Caballero B. (ed.) Encyclopaedia of Food Science, Food Technology and Nutrition. Ten-Volume Set

Подождите немного. Документ загружается.

such as abundant aldehydes and alcohols, as well as

of aroma-damaging substances like volatile phenols,

are also removed. As a result the typical and slightly

volatile aroma becomes more pronounced.

0057 Sorption phenomena of aroma materials at particle

surfaces of the heterogeneous chocolate mass of fat

and fat-free portions have considerable influence on

the flavor.

0058 Thin-layer treatment of cocoa mass During modern

chocolate processing the cocoa liquor is degassed

before being put into the chocolate recipe. By using

special equipment, degassing can be performed more

effectively than in the conche. Thin-layer vaporizers

are generally used, which may shorten the remaining

conching time considerably. The process conditions

have to be coordinated with the quality of the cocoa

liquor, so that the valuable aroma is not removed

together with the more volatile unknown materials,

thereby flattening the taste of the product. A reduc-

tion of 10–30% of acids and low-boiling compounds

seems to be optimal.

0059 Alkalization of cocoa Alkalization (‘Dutching’)is

mainly used to intensify the color, to modify the flavor,

and to improve the dispersability of cocoa powder in

water or milk. In this process, beans, nibs, or mass are

treated with solutions or suspensions of alkali, usu-

ally potassium or sodium carbonate, sometimes hy-

droxides or ammonia. The permitted maximum is

generally 2.5–3 parts of potassium carbonate (or

equivalent alkali) per 100 parts of cocoa. The pH

of a natural cocoa is about 5.5, whereas the pH of

commercially available Dutch process cocoas will fall

between 6.8 and 8.0.

0060 Color and flavor changes during alkalization are

mainly based on modifications of polyphenolic sub-

stances. Dutch processing results in a milder and

altered flavor, and a darker color. Under alkaline

conditions, soluble monomeric flavonoids (mainly

epicatechin) are oxidized and condensed to give poly-

mer tannins. Red and brown pigments are developed

and the astringency is reduced. Volatile acids are

partially neutralized.

Cocoa Aroma Analysis

0061 The cocoa aroma volatiles only occur in milli- or

microgram per kilogram quantities. They are dis-

solved in the fat phase or physically bound to a

heterogeneous matrix of constituents of differing

polarity. Isolation and enrichment techniques are

necessary to obtain the flavor volatiles in sufficiently

high concentrations. Distillation methods, head-

space techniques, and extractions, or combinations

of these techniques are used. The separation of the

volatiles is performed by means of capillary gas chro-

matography (GC) or high-performance liquid

chromatography (HPLC). (See Chromatography:

High-performance Liquid Chromatography; Gas

Chromatography.)

0062Isolation and Analytical Techniques 0062

Extraction Extraction techniques have several dis-

advantages (time-consuming and the possibility of

introducing artifacts) but have the advantage of quan-

titatively isolating volatiles independent of their

boiling points. Usually, cocoa is defatted with petro-

leum ether and the residue is extracted with a polar

solvent such as water, ethanol, acetonitrile, ethyl

ether, or subcritical carbon dioxide. By means of

extraction, specific high-boiling compounds have

been isolated from cocoa: diketopiperazines, pyrones,

amines, amides, aromatic acids and phenols (these

compounds are mainly for subsequent GC separ-

ation), and theobromine, caffeine, and monomeric

flavonoids (for HPLC measurement).

0063Distillation Most of the known aroma volatiles

have been isolated from cocoa by means of distillation

(steam distillation, vacuum condensation, simultan-

eous steam distillation–extraction (SDE), head-space

enrichment). Nonvolatile impurities can thus be

avoided and a defatting step is not necessary, but

high-boiling compounds are partially lost and one

has to be aware of possible reactions during the heat

treatment.

0064The aqueous distillates are usually extracted with

small portions of ethyl ether for performing GC or

may be injected directly into an HPLC system.

0065Table 6 shows typical volatiles of roasted cocoa.

The flavor chromatograms of different cocoas are

rather similar, but the quantitative distribution of

the single constituents and their total amounts vary

distinctly with origin, fermentation, drying, roasting,

degassing, alkalization, or conching. Distillates of

chocolate contain similar compounds in lower con-

centration and altered distribution.

0066Analysis of indicative compounds and key odorants

Objective analytical data are necessary to relieve the

industrial quality control of time-consuming sensory

tests. For that purpose suitable indicators are needed

to evaluate the flavor. Several examples of indicative

compounds in cocoa aroma follow, together with

methods suitable for their estimation (GC; HPLC;

ultraviolet (UV) detector; electrochemical detector

(ELCD)): Fermentation degree – epicatechin

(HPLC–ELCD), tetramethylpyrazine (GC or HPLC–

UV); flavor-grade cocoas linalool (GC); roasting in-

tensity ratios of 2,5-dimethyl- to tetramethylpyrazine

1446 COCOA/Chemistry of Processing

or trimethyl- to tetramethylpyrazine (GC or HPLC–

UV), ratio of 5-methyl-2-phenyl-hex-2-enal to 2-phe-

nylethanol (GC), dihydrohydroxymaltol (GC or

HPLC–ELCD–UV); intensity of degassing of cocoa

masses – 3-methylbutanal (head-space GC).

0067Aroma extract dilution analyses (AEDA) is a potent

tool to screen the most odor-active compounds in

foods. The application of AEDA on the cocoa volatiles

revealed about 40 main odorants in cocoa mass.

Sensory Evaluation

0068Sensory evaluation is extremely important for cocoa

process control and for the quality assurance of cocoa

and chocolate products. Based on a tasting panel,

which consists of trained and selected members, dif-

ferent sensory tests (flavor profile, difference, and

scoring tests) are used and the results are statistically

evaluated. (See Sensory Evaluation: Taste.)

0069The flavor profile test This test is essential to obtain

an assessment of the development of taste and flavor

within the mouth as a result of individual sensory

perceptions. The varying intensities of the specific

characteristics are observed and evaluated. An evalu-

ation plan is used which considers the flavor and

texture characteristics and is completed by a numeric

scale. Some characteristics of cocoa mass are raw,

acid, flavorful, typical cocoa bitterness, burnt/bitter,

astringent, off-taste (cooked, moldy, smoky, ham,

chemical, etc.). In addition, descriptors for plain

chocolate are sweetness, aromatic, cocoa intensity,

harmonic, persistent, caramel, vanilla, off-taste (card-

board, metallic, rancid, etc.). Intensities range from 0

to 4: absent, barely perceptible, weak, medium,

strong. Texture descriptors are sticky/neat, coarse/

smooth, dry/mellow, snappy/soft. (See Sensory Evalu-

ation: Descriptive Analysis.)

0070Difference tests These are preference tests to com-

pare a sample against a control: in a simple form it is a

paired test or, with one sample duplicated, a triangu-

lar test. Preferences and the intensity of deviations are

reported. Triangular tests are used to check samples

which have slight deviations from the standard.

0071Scoring test This test is used for production quality

control and for assessing chocolate samples after

periods of storage under different conditions. Sensory

evaluation: 9–7, optimal to good; 6–4, satisfactory to

sufficient; 3–1, lacking to unsatisfactory. (See Sensory

Evaluation: Sensory Rating and Scoring Methods.)

0072Tasting method Solid cocoa masses are chopped

and tasted with a spoon. Cocoa powders are

tbl0006 Table 6 Selected aroma compounds identified in extracts of

roasted cocoas

Compound Concentration

range

3-Methylbutanal IV

2,3-Butandione III

3-Methyl-butan-1-oe IV

Acetone II

2-Methylpyrazine III

2,5-Dimethylpyrazine III

2,6-Dimethylpyrazine II

2-Ethylpyrazine I

2,3-Dimethylpyrazine II

Methyltrithiomethane I

2-Ethyl-5-methylpyrazine II

2-Ethyl-3-methylpyrazine II

Acetic acid V

Trimethylpyrazine III

Linalool oxide (cis-furanoid) I

2,5-Dimethyl-3-ethylpyrazine II

2-Furfural II

Linalool oxide (trans-furanoid) I

Tetramethylpyrazine IV

Pyrrole I

Benzyaldehyde III

Linalool II

2,5-Diethyl-3-methylpyrazine I

5-Methyl-2-furfural I

3-Methyl butanoic acid V

Phenylacetaldehyde IV

Acetophenone III

Furfuryl alcohol I

1-Phenylethanol I

2-Phenethyl acetate II

2-Hydroxy-3-methyl-2-cyclo penten-1-one

(cyclotene)

Guajacol I

Benzylalcohol II

2-Phenylethanol III

2-Phenylbut-2-enal III

2-Phenethylamine II

2-Acetylpyrrole III

4-Methyl-2-phenylpent-2-enal II

Phenol I

3-Hydroxy-2-methyl-4-pyron (maltol) II

4-Hydroxy-2,5-dimethyl-3-(2H)-furanone (furanol) III

5-Methyl-2-phenylhex-2-enal III

2-Pyrrolecarbaldehyde II

5-Methyl-2-pyrrolecarbaldehyde II

2,3-Dihydro-3,5-dihydroxy-6-methyl-4-pyrone

(dihydrohydroxymaltol)

3,5-Dihydroxy-6-methyl-4-pyrone (hydroxymaltol) II

5-(2-Hydroxyethyl)-4 -methylthiazole III

1,2-Benzenediol III

Benzoic acid II

2-Phenylacetic acid IV

2-Phenylacetamide IV

Usual concentration range (mg kg

1

): I, <01; II, 01–05, III, 05–2; IV, 2–10;

V, >10. Succession of gas chromatographic retention times on a DB-wax

capillary column.

COCOA/Chemistry of Processing 1447

suspended to 10% in warm water, plain chocolates

are broken into pieces and tasted.

See also: Amino Acids: Properties and Occurrence;

Browning: Enzymatic – Biochemical Aspects; Caffeine;

Carbohydrates: Classification and Properties;

Chromatography: Gas Chromatography; High-

performance Liquid Chromatography; Cocoa:

Production, Products, and Use; Colorants (Colourants):

Properties and Determination of Natural Pigments;

Enzymes: Functions and Characteristics; Fatty Acids:

Properties; Lactic Acid Bacteria; Fats: Classification;

Sensory Evaluation: Sensory Rating and Scoring

Methods; Descriptive Analysis; Taste

Further Reading

Berbert PRF (1979) Contribuic¸a

o para o conhecimento dos

ac¸ucares componentes da ame

ndoa e do mel do cacau.

Revista Theobroma (Brasil) 9: 55–61.

Biehl B, Heinrichs H, Ziegeler A, Berghausen H, Hammoor

M and Senyuk V (1993) The proteases of ungerminated

cocoa seeds and their role in the fermentation process.

Angewandte Botanik (Journal of Applied Botany) 67:

59–65.

Fincke A (1989) Analytische und technologische Entwick-

lungen auf dem Gebiet der Kakaoerzeugnissc. Lehens-

mittelchemie und gerichtliche Chemie 43: 49–55.

Forsyth WGC and Quesnel VC (1963) The mechanism of

cocoa curing. Advances in Enzymology 25: 457–495.

Hardy F and Rodriguez G (1953) Quantitative variations in

nitrogenous components of the cacao bean: effect of

genetic type and soil type. Annual Report on Cacao

Research, Imperial College of Tropical Agriculture

(Trinidad), 1945–51: 89–91.

Lehrian DW and Patterson GR (1983) Cocoa fermentation.

In: Rehm HJ and Reed G (eds) Food and Feed Produc-

tion with Microorganisms. Biotechnology, vol. 5, pp.

529–575. Weinheim: Verlag Chemie.

Martin RA Jr (1987) Chocolate. Advances in Food Re-

search 31: 211–342.

Oberparleiter S and Ziegleder G (1997) Amadori-Verbin-

dungen als Aromavorstufen in Kakao. Nahrung 41:

142–145.

Rohan TA (1963) Processing of Raw Cocoa for the Market.

FAO Agricultural Studies, no. 60. Geneva: Food and

Agriculture Organization.

Schnermann P and Schieberle P (1997) Evaluation of key

odorants in milk chocolate and cocoa mass by aroma

extract dilution analysis. Journal of Agriculture and

Food Chemistry 45: 867–872.

Schwan RF (1998) Cocoa fermentations conducted with a

defined microbial cocktail inoculum. Applied and Envir-

onmental Microbiology 64: 1477–1483.

Spencer ME and Hodge R (1992) Cloning and sequencing

of a cDNA encoding the major storage proteins of Theo-

broma cacao: identification of the proteins as members

of the vicilin class of storage proteins. Planta 186: 567–

576.

Voigt J and Biehl B (1995) Precursors of the cocoa-specific

aroma components are derived from the vicilin-class (7S)

globulin of the cocoa seeds by proteolytic processing.

Botanica Acta 108: 283–289.

Ziegleder G (1991) Composition of flavor extracts of raw

and roasted cocoas. Zeitschrift fu

r Lehensmittel Unter-

suchung und Forschung 192: 521–525.

Ziegleder G and Biehl B (1988) Analysis of cocoa flavour

components and flavour precursors. In: Linskens HF,

Jackson JF (eds) Analysis of Nonalcoholic Beverages.

Modern Methods of Plant Analysis, new series, vol. 8,

pp. 321–393. Berlin: Springer-Verlag.

Production, Products, and Use

B Biehl, Technische Universita

t Braunschweig,

Braunschweig, Germany

G Ziegleder, Fraunhofer Institut fu

r Verfahrenstechnik

und Verpackung, Freising, Germany

Introduction

0001This article describes raw cocoa processing and the

nature, the technologies, and the use of cocoa powder

and chocolate. These are made from defatted and

nondefatted raw cocoa beans, respectively, which

come from the fermented seeds of the tropical tree

Theobroma cacao L. ‘Cocoa’ here is used as the over-

all term covering the finished products (cocoa

powders and chocolate), the semifinished product

(the raw cocoa beans), and the original seeds in the

pods on the tree.

0002Although cocoa products contain valuable nutri-

ents, their special nature lies in the unique flavor

which is not obtained from other plants. The unpro-

cessed seeds, however, do not have this desired flavor,

but instead an astringent unpleasant taste. Hence the

constituents expressed by the unique genetic potential

of the cocoa seeds are profoundly changed by pro-

cessing and production techniques, and these are

responsible for realizing or destroying the flavor po-

tential, as described in the previous article.

0003‘Processing’ covers all steps from harvesting the

pods up to the stage where the raw cocoa beans

come to the industrial manufacturer’s plant for ‘pro-

duction’ of the final goods.

Raw Cocoa Processing

General Outline

0004Processing usually includes the harvesting and break-

ing of the pods, extraction of the wet beans, transport

1448 COCOA/Production, Products, and Use

(of pods or wet beans), fermentation in covered heaps

or containers for some days, drying of the fermented

beans, sorting, grading, sacking, and storage. ‘Beans’

is a traditional term used commercially for seeds of

cocoa at any stage of their processing. Wet beans are

the seeds covered by the mucilaginous, sweet and

acidic pulp or its remainder during fermentation.

Raw cocoa beans are fermented and dried. The term

‘fermentation’ is used to describe the overall process

of heaping wet cocoa beans for several days, micro-

bial pulp degradation, and also endogenous postmor-

tem reactions, both enzymatic and nonenzymatic,

that take place in the seeds and which are induced

by pulp fermentation.

0005 At the cocoa plantations, fermentation and drying

may primarily be looked upon as a curing process in

order to stabilize the fresh beans by microbial degrad-

ation of the firmly adhering, perishable pulp, and by

drying.

0006 From this point of view the resulting qualities,

especially the flavor potential, may be subordinate

aspects. However, the quality is essential for subse-

quent manufacture of finished products. Raw cocoa

quality may vary considerably depending on the

manner of processing. First grades command an

extra premium. However, since grading takes into

account defects of raw cocoa, but not flavor inten-

sities or flavor notes, there is no way of providing an

extra premium for raw cocoa with superior flavor

potential. Certain raw cocoa origins are given prefer-

ence (and a higher price) by the industry and the

market in virtue of experienced quality differences

compared to less desirable origins.

Origin of Flavor Differences

0007 Some experts assume flavor differences to be due to

edaphic, climatic, or plant genetic particularities in

the cocoa-growing countries. However, these differ-

ences may partially or even predominantly be due to

traditional processing practices which are not uni-

form in the various producing countries.

0008 Soil nutrients and climate generally affect the

physiology of plants. Climatic changes have influence

on pod growth, seed size, the amount of seed con-

stituents, and thus on the starting material used for

cocoa processing. Apart from the variable response of

cocoa cultivars to climate, there is inadequate know-

ledge about the significance of genetic diversity of the

planted cocoa trees and harvested seeds with respect

to its influence on flavor quality. Criollo (fine) and

Forastero (bulk) raw cocoas reveal flavor differences

which are attributed to genetic influences. A huge

number of different genotypes and hybrids, derived

from one or both of these varieties and their forms, are

planted. During the last decade genotype-depending

flavors in raw cocoa beans have been reported which

were harvested from trees of different vegetative

clones and which were processed under empirical

standard conditions. Unequivocal conclusions are

complicated since flavor determinant constituents in

raw cocoa (like acids, catechins, and peptides) under-

went profound transformations during processing.

Physiological effects like variation of the seed’s indi-

vidual pulp layers also interfere. In the case of clonal

effects it is likely that flavor differences do not depend

on the genetic expression of the cotyledons (nibs), but

rather on the genotype of the vegetative tree control-

ling pod growth and the property of pulp and testa.

All vegetative parts of a cloned hybrid represent the

same genotype but the generative progeny, the

embryos, does not.

Practice of Fermentation

0009Handling and care during processing significantly

affect raw cocoa quality and flavor. To improve raw

cocoa quality, repeated attempts have been made to

modify the methods. Few met general application.

The different methods of cocoa fermentation are

mostly developed from local traditions under particu-

lar conditions in the various cocoa-growing coun-

tries. They will not be described in this article. The

reader is referred to earlier reviews cited below for

detailed and valuable information. Instead, repre-

sentative processing methods as practiced in Brazil,

Ghana, and Malaysia, variable courses of fermenta-

tion, and effective modifications will be described.

Most of the common practices correspond to the

description given in the following sections. Heap fer-

mentation and box fermentation as the prevailing

methods will be described briefly.

0010In both cases, ripe pods are harvested and the husks

are opened by means of a cutlass or wooden billet.

The wet seeds are removed from the husk and pla-

centa and are collected for transport. Abnormal,

black, infested or clustered beans and foreign matter

are removed. The wet cocoa is piled up in heaps or

boxes of variable capacity, allowing drainage of

sweatings through the bottom. Care is taken that

mixing can be carried out by turning the fermenting

cocoa upside down and inside out. The wet cocoa is

covered with banana leaves or jute bags and is left by

itself for 5–7 days, interrupted by one or more turn-

ings after 24, 48, or 72 h. After a given time, or

following an assessment of its appearance and odor,

the fermented cocoa is transferred to a drier.

0011Standardized methods, such as the heap and box

fermentations, are used at well-established farms or

estates but they are not in common use elsewhere.

Any kind of box or other receptacle such as troughs,

frames, baskets, boats, or bags can be found to serve

COCOA/Production, Products, and Use 1449

for curing widely varying quantities of wet cocoa.

Alternatively, the wet cocoa is left on the ground or

on a drying platform where a partial fermentation

may occur, depending on the depth of the layer, occa-

sional heaping, and weather conditions. These irregu-

lar types of curing probably play a subordinate role in

most of the raw cocoa-producing countries, but they

contribute to the nonuniform low-quality cocoas as

they appear on the market. However, some special

methods are used to produce high-quality cocoas, as

is reported from Ecuador, where in some estates the

wet beans are spread at day but heaped overnight.

This reminds us that cocoa fermentation is still basic-

ally an empirical procedure.

0012 Heap fermentation (Figure 1) The typical succes-

sion of microbial pulp degradation occurs whenever

wet cocoa is piled up for several days. Heap fermen-

tation is used in West Africa. The heap is prepared

outdoors on banana leaves layered on wooden sticks

to allow better drainage. The cocoa is covered by the

same leaves by folding up the peripheral ends in such

a way that rain is kept off. The amount of wet cocoa

(80–1000 kg) depends on the harvest. Due to the flat,

conical shape of the heap compared to boxes, the

bottom layer and the surface layer are large, which

promotes not only aeration but also cooling. Thus, at

least one turning after 48 h is necessary. The highest

temperature may normally be found 5 cm beneath the

surface.

0013Box fermentation (Figure 2) Boxes are placed under

roofs and are made from locally indigenous timbers.

They are jacked up and their bottom (sometimes also

the walls) are perforated to allow drainage. On large

plantations the fermentary would consist of batteries

of several open boxes (1.0 1.0 1.0 m) which facili-

tate turning from one box to the next. Alternatively,

long boxes (e.g., 10 m length) can be divided by

partitions. Depending on the harvest, the cocoa is

filled up to 100 cm in depth.

0014Compared to average-sized heaps (about 50–80 cm

top height), the deep box (100 cm) depth requires a

longer time for fermentation because of reduced aer-

ation (almost exclusively from the top).

0015In Malaysia, shallow boxes have been introduced

to overcome the lack of uniformity (see below and

Figure 2). Cascades of shallow boxes are used in large

estates, each of which may have a capacity of several

tonnes, but a height not exceeding 35 cm.

0111Revolving barrels which are provided with perfor-

ations for aeration and with some device for effective

mixing during gentle rotations allow quite uniform

fermentations. Aeration, pulp loss, and water evapor-

ation are significantly increased by frequent turnings.

Whether this is favorable or not depends on the ripe-

ness and constitution of the pulp.

Standard Course of Fermentation

0016The progress of fermentation is characterized by

visual changes and changes of odor. Most typical are

fig0001 Figure 1 Preparing a heap fermentation (Cocoa Research Institute of Ghana, Ghana). Reproduced from Cocoa: Production, Products

and Uses. Encyclopaedia of Food Science, Food Technology and Nutrition, Macrae R, Robinson RK and Sadler MJ (eds), 1993, Academic

Press.

1450 COCOA/Production, Products, and Use

color changes and the accumulation of a red sap in the

beans penetrating the shell when the beans die. The

internal appearance of dried beans taken at regular

stages from the fermenter also indicates the progress

of fermentation. Seeds appear slatey as long as antho-

cyanins have not yet been released from their intact

storage cells. After seed death they would dry to

compact beans, violet in color, with the shells sticking

strongly to the nibs. When fermentation is completed

the nibs would turn brown and the beans retain a

plump, porous structure, with a loosened shell. This

is the type of bean required for marketing. For

grading on the market, the internal appearance of

beans is taken to estimate the so-called ‘degree of

fermentation’ in a cut test. (See Colorants (Colour-

ants): Properties and Determination of Natural

Pigments.)

0017 After harvested pods have been broken, the

wet cocoa is inoculated by spores of an ubiquitous

range of airborne microorganisms. The course of

fermentation may be subdivided into four stages (I–

IV). The succession of events is generally as follows

(cf. solid line in Figure 3).

0018Stage I 0018

Pulp In the mucous, acidic (pH 3.5–4.0) pulp, an-

aerobic, microbial metabolism of sugar starts imme-

diately, with alcoholic fermentation first dominating

over the lactic acid fermentation. Pulp is drained off.

Carbon dioxide is produced and this displaces air

from the box. The odor changes from sweet fruits to

ethanol.

0019Seeds Most seeds at this stage are still alive (germin-

ation test).

0020Stage II 0020

Pulp As more pulp has been drained off and sugar

metabolized, less carbon dioxide is produced. Conse-

quently, more air is taken up by the pulp, giving rise

to oxidative formation of acetic acid, an exothermic

process heating up the cocoa from the ambient tem-

perature to 45–52



C. Aeration would now be ran-

domized and accelerated by turning. There is an odor

of acetic acid.

0021Seeds Both the temperature increase to > 45



Cand

the uptake of acetic acid kill the seeds. A maximum

level of acetic acid produced in the pulp is followed by

a maximum amount of acetic acid absorbed in the

beans. The average pH value in the nibs drops from

6.4 to 4.0–4.7. Oxygen is quantitatively consumed in

the fermenting pulp, keeping the nib under anaerobic

conditions and allowing postmortem reactions in the

absence of oxygen.

0022Stage III 0022

Pulp As acetic acid production slows down because

of exhaustion of substrate, bacterial oxidation of

acetic acid causes a slow increase in pulp pH. The

temperature may drop because of lack of substrates

and reduced microbial activity in the pulp, which is

still acidic (about pH 5.0). Lactic acid formation,

which has been suppressed during stage II, may in-

crease once more. The odor is still acidic but changes

to a bread-like impression.

0023Seeds Most of the postmortem enzymatic and non-

enzymatic reactions in the seeds are completed by this

time.

0024Stage IV 0024

Pulp Now, after pulp degradation, a buoyant air

flow from the bottom to the top of the hot cocoa

fig0002 Figure 2 Experimental shallow box fermentation (Malaysian

Agricultural Research and Development Institute, Malaysia). Re-

produced from Cocoa: Production, Products and Uses. Encyclo-

paedia of Food Science, Food Technology and Nutrition, Macrae R,

Robinson RK and Sadler MJ (eds), 1993, Academic Press.

COCOA/Production, Products, and Use 1451

1234567

123456

4.0

5.0

6.0

Acetic acid (nibs)

(mg seed

−1

)

pH (nibs)

Temperature (8C)

(a)

(b)

(c)

(d)

Oxygen (% saturation in the pulp)

Days

fig0003 Figure 3 Differences in the course of shallow box fermentations. Solid lines, well-aerated top layer; seeds from unstored ripe pods

(up to 8 days of fermentation). Dotted lines, less aerated middle layer; seeds from unstored ripe pods (up to 8 days of fermentation).

1452 COCOA/Production, Products, and Use

provides the beans with access to air. A second tem-

perature increase indicates a new aerobic microbial

activity on the shells, which will spread to the nibs if

drying is delayed.

0025 Seeds Stage IV seems to be important to insure

browning of the beans and to harmonize flavor

during subsequent drying. Although no experimental

explanation is available, it is probable that develop-

ing aerophilic organisms may now help to increase

shell permeability. (See Lactic Acid Bacteria.)

0026 Although this fourth stage cannot clearly be dis-

tinguished from the preceding one, it merits special

attention for correct termination of fermentation to

avoid subsequent overfermentation.

0027 Overfermentation subsequent to stage IV is charac-

terized by a steep increase in pH value, first at the

surface of the beans, and later in the nibs. It is

accompanied by a pronounced darkening or even

blackening of the beans and a characteristic hammy

off-flavor which persists in the raw cocoa to an extent

which depends on how severely this process took

place before drying. Overfermentation may be looked

upon as a direct aerophilic microbial attack on the

nibs, destroying the cocoa flavor potential. Depending

on the bean surface moisture at stage IV, the transi-

tion to overfermentation may be slow or very rapid.

Wet beans may turn black overnight and may evolve a

typical strong hammy off-flavor. It appears that the

oxygen concentration in the fermenting mass at this

stage is high enough for saturating respiration of

aerobes and microaerophilic organisms, but too low

for sufficient polyphenoloxidase-triggered browning

reactions which would impair microbial growth.

Variables in the Course of Fermentation

0028 Rate of fermentation The above description may

give the incorrect impression of a uniform process

running on a constant timescale. The stages described

in the previous section and illustrated in Figure 3,

however, are not bound to a fixed time but may be

considerably shorter or longer.

0029 The main reason is aeration: the more the wet

cocoa is aerated in stages I and II, the quicker the

temperature increases, the quicker the peak of acetic

acid appears first and decreases later on (solid and

dotted lines in Figure 3). These differences are

much more pronounced in deep boxes. The succes-

sion of events described is not changed. Wet cocoa

containing > 1.0 ml of pulp per seed would pass

through an ethanol fermentation even under con-

tinued aeration during stage I. In contrast, if no aer-

ation is allowed at stage I, the pulp of wet cocoa may

appear unchanged for several days. Lactic acid fer-

mentation would slowly acidify the seeds at a low

temperature. Later on, forced aeration at stages III

or IV causes a quick pH increase in the nibs.

0030Thus, if fermentations are terminated at a prefixed

time, e.g., after 6 days, the cocoa may be at stage III,

stage IV, or overfermented, revealing a low or a high

pH value in the nibs, respectively. In practice, aer-

ation is intensified by turning, batch size, type of

fermenter, and pulp volume per seed: the smaller the

batch size, the larger is the specific surface for air

uptake. The surface area is also influenced by the

type of fermenter. A small conical heap (e.g., 250 kg)

is much more aerated than a deep box (100 cm

depth). A shallow box (30 cm depth) helps to increase

surface aeration.

0031Nonuniformity in the course of fermentation Not

only variation in rates of fermentation but also local

differences in the fermenting heaps or boxes are due

to inhomogeneous aeration. This applies especially to

deep boxes. At stage I, oxygen is only accessible at the

surface layer, where the process is accelerated, while

the lower layers are still under a carbon dioxide at-

mosphere. The hot zone of pulp oxidation and acetic

acid production slowly moves down to the bottom. If

pulp volume is large and the cocoa is not turned, the

anaerobic, cold situation (stage I) in the bottom layer

may persist during the time the top layer goes through

stages I–IV. After drying, the bottom layer would

produce slatey or violet beans, whereas the top layer

would consist of properly fermented or overfermen-

ted brown beans. Turning after 24 and 48 h reduces

this inhomogeneity, but not enough to eliminate this

effect entirely. As shown in Figure 3, even in shallow

boxes acetic acid accumulation and consumption are

delayed in a less aerated middle layer compared to a

top layer. While this type of nonuniformity is found

during fermentation of wet cocoa beans with a fully

developed pulp layer, the situation may be different

with beans from overripe pods covered by a thin layer

of pulp. Too many turnings would reduce the capacity

of the pulp to fulfill the entire sequence of stages I–IV

and a premature overfermentation would result.

Additionally, beans in unturned, aerated niches of

Dashed lines, middle layer; seeds from ripe pods stored for 10 days (up to 5 days of fermentation). (a) Acetic acid in the nibs; (b) pH

values in the nibs; (c) temperature in the fermenting mass; (d) available oxygen between the beans. Fermentations were performed

with wet beans from Malaysian hybrids. Simplified curves: turnings in (a) after 44, 92, and 139 h; (b) and (c) after 67 and 120 h. Data

derived from Meyer B, Biehl B, Said M, Samarakoddy RJ (1989) Post harvest pod storage: a method for pulp preconditioning to impair

strong nib acidification during cocoa fermentation in Malaysia. Journal of Science and Food Agriculture 48: 285–304.

COCOA/Production, Products, and Use 1453

the batches in boxes or heaps may undergo abnormal

changes due to heat loss, drying out, or fungal attack.

0032 Effect of harvest and postharvest treatment on fer-

mentation and raw cocoa quality Controlled aer-

ation is helpful in overcoming nonuniformity but

does not help in avoiding strong nib acidification

(Figure 3). The minimal pH value in the beans during

enzymatic reactions at stage II is essential for the

formation of flavor precursors rather than the nib

pH when terminating fermentation. With a large

pulp volume per seed, an excess of aeration during

the first stages accelerates acidification, but does nei-

ther limit acetic acid production nor decrease of nib

pH values to less than pH 4.5 (Figure 3, solid line).

However, nib acidification is moderate if the pulp

volume per seed is small (Figure 3, dashed line).

0033 The composition and volume of the pulp depend

on pod ripeness at harvest and postharvest treat-

ments. Ten days of open-air postharvest pod storage

or several hours of bean spreading for pulp drying

after breaking the pods were found significantly to

reduce pulp volume, pulp water, and pulp sugar per

seed (Table 1). These methods are particularly helpful

prior to fermentation of small seeds with a strong

layer of pulp. Subsequent shallow box fermentation

proceeds with well-aerated pulp during stage I. Cor-

respondingly, there is an early steep temperature in-

crease, but the decrease of nib pH value and the

formation and accumulation of acetic acid in pulp

and nibs is considerably reduced (Figure 3). It is as-

sumed that early aeration of a thin pulp surface layer

enhances respiration of sugars by yeasts and reduces

alcoholic fermentation beneath the surface, in con-

trast to a voluminous pulp.

0034 The important pulp/seed relation depends on both

genetic and physiological parameters: large seeds bear

less pulp per seed surface area than small seeds. How-

ever, seed size does not only depend on genetics. Pods

and seeds may be considerably smaller than usual

when growing under adverse conditions, e.g., during

severe drought. Furthermore, the pulp-to-seed ratio is

significantly reduced during ripening. Large pulp

volumes per seed from unripe seeds cause badly aer-

ated slimy fermentations. In overripe pods the pulp

volume is reduced as with stored pods, leading to

well-aerated, rapid fermentations. Raw cocoa flavor

characteristics with particular origins may depend to

a large extent on these physiological stages of wet

cocoa due to local traditions in postharvest treat-

ments. In average, bean size and ripeness of pods

harvested in Malaysia or in Ghana are quite different.

0035Seasonal effects on fermentation and raw cocoa

quality are often due to these differences in the pulp.

In practice, it depends on the circumstances discussed

above whether or not pulp preconditioning is advis-

able.

0036Finally, if wet cocoa is exposed to rain, subsequent

fermentation may lack sugar for heating and acetic

acid production, and would proceed in an erratic way.

Drying

0037Enzyme-controlled browning reactions which do not

occur before drying are essential for quality.

0038For sun-drying, the beans are spread on the ground,

on mats or on wooden floors raised from the ground

and protected against rain. Care is taken to move the

drying cocoa. The layers of cocoa are made shallow

or high to regulate the time of drying. Five to seven

days are usually necessary to reduce the water content

to less than 7.0%, a limit for fungal growth. (See

Drying: Drying Using Natural Radiation.)

0039For artificial drying, wood fires or oil burners are

connected to a flue under a floor of closely spaced

slats. There are many devices of this type for passing a

stream of hot air, but not of smoke, through the layer

of cocoa beans. The industry rejects smoky beans but

the smell of smoke may be confused with a hammy

off-flavor of overfermented beans. On large estates,

different types of industrial driers are used for

tbl0001 Table 1 Changes in the pulp during pod storage and bean spreading

Wet beans Pulp volume

(ml seed

1

)

Pulp water

(g seed

1

)

Sucrose þglucose

(mgseed

1

)

Reducing sugars

(mgseed

1

)

Malaysia, ripe (from unstored pods)

1.19 + 0.14 0.95 + 0.12 121.3 + 57.1

Malaysia, ripe (from pods stored for 10 days)

0.74 + 0.06 0.58 + 0.05 39.9 + 5.4

Malaysia (from unstored pods after spreading)

0.46 + 0.06 0.34 + 0.12 79.2 + 32.3

Ghana, ripe (from unstored pods)

0.84 0.69 91.2

Ghana (from pods stored for 7 days)

0.55 0.41 68.6

Mean values and standard deviations from samples of

10 and

four experiments with 1000 pods each.

Samples from four experiments with 130 kg of wet cocoa, each of different ripeness, sun spreading and surface drying for several hours.

From one experiment with 10 pods each.

Estimated polarographically.

3,5-dinitrosalicylic acid method (Biehl B, Meyer B, Crone G, Pollmann L (1989) Chemical and physical changes in the pulp during ripening and post

harvest storage of cocoa pods. Journal of Science and Food Agriculture 48: 189–208).

1454 COCOA/Production, Products, and Use

effective, short-time drying. (See Drying: Theory of

Air-drying.)

0040 The first stage of drying should be effective in order

to avoid overfermentation or fungal growth on the

wet cocoa, if the internal pH value is high. At this

stage, there is no danger in using hot air (e.g., 100



C),

although the temperature of the beans should not

exceed 40–60



C. In a shallow layer of cocoa beans,

water evaporation will keep the beans cool. However,

hot air flow through a deep layer (10 cm) would

heat the beans excessively. After external dryness of

the beans has been attained, the second stage of

drying can be continued slowly to facilitate enzymatic

browning in the nibs and to allow moisture equilib-

rium in the beans from the wet core to the dry surface.

Artificial drying may be interrupted for a while by a

resting period. In a third stage, the cocoa should

effectively be dried to < 7.5% water (w/w), preferably

in a stream of warm (< 60



C) but not hot air.

0041 Rewetting, polishing, and drying raw cocoa are

practiced locally to improve the external appearance,

especially of moldy beans, but shell breaking and

reinfection of the nibs may result. Rewetting also

increases browning of violet beans.

0042 After drying, the raw cocoa is sorted by removing

flat, broken, externally moldy beans and foreign

matter and by sieving (in the case of nonuniform

bean size) to meet the quality standards. Locally,

batches from different origins may be blended to

reach the maximum percentage of the particular vis-

ible bean defects which are allowed in the official

grading system.

Storage

0043 Provided there are good storage conditions, properly

processed raw cocoa can be stored for years without

any unwanted, serious sensory or analytical indica-

tion of loss of quality or of spoilage. However, once

processed, raw cocoa usually travels a long way from

the tropical origin to the final manufacturing plant,

being subjected to several risks underway. These risks

are primarily water and high air humidity, insect

infestation, and consequences of improper process-

ing. (See Storage Stability: Parameters Affecting Stor-

age Stability.)

0044 In addition to general rules for storage in tropical

and temperate climates, the hygroscopy of raw cocoa

beans must be considered. Dry raw cocoa containing

< 7.5% water will not be attacked by fungi.

According to moisture absorption isotherms, this

bean moisture would not be exceeded at < 80%

relative humidity. However, at a high level of air

humidity, any reduction in air temperature (e.g.,

during shipping) would cause damage by water

condensation and mold growth. Residual humidity

from fast drying must be considered. Repeated mois-

ture determinations are necessary in every lot. Fungal

spores are ubiquitous. However, moldy beans

resulting from fermentation and drying increase the

danger of fungal growth at a critical level of moisture.

Thermophilic and xerophilic fungi such as Asper-

gillus glaucus, A. fumigatus, Penicillium spp. and

Mucor sp. have been found to occur during both

processing and storage. Internal mold is further facili-

tated on broken beans and on injured shells. Brittle

shells, particularly from overfermented or rewetted

cocoa, not only increase the danger of mold but also

increase the amount of broken beans during transport

and stackingof bags. (See Spoilage: Molds in Spoilage.)

0045Out of nine insect species listed by Rohan, four

were mentioned to be important threats to stored

raw cocoa in the tropics: Cadra cautella, Lasioderma

serricorne, Araecerus fasciculatus and Tribolium cas-

taneum. Ephistia elutella, the cocoa moth, is only

found in temperate climates. In tropical countries,

insects are controlled by fumigation of raw cocoa

using the gases methyl bromide or phosphine applied

to stacks of stored cocoa under a gas-proof and water-

proof polythene sheet for 1.5 h. The gas is then ex-

tracted by fans. The sheets are left to protect the cocoa

against uptake of moisture and insects. (See Fumi-

gants; Insect Pests: Problems Caused by Insects and

Mites.)

Production of Cocoa Powder and

Semifinished Products

Cocoa Products

0046Two products are derived solely from cocoa beans:

chocolate and cocoa powder. These products are

interrelated, and two intermediate products play an

important role – cocoa mass and cocoa butter. This

relationship is made clear in Figure 4. The system

Cocoa powder

600

Cocoa butter

600

Chocolate

3800

1200

600

Cocoa mass

1800

Cocoa beans

2100

Cocoa shell

Sugar

1800

Milk powder

800

fig0004Figure 4 Products derived from cocoa beans. The figures are

approximate and in 1000 tonnes. Reproduced from Cocoa: Pro-

duction, Products and Uses. Encyclopaedia of Food Science, Food

Technology and Nutrition, Macrae R, Robinson RK and Sadler MJ

(eds), 1993, Academic Press.

COCOA/Production, Products, and Use 1455