Food Research International 119 (2019) 84–98
Contents lists available at ScienceDirect
Food Research International
journal homepage: www.elsevier.com/locate/foodres
Pod storage with roasting: A tool to diversifying the flavor profiles of dark T
chocolates produced from ‘bulk’ cocoa beans? (part I: aroma profiling of
chocolates)
Michael Hinneha,b,c,⁎, Enoch Enorkplim Abotsia, Davy Van de Wallea,
Daylan Amelia Tzompa-Sosaa, Ann De Winnec, Julien Simonisd, Kathy Messense,
Jim Van Durmec, Emmanuel Ohene Afoakwab, Luc De Coomanc, Koen Dewettincka
a Department of Food Technology, Safety and Health, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, 9000 Ghent, Belgium
bDepartment of Nutrition & Food Science, University of Ghana, P. O. Box LG 134, Legon, Accra, Ghana
c Research Group Molecular Odor Chemistry, Department of Microbial and Molecular Systems (M2S), Research Cluster Food and Biotechnology, KU Leuven Technology
Campus, 9000 Ghent, Belgium
d Puratos – Belcolade, Industrielaan 16, Industriezone Zuid III, B-9320 Erembodegem, Belgium
e Department of Biotechnology, Faculty of Bioscience Engineering, Ghent University, Valentin Vaerwyckweg, 1, 9000 Ghent, Belgium
A R T I C L E I N F O A B S T R A C T
Keywords: The impact of pod storage (PS) and roasting temperature (RT) on the aroma profiles of dark chocolates were
Aroma profile evaluated. Cocoa liquor samples comprised of ten different combinations of PS and RT, whilst keeping the
Chocolate roasting time fixed at 35min. Additionally, commercial cocoa liquors from renowned origins (Ecuador,
Cocoa Madagascar, Venezuela, Vietnam, Ivory Coast and Ghana) were acquired for comparison. From these, 70% dark
HS-SPME-GC–MS chocolates were produced under the same conditions after which they were subjected to headspace solid-phase
Pod storage
Roasting microextraction-gas chromatography–mass spectrometry (HS-SPME-GC–MS) analysis. Although both PS and RT
were found to influence the aroma volatile concentrations, the impact of RT over PS seemed to be greater. An
agglomerative hierarchical clustering (AHC) of all chocolates on the basis of their aroma profiles revealed a
similar impact as earlier observed, where major clustering of the chocolates was in accordance with the intensity
of the roasting process applied. However, within each group, the dissimilarities owing to PS among the cho-
colates was clearly depicted. Comparatively, chocolates with low (100–120 °C), instead of moderate to high
(135–160 °C) RT's, rather showed a low dissimilarity with those from the commercial cocoa liquors of the dif-
ferent origins. Although from the same beans, the diversity of aroma profiles of these chocolates as well as the
similitude of some treatments to some chocolates from commercial grade cocoa liquors, unequivocally under-
scores the possibility for steering diverse distinct flavors from ‘bulk’ cocoa through PS and roasting, with ben-
eficial implications, both from an application and an economic point of view.
1. Introduction notes. They are mainly of the Criollo, Trinitario and Nacional varieties.
However, the ‘bulk’ flavor cocoa mainly consists of the Forastero
The cocoa bean is the main raw material of chocolate. As such, its variety. It is generally characterized by its typically strong acidic, as-
flavor potential is key in determining the final flavor of the chocolate. tringent, intense cocoa and less fruity/floral notes, and constitutes 95%
The variations in flavor precursors contained within the bean have been of global cocoa production. For their unique aroma and flavor char-
identified to be greatly linked to its origin, soil type, age of cocoa tree acteristics, ‘fine’ flavor cocoa are highly sought after and thus, attract a
and the genotype or variety (Kongor et al., 2016). From the latter, two high premium price in comparison to ‘bulk’ cocoa. Yet, the soaring
main classifications are often discussed in relation to the cocoa flavor; global demand for cocoa and chocolate products exhibiting more di-
‘fine’ and ‘bulk’ or ‘basic’ cocoa (Afoakwa, 2016). The ‘fine’ flavor cocoa verse and distinct flavor profiles, can be partly addressed through
are generally known, among others, by their fruity, floral and spicy various cost-effective and sustainable technological interventions by
⁎ Corresponding author at: Department of Food Technology, Safety and Health, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, 9000
Ghent, Belgium.
E-mail address: hinnehmichael@gmail.com (M. Hinneh).
https://doi.org/10.1016/j.foodres.2019.01.057
Received 22 November 2018; Received in revised form 22 January 2019; Accepted 23 January 2019
Available online 24 January 2019
0963-9969/ © 2019 Elsevier Ltd. All rights reserved.
M. Hinneh et al. Food Research International 119 (2019) 84–98
which not only ‘fine’ flavor cocoa, but also ‘bulk’ cocoa can be valorized beans. These chocolates were then assessed on the basis of the un-
to meet this ever-growing demand. iqueness of their aroma profiles with respect to the applied treatments
Through decades of investigations, the impact of various post- as well as in comparison to equally manufactured chocolates from
harvest and processing techniques on the final flavor of the end-product commercial cocoa liquors of recognized origins.
has been established (Aculey et al., 2010; Afoakwa, 2010; Afoakwa,
2014; Beckett, 2009; Frauendorfer & Schieberle, 2008; Hinneh et al.,
2018; Kadow, Bohlmann, Phillips, & Lieberei, 2013; Meyer, Biehl, Said, 2. Materials and methods
& Samarakoddy, 1989; Sulaiman, Yang, & Ariffin, 2017). Thus, offering
various possibilities through which the flavor of cocoa/chocolate can be 2.1. Preliminary trials for sample selection
tuned. Spanning from bean to bar, a range of such postharvest/pro-
cessing factors can be envisioned; including pod storage (PS), fermen- A preliminary study on the impact of PS and RT on the aroma
tation, drying, alkalization, roasting (temperature and time) and con- profiles of Ghanaian cocoa liquors was performed to select the PS and
ching. Of these, the possible synergistic impact of PS and roasting RT to be applied in the present study. In this preliminary study, a 3×5
temperature is yet to be explored. full factorial experiment comprising of PS; 0, 3 and 7 days and roasting
Pod storage is an on/off-farm practice of storing harvested cocoa conditions; 100, 120, 135, 140 and 160 °C each with a constant roasting
pods under specific conditions for a specified duration of time prior to duration of 35min was used. Cocoa liquors from these samples were
fermentation (Hinneh et al., 2018). This technique has become a analyzed using headspace solid-phase microextraction-gas chromato-
common practice in some cocoa-producing countries, although pre- graphy–mass spectrometry (HS-SPME-GC–MS) to identify and quantify
viously very little was known about its impact on the flavor potential of the volatile compounds related to aroma. Ten out of these fifteen cocoa
the cocoa beans. For instance, as observed by Duncan (1984), Ghanaian liquors were then selected for the present study. The basis for this se-
farmers unknowingly practiced PS as a means of reducing labor cost by lection was the interest to identify cocoa liquors of the treatments
first gathering enough ripe pods over a period of days before organizing showing diverse and distinct aroma profiles from the others. However,
family and friends to assist in splitting them. Meanwhile, in Malaysia, in order to understand the influence of the factors on possibly identified
this was adopted as an intervention for curbing the problem of over- trends, the following criteria was also considered; (1) for each selected
acidity in their fermented cocoa beans (Meyer et al., 1989). It is evident PS – RT combination, a corresponding reference with unstored pods
that a series of biochemical changes occurring during the PS including (0PS) was also present, (2) the impact of PS or RT could be traced
pulp volume and moisture content reductions have consequent impact among the samples given that at least one RT (ie. 135 °C) or PS (ie. 0PS)
on the chemical composition of the beans (Afoakwa, Quao, Budu, was fixed respectively. Table 1 shows a summary of these ten selected
Takrama, & Saalia, 2011b; Saltini, Akkerman, & Frosch, 2013). In PS – RT combinations to be investigated in this present study.
dealing with Ghanaian cocoa beans, Afoakwa, Quao, Budu, Takrama, Finally, by means of situating the different liquors samples in a ty-
and Saalia (2011a) reported that PS duration between three and seven pical industrial context, six extra cocoa liquors from renowned origins
days along with adequate fermentation resulted in appreciable reduc- (Ecuador, Ghana, Ivory coast, Madagascar, Venezuela and Vietnam)
tions in nib acidification, sugars (non-reducing and total sugars) and were obtained from a commercial supplier. Information as obtained
protein with resultant increase in reducing sugars and acceptable free from the supplier are summarized in Table 2. It must be indicated that
fatty acid levels. In agreement, Hinneh et al. (2018) recently reported for these commercial cocoa liquors, information concerning PS, specific
detailed analyses of some flavor precursors (sugars and free amino acid type and conditions of roasting were not provided. However, within the
profiles), where they observed an increasing total concertation of pre- framework of this study, they were viewed as representation of the
cursors with increasing PS. More so, of the Maillard reaction-related typical industrial grade liquor. As such, the basis of comparison of these
aroma volatiles investigated, PS up to seven days was associated with with our liquor samples, given the goal of this study, was irrespective of
more than twice the total concentration of aroma volatiles compared to the specifics of the applied industrial process. Hitherto, the basic idea
the reference. Thus, it is possible to boost or modify the aroma of cocoa was to see, of the diverse aroma profiles of our samples, if some of these
through PS. chocolates could mimic typical aroma profiles of those from a com-
After fermentation and drying, a very crucial stage which ensures mercial grade liquor given the same manufacturing processing condi-
the formation of desirable flavor volatiles from precursors (sugars and tions. From all sixteen cocoa liquors, 70% dark chocolates were then
free amino acids) through the Maillard reaction is the roasting process produced and thereby compared on the basis of their aroma profiles
(Counet, Callemien, Ouwerx, & Collin, 2002; Frauendorfer & using the HS-SPME-GC–MS analytical technique.
Schieberle, 2008; Nazaruddin, Seng, Hassan, & Said, 2006). Here, the
free amino acids interact with the reducing sugars through a series of 2.2. From pod storage to roasted cocoa liquor
chemical reactions to form the Amadori compounds. This is a precursor
for the formation of 3-deoxyhexuloses and 2,3-enediol with dehydror- Ripe cocoa pods were harvested during the September–October
eductone intermediates under acidic and basic/neutral conditions, re- peak season from a cocoa farm in Jachere, Ghana (N7.088525,
spectively. From the latter, α-dicarbonyl compounds are formed, which W2.110127833333333). This was representative of a cluster of farms
undergoes Strecker degradation and heterocyclization to give rise to
various kinds of aldehydes, ketones, pyrazines, pyrroles and pyridines Table 1
among others (Afoakwa, Paterson, Fowler, & Ryan, 2008; Aprotosoaie, Pod storage and roasting conditions of selected samples.
Luca, & Miron, 2016). For this reason, roasting is also regarded as one Pod storage (day) Temperature (°C)
of the most important stages in the processing of cocoa beans, espe-
cially, in the context of the flavor quality of the end product. 0 100
To the best of our knowledge, in spite of the immense contributions 0 120
0 135
of PS and roasting temperature (RT) on the cocoa/chocolate flavor, the 0 140
impact of both processes on the aroma profile of the final chocolate has 0 160
not been investigated. Thus, the aim of this study was to explore the 3 120
potential of PS with RT for the diversification and/or tuning of the 3 135
aroma profiles of dark chocolates produced from ‘bulk’ cocoa beans. 7 1007 135
This was done by applying different pre-selected treatments of PS and 7 160
RT and thereafter, the production of 70% dark chocolates from these
85
M. Hinneh et al. Food Research International 119 (2019) 84–98
Table 2
Information about commercial cocoa liquors as received from supplier.
Origin Variety Fermentation Drying Roasting
Ecuador Nacional Box Combined drying Nib roasting
Ghana Forastero Heap Solar drying Nib roasting
Ivory coast Forastero Heap Solar drying Nib roasting
Madagascar Forastero+ Criollo+ Trinitario Box Combined drying Nib roasting
Venezuela Criollo+ Trinitario Box Combined drying Nib roasting
Vietnam Trinitario Box Solar drying Nib roasting
demonstrating high potential for quality cocoa beans following an 2.4. Aroma profiling of chocolates
earlier survey in this region (Kongor et al., 2017). Cocoa trees were
approximately 31 years old. The pods were stored in small heaps (ca The aroma analysis was carried out using HS-SPME-GC–MS ac-
80–100 pods) on the farm for three different storage times (0, 3, and cording to Tran et al. (2015) and modified by Hinneh et al. (2018).
7 days), at ambient temperature (28–30 °C) and relative humidity of Identification of aroma volatiles in the headspace was performed by
77–85%. At the end of each duration of PS, the pods were opened, comparing the MS-spectrum of each peak to those from the Wiley 275
beans were gathered in heaps (ca 45 kg) and covered with fresh banana library. The identification was verified by determination of Kovat in-
leaves to begin the spontaneous heap fermentation process. This oc- dices (KI's) after injection of a series of n-alkane homologues (C5 –
curred for six days with two turnings at 48 h intervals as recommended C13). From these, the Kovat indices of the confirmed aroma compounds
for local farmers by the Cocoa Research Institute of Ghana (CRIG). After were calculated based on their respective retention times and compared
this, the fermented cocoa beans were sun-dried on raised platforms with Kovat indices from literature.
until the required moisture content of< 7% was achieved, as described The semi-quantitative concentrations of the identified volatile
by Afoakwa (2010). The dried beans were then stored in jute bags of compounds were expressed as nanograms of the internal standard
64 kg gross weight. They were then air-freighted to the Faculty of equivalents per gram of cocoa liquor and calculated as the area of the
Bioscience Engineering, Ghent University, Belgium. Here, the bags were compound of interest divided by the response factor of the internal
stored in a cool well-ventilated room on odorless wooden racks. standard. For each chocolate, isolation, separation, identification and
Roasting and winnowing (2 kg per batch) were carried out in the Selmi quantification of the aroma volatiles were performed in triplicate.
roaster (Selmi-group, Italy) and Winn-15 Winnower (Cacao Cucina, Finally, in order to evaluate a volatile's contribution to the overall
U.S.A), respectively. Cocoa nibs (1.5 kg per batch) were first pre-broken flavor profile of the cocoa liquor, odor activity values (OAV's) were
using the Stephan mixer at 45 °C. First, 8 min at 50% speed, then, 6min calculated using odor threshold values (OTV's) documented in litera-
at 75% speed. Thereafter, the ECGC-12SLTA melanger (CocoaTown, ture. Here, the OTV in oil media were used since chocolate is a fat
Roswell, USA) was used to further grind the nibs into liquor. For this, continuous suspension (43% fat). These values were sourced from Van
about 300min grinding time was required to obtain liquor of particle Gemert (2011). From these, OAV's were calculated by dividing the
size; D (v,0.9) of 21 μm. The particle size was determined using a detected headspace concentration by their respective OTV's. Hitherto, a
Malvern Mastersizer according to Saputro et al. (2017). volatile with OAV≥1 may be considered as an odor-active volatile and
vice versa. It is worth mentioning that depending on its OTV, a volatile
2.3. Chocolate production with a relatively higher headspace concentration may not necessarily
contribute to the overall flavor profile of a particular liquor. Moreover,
Sixteen batches of 70% dark chocolate (total fat= 43%) consisting as can be seen from Table 4a, 4b, not all OTV references were found.
of 30.00% pre-broken sugar (Barry Callebaut Belgium, Wieze, Hence, not all potentially odor-active volatiles could be elucidated.
Belgium), 64.65% cocoa liquor, 5.00% cocoa butter (Puratos - Thus, even though the calculated OAV's may provide some insight, they
Belcolade, Erembodegem, Belgium) and 0.35% soy lecithin (Soya must be interpreted with caution.
International Ltd., Cheshire, U.K.) were produced on a 5 kg scale.
Mixing was carried out using the VEMA BM 30/20 planetary mixer 2.5. Statistical analysis
(Machinery Verhoest NV/Vema Construct, Izegem, Belgium) for a
duration of 20min at 45 °C. The mixed ingredients (27% fat) was re- The concentrations of the various volatile compounds as well as
fined with the Exakt 80S 3-roll refiner (Exakt Technologies, inc., USA) their totals were subjected to Analysis of Variance (ANOVA) at a 5%
at gap setting 2–1, roller speed of 400 rpm and temperature of 35 °C. significance level. Assumptions of normality and equality of variance
The refined chocolate mass was then conched in a Bühler ELK'olino were tested prior to the analysis using Kolmogorov-Smirnov test and
conche (Richard Frisse GmbH, Bad Salzuflen, Germany) in two phases. Modified Levene's test, respectively. Where assumptions were fulfilled,
The dry phase was carried out at 60 °C with 1200 rpm for two hours a post-hoc Tukey's test was used to investigate significant differences
(clockwise) and 80 °C with 1200 rpm for four hours (anti-clockwise). At among samples. However, when assumption was not fulfilled, a non-
the liquid phase, calculated amounts of pre-conched cocoa liquor, cocoa parametric alternative, Welch was used along with Games Howell post-
butter and the soy lecithin were added, such that, the final fat content hoc test. ANOVA was performed with Minitab 18 (Minitab Inc., USA).
of the chocolate was 43%. Here, the process was carried out as follows; Agglomerative Hierarchical Clustering (AHC) of the chocolates on the
45 °C with 2400 rpm for 15min (clockwise) and 15min (anti-clock- basis of their aroma volatiles was carried out using XLSTAT 2014.5.03
wise). Pre-conching of part of the cocoa liquor was necessary since the (Addinsoft, USA).
entire amount of cocoa liquor required to produce final chocolate
consisting of 70% cocoa could not be included in the recipe prior to the 3. Results and discussion
dry conching phase. Hence, for each batch, equal amounts of cocoa
liquor was previously dry-conched using the same dry conching pro- Table 3a and 3b display the full list of aroma volatiles identified in
cedure, of which specific required amounts were later added at the all 16 chocolates. The odor activity values (OAV's) were calculated in
stage of liquid conching in order to make up for this final concentration. order to estimate the contribution of each volatile to the overall aroma
This production method was inspired by the Cocoa of Excellence of the chocolates (Table 4a and 4b). Furthermore, dendrograms (Fig. 2a
Program. and b) were also constructed by agglomerative hierarchical clustering
86
M. Hinneh et al. Food Research International 119 (2019) 84–98
87
Table 3a
Concentrations (ng/g cocoa) of aroma volatiles identified from dark chocolates produced from cocoa beans of different pod storage – roasting conditions.
N- Volatile 0PS-100 °C 7PS-100 °C 0PS-120 °C 3PS-120 °C 0PS-135 °C 3PS-135 °C 7PS-135 °C 0PS-140 °C 0PS-160 °C 7PS-160 °C
o.
Acids
1 Acetic acid 2822.26±195.13C- 2090.90± 95.39FG 2834.36± 86.31CD 2169.45± 81.26FG 3609.52± 217.97- 2107.52±70.26FG 3516.61±204.88- 2902.55±33.42C 3885.59±109.66- 4312.15±110.39A
DE B B B
2 Isovaleric acid 213.77± 3.21CD 171.26± 24.58DE 380.86± 43.62A 231.45±15.04CD 331.45± 40.25AB 223.73± 15.88CD 277.98± 62.96BC 381.40± 32.92A 261.07± 26.23BCD 268.21± 19.97BC
3 Oxalic acid DITMS 152.88± 28.94AB 167.32± 30.71AB 104.78± 9.42BC 102.78±4.04BC 115.95± 14.20BC 157.78± 6.79AB 139.82± 65.83ABC 101.33± 10.93BC 175.87± 19.13AB 218.59± 54.44A
Total 3188.91±221.50C 2429.48± 78.04D 3319.99± 111.11C 2503.68± 70.44D 4056.93± 261.98- 2489.03±53.32D 3934.42±173.43- 3385.27±75.86C 4322.53±122.78- 4798.95±122.95A
B B B
Alcohols
4 Amyl alcohol 40.99± 0.89BC 39.49±7.88BC 46.64±3.11AB 41.43± 2.77BC 34.38± 2.02CD 34.84± 0.73CD 25.26± 0.77EF 18.77± 2.57F 19.24± 1.33EF 23.20± 1.61EF
5 1-Hexanol 0.00±0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 0.00±0.00D 0.00±0.00D 0.00±0.00D 0.00±0.00D
6 2,3-Butanediol 71.89± 7.07DEF 133.42± 29.33BC 46.00±10.13EF 50.81± 13.40DEF 164.02± 21.11B 106.29± 12.63BCD 410.45± 39.94A 104.53± 19.09CDE 434.76± 30.24A 433.56± 30.92A
7 1,3-Butanediol 116.55± 19.76EF 103.57± 6.30F 169.21± 8.20BCDE 189.85±18.88BCD 351.41± 49.75A 184.09± 7.25BCD 368.60± 6.89A 218.97± 45.64B 227.35± 27.38B 200.75± 16.73BC
8 Benzyl alcohol 126.46± 10.48BCDE 128.10± 7.78BCDE 120.06± 12.48CDEF 145.05±13.47ABC 142.20± 3.49BCD 108.49± 34.35CD- 121.10± 14.39CD- 170.75± 33.19AB 122.76± 6.91CDEF 188.15± 14.47A
EF EF
9 2-Phenylethyl 382.59± 21.94EF 98.05±3.62H 510.18± 14.16D 296.66±34.00F 887.19± 22.30B 361.74± 9.94EF 397.44± 55.17E 828.68± 37.29B 1083.60±59.08A 614.51± 36.63C
alcohol
Total 738.48± 46.57E 502.64± 15.13FG 892.08± 43.02D 723.81±75.55E 1579.19± 53.37B 795.46± 28.18DE 1322.85±98.80C 1341.70±94.88C 1887.71±14.89A 1460.17±17.57BC
Aldehydes
10 2-Methylbutanal 12.87± 2.36DEF 13.63±1.07CDEF 18.58±1.32BCDE 14.78± 2.85CDEF 19.73± 1.46BC 25.11± 3.21B 18.83± 2.86BCDE 24.48± 0.72B 44.77± 2.94A 40.71± 4.05A
11 3-Methylbutanal 47.18± 4.76EFG 39.76±3.71GH 89.00±3.45D 61.16± 8.44E 113.90± 6.01BC 104.47± 2.92CD 96.20± 3.84D 125.18± 4.12AB 138.82± 2.11A 121.71± 8.29B
12 Pentanal 128.54± 6.19BC 135.89± 6.09B 115.66± 5.25BCD 104.88±4.09CDE 110.01± 5.29CDE 89.10± 0.95E 113.41± 13.36BC- 128.79± 6.44BC 95.52± 3.86DE 117.33± 7.99BCD
DE
13 Hexanal 191.78± 14.84DEF 162.76± 4.62FGH 178.35± 8.71EFG 153.03±3.27GHI 138.56± 11.29HIJ 128.82± 2.46IJ 155.26± 15.99GHI 178.77± 11.87EFG 113.75± 5.49J 135.14± 8.35HIJ
14 Heptanal 24.46± 2.11FGH 19.72±1.18H 27.89±0.37CDEFG 21.93± 2.49GH 24.34± 2.46FGH 24.15± 2.59FGH 27.06± 2.66EFGH 33.97± 0.91BCDE 27.44± 3.63DEFG 30.54± 2.21BCDEF
15 Octanal 29.99± 0.73EFG 22.40±1.00G 38.97±1.00BC 36.77± 3.53CDE 38.93± 1.22BC 36.89± 5.74CDE 37.63± 4.40CDE 40.63± 1.77BC 38.71± 3.36BCD 30.67± 0.40DEF
16 Nonanal 58.80± 6.31BC 41.50±4.00C 67.74±6.13BC 52.21± 8.14BC 46.94± 2.26BC 83.12± 27.26ABC 66.21± 18.27BC 61.49± 6.52BC 131.49± 27.89A 63.11± 6.44BC
17 Benzaldehyde 86.48± 3.83EFGH 81.56±6.82FGH 110.03± 2.86BCD 89.33± 8.18EFGH 112.20± 4.36ABC 99.05± 5.63CDEF 114.24± 8.22ABC 115.94± 4.16ABC 126.94± 13.90AB 130.59± 4.14A
18 Benzeneacetaldeh- 163.81± 12.86C 40.34±1.96F 207.65± 5.54B 82.21± 8.73E 264.64± 14.95A 77.04± 3.08E 110.05± 16.04D 229.86± 8.14B 159.16± 8.52C 79.04± 2.12E
yde
19 Methyl phenyl 16.69± 2.22FG 11.30±2.05GH 24.84±2.14EF 32.45± 3.07DE 45.69± 4.68B 35.78± 4.74CD 37.16± 4.19CD 42.45± 1.24BC 50.40± 2.05AB 55.38± 3.77A
pentenal
20 5-Methyl-2-phenyl- 74.90± 7.82EF 41.61±8.94FG 126.73± 9.41DE 151.65±35.42CD 307.41± 47.56A 170.50± 14.16CD 184.43± 22.13BC 228.96± 13.64B 135.34± 11.14CD 162.10± 11.41CD
2-hexenal
Total 835.52± 43.38E 610.47± 34.10F 1005.42± 10.46BC 800.40±56.86E 1222.36± 34.22A 874.03± 31.49DE 960.50± 35.84BCD 1210.53±41.90A 1062.34±73.68B 966.34± 26.49BCD
Ketones
21 2-Heptanone 28.56± 2.89CDEF 32.25±1.05BCDE 27.84±2.30CDEF 28.27± 2.08CDEF 23.96± 2.47F 25.87± 1.69DEF 25.12± 4.18EF 26.23± 2.16DEF 31.63± 3.09BCDEF 35.78± 4.03BC
22 3-Octanone 0.00±0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00±0.00B 0.00±0.00B 0.00±0.00B 0.00±0.00B
23 3-Hydroxy-2- 57.44± 4.51A 56.38±6.04A 34.93±4.85B 27.48± 2.14BC 0.00± 0.00# 0.00± 0.00# 0.00±0.00# 0.00±0.00F 0.00±0.00F 0.00±0.00F
butanone
24 2-Octanone 0.00±0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00±0.00B 0.00±0.00B 0.00±0.00B 0.00±0.00B
25 3-Octen-2-one 0.00±0.00F 0.00± 0.00F 28.39±1.16D 24.58± 3.67D 65.88± 3.71B 47.87± 6.46C 55.02± 2.41C 70.41± 2.63B 96.63± 4.61A 70.86± 4.19B
26 2-Nonanone 14.52± 0.27C 15.90±0.42C 20.76±0.91B 21.58± 1.55B 0.00± 0.00# 0.00± 0.00# 0.00±0.00# 0.00±0.00D 0.00±0.00D 0.00±0.00D
27 Acetophenone 28.73± 1.48EFGH 32.36±0.99CDEF 39.31±0.64BCDE 45.30± 4.68BCD 48.47± 1.59B 54.49± 5.07B 53.99± 1.23B 47.05± 2.50BC 86.02± 9.08A 76.51± 14.98A
Total 129.25± 2.05C 136.89± 6.00C 151.22± 4.16C 147.21±11.83C 138.31± 6.90C 128.23± 12.49C 134.14± 5.59C 143.69± 7.29C 214.27± 16.41A 183.15± 22.94B
Pyrazines
28 Methylpyrazine 0.00±0.00C 0.00± 0.00C 0.00± 0.00C 0.00± 0.00C 12.84± 1.35B 15.13± 1.81B 16.13± 2.80B 22.79± 3.16B 38.57± 11.87A 22.46± 3.57B
29 2,5- 0.00±0.00H 0.00± 0.00H 12.99±0.49FG 10.44± 1.65G 19.63± 1.18DE 18.71± 1.44E 26.12± 0.80C 22.21± 0.89D 33.41± 2.59B 40.19± 1.67A
Dimethylpyrazine
30 2,6- 0.00±0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 17.53± 0.65B 19.62± 1.53B 18.57± 1.89B 19.30± 1.59B 37.12± 3.49A 33.76± 1.97A
Dimethylpyrazine
(continued on next page)
M. Hinneh et al. Food Research International 119 (2019) 84–98
88
Table 3a (continued)
N- Volatile 0PS-100 °C 7PS-100 °C 0PS-120 °C 3PS-120 °C 0PS-135 °C 3PS-135 °C 7PS-135 °C 0PS-140 °C 0PS-160 °C 7PS-160 °C
o.
31 2-Ethyl-6- 0.00±0.00E 0.00± 0.00E 6.01± 0.45DE 4.50± 0.69DE 14.58± 1.34C 9.76± 1.66CD 14.57± 1.51C 15.95± 1.35C 38.41± 1.78B 34.98± 3.79B
methylpyrazine
32 2-Ethyl-5- 0.00±0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 37.81± 1.16BC 31.82± 2.89C 41.74± 1.51B 45.71± 2.77B 69.77± 2.64A 74.19± 9.10A
methylpyrazine
33 Trimethylpyrazine 10.95± 0.97H 13.71±1.10GH 25.76±0.81EFGH 21.08± 1.65FGH 63.77± 4.13C 47.84± 3.03CDE 134.19± 2.70B 63.11± 5.09C 151.72± 8.46B 345.28± 28.80A
34 2,6-Diethyl-3- 0.00±0.00F 0.00± 0.00F 0.00± 0.00F 0.00± 0.00F 209.63± 19.22C 116.82± 23.22E 183.51± 31.55CD 166.80± 8.52D 349.66± 19.98B 488.17± 15.51A
ethylpyrazine
35 Tetramethylpyrazi- 178.26± 23.40FG 581.62± 9.37C 197.79± 8.92EF 150.22±29.12FGH 277.23± 3.21DE 103.26± 7.94GH 966.32± 63.36B 223.28± 7.86DEF 306.64± 27.86D 1467.80±79.11A
ne
36 3,5-Diethyl-2- 0.00±0.00F 0.00± 0.00F 20.13±0.90DE 13.15± 2.25E 39.37± 5.40B 27.13± 1.21CD 36.67± 4.64BC 35.74± 0.76BC 87.60± 10.30A 78.01± 6.40A
methylpyrazine
37 2,3,5-Trimethyl-6- 30.72± 3.87D 37.24±2.06D 49.86±1.72D 36.29± 2.13D 107.94± 1.38C 36.58± 5.38D 136.54± 7.28B 95.57± 4.69C 154.58± 14.28B 305.20± 21.20A
ethylpyrazine
38 2,5-Dimethyl-3-n- 0.00±0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 35.15± 8.84C 24.88± 1.34C 34.45± 1.24C 27.22± 8.81C 64.47± 8.92B 99.42± 18.32A
pentylpyrazine
Total 219.93± 22.79GH 632.58± 11.08E 312.54± 11.95G 235.69±30.95GH 835.48± 24.91D 451.54± 41.22F 1608.82±63.47B 737.68± 20.06DE 1331.95±53.81C 2989.47±134.56A
Esters
39 Methyl acetate 32.38± 3.61BCD 25.70±5.58DEF 19.85±0.88F 20.32± 2.76EF 24.70± 1.97DEF 25.57± 1.94DEF 26.39± 2.05DEF 28.72± 2.71CDEF 54.67± 3.99A 41.19± 3.62B
40 Ethyl octanoate 11.68± 1.34CD 9.67± 0.69DE 17.82±0.38A 15.95± 1.74AB 0.00± 0.00G 0.00± 0.00G 0.00±0.00G 0.00±0.00G 0.00±0.00G 0.00±0.00G
41 Methyl phenyl 0.00±0.00C 0.00± 0.00C 0.00± 0.00C 0.00± 0.00C 18.67± 3.84B 22.67± 7.43B 17.72± 3.64B 20.17± 0.43B 44.66± 0.97A 46.71± 2.91A
acetate
42 Ethyl phenyl 40.35± 4.46EFG 56.45±3.51CDE 44.85±3.86EF 49.32± 13.73DEF 70.12± 9.51C 65.55± 7.59CD 97.99± 5.01B 75.15± 5.65C 74.31± 9.27C 118.35± 7.06A
acetate
43 Phenyl ethyl 99.28± 6.74EFGH 83.45±5.60FGHI 103.91± 2.26DEFG 64.51± 12.44IJ 123.59± 7.07BCDE 53.24± 2.37J 106.95± 14.60DEF 109.98± 8.94DEF 128.75± 8.61BCD 139.89± 9.72BC
acetate
44 Amyl benzoate 56.76± 6.76BCDE 71.27±6.35ABC 51.16±1.52CDEF 76.55± 11.67ABC 57.74± 5.01BCDE 70.54± 7.15BCD 78.73± 14.05AB 60.22± 3.42BCDE 73.90± 15.46ABC 97.22± 12.84A
Total 240.45± 8.11DEFG 246.54± 4.44DEF 237.59± 8.03EFGH 226.65±39.62FGH 294.81± 9.20CD 237.58± 20.76EF- 327.78± 37.45BC 294.24± 12.20CDE 376.30± 8.23B 443.37± 17.32A
GH
Terpenes and terpenoids
45 Delta-3-carene 9.96±1.26ABCD 8.28± 0.10BCDE 12.00±0.49AB 8.28± 1.52BCDE 7.32± 0.32DE 8.30± 0.59BCDE 7.35±1.06CDE 10.11± 0.61ABCD 13.20± 1.35A 12.84± 4.02A
46 Myrcene 5.85±1.19EF 6.41± 0.45DEF 10.24±0.84A 8.12± 0.61BCD 8.65± 0.46ABC 8.85± 0.89AB 8.58±0.47ABC 10.06± 1.47AB 9.66±0.89AB 10.58± 0.17A
47 Limonene 14.94± 1.03CDE 10.78±1.15EF 11.49±0.28DEF 10.30± 2.38F 9.66± 1.01F 12.68± 0.79DEF 11.28± 0.73EF 12.18± 1.46DEF 19.84± 1.49B 17.73± 0.84BC
48 Cis-ocimene 11.23± 1.44E 11.10±0.90E 19.15±1.64BC 16.29± 0.90CD 20.69± 2.14ABC 21.24± 2.63AB 22.15± 1.00AB 22.73± 2.16AB 24.43± 0.73A 24.22± 1.77A
49 Trans-ocimene 0.00±0.00B 0.00± 0.00B 5.37± 0.58AB 4.88± 0.84AB 6.28± 1.02AB 6.11± 0.19AB 5.03±0.26AB 5.24±0.59AB 5.13±1.60AB 10.67± 8.56A
50 Linalool oxide 23.55± 6.49CD 22.62±1.82CD 27.84±4.07ABC 37.33± 8.99ABC 43.25± 11.46A 25.47± 2.59BCD 32.30± 5.09ABC 28.99± 6.78ABC 29.15± 7.83ABC 39.76± 2.41AB
51 Dihydromyrcenol 0.00±0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00±0.00B 0.00±0.00B 0.00±0.00B 0.00±0.00B
52 Linalool 74.91± 10.99CD 69.19±3.91D 93.47±7.97BCD 91.92± 5.17BCD 113.34± 5.72AB 93.09± 7.36BCD 117.01± 5.72AB 99.75± 6.02ABC 113.72± 7.08AB 121.33± 24.83A
53 Epoxylinalool 42.23± 5.69B 36.47±9.42B 31.81±3.30B 39.04± 8.06B 40.66± 3.60B 30.64± 7.39BC 39.56± 5.60B 36.16± 6.58B 45.82± 5.16B 62.66± 8.17A
Total 182.68± 22.76DEF 164.85± 13.77EF 211.37± 8.13CD 216.16±10.99BCD 249.85± 14.42BC 206.38± 16.13CDE 243.26± 6.83BC 225.21± 9.30BCD 260.95± 14.76AB 299.79± 41.18A
Furans, furanones, pyrans, pyrones
54 2-Pentyl furan 47.17± 2.38GH 53.87±5.08GH 65.96±3.08DEF 43.10± 1.69H 43.54± 4.84H 55.82± 2.24FG 45.82± 3.46GH 69.68± 5.13CD 70.44± 3.83CD 68.91± 0.50CD
55 Rose oxide 0.00±0.00F 0.00± 0.00F 65.26±1.53D 52.58± 5.05D 134.24± 12.68B 114.38± 10.77C 138.39± 7.35B 147.14± 5.55B 211.72± 10.45A 202.66± 12.08A
56 γ-Butyrolactone 0.00±0.00E 0.00± 0.00E 24.64±10.57CDE 19.47± 5.04DE 47.40± 16.72BCD 58.54± 20.59B 54.64± 25.01BC 39.62± 1.13BCD 99.05± 9.43A 93.96± 15.95A
57 Furfuryl alcohol 0.00±0.00E 0.00± 0.00E 49.85±2.49D 52.63± 13.64CD 76.58± 4.99BC 78.92± 7.49B 94.76± 5.52AB 86.39± 9.57B 114.05± 5.63A 91.61± 9.93AB
58 5-Ethyl-2(5H)- 0.00±0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 0.00±0.00B 0.00±0.00B 0.00±0.00B 0.00±0.00B
furanone
59 Tetrahydro-2H- 0.00±0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 27.68± 4.92BC 24.72± 3.45C 33.34± 4.94B 29.81± 3.17BC 55.34± 1.47A 31.94± 1.13B
pyran-2-one
60 Pantoic lactone 0.00±0.00F 0.00± 0.00F 36.09±1.41CDE 35.06± 4.09CDE 30.63± 4.00CDE 23.68± 1.54E 50.54± 23.16BC 46.25± 2.93BCD 65.80± 2.22B 101.24± 5.52A
(continued on next page)
M. Hinneh et al. Food Research International 119 (2019) 84–98
(AHC) to illustrate the (dis)similarities between the different chocolates
in terms of their aroma profiles. Hereby, Fig. 1a and b represent clus-
tering on the basis of all identified volatiles and odor-active volatiles,
respectively.
3.1. Acids
The fermentation of cocoa beans facilitates the formation of both
volatile and non-volatile acids which imparts sourness and astringency
with inimical implications on the flavor quality of cocoa/chocolates
(Afoakwa et al., 2008; Frauendorfer & Schieberle, 2008). Table 3a and
3b show the concentrations of aroma volatiles identified from both
sample chocolates and those from commercial cocoa liquors respec-
tively. Generally, chocolates from unstored pods expressed higher total
acid concentrations than those from stored pods. Hinneh et al. (2018)
attributed this to the impact of pulp volume reduction on the quality of
the fermentation process, thus, resulting in minimal acid production
after PS. However, for a rise in acidity between 3PS and 7PS, they at-
tributed this to the additional consequence of an onset of cellular de-
gradation of both pod and pulp (fermentation-like process), in which
case it was associated with a gradual rise in acidity. This trend was
consistent at all temperatures with an exception of RT of 160 °C where a
leveling-off effect was observed, probably due to the influence of this
elevated temperature. Generally, the total concentration of volatile
acids increased insignificantly (p > .05) with increasing RT, except at
temperatures of 135 °C and 160 °C where a significant increase
(p < .05) was observed. Through the mechanism of volatilization, the
roasting processes has been claimed to reduce acidity in cocoa beans
(Afoakwa et al., 2008; Rodriguez-Campos et al., 2012). However, it is
likely that the heating process may have resulted in unlocking these
volatiles from the matrix, thereby making them more ‘available’ in the
liquor and subsequently in the chocolate. Afoakwa (2016), described
the impact of roasting on the structural changes within the bean,
whereby, the extensive loss of moisture ultimately renders the beans
friable, accompanied with loose and a porous matrix. Therefore, the
degree of change along with the liberation of volatiles from the matrix
can be said to be proportional to the roasting intensity. Meanwhile, the
chocolates from the commercial cocoa liquors possessed generally
lower concentrations of volatile acids. Of these, the Ghanaian, Vietna-
mese and Madagascan chocolates possessed significantly similar
(p > .05) level of acids with 3PS-120 °C, 3PS-135 °C and 7PS-100 °C
chocolates. Meanwhile, both Venezuelan and Ivorian chocolates re-
corded the lowest (p < .05) total acid concentrations. Among the
identified acids, only acetic acid and isovaleric acid were identified as
odor-active volatiles (Table 4). However, due to its higher concentra-
tions, the major contribution of acetic acid to the overall sourness of the
chocolates could be suspected.
3.2. Alcohols
The trend in the total alcohol concentration was similar to the vo-
latile acids, where, for any given RT, chocolates with pod stored cocoa
beans possessed significantly lower (p < .05) total alcohol concentra-
tion compared to chocolates from the unstored pods (Table 3a). For the
same reason aforementioned, the significant rise (p < .05) in total
alcohols from 3PS to 7PS was also logical. High alcohol contents are
essential in developing chocolates with flowery and candy notes
(Rodriguez-Campos et al., 2012). Thus, the results suggest that very
minimal or no PS may be ideal in conserving more alcohols for any RT.
Likewise, a significant increasing (p < .05) trend in total alcohol with
increasing RT was also observed. Apart from the possible impact of
roasting on the liberation of volatiles as suggested earlier, Ramli,
Hassan, Said, Samsudin, and Idris (2006) also observed an increasing
linear relationship between alcohol concentration and RT. Of these,
they speculated that alcohols such as 2-heptanol and linalool may be
recognized as volatiles of thermally degradable amino acids.
89
Table 3a (continued)
N- Volatile 0PS-100 °C 7PS-100 °C 0PS-120 °C 3PS-120 °C 0PS-135 °C 3PS-135 °C 7PS-135 °C 0PS-140 °C 0PS-160 °C 7PS-160 °C
o.
61 2,3-Dihydro-3,5- 0.00±0.00E 0.00± 0.00E 39.70±10.27CD 0.00± 0.00E 37.31± 3.15CDE 33.62± 4.31DE 53.44± 20.49BCD 170.03± 40.69A 74.26± 8.79BC 82.10± 11.52B
dihydroxy-6-
methyl-4H-pyran-
4-one
Total 47.17± 2.38H 53.87±5.08H 281.51± 17.41EF 202.85±22.45FG 397.38± 14.30CD 389.67± 29.31D 470.92± 76.68C 588.92± 49.31B 690.66± 11.12A 672.44± 9.31A
Pyrroles
62 Isoamyl-2-formyl 0.00±0.00E 0.00± 0.00E 0.00± 0.00E 0.00± 0.00E 44.46± 8.60BC 25.37± 10.37D 37.03± 3.18BCD 31.64± 4.27CD 52.19± 10.01B 148.55± 11.52A
pyrrole
63 2-Acetylpyrrole 19.29± 0.65E 19.29±0.70E 42.71±1.02CD 43.39± 2.88CD 87.08± 3.59B 55.29± 1.74C 92.22± 10.26B 83.22± 4.08B 116.65± 6.07A 126.93± 7.92A
64 2-Formylpyrrole 0.00±0.00C 0.00± 0.00C 0.00± 0.00C 0.00± 0.00C 64.31± 16.11A 44.13± 4.46B 61.95± 8.81A 51.82± 2.56AB 59.49± 2.71AB 57.90± 7.57AB
Total 19.29± 0.65G 19.29±0.70G 42.71±1.02FG 43.39± 2.88FG 195.85± 26.16BC 124.79± 14.78D 191.20± 16.15C 166.68± 6.52C 228.33± 17.33B 333.38± 22.10A
Others
65 Dimethyldisulfide 0.00±0.00C 0.00± 0.00C 13.23±0.77B 12.74± 1.75B 13.01± 1.45B 15.18± 1.47B 14.42± 1.63B 13.24± 0.73B 24.45± 2.40A 22.52± 1.47A
66 o-Xylene 9.17±2.01ABCD 8.77± 0.38ABCDE 9.43± 2.27ABCD 7.74± 1.79BCDE 7.50± 0.13CDE 8.41± 1.11ABCDE 6.50±0.62DE 13.30± 2.40A 12.47± 2.96AB 12.36± 2.93ABC
67 Pyridine 0.00±0.00D 0.00± 0.00D 3.35± 0.87ABC 2.81± 0.79BC 2.82± 1.07BC 2.60± 0.85BC 2.53±0.34C 2.90±0.33BC 4.94±0.79A 4.44±0.92AB
68 2-Methoxyphenol 0.00±0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 24.13± 7.85BC 16.93± 3.97C 29.61± 1.95BC 31.66± 4.91B 36.85± 1.64B 52.07± 15.80A
69 Pyrrolidinone 0.00±0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 31.44± 2.88C 25.84± 1.96CD 30.17± 3.72C 29.51± 4.29CD 65.10± 10.05B 125.96± 36.43A
Total 9.17±2.01G 8.77± 0.38G 26.01±1.37FG 23.29± 4.25FG 78.90± 5.01CD 68.97± 5.12CDE 83.23± 2.69CD 90.61± 3.59C 143.81± 9.76B 217.33± 44.99A
Grand total 5610.87±346.36F 4805.39± 123.78- 6480.46± 124.59E 5123.13± 244.52- 9049.06± 400.27- 5765.68±182.63- 9277.11±458.77- 8184.53±174.66- 10,518.86± 283.- 12,364.38± 86.20A
G FG C EF C D 81B
For each row different alphabets represent significant differences (α= 0.05) among samples. (#) peak was present but too small to be integrated.
M. Hinneh et al. Food Research International 119 (2019) 84–98
Additionally, they also noted that the prolonged heat duration of the beans due to PS and its link with the abundance of pyrazines formed
roasting may be responsible for the volatilization of some alcohols, yet, during the roasting process. Remarkably, the concentration of pyrazines
this was only observed within the range of 160 °C to 170 °C. Comparing in 7PS increased over two folds at most RT's. Suggestively, the pro-
the chocolates from the commercial cocoa liquors to the sample cho- longed period of PS may have facilitated the formation of more pre-
colates, significantly lower (p < .05) concentrations of the alcohols cursors which enhanced the formation of these pyrazines during the
were recorded. Furthermore, of these chocolates from the commercial subsequent roasting process. More so, the same increasing effect of RT
liquors, the differences were insignificant (p > .05), except for Ghana on the pyrazine concentration was observed. Apart from tetra-
and Venezuela. Among all the alcohols, 2-phenylethyl alcohol was the methylpyrazine and trimethylpyrazine which are reported to partly
only odor-active volatile, thus, it is expected to impart its typical floral evolve from microbiological activities during fermentation (Jinap, Siti,
note to the chocolates (Tables 4a and 4b). & Norsiati, 1994; Schwan & Wheals, 2004), the majority of pyrazines
originate from the Strecker degradation in the Maillard reaction.
3.3. Aldehydes and ketones Temperature as well as abundance of aroma precursors are critical
factors that influence their concentrations (Afoakwa et al., 2008;
Aldehydes and ketones which are formed from the Strecker de- Aprotosoaie et al., 2016). This may explain why the RT appeared to
gradation reaction during roasting, are crucial for the expression of play a substantial role in in the concentration of the pyrazines. Tetra-
good cocoa flavor. Table 3a shows a similar trend of total aldehydes methylpyrazine was the most dominant pyrazine in all chocolates with
with PS as seen for the alcohols. Hereby, the concentrations of the al- its highest concentration observed in 7PS-160 °C. Nevertheless, the
dehydes were higher in chocolates of 0PS compared to other days of PS contributing effect of this volatile to the flavor quality of the cocoa was
for any given RT, except at 160 °C where no significant (p > .05) dif- negligible due to its low OAV (Table 4a and 4b). Trimethypyrazine was
ference was seen for the unstored and stored pods. More so, the impact odor-active exclusively in 7PS-160 °C chocolate. Methylpyrazine and
of the RT was also seen as the aldehyde concentration increased with 2,6-dimethyl pyrazine were among the pyrazines absent in chocolates
increasing RT until a peak between 135 and 140 °C. Between 140 °C and whose initial cocoa beans were roasted at temperatures below 135 °C.
160 °C, the total concentration declined, possibly, due to the higher This is likely due to the lower concentrations of the volatiles produced
temperature. Accordingly, Aprotosoaie et al. (2016), stated that high in the liquors after roasting, more so, the heating effect and the pro-
temperatures and a longer roasting conditions are known to decrease cessing conditions during the conching may have enhanced their
the content of aldehydes. It can be seen from Tables 3a and 3b that a complete elimination from these chocolates. Some pyrazines were ab-
total of eleven aldehydes were identified. Of these, key aldehydes, sent in many of the chocolates from the commercial liquors. Of these,
namely; 2-methylbutanal and 3-methylbutanal, which were both the Ghanaian and Vietnamese chocolates were marked among the ori-
dominant in 0PS-160 °C chocolate, increased trivially with increasing gins as they exhibited significantly higher (p < .05) total pyrazine
RT. Both aroma volatiles were also found to be odor-active and most concentration (Table 3b).
likely to impart their characteristic cocoa and chocolate notes to the
final flavor (Frauendorfer & Schieberle, 2008). It is worth mentioning 3.5. Esters, terpenes and terpenoids
that despite registering generally lower concentrations in other key
aldehydes, the concentrations of odor-active hexanal and heptanal were Esters are associated with fruity/floral notes and represent the
generally higher in the chocolates from commercial cocoa liquors than second most important group of volatile compounds after pyrazines in
in the sample chocolates. Here, the Ecuadorian chocolate recorded roasted cocoa nibs (Jinap, Wan-Rosli, Russly, & Nordin, 1998). Apart
significantly higher (p < .05) concentrations for both volatiles. Not from methyl acetate, ethyl octanoate and phenyl ethyl acetate, 7PS-
only was the Madagascan chocolate also dominant with respect to 160 °C chocolates exhibited the highest concentrations of all other es-
heptanal, but together with the Ecuadorian chocolate, also recorded the ters (Table 3a). Ethyl octanoate was exclusive to the chocolates from
highest total concentration of the aldehydes among the chocolates from commercial liquors and sample chocolates with lower RT's
the commercial liquors. (100–120 °C). Ethyl octanoate was significantly pronounced (p < .05)
Except at a temperature of 160 °C, both RT and PS had no significant in both 0PS-120 °C and 3PS-120 °C chocolates. It was odor-active and
(p > .05) effect on the total concentration of ketones. For the sample thus expected to contribute to the final chocolate flavor (Table 4a and
chocolates, identified ketones including; 2-nonanone and 3-hydroxy-2- 4b). Phenyl ethyl acetate was the most abundant ester, being more
butanone, appeared dominant in chocolates with lower RT's. For these pronounced in the Vietnamese chocolate (Table 3b). This volatile
volatiles, quantifiable amounts were observed at 100 °C and 120 °C. confers honey and floral notes (Rodriguez-Campos et al., 2012), but was
This may be attributed to the high volatility of these aroma compounds. only odor-active in the Ecuadorian, Vietnamese and 7PS-160 °C cho-
On the contrary, acetophenone and 3-octen-2-one were produced in colates. It is worth mentioning that most esters occur as microbial
abundance at higher temperatures. Although the trend underlying the metabolites which may arise from the fermentation process
effect of both PS and RT were indefinite, the highest concentrations of (Aprotosoaie et al., 2016). Yet, it seemed that both the PS and the RT
the ketones were obtained at higher temperatures (~ 160 °C) for 0PS. impacted the concentrations of these esters as they may have promoted
From Table 3b, the Ghanaian chocolate was found to possess higher their formation and release respectively. However, these effects were
total concentration of ketones among the chocolates from commercial significant above 135 °C, beyond which the abundance of esters asso-
liquors. ciated with 7PS chocolate was conspicuous. More so, the abundance of
esters in the chocolates from commercial liquors were fairly compar-
3.4. Pyrazines able to the sample chocolates. Of the former, the Vietnamese chocolate
recorded generally higher concentration of esters.
Pyrazines are a major group of volatiles which are formed through No significant effect (p > .05) of PS on the total concentration of
various heterocyclization reactions following the Strecker degradation terpenes and terpenoids was found. Likewise, a trivial increase in the
reaction during the roasting process. They are typically known to confer total concentration these volatiles with RT was also found (Table 3a).
cocoa, chocolate and nutty flavor notes. From the sample chocolates in Apart from linalool, other terpenes occurred in minimal amounts in the
Table 3a, a trend revealing a direct relationship between PS and the chocolates with 7PS-160 °C chocolate recording the highest total con-
total pyrazine concentration was found, except for the short duration of centration. It also increased with increasing RT. Just like the other
PS (ie. 0–3 days) where the opposite was found. This was synonymous terpenes, it is probable that linalool became less bound within the
with the findings of Hinneh et al. (2018) who demonstrated the mod- matric following an increase in temperature. It was identified as odor-
ification of aroma precursors (reducing sugars and free amino acids) in active in most chocolates and likely to impart its characteristic floral
90
M. Hinneh et al. Food Research International 119 (2019) 84–98
91
Table 3b
Concentrations (ng/g cocoa) of aroma volatiles identified from dark chocolates produced from commercial cocoa liquors.
No. Volatile Ecuador Ghana Ivory coast Madagascar Venezuela Vietnam Description* KI
Acids
1 Acetic acid 3000.67± 118.90C 2405.33± 180.97EF 1479.63± 95.61H 2433.68± 199.92DEF 1821.11±134.11GH 2478.24± 106.48DEF Sour, vinegar 1418.21
2 Isovaleric acid 47.69± 3.58F 177.29±54.46D 14.14± 0.33F 56.29± 9.63F 53.28± 7.09F 86.37±18.30EF Sweat, rancid 1607.91
3 Oxalic acid DITMS 112.94± 10.25BC 124.14±1.84BC 113.91±24.46BC 112.13±21.25BC 67.35± 8.08C 132.23± 5.63BC – 1357.99
Total 3161.30± 130.83C 2706.76± 235.68D 1607.69± 120.28E 2602.10± 172.15D 1941.73±133.99E 2696.84± 127.92D
Alcohols
4 Amyl alcohol 36.04± 1.23CD 54.43± 1.01A 41.98± 4.25BC 27.58± 2.91DE 35.64± 0.87CD 20.94±1.53EF Banana 1236.36
5 1-Hexanol 8.87± 0.17B 0.00± 0.00D 0.00± 0.00D 9.74± 0.15A 7.65±0.16C 7.70± 0.79C Fruity, green 1340.18
6 2,3-Butanediol 21.89± 2.25F 68.92± 19.15DEF 69.53± 9.71DEF 42.55± 0.68F 55.12± 0.85DEF 56.95±3.69DEF Cocoa butter 1507.95
7 1,3-Butanediol 87.29± 7.23F 137.10±14.34CDEF 132.89±19.34DEF 115.32±11.16EF 83.34± 2.39F 138.52± 5.14CDEF Sweet, flowery, caramel 1540.13
8 Benzyl alcohol 86.16± 1.28EFGH 96.63± 7.36EFGH 55.05± 9.91GH 97.78± 6.92DEFG 53.07± 1.98H 79.02±3.54FGH Sweet, floral 1767.54
9 2-Phenylethyl alcohol 181.16± 6.36GH 190.10±6.88G 137.21±33.00GH 134.11±17.90GH 119.80± 9.29GH 174.42± 3.38GH Floral 1793.64
Total 421.41± 3.43FG 547.18±43.20F 436.66±68.35FG 427.08±14.70FG 354.60± 12.08G 477.55± 10.43FG
Aldehydes
10 2-Methylbutanal 12.79± 1.85EF 16.25± 1.37CDEF 19.41± 1.57BCD 12.72± 1.25EF 11.96± 0.67F 16.49±1.55CDEF Cocoa, chocolate 915.12
11 3-Methylbutanal 37.77± 3.40GH 58.12± 6.46EF 43.97± 0.81FGH 33.14± 6.68GH 31.17± 2.80H 55.64±7.24EF Cocoa, chocolate 917.15
12 Pentanal 186.76± 7.65A 164.55±0.67A 89.50± 7.30E 162.26±23.68A 96.15± 1.91DE 122.76± 3.15BC Pungent 953.13
13 Hexanal 376.68± 14.26A 253.21±6.34C 171.20±6.38FG 325.30±19.30B 207.47± 1.46DE 212.98± 0.63D Green 1047.75
14 Heptanal 52.01± 2.39A 35.84± 1.45B 34.75± 5.17BCD 54.36± 1.94A 37.32± 1.90B 35.09±2.15BC – 1158.50
15 Octanal 50.89± 2.29A 34.25± 2.89CDEF 32.57± 3.58CDEF 46.12± 2.32AB 33.59± 1.21CDEF 28.22±1.28FG Fatty, waxy 1268.20
16 Nonanal 71.80± 1.95BC 66.53± 11.96BC 84.82± 36.40ABC 97.86± 36.06AB 77.37± 10.63BC 54.70±4.54BC Soapy 1372.10
17 Benzaldehyde 71.53± 5.23HI 104.33±2.62CDE 93.61± 6.65DEFG 75.47± 5.85GHI 62.85± 1.42I 71.47±0.70HI Almond, burnt sugar 1477.26
18 Benzeneacetaldehyde 16.44± 1.06G 36.31± 3.49FG 17.33± 3.46FG 26.46± 3.48FG 16.75± 1.59G 31.71±1.43FG Honey, sweet, rose, flora 1804.38
19 Methyl phenyl pentenal 8.62± 0.78GH 10.74± 0.77GH 8.90± 1.19GH 16.50± 3.53FG 7.55±0.93H 10.05±1.23GH Cocoa 1814.71
20 5-Methyl-2-phenyl-2-hexenal 21.11± 2.25FG 36.38± 4.24FG 15.44± 1.20G 60.42± 15.84FG 11.13± 0.77G 30.47±1.24FG Cocoa 1908.33
Total 906.41± 25.88CDE 816.50±15.63E 611.51±47.08F 910.62±18.15CDE 593.31± 19.65F 669.58± 16.43F
Ketones
21 2-Heptanone 36.94± 2.59B 33.87± 3.96BCD 29.70± 2.02BCDEF 48.78± 2.64A 24.19± 1.19EF 35.33±2.67BC Fruity, floral 1155.99
22 3-Octanone 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 11.59± 1.62A 0.00±0.00B 0.00± 0.00B – 1233.48
23 3-Hydroxy-2-butanone 7.87± 1.11E 26.83± 2.13C 19.18± 2.58D 0.00± 0.00F 0.00±0.00F 0.00± 0.00F Butter, cream 1254.32
24 2-Octanone 0.00± 0.00B 0.00± 0.00B 9.52± 0.62A 0.00± 0.00B 0.00±0.00B 0.00± 0.00B – 1264.97
25 3-Octen-2-one 0.00± 0.00F 10.19± 0.77E 0.00± 0.00F 0.00± 0.00F 0.00±0.00F 0.00± 0.00F – 1343.52
26 2-Nonanone 27.85± 1.93A 28.96± 1.61A 20.57± 3.14B 23.24± 0.18B 21.46± 2.01B 22.50±0.57B Milk, green, fruity 1368.61
27 Acetophenone 14.86± 1.58HI 30.25± 2.82DEFG 15.75± 1.60GHI 14.58± 2.24HI 13.41± 0.40I 22.68±4.42FGHI Floral, almond 1587.15
Total 87.52± 4.98DE 130.09±6.42C 94.73± 5.59D 98.19± 4.87D 59.05± 1.01E 80.52±6.64DE
Pyrazines
28 Methylpyrazine 0.00± 0.00C 0.00± 0.00C 13.58± 5.76B 0.00± 0.00C 0.00±0.00C 0.00± 0.00C Nutty, cocoa, roasted-nuts 1239.39
29 2,5-Dimethylpyrazine 18.21± 1.00E 17.69± 0.58E 0.00± 0.00H 17.57± 0.94E 11.80± 0.61FG 14.00±0.60F Cocoa, roasted nuts 1299.09
30 2,6-Dimethylpyrazine 0.00± 0.00D 0.00± 0.00D 12.21± 1.06C 0.00± 0.00D 0.00±0.00D 0.00± 0.00D Nutty, coffee, green 1304.66
31 2-Ethyl-6-methylpyrazine 3.66± 0.57DE 121.96±4.93A 4.37± 0.59DE 3.84± 0.61DE 2.90±0.25DE 124.64± 7.75A Roasted, coffee, hazelnut 1362.22
32 2-Ethyl-5-methylpyrazine 0.00± 0.00D 4.51± 0.37D 0.00± 0.00D 0.00± 0.00D 0.00±0.00D 6.12± 0.57D Nutty, raw potato 1368.64
33 Trimethylpyrazine 45.73± 3.57CDE 41.71± 3.66CDEF 19.31± 1.04FGH 35.10± 2.57DEFG 28.92± 0.80EFGH 54.90±1.72CD Cocoa, roasted nuts, sweet, 1381.21
smoky
34 2,6-Diethyl-3-ethylpyrazine 0.00± 0.00F 0.00± 0.00F 0.00± 0.00F 0.00± 0.00F 0.00±0.00F 0.00± 0.00F – 1435.11
35 Tetramethylpyrazine 189.34± 11.52EFG 626.85±36.22C 76.85± 11.61H 65.41± 4.68H 104.55± 6.30GH 552.69± 5.51C Cocoa, coffee, roasted 1448.42
36 3,5-Diethyl-2-methylpyrazine 18.96± 1.68DE 16.44± 2.05DE 16.63± 4.92DE 0.00± 0.00F 0.00±0.00F 19.56±0.86DE – 1465.49
37 2,3,5-Trimethyl-6-ethylpyrazine 47.28± 3.65D 43.36± 1.04D 38.76± 7.05D 0.00± 0.00E 41.53± 1.46D 48.35±2.03D Candy, sweet 1484.11
38 2,5-Dimethyl-3-n-pentylpyrazine 0.00± 0.00D 0.00± 0.00D 27.63± 1.95C 0.00± 0.00D 0.00±0.00D 0.00± 0.00D – 1602.94
Total 323.18± 18.51FG 872.53±45.55D 209.34±27.55GH 121.92±6.70H 189.70± 7.54GH 820.26± 5.12D
Esters
39 Methyl acetate 25.80± 1.22DEF 23.47± 3.45DEF 29.29± 5.05CDE 25.64± 3.61DEF 35.93± 0.75BC 25.48±1.17DEF – 888.28
40 Ethyl octanoate 11.76± 1.16CD 12.08± 1.18C 8.99± 0.98EF 7.11± 0.18F 9.45±0.34DEF 14.83±0.83B Fruity, floral 1415.50
(continued on next page)
M. Hinneh et al. Food Research International 119 (2019) 84–98
92
Table 3b (continued)
No. Volatile Ecuador Ghana Ivory coast Madagascar Venezuela Vietnam Description* KI
41 Methyl phenyl acetate 0.00± 0.00C 0.00± 0.00C 0.00± 0.00C 0.00± 0.00C 0.00±0.00C 0.00± 0.00C Sweet, honey, jasmine 1678.38
42 Ethyl phenyl acetate 31.12± 2.87FG 34.48± 1.12FG 0.00± 0.00H 39.45± 4.68EFG 25.87± 2.14G 39.84±1.17EFG Fruity, sweet 1700.45
43 Phenyl ethyl acetate 143.28± 13.05B 114.83±1.19CDE 56.81± 12.21IJ 78.61± 10.10GHIJ 76.75± 5.41HIJ 203.56± 8.57A honey, floral 1722.77
44 Amyl benzoate 38.91± 4.43EF 44.57± 2.52DEF 36.34± 8.27EF 30.14± 6.60F 41.21± 5.33EF 37.15±9.96EF Balsam, sweet 1742.59
Total 250.88± 16.37DEF 229.43±5.02FGH 131.43±19.77I 180.94±20.49HI 189.21± 12.78GH 320.85± 16.91BC
Terpenes and terpenoids
45 Delta-3-carene 11.20± 0.44ABC 12.60± 0.32A 6.00± 0.20E 6.00± 0.05E 10.15± 1.26ABCD 6.57± 0.64DE – 1115.82
46 Myrcene 4.64± 0.30F 6.80± 0.12CDE 0.00± 0.00G 4.50± 0.27F 5.22±0.47EF 0.00± 0.00G Balsamic, must, spicy, sweet 1139.26
47 Limonene 12.56± 0.53DEF 11.61± 1.61DEF 16.02± 2.06BCD 48.47± 2.51A 12.21± 2.13DEF 9.83± 1.90F Citrus-like 1169.41
48 Cis-ocimene 0.00± 0.00F 13.17± 1.70DE 10.16± 2.18E 0.00± 0.00F 0.00±0.00F 0.00± 0.00F Balsamic, peppery, spicy 1214.52
49 Trans-ocimene 0.00± 0.00B 5.38± 0.43AB 0.00± 0.00B 0.00± 0.00B 0.00±0.00B 0.00± 0.00B Balsamic, must, spicy, sweet 1230.91
50 Linalool oxide 0.00± 0.00E 9.89± 2.49DE 0.00± 0.00E 0.00± 0.00E 0.00±0.00E 0.00± 0.00E Fruity: citrus, floral 1446.42
51 Dihydromyrcenol 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 13.22± 1.52A 0.00±0.00B 0.00± 0.00B – 1452.19
52 Linalool 16.94± 2.11EF 70.17± 2.59D 38.32± 6.16E 0.00± 0.00F 16.94± 2.10EF 25.66±2.68E Floral, sweet 1517.66
53 Epoxylinalool 0.00± 0.00D 15.47± 2.43CD 0.00± 0.00D 0.00± 0.00D 0.00±0.00D 0.00± 0.00D Floral 1690.80
Total 45.33 145.09±5.23F 70.50± 9.12G 72.19± 3.18G 44.52± 1.25G 42.07±4.91G
± 1.85G
Furans, furanones, pyrans, pyrones
54 2-Pentyl furan 112.91± 7.48A 74.65± 2.38CD 67.73± 6.71CDE 99.34± 1.27B 78.65± 2.45C 56.70±0.98EFG Green bean, vegetable 1209.21
55 Rose oxide 0.00± 0.00F 27.64± 1.62E 0.00± 0.00F 9.12± 1.45EF 6.79±1.10F 18.47±1.57EF Floral: geranium-like 1337.81
56 γ-Butyrolactone 25.12± 0.87CDE 25.51± 0.25BCDE 27.59± 2.26BCDE 29.76± 5.79BCDE 24.57± 6.58CDE 27.42±4.67BCDE Cheesy 1564.28
57 Furfuryl alcohol 38.44± 6.73D 37.00± 7.94D 45.80± 19.26D 45.12± 2.67D 30.95± 2.91D 50.72±1.00D Faint burning 1603.11
58 5-Ethyl-2(5H)-furanone 0.00± 0.00B 0.00± 0.00B 0.00± 0.00B 14.32± 2.13A 0.00±0.00B 0.00± 0.00B – 1673.04
59 Tetrahydro-2H-pyran-2-one 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 0.00±0.00D 0.00± 0.00D – 1705.16
60 Pantoic lactone 29.68± 0.93DE 29.28± 3.66DE 30.81± 5.18CDE 32.55± 3.89CDE 23.45± 1.21E 47.21±5.38BCD – 1874.47
61 2,3-Dihydro-3,5-dihydroxy-6-methyl-4H-pyran-4- 37.75± 7.28CDE 19.57± 0.11DE 20.15± 6.70DE 55.73± 4.91BCD 25.20± 0.43DE 35.84±2.52DE Roasted 2030.28
one
Total 243.90± 4.68EFG 213.65±8.54EFG 192.08±14.04G 285.94±5.22E 189.61± 9.91G 236.36± 13.35EFG
Pyrroles
62 Isoamyl-2-formyl pyrrole 0.00± 0.00E 0.00± 0.00E 0.00± 0.00E 0.00± 0.00E 0.00±0.00E 0.00± 0.00E – 1714.35
63 2-Acetylpyrrole 44.68± 2.28C 42.18± 2.99CD 29.57± 6.49DE 55.11± 5.74C 29.77± 2.80DE 88.94±4.23B Caramel 1833.53
64 2-Formylpyrrole 0.00± 0.00C 0.00± 0.00C 0.00± 0.00C 0.00± 0.00C 0.00±0.00C 0.00± 0.00C – 1867.79
Total 44.68± 2.28FG 42.18± 2.99FG 29.57± 6.49FG 55.11± 5.74EF 29.77± 2.80FG 88.94±4.23E
Others
65 Dimethyldisulfide 12.16± 0.56B 12.93± 0.92B 0.00± 0.00C 0.00± 0.00C 0.00±0.00C 13.44±2.22B Sulfurous 1030.73
66 o-Xylene 12.30± 1.06ABC 9.15± 0.82ABCD 4.23± 0.57E 6.21± 0.57DE 5.18±0.72DE 7.22± 1.28DE – 1103.28
67 Pyridine 0.00± 0.00D 2.22± 0.40C 3.68± 0.13ABC 3.08± 0.23ABC 2.43±0.16C 3.01± 0.98BC – 1153.16
68 2-Methoxyphenol 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 0.00± 0.00D 0.00±0.00D 0.00± 0.00D Smoky 1752.67
69 Pyrrolidinone 42.21± 0.59BC 0.00± 0.00D 30.58± 8.26C 45.76± 4.78BC 43.93± 0.67BC 45.67±4.83BC – 1882.01
Total 66.66± 1.21CDE 24.30± 1.75FG 38.49± 8.89EFG 55.05± 4.84CDEF 51.54± 1.39DEF 69.35±2.03CDE
Grand total 5551.27± 64.78F 5727.71± 241.74F 3421.99± 151.10H 4809.14± 160.63G 3643.05±186.94H 5502.32± 176.40FG
For each row different alphabets represent significant differences (α= 0.05) among samples. (*) odor description from Afoakwa et al. (2008), Aprotosoaie et al. (2016), Bonvehi (2005), Rodriguez-Campos et al. (2012)
and (Tran, Van de Walle, et al., 2015, Tran, Van Durme, et al., 2015).
M. Hinneh et al. Food Research International 119 (2019) 84–98
93
Table 4a
Odor activity values of aroma volatiles identified from dark chocolates produced from cocoa beans of different pod storage – roasting conditions.
No. Volatile OTV (ng/g) 0PS-100 °C 7PS-100 °C 0PS-120 °C 3PS-120 °C 0PS-135 °C 3PS-135 °C 7PS-135 °C 0PS-140 °C 0PS-160 °C 7PS-160 °C
Acids
1 Acetic acid 124–750 3.76–22.76 2.79–16.86 3.78–22.86 2.89–17.50 4.81–29.11 2.81–17.00 4.69–28.36 3.87–23.41 5.18–31.34 5.75–34.78
2 Isovaleric acid 22–1000 0.21–9.72 0.17–7.78 0.38–17.31 0.23–10.52 0.33–15.07 0.22–10.17 0.28–12.64 0.38–17.34 0.26–11.87 0.27–12.19
Alcohols
3 Amyl alcohol 470 0.09 0.08 0.10 0.09 0.07 0.07 0.05 0.04 0.04 0.05
4 Hexanol 400 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
5 Benzyl alcohol 5100–900,000 ≤ 0.02 ≤ 0.03 ≤ 0.02 ≤ 0.03 ≤ 0.03 ≤ 0.02 ≤ 0.02 ≤ 0.03 ≤ 0.02 ≤ 0.04
6 2-Phenyl ethyl alcohol 10–590 0.65–38.26 0.17–9.81 0.86–51.02 0.50–29.67 1.50–88.72 0.61–36.17 0.67–39.74 1.40–82.87 1.84–108.36 1.04–61.45
Aldehydes
7 2-Methylbutanal 2.2–152 0.08–5.85 0.09–6.19 0.12–8.44 0.10–6.72 0.13–8.97 0.17–11.41 0.12–8.56 0.16–11.13 0.29–20.35 0.27–18.50
8 3-Methylbutanal 5.4–80 0.59–8.74 0.50–7.36 1.11–16.48 0.76–11.33 1.42–21.09 1.31–19.35 1.20–17.82 1.56–23.18 1.74–25.71 1.52–22.54
9 Pentanal 240 0.54 0.57 0.48 0.44 0.46 0.37 0.47 0.54 0.40 0.49
10 Hexanal 75 2.56 2.17 2.38 2.04 1.85 1.72 2.07 2.38 1.52 1.80
11 Heptanal 64.8 0.38 0.30 0.43 0.34 0.38 0.37 0.42 0.52 0.42 0.47
12 Octanal 320 0.09 0.07 0.12 0.11 0.12 0.12 0.12 0.13 0.12 0.10
13 Nonanal 56 1.05 0.74 1.21 0.93 0.84 1.48 1.18 1.10 2.35 1.13
14 Benzaldehyde 60 1.44 1.36 1.83 1.49 1.87 1.65 1.90 1.93 2.12 2.18
15 Benzeneacetaldehyde 22–154 1.06–7.45 0.26–1.83 1.35–9.44 0.53–3.74 1.72–12.03 0.50–3.50 0.71–5.00 1.49–10.45 1.03–7.23 0.51–3.59
Ketones
16 2-Heptanone 1500 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02
17 2-Octanone 500–510 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
18 3-Octen-2-one 250 0.00 0.00 0.11 0.10 0.26 0.19 0.22 0.28 0.39 0.28
19 2-Nonanone 100 0.15 0.16 0.21 0.22 0.00 0.00 0.00 0.00 0.00 0.00
20 Acetophenone 5629 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.02 0.01
Pyrazines
21 Methylpyrazine 27,000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
22 2,5-Dimethylpyrazine 2600–17,000 0.00 0.00 0.00 0.00 ≤ 0.01 ≤ 0.01 ≤ 0.01 ≤ 0.01 ≤ 0.01 ≤ 0.02
23 2,6-Dimethylpyrazine 1021–8000 0.00 0.00 0.00 0.00 ≤ 0.02 ≤ 0.02 ≤ 0.02 ≤ 0.02 ≤ 0.04 ≤ 0.03
24 2-Ethyl-6-methylpyrazine 320 0.00 0.00 0.02 0.01 0.05 0.03 0.05 0.05 0.12 0.11
25 Trimethylpyrazine 290 0.04 0.05 0.09 0.07 0.22 0.16 0.46 0.22 0.52 1.19
26 Tetramethylpyrazine 38,000 0.00 0.02 0.01 0.00 0.01 0.00 0.03 0.01 0.01 0.04
Esters
27 Methyl acetate 3.4 9.52 7.56 5.84 5.98 7.26 7.52 7.76 8.45 16.08 12.11
28 Ethyl octanoate 1.5–5.5 2.12–7.79 1.76–6.44 3.24–11.88 2.90–10.63 0.00 0.00 0.00 0.00 0.00 0.00
29 Phenyl ethyl acetate 137–233 0.43–0.72 0.36–0.61 0.45–0.76 0.28–0.47 0.53–0.90 0.23–0.39 0.46–0.78 0.47–0.80 0.55–0.94 0.60–1.02
Terpenes and terpenoids
30 Myrcene 9.18 0.64 0.70 1.12 0.88 0.94 0.96 0.93 1.10 1.05 1.15
31 Limonene 250–14,700 ≤ 0.06 ≤ 0.04 ≤ 0.05 ≤ 0.04 ≤ 0.04 ≤ 0.05 ≤ 0.05 ≤ 0.05 ≤ 0.08 ≤ 0.07
32 Linalool 37 2.02 1.87 2.53 2.48 3.06 2.52 3.16 2.70 3.07 3.28
Furans and furanones
33 2-Pentyl furan 100–2000 0.02–0.47 0.03–0.54 0.03–0.66 0.02–0.43 0.02–0.44 0.03–0.56 0.02–0.46 0.03–0.70 0.04–0.70 0.03–0.69
34 γ-Butyrolactone 35 0.00 0.00 0.70 0.56 1.35 1.67 1.56 1.13 2.83 2.68
Others
35 Dimethyldisulfide 12 0.00 0.00 1.10 1.06 1.08 1.26 1.20 1.10 2.04 1.88
36 Pyridine 920–3000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 ≤ 0.01 0.00
37 2-Methoxyphenol 10 0.00 0.00 0.00 0.00 2.41 1.69 2.96 3.17 3.69 5.21
M. Hinneh et al. Food Research International 119 (2019) 84–98
Table 4b
Odor activity values of aroma volatiles identified from dark chocolates produced from commercial cocoa liquors.
No. Volatile OTV (ng/g) Ecuador Ghana Ivory coast Madagascar Venezuela Vietnam
Acids
1 Acetic acid 124–750 4.00–24.20 3.21–19.40 1.97–11.93 3.24–19.63 2.43–14.69 3.30–19.99
2 Isovaleric acid 22–1000 0.05–2.17 0.18–8.06 0.01–0.64 0.06–2.56 0.05–2.42 0.09–3.93
Alcohols
3 Amyl alcohol 470 0.08 0.12 0.09 0.06 0.08 0.04
4 Hexanol 400 0.02 0.00 0.00 0.02 0.02 0.02
5 Benzyl alcohol 5100–900,000 ≤ 0.02 ≤ 0.02 ≤ 0.01 ≤ 0.02 ≤ 0.01 ≤ 0.02
6 2-Phenyl ethyl alcohol 10–590 0.31–18.12 0.32–19.01 0.23–13.72 0.23–13.41 0.20–11.98 0.30–17.44
Aldehydes
7 2-Methylbutanal 2.2–152 0.08–5.81 0.11–7.38 0.13–8.82 0.08–5.78 0.08–5.44 0.11–7.49
8 3-Methylbutanal 5.4–80 0.47–7.00 0.73–10.76 0.55–8.14 0.41–6.14 0.39–5.77 0.70–10.30
9 Pentanal 240 0.78 0.69 0.37 0.68 0.40 0.51
10 Hexanal 75 5.02 3.38 2.28 4.34 2.77 2.84
11 Heptanal 64.8 0.80 0.55 0.54 0.84 0.58 0.54
12 Octanal 320 0.16 0.11 0.10 0.14 0.10 0.09
13 Nonanal 56 1.28 1.19 1.51 1.75 1.38 0.98
14 Benzaldehyde 60 1.19 1.74 1.56 1.26 1.05 1.19
15 Benzeneacetaldehyde 22–154 0.11–0.75 0.24–1.65 0.11–0.79 0.17–1.20 0.11–0.76 0.21–1.44
Ketones
16 2-Heptanone 1500 0.02 0.02 0.02 0.03 0.02 0.02
17 2-Octanone 500–510 0.00 0.00 0.02 0.00 0.00 0.00
18 3-Octen-2-one 250 0.00 0.04 0.00 0.00 0.00 0.00
19 2-Nonanone 100 0.28 0.29 0.21 0.23 0.21 0.22
20 Acetophenone 5629 0.00 0.01 0.00 0.00 0.00 0.00
Pyrazines
21 Methylpyrazine 27,000 0.00 0.00 0.00 0.00 0.00 0.00
22 2,5-Dimethylpyrazine 2600–17,000 ≤ 0.01 ≤ 0.01 0.00 ≤ 0.01 0.00 ≤ 0.01
23 2,6-Dimethylpyrazine 1021–8000 0.00 0.00 ≤ 0.01 0.00 0.00 0.00
24 2-Ethyl-6-methylpyrazine 320 0.01 0.38 0.01 0.01 0.01 0.39
25 Trimethylpyrazine 290 0.16 0.14 0.07 0.12 0.10 0.19
26 Tetramethylpyrazine 38,000 0.00 0.02 0.00 0.00 0.00 0.01
Esters
27 Methyl acetate 3.4 7.59 6.90 8.62 7.54 10.57 7.49
28 Ethyl octanoate 1.5–5.5 2.14–7.84 2.20–8.05 1.63–5.99 1.29–4.74 1.72–6.30 2.70–9.89
29 Phenyl ethyl acetate 137–233 0.61–1.05 0.49–0.84 0.24–0.41 0.34–0.57 0.33–0.56 0.87–1.49
Terpenes and terpenoids
30 Myrcene 9.18 0.51 0.74 0.00 0.49 0.57 0.00
31 Limonene 250–14,700 ≤ 0.05 ≤ 0.05 ≤ 0.06 ≤ 0.19 ≤ 0.05 ≤ 0.04
32 Linalool 37 0.46 1.90 1.04 0.00 0.46 0.69
Furans and furanones
33 2-Pentyl furan 100–2000 0.06–1.13 0.04–0.75 0.03–0.68 0.05–0.99 0.04–0.79 0.03–0.57
34 γ-Butyrolactone 35 0.72 0.73 0.79 0.85 0.70 0.78
Others
35 Dimethyldisulfide 12 1.01 1.08 0.00 0.00 0.00 1.12
36 Pyridine 920–3000 0.00 0.00 0.00 0.00 0.00 0.00
37 2-Methoxyphenol 10 0.00 0.00 0.00 0.00 0.00 0.00
and sweet notes to the final flavor. The chocolates from commercial chocolates with RT of 100 °C. Arguably, due to the low temperature,
liquors recorded lower concentrations than the sample chocolates al- most of these volatiles may have been formed in very minimal con-
though, although among these, the Ghanaian chocolate was sig- centrations which were eventually lost during the chocolate manu-
nificantly highest (p < .05). Myrcene being present in marginal con- facturing process. Rose oxide, which was identified in substantial
centrations was only odor-active in most sample chocolates. amounts in the chocolates, is known for its geranium-like floral note
(Bonvehí, 2005). It was however not odor-active (Table 4a). γ-Butyr-
3.6. Furans, furanones, pyrans, pyrones, pyrroles and others olactone which was most dominant in the chocolates roasted at 160 °C,
was also odor-active in some chocolates and is thus expected to confer
Generally, these groups of volatiles showed a similar pattern sweet and caramel notes towards the final flavor. Also, among the
whereby their total concentrations were largely influenced by the RT. It chocolates from the commercial liquor, the Madagascan, Ecuadorian,
was evident from Table 3a, that these concentrations increased with Vietnamese and Ghanaian chocolates recorded the most abundant (but
increasing temperature. Withal, at low to moderate RT's (100–135 °C), significantly equal) concentrations of this group of volatiles (Table 3b).
no clear trends were seen with respect to PS. However, at 160 °C, 7PS The pyrroles were the least abundant group identified. Apart from
recorded significantly higher (p < .05) total concentrations of pyrroles their absence in the chocolates with lower RT's (< 135 °C), their con-
and the other unclassified volatiles. This may be due to their higher centrations seemed to increase as RT increased. Both 0PS and 7PS ap-
reducing sugar and free amino acid contents which are required for the peared to have comparable concentrations at RT's below 160 °C.
Maillard reaction as reported by Hinneh et al. (2018). Although, isoamyl-2-formyl pyrrole was the most abundant in 7PS-
Apart from 2-pentyl furan – which is a product from fat degradation, 160 °C chocolate, 2-acetyl pyrrole was the most prevalent pyrrole
all other furans, furanones, pyrans and pyrones were absent in identified in all the chocolates. The latter is formed from proline
94
M. Hinneh et al. Food Research International 119 (2019) 84–98
Fig. 1. Total concentrations of various groups of volatiles identified in dark chocolates produced from cocoa beans of different pod storage – roasting conditions and
commercial cocoa liquors.
through the Strecker degradation reaction (Afoakwa, Paterson, Fowler, lower roasting intensities or lost to the extreme industrial process often
& Ryan, 2009; Mottram, 2007) and is characterized by its caramel applied. Finally, it is also possible that on a mass industrial scale, the
flavor notes. Notwithstanding, this volatile was not odor-active mixing together of different cocoa beans of different qualities (in terms
(Table 4a). Pyrroles were absent in the chocolates from commercial of pod storage, fermentation and drying) from the different producers
liquors, except for 2-acetyl pyrrole, which was the highest in the of the beans could have inevitably influenced the aroma quality of these
Vietnamese chocolate. commercial liquors. However, for the purpose of comparison, these li-
Some other (unclassified) volatiles such as 2-methoxyphenol and quors, being representative of the typical industrial quality, were ac-
dimethyldisulfide were also marked for their notable presence in cho- cepted and utilized ‘as-is’.
colates (Aprotosoaie et al., 2016). These two volatiles are generated at
intense heating processes and thus, associated with undesirable flavor 3.8. Comparing chocolates on the basis of their aroma volatile profiles
notes. They were odor-active and more dominant in chocolates with
beans roasted at 160 °C. Clustering analysis can be viewed as a means of grouping a set of
data such that the samples in a particular group (here referred to as
3.7. Trends in the overall aroma concentration of chocolates as influenced cluster) do share certain similarities in comparison to others from a
by pod storage and roasting temperature different group. AHC uses a bottom-up approach, thus, each sample
starts as its own unique cluster, of which pairs of other clusters are
Fig. 1 shows the general trends in the overall aroma volatile con- merged systematically in a hierarchical order. The process begins by
centration in the various chocolates. Overall, the impact of PS and RT computing the dissimilarity between the available number (N) of
on the volatile concentrations were also depicted. Particularly, at lower samples. Then, two samples with a low dissimilarity that minimizes the
temperatures (100–120 °C), chocolates with unstored pods possessed given agglomeration criterion are clustered. Next, the dissimilarity
higher concentrations than those from stored pods. However, at 160 °C, between this cluster and the N-2 other samples is also computed ac-
the opposite was observed. This is probably because of the high rate of cording to the same agglomeration criterion. These two samples whose
Maillard reaction occurring amidst the excess of flavor precursors (re- clustering again minimizes the agglomeration criterion are then clus-
ducing sugars and free amino acids) in 7PS (Hinneh et al., 2018). tered together. Consequently, the process continues until all samples
Comparatively, a temperature of 135 °C seemed more optimal for have been clustered (XLSTAT, 2019). The AHC was therefore useful for
maximal aroma volatile release than at 140 °C. Aside the trend from the exploring the possible existing (dis)similarities among the chocolates in
overall volatile concentrations, similar observations were made from terms of their aroma profiles.
the total acid and alcohol concentrations. At this temperature (135 °C), Fig. 2a and 2b depict the similarities between the chocolates, first,
both 0PS and 7PS contained significantly similar (p > .05) overall taking into consideration all the identified volatiles and next, on the
volatile concentrations whereas the concentration of 3PS was sig- basis of their odor-active volatiles respectively. From Fig. 2a, with a
nificantly lower (p < .05). Generally, except at lower RT's dissimilarity of 400, three main clusters were obtained. The first group
(100–120 °C), the concentrations of aroma volatiles from the sample comprised of chocolates of various PS with moderate to high RT's
chocolates were seen to be higher than those from the commercial (135–160 °C). Remarkably, these were chocolates with generally higher
cocoa liquors. Whereas this could be due to the type (whole bean overall aroma volatile concentration, owing to the intensity of roasting
roasting for treated samples as opposed to nib roasting for the com- (Table 3a). These were further distinguished between chocolates of RT's
mercial liquors) and intensities of roasting applied in the latter, dif- 135–140 °C on one hand and 160 °C on the other. From these, the dif-
ference in the industrial treatments applied from bean to liquor could ferent groupings arising from the different PS treatments were also
also be implicated to have had a detrimental impact on the levels of the seen. The second major group consisted of chocolates from Madagascar.
various volatiles. Consequently, desirable volatiles such as esters, al- Suggestively, this separation was as a result of its relatively higher
dehydes, ketones, terpenes and pyrazines which impart the overall concentrations of aldehydes, furans, furanones, pyrans and pyrroles as
aroma expression were either produced at lower concentrations due to well as the lower pyrazine concentration among the chocolates from
95
M. Hinneh et al. Food Research International 119 (2019) 84–98
Fig. 2. a. Agglomerative hierarchical clustering of all identified aroma volatiles in various dark chocolates produced from cocoa beans of different pod storage –
roasting conditions and commercial cocoa liquors. 2b: Agglomerative hierarchical clustering on the basis of odor activity values of various dark chocolates produced
from cocoa beans of different pod storage – roasting conditions and commercial cocoa liquors.
96
M. Hinneh et al. Food Research International 119 (2019) 84–98
commercial liquors. More so, from Table 2, this was the only chocolate acids (sour), esters (fruity/floral/sweet), furans (roasted/cheesy/green)
which was produced with liquor comprising of all three major varieties and pyrroles (caramel). The latter group of chocolates could also be
of cocoa (Forastero, Criollo and Trinitario). However, the final group marked by the significantly higher (p < .05) levels of undesirable
encompassed all remaining chocolates from commercial liquors in- volatiles such as dimethyldisulfide (sulfurous) and 2-methoxyphenol
cluding sample chocolates with minimal roasting intensity (100 °C). (smoky). Conversely, all chocolates with low (100–120 °C) RT's in-
The chocolates from this cluster where generally known for their lower cluding the Ghanaian origin chocolate also showed high similarity. Of
overall volatile concentrations (Table 3a and 3b). Hitherto, a keen si- these, chocolates from the unstored pods, were significantly higher
milarity is observed between the Ghanaian chocolate and those roasted (p < .05) in alcohols (fruity/floral), acids (sour) and aldehydes
at 100 °C and 120 °C. Meanwhile, the Ivorian chocolates, showed more (cocoa/chocolate), whereas, 7PS chocolates were significantly
similarity with the Ecuadorian, Venezuelan and Vietnamese chocolates. (p < .05) rich in pyrazines. Suggestively, with similar levels of ter-
Given the clear differences in the variety of cocoa beans used to man- penes/terpenoids (fruity/floral/spicy) in both unstored and pod stored
ufacture these chocolates (Table 2), it is possible that the different chocolates, the latter is likely to express more fruity/floral notes.
roasting protocols (time – temperature) applied in each scenario may Generally, at the different RT's applied, 3PS chocolates mostly recorded
have play a more important role in fashioning their individual profiles, the lowest volatile concentrations. It is possible that the suppression of
although unfortunately, this information was not provided by the some dominant aroma volatiles (such as acids, pyrazines, aldehydes and
supplier. Yet, for the sole purpose of comparison within the context of ketones) due to the limited PS treatment could have led to a more ba-
this study, these commercial cocoa liquors were envisioned as re- lanced aroma profile, with possible expressions of more subtle volatiles
presentative of what was available on the market irrespective of the including esters and terpenes/terpenoids. This study, thus, reveals the
specific pre-processing technique applied by the producer. It is, thus possibility of harnessing both PS and RT in order to steer the aroma
clear, that regardless of the cocoa bean variety, it is possible to steer the profiles of dark chocolates produced from ‘bulk’ cocoa beans.
aroma profiles of chocolates through PS and roasting. However, from
these clustering, it seemed that the impact of the latter may have been Acknowledgements
greater.
From Fig. 2b, the dissimilarities between the chocolates on the basis The authors would like to express their profound gratitude to the
of their odor-active volatiles were rather low, given that with a dis- Belgian Government through VLIR-UOS for funding this project (ICP
similarity of 100, only three clusters were obtained. First, a cluster PhD 2014–001). A big thanks to the technical staff of Puratos –
comprising of chocolates with a high roasting temperature of 160 °C. Belcolade, Belgium for the assistance with roasting and winnowing. We
The second cluster comprised of chocolates within the range of are also grateful to Cacaolab bvba, Belgium for the chocolate produc-
135–140 °C. Of these, chocolates; 0PS-135 °C and 0PS-140 °C showed a tion.
great similarity, followed by chocolates; 7PS-135 °C and 3PS-135 °C.
Particularly within this cluster, the distinguishing effect of PS seemed Conflict of interest
more pronounced than that of the RT. From the third cluster, two main
groupings were observed. On the one hand, the Ghanaian chocolate The authors declare that they have no conflict of interest.
together with sample chocolates with low roasting intensities
(100–120 °C), whereas, on the other hand, all the other chocolates from References
commercial cocoa liquors were grouped. Here, the Ivorian, the Mada-
gascan and the Venezuelan chocolates showed some similarities. Aculey, P. C., Snitkjaer, P., Owusu, M., Bassompiere, M., Takrama, J. S., Nørgaard, L., &
However, these were slightly different from both the Vietnamese and Nielsen, D. S. (2010). Ghanaian cocoa bean fermentation characterized by spectro-
Ecuadorian chocolates. The observed similarities may be ascribed to the scopic and chromatographic methods and chemometrics. Journal of Food Science,75(6), 300–307.
varieties of beans utilized for the production of the various liquors. Afoakwa, E. O. (2010). Chocolate science and technology. UK: Wiley-Blackwell
However, the closeness of the Venezuelan and Madagascan chocolates Publishers3–82.
to the Ivorian chocolate in spite of their different cocoa varieties, may Afoakwa, E. O. (2014). Cocoa Production and Processing Technology (1st ed.). Boca Raton,Fl: CRC Press.
be due to the additional difference in the roasting process/protocol Afoakwa, E. O. (2016). Chocolate Science and Technology (2nd ed.). UK: John Wiley &
applied. Sons, Ltd.
Afoakwa, E. O., Paterson, A., Fowler, M., & Ryan, A. (2008). Flavor formation and
character in cocoa and chocolate: a critical review. Critical Reviews in Food Science and
4. Conclusions Nutrition, 48(9), 840–857.
Afoakwa, E. O., Paterson, A., Fowler, M., & Ryan, A. (2009). Matrix effects on flavor
A decreasing effect of PS on the overall aroma concentration was volatiles release in dark chocolates varying in particle size distribution and fat con-
tent using GC–mass spectrometry and GC–olfactometry. Food Chemistry, 113,
observed. This has often been ascribed to the partial modification of the 208–215.
pulp leading to reduction in pulp volume during the storage process. Afoakwa, E. O., Quao, J., Budu, A. S., Takrama, J., & Saalia, F. K. (2011a). Effect of pulp
However, for long durations of PS (~ 7 days), an increase in the overall pre-conditioning on acidification, proteolysis, sugars and free fatty acids concentra-
tion during fermentation of cocoa (Theobroma cacao) beans. International Journal of
volatile concentration could be attributed to the on-set of an extensive Food Science and Nutrition, 62(7), 755–764.
degradation of the pod and pulp material, thus, making available more Afoakwa, E. O., Quao, J., Budu, A. S., Takrama, J., & Saalia, F. K. (2011b). Chemical
aroma precursors. RT also contributed significantly (p < .05) to an composition and physical aquality characteristics of Ghanaian cocoa beans as af-
increase in the overall volatile concentration. This increase is either as a fected by pulp pre-conditioning and fermentation. Journal of Food Science andTechnology, 47(1), 67–75.
result of the unlocking effect of the heating process on the volatiles Aprotosoaie, A. C., Luca, S. V., & Miron, A. (2016). Flavor Chemistry of Cocoa and Cocoa
trapped within the matrix or the promotion of new volatile formation Products-an Overview. Comprehensive Reviews in Food Science and Food Safety, 15,
through the Maillard reaction during roasting. From an AHC, choco- 73–91.Beckett, S. T. (2009). Industrial chocolate manufacture and use (4th ed.). UK: Wiley-
lates with a moderate to high (135–160 °C) RT's showed high similarity Blackwell, York.
among each other, such that, these chocolates possessed more volatiles Bonvehí, J. S. (2005). Investigation of aromatic compounds in roasted cocoa powder.
with relatively higher concentrations. However, the diversity in the European Food Research and Technology, 221(1–2), 19–29.Counet, C., Callemien, D., Ouwerx, C., & Collin, S. (2002). Use of gas chromatography
aroma profiles among these chocolates was evident as chocolates from olfactometry to identify key odorant compounds in dark chocolate. Comparison of
unstored pods were significantly dominant (p < .05) in alcohols samples before and after conching. Journal of Agricultural and Food Chemistry, 50(8),
(fruity/floral), aldehydes (cocoa/chocolate) and ketones (fruity/ 2385–2391.
Duncan, R. J. E. (1984). A survey of Ghanaian cocoa farmers' fermentation and drying
creamy). Whereas those from prolonged PS (~ 7 days) were sig- practices and their implication for Malaysian practices. Proceedings of International
nificantly dominant (p < .05) in pyrazines (cocoa/nutty/roasted), Conference on Cocoa and Coconut, October 15–17, Malaysia.
97
M. Hinneh et al. Food Research International 119 (2019) 84–98
Frauendorfer, F., & Schieberle, P. (2008). Changes in key aroma compounds of Criollo Ramli, N., Hassan, O., Said, M., Samsudin, W., & Idris, N. (2006). Influence of roasting
cocoa beans during roasting. Journal of Agricultural and Food Chemistry, 56, conditions on volatile flavor of roasted Malaysian cocoa beans. Journal of Food
10244–10251. Processing and Preservation, 30(3), 280–298.
Hinneh, M., Semanhyia, E., Van de Walle, D., De Winne, A., Tzompa-Sosa, D. A., Scalone, Rodriguez-Campos, J., Escalona-Buendia, H. B., Contreras-Ramos, S. M., Orozco-Avila, I.,
G. L. L., ... Dewettinck, K. (2018). Assessing the influence of pod storage on sugar and Jaramillo-Flores, E., & Lugo-Cervantes, E. (2012). Effect of fermentation time and
free amino acid profiles and the implications on some Maillard reaction related flavor drying temperature on volatile compounds in cocoa. Food Chemistry, 132(1),
volatiles in Forastero cocoa beans. Food Research International, 111, 607–620. 277–288.
Jinap, S., Siti, M. H., & Norsiati, M. G. (1994). Formation of methyl pyrazine during cocoa Saltini, R., Akkerman, R., & Frosch, S. (2013). Optimizing chocolate production through
bean fermentation. Pertanika Journal of Tropical Agricultural Science, 17, 27–32. traceability: A review of the influence of farming practices on cocoa bean quality.
Jinap, S., Wan-Rosli, W. I., Russly, A. R., & Nordin, L. M. (1998). Effect of roasting time Food Control, 29, 167–187.
and temperature on volatile component profiles during nib roasting of cocoa beans Saputro, A. D., Van de Walle, D., Aidoo, R. P., Mensah, M. A., Delbaere, C., De Clercq, N.,
(Theobroma cacao). Journal of the Science of Food Agriculture, 77, 441–448. & Dewettinck, K. (2017). Quality attributes of dark chocolates formulated with palm
Kadow, D., Bohlmann, J., Phillips, W., & Lieberei, R. (2013). Identification of main fine or sap-based sugar as nutritious and natural alternative sweetener. European Food
flavour components in two genotypes of the cocoa tree (Theobroma cacao L.). Journal Research and Technology, 243(2), 177–191.
of Applied Botany and Food Quality, 86, 90–98. Schwan, R. F., & Wheals, A. E. (2004). The microbiology of cocoa fermentation and its
Kongor, J. E., De Steur, H., Van de Walle, D., Gellynck, X., Afoakwa, E. O., Boeckx, P., & role in chocolate quality. Critical Reviews in Food Science and Nutrition, 44, 205–221.
Dewettinck, K. (2017). Constraints for future cocoa production in Ghana. Springer; Sulaiman, B. K., Yang, A. T., & Ariffin, F. (2017). Color and antioxidant properties of
Agroforestry Systems. cocoa beans from pods storage and fermentation using shallow box. MATTER:
Kongor, J. E., Hinneh, M., Van de Walle, D., Afoakwa, E. O., Boeckx, P., & Dewettinck, K. International Journal of Science and Technology, 3(2), 295–307.
(2016). Factors influencing quality variation in cocoa (Theobroma cacao) bean flavour Tran, P. D., Van de Walle, D., De Clercq, N., De Winne, A., Kadow, D., Lieberei, R., ... Van
profile — A review. Food Research International, 82, 44–52. Durme, J. (2015). Assessing cocoa aroma quality by multiple analytical approaches.
Meyer, B., Biehl, B., Said, M., & Samarakoddy, R. J. (1989). Postharvest pod storage: A Food Research International, 77, 657–669.
method for pulp preconditioning to impair strong nib acidification during cocoa Tran, P. D., Van Durme, J., Van de Walle, D., De Winne, A., Delbaere, C., De Clercq, N., ...
fermentation in Malaysia. Journal of Science of Food and Agriculture, 48, 285–304. Dewettinck, K. (2015). Quality attributes of dark chocolate produced from
Mottram, D. S. (2007). The Maillard reaction: Source of flavor in thermally processed Vietnamese cocoa liquors. Journal of Food Quality, 39(4), 311–322.
foods. In R. G. Berger (Ed.). Flavors and fragrances: Chemistry, bioprocessing and sus- Van Gemert, L. J. (2011). ODOUR THRESHOLDS: Compilations of odour threshold values in
tainability (pp. 269–284). Hannover, Germany: Springer-Verlag. air, water and other media (2nd ed.). BV, Utrecht: Oliemans, Punter & Partners.
Nazaruddin, R., Seng, L., Hassan, O., & Said, M. (2006). Effect of pulp preconditioning on XLSTAT. Agglomerative Hierarchical Clustering. (2019). https://www.xlstat.com/en/
the content of polyphenols in cocoa beans (Theobroma cacao) during fermentation. solutions/features/agglomerative-hierarchical-clustering-ahc Accessed 22 January
Industrial Crops and Products, 24, 87–94. 2019.
98