Review of Polyphenols in Theobroma Cacao

Food Research International 33 (2000) 423447
www.elsevier.com/locate/foodres
Review on polyphenols in Theobroma cacao: changes in composition

during the manufacture of chocolate and
methodology for identication and quantication
Jan Wollgast, Elke Anklam *
European Commission, DG Joint Research Centre, Institute for Health and Consumer Protection,
Food Products and Consumer Goods Unit, I-21020 Ispra (Va), Italy
Received 22 December 1999; accepted 18 January 2000
Abstract
Polyphenols have become an intense focus of research interest because of their perceived health-benecial eects, such as anticarcinogenic, anti-atherogenic, anti-inammatory, anti-microbial, etc. Polyphenols in green and black tea, grape seeds, grapes and
(red) wine have raised much attention but chocolate has not been investigated intensively up to now. This review is concerned with
polyphenols in Theobroma cacao, the change in composition and quantity during fermentation, drying, and the manufacture of
chocolate, as well as with analytical methods for isolation, characterisation and quantication. Cocoa beans are rich in polyphenols
in particular catechins and proanthocyanidins. However, a sharp decrease in quantity occurs during fermentation and drying of
cocoa beans and further retention has been reported during roasting. Characterisation and in particular quantication of polyphenols in chocolate has only been developed relatively recently. This work reviews further on the literature on the available
methodology for analysis, quantication, isolation, purication, and structure elucidation of polyphenols in cocoa components and
other commodities. Concerning the analytical methods main emphasis is put on HPLC as it is usually the method of choice due to
its high resolution, high eciency, high reproducibility and relatively short analysis time without restriction on sample volatility.
Moreover, HPLC can be coupled to a variety of detectors such as UVVis, photodiode array (PDA), uorescence, electrochemical
(ECD), and mass spectrometry (MS). However, TLC as a screening method and capillary electrophoresis (CE) as a promising tool
is taken into consideration as well. The characterisation and quantication of the polyphenol composition is amongst the rst steps
to be done to evaluate a putative contribution of chocolate to human health. # 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction
The signicance of antioxidants in preventive medicine is recognised since several years. Diseases that are
believed to be caused or at least enhanced by oxidative
stress include cardiovascular and cerebrovascular disease, some forms of cancer and several other disorders,
such as diabetes and rheumatoid disease, many of which
may be age-related. Although the aetiology of these
conditions is complex, it is recognised that diet, or the
intake of certain dietary components is playing an
essential role in the prevention or management of
* Corresponding author. Tel.: +39-0332-785390; fax: +39-0332785930.

E-mail address: elke.anklam@jrc.it (E. Anklam).
these diseases. Most of these conditions are seen as a

result of an excessive western lifestyle based around
a highly rened diet rich in saturated fat and non-milk
extrinsic sugars, but low in complex plant carbohydrate and sometimes in minerals and vitamins (Zumbe,
1998).
However, increasing interest has been devoted to
naturally occurring compounds, for a long time being
considered non-nutritive. They are produced in the secondary metabolism of many plants and play a role for
instance in the defence against micro-organisms, as
signalling compounds, etc. Therefore, they are called
``secondary plant products'' (Watzl & Leitzmann, 1995),
``phytochemicals'' (Agarwal & Mukhtar, 1996; Newmark, 1996; Yang, Laihshun, Lee & Landau, 1996),
or ``chemo-preventers'' (Zumbe, 1998). Such chemopreventers are believed to have the potential to delay,
0963-9969/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0963-9969(00)00068-5
424
J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447
prevent, or even reverse many conditions from cancer to

dental caries. They have the potential for inclusion into
manufactured foods as an added ingredient, or exist
as an intrinsic ingredient of the food in questions
(Zumbe).
Polyphenols are one such class of compounds. They
occur in a variety of fruits, vegetables, nuts, seeds,
owers, bark, beverages, and even some manufactured
food, as a component of the natural ingredients used.
Polyphenols have become an intense focus of research
interest because of their perceived health-benecial
eects. They have been reported to exhibit anti-carcinogenic (Agarwal & Mukhtar, 1996; Fotsis et al., 1997;
Han, 1997; Kuo, 1996; Newmark, 1996; Suschetet,
Siess, Le Bon & Canivenc-Lavier, 1998; Williamson,
Plumb, Uda, Price & Rhodes, 1996; Yamane et al.,
1996; Yang et al., 1996; Yang, Liao, Kim, Yorkow &
Yang, 1998), anti-atherogenic (Kondo, Hirano, Matsumoto, Igarashi & Itakura, 1996; Konneh & Caen, 1998;
Leake, 1998; Luo, Kannar, Wahlqvist & O'Brien, 1997;
Pearson, Frankel, Aeschbach & German, 1998; Polette,
Lemaitre, Lagarde & Vericel, 1996), anti-ulcer (Saito,
Hosoyama, Ariga, Kataoka & Yamaji, 1998), antithrombotic (Samman, Lyons Wall & Cook, 1998),
anti-inammatory (Benavente-Garcia, Castillo, Marin,
Ortuno & Del Rio, 1997; Middleton, 1998; Rohdewald,
1998), anti-allergenic (Benavente-Garcia et al., 1997;
Middleton, 1998), immune modulating (Rohdewald;
Sanbongi, Suzuki & Sakane, 1997), anti-microbial
(Benavente-Garcia, et al., 1997; Chung, Lu & Chou,
1998), vasodilatory (Fitzpatrick, Bing & Rohdewald,
1998), and analgesic (Benavente-Garcia, et al., 1997)
eects. Polyphenols exert these eects as antioxidants
(Catapano, 1997; Hagerman, et al., 1998; Ioku, Tsushida, Takei, Nakatani & Terao, 1995; Salah, Miller,
Paganga, Tijburg, Bolwell & Rice-Evans, 1995; Serani,
Ghiselli & Ferro-Luzzi, 1996; Vinson, 1998; Williamson
et al., 1996), chelators of divalent cations (Cook & Samman, 1996), inhibitors of the activity of enzymes including
DNA topoisomerase II (Romanczyk et al., 1997), protein
kinase C (Huang & Ferraro, 1992) and protein tyrosine
kinases (Groundwater, Solomons, Drewe & Munawar,
1996), inducers of hepatic electrophile-processing (Phase
II) enzymes (Prochaska & Talalay, 1992), and as modulators of the activity of enzymes such as cytochrome P450 isozyme (Huang & Ferraro, 1992), nitric oxide synthase (Romanczyk et al., 1997), cyclo-oxygenase and
lipoxygenase (Cook & Samman, 1996; Romanczyk et al.,
1997).
Cocoa is unusually rich in polyphenols, but accurate
characterisation, let alone quantication of the polyphenol content has only been developed relatively
recently (Zumbe, 1998). Moreover, chocolate as the
major source of cocoa consumed by humans has been
the least object of investigation.
Therefore, this review is focused on: (1) the chemistry

and biosynthesis of polyphenols; (2) their occurrence in
cocoa and the state of knowledge of changes in composition and quantity of polyphenols during chocolate
manufacturing; and (3) a summary of available methodology for isolation and characterisation of polyphenols
in cocoa components and other food sources with similar polyphenol composition.
The evaluation of the health-protective potential of
polyphenols present in chocolate is a complex problem
and needs an integrated approach. Besides qualitative
and quantitative analysis, this includes an investigation
of the bioavailability of polyphenols from the food matrix
chocolate, the metabolism of absorbed polyphenols
mainly in enterocytes and hepatocytes, the degradation
of polyphenols by the (colonic) microora, molecular
targets, the eective dosage in vivo, and the interaction with other nutrients such as proteins, polysaccharides, fats and some micronutrients. Cocoa
products or cocoa polyphenols have not been investigated intensively in this context up to now, whereas
green and black tea as well as red wine having a
quite similar polyphenol composition have had raised
attention for investigation.
2. Chemistry and biosynthesis of polyphenols
Phenolic compounds or polyphenols constitute one of
the most numerous and widely distributed groups of
substances in the plant kingdom, with more than 8000
phenolic structures currently known (Bravo, 1998).
Polyphenols are products of the secondary metabolism of plants. They arise biogenetically from two main
primary synthetic pathways: the shikimate pathway and
the acetate pathway (Bravo, 1998). Both acetic acid and
shikimic acid are derived from glucose metabolism
(Formica & Regelson, 1995; Schwarze, 1958). Acetic
acid in its active form acetyl-CoA or later in the pathway as malonyl-CoA is the starting point of fatty acid
synthesis in a primary pathway but is also the starting
point in a secondary pathway of the synthesis of the A
ring of the avonoids. Products of the primary shikimate pathway are the aromatic amino acids (phenylalanine, tyrosine) but their degradation leads also into the
phenylpropanoid pathway considered as a secondary
pathway. However, the phenylpropanoid pathway is
ubiquitously present in higher plants forming the core
of a series of related pathways leading to diverse products including avonoids and stilbens. The phenylpropanoid pathway appears essential to the survival of
terrestrial plants providing plant constituents such as
lignin with its important mechanical and structural role.
Furthermore, phenylpropanoid-derived compounds have
distinct roles in the physiology of plants (e.g. as signalling compounds within the plant and as factors con-
trolling male sterility and regulating hormonal activity)

and their inter-relations with other organisms (e.g. as
defensive compounds against micro-organisms and signalling compounds between plants and other organisms) (Rhodes, 1998). The accumulation of compounds
derived from the phenylpropanoid pathway is controlled in a way sensitive to the plant environment
involving a hierarchy of controls at the genomic level
and regulation of a highly specic set of enzyme proteins. Moreover, a spacial compartition allows for a
series of parallel pathways leading to specic products
and for dierent regulatory systems to operate. Thus,
the distinction that is drawn between a primary and
secondary metabolism is somewhat arbitrary and
Rhodes (1998) suggested to think in terms of an integrated metabolism which is unique and characteristic in
plants.
Polyphenols can be divided into at least 10 dierent
classes depending on their basic structure. Table 1 illustrates the basic chemical structure of the main polyphenolic compounds.
Flavonoids, which constitute the most important single group, can be further divided into 13 classes, with
more than 5000 compounds described by 1990 (Table 2).
Their common structure is that of diphenylpropanes
(C6C3C6) and consists of two aromatic rings linked
through three carbons that usually form an oxygenated
heterocycle (Fig. 1).
As mentioned above, the A ring is biosynthesised by
the condensation of three moles of malonyl-CoA
derived from the metabolism of glucose. The C and B
rings are also derived from glucose mechanism by way
of the shikimate pathway and phenylpropanoid pathway, respectively to yield C-9 acids (e.g. cinnamic acid,
hydroxycinnamic acid, and coumaric acid). As CoA
derivatives these C-9 acids condense with the C-6 product from malonate to form a C-15 chalcone. Subsequent ring closure and hydratation gives rise to the
diverse avonoids (Formica & Regelson, 1995).
Summarising, Fig. 2 shows a representation of the
avonoid biosynthesis according to Formica and
Regelson (1995), Rice-Evans, Miller and Paganga
(1996), Rhodes (1998) and Schwarze (1958).
Along with phenylpropanoids or hydroxycinnamic
acid derivatives, avonols and to a lesser extent avones
are found in almost every plant. While avanones and
avones are often found together (e.g. in citrus fruits)
and are connected by specic enzymes, there is a certain
mutual exclusion between avones and avonols in
many plant families and anthocyanins are almost absent
in avanone-rich plants (Rice-Evans et al., 1996).
Flavonoids occasionally occur in plants as aglycones,
although they are most commonly found as glycoside
derivatives (Bravo, 1998; Shahidi & Naczk, 1995). The
preferred glycosylation site is the 3 position and less
frequently the 7 position. Glucose is the most usual
425
Table 1
Main classes of polyphenolic compounds (Bravo, 1998)
Class
Basic skeleton
Simple phenols
C6
Benzoquinones
C6
Phenolic acids
C6C1
Acetophenones
C6C2
Phenylacetic acids
C6C2
Hydoxycinnamic acids
C6C3
Phenylpropenes
C6C3
Coumarines,
isocoumarines
C6C3
Chromones
C6C3
Naphtoquinones
C6C4
Xanthones
C6C1C6
Stilbenes
C6C2C6
Anthraquinones
C6C2C6
Flavonoids
C6C3C6
Lignans, neolignans
(C6C3)2
Lignins
(C6C3)n
Basic structure
sugar residue but others include galactose, rhamnose,

and xylose (Rice-Evans et al., 1996).
Furthermore, individual dierences within each group
of avonoids result from the variation in number and
arrangement of the hydroxyl groups with the most
commonly occurring being those with dihydroxylation
in the 30 and 40 positions (Rice-Evans et al., 1996).
Flavonoids, especially the avan-3-ols catechin, epicatechin, gallocatechin, and epigallocatechin, are the
monomeric constituents of the condensed tannins,
although they are also very common as free monomers
(Bravo, 1998; Herrmann, 1995; Kealey et al., 1998).
426
Table 2
Classication of food avonoids (Bravo, 1998)
Flavonoid
Basic structure
Chalcones
Fig. 1. Basic structure and numbering system of avonoids.
Dihydrochalcones
Aurones
Flavones
Flavonols
Dihydroavonols
Flavanones
Flavanols
Flavandiol or
leucoanthocyanidin
Anthocyanidin
Isoavonoids
Biavonoids
Proanthocyanidins or
condensed tannins
Anthocyanins are the most important group of watersoluble plant pigments and are responsible for the colour of owers and fruits of higher plants. The term
anthocyanin refers to the glycosides of anthocyanidin
(e.g. malvidin, cyanidin). Anthocyanins and polymeric
pigments formed from anthocyanins by condensation
with other avonoids are responsible for the colour of
red wine (Bravo, 1998).
Unlike previously described groups of plant phenolics, tannins are compounds of intermediate to high
molecular weight. Tannins with a molecular mass up to
30,000 Da have been found in carob pods. Tannins are
highly hydroxylated molecules and can form insoluble
complexes with carbohydrates and protein. This function of plant tannins is responsible for the astringency of
tannin-rich foods, because of the precipitation of salivary
proteins. The term ``tannin'' derives from the tanning
capacity of these compounds in transforming animal
hides into leather by forming stable tanninprotein
complexes with skin collagen (Bravo, 1998; Herrmann,
1994; Makkar, 1989; Makkar, Dawra & Singh, 1987).
Plant tannins can be subdivided into two major
groups: hydrolysable and condensed tannins. Hydrolysable tannins consist of gallic acid and its dimeric
condensation product, hexahydroxydiphenic acid,
estered to a polyol, which is mainly glucose. As their
name indicates, these tannins are easily hydrolysed with
acid, alkali, and hot water and by enzymatic action
yielding to polyhydric alcohol and phenylcarboxic acid
(Bravo, 1998; Herrmann, 1994).
Condensed tannins or proanthocyanidins are highmolecular-weight polymers. The monomeric unit is a
avan-3-ol (e.g. catechin, and epicatechin), with a avan-3,4-diol or leucoanthocyanidin molecule as its precursor. Oxidative condensation occurs between carbon
C-4 of the heterocycle and carbons C-6 or C-8 of adjacent units. Most of the literature on the condensed tannin
content refers only to oligomeric proanthocyanidins
(dimers, trimers, and tetramers), because of the diculty
in analysing highly polymerised molecules. Proanthocyanidins, however, can occur as polymers with a degree
of polymerisation of 50 and more. Auto-oxidative or
enzymatic polymerisation of avan-3-ol and avan-3,4diol units have been suggested as the process leading
to the formation of condensed tannins (Bravo, 1998;
Herrmann, 1994; Kealey et al., 1998; Romanczyk et al.,
1997).
427
Fig. 2. Schematic outline of the biosynthesis and interconnections of avonoids.
Interavanoid linkages are acid labile and yield to

anthocyanidins during acid hydrolysis in alcoholic
solutions (e.g. hydrochloric acidbutanol). This reaction
is used for determination of proanthocyanidin molecules. If the sub-units consist only of catechin and epicatechin, cyanidin is the only resulting product of
hydrolysis. Those proanthocyanidins are then called
specically procyanidins, however, this dierentiation is
not followed consistently in the literature. Phlobaphenelike substances are also formed when condensed
catechins are heated in mineral acid solutions from the
further polymerisation of these compounds (Bravo,
1998; Escribano-Bailon, Gutierrez-Fernandez, RivasGonzalo & Santos-Buelga, 1992; Pascual-Teresa,
Gutierrez-Fernandez, Rivas-Gonzalo & Santos-Buelga,
1998a; Prieur, Rigaud, Cheynier & Moutounet, 1995;
Rigaud, Perez-Ilzarbe, Ricardo da Silva & Cheynier,
1991; Romanczyk et al., 1997).
Oligomeric proanthocyanidins are soluble in dierent
aqueous and organic solvents, such as acetone and
methanol. However, high-molecular-weight condensed
tannins are insoluble. In addition, when tannins form
complexes with protein or cell wall polysaccharides, they
remain insoluble. This insolubility is responsible for signicant errors in the quantication of the polyphenolic
content of plants, because polyphenols usually are analysed in extracts, often omitting the quantication of
insoluble or non-extractable tannins (Bravo, 1998).
3. Polyphenols in cocoa: alterations in composition and

quantity from cocoa bean to chocolate state of
knowledge
3.1. The cocoa bean
Cocoa beans are derived from cocoa trees, which are
found in warm, moist climates in areas about 20 latitude north and south of the equator. In general, the
seeds of the Theobroma cacao (of the order Sterculiacae)
are known chiey in two varieties: Criollo and Forastero, with Forastero divided into several subvarieties.
A third group called Trinitario, is essentially a cross
between Criollo and Forastero and is not found in the
wild. The cocoa bean is comprised of an inner nib portion covered by an outer shell (Lange & Fincke, 1970;
Minie, 1989).
The polyphenols in cocoa beans are stored in the pigment cells of the cotyledons. Depending on the amount
of anthocyanins those pigment cells, also called polyphenol-storage cells, are white to deep purpur. Three
groups of polyphenols can be distinguished: catechins or
avan-3-ols (ca. 37%), anthocyanins (ca. 4%) and
proanthocyanidins (ca. 58%). The main catechin is ()epicatechin with up to 35% of polyphenol content. In
one study (Kim & Keeney, 1984) the ()-epicatechin
contents ranged from 34.65 to 43.27 mg per g of defatted sample in freshly harvested Catongo and Forastero
428
cocoa beans. In smaller amounts, (+)-catechin as well

as traces of (+)-gallocatechin and ()-epigallocatechin
have been found. The anthocyanin fraction consists
mainly of cyanidin-3-a-l-arabinosid and cyanidin-3-bd-galactosid. Procyanidins are mostly avan-3,4-diols,
that are 4!8 or 4!6 bound to condensed dimers, trimers or oligomers with epicatechin as the main extension
sub-unit (Belitz & Grosch, 1992; Romanczyk et al., 1997).
The total amount of soluble polyphenols in the dried
fat-free mass of fresh cocoa beans is 15 to 20% (equals
approx. 6% in air dried cocoa beans, containing 54%
fat and 6% water), in fermented beans approx. 5%
(10% and more being considered a sign of a bad fermentation). These values are valid for Forastero beans,
Criollo cocoa beans have approx. 2/3 of the amount of
polyphenols, anthocyanins have not been found (Lange
& Fincke, 1970).
In summary, the dierent polyphenols that have been
identied in cocoa beans or cocoa-products are listed
below (Herrmann, 1995; Kealey et al., 1998; Kim &
Keeney, 1984; Osakabe, Yamagishi, Sanbongi, Natsume, Takizawa & Osawa, 1998; Porter, Ma & Chan,
1991; Rigaud, Escribano-Bailon, Prieur & Souquet,
1993; Romanczyk et al., 1997; Sanbongi, Osakabe,
Natsume, Takizwaw, Gomi & Osawa, 1998):
3.1.1. Catechins
.
.
.
.
()-epicatechin
(+)-catechin
(+)-gallocatechin
()-epigallocatechin
3.1.2. Procyanidins
.
.
.
.
.
.
procyanidin B1=epicatechin-(4b!8)-catechin
procyanidin B2=epicatechin-(4b!8)-epicatechin
procyanidin B3=catechin-(4a!8)-catechin
procyanidin B4=catechin-(4a!8)-epicatechin
procyanidin B5=epicatechin-(4b!6)-epicatechin
procyanidin C1=epicatechin-(4b!8)-epicatechin(4b!8)-epicatechin
. procyanidin D=epicatechin-(4b!8)- epicatechin(4b!8)-epicatechin-(4b!8)-epicatechin
. higher oligo- and polymers, mostly homologues of
epicatechin with 2 to 18 monomeric units
3.1.3. Anthocyanins
. cyanidin-3-a-l-arabinosid
. cyanidin-3-b-d-galactosid
3.1.4. Flavonol glycosides
. quercetin-3-O-a-d-arabinosid
. quercetin-3-O-b-d-glucopuranosid
3.1.5. Others
. clovamide
. dideoxyclovamide
3.2. Cocoa process in countries of origin: eect of
fermentation and drying on polyphenol content and
composition
The correct fermentation and drying of cocoa beans is
essential to the development of suitable avours and/or
avour precursors. After the pods are cut from the
trees, the beans with the adhering pulp are removed and
transferred to heaps, boxes, or baskets for fermentation
to take place. Fermentation lasts from 5 to 6 days with
Forastero beans taking rather longer than Criollo.
During the rst day the adhering pulp becomes liquid
and drains away, with the temperature rising steadily.
Under anaerobic conditions micro-organisms produce
acetic acid and ethanol. These processes inhibit germination of seeds and contribute to structural changes in
fermented beans such as the removal of the compartimentation of enzymes and substrates. Cell liquids move
across cell walls and are spread all over the cocoa nib.
This occurs generally after 2448 h of bean fermentation. By the third day the mass of beans will have been
fairly evenly heated to 45 C and will remain between
this temperature and 50 C until the fermentation is
complete. It is necessary to mix the beans occasionally
for aeration and to ensure that those being initially in
the outside of the heap are exposed to the temperature
in the interior (Kealey et al., 1998; Kim & Keeney, 1984;
Lange & Fincke, 1970).
After fermentation, the beans are placed in shallow
trays to dry. In some growing areas where the main
harvest coincides with the dry season, sun drying is
adequate. In areas where rainfall and humidity do not
permit sun drying, articial drying becomes necessary
(Kim & Keeney, 1984; Minie, 1989).
After fermentation and drying, the cocoa beans
should have moisture content of ca. 57%. This is of
great importance for a correct storage and transport
as above a critical moisture content of 8%, moulds are
likely to develop (Kealey et al., 1998; Kim & Keeney,
1984).
During fermentation of cocoa beans, polyphenols
diuse with cell liquids from their storage cells and
undergo oxidation to condensed high molecular mostly
insoluble tannins. These reactions are both non-enzymatic and catalysed by the enzyme polyphenol oxidase,
even though this enzyme is strongly inactivated during
the rst days of fermentation, remaining only 50 and
6% of enzyme activity after 1 and 2 days, respectively
(Hansen, del Olmo & Burri, 1998). The occurrence of
condensation reactions is conrmed by the sharp
decrease of epicatechin content between the second and
third day of fermentation. Epicatechin and soluble

polyphenol content, respectively, is reduced to approx.
10 to 20% during fermentation. This is not only due to
the oxidation process but also caused by diusing of
polyphenols into fermentation sweatings (Biehl, 1973;
Bracco, Grailhe, Rostango & Egli, 1969; Hansen et al.,
1998; Herrmann, 1995; Kim & Keeney, 1984).
Polyphenol oxidase is also sensitive to drying so that
remaining enzyme activity after fermentation and drying of beans is only about 2%. It is believed that also nonenzymatic oxidation of polyphenols could be important
during the drying process (Hansen et al., 1998).
It has been shown that 2 days of sun-drying of fresh
unfermented cocoa beans alone (without fermentation)
causes a 50% decrease in epicatechin content (ca. 22
instead of approx. 40 mg/g defatted sample). Thus, to
investigate the epicatechin content of fresh cocoa beans,
these had to be freeze-dried immediately after removal
from the pods. Most beans for chocolate manufacture
are fermented; however, it is far from being a standardised process throughout the world, or even within a
region. This is evidenced by a 6-fold variation in epicatechin concentration of 10 samples of fermented cocoa
beans from dierent regions (see Table 3) (Kim & Keeney,
1984).
During the fermentation process, anthocyanins are
hydrolysed to anthocyanidins. The latter compounds
polymerise along with simple catechins to form complex
tannins. Anthocyanins usually disappear rapidly during
fermentation process (93% loss after 4 days). Thus,
anthocyanin content has been considered as a good
index for determination of the degree of cocoa bean
fermentation (Lange & Fincke, 1970; Pettipher, 1986;
Shahidi & Naczk, 1995).
In a recent patent application, procyanidin levels have
been shown to decrease 3- to 5-fold during fermentation
(see Table 4). There is a negative correlation of the
procyanidin content with the degree of fermentation
Table 3
Concentration of ()-epicatechin in fermented cocoa beans from
shipments representing several countries of productiona (mg/g) (Kim
& Keeney, 1984)
Bean source
()-Epicatechin in defatted sample
Ivory Coast
Maracaibo (Venezuela)
Samoa
Trinidad
Bahia (Brazil)
Ghana
Lagos (Nigeria)
Costa Rica
Arriba (Ecuador)
Jamaica
6.22
3.62
10.64
4.68
8.23
4.52
4.68
16.52
8.08
2.66
a
The results are mean values of duplicate injections of a single
extract.
429
and the change in colour from slaty over purple to

brown beans (Kealey et al., 1998).
3.3. Cocoa process for chocolate manufacture in user
countries: eect of roasting, nib-grinding, alkalising, and
conching
The rst process in user countries that must precede
the manufacture of chocolate or cocoa is that of rawbean cleaning. The machinery consists of a series of
operations, which removes bre (from the jute sacks),
stones and grit, metal, bean clusters, and immatures
(Minie, 1989).
Roasting of the whole bean or nib is an essential step
in the manufacture of chocolate liquor or partially
defatted cocoa solids. Cocoa beans are roasted to
develop further the chocolate avour, which should
already exist in the form of precursors arising from the
correct fermentation and drying of the original beans.
Whole bean roasting also loosens the shell so that it can
be readily removed during the winnowing process. The
degree of cocoa roast is a time/temperature dependent
relationship, where the time can vary from 5 to 120 min
and the temperature of whole bean can typically vary
from 120 to 150 C. Lower-temperature roasts are usual
for milk-chocolate and for some dark chocolates. In the
pre-roasting of whole beans, an initial heating step can
be performed at just below 100 C, followed by roasting
of the nibs at elevated temperatures up to about 130 C.
Other heat pre-treatment steps to loosen the shell can be
a thermal shock of the beans given by hot air, steam or
infra-red heat (Kealey et al., 1998; Lange & Fincke,
1970; Minie, 1989).
The next step in conventional cocoa processing
involves nib grinding. Nib grinding is typically performed in two stages, an initial grinding stage to convert
the solid nibs into a uid paste and a nal grinding stage
to achieve the desired particle size. During the grinding,
the nib is ground, for instance by milling, into a uid,
dark brown ``liquor''. The uidity is due to the breakdown of the cell walls and the release of the cocoa butter
during the processing (Kealey et al., 1998; Lange &
Fincke, 1970; Minie, 1989).
Other conventional cocoa processing includes the
separation of cocoa butter from the liquor by either
hydraulic presses or screw presses. For chocolate manufacture this has importance for the cocoa butter that is
added when mixing all ingredients for chocolate and/or
in the end of the conching process (described later)
(Kealey et al., 1998; Lange & Fincke, 1970; Minie,
1989).
Another step may involve alkalising of the beans,
liquor, nibs, or powder with solutions or suspensions of
alkali, mainly to change colour. Alkalisation also aects
avour but it is dubious whether there is any improvement. However, alkalising is not an indispensable step
430
Table 4
Procyanidin levels ppm (mg/g) in defatted samples of cocoa beans (T. cocoa, SIAL 659) with varying degrees of fermentation (Kealey et al., 1998)
Hours of
fermentation
Monomer
Dimer
Trimer
Tetramer
Pentamer
Hexamer
Heptamer
Octamer
Nonamer
Decamer
Undecamer
Total
0
24
48
96
120
21929
21088
20887
9552
8581
10072
9762
9892
5780
4665
10196
9119
9474
5062
4070
7788
7064
7337
3360
2527
5311
4744
4906
2140
1628
3242
2906
2929
1160
888
1311
1364
1334
464
326
626
608
692
254
166
422
361
412
138
123
146
176
302
Tr.
Tr.
Tr.a
Tr.
Tr.
N.D.b
N.D.
60753
57252
58165
27910
22974
a
b
Tr., traces.
N.D., none detected.).
in chocolate manufacture but more common for other

cocoa products such as dark cocoa powder, cocoa
drinks or as an ingredient in a coating or a cookie
(Lange & Fincke, 1970; Minie, 1989).
The basic ingredients required for the manufacture of
chocolate are cocoa nibs, cocoa liquor, sugar, other
sweeteners, cocoa butter, butter fat (oil), milk powder,
milk crumb, and emulsiers. These ingredients have to
be mixed rst either continuously or in batch mixers.
This should produce a chocolate paste of somewhat
rough texture and plastic consistency (Minie, 1989).
The rening of chocolate paste as next step is an
important operation and produces the smooth texture
being desirable in modern chocolate confectionery.
Today, the rening is conducted on multi-step rening
systems using roll reners. In modern reners, the pressure between the rolls is controlled and each roll is
cooled internally so that the desired temperature can be
achieved. Temperatures on the rolls usually vary from
25 to 50 C (Kealey et al., 1998; Lange & Fincke, 1970;
Minie, 1989).
The rened chocolate paste is stored for 24 h at 45
50 C (``ripening''), getting a doughy texture. It may be
used as baker's chocolate but for ne chocolate conching is required (Minie, 1989).
Conching may be regarded as the last process in the
manufacture of bulk chocolate, whether dark or milk
chocolate. It is certainly an essential process for the
development of the nal texture and avour. It is
usually a two step process with the rst to decrease
moisture, drive out volatile substances, and distribute
the fat equally so that all particles are dispersed in a
continuous fat phase. In the second step, cocoa butter is
added and nally lecithin to get a liquid homogenised
paste. The time/temperature conditions during conching
can vary reasonably for the type of chocolate to be
processed. With crumb milk chocolate 1016 h at 49
52 C is frequently be used, 1624 h at 60 C is more
likely with milk powder chocolates, and dark chocolate
is generally conched at higher temperatures, 70 C,
sometimes up to 82 C (Kealey et al., 1998; Lange &
Fincke, 1970; Minie, 1989).
Conching conditions may be modied (shortened) by

pre-treatment of the chocolate liquor, e.g. heat treatment (temperatures above 100 C) as a thin lm (Minie,
1989).
Before lling into forms, the chocolate paste has to be
cooled down to 10 C and reheated several times to 29
31 C for good crystallisation (Lange & Fincke, 1970;
Minie, 1989).
Alteration in content and composition of polyphenol
compounds in the process of chocolate manufacturing,
preferable during roasting, grinding, rening, and conching where rather high temperatures are achieved and
air (oxygen) is present must be anticipated due to the
high redox-activity of polyphenols. However, knowledge of these changes is limited.
Only in a patent application (Kealey et al., 1998) have
such changes been reported in relation to process
parameters. Generally, higher processing temperatures
and/or longer processing times reduce the amount of
polyphenols available in cocoa components. If an alkalising step is present in the process, this also leads to a
remarkable decrease in polyphenol content.
To study the changes in polyphenol composition and
quantity, under-fermented cocoa beans had been roasted at three dierent roasting temperatures (127, 159,
and 181 C) and subsequently milled in two steps into
chocolate liquor. The changes found in total polyphenol
Table 5
Under-fermented bean process results (Kealey et al., 1998)
Product
temperature
Pentamer
content
of total
weight (mg/g)
Total
procyanidin
of total
weight (mg/g)
127 C roasted nibs

Finished liquor
119 C, IBTa
8295 C
1953
1943
24,618
23,710
159 C roasted nibs

Finished liquor
142 C, IBT
5992 C
810
727
21,234
16,826
181 C roasted nibs

Finished liquor
162 C, IBT
5983 C
425
408
12,786
11,656
IBT, internal bean temperature.
content and of a procyanidin pentamer in relation to

roasting temperatures or internal bean temperature
(IBT), respectively, and milling are listed in Table 5.
As the roasting temperature is increased from 127 to
181 C the level of total polyphenols decreases from
24,618 to 12,786 mg/g and that of the procyanidin pentamer from 1953 to 425 mg/g. Similar decreases in polyphenol and pentamer contents are achieved using an
infra-red heating apparatus with the purpose to heat the
whole beans and loosen the shell from the nib. Accordingly, the temperature is an important factor in the
retention of cocoa polyphenols, especially the higher
oligomers (Kealey et al. 1998).
Since it is an objective of the invention to provide
methods of selecting and/or processing cocoa beans for
producing cocoa components or chocolate, respectively,
having enhanced levels of cocoa polyphenols, there is
very limited information on the polyphenol content of
conventional manufactured chocolate. One embodiment
of the invention relates to a ``standard of identity'' dark
chocolate comprising at least 3600 mg (up to more than
8000 mg) cocoa polyphenols per gramme chocolate.
Another embodiment relates to a ``standard of identity''
milk chocolate comprising at least 1000 mg (up to more
than 5000 mg) per gramme milk chocolate. Therefore, it
can be suggested as it has been previously determined
that conventional ``standard of identity'' dark and milk
chocolate have levels below 3600 and 1000 mg, respectively. Similar levels for dark chocolate have been found
by Waterhouse, Sirley and Donovan (1996) where the
amount of total polyphenols, measured by the Folin
method (described below) given as gallic acid equivalents (GAE), was 5 mg/g equals 205 mg polyphenols in a
41 g (1.5 oz) piece of chocolate. These slightly higher
values could be due to the dierent methods used for
analysing the total polyphenol content. In the patent
application (Kealey et al., 1998) only monomeric and
oligomeric avan-3-ols are described, the result given by
weight whereas the Folin-method measures all polyphenols given as equivalents of a standard curve
obtained with gallic acid.
Principally the way by which a higher cocoa polyphenols content in cocoa components or chocolate is
conserved is by choosing cocoa beans rich in polyphenols, using under-fermented beans, and reduce the
time and/or temperature of heat treatment, e.g. in the
roasting or heat treatment of cocoa liquor. However,
the higher polyphenol content is typically associated
with a bitter, astringent avour. Various methods are
presented to reduce the bitter, astringent note, such as
avour additives, milk solids in an amount greater than
12% by weight for milk chocolates, and various variations in the processing of chocolate. For instance, at
least two chocolate liquors with varying levels of cocoa
polyphenols have been used. The ones with a lower
polyphenol content are subjected to a higher tempera-
431
ture whereas the ones with a higher polyphenol content

to a lower temperature to conserve the elevated polyphenol content. Subsequently the two liquors are combined for processing into the nal chocolate product
(Kealey et al. 1998).
4. Available methodology for analysis, isolation,
purication, and identication
The methodology applied to the study of avonoids
depends to a certain extent on the purpose of the investigation. This may be (1) to screen a certain plant group
for the presence of avonoids; (2) to isolate avonoids
from a plant which is known to contain this type of
substance; (3) to determine the concentration of a certain avonoid in a particular plant; or (4) to identify an
isolated and puried avonoid (Markham & Bloor,
1998).
For the rough quantication of polyphenols colourimetric methods are widely used mainly due to their
simplicity and high sensitivity. These include the FolinCiocalteu and Prussian-Blue methods for total polyphenols, the vanillin-HCl assay for catechins and butanol-HCl assay for proanthocyanidins (Lee & Widmer,
1996; Makkar, 1989; Markham & Bloor, 1998;
McMurrough & Byrne, 1992).
The protein precipitation capacities of tannins have
relevance for sensory aspects, nutritional physiology
(e.g. bioavailability of macronutrients such as proteins
and polysaccharides and micronutrients such as iron
and other trace elements), and food processing. Therefore, methods have been developed to determine the
percentage of tannins and non-tannins, respectively, by
so-called protein precipitation methods. Principally,
bovine serum albumin (BSA) or gelatine is added to a
crude polyphenol extract to yield a tannin-protein precipitation. Then either the remaining protein in the
supernatate or the protein of the precipitation after
resolubilisation is determined by the ninhydrin method
(Makkar et al., 1987; Makkar, 1989).
In both cases, a polyphenol extraction procedure is
required, and thus, results depend much on the extraction solvent and procedure. Moreover, a standard curve
with either gallic acid or catechin is common for the
colourimetric methods. Therefore, results can only be
given as equivalents of those standards and results are
rather unspecic for the types of avonoids present
(Makkar, 1989).
More specic results and nowadays more commonly
used are chromatographic techniques, such as thin-layer
chromatography (TLC) and high-performance liquid
chromatography (HPLC), and more recently capillary
electrophoresis (CE). These techniques can be applied
for qualitative and quantitative analysis as well as for
isolation and purication procedures.
432
Fig. 3. Outline of general procedure to be worked through for analysis, quantication, isolation and structure elucidation of polyphenols
in food.
Column chromatography (CC) over Sephadex LH-20

is frequently used for nal purication, since it gives
residue-free solutions for structural analysis by degradation or instrumental techniques, such as mass spectrometry (MS) and nuclear magnetic resonance (NMR).
Generally, the steps that have to be worked through
are indicated below (Fig. 3) (Grayer, 1989; Lee & Widmer, 1996; Makkar, 1989; Markham & Bloor, 1998;
McMurrough & Byrne, 1992):
4.1. Sample preparation
Since some avonoids are unstable, as in the case of
cocoa polyphenols anthocyanins and procyanidins, care
is required during storage and sample preparation. If
quantication is the objective of the subsequent analysis, freezing, most preferable snap freezing in liquid
nitrogen is advisable (Grayer, 1989; Markham & Bloor,
1998).
The rst step in analysis is to crush, mill or grind raw
food (e.g. cocoa beans), or to mix or homogenise processed food (e.g. chocolate) in order to allow a better
contact of the extracting solvent with the sample and to
ensure that the extracted portion is representative for
the entire sample. Since many polyphenols occur in a
bound form as glycosides or esters, the sample preparation may include alkaline or acid hydrolysis to free
bound phenolics before or after the solvent extraction
step. However, the hydrolysis step is omitted if the
measurement of phenolics in the bound forms is desired.
Moreover, procyanidins might undergo partial hydrolysis (Grayer, 1989; Lee & Widmer, 1996; Markham &
Bloor, 1998; McMurrough & Byrne, 1992).
4.2. Extraction procedure
For cocoa polyphenols, 7080% aqueous methanol or
70% aqueous acetone or combinations thereof are most
commonly used for extraction. Water and ethanol have
also been used, however, oligomeric procyanidins are
extracted only partially, high-molecular polymers are

not extracted at all (Grayer, 1989; Lee & Widmer,
1996).
After an exhaustive extraction for a minor component, components may be present in a very dilute level.
Concentration is almost always done at low temperatures (below 40 C) and under reduced pressure to minimise the component degradation. At this stage the
analytes risk to be oxidised (Lee & Widmer, 1996).
Sample clean-up procedures are generally being used
to clean-up extracts prior to further analysis by HPLC
(or CE). The clean-up stage is a critical part of a
method, removing potential interfering components
(e.g. theobromine and caeine in the case of cocoa) and
necessarily varies according to the type of food matrix
to be analysed. These include liquidliquid partitioning
with a non-miscible solvent and column chromatography on Sephadex LH-20, polyamide, Amberlite
XAD-2, preparative HPLC, and solid phase extraction
(SPE) using commercially available disposable cartridges (Grayer, 1989; Markham & Bloor, 1998).
Removal of theobromine and caeine can usually be
accomplished by extraction with chloroform or dichloromethane, since most avonoids have very limited
solubility in these solvents. The most important factors
to consider in selecting the clean-up adsorbent are eciency, recovery, and contamination (Lee & Widmer,
1996).
Sephadex LH-20 was successfully used for further
purication of avonoids and semi-preparative HPLC
on C18 bonded phase prior to the analytical HPLC can
be applied as well. Since the introduction of disposable
solid phase extraction (SPE) cartridges (small packed
chromatography columns) with HPLC packings, SPE
has become the preferred method for clean-up of crude
extracts prior to analysis. The full range of silica-based
polar and non-polar stationary phases in small cartridges is commercially available and SPE on C18 bonded phase is mostly used for isolation of phenolics,
replacing the use of Sephadex for purication steps. By
preconditioning the C18 cartridge sequentially with
neutral or acid solvents fractionation of phenolics into
acid and neutral groups can be achieved as well as
separation of phenolics by further modication by eluting sequentially using dierent euents (Grayer, 1989;
Lee & Widmer, 1996; Markham & Bloor, 1998).
4.3. Analytical methods
Although other methods like colourimetric methods
(see above), paper chromatography (PC), countercurrent chromatography (CCC) or gas chromatography
(GC) may principally be used the most common method
is HPLC due to its high resolution, high eciency,
high reproducibility, and relatively short analysis time
without derivatisation and no restriction on sample
volatility. HPLC is also easily coupled to a variety of

detectors (Lee & Widmer, 1996; Markham & Bloor, 1998).
Thus on HPLC methods will be put the main emphasis. However, TLC as a screening method and capillary
electrophoresis as a promising tool will be taken into
consideration as well.
4.3.1. Thin-layer chromatography (TLC)
TLC is predominantly used as an analytic technique.
Its main advantages over PC for this purpose are its
speed and its versatility, resulting from the range of
stationary phases available. These include alumina,
silica, cellulose, polyamide, reversed phase silica, etc.
Literally, there is an adsorbent and solvent combination
to suit every conceivable type of avonoid. TLC is
especially used for screening of avonoid containing
extracts or fractions from other chromatographic
separations, e.g. PC and CC, but also for qualitative
analysis. Table 6 shows a selected range of solvent
adsorbent combinations that have been used for procyanidins and/or anthocyanins from cocoa beans or
other commodities with a similar polyphenol composition (Grayer, 1989; Markham & Bloor, 1998).
As with PC, spots may be visualised using an UV
lamp. This is a very sensitive method, especially when
using UV uorescent silica. Using various spray
reagents, such as vanillin-HCl and alcoholic ferrous
chloride may enhance the sensitivity on other media
(Grayer, 1989; Markham & Bloor, 1998).
433
4.3.2. High-performance liquid chromatography

(HPLC)
HPLC can be used for separation, quantitative determination, and identication of avonoids. In most
cases, the reported systems for the separation of phenolics and their glycosides in food were carried out by
reverse phase (RP) chromatography on silica-based
C18-bonded phase columns. The average particle diameter of HPLC packings is typically 310 mm. Columns
of smaller particle size usually provide a larger number
of plates per unit time than do columns with larger
particle size. However, 3-mm columns are somewhat
more dicult to work with and become plugged more
easily (Grayer, 1989; Lee & Widmer, 1996; Markham &
Bloor, 1998; McMurrough & Byrne, 1992).
Most of the solvent systems used in analytical HPLC
include binary gradient elutions and, occasionally, isocratic elution. Gradient or isocratic elution using solvents of aqueous acetic, formic, or phosphoric acids
with methanol (MeOH) or acetonitrile (ACN) as
organic modier is common. The range of solvent
strength used in gradient elutions and time required for
analytical separation is dependent on the number and
type of phenolics in the mixture. Oftentimes multiplestep gradients are employed with complex mixtures.
Isocratic methods can be used for partially puried
extracts or crude extracts containing only a few components of similar polarity (Grayer, 1989; Lee & Widmer, 1996; Markham & Bloor, 1998).
Table 6
TLC combinations for polyphenols from cocoa beans and other sources with a similar polyphenol composition
Source
Stationary phase
Solvent optionsa
Reference
Cocoa beans
Silica/cellulose
TLC on silica gel using: TAF

(tolueneacetoneformic acid 3:6:1);
2-D TLC on cellulose: A: TBA
(t-butanolacetic acidwater 3:1:1)
and B: 6% acetic acid
Porter et al., 1991
Cocoa beans (cotyledon)
Cellulose
2-D TLC: A: 5% acetic acid;

B: n-butanolacetic acidwater 4:1:5
Jalal and Collin, 1977
Coee cherries
Silica
Tolueneacetoneformic acid 6:6:1
Colmenarez et al., 1998
Bark from Guazuma ulmifolia
Silica/cellulose
1-D TLC on silica gel:

ethyl acetateformic acidwater
18:1:1; 2-D TLC on cellulose:
A: 6% acetic acid and
B: 2-butanol-acetic acidwater 14:1:5
Hoer, Rimpler and Heinrich, 1995
Beverages; grape seeds; almonds
Silica/cellulose
On silica gel: tolueneacetoneacetic acid

6:3:1; on cellulose: 10% formic acid
Pascual-Teresa, Treutter, Rivas-Gonzalo

and Santos-Buelga, 1998b
Grape and wine
Silica
Tolueneacetoneacetic acid 3:3:1
Flavan-3-ols (in general)
Cellulose
1-D TLC: iso-butanolwateracetic acid

14:5:1 or n-butanolwateracetic acid 4:5:1;
2-D TLC: A: 5% acetic acid;
B: n-butanolacetic acidwater 4:1:5
Sun, Leandro, Ricardo da Silva

and Spranger, 1998
Grayer, 1989
1-D TLC, one-dimensional TLC; 2-D TLC, two-dimensional TLC.
434
The nature of the solvent such as solvent strength and

viscosity of the organic modier has an important
inuence on the chromatic separation. ACN was found
to give better resolution and to give better and sharper
peaks resulting in higher plate number than did MeOH.
However, MeOH may be preferable to ACN because of
its less toxic nature. In addition, a higher percentage of
organic solvent can be used in the mobile phase to prevent deterioration of the reversed phase column. In
some cases, substitution of MeOH with Tetrahydrofuran (THF) which has even higher solvent elution strength value (elutropic strength) than both ACN
and MeOH, improved resolution (Lee & Widmer,
1996).
The pH and ionic strength of the mobile phase are
known to inuence the retention of phenolics on the
column depending on the occurrence of protonation,
dissociation, or a partial dissociation. A change in pH
that increases the ionisation of a sample could reduce
the retention in a reversed phase separation. Thus, small
amounts of acetic acid (25%), phosphoric or triuoroacetic acid (TFA) (0.1%) are included in the solvent system to suppress ionisation of phenolic and
carboxylic groups and hence improve resolution and
reproducibility of runs. Elution of phenolics for RPHPLC is in order of decreasing polarity, so that glycosides are eluted before aglycones, and aglycones with
more hydroxyl groups before those with less (Lee &
Widmer, 1996; Markham & Bloor, 1998).
Postcolumn derivatisation techniques have not been
used much for detecting phenolics, but there is clearly a
great potential therein. One application was the detection of catechins and oligomeric proanthocyanidins
with 4-dimethylaminocinnamaldehyde (McMurrough &
Byrne, 1992; Pascual-Teresa, Treutter, Rivas-Gonzalo
& Santos-Buelga, 1998; Treutter, Santos-Buelga, Gutmann & Kolodziej, 1994).
Rigaud et al. (1993) developed an HPLC method
using a normal phase (NP) silica column and a gradient
of dichloromethane into methanol with constant 4% of
formic acidwater (1:1) mixture as the eluent to separate
procyanidins on a molecular mass basis, without derivatisation. It was repeatedly applied to the analysis of
procyanidin extracts from cocoa as well as for semipreparative and preparative HPLC for isolation, purication, and subsequent structure analysis (Hammerstone, Lazarus, Mitchell, Rucker & Schmitz, 1999;
Kealey et al., 1998; Romanczyk et al. 1997).
Phenolics absorb well in the UV region and the most
commonly used detector for HPLC is a variable-wavelength UV or UVVis detector. No single wavelength is
ideal for monitoring all classes of phenolics since they
display absorbency maxima at dierent wavelengths.
For maximum sensitivity, usually a wavelength near the
maximum is desired, however, in practice, wavelengths
are set for the best overall detection of all components,
which is in case of avonoids mostly around 280 nm

(Lee & Widmer, 1996; McMurrough & Byrne, 1992).
In modern instruments, multichannel, fast scanning,
or photodiode array detectors (PDA) have become the
norm. PDA can yield data both in the time and spectral
domains. It demonstrated the usefulness of qualitative
information in phenolics analysed based on the absorption spectrum. PDA has three major advantages for
HPLC analysis: multiple wavelength detection, peak
identication, and peak purity determination. Since
PDA can record the characteristic UV spectra of the
dierent phenolics as they elute from the column, characterisation and providing of information on the purity
of a peak can be facilitated through comparison of the
spectra at the front, apex, and tail of each peak. Furthermore, the rapid calculation of absorbency ratios
between dierent wavelengths is possible. Commonly,
identication of phenolic compounds in HPLC analysis
was often performed by comparing the retention times
and spectral characteristics of their peaks with those of
standards. Standard samples of aglycones and some of
the more common glycosides are commercially available, however, procyanidins especially oligomers with
three or more monomeric units are not. The order of
elution is largely independent of minor variations in the
solvent system of an RP-HPLC, and so it is also possible to make tentative identications by comparing relative retention times with published lists of such data
(Lee & Widmer, 1996; Markham & Bloor, 1998).
For the peak purity check, spectra from upslope,
downslope, and peak apex are used to compare. However, for isomers, UVVis spectral information alone
cannot be used for positive identication because coeluting isomers can lead to spectra representing a mixture of isomers. In this case and for compounds, of
which standards are not available, additional means of
identication should be used in interpreting HPLC
separation, such as mass spectrometry (MS), nuclear
magnetic resonance (NMR), and fourier transform
infrared (FTIR) (see below) (Grayer, 1989; Lee &
Fluorescence detectors are also used for phenolics,
but have not been applied widely to the detection of
avonoids. Fluorescence detection may oer advantage
over UV detection in terms of enhanced selectivity and
greater sensitivity. Moreover, coupled with UV detection uorescence detection may be useful in peak identication and peak purity determination (Lee &
Widmer, 1996).
More recently, the use of electrochemical detection
(ED) has received much attention. ED is very sensitive
for compounds that can be oxidised or reduced at
low-voltage potentials, such as avonoids. It shows
enhanced sensitivity and selectivity over UV detection,
and thus, is very useful when analysing real samples as it
reduces matrix eects and consequently improves the
quantication and identication of analyte peaks. Time

consuming sample preparation steps, which in addition
are possible sources of error, can be reduced. Furthermore, ED is almost insensitive to changes in mobile
phase conditions associated with gradient elution.
Steady baselines can be achieved at high detector sensitivity settings. In one case (Madigan, McMurrough &
Smyth, 1994) rather complex methanol-phosphoric acid
gradients could have been replaced by an acetic acid
(2.5 to 10%) in water linear gradient achieving a
smoother baseline, along with obviously increased resolution of avanols (Hayes, Smyth & McMourrough,
1987a, 1987b; Lee & Widmer, 1996; Lunte, Blankenship
& Read, 1988; McMurrough & Byrne, 1992).
The use of dual-channel electrochemical detection with
HPLC for the analysis of phenolic compounds has generated particular interest. This is because the use of
working electrodes in series or parallel at dierent operating potentials can be used to provide unambiguous
identication of sample components (Madigan et al.,
1994).
The coupling of a modied NP-HPLC method with
on-line mass spectrometry (MS) analysis using an
atmospheric pressure ionisation electrospray chamber
has been shown to be a powerful tool in the qualitative
analysis and identication of procyanidins in cocoa
components and chocolate (Hammerstone, et al., 1999;
Romanczyk et al., 1997). Combining the data of mass
and retention time is sucient to pre-identify
or conrm identity of substances without previous
isolation. Conrmation of this method as a reliable
quantitative tool for determining levels of procyanidins in cocoa, chocolate, and other food products is
currently being investigated. Since this method does
not require the isolation of substances in order to
characterise their size, it is more rapid and can further provide a check of the reliability of the techniques based on hydrolysis for the estimation of the
average degree of polymerisation (DG) in proanthocyanidins (described later on) (Hammerstone et al.,
1999; Pascual-Teresa et al., 1998a).
4.3.3. Capillary electrophoresis (CE)
CE is a renement of traditional electrophoresis in
which separating power has increased to an extent (up
to 120,000 theoretical plates). It is able to exceed HPLC
by up to 65 times. Sensitivity is also 1012 times higher
than in HPLC, for although sample detection is commonly also by UV absorption, the sample size injected is
only ca. 1 ng. Moreover, analysis times are much
shorter in CE than in HPLC. HPLC, however, often
remains the method of choice for avonoid analysis.
This is due to the fact that in CE collected UV spectra
are often noisy because of the small amount injected
and detection limits for components in the original
sample are higher. CE oers a method for separating
435
avonoids that utilises quite dierent molecular properties. As such CE will sometimes produce separations
where HPLC will not (Lee & Widmer, 1996; Markham
& Bloor, 1998).
The principle of CE is based on a buer-lled capillary tube whose ends are in buer-lled reservoirs containing electrodes. The sample is applied by various
means to one end of the tube and a tension voltage
applied between the electrodes. Either positively or
negatively charged species can be selected by changing
electrode polarity. Neutral components travel together
through the capillary with the electro-osmotic ow of
the buer induced by the potential dierence between
electrodes. Charged sample molecules are separated as
they migrate through the capillary (against the ow)
(Lee & Widmer, 1996).
Flavonoid mobility in CE is determined essentially
by the charge-to-mass ratio of the molecule. Thus,
avonoids not carrying a charge must be ionised by use
of a suitable buer. Borate buers with a pH of 811
and a concentration of 25200 mM are commonly used,
although sodium borate buers can form complexes
with ortho-dihydroxyl groups on the avonoid nucleus
and with vicinal cis-dihydroxyl groups on sugars (Lee &
Flavonoids that are not readily ionised or are hydrophobic may be eectively handled using micellar electrokinetic capillary chromatography (MECC), in which
a surfactant, often sodium-dodecyl sulphate (SDS), is
introduced to produce charged micelles in which the
neutral avonoid migrates. Alternatively, the addition
to the buer of about 20% of an organic solvent such as
methanol or acetonitrile may suce (Markham &
Bloor, 1998).
In the majority of separations, the endo-osmotic ow
of water causes the avonoid to be driven toward the
cathode, and the extent of this eect is determined by
molecular size. This is the major and overriding inuence on mobility. However, avonoid anions, which are
produced by the alkaline pH of the buer, are also
attracted to the anode, with the strength of this attraction being determined by the degree of ionisation. Thus,
the balance of these two eects determines the net rate
at which avonoids migrate along the capillary column
to the cathode. While molecular size is generally selfevident, i.e. from the molecular weight, the degree of
ionisation is dependent upon the pKa of the (most
acidic) hydroxyl group (Lee & Widmer, 1996; Markham
& Bloor, 1998).
MECC has been used as a rapid method to separate
cocoa procyanidin oligomers (Romanczyk et al., 1997).
It has been demonstrated that this method requires only
12 min to achieve the same separation as that obtained
by a 70 min normal phase HPLC analysis. The MECC
buer consisted of 200 mM boric acid, 50 mM SDS and
sodium hydroxide to adjust to pH 8.5. The capillary was
436
a 50 cm 75 mm i.d. uncoated fused silica and the

analytes detected by PDA.
Treutter et al., 1998; Prieur et al., 1995; Romanczyk et

al., 1997).
4.4. Flavonoid quantication
4.5. Flavonoid isolation procedures
A crude quantication of pure single components or

mixtures is possible using colourimetric techniques
(described above) and/or UV/visible spectrophotometry. In cases where the components of a mixture are
of a similar type, e.g. all anthocyanins or all avanols
these methods can give reasonable results using standard curves with commercially available avonoids. In
the case of cocoa components this requires at least
fractionation of anthocyanins and avanols (see
above) or for more accurate quantication even more
tedious isolation procedures using PC, CC, gel permeation chromatography (GPC), and HPLC or combinations thereof as described later (Markham &
Bloor, 1998).
HPLC with a UV/visible detector can be used as such
for quantication. The chromatogramme shows all of
the components that adsorb at the wavelength of interest and the peak area of each component. The peak area
is dependent upon the absorption coecient of that
component at that wavelength. For quantication these
peak areas are then compared with those of standard
compounds, which are either included in the sample
before injection (internal standard) or chromatographed
separately (external standard). The primary functions of
an internal standard are in the determination of the
reliability of extraction, sample preparation, and chromatographic procedures. When the compound class of
each compound of the mixture under study is known
(e.g. from absorption spectra), their levels can be calculated from their relativity to the internal standard peak.
These relatives, at a specic wavelength, are established
by separate chromatography of the internal standard
with suitable standard representatives of each class
(Markham & Bloor, 1998).
In the case of cocoa polyphenols, it is quite well
known that avanols are the predominant group over
avonol (quercetin) glycosides and anthocyanins (e.g.
Belitz & Grosch, 1992; Herrmann, 1995). However, it
has to be established whether procyanidin oligomers
have similar UV responses as their monomeric units
epicatechin and catechin for which standards are
available. An option is the hydrolytic degradation of
the procyanidins in the presence of phloroglucinol or
phenylmethanethiol to yield the monomeric catechins
and thus the average degree of polymerisation (DG) (see
more precise description below). This would give a
quantication of procyanidins as a group as equivalents of the monomeric standard. If oligomeric procyanidin fractions have been isolated previously, the
amount of dimers, trimers, tetramers, etc., could be
achieved (Escribano-Bailon et al., 1992; Pascual-Teresa,
The isolation and purication of individual avonoids

is often required because structures are unknown or as
in the case of cocoa procyanidins standards are not
commercially available. Pure material may also be needed for activity measurements, e.g. antioxidant or anticarcinogenic activity, or for bioavailability studies in
animals, humans or using cell cultures (e.g. the human
intestinal epithelial cell line Caco-2).
A generalised scheme for avonoid isolation that has
often proved to be useful is presented below (Markham
& Bloor, 1998):
1. Initial clean-up of extract
2. Large-scale fractionation using column chromatography
3. Final purication (usually small-scale)
Most of the aqueous methanol or acetone extract
weight is due to carbohydrates. A primary crude
separation of these carbohydrates from the rest of the
extract can be achieved by CC using lling materials
such as the Amberlite XAD resins, derivatised silica gel
(e.g. RP-18) or the Diaion HP products. For this purpose the extract is dissolved in water (or the organic
solvent removed by rotary evaporation) and passed
through the column. The sugars are not adsorbed and
are totally washed from the column with additional
water. The retained less polar compounds, including the
avonoids, are then washed from the column with aqueous or neat alcohol. If desired, some separation can be
achieved at this stage by a stepped increase in the
methanol content (Markham & Bloor, 1998).
Further separation of the avonoid-containing fraction(s) can now proceed with another form of CC.
Polyamide (e.g. MN SC-6) and Sephadex LH-20 are
useful media fur this purpose. Using acidic water, 10
60% aqueous methanol, and 60100% methanol on a
previously acidied polyamide column, anthocyanins,
avonol glycosides and avanol aglycones (and proanthocyanidins), respectively, can be fractionated. Separation by GPC may be used for further fractions from the
larger column separations. Sephadex LH-20 is most
commonly employed and produces separations based
not only on molecular size but also on the H-bonding
interaction. Most avonoids can eventually be isolated
in pure form by GPC, although in some cases the chromatography will need to be repeated to obtain fractions
of sucient purity for nal purication or structure
determination. These columns are usually run with a
single solvent system. For anthocyanins, methanol-acetic
acidwater (10:1:9) is useful, while for other avonoids
watermethanol (or ethanol) or alcohol alone result

in good separations. Acetone can be added to assist
the elution of some tannins (Markham & Bloor,
1998).
For more dicult separations, or where only a small
amount of pure compound is required, a preparative
form of one of the analytical techniques can be used.
Semi-preparative HPLC is often used. A larger (10-mm
diameter) column is required and a higher ow rate
than that used in the analytical work is used. Since only
small amounts can be collected from each run, the pure
compounds are accumulated from several injections by
collection of appropriate peaks as detected by UV/visible absorption or other detection. To ensure the reproducibility over many runs, an isocratic solvent system is
preferred and the column temperature should be accurately controlled through use of a column heater. GPC
on Sephadex LH-20 and both semipreparative RPHPLC and NP-HPLC have been used for isolation and
purication of cocoa polyphenols (Romanczyk et al.,
1997). Alternatively, one-dimensional PC may be used.
Quantities of mg can be obtained by running several
one-dimensional paper chromatogrammes. The appropriate bands are excised, eluted, and the eluates combined. Contaminating polysaccharide material may be
removed subsequently using CC as described above
(Escribano-Bailon et al., 1992; Markham & Bloor, 1998;
Porter et al., 1991).
4.6. Structure analysis by degradation
The most important technique for avonoid analysis
by degradation is that of hydrolysis. For cocoa components, polyphenol composition is quite well known and
standards for the occurring monomers are commercially
available. Since this is not the case for oligomeric and
higher-polymer proanthocyanidins, hydrolytic degradation can give useful information on the type of proanthocyanidin present in the extract, on proanthocyanidin
structure (of isolated and puried components), and on
the average degree of polymerisation (DG).
Total acid hydrolysis in the presence of butanol-HCl
can be used to establish the type of proanthocyanidin.
In the case of cocoa proanthocyanidins this has been
demonstrated to be the procyanidin type, not gallated
and with epicatechin over catechin being the dominating
extension unit (Escribano-Bailon et al., 1992; PascualTeresa, Treutter et al., 1998).
Partial acid hydrolysis is performed to identify the
``lower'' (initiation or terminal) sub-units. Only part of
the inter-avan bond is broken, releasing free lower
subunits; thus, for example, from a trimer procyanidin
the non-hydrolysed trimer, the lower dimer and monomer are obtained (as can be identied by TLC and/or
HPLC) (Escribano-Bailon et al., 1992; Pascual-Teresa,
Treutter et al., 1998).
437
Acid hydrolysis in the presence of phloroglucinol and

of phenylmethanethiol, with later desulphuration of the
thioethers (using Raney Nickel in methanol), are carried
out for the identication of ``upper'' extension units. In
this reaction, the lower sub-units are released as free
catechins, while the upper sub-units in form of phloroglucinol adducts or benzylthioethers, respectively. The
ratio between amounts of adducts or thioethers and of
free catechins released can thus be related to the degree
of polymerisation of the compounds present in the
extract. The use of phenylmethanethiol is of disadvantage
due to its strong and persistent smell and, moreover, the
chromatographic analysis of the phloroglucinol adducts
have been reported to be more easy (Escribano-Bailon
et al., 1992; Pascual-Teresa, Treutter et al., 1998; Prieur
et al., 1995; Rigaud et al., 1991; Romanczyk et al., 1997).
4.7. Structure analysis using instrumental techniques
4.7.1. Absorption spectroscopy
Absorption spectroscopy is used in particular for the
quantication of avonoids and for preliminary studies
of the structures of avonoids isolated by chromatography or detected by HPLC. A major advantage of this
technique is that it requires only very small quantities of
avonoids, e.g. the amount commonly available from
one two-dimensional PC spot is normally adequate.
The absorption spectra of a wide selection of avonoids is now available in the literature, and these spectra
can provide a useful means of determining the avonoid
type, the oxygenation pattern, and even occasionally the
glycosalation pattern. However, cocoa polyphenols
consist mainly of avanols (catechins, procyanidins) and
only small amounts of avonol and anthocyanidin glycosides. Thus absorption spectroscopy could only conrm the presence of a avanol, but could not give
further information for the identication whether epicatechin or catechin or an oligomeric procyanidin. For
this purpose and positive identication in general, techniques such as MS and NMR are indispensable (Grayer,
1989; Markham & Bloor, 1998).
4.7.2. Mass spectrometry (MS)
To-date, MS is used mainly in avonoid analysis for
the conrmation of molecular weight (Markham &
Bloor, 1998). Thus, this technique could give very useful
information on the number of catechin sub-units in
cocoa procyanidins.
Using electron impact mass spectrometry (EIMS) on
sub-milligramme quantities, other than the molecular
weights, chemical formulae of avonoids can be determined, and valuable information is obtained as to the
substitution patterns of A- and B-rings. During the electron impact procedure, avonoids are cleaved into a
number of fragments according to certain pathways. For
instance, the fragmentation of avanones and dihydro-
Cold 70%
aqueous methanol
70% aqueous acetone

+70% aqueous
methanol (beans);
acetonewateracetic
acid 70:29.5:0.5
70% aqueous acetone; Ethyl-acetate extract on

70% aqueous methanol Sephadex LH20 eluted with
step wise gradient of
water into methanol
70% methanol
60% acetonewater
80% aqueous acetone
Boiling water
Cocoa beans
(cotyledon)
Cocoa beans;
dark chocolate
Cocoa beans;
chocolate;
chocolate raw
products
Cocoa beans
and grape seeds
Cocoa beans
Cocoa beans
(fermented and
unfermented)
Cocoa liquor
Sephadex LH-20 eluted with

water-acetone step wise
gradient
RP-HPLC:
25% methanol containing
0.03% triouroacetic acid
(TFA) isocratic
UVVis at 280 nm
UVVis at 280 nm
FABMS and
C NMR; thiolytic
degradation
Osakabe et al.,
1998
Kim and Keeney,

1984
Porter and Chan,

1991.
Rigaud et al.,
1993
Romanczyk et al.,
1997
Hammer-stone
et al.,
1999
Jalal and Collin,

1977
Kealey et al.,
1998
Ref.
(continued on next page)
MS and 1H and 13C NMR
13
TLC on silica gel using:

Vanillin-HCl for TLC;
TAF (tolueneacetoneformic
UVVis at
acid 3:6:1);
280 nm for HPLC
2-D TLC on cellulose: A: TBA
(t-butanol-acetic acidwater 3:1:1)
and B:
6% acetic acid; RP-HPLC:
methanolacetic acid (1%)
1:4 isocratic
FABMS using a
liquid secondary ion mass
spectrometry (LSIMS)
technique in positive and
negative ion modes or
MALDITOF/MS;
13
C NMR and 1H NMR
spectra; HOHAHA,
HMQC; COSY, APT,
XHCORR spectras;
thiolysis and
desulfuration
Due to molecular mass

in combination with
DAD spectra
Structure elucidation
Microthiolysis to
identify procyanidins
DAD 280 nm;

uorescence l
ex 276 nm, l em 316 nm
DAD and MS with

API-ES ionisation
chamber using both
scan mode and
selected ion monitoring
(SIM)
UV-light, vanillin-HCl,
titanium oxalate
DAD 280 nm; uorescence l

ex 276 nm,
l em 316 nm
Detection
NP-HPLC
UVVis at 280 nm
dichlormethanemethanolformic
acidwater with ratios A: 4:43:1:1
and B: 41:7:1:1
NP-HPLC:
dichlormethanemethanol
gradient with const.
4% acetic acidwater (1:1);
RP-HPLC watermethanol
gradients (const. 0.5%
acetic acid); MECC:
200 mM boric acid, 50 mM
SDS at pH 8.5
NP-HPLC;
gradient with const. 4%
acetic acidwater (1:1)
2-D TLC on cellulose:

A: 5% acetic acid; B:
n-butanolacetic acidwater
4:1:5
NP-HPLC;
gradient with const. 4%
acetic acidwater (1:1)
Analysis method
RP-C18 SEP-PAK cartridge

RP-HPLC:
eluted with 40% aqueous methanol watermethanolacetic acid
(87:8:5) isocratic
Extracted with ethyl acetate;

1. aqueousueous phase:
A: on Sephadex LH-20
(polymer fraction)
B: on TSK-HW-40(F) eluted
with methanol (procyanidins
with increasing Mr-weight);
2. ethyl acetate phase on
water into methanol step
gradient
Evaporated and redissolved

in 60% acetone; salted out
with NaCl; washed with
chloroform, extracted
with ethyl acetate;
SPE on C18; eluted with

water to remove sugars then
acetone water acetic acid
70:29.5:0.5 for procyanidins
Acidied ethyl acetate

extract further puried by
several runs of PC
Acetonewateracetic
acid 70:29.5:0.5
Cocoa beans or
cocoa liquor
Purication and/or
fractionation
Extraction solvent
Source
Table 7
Overview on methods for analysis, isolation, purication, identication and quantication of polyphenols from cocoa beans and products and sources with similar polyphenol composition (procyanidins)a
438
70% aqueous
methanol (for
total polyphenols
and tannins);
75% aqueous
acetone
(for HPLC)
80% aqueous ethanol
95% methanol
90% aqueous methanol

(apples/grapes);
(beans)
Cocoa beans;
non-alkalised
low fat
cocoa powder;
and instant
cocoa powder
Cocoa liquor
Chocolate and
cocoa powder
Apple; black
grapes, canned
kidney beans
TLC on silica plates:

tolueneacetoneacetic acid
3:3:1; RP-HPLC: water10%
acetic acid gradients
C18-Sep-Pak cartridges eluted

with water (phenolic acids),
ethyl acetate (catechins and
oligomeric procyanidins),
methanol (polymeric
procyanidins and anthocyanins);
ethyl acetate fraction again on
Sep-Pak, eluted with
diethyl-ether (monomeric
catechins) and methanol
(oligomeric procyanidins)
Grape and wine
RP-HPLC:
wateracetonitrile gradients
in 0.05 M
ammonium phosphate buer
(pH 2.5)
RP-HPLC: 5 to 25%
acetonitrile in phosphate
buer (pH 2.4)
FolinCiocalteau reagent
Total phenols and tannins

with FolinCiocalteau
reagent (RFC);
RP-HPLC (only epicatechin):
water at pH 2.6
(phosphoric acid)methanol
gradients
Prep RP-HPLC: 40%

methanol containing
0.1% TFA isocratic
Analysis method
RP-HPLC:
2.5% acetic acid-acetonitrile
gradients
C18 Sep-Pak cartridge

eluted with water (organic
acids and sugars), ammonia
solution at pH 9.5 (phenolic
acids), and methanol
(for avonoids)
RP-chromatography on
Diaion HP-2MG eluted
with 20% ethanol (impurities
such as sugars and protein)
and 80% ethanol
(polyphenols); air dried and
dissolved in DMSO
Precipitation of tannins in
methanol extract at pH 4
using 1% NaCl in 10%
gelatine solution;
acetone extract (for HPLC):
ltration and saturation
with NaCl (separation in
two phases); use of acetone
phase, evaporated to dryness;
dissolved again in
watermethanol
Diaion HP2MG column

eluted with wateracetone
(containing 0.05% TFA)
step-wise gradient
Purication and/or
fractionation
Pure solutions
of procyanidin
or isolated from
grape seeds
Methanol (for grape)
80% aqueous ethanol
Cocoa liquor
Wine and
grape seeds
Extraction solvent
Source
Table 7 (continued)
Vanillin-HCl for TLC;

UVVis at 280 nm
for HPLC
UVVis at 280 nm
Dual-electrode LC-ECD
UVVis at 280 nm and

uorescence in series
Absorbance at 760 nm
Absorbance at 760 nm
RFC: absorbance at
760 nm
(gallic acid equivalents);
HPLC: DAD
at 278282 nm
UVVis at 280 nm
Detection
Thiolytic degradation
with toluene-a-thiol
(+HPLC) for
mean DP
Sun et al., 1998
Rigaud et al.,
1991
Lunte et al., 1998
Arts and
Hollmann, 1998
Waterhouse
et al., 1996
Sanbongi et al.,
1997
Thiolytic degradation
with phenylmethanethiol
in sulphurous acid
(+RP-HPLC);
desulphuration with
Raney Nickel (+HPLC)
Sanbongi et al.,
1998
MS and 1H and 13C NMR
Serra Bonvehi
and Ventura
Coll, 1997
Ref.

439
75% aqueous acetone
Methanol
70% aqueous acetone
50% aqueous methanol Solka oc cellulose column

eluted with methanol and
acetone; further fractionation
on Sephadex LH-20 eluted with
20, 40, and 80% acetone
Grape seeds
Grape seeds
Coee cherries
Decaeinated
black tea
Ethylacetate extract washed

with dichlormethane (to
eliminate caeine); on
Sephadex LH-20 column
eluted with 80% aqueous acetone
Sephadex LH 20 eluted
with 96% ethanol
Prep-RP-HPLC:
acetonitrileacetic acidwater
10:0.5:89.5
TLC:
tolueneacetoneformic
acid 6:6:1; RP-HPLC 0.5%
phosphoric acidmethanol
gradients
RP-HPLC:
4,5% formic acidacetonitrile
gradient
Prep-RP-HPLC:
15% methanol
RP-HPLC: 2%
acetic acidmethanol
gradient
Ethyl acetate
extract on Sephadex
LH 20 eluted
with 96% ethanol
Cold methanol
containing 0.05%
ascorbic acid
(20 C)
Grape seeds,
apple skin,
lentil, almond
esh
Diaion HP-20 resin
TLC on silica gel:

tolueneacetoneacetic acid
6:3:1; TLC on cellulose:
10% formic acid; RP-HPLC:
watermethanol4.5%
formic acid gradients
Sephadex LH-20
Methanol
(for grape seeds,
almonds)
RP-HPLC: 5%
acetic acid-acetonitrile
gradients
Analysis method
Beverages;
grape seeds;
almonds
Purication and/or
fractionation
Liquid-liquid extraction:
1. ethyl acetate at pH 2.0;
2. ethyl acetate phase
evaporated and
redissolved in water
extracted with ethyl
acetate at pH 7.0
Extraction solvent
Red wine
Source
Table 7 (continued)
DAD 280 nm
DAD 280 nm
DAD at 280 nm
UVVis at 280 nm;

purity of the compounds
checked by HPLC-DAD
and LCMS
DAD double online

detection: at 280 nm and
after derivatisation
with DMACA at 640 nm
DAD at
535 330 310 280 nm
Detection
Ref.
Negative ion
electrospray-MS; 1H,
C NMR
13
Davis et al.,
1997
Colmenarez
et al., 1997
Escribano-Bailon
et al., 1992
Takahashi,
Kamiya and
Yakoo, 1998
Plumb,
Pascual-Teresa,
Santos-Buelga,
Cheynier and
Williamson, 1998
FTIR using KBr discs;

C NMR; autooxidation
reaction with
n-butanol-HCl-Fe(III) to
assess homogeneity of
procyanidin-rich fractions
13
TLC on silica
(tolueneacetoneacetic
acid 3:7.5:1); complete
acid hydrolysis
with butanol-HCl ; partial
hydrolysis with 0.1 N HCl
(+ HPLC); thiolysis and
desulfuration (+HPLC);
enzymatic hydrolysis
(only for galloyl esters)
(+HPLC)
Hydrolysis with tannase

(galloylated procyanidins);
thiolytic degradation with
toluene-a-thiol
in acetic acid (+HPLC)
LCESIMS
(pure procyanidins);
acid hydrolysis in the
presence of phloroglucinol
and phenylmethanethiol
with subsequent
desulphuration of
thiolethers; MS and
2-D NMR (for
glycosylated catechins)
Pascual
Teresa et al., 1998b
MS (PE-API/Sciex ion spray Ghiselli, Nardini,

interface and a
Baldi and
triplequadrupole MS Sciex Scaccini, 1998
Taga 6000E) with positive
ion mode for anthocyanins,
negative ion mode for
catechins, phenolic acids
and avonols
440
Water
Extracted 75%
aqueous acetone (barley)
Black tea
Beer and
barley
Standard
solutions and
oak leaves

at 90 C
( unconjugated
phenolics);
1.2 M HCl in 50%
methanol at 90 C
(total phenolics)
23 vegetables
Protein precipitation
with BSA; alkaline hydrolysis
of tanninBSA complexes
Ninhydrin method for

BSA-amino acids of
hydrolysed complex
Colourimetric analysis with

using catechin as a standard
1-D TLC on Silica gel:

ethyl acetateformic acidwater 18:1:1; 2-D TLC
on cellulose: A: 6%
acetic acid and
B: 2-butanolacetic
acidwater 14:1:5
Washed with
dichloromethane;
ethyl-acetate extract on
Sephadex LH-20
eluted with A: 50%
ethanol and B:
ethanolwateracetone
9:9:2
Cold 70%
aqueous ethanol;
96% ethanol
and 70% ethanol
(reux)
Bark from
Guazuma
ulmifolia
RP-HPLC: 4.5%
formic acid-acetonitrile
gradients or
10 mM NH4OAc0.1%
formic acidmethanol
gradients
RP-HPLC: 3.5% acetic

acidmethanol gradients
RP-HPLC: 2.5 to 10%

acetic acid
RP-HPLC: 2% acetic
acid-acetonitrile or
1% citric acid
(pH 2.8 with
NaOH)-acetonitrile
or 2% acetic acid in 2%
EDTA sodium
salt-acetonitrile
RP-HPLC 0.5%
acetic acid-acetonitrile
gradients
Analysis method
RP-HPLC
acetonitrilewater
(containing 0.03%
formic acid) gradient

96% ethanol and acetone
Extraction of caeine and

chlorophyll with chloroform
Purication and/or
fractionation
Plant leaves of
70% aqueous methanol Ethyl-acetate extract;
Ameyena scandens
CC on silica-gel
Unripe
almond fruits
Cold methanol
containing 0.05%
ascorbic acid
(20 C)
Hot water
Black tea
Beer
Extraction solvent
Source
Table 7 (continued)
Photometric
(absorbance at 570 nm)
Vanillin sulphuric acid or

vanillinethanolwaterHCl
0.5:5:95:25)
DAD at 280 nm; ESIMS
DAD at 280 nm;

HPLCMS (ESIMS)
UV at 254 nm; ECD;

cyclic voltammetry
Coulochem II ECD
(colourimetric electrode
in series with
amperometric electrode
DAD at 280, 380, 460 nm
DAD in two runs between

190390 and 390-590 nm
Detection
Hayes et al.,
1987a, 1987b
Madigan et al.,
1994
Bailey, Nursten
and McDowell,
1991
Opie, Robertson
and Cliord, 1990
Ref.
Makkar et al.,
1987
Vinson, Hao,
Su and Zubick,
1998
Hoer et al., 1995
n-butanol-HCl-Fe(III)
to assess homogeneity of
procyanidin-rich fractions
Gariboldi,
Mascetti, Galli,
Caballion and
Bosiso, 1998
HPLC-retention times;
Pascual Teresa
Rf-values of 1-D TLC
et al., 1998a
and 2-D TLC; total
hydrolysis (butanol-HCl);
partial hydrolysis (1 N HCl);
acid hydrolysis
in the presence of
phloroglucinol and
phenylmethanethiol with
subsequent desulphuration;
MS

441
75% aqueous acetone;

aqueous ethanol or
methanol; ethyl
acetate; C18-SPE
or precipitation
with PVPP
(liquids)
Flavonoids in
general
UVVis at 280 nm (avan-3-ols)

or 270280 and 475555 nm
(anthocyanins) or 350365 nm
(avonol glycosides) or at
640 nm (after derivatisation
of avan-3-ols with DMACA);
DAD and ECD (single working
or dual electrode amperic
detectors); uorescence
McMurrough
and Byrne, 1992
Makkar, 1989
Photometric (at 570 nm

for determination of proteins;
at 510 nm for tannins)
Ref.
Treutter et al.,
1994
DAD at 280 nm and at 640 nm

after derivatisation with DMACA
Detection
a
Abbreviations used here that are not explained in the text or are dierent: SPE, solid phase extraction; DAD, diode array detection; API-ES, atmospheric pressure ionisation electrospray;
MALDITOF, matrix assisted laser disorption ionisationtime of ight; HOHAHA, homonuclear HartmannHahn; HMQC, heteronuclear multiple quantum coherence; COSY, homonuclear correlation spectroscopy; APT, attached proton test; DMACA, p-dimethylaminocinnamaldehyde; ECD, electrochemical detection; PVPP, polyvinylpolypyrrolidone.
Aq acetone extracts
saturated with NaCl
(two phases): the aqueous phase
contains tannins; the acetone
phase (avonol glycosides, lower
avanol oligomers) can be
applied on: C18-SPE, SEP-PAK
cartridges or Sephadex LH-20
A: protein determination
with ninhydrin;
B: tannin determination
with 0.01 M FeCl3 in
0.01 M HCl
Ferric ammonium citrate
in alkaline solution or
(total phenolics);
vanillin HCl (avan-3-ols);
RP-HPLC:
A: steep gradient of
methanol
095 or 050%
in 2.5% acetic acid,
B: gradient of
010% acetic acid
Precipitation with BSA;

pellet washed and dissolved
in SDS (1%)triethanolamine
(5%) solution or just in 1% SDS
Water or
methanol
Analysis method
Tannins in
general
Purication and/or
fractionation
RP-HPLC:
5% formic acidmethanol
gradients
Extraction solvent
Reference
compounds
(commercially
available or
isolated from
horse chestnut)
Source
Table 7 (continued)
442
avonols often involves retro-DielsAlder processes.

During cleavage, many intact A- and B-ring fragments are
formed, the combination of which tends to be characteristic for the class of avonoids to which the compound
under study belongs. In order to give a good EI mass
spectrum and molecular ion, the compound under study
must be volatile at the probe temperatures used in high
vacuum within the mass spectrometer (100230 C).
Flavonoid aglycones meet this requirement, but their
glycosides (and probably oligomeric procyanidins) do
not. However, other mass spectral techniques are available, e.g. negative ion fast atom bombardment MS (FAB
MS) and eld desorption MS (FDMS) (Grayer, 1989;
Lee & Widmer, 1996; Markham & Bloor, 1998, pp. 15, 30,
35).
As mentioned above, the coupling of HPLC or CE
with on-line mass spectrometry (MS) analysis can be a
powerful tool in the qualitative analysis and identication as well as quantitative determination of avonoids,
in particular of procyanidins in cocoa components and
chocolate, without previous isolation (Escribano-Bailon
et al., 1992; Hammerstone et al., 1999; Kealey et al.,
1998; Romanczyk et al., 1997).
4.7.3. Nuclear magnetic resonance (NMR)
NMR spectroscopy is a powerful tool in avonoid
structure determination. As well as providing information on the chemical environment of each proton or
carbon nucleus in the molecule, the technique can also
be employed to determine linkages between nearby
nuclei, often enabling a complete structure to be assembled (Grayer, 1989; Lee & Widmer, 1996; Markham &
Bloor, 1998).
Basic proton and 13C spectra are usually obtained in
separate experiments. Due to the relatively low natural
abundance of the 13C isotope, 13C experiments take
considerably longer to acquire sucient data to yield a
presentable spectrum (hours), whereas proton spectra
can be obtained in a matter of minutes. A decipherable
proton spectrum can be obtained with as little as 0.3 mg
of sample, whereas 13C spectra generally require larger
samples, typically more than 1 mg (Markham & Bloor,
1998), one author suggests even more than 15 mg
(Grayer, 1989).
Most solvents produce their own proton and/or 13C
signals, and this may inuence the choice of solvent.
Solvents also exert some eect on the spectra. Therefore, the same solvent should be used when comparing
data. DMSO-d6 (dimethyl sulfoxide) is the solvent of
choice for many avonoids; however, for avans and
proanthocyanidins, methanol-d4 has been reported to
give the best results (Markham & Bloor, 1998).
In many cases of well-studied compound classes as
the avonoids, basic one-dimensional proton and 13C
spectra are all that is required to conrm a suspected
structural type. Interpretation of the spectral data can
443
often be accomplished through comparison with

published data for known compounds. Thus, proton
NMR spectroscopy can be used: (1) to determine the
oxygenation pattern of the molecule; (2) to determine
the number and structure of functional groups other
than hydroxyls; (3) to establish the number of sugars
present in glycosides; and (4) to distinguish between
dierent classes of avonoids. 13C NMR spectroscopy
gives valuable information as to the carbon skeleton of
a compound, e.g. the number of carbon atoms and
which of these carbons is oxygenated. It can also be
used to distinguish between the various classes of
avonoids. However, the most valuable contribution
of this technique has been that it has provided a relatively easy means of identifying the sugar moiety of
avonoid C-glycosides (and of O-glycosides), of establishing the presence and identity of any acyl groups, and
of determining the positions of linkages between the
various moieties (Grayer, 1989; Lee & Widmer, 1996;
Markham & Bloor, 1998; Romanczyk et al., 1997).
If a review of known compounds fails to provide
suitable data for structure determination by comparison, then a number of more sophisticated NMR techniques are available to help to determine linkages within
the molecule. In most of these experiments the instrument automatically combines the results of many
experiments and the data is presented as a two-dimensional contour plot (Grayer, 1989; Markham & Bloor,
1998; Romanczyk et al., 1997). However, it is beyond
the scope of this overview to discuss these techniques in
detail.
4.7.4. Fourier transform infrared (FTIR)
IR spectroscopy has been less frequently used for
identication of phenolic substances. However, after
GC or HPLC, the combination of the two detection
techniques IR and MS can be a powerful tool for identication of individual phenolics, especially for positional isomers. For such compounds, mass spectra are
usually virtually identical but IR spectra can be very
dierent (Lee & Widmer, 1996).
4.8. Overview
Table 7 gives an overview of the procedures used for
analysis, isolation, purication, identication and
quantication of polyphenols in cocoa products and in
other sources with similar polyphenol composition,
namely rich in procyanidins.
5. Conclusion
The importance and actuality of phytochemicals in
general and of polyphenols in particular as promising
chemo-preventers in human nutrition and medicine
444
have been touched in the introduction. This will be

handled in more detail in another review study.
Polyphenols exist as an intrinsic ingredient in cocoa,
and thus, cocoa products such as chocolate could
become a functional food to confer benecial health
eects to the consumer like it is suggested for red wine
and green or black tea (Zumbe, 1998). However, during
processing of cocoa beans and the further chocolate
manufacture there is a remarkable decrease in the polyphenol content. Such a decrease has been described quite
well for the fermentation and drying of cocoa beans in
the countries of origin (e.g. Biehl, 1973; Bracco et al.,
1969; Kealey et al., 1998; Kim & Keeney, 1984; Lange &
Fincke, 1970). On the other hand, cocoa processing in
user countries and in particular the manufacture of chocolate in relation to changes in composition and content
of polyphenols has so far not been much object of investigation with one exception of a patent application
(Kealey et al., 1998). Moreover, there is even very little
knowledge on the polyphenol content in chocolate,
especially milk chocolate, although these cocoa products
are widely consumed in western countries. Variations in
process parameters (e.g. the time/temperature condition
of roasting) may provide cocoa products with enhanced
levels of polyphenols.
Nevertheless, to get an answer to the question, whether
chocolate might contribute to human health as a functional food according to its polyphenol content, one has
to get answers to more concrete questions such as:
. Which cocoa polyphenols are present and in what
quantity can they be still found in chocolate, whether
dark or milk, and what are the key process factors
responsible for their retention?
. Do those polyphenols already exert bioactivity in
the digestive tract?
. Are they absorbed from the intestine and if so to
what extent (bioavailability) and for how long do
they remain in the body before excretion (biokinetics)? What inuence on bioavailability has the
food matrix chocolate, e.g. the milk proteins in
milk chocolate?
. What happens with the unabsorbed polyphenols in
the colon exposed to the micro-organisms?
. Absorption pre-supposed are the polyphenols
metabolised, and if so what bioactivity do those
metabolites exert?
. What interactions with other nutrients or nonnutrients do exist?
. What are the molecular targets of cocoa polyphenols and their metabolites?
. What dosage can be considered as eective for
either short-term or long-term eects?
References
Agarwal, R., & Mukhtar, H. (1996). Cancer Chemoprevention by
Polyphenols in Green Tea and Artichoke. In N. Back, I. R. Cohen,
D. Lajtha, A. Lajtha, & R. Paoletti, Dietary phytochemicals in cancer prevention and treatment (pp. 3550). New York: Plenum Press.
Bailey, R. G., Nursten, H. E., & McDowell, I. (1991). Comparative
study of the reversed-phase high-performance liquid chromatography of black tea liquors with special reference to the thearubigins. Journal of Chromatography, 542, 115128.
Belitz, H. D., & Grosch, W. (1992). Kakao und Schokolade. In
H. D. Belitz, & W. Grosch, Lehrbuch der Lebensmittelchemie (pp.
869879). New York: Berlin Heidelberg Springer Verlag.
Benavente-Garcia, O., Castillo, J., Marin, F. R., Ortuno, A., & Del
Rio, J. A. (1997). Uses and Properties of Citrus Flavonoids. Journal
of Agricultural and Food Chemistry, 45(12), 45054515.
Biehl, B. (1973). Veranderungen der subcellularen Struktur in Keimblattern von Kakaosamen (Theobroma cacao L.) wahrend der Fermentation und Trocknung. Zeitschrift fuer LebensmittelUntersuchung und-Forschung, 153, 137150.
Bracco, U., Grailhe, N., Rostango, W., & Egli, R. (1969). Analytical
Evaluation of cocoa curing in the Ivory Coast. Journal of the Science of Food and Agriculture, 20, 713717.
Bravo, L. (1998). Polyphenols: chemistry, dietary sources, metabolism,
and nutritional signicance. Nutrition Reviews, 56(11), 317333.
Catapano, A. L. (1997). Antioxidant eect of avonoids. Angiology,
48(1), 3944.
Chung, K.-T., Lu, Z., & Chou, M. W. (1998). Mechanism of inhibition of tannic acid and related compounds on the growth of intestinal bacteria. Food and Chemical Toxicology, 36, 10531060.
Colmenarez, N. G. D., Ramirez-Martinez, J. R., Aldana, J. O.,
Ramos-Nino, M. E., Cliord, M. N., Pekerar, S., & Mendez, B.
(1998). Isolation, characterization and determination of biological
activity of coee proanthocyanidins. Journal of the Science of Food
and Agriculture, 77, 368372.
Cook, N. C., & Samman, S. (1996). Flavonoids chemistry, metabolism, cardioprotective eects, and dietary sources. Journal of
Nutritional Biochemistry, 7, 6676.
Davis, A. L., Lewis, J. R., Cai, Y., Powel, C., Davis, A. P., Wilkins,
J. P. G., Pudney, P., & Cliord, M. N. (1997). A polyphenolic pigment from black tea. Phytochemistry, 46(8), 13971402.
Escribano-Bailon, T., Gutierrez-Fernandez, Y., Rivas-Gonzalo, J. C.,
& Santos-Buelga, C. (1992). Characterization of procyanidins of
Vitis vinifera variety Tinta del Pais grape seeds. Journal of Agricultural and Food Chemistry, 40(10), 17941799.
Fitzpatrick, D. F., Bing, B., & Rohdewald, P. (1998). Endotheliumdependent vascular eects of pycnogenol. Journal of Cardiovascular
Pharmacology, 32, 509515.
Formica, J. V., & Regelson, W. (1995). Review of the biology of
quercetin and related bioavonoids. Food and Chemical Toxicology,
33(12), 10611080.
Fotsis, T., Pepper, M. S., Aktas, E., Breit, S., Rasku, S., Adlercreutz,
H., Waehaelae, K., Montesano, R., & Schweigerer, L. (1997). Flavonoids, dietary-derived inhibitors of cell proliferation and in vitro
angiogenesis. Cancer Research, 57, 29162921.
Gariboldi, E., Mascetti, D., Galli, G., Caballion, P., & Bosiso, E.
(1998). LCUVelectrosprayMSMS mass spectrometry analysis
of plant constituents inhibiting xanthine oxidase. Pharmaceutical
Research, 15(6), 936943.
Ghiselli, A., Nardini, M., Baldi, A., & Scaccini, C. (1998). Antioxidant
activity of dierent phenolic fractions separated from Italian red
wine. Journal of Agricultural and Food Chemistry, 46(2), 361367.

Grayer, R. J. (1989). Flavanoids. In J. B. Harborne, Plant phenolics
(Vol. 1) (pp. 283323). San Diego: Academic Press Inc.
Groundwater, P. W., Solomons, K. R. H., Drewe, J. A., & Munawar,
M. A. (1996). Protein tyrosine kinase inhibitors. In G. P. Ellis and
D. K. Luscombe, Progress in medicinial chemistry (Vol. 33) (pp.
233329). Elsevier Science B.V.
Hagerman, A. E., Riedl, K. M., Jones, G. A., Sovik, K. N., Ritchard,
N. T., Hartzfeld, P. W., & Riechel, T. L. (1998). High molecular
weight plant phenolics (tannins) as biological antioxidants. Journal
of Agricultural and Food Chemistry, 46(5), 18871892.
Hammerstone, J. F., Lazarus, S. A., Mitchell, A. E., Rucker, R., &
Schmitz, H. H. (1999). Identication of procyanidins in cocoa
(Theobroma cacao) and chocolate using high-performance liquid
chromatography/mass spectrometry. Journal of Agricultural and
Food Chemistry, 47(2), 490496.
Han, C. (1997). Screening of anti-carcinogenic ingredients in tea
polyphenols. Cancer Letters, 114, 153158.
Hansen, C. E., del Olmo, M., & Burri, C. (1998). Enzyme activities in
cocoa beans during fermentation. Journal of the Science of Food and
Agriculture, 77, 273281.
Hayes, P. J., Smyth, M. R., & McMourrough, I. (1987a). Comparision
of electrochemical and ultraviolet detection methods in high-performance liquid chromatography for the determination of phenolic
compounds commonly found in beers part 1. Optimisation of
operating parameters. Analyst, 112, 11971204.
Hayes, P. J., Smyth, M. R., & McMourrough, I. (1987b). Comparision
of electrochemical and ultraviolet detection methods in high-performance liquid chromatography for the determination of phenolic
compounds commonly found in beers part 2. Analysis of beers.
Analyst, 112, 12051207.
Herrmann, K. (1994). Bemerkungen zum ``Gerbsto''-Gehalt in
Lebensmitteln. Gordian, 94(12), 1214.
Herrmann, K. (1995). Neuere Erkenntnisse ueber Kakaoinhaltsstoe.
II. Catechine und Procyanidine und deren oxidative Kondensation
sowie Ballaststoe der Kakaoschalen. Gordian, 95(9), 141142, 144.
Hoer, M., Rimpler, H., & Heinrich, M. (1995). Inhibition of intestinal
chloride secretion by proanthocyanidins from Guazuma ulmifolia.
Planta Medica, 61, 208212.
Huang, M.-T., & Ferraro, T. (1992). Phenolic compounds in food and
cancer prevention. In M.-T. Hunag, C.-T. Ho, & C. Y. Lee, Phenolic compounds in food and their eects on health II antioxidants
and cancer prevention (pp. 834). Washington DC: American Chemical Society.
Ioku, K., Tsushida, T., Takei, Y., Nakatani, N., & Terao, J. (1995).
Antioxidative activity of quercetin and quercetin monoglucosides in
solution and phospholipid bilayers. Biochimica et Biophysica Acta,
1234, 99104.
Jalal, M. A. F., & Collin, H. A. (1977). Polyphenols of mature plant,
seedling and tissue cultures of Theobroma cacao. Phytochemistry,
16, 13771380.
Kealey, K. S., Snyder, R. M., Romanczyk, L. J., Geyer, H. M., Myers,
M. E., Withcare, E. J., Hammerstone, J. F., & Schmitz, H. H. (1998).
Cocoa components, edible products having enhanced polyphenol
content, methods of making same and medical uses. Patent Cooperation Treaty (PCT) WO 98/09533, Mars Incorporated, USA.
Kim, H., & Keeney, P. G. (1984). ()Epicatechin content in fermented
and unfermented cocoa beans. Journal of Food Science, 49, 1090
1092.
Kondo, K., Hirano, R., Matsumoto, A., Igarashi, O., & Itakura, H.
(1996). Inhibition of LDL oxidation by cocoa. The Lancet, 348,
1514.
445
Konneh, M., & Caen, J. (1988). Red wine derived compounds and
their putative antiatherogenic properties. In INRA Polyphenols 96
(pp. 105115). Paris.
Kuo, S.-M. (1996). Antiproliferative potency of structurally distinct
dietary avonoids on human colon cancer cells. Cancer Letters, 110,
4148.
Lange, H., & Fincke, A. (1970). Kakao und Schokolade. In L. Acker,
K.-G. Bergner, & W. Diemair, Handbuch der Lebensmittel Band VI:
Alkaloidhaltige Genussmittel, Gewurze, Kochsalz (pp. 210309). New
York: Berlin, Heidelberg Springer Verlag.
Leake, D. S. (1998). Eects of avonoids on the oxidaion of low-density lipoproteins. In C. A. Rice-Evansand, & L. Packer, Flavonoids
in health and disease (pp. 253276). New York, Basel, Hong Kong:
Marcel Dekker Inc.
Lee, H. S., & Widmer, B. W. (1996). Phenolic compounds. In L. M. L.
Nollet, Handbook of food analysis (Vol. 1) (pp. 821894). Basel,
New York, Hong Kong: Marcel Dekker Inc.
Lunte, S. M., Blankenship, K. D., & Read, S. A. (1988). Detection
and identication of procyanidins and avanols in wine by dualelectrode liquid chromatography electrochemistry. Analyst, 113,
99102.
Luo, M., Kannar, K., Wahlqvist, M. L., & O'Brien, R. C. (1997).
Inhibition of LDL oxidation by green tea extract. Lancet, 349, 360
361.
Madigan, D., McMurrough, I., & Smyth, M. R. (1994). Determination of procyanidins and catechins in beer and barley by high-performance liquid chromatography with dual-electrode electrochemical
detection. Analyst, 119(5), 863868.
Makkar, H. P. S. (1989). Protein precipitation methods for quantitation of tannins: a review. Journal of Agricultural and Food Chemistry, 37(4), 11971202.
Makkar, H. P. S., Dawra, R. K., & Singh, B. (1987). Protein precipitation assay for quantitation of tannins: determination of protein in
tanninprotein complex. Analytical Biochemistry, 166, 435439.
Markham, K. R., & Bloor, S. J. (1998). Analysis and identication of
avonoids in practice. In C. A. Rice-Evans, & L. Packer, Flavonoids
in health and disease (pp. 133). New York, Basel, Hong Kong:
Marcel Dekker Inc.
McMurrough, I., & Byrne, J. R. (1992). HPLC analysis of bittering
substances, phenolic compounds, and various compounds of
alcoholic beverages. In L. M. L. Nollet, Food analysis by HPLC (pp.
579641). New York, Basel, Hong Kong: Marcel Dekker Inc.
Middleton, E. J. (1998). Eect of plant avonoids on immune and
inammatory cell function. In J. A. Mantheyand, & B. S. Buslig, Flavonoids in the living system (pp. 175182). New York: Plenum Press.
Minie, B. W. (1989). Chocolate, cocoa and confectionary: Science and
technology. New York: Chapman & Hall.
Newmark, H. L. (1996). Plant phenolics as potential cancer prevention
agents. In N. Back, I. R. Cohen, D. Kritchevsky, A. Lajthaand, &
R. Paoletti, Dietary phytochemicals in cancer prevention and treatment (pp. 2534). New York: Plenum Press.
Opie, S. C., Robertson, A., & Cliord, M. N. (1990). Black tea
thearubigens their HPLC separation and preparation during in-vitro
oxidation. Journal of the Science of Food and Agriculture, 50, 547561.
Osakabe, N., Yamagishi, M., Sanbongi, C., Natsume, M., Takizawa,
T., & Osawa, S. T. (1998). The antioxidative substances in cacao
liquor. Journal of Nutritional Science and Vitaminology, 44, 313321.
Pascual-Teresa, S. D., Gutierrez-Fernandez, Y., Rivas-Gonzalo, J. C.,
& Santos-Buelga, C. (1998a). Characterization of monomeric and
oligomeric avan-3-ols from unripe almond fruits. Phytochemical
Analysis, 9, 2127.
446
Pascual-Teresa, S. D., Treutter, D., Rivas-Gonzalo, J. C., & SantosBuelga, C. (1998b). Analysis of avanols in beverages by high-performance liquid chromatography with chemical reaction detection.
Journal of Agricultural and Food Chemistry, 46(10), 42094213.
Pearson, D. A., Frankel, E. N., Aeschbach, R., & German, J. B.
(1998). Inhibition of endothelial cell mediated low-density lipoprotein oxidation by green tea extracts. Journal of Agricultural and Food
Chemistry, 46(4), 14451449.
Pettipher, G. L. (1986). An improved method for the extraction and
quantitation of anthocyanins in cocoa beans and its use as an index
of the degree of fermentation. Journal of the Science of Food and
Agriculture, 37, 289296.
Plumb, G. W., Pascual-Teresa, S. D., Santos-Buelga, C., Cheynier, V.,
& Williamson, G. (1998). Antioxidant properties of catechins and
proanthocyanidins: eect of polymerisation, galloylation and glycosalation. Free Radical Research, 29, 351358.
Polette, A., Lemaitre, D., Lagarde, M., & Vericel, E. (1996). N-3 fatty
acid-induced lipid peroxidation in human platelets is prevented by
catechins. Thrombosis and Haemostasis, 75(6), 945949.
Porter, L. J., Ma, Z., & Chan, G. (1991). Cacao procyanidins: major
avonoids and identication of some minor metabolites. Phytochemistry, 5, 16571663.
Prieur, C., Rigaud, J., Cheynier, V. & Moutounet, M. (1995). Characterization of grape seed oligomeric and polymeric tannins using thiolysis.
In R. Brouillard, M. Jay, & A. Scalbert, Polyphenols 94; 17th international conference on polyphenols (pp. 155156). Palma de Mallorca,
Spain, 2327 May, 1994Versailles Cedex, France: INRA Editions.
Prochaska, H. J., & Talalay, P. (1992). Phenolic antioxidants as inducers of anticarcinogenic enzymes. In M.-T. Hunag, C.-T. Ho, &
C. Y. Lee, Phenolic compounds in food and their eects on health II
antioxidants and cancer prevention (pp. 150159). Washington, DC:
American Chemical Society.
Rhodes, M. J. C. (1998). Physiological roles of phenolic compounds in
plants and their interactions with microorganisms and humans. In
INRA Polyphenols 96 (pp. 1330). Paris: INRA.
Rice-Evans, C. A., Miller, N. J., & Paganga, G. (1996). Structure
antioxidant activity relationships of avonoids and phenolic acids.
Free Radical Biology and Medicine, 20(7), 933956.
Rigaud, J., Escribano-Bailon, M., Prieur, C., & Souquet, J.-M. (1993).
Normal-phase high-performance liquid chromatographic separation
of procyanidins from cacao beans and grape seeds. Journal of
Chromatography, 654, 255260.
Rigaud, J., Perez-Ilzarbe, J., Ricardo da Silva, J. M., & Cheynier, V.
(1991). Micro method for the identication of proanthocyanidin
using thiolysis monitored by high-performance liquid chromatography. Journal of Chromatography, 540, 401405.
Rohdewald, P. (1998). Pycnogenol. In C. A. Rice-Evans, & L. Packer,
Flavonoids in health and disease (pp. 405417). New York, Basel,
Hong Kong: Marcel Dekker Inc.
Romanczyk, L. J., Hammerstone, J. F., Buck, M. M., Post, L. S.,
Cipolla, G. G., Micceland, C. A., Mundt, J. A., & Schmitz, H. H.
(1997). Cocoa extract compounds and methods for making and
using the same. Patent Cooperation Treaty (PCT) WO 97/36497,
Mars incorporated, USA.
Saito, M., Hosoyama, H., Ariga, T., Kataoka, S., & Yamaji, N.
(1998). Antiulcer activity of grape seed extract and procyanidins.
Journal of Agricultural and Food Chemistry, 46(4), 14601464.
Salah, N., Miller, N. J., Paganga, G., Tijburg, L., Bolwell, G. P., &
Rice-Evans, C. (1995). Polyphenolic avonols as scavengers of
aqueous phase radicals and as chain-breaking antioxidants.
Archieves of Biochemistry and Biophysics, 322(2), 339346.
Samman, S., Lyons Wall, P. M., & Cook, N. C. (1998). Flavonoids

and coronary heart disease: dietary perspectives. In C. A. RiceEvans, & L. Packer, Flavonoids in health and disease (pp. 469481).
New York, Basel, Hong Kong: Marcel Dekker Inc.
Sanbongi, C., Suzuki, N., & Sakane, T. (1997). Polyphenols in chocolate, which have antioxidant activity, modulate immune functions in
humans in vitro. Cellular Immunology, 177, 129136.
Sanbongi, C., Osakabe, N., Natsume, M., Takizawa, T., Gomi, S., &
Osawa, T. (1998). Antioxidative polyphenols isolated from Theobroma cacao. Journal of Agricultural and Food Chemistry, 46(2),
454457.
Schwarze, P. (1958). Phenole und Chinone und die biogene Bildung
von Benzolkernen bei hoeheren Panzen. In W. Ruhland, Encyclopedia of plant physiology (Vol. X) (pp. 507542). Berlin, Goettingen,
Heidelberg: Springer Verlag.
Serani, M., Ghiselli, A., & Ferro-Luzzi, A. (1996). In vivo antioxidant eect of green and black tea in man. European Journal of
Clinical Nutrition, 50, 2832.
Serra Bonvehi, J., & Ventura Coll, F. (1997). Evaluation of bitterness
and astringency of polyphenolic compounds in cocoa powder. Food
Chemistry, 60(3), 365370.
Shahidi, F., & Naczk, M. (1995). Food phenolics: sources, chemistry,
eects, and applications. Lancaster, PA: Technomic Publishing
Company Inc.
Sun, B., Leandro, C., Ricardo da Silva, J., & Spranger, I. (1998).
Separation of grape and wine proanthocyanidins according to their
degree of polymerization. Journal of Agricultural and Food Chemistry, 46(4), 13901396.
Suschetet, M., Siess, M.-H., Le Bon, A.-M. & Canivenc-Lavier, M.-C.
(1998). Anticarcinogenic properties of some avonoids. In INRA,
Polyphenols 96 (pp. 165204). Paris: INRA.
Takahashi, T., Kamiya, T., & Yokoo, Y. (1998). Proanthocyanidins from
grape seeds promote proliferation of mouse hair follicle cells in vitro
and convert hair cycle in vivo. Acta Dermatovenereologica, 78, 428432.
Treutter, D., Santos-Buelga, C., Gutmann, M., & Kolodziej, H.
(1994). Identication of avan-3-ols and procyanidins by high-performance liquid chromatography and chemical reaction detection.
Journal of Chromatography A, 667, 290297.
Vinson, J. A., Hao, Y., Su, X., & Zubik, L. (1998). Phenol antioxidant
quantity and quality in foods: vegetables. Journal of Agricultural
and Food Chemistry, 46(9), 36303634.
Vinson, J. A. (1998). Flavonoids in foods as in vitro and in vivo antioxidants. In J. A. Manthey, & B. S. Buslig, Flavonoids in the living
system (pp. 151164). New York: Plenum Press.
Waterhouse, A. L., Sirley, J. R., & Donovan, J. L. (1996). Antioxidants in chocolate. Lancet, 348, 834.
Watzl, B., & Leitzman, C. (1995). Bioaktive Substanzen in Lebensmitteln. Stuttgart: Hippokrates Verlag.
Williamson, G., Plumb, G. W., Uda, Y., Price, K. R., & Rhodes,
M. J. C. (1996). Dietary quercetin glycosides: antioxidant activity and
induction of the anticarcinogenic phase II marker enzyme quinone
reductase in Hepalclc7 cells. Carcinogenesis, 17(11), 23852387.
Yamane, T., Nakatani, H., Kikuoka, N., Matsumoto, H., Iwata, Y.,
Kitao, Y., Oya, K., & Takahshi, T. (1996). Inhibitory eects and
toxicity of green tea polyphenols for gastrointestinal carcinogenesis.
American Cancer Society, 77, 16621667.
Yang, C. S., Laihshun, C., Lee, M.-J., & Landau, J. M. (1996). Eects
of tea on cacinogenesis in animal models and humans. In N. Back,
I. R. Kritchevsky, D. Kritchevsky, A. Lajtha, & R. Paoletti, Dietary
phytochemicals in cancer prevention and treatment (pp. 5161). New
York: Plenum Press.

Yang, G.-Y., Liao, J., Kim, K., Yurkow, E. J., & Yang, C. S. (1998).
Inhibition of growth and induction of apoptosis in human cancer cell
lines by tea polyphenols. Carcinogenesis, 19(4), 611616.
447
Zumbe, A. (1998). Polyphenols in cocoa: are there health benets?

BNF Nutrition Bulletin, 23, 94102.

Review of Polyphenols in Theobroma Cacao

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Review of Polyphenols in Theobroma Cacao

Caricato da

Copyright:

Formati disponibili

Food Research International 33 (2000) 423447

Review on polyphenols in Theobroma cacao: changes in composition

* Corresponding author. Tel.: +39-0332-785390; fax: +39-0332785930.

these diseases. Most of these conditions are seen as a

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

prevent, or even reverse many conditions from cancer to

Therefore, this review is focused on: (1) the chemistry

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

trolling male sterility and regulating hormonal activity)

sugar residue but others include galactose, rhamnose,

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

Fig. 2. Schematic outline of the biosynthesis and interconnections of avonoids.

Interavanoid linkages are acid labile and yield to

3. Polyphenols in cocoa: alterations in composition and

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

cocoa beans. In smaller amounts, (+)-catechin as well

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

third day of fermentation. Epicatechin and soluble

()-Epicatechin in defatted sample

and the change in colour from slaty over purple to

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

in chocolate manufacture but more common for other

Conching conditions may be modied (shortened) by

127 C roasted nibs

159 C roasted nibs

181 C roasted nibs

IBT, internal bean temperature.

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

content and of a procyanidin pentamer in relation to

ture whereas the ones with a higher polyphenol content

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

Column chromatography (CC) over Sephadex LH-20

extracted only partially, high-molecular polymers are

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

volatility. HPLC is also easily coupled to a variety of

4.3.2. High-performance liquid chromatography

TLC on silica gel using: TAF

Porter et al., 1991

Cocoa beans (cotyledon)

2-D TLC: A: 5% acetic acid;

Jalal and Collin, 1977

Tolueneacetoneformic acid 6:6:1

Colmenarez et al., 1998

Bark from Guazuma ulmifolia

1-D TLC on silica gel:

Hoer, Rimpler and Heinrich, 1995

Beverages; grape seeds; almonds

On silica gel: tolueneacetoneacetic acid

Pascual-Teresa, Treutter, Rivas-Gonzalo

Grape and wine

Tolueneacetoneacetic acid 3:3:1

Flavan-3-ols (in general)

1-D TLC: iso-butanolwateracetic acid

Sun, Leandro, Ricardo da Silva

1-D TLC, one-dimensional TLC; 2-D TLC, two-dimensional TLC.

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

The nature of the solvent such as solvent strength and

which is in case of avonoids mostly around 280 nm

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

quantication and identication of analyte peaks. Time

J. Wollgast, E. Anklam / Food Research International 33 (2000) 423447

a 50 cm 75 mm i.d. uncoated fused silica and the

Treutter et al., 1998; Prieur et al., 1995; Romanczyk et

127 C roasted nibs

159 C roasted nibs

181 C roasted nibs