoa butter upon cooling to
room temperature. If not, a grainy texture appears in the finished
ganache (Greweling, 2007).
Sugar crystals and any milk powders in the chocolate dis-
solve, at least partially, in water in the cream. The fate of cocoa

C 2018 Institute of Food Technologists

 R

doi: 10.1111/1750-3841.14053 Vol. 00, Nr. 00, 2018 r Journal of Food Science 1
Further reproduction without permission is prohibited
 Investigation of chocolate ganache . . .

particles from the chocolate is unknown, but one might sus- Table 1–Butter and cream ganache formulations. Ratios are ex-
pect they partition between oil and water phases based on their pressed as the portion of milk fat sourced from either C – cream
or B – butter.
hydrophobic/hydrophilic regions. The exact microstructure of
ganache and how the distribution of these structures influence Ingredient (%) 1:0 C:B 3:2 C:B 3:7 C:B 0:1 C:B
textural properties has largely been unstudied. Semisweet chocolate 66.67 66.67 66.67 66.67
Understanding ganache at a more fundamental level would al- Cream 33.33 20.00 10.00 0.00
low for those who make it to have more control over the finished Butter 0.00 5.93 10.38 14.83
product. Within the confectionery community, there has been a Deionized water 0.00 6.64 11.63 16.61
MPCa 0.00 0.33 0.57 0.82
call to expand the scientific knowledge of ganache precisely for Lactose 0.00 0.40 0.70 1.00
this reason (Greweling, 2007). Therefore, the aims of this research Milk minerals 0.00 0.03 0.05 0.07
were to (1) determine if dairy fats and cocoa butter mix in ganache, Total 100.00 100.00 100.00 100.00
(2) characterize ganache microstructure, and how structure affects a
MPC, milk protein concentrate.
texture and rheology, and (3) identify how changes in chocolate
composition, either to the fat phase or to dispersed particles in Table 2–Milk fat modified chocolate formulations. CB, cocoa
chocolate, affect the physical properties of ganache. butter. MFF, milk fat fraction with 15 °C melting point. Column
headers refer to the ratios of fat added to the chocolate mass,
which contained 28% cocoa butter.
Materials and Methods
1:0 3:2 2:3 0:1
Materials Ingredient (%) CB:MFF CB:MFF CB:MFF CB:MFF
Chocolate mass, cocoa butter, and semisweet, milk, and
Chocolate mass 90.98 90.98 90.98 90.98
white chocolate were supplied by Guittard Chocolate Company Cocoa butter 9.02 5.41 3.61 0.00
(Burlingame, CA, U.S.A.). Unsalted butter was purchased from the MFF 0.00 3.61 5.41 9.02
Food Engineering, Materials
Science, & Nanotechnology

same lot at Metcalfes Sentry Grocery Store (Madison, WI, U.S.A.) Total 100.00 100.00 100.00 100.00
and kept frozen until needed. Fresh heavy cream (Kemps LLC,
Cedarburg, WI, U.S.A.) was procured from Babcock Dairy Hall Table 3–Formulations altering ratios of sugar and cocoa powder
(Madison, WI, U.S.A.). Milk minerals were provided by Glanbia in model chocolates. Batch size = 700 g. CP, cocoa powder; S,
Nutritionals (Fitchburg, WI, U.S.A.). Milk protein concentrate sugar.
(90%), nonfat milk powder, and lactose were provided by Food In- Ingredient (%) 0:1 CP:S 1:2 CP:S 2:1 CP:S 1:0 CP:S
gredients Inc. (Waukesha, WI, U.S.A.). A milk fat fraction (MFF)
with a melting point of 15 °C was provided by Nutrical, S.A. de Cocoa butter 37.00 37.00 37.00 37.00
Sugar 62.50 41.67 20.83 0.00
C.V. (Mexico City, Mexico). Additional cocoa butter for the milk Cocoa powder 0.00 20.83 41.67 62.50
fat modified ganache experiments was provided by Clasen Quality Lecithin 0.50 0.50 0.50 0.50
Chocolate (Madison, WI, U.S.A.). ADM Cocoa supplied defat- Total 100.00 100.00 100.00 100.00
ted cocoa powder (Milwaukee, WI, U.S.A.). The cane sugar used
was Domino brand extra fine granulated (Domino Foods Inc., fat at 20 °C) milk fat fraction (MFF) to create chocolates of 34.5%
Yonkers, NY, U.S.A.). Clearate R
B-60 soy lecithin was provided total fat but with a range of SFC. Here, 26% of the fat phase in the
by W.A. Clearly Product, Inc. (Somerset, NJ, U.S.A.). Water was final chocolate was made up of the added fat. These chocolates
deionized. were used to make ganache using the cream formulation used in
the previous Cream/Butter ganache section (Table 2). Although
Ganache variations actual SFC of the final chocolates could not be measured, the
To better understand the roles of the microstructural elements sample with the highest MFF addition had the lowest SFC and
on ganache properties, several different formulations were evalu- this increased as the ratio of cocoa butter added was increased.
ated with specific ratios of components. Chocolate model system ganache. Model chocolate sys-
Cream/butter ganache. To determine if the milk fat glob- tems were made to examine the effects of ratios of different par-
ules in cream rupture and mix with cocoa butter during ganache ticles on ganache structure, rheology, and texture. Defatted cocoa
manufacturing, mass balanced formulations with fat sourced from powder (CP) and nonfat milk powder (NFMP) were used so that
cream (intact fat globules) and/or butter (fragmented milk fat glob- cocoa butter was the only fat in the model chocolate systems, and
ules) were compared. A traditional 2:1 semisweet chocolate: cream the total amount of fat remained constant. In one set of experi-
formulation was used as the base for all experiments. Cream con- ments, the ratio of CP to sugar (S) was altered with cocoa butter
tains 36.1% fat and butter contains 81.1% fat. Formulations were and lecithin remaining constant (Table 3). These spanned from no
designed to contain the same 36.1% fat contributed by dairy, ei- sugar (akin to unsweetened chocolate), to no CP (a simple sugar-
ther from all cream or butter, or a combination of both (Table 1). cocoa butter system). In the second set of experiments, NFMP
Water, milk proteins, milk minerals, and lactose were balanced to to CP ratios were altered with sugar, cocoa butter and lecithin
account for differences between cream and butter using data from quantities remaining constant (Table 4). The formulation with no
the USDA Food Composition Database. NFMP approximates a dark chocolate, that with all CP approxi-
Milk fat modified ganache. Different chocolates can vary mates a white chocolate, and intermediate formulations with both
in solid fat content (SFC) based on ingredient composition and CP and NFMP were akin to milk chocolate.
manufacturing practices and these differences in SFC could alter Particle size is known to affect chocolate rheology (Beckett,
ganache attributes. To determine the effects of changing SFC, a 2008). To minimize the effect of particle size, chocolate systems
low-fat chocolate mass (28% cocoa butter) was mixed with differ- were produced in a Premier Wonder Table Top Wet Grinder
ent ratios of either cocoa butter (approximately 69% solid fat at (SS Premier, Chennai, India), which can be used to refine small
20 °C as measured by pulsed-NMR) and a low melting (0% solid batches of chocolate. Cocoa butter was fully melted to 45 °C. This

2 Journal of Food Science r Vol. 00, Nr. 00, 2018

Investigation of chocolate ganache . . .
Table 4–Formulations altering ratios of milk powder and cocoa equal to the heat removed by the system. Slopes within the range of
powder in model chocolates. Batch size was 700 g. CP, cocoa ±1.0 are generally considered adequately tempered; for increased
powder; MP, milk powder.
Ingredient (%) 0:1 CP:MP 1:2 CP:MP 2:1 CP:MP 1:0 CP:MP
Cocoa butter 37.00 37.00 37.00 37.00
Sugar 41.67 41.67 41.67 41.67
Cocoa powder 0.00 6.94 13.89 20.83
Milk powder 20.83 13.89 6.94 0.00
Lecithin 0.50 0.50 0.50 0.50
Total 100.00 100.00 100.00 100.00
and lecithin were added to the wet grinder, which was positioned
under a 120 V/60 Hz fluorescent light bulb to maintain fluid-
ity while processing. With the machine running, CP, sugar, and
NFMP were gradually added to the wet grinder. A Master heat
gun (Master Appliance Corp., Racine, WI, U.S.A.) warmed the
bowl of the wet grinder until all ingredients were incorporated,
and friction plus lamp heat was sufficient to maintain fluidity. The
grinder was gradually adjusted to the finest level of refining and left
to run for 24 hours, covered. The sides of the bowl and grinding
stones were scraped down occasionally as needed. Finished choco-
late was cooled at room temperature on parchment paper, then
stored in plastic bags until needed. One batch of each formulation
was made.
Semisweet, milk, and white chocolate ganache. It is well
established in the pastry world that ganache made with milk and
white chocolate are softer than those using dark chocolate if the
same ratios of chocolate to cream are used. However, this effect
has not been quantified, and the reasons for these differences are
unknown. Commercial semisweet, milk, and white chocolates
were used to make ganache with the same 2:1 chocolate: cream
ratio used in other experiments in this study.
Ganache procedure
Tempering. Both commercial chocolate and chocolate model
systems were tempered before incorporating additional warmed
ganache ingredients, as recommended for optimal texture
(Greweling, 2007). Chocolate was table tempered, using trial and
error to determine the appropriate length of time of tempering.
Chocolate was melted in a 50 °C oven for 1 hour to melt out pre-
existing crystals, then cooled to 45 °C. Approximately 2/3 of the
chocolate was poured onto a marble slab, where it was agitated
with a bench scraper for a set time (which varied by formula-
tion). This was mixed with the remaining 1/3 chocolate, then
transferred to a jacketed beaker (typically 32 °C; 31 °C for milk
chocolate; 30 °C for white chocolate and chocolate systems with-
out CP) fitted with a metal paddle. A Master Servodyne electronic
mixer (Cole-Palmer Instrument Company, LLC, Vernon Hills, IL,
U.S.A.) stirred the chocolate at 20 rpm for 5 min to ensure ho-
mogeneity, and the elimination of undesirable low-melting crystal
forms.
Temper measurement. To ensure that chocolate and model
systems were well-tempered, 2 methods were used. First, a strip of
wax paper was dipped into the chocolate and allowed to set at room
temperature. Well-tempered samples set within a few minutes and
had a glossy appearance. Additionally, an automatic Temper Meter,
model 505A (Tricor Systems Inc., Elgin, IL, U.S.A.) was used
to measure the temperature-time cooling curve. Well-tempered
samples have cooling curves with a steep decrease in temperature, a
plateau region, and another temperature drop. Generally, perfectly
tempered chocolate is regarded as one with a slope of 0 at the
inflection point, where the heat produced by crystallization is
accuracy, slopes of ±0.55 were used.
Ganache production. All formulations were made in a sim-
ilar manner. Generally, cream was placed in a 150 mL beaker
with a stir bar and covered with Parafilm. A thermometer
was inserted through the film to obtain temperature readings
while minimizing moisture loss. The beaker was placed on a
Corning Laboratory Stirrer/Hot plate (Model PC-420; Corn-
ing Inc., Corning, N.Y., U.S.A.) on a low temperature setting
(2 to 3) and high speed setting (10). Once the cream mixture
reached 40 °C, it was poured over the tempered chocolate in
a jacketed beaker (typically 32 °C; 31 °C for milk chocolate;
30 °C for white chocolate and chocolate systems without CP).
This was mixed at 60 rpm for 8 min, occasionally stopping to
scrape down the walls and paddle. Prepared ganache was stored in
the beaker, covered, while measurements were taken.
For the butter/cream formulations, butter (32 °C) was first
added to the chocolate and mixed for 8 min. Cream and/or other
dairy ingredients were mixed together, warmed to 40 °C, and
mixed with the chocolate-butter mixture for an additional 4 min.
The 100% cream formulation was mixed for 12 min to keep total
Food Engineering, Materials
Science, & Nanotechnology
mixing time constant.
All ganache formulations were prepared in triplicate.
Analyses
Ganache samples were analyzed for texture, rheological prop-
erties, melting properties, and microstructure.
Texture analysis. A TA.XT2 texture analyzer (Texture Tech-
nologies Corp, Hamilton, MA, U.S.A.) equipped with Exponent
software (Stable Micro Systems Ltd, Surrey, U.K.) was used for
spreadability and hardness tests. Force was calibrated with a 2-kg
weight before running experiments. Quadruplicate samples were
run at room temperature (approximately 22 °C).
Spreadability was measured using a set of matched male and fe-
male 90° acrylic cones (TTC Spreadability Fixture, Texture Tech-
nologies Corp, Scarsdale, NY, U.S.A.). Freshly made ganache was
poured or spread into female cups and the top was leveled. Samples
were placed in a lidded container to set overnight at room tem-
perature. Before running the first experiment, the machine was
calibrated with an empty female cone. The program for “con-
fectionery: chocolate spreads” was used, which involved the male
cone moving through the sample at a 3.0 mm/s until it reached
2 mm above the bottom of the female cone. Maximum force
(spreadability hardness), and the work of shear (the work to spread
samples along the female cone surface), were recorded.
For hardness tests, freshly-made ganache was spread or poured
into rectangular plastic molds (3 cm × 10.8 cm × 1 cm) that
were lined with plastic film (SaranTM Cling Plus, S.C. Johnson
and Son, Inc, Racine, WI, U.S.A.) for easy removal. Samples were
left uncovered for 10 min to partially set, and then were covered
with plastic film and allowed to set overnight at room temperature.
The following day, samples were unmolded and cut into fourths.
A stainless steel cylindrical puncture probe 5 mm in diameter (TA-
55, Texture Technologies Ltd.) was used, with the height calibrated
before running the first test. The program “confectionery: chewy
confectionery” was used, which moved the rod 2 mm into the
sample from a height of 25 cm. The maximum force was recorded
as hardness of the ganache.
Differential scanning calorimetry (DSC). DSC was used
to obtain melting profiles for chocolate and ganache samples.
Freshly made ganache (10 to 25 mg) and small pieces (10 to
Vol. 00, Nr. 00, 2018 r Journal of Food Science 3
 Investigation of chocolate ganache . . .
20 mg) of solidified tempered chocolate were placed in stan-
dard aluminum hermetic pans (TA Instruments, New Castle, DE,
U.S.A., and DSC consumables Inc., Austin, MN, U.S.A.). Sam-
ples were run on a PerkinElmer DSC 8500 Differential Scanning
Calorimeter (DSC) with Hyper DSCTM (PerkinElmer, Waltham,
MA, U.S.A.). Before running tests, the DSC was calibrated using
an indium sample, and a daily calibration factor was calculated
based on the expected and experimentally obtained melting onset
values.
For ganache and chocolate samples the following program was
used: conditioning at 5 °C for 3 min, then heat at 10 °C/min to
50 °C. The onset melting temperature, peak melting temperature,
end melting temperature and enthalpy of melting (H; area under
the curve) were calculated using the Peak Area function of the
Pyris software. Samples were run in quadruplicate for ganache,
and triplicate for chocolate.
Rheology. A Discovery Hybrid Rheometer 2 (TA Instrument,
New Castle, DE, U.S.A.) equipped with TRIOS software (TA
Instrument) was used to evaluate rheological properties of ganache.
Parallel-plate geometry was used, with a 2.5 cm diameter cross-
hatched plate as the upper fixture, and a temperature controlled
Peltier plate with a sand-blasted, cross-hatched plate attached (to
Food Engineering, Materials
Science, & Nanotechnology
minimize slippage) as the lower fixture. The gap distance was
zeroed before each sample. Warm (30 to 32 °C) freshly made
ganache samples were pipetted onto the bottom plate. The top
plate was lowered to 1 mm above the bottom plate to compress
the sample. Excess sample was removed from the edges, leaving an
even disc of ganache between the plates. The bottom plate was set
to 21 °C, approximately room temperature, and the temperature at
which ganache is typically formed and consumed. A strain sweep
was conducted using the following program: conditioning for 5
min at 21 °C, then a logarithmic increase in strain from 0.001%
to 1000% with oscillation amplitude = 1.0 Hz.
TRIOS software was used to calculate the crossover modulus
(where storage modulus, G’, and the loss modulus, G’’, crossed).
The G’ and G’’ of the linear viscoelastic regions (LVE) were calcu-
lated as an average, encompassing the third recorded value (to avoid
initial inconsistencies) to the largest value, after which moduli be-
gan to decline. Yield stress, σ y, was calculated using σ y = γ y G ,
with the yield strain γ y and storage modulus G’ chosen at the point
where G’ has decreased 5% from linearity (Vaart et al., 2013). Runs
were completed in duplicate for each replicate.
Microscopy. Freshly-made, warm (30 to 32 °C) ganache was
gently mixed with deionized water in a small beaker placed in
a warm (30 to 32 °C) water bath. The ratio of ganache to wa-
ter varied from 1:5 to 1:20 by weight, as needed to be diluted
sufficiently to clearly see structures. A drop of diluted ganache
was placed on a microscope slide with a cover slip and exam-
ined with a Nikon Labophot-2 light microscope (Nikon, Minato,
Tokyo, Japan) at 400× magnification. Images were taken with a
Qimagin QI camera (Surrey, BC, Canada) and analyzed with Im-
age Pro Plus version 7.0 software (Media Cybernetics, Rockville,
MD, U.S.A.). To help elucidate structures observed in ganache,
individual components were also examined with the microscope.
Cream was diluted with water and observed as described above.
Images were taken of all ganache formulations. However, as
formulations appeared similar to each other, only images from
one set of experiments are presented.
Particle size distribution. Particle size distribution (PSD)
analysis was completed for the chocolate model systems to validate
that the wet grinding method produced sufficiently fine particles,
and to determine if chocolate model systems had comparable PSD.
4 Journal of Food Science r Vol. 00, Nr. 00, 2018
Table 5–Texture analysis of mass balanced ganache formula-
tions made with different ratios of milk fat sourced from cream
(C, intact milk fat globules) and butter (B, fragmented milk fat
globules) (n = 12).1
Spreadability-work Spreadability- Penetration-
Formula of shear (kg/s) hardness (kg) hardness (kg)
1:0 C:B 36.3 ± 2.0a 24.5 ± 1.3a 0.247 ± 0.002a
3:2 C:B 38.7 ± 2.7a 22.2 ± 2.4a 0.65 ± 0.18b
3:7 C:B 54.9 ± 6.6b 29.6 ± 3.3b 0.97 ± 0.16bc
0:1 C:B 51.4 ± 7.2b 28.1 ± 2.9b 0.80 ± 0.21c
1
Values with the same letter are not significantly different at α < 0.05.
A MasterSizer 3000 laser particle size analyzer (Malvern Instru-
ments Ltd, Malvern, U.K.) was used. A small amount of each
chocolate model system was dispersed in a mixture of lecithin and
Isopar-G at ambient temperature until an obscuration of 0.2 was
obtained. To maintain the dispersion during measurement, the
unit remained stirring at 2000 rpm. Malvern MasterSizer Micro
Software was used to quantify the size distribution as relative vol-
ume of particles in each band size, presented as size distribution
curves. PSD parameters obtained included the particle size at the
10th percentile (D10 ), mean particle size (D50 ) and the particle size
at the 94th percentile (D94 ).
Statistics
All statistical analysis was performed with JMP Pro 11 software
(SAS, Cary, N.C., U.S.A.). Texture analyzer and PSD data were
evaluated using one-way ANOVA with Tukey HSD tests. DSC
data were analyzed using two-way ANOVA, to compare chocolate
vs. ganache, and formulation on the melting profile. Post hoc tests
included Student’s t-test to compare chocolate and ganache, and
Tukey HSD tests to compare formulations. For the butter-cream
experiment, which used only 1 type of chocolate, a one-way
ANOVA with Tukey HSD was used. Rheological data in this
study expressed unequal variances, making a standard ANOVA test
unsuitable. Therefore, Welch’s ANOVA, which is most commonly
recommended to correct for variance heterogeneity (Jan & Shieh,
2014), was used to compare values from rheological curves.
Results and Discussion
Cream/butter experiments
The results of the spreadability and penetration tests for cream
and butter ganache formulations are listed in Table 5. One-way
ANOVAs found significant differences between the means for
all texture analysis measures (spreadability-work of shear [F (3,
44) = 37.9306, P < 0.0001]; spreadability- hardness [F (3, 44) =
20.3991, P < 0.0001]; penetration- hardness [F (3, 44) = 44.7603,
P < 0.0001]. Post hoc tests using Tukey–Kramer HSD revealed
significant differences between formulations with high levels of
cream compared to those made with butter. Generally, samples
with higher levels of cream exhibited lower hardness scores and
were more spreadable (lower work of shear). This is likely due
to differences in the temperatures of the cream and butter when
added to the chocolate. Butter was added at 32 °C, as for typical
ganache manufacture. However, at this temperature, some milk
fat in butter is unmelted, whereas with cream warmed to 40 °C,
all the fat is melted. The differences in SFC of cream and butter
when added to the chocolate would explain why formulations
with more butter were harder than those with more cream.
DSC results for chocolate and butter/cream ganache formu-
lations are in Table 6. One-way ANOVA analysis with Tukey
Investigation of chocolate ganache . . .

Table 6–Melting profiles for chocolate and for mass balanced ganache formulations made with different ratios of milk fat sourced
from cream (C, intact milk fat globules) and butter (B, fragmented milk fat globules) (n = 12).1

Formula Onset (°C) Peak (°C) End (°C) H (J/g) Index

Chocolate 29.6 ± 1.5a 35.3 ± 1.0a 38.6 ± 0.8a 40.0 ± 1.7a 9.0 ± 2.0c
1:0 C:B 24.9 ± 0.7b 31.1 ± 0.7b 36.1 ± 1.0b 21.7 ± 1.0b 11.3 ± 0.7a
3:2 C:B 25.0 ± 0.8b 31.7 ± 0.8b 36.2 ± 1.2b 22.3 ± 1.1b 11.2 ± 0.5a
3:7 C:B 24.7 ± 0.5b 31.1 ± 0.9b 34.1 ± 2.0c 22.2 ± 0.8b 9.4 ± 2.0bc
0:1 C:B 25.1 ± 0.6b 31.5 ± 1.1b 35.9 ± 1.5b 22.0 ± 0.5b 10.9 ± 1.2ab
1
Values with the same letter are not significantly different at α < 0.05.

HSD post hoc tests revealed that there were few statistically sig-
nificant differences among ganache formulations, but that values
for all measures were significantly higher for chocolate than for all
ganache formulations (Onset [F (4,57) = 73.5698, P < 0.0001];
peak [F (4,57) = 54.2645, P < 0.0001]; End [F (4,57) = 19.6432,
P < 0.0001]; deltaH [F (4,57) = 688.163, P < 0.0001]).
The differences in onset, peak, and melting temperatures be-
tween chocolate and all ganache formulations give insight into
the behavior of ganache as a system. If there was no interaction
between chocolate and milk fat, cocoa butter melting tempera-
tures would be the same in ganache and chocolate. However, the
The crossover modulus is where G’ and G’’ intersect- beyond this
strain, the product will exhibit more viscous than elastic behavior.
No significant differences were found for any measure between
ganache formulations (Storage LVR [F (3, 9.9267) = 2.660, P =
0.1052]; Loss LVR [F (3, 9.635) = 2.4774, P = 0.1234], Yield
stress [F (3, 10.523) = 0.8475, P = 0.4975]; Crossover modulus
[F (3,10.012) = 1.6891, P = 0.2318]).
For all samples, the storage modulus (G’) was greater than the
loss modulus (G’’) in the LVR, indicating that the ganache exhib-
ited more elastic than viscous behavior at room temperature. Once
the strain amplitude increased to the point where the ganache be-

Food Engineering, Materials

melting profile of ganache is shifted lower in temperature. This
shift indicates an interaction between the milk fat and chocolate,
causing the softening effect that is known to occur when milk fat
is mixed with cocoa butter (Timms, 1980; Hartel, 1996).
Average curves for strain sweep tests are shown in Figure 1. Four
characteristics from the strain sweep curves were quantified and
compared (Table 7): the linear viscoelastic region (LVR) of the
storage modulus (G’) and loss modulus (G’’), yield stress, and the
crossover modulus. The LVR is where G’ and G’’ are unaffected
by changes in strain (the plateau region of the graph). The yield
stress is the stress required to initiate flow in a structured fluid.
Science, & Nanotechnology
gan to yield the viscoelastic moduli decreased, indicating shear
thinning. Although chocolate has been shown to exhibit increases
in moduli at large strain amplitudes after shear thinning behavior
is observed at moderate strain amplitudes (Vaart et al., 2013), this
behavior was not observed in ganache.
Ganache formulations were diluted with water and viewed with
a light microscope. Representative images of each ganache for-
mulation are shown in Figure 2. Formulations appeared simi-
lar, regardless of milk fat origin. The ganache appeared to be
comprised primarily of fat globules of irregular sizes and shapes.
Sugar appeared to be dissolved as no clear sugar crystals were

1000000

100000

10000
Modulus (Pa)

1000

100

10

1
0.0001 0.001 0.01 0.1 1 10 100 1000 10000
Oscillaon (%)

Figure 1–Average strain sweeps for mass balanced ganache formulations made with different ratios of milk fat sourced from cream (C, intact milk fat
globules) and butter (B, fragmented milk fat globules). Filled symbols—G’ (storage modulus); Open symbols—G’’ (loss modulus). ♦ , 0:1 C:B;  , 3:7 C:B;
 , 3:2 C:B; •, 1:0 C:B.

Vol. 00, Nr. 00, 2018 r Journal of Food Science 5

 Investigation of chocolate ganache . . .

A B

C D
Food Engineering, Materials
Science, & Nanotechnology

Figure 2–Representative microscopy imagery of diluted mass balanced ganache formulations made with different ratios of milk fat sourced from cream
(C, intact milk fat globules) and butter (B, fragmented milk fat globules), and diluted cream only. (A) 0:1 C:B; (B) 3:7 C:B; (C) 3:2 C:B; (D) 1:0 C:B; (E)
cream only. Scale bars are 10 μm. Arrows point to cocoa powder in ganache (A, D), and to milk fat globules in cream (E).

differences among each of the textural measures (Spreadability-

identifiable. Cocoa powder particles appeared reddish brown, with
work of shear [F (3, 44) = 62.6936, P < 0.0001]; Spreadability-
some remaining within the fat globules and some partitioning into
the aqueous phase. This partitioning between fat and aqueous hardness- [F (3, 44) = 33.5473, P < 0.0001]; Penetration-hardness
phases must be related to their hydrophobic/hydrophilic nature. [F (3, 44) = 84.8594, P < 0.0001]). Post hoc Tukey–Kramer HSD
No clear trends were observed for cocoa particle partitioning. tests (α < 0.05) showed differences between all formulations, for
Cream diluted with water is also shown in Figure 2(E). Intact all measurements. As the level of MFF increased (SFC decreased),
ganache became more spreadable, with lower levels of hardness in
milk fat globules can be clearly seen as the only visible structures.
both spreadability and penetration tests.
Although similar looking structures can be seen in some images of
ganache, these are no more apparent in ganache made with more Although no previous work has examined SFC in ganache, sev-
cream, and can be found in ganache made with only butter, or eral studies have investigated how changing the SFC of chocolate
even in water “ganache” made with no dairy at all (not shown). affects measures of hardness. Regardless of whether SFC levels
were changed by adjusting the types of cocoa butter used or by
Therefore, microscopy evidence also indicates that cream interacts
adding anhydrous milk fat, lower SFC chocolates were found to
with cocoa butter, and is likely present within the main irregular
globular structures. be softer (Kattenberg, 1989; Full et al., 1996; Hartel, 1996). The
results of texture analysis on milk fat modified ganache align with
Milk fat modified chocolate ganache those previously made on chocolate.
Results from spreadability and penetration tests are shown in Melting profile values from DSC analysis for milk fat modified
Table 8. One-way ANOVAs revealed that there were significant chocolate and ganache can be found in Table 9. Two-way ANOVA

6 Journal of Food Science r Vol. 00, Nr. 00, 2018

Investigation of chocolate ganache . . .

Table 7–Rheological measures from strain sweep tests for mass Table 10–Rheological values from strain sweep tests for ganache
balanced ganache formulations made with different ratios of made with chocolate systems with different ratios of a low melt-
milk fat sourced from cream (C, intact milk fat globules) and ing milk fat fraction (MFF) and cocoa butter (CB) added to
butter (B, fragmented milk fat globules). G’ and G’ measured in chocolate mass (n = 6).1
linear viscoelastic region (n = 6).1
Yield Crossover
Yield Crossover Formulation G’ (kPa) G” (kPa) stress (kPa) modulus (%)
Formulation G’ (kPa) G” (kPa) stress (kPa) modulus (%)
1:0 MFF:CB 18 ± 9.1a 4.1 ± 1.6a 0.025 ± 0.007a 82 ± 23a
0:1 C:B 75.1 ± 5.5a 19.0 ± 14.4a 0.45 ± 0.63a 25.9 ± 7.7a 3:2 MFF:CB 15 ± 4.5a 3.4 ± 0.6a 0.029 ± 0.009a 67 ± 26ab
3:7 C:B 506 ± 377a147.8 ± 118.2a 1.50 ± 1.50a 22.7 ± 3.3a 2:3 MFF:CB 16 ± 32a 3.9 ± 0.7a 0.038 ± 0.012a 55 ± 11ab
3:2 C:B 80 ± 76a 26.1 ± 25.4a 0.48 ± 0.69a 21.6 ± 2.3a 0:1 MFF:CB 210 ± 158a 43 ± 30b 0.61 ± 0.49b 38 ± 18b
1:0 C:B 251 ± 350a 62.6 ± 87.3a 0.96 ± 1.63a 46 ± 30a
1
Values with the same letter are not significantly different at α < 0.05.
1
Values with the same letter are not significantly different at α < 0.05.

Table 8–Texture analysis of ganache made with chocolate sys- it takes less energy to melt than formulations that have more MFF.
tems with different ratios of a low melting milk fat fraction Adding cream adds more anhydrous milk fat, which has a lower
(MFF) and cocoa butter (CB) added to chocolate mass. Added melting enthalpy than cocoa butter (Md.Ali & Dimick, 1994),
fat is 26% (w/w) of the total fat in the chocolate systems (n =
12).1 further contributing to the lower melting enthalpies in ganache as
compared to chocolate.
Spreadability-work Spreadability- Penetration- As with the cream and butter ganache formulations, all SFC-
Formula of shear (kg/s) hardness (kg) hardness (kg)
modified ganache exhibit G’ > G’’ at low strain values (Table 10).
1:0 MFF:CB 13.8 ± 5.2a 10.6 ± 3.7a 0.079 ± 0.02a Welch’s one-way ANOVA showed significant differences between
3:2 MFF:CB 21.3 ± 2.0b 14.6 ± 4.7b 0.166 ± 0.038b means for G’ ([F (3, 10.208) = 3.7297, P = 0.0484]), yield stress
2:3 MFF:CB 25.6 ± 3.7c 19.3 ± 3.0c 0.201 ± 0.032c
([F (3, 10.382) = 4.1053, P = 0.0372]) and crossover modulus ([F

Food Engineering, Materials

0:1 MFF:CB 33.1 ± 2.0d 24.0 ± 1.7d 0.260
1
Values with the same letter are not significantly different at α < 0.05.
tests showed significant differences among the means for all
measures (onset [F (7,74) = 33.2638, P < 0.0001]; peak [F (7,74)
= 102.7824, P < 0.0001]; end [F (7,74) = 46.0635, P < 0.0001];
delta H [F (7,74) = 504.7646, P < 0.0001]). Student’s T-tests
(α < 0.05) showed significant differences between the means of
chocolate and ganache for all measures. As with the cream-butter
experiment, chocolate exhibited higher melting temperatures and
melting enthalpies than ganache. Formulation also had an effect,
which was more prominent between milk fat modified chocolates
than the associated ganache for the onset, peak, and end tem-
peratures. As the level of MFF increased, onset, peak and end
temperatures decreased for chocolate. However, this trend was
not observed for ganache. Melting enthalpies also decreased with
higher addition of MFF (decreased SFC), as expected, across both
chocolate and ganache.
There was a strong correlation between the enthalpy of melting
and the quantity of MFF added, in both chocolate and ganache. As
SFC decreases with increasing addition of MFF, a higher propor-
tion of the product is already liquid at room temperature; therefore,
Science, & Nanotechnology
± 0.016d
(3, 10.501) = 4.3511, P = 0.0314]), but not for G” [F (3, 10.097)
= 3.0233, P = 0.0799]. Post hoc tests indicated that ganache with-
out MFF had significantly higher loss modulus in the LVR, and
a higher yield stress than any formulation that contained MFF,
which did not differ significantly from each other.
A small (2.7% of the total mass) amount of added MFF caused
changes in the rheological properties of the ganache, but the ad-
dition of more MFF, within the range used in this study (2.7% to
9.0%) did not appear to have increasing effects. Research on several
types of fats has found that for most, the relationship between SFC
and the elastic modulus is not linear (Narine & Marangoni, 1999).
Although Narine and Marangoni studied only simple fats, and
the systems in this research are more complex, this may partially
explain why the moduli did not correlate with MFF content.
Ganache, as an oil-in-water emulsion, consists of an aqueous
phase with globules of fat dispersed throughout. Both aqueous
and dispersed phases potentially influence textural and rheologi-
cal properties of ganache. In this case, changes to the SFC in the
dispersed phase clearly affect both textural and flow properties of
the system. The differences in rheological measures as MFF was
added to ganache can be explained as follows. As the SFC in-
creases within the dispersed phase, there is more internal structure
in the fat globules, and more force is needed. Likewise, because

Table 9–Melting profiles for chocolate made by adding different ratios of a low melting milk fat fraction (MFF) or cocoa butter
(CB) to chocolate mass, and for the ganache made with those chocolate systems. Added fat is 26% (w/w) of the total fat in the
chocolate systems (ganache, n = 12; chocolate n = 9).1

Formula Onset (°C) Peak (°C) End (°C) H (J/g) Index

Ganache
1:0 MFF:CB 25.1 ± 0.9ab 30.6 ± 0.7a 35.8 ± 0.9a 10.5 ± 1.4a 9.6 ± 0.6a
3:2 MFF:CB 25.6 ± 0.8a 30.9 ± 0.6a 35.9 ± 0.8a 15.7 ± 2.2b 10.3 ± 0.9ab
2:3 MFF:CB 25.3 ± 0.3a 31.0 ± 0.6a 35.9 ± 1.0a 19.3 ± 0.8c 10.6 ± 0.7abc
0:1 MFF:CB 25.5 ± 0.6a 31.4 ± 0.8a 36.4 ± 1.0ab 21.4 ± 1.3d 10.9 ± 0.6bc
Chocolate
1:0 MFF:CB 26.2 ± 0.6ab 33.7 ± 0.8b 37.6 ± 1.0bc 24.5 ± 1.3e 11.5 ± 1.1cd
3:2 MFF:CB 26.9 ± 0.4bc 34.5 ± 0.7bc 38.6 ± 0.7cd 31.1 ± 1.1f 11.7 ± 0.9cd
2:3 MFF:CB 27.8 ± 0.8cd 35.3 ± 0.6c 39.2 ± 0.4d 36.1 ± 2.1g 11.4 ± 0.9cd
0:1 MFF:CB 28.7 ± 0.5d 36.8 ± 1.1d 41.0 ± 0.9e 41.2 ± 1.1h 12.3 ± 0.8d
1
Values with the same letter are not significantly different at α < 0.05.

Vol. 00, Nr. 00, 2018 r Journal of Food Science 7

 Investigation of chocolate ganache . . .

Table 11–Texture analysis of ganache made with chocolate with from penetrating fully. Conversely, the no CP ganache was too soft
different ratios of sugar (S) and defatted cocoa powder (CP) or for the penetrating rod to detect it.
milk powder (MP) and cocoa powder (CP) (n = 12).1
Texture analyzer results for formulations varying CP and NFMP
Spreadability- Penetration- are also shown in Table 11. One-way ANOVAs indicated that
Spreadability-work hardness hardness there were significant differences between the means for all mea-
Formula of shear (kg/s) (kg) (kg)
sures (Spreadability-work of shear [F (3, 44) = 760.8313, P
1:0 S:CP 0.9 ± 0.2a 1.3 ± 0.7a –2 < 0.0001]; spreadability-hardness [F (3, 44) = 723.7847, P <
2:1 S:CP 27.2 ± 2.3b 20.6 ± 1.6b 0.17 ± 0.02a 0.0001]; penetration-hardness [F (2, 31) = 134.6068, P < 0.0001].
1:2 S:CP 95.6 ± 20.4c 47.2 ± 7.4c 1.0 ± 0.2b As with CP-sugar formulations, Tukey–Kramer HSD post hoc tests
0:1 S:CP –3 –3 2.4 ± 0.3c
showed significant (α < 0.05) differences among all NFMP-CP
1:0 MP:CP 2.4 ± 0.6a 2.9 ± 0.4a –2
2:1 MP:CP 4.9 ± 0.4b 5.6 ± 0.5b 0.029 ± 0.003a formulations, with those with more CP registering as harder and
1:2 MP:CP 12.9 ± 1.5c 11.5 ± 1.1c 0.07 ± 0.002b less spreadable. As for the other formulation without CP, the
0:1 MP:CP 27.2 ± 2.3d 20.6 ± 1.6d 0.17 ± 0.002c NFMP-sugar-cocoa butter formulation also was too soft for the
1
penetration test to register.
Values with the same letter within a comparison category are not significantly different
at α < 0.05. In both sets of experiments, increasing ratios of CP resulted in
2
Test failed because material was too soft for force to register on machine. firmer, less spreadable ganache. One likely cause for this differ-
3
Test failed because material was too hard for male cone to penetrate to within 2 mm of
the bottom of the female cone. ence is due to disparities in the densities of different particulate
components. Although the densities of the exact materials used in
this study were not measured, the true densities of cocoa powder,
hardness is determined by the force required to go through both nonfat milk powder, and sucrose have been reported as 1.44, 1.48,
the continuous and dispersed phases, an increase in the SFC of and 1.58 g/cm3 , respectively (Berlin & Pallansch, 1963; Mullarney
the dispersed phase results in an increase in hardness of the system. et al., 2003; Vu, Galet, Fages, & Oulahna, 2003). Because model
Higher levels of cocoa butter in these ganache systems, and there-
Food Engineering, Materials
Science, & Nanotechnology

systems were made on a weight basis, the volume of CP added to

fore more crystalline fat in the dispersed phase, resulted in firmer
products. Similar results for the effects of SFC in the fat phase
on textural properties of caramel were found by Mendenhall and
Hartel (2016).
Model chocolate systems
Based on particle size measurements, the wet grinding process
was successful in reducing the maximum size of sugar granules to
an acceptable level, typically regarded as between 15 and 30 μm
(Ziegler & Hogg, 2009). In general, D10 values were about 1.5 μm,
D50 values were between 5 and 7 μm, and D90 values were gen-
erally 20 μm or lower.
A summary of textural measurements for CP-sugar formulations
is shown in Table 11. One-way ANOVA indicated significant
differences among the means for all measures (Spreadability- work
of shear [F (2, 33) = 203.9657, P < 0.0001]; spreadability-hardness
[F (2, 33) = 329.1524, P < 0.0001]; penetration-hardness [F (2,
33) = 342.9639, P < 0.0001]). Tukey–Kramer HSD post hoc tests
showed significant (α < 0.05) differences among all formulations,
with those with more CP harder and less spreadable. The extremes,
no CP and no sugar, were vastly different from each other, to the
degree that one of the texture tests failed for each of them. The
high CP ganache was firm and crumbly, as opposed to the paste-
like consistency of typical ganache; its hardness prevented the cone
the system was significantly greater than for sucrose, and slightly
greater than for NFMP for the same mass. The PSD is similar
between systems with sugar, NFMP, and CP, indicating that the
greater phase volume of CP comes from more particles rather than
larger particles. The presence of more particles in the same amount
of continuous phase increases particle-particle interactions, which
increases the hardness of the system. Additionally, NFMP and su-
crose are both hydrophilic, and likely dissolve to some degree in
the aqueous phase of the cream. This dissolution decreases the
actual quantity of particles present in finished ganache, thereby
increasing the difference in particle phase volume relative to CP,
and resulting in a softer product.
A summary of melting profile characteristics for both choco-
late and ganache with changing ratios of sugar to CP is shown
in Table 12. Two- way ANOVA found significant differences
among the means for all melting profile measurements (onset
[F (7,74) = 29.6492, P < 0.0001]; peak F (7,74) = 82.6186,
P < 0.0001]; end [F (7,74) = 43.9531, P < 0.0001]; H [F
(7,74) = 585.2205, P < 0.0001]). Post hoc Student’s T-test found
significant differences between chocolate and ganache for the
onset, peak, and end temperatures, as well as the melting en-
thalpy. There was a further trend for ganache made with more
CP to have higher melting peak and end temperatures, as well as
increased H.

Table 12–Melting profiles for chocolate model systems made with different ratios of defatted cocoa powder (CP) to sugar (S), and
for ganache made with that chocolate (ganache, n = 12; chocolate n = 9).1

Formula Onset (°C) Peak (°C) End (°C) H (J/g) Index

Ganache
0:1 CP:S 24.7 ± 0.5a 31.1 ± 0.6a 35.2 ± 0.8a 21.7 ± 1.8a 10.6 ± 0.7a
1:2 CP:S 24.3 ± 0.7a 30.9 ± 0.4a 35.4 ± 0.5ab 24.3 ± 1.2b 11.1 ± 0.6ab
2:1 CP:S 24.7 ± 0.6a 31.7 ± 1.0ab 36.4 ± 1.1bc 24.0 ± 1.5b 11.7 ± 0.8ab
1:0 CP:S 25.1 ± 0.9a 32.6 ± 0.6b 37.3 ± 0.8c 24.4 ± 2.1b 12.2 ± 0.7b
Chocolate
0:1 CP:S 27.6 ± 0.9b 35.1 ± 0.5c 38.6 ± 0.6d 43.6 ± 0.8c 11.0 ± 1.1ab
1:2 CP:S 28.7 ± 2.1b 35.1 ± 1.2c 39.1 ± 1.3d 43.5 ± 1.5c 10.4 ± 2.1a
2:1 CP:S 29.0 ± 2.0b 35.9 ± 1.2c 39.8 ± 1.2d 44.8 ± 0.8c 10.7 ± 1.5ab
1:0 CP:S 28.3 ± 1.1b 35.9 ± 0.3c 39.7 ± 0.6d 45.3 ± 1.2c 11.4 ± 1.5ab
1
Values with the same letter are not significantly different at α < 0.05.

8 Journal of Food Science r Vol. 00, Nr. 00, 2018

Investigation of chocolate ganache . . .

Table 13–Melting profiles for chocolate model systems made with different ratios of defatted cocoa powder (CP) and nonfat milk
powder (MP), and ganache made with that chocolate (ganache, n = 12; chocolate n = 9).1

Formula Onset (°C) Peak (°C) End (°C) H (J/g) Index

Ganache
1:0 MP:CP 24.7 ± 0.7a 31.7 ± 0.8a 34.6 ± 2.8a 24.5 ± 1.5a 9.9 ± 2.6a
2:1 MP:CP 24.9 ± 0.7a 31.4 ± 0.7a 35.5 ± 0.9a 24.1 ± 1.6a 11.1 ± 0.8a
1:2 MP:CP 24.1 ± 0.4a 30.9 ± 0.5a 35.2 ± 0.7a 24.8 ± 0.9a 10.6 ± 0.6a
0:1 MP:CP 24.3 ± 0.7a 30.9 ± 0.4a 35.4 ± 0.5a 24.3 ± 1.2a 11.1 ± 0.5a
Chocolate
1:0 MP:CP 28.0 ± 1.4b 35.5 ± 0.8b 39.6 ± 0.6b 43.9 ± 0.9b 11.5 ± 1.7b
2:1 MP:CP 27.3 ± 0.3b 35.6 ± 0.9b 39.0 ± 0.5b 44.5 ± 1.3b 11.6 ± 0.5b
1:2 MP:CP 27.6 ± 1.1b 35.5 ± 0.8b 39.4 ± 0.6b 44.2 ± 0.5b 11.8 ± 1.1b
0:1 MP:CP 28.6 ± 2.1b 35.2 ± 1.2b 39.1 ± 1.3b 43.5 ± 1.5b 10.4 ± 2.1b
1
Values with the same letter are not significantly different at α < 0.05.

Table 14–Rheological values from strain sweep tests for ganache Table 15–Texture analysis of ganache made from different types
made with chocolate model systems with different ratios of de- of commercially made chocolates (n = 12).1
fatted cocoa powder (CP) to sugar (S) or of nonfat milk powder
(MP) to defatted cocoa powder (CP) (n = 6).1 Spreadability- Spreadability- Penetration-
work of shear hardness hardness
Yield Crossover Formula (kg/s) (kg) (kg)
Formulation G’ (kPa) G” (kPa) stress (kPa) modulus (%)
Dark 29.1 ± 1.1 a 20.5 ± 2.5 a 0.28 ± 0.02a
0:1 CP:S 106 ± 16a 19.4 ± 2.5a 0.07 ± 0.02a 29 ± 13a Milk 6.8 ± 0.1b 6.8 ± 0.7b 0.031 ± 0.003b
345 ± 23a 8.7 ± 4.9a 0.12 ± 0.07a 65 ± 23bc

Food Engineering, Materials

2.7 ± 0.1c 3.6 ± 0.3c

Science, & Nanotechnology

1:2 CP:S White –2
2:1 CP:S 178 ± 53a 38 ± 10a 1.1 ± 0.3a 85 ± 15c
1:0 CP:S 833 ± 283b 174 ± 52b 2.8 ± 1.5b 42 ± 12ab
1
Values with the same letter are not significantly different at α < 0.05.
2
1:0 MP:CP 12.7 ± 5.8a 3.9 ± 1.4a 0.010 ± 0.008a 1.3 ± 0.9a Test failed because material was too soft for force to register on machine.
2:1 MP:CP 2.1 ± 1.7a 1.0 ± 0.6a 0.003 ± 0.002a 0.4 ± 0.3a
1:2 MP:CP 1.7 ± 0.5a 0.9 ± 0.2a 0.011 ± 0.004a 2.3 ± 0.7a
0:1 MP:CP 35 ± 23b 8.7 ± 4.9b 0.12 ± 0.07b 65 ± 23b Differences in the densities of CP, sugar and NFMP, as discussed
previously, likely account for some of the observed differences.
1
Values with the same letter within each category are not significantly different at α <
0.05. With a higher total volume of particulates in mixtures with more
CP, there are more contact points for particle–particle interactions,
which increases resistance to flow. Afoakwa, Paterson, and Fowler
Melting profile characteristics for chocolate model systems with (2008) showed that as chocolate particle size decreases, both the
varying ratios of NFMP and CP, and the ganache made with those number of particles and the particle surface area increases, as does
systems, are shown in Table 13. Two-way ANOVA indicated that plastic viscosity and yield values. In the case of these ganache sys-
there were significant differences between the means (onset [F (7, tems, the particle surface area for high CP systems is also higher
74) = 32.1907, P < 0.0001]; peak [F (7,74) = 91.0869, P < than for those with less or no CP, and accounts for some of the
0.0001]; end [F (7,74) = 28.8254, P < 0.0001]; H [F (7,74) = observed differences.
730.9986, P < 0.0001]). The interaction of chocolate and ganache The nature of the particles themselves may also have an effect.
was significant for all measures, with chocolates expressing higher Do, Vieira, Hargreaves, Mitchell, and Wolf (2011) investigated
values than ganache, but formulation had no significant effect. chocolates made with the same amount of fat, but with different
As with other experiments, G’ > G’’ for CP-sugar and CP- quantities from CP or cocoa butter. Counterintuitively, chocolates
NFMP formulations. Welch’s ANOVA indicated significant dif- with highly defatted CP and more cocoa butter were less fluid than
ferences between means for all rheological measures for CP-sugar chocolates with less defatted CP and less cocoa butter, although
formulations (Table 14; LVR storage [F (3, 10.047) = 29.5453, P more cocoa butter should make for a less viscous product due to
< 0.0001]; LVR loss [F (3, 9.5725) = 28.6859, P < 0.0001]; yield increased lubrication and reduced particle–particle interactions.
stress [F (3, 8.7278) = 23.9127, P = 0.0002]; crossover modulus Confocal laser scanning microscopy revealed pores within the CP,
[F (3, 10.911) = 15.8280, P = 0.0003]) and for CP-NFMP for- which could trap bulk fat, and resulted in higher viscosities. A
mulations (Table 14; LVR storage [F (3, 8.8584) = 9.9761, P = similar fat trapping phenomenon could occur in ganache with
0.0033]; LVR loss [F (3, 9.1452) = 12.1374, P = 0.0015]; Yield defatted CP.
stress [F (3, 9.2663) = 12.0466, P = 0.0015]; crossover modulus Only the systems with the highest level of CP, which exclude
[F (3, 9.5715) = 25.8071, P < 0.0001]). Post hoc tests found the either sugar (in the first set of formulations) or NFMP (in the
all CP/no sugar formulation to have significantly higher LVR G’, second set of formulations), were significantly different from the
LVR G’’, and yield stress values than other formulations. Although systems that did include these ingredients. This is likely due to
there were significant differences in the crossover modulus, there the hydrophilicity of sugar and NFMP, which dissolve to some de-
was no particular pattern with respect to increasing levels of CP gree in the aqueous component of cream. This decreases the num-
or sugar. Additionally, for the CP:NFMP experiments, the no- ber of particles in the system, which reduces particle-particle inter-
NFMP is significantly different for all measured attributes, but actions, and explains some of the observed rheological differences.
there are no significant differences among the formulations with
NFMP. Dark, milk, and white chocolate ganache
For both sets of formulations, only the ones with the highest lev- Results from spreadability and penetration tests for dark,
els of CP were found to be significantly different from the others. milk, and white chocolate ganache are shown in Table 15.

Vol. 00, Nr. 00, 2018 r Journal of Food Science 9

 Investigation of chocolate ganache . . .

Table 16–Melting profiles for commercial dark, milk, and white chocolate, and for ganache made with those chocolates (n = 12).1

Formula Onset (°C) Peak (°C) End (°C) H (J/g) Index

Ganache
Dark 24.9 ± 0.7a 31.1 ± 0.7b 36.1 ± 1.0b 21.7 ± 1.0b 11.3 ± 0.7bc
Milk 24.6 ± 0.5a 30.4 ± 0.5a 34.8 ± 0.7a 17.0 ± 0.6a 10.3 ± 0.7ab
White 25.4 ± 0.8ab 30.9 ± 0.6ab 35.3 ± 1.0ab 16.9 ± 2.8a 9.9 ± 1.3ab
Chocolate
Dark 29.6 ± 1.5d 35.3 ± 1.0c 38.6 ± 0.8cd 40.0 ± 1.7d 9.0 ± 2.0a
Milk 27.8 ± 2.4c 34.8 ± 0.8c 37.6 ± 0.7c 31.1 ± 0.9c 9.9 ± 2.4ab
White 26.7 ± 1.1bc 34.9 ± 0.8c 39.0 ± 0.9d 31.8 ± 0.4c 12.3 ± 0.8c
1
Values with the same letter are not significantly different at α < 0.05.

One-way ANOVAs, or t-test in the case of the penetration Table 17–Rheological values from strain sweep tests for ganache
test, indicated significant differences between the means for all made with dark, milk, and white chocolate (n = 6).1
measures (Spreadability-work of shear [F (2, 31) = 433.2998, Yield stress Crossover
P < 0.0001]; Spreadability-hardness [F (2, 31) = 383.9228, P < Formulation G’ (kPa) G” (kPa) (kPa) modulus (%)
0.0001]; Penetration-Hardness [t(22) = –39.8279, P < 0.0001]. Dark 251 ± 350a 63 ± 87a 1.0 ± 1.6a 46 ± 30a
Post hoc tests indicated significant differences between all 3 types Milk 2.5 ± 1.0b 0.7 ± 0.3b 0.006 ± 0.002b 1.8 ± 0.6b
of ganache, with dark chocolate the hardest and least spreadable, White 9.9 ± 4.9b 2.3 ± 0.9b 0.019 ± 0.006b 7.2 ± 3.5b
and white chocolate the softest and easiest to spread. As with 1
Values with the same letter are not significantly different at α < 0.05.
Food Engineering, Materials
Science, & Nanotechnology

model systems lacking cocoa powder, white chocolate ganache

was too soft to run a penetration test, and the probe moved
through the product without registering any force.
Two-way ANOVAs indicated significant differences among the
means for all melting profile measures for commercial dark, milk,
and white chocolates, and ganache (Table 16; onset [F (5, 68) =
28.2395, P < 0.0001]; peak [F (5, 68) = 114.2217, P < 0.0001];
end [F (5, 68) = 50.6140, P < 0.0001]; H [F (5,68) = 508.7535,
P < 0.0001]). Post hoc Tukey tests revealed significant differences
between the chocolate and ganache for all measures. The effect of
chocolate type was also significant for all measures. Though dark
chocolate had higher melting temperatures for some measures in
ganache and chocolate, there was a high degree of overlap with
other types of chocolate, and no distinct patterns. Dark choco-
late required significantly more energy to melt (higher H) for
both ganache and chocolate than either milk or white chocolate
formulations. This is likely due to the softening effect of milk fat
present in milk and white chocolates, which is known to decrease
melting enthalpy (Full, Reddy, Dimick, & Ziegler, 1996).
As for other ganache formulations, the elastic modulus was
greater than the viscous modulus for all samples. Welch’s ANOVA
indicated significant differences between means for all measures
(Table 17; LVR storage [F (2, 6.9255) = 7.3127, P = 0.0196];
LVR loss [F (2, 7.2647) = 9.2928, P = 0.0099]; Yield stress [F
(2, 7.1181) = 12.6267, P = 0.0046]; Crossover modulus [F (2,
6.8458) = 12.1970, P = 0.0055]. For all measures, dark chocolate
ganache had higher values than either milk or white chocolate
ganache, which were not significantly different from each other.
The results from this set of experiments reiterate the differences
between ganache made with different types of chocolate. Varia-
tions are due to compositional differences between different types
of chocolate. Glicerina, Balestra, Dalla Rosa, and Romani (2016)
examined the microstructural and rheological properties of typical
commercially produced dark, milk, and white chocolates. They
found that white chocolate had the smallest particle size, and least
dense crystalline network with reduced particle-particle interac-
tions due to the greatest amount of fat. It also was the least viscous,
due to low flow resistance with such large distances between par-
ticles. Milk chocolate, although similar in fat to white chocolate,
had a denser microstructure due to the presence of cocoa powder,
and was found to be slightly more viscous. The dark chocolate
had the least amount of fat and also exhibited the least distance
between particles. It was the most viscous of the 3.
Although only the approximate composition of the commercial
chocolates used in this study could be obtained, as in the Glicerina
et al. (2016) study, the milk and white chocolates used contained
greater amounts of fat than dark chocolate. Higher levels of fat
increases the distance between suspended particles, thereby reduc-
ing particle-particle interaction and leading to a softer product.
Additionally, the milk fat present in milk and white chocolate
is at a great enough level (>30% of the fat phase) to cause eu-
tectic effects, which would lead to increased softening (Timms,
1980). The patterns in ganache mirror those found in choco-
late with regards to texture and rheology. Therefore, fat con-
tent and particle size of chocolate, and how those factors affect
the distance between particles, are important to ganache physical
properties.
Conclusion
Ganache is a multi-phase, complex food that contains a fat
globule emulsion with various particulates distributed both within
the emulsion and in the continuous aqueous phase. In the emul-
sion phase, both cocoa powder particles and cocoa butter crystals
provide structure to the nonuniform fat globules. The aqueous
phase contains some cocoa powder particles and perhaps some
crystalline sucrose. This complex mixture of structures provides a
smooth texture to ganache along with a rapid flavor release in the
mouth.
Manipulating these structures allows for some level of control of
ganache textural properties, as documented by the results of this
study. Increasing the SFC of the emulsion (based on the SFC of
the chocolate) resulted in harder and less spreadable ganache, as did
higher levels of cocoa particles. Thus, ganache made with choco-
late with elevated levels of cocoa particles (namely bittersweet
chocolate) gave a firmer ganache while, at the same chocolate to
cream ratio, milk and white chocolate, with reduced cocoa parti-
cles gave softer texture. These results provide quantitative guidance
and mechanistic understanding for controlling ganache texture.

10 Journal of Food Science r Vol. 00, Nr. 00, 2018

Investigation of chocolate ganache . . .
Acknowledgments
Thanks to Nico Tomaselli and Jen Burke for their assistance
with PSD analysis.
Author Contributions
JM designed the study, collected and interpreted data, and
drafted the manuscript. RWH gave feedback on experimental de-
sign and data analysis, and edited the manuscript.
References
Afoakwa, E. O., Paterson, A., & Fowler, M. (2008). Effects of particle size distribution and
composition on rheological properties of dark chocolate. European Food Research and Technology,
226, 1259–1268.
Beckett, S. T. (2008). Chocolate flow properties. In S. T. Beckett (Ed.), Industrial chocolate
manufacture and use (4th ed., pp. 224–246). West Sussex, UK: Wiley-Blackwell.
Bensoussan, M., Alcaraz, G., & Tourtel, V. (1997). Growth characteristics of Aspergillus chevalieri
and other fungi from under-coating of chocolate truffles. In Advances in solid state fermentation
(pp. 39–47). Dordrecht: Springer.
Berlin, E., & Pallansch, M. J. (1963). Influence of drying methods on density and porosity of
milk powder granules. Journal of Dairy Science, 46, 780–784.
Dias, J., Alvarenga, N., & Sousa, I. (2015). Effect of hydrocolloids on low-fat chocolate fillings.
Journal of Food Science and Technology, 52, 7209–7217.
Do, T-A. L., Vieira, J., Hargreaves, J. M., Mitchell, J. R., & Wolf, B. (2011). Structural char-
acteristics of cocoa particles and their effect on the viscosity of reduced fat chocolate. LWT -
Food Science and Technology, 44, 1207–1211.
Full, N. A., Reddy, S. Y., Dimick, P. S., & Ziegler, G. R. (1996). Physical and sensory properties
of milk chocolate formulated with anhydrous milk fat fractions. Journal of Food Science, 61,
1068–1073.
Glicerina, V., Balestra, F., Dalla Rosa, M., & Romani, S. (2016). Microstructural and rheolog-
ical characteristics of dark, milk and white chocolate: A comparative study. Journal of Food
Engineering, 169, 165–171.
Greweling, P. (2007). The crystallization of ganache. Manufacturing Confectioner, 87, 53–56.
Greweling, P. (2013). Chocolates and confections (2nd ed.). Hoboken, N.J.: Wiley.
Hartel, R. W. (1996). Applications of milk-fat fractions in confectionery products. Journal of the
American Oil Chemists’ Society, 73, 945–953.
Jan, S-L., & Shieh, G. (2014). Sample size determinations for Welch’s test in one-way het-
eroscedastic ANOVA. British Journal of Mathematical and Statistical Psychology, 67, 72–93.
Kattenberg, H. (1989). The effect of cocoa butter on chocolate tempering and bloom. Manufac-
turing Confectioner, 69, 77–81.
Md.Ali, A. R., & Dimick, P. S. (1994). Thermal analysis of palm mid-fraction, cocoa butter
and milk fat blends by differential scanning calorimetry. Journal of the American Oil Chemists’
Society, 71, 299–302.
Mendenhall, H., & Hartel, R. W. (2016). Effects of fat content and solid fat content on caramel
texture attributes. Journal of the American Oil Chemists’ Society, 93, 1191–1199.
Minifie, B. (1989). Chocolate, cocoa, and confectionery (3rd ed., Vol. 109, pp. 35–84). New York,
N.Y.: Van Nostrand Reinhold.
Mullarney, M. P., Hancock, B. C., Carlson, G. T., Ladipo, D. D., & Langdon, B. A.
(2003). The powder flow and compact mechanical properties of sucrose and three high-
intensity sweeteners used in chewable tablets. International Journal of Pharmaceutics, 257, 227–
236.
Narine, S. S., & Marangoni, A. G. (1999). Mechanical and structural model of fractal networks
of fat crystals at low deformations. Physical Review E, 60, 6991–7000.
Robuchon, J. (2009). Larousse Gastronomique (pp. 488). New York, N.Y.: Crown Publishing
Group.
Timms, R. (1980). The phase behaviour of mixtures of cocoa butter and milk fat. Lebensm-
Wissenshaft Technol, 13, 61–65.
van der Vaart, K., Depypere, F., Graef, V. D., Schall, P., Fall, A., Bonn, D., & Dewettinck, K.
(2013). Dark chocolate’s compositional effects revealed by oscillatory rheology. European Food
Research and Technology, 236, 931–942.
Vu, T., Galet, L., Fages, J., & Oulahna, D. (2003). Improving the dispersion kinetics of a cocoa
powder by size enlargement. Powder Technology, 130, 400–406.
Ziegler, G. R., & Hogg, R. (2009). Particle size reduction. In S. T. Beckett (Ed.), Industrial
chocolate manufacture and use ( 4th ed., pp. 142–166). West Sussex, UK: Wiley-Blackwell.
Food Engineering, Materials
Science, & Nanotechnology
Vol. 00, Nr. 00, 2018 r Journal of Food Science 11