Trends in Food Science & Technology 13 (2002) 361–371
Review
Chlorophyll and
carotenoid analysis
in food products.
Properties of the
pigments and
methods of analysis
Benoı̂t Schoefs
Laboratoire de Plasticité et Expression des Génomes
Microbiens, CNRS- FRE 2383, Université Joseph
Fourier, 260 Rue de la Piscine, BP 53, F-38041 Grenoble
Cedex 09, France (tel.: +33-4-76-63-56-64;
fax: +33-4-76-63-56-63
e-mail: benoit.schoefs@ujf-grenoble.fr)
Chlorophylls and carotenoids are very common pigments,
which give colour to vegetables and several fruits. Because
of their colour and their physico-chemical properties, they
are also used as additives to food products. Both pigment
types are fragile and can be easily altered or destroyed,
modifying the perception and the quality of the products.
To control the pigment content of food products, analytical
processes, including nondestructive methods have been
developed. The aim of this paper is to give a brief overview
on these methods.
# 2002 Elsevier Science Ltd. All rights reserved.
Introduction
‘Plant pigment’ is a generic expression used to desig-
nate a large number of coloured molecules. On the basis
of their chemical structure, they can be classed into 5
* Tel.: +33 4 76 63 56 64; fax: +33 4 76 63 56 63.
0924-2244/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 92 4 - 2 24 4 ( 0 2 ) 0 0 18 2 - 6
families i.e. tetrapyrroles (e.g. chlorophyll), carotenoids
(e.g. b-carotene), flavonoids (e.g. anthocyanins), phe-
nolic compounds (e.g. teaflavin) and N-heterocyclic
compounds (e.g. betalains). The number of members
belonging to each family is so high that several volumes
would be necessary to describe their particular proper-
ties and methods of analysis. Therefore, this review is
limited to the analytical methods of chlorophylls and
carotenoids.
Chlorophylls and carotenoids are frequent organic
food components because they are naturally present in
plants, giving their specific colouration. In vivo, these
pigments play key roles in photosynthesis.
Several reports have demonstrated that plant pig-
ments play important roles in health (Mayne et al.,
1994; Franceschi et al., 1994). Neither chlorophylls nor
carotenoids can be synthesized by animal tissues,
though animal cells can chemically modify them for
assimilation. Thus, these molecules must be obtained
from food. Carotenoid and chlorophyll molecules are
also extracted and used as natural colorants and anti-
oxidants (Burton & Ingold, 1984; Palozza & Krinsky,
1991; Britton, 1995) to restore the natural level of these
molecules in food products or to prepare fortified pro-
ducts. They also can be chemically modified before
being incorporated into food products (e.g. glycosyla-
tion of crocetin (Dufresne et al., 1999)), replacement of
Mg2+ by Cu2+ in chlorophyll. Chlorophyll-precursors
and chlorophyll-derivatives are also used in medicine
for photodynamic treatments.
It is necessary to have analytical procedures that are
able to identify rapidly and precisely, and quantify the
pigments present in food products. Addressing these
issues is never straightforward. With some foods, like
bottled waters, analysis is very simple. Other foods are
highly complex mixtures, requiring care with process
and storage conditions to ensure that the end product
has unchanged level(s) of pigments, colour and taste. In
fact, chemical modifications of the pigments can affect
their perception and/or their properties. This is espe-
cially important in the case of fortified food, which cre-
ates nutritional expectation from both the health
authorities and the consumers. The development of
powerful analytical methods is also of prime importance
in the quality control of food products. Obviously, this
requires a precise knowledge of the pigment composi-
tion of original products (see Minguez-Mosquera,
362 B. Schoefs / Trends in Food Science & Technology 13 (2002) 361–371
Homero-Mendez, & Gandul-Rojas, 1995; Mercandante,
Britton, & Rodriguez-Amaya, 1998). In this contribu-
tion, the molecular structures and the general properties
of chlorophylls and carotenoids are described first,
followed by a review of invasive and noninvasive tech-
niques used for pigment analysis. Several examples of
analytical procedures are presented elsewhere (Schoefs,
2003).
Molecular structures, and spectroscopic and
chemical properties of carotenoids, chlorophylls
and chlorophyll-related pigments
Carotenoids
The basic structure of a carotenoid molecule is a
symmetrical tetraterpene skeleton formed by tail-to-tail
linkages of two geranylgeranyl diphosphate molecules
(C20 unit). This basic skeleton is illustrated by the
acyclic carotenoid lycopene (Fig. 1), the pigment, which
gives the red colour to tomatoes. The end-groups of the
basic skeleton are modified into six-membered rings
yielding dicyclic carotenoid like b-carotene, the most
ubiquitous carotenoid (Fig. 1). The carotenoid group of
molecules can be divided into two families of com-
pounds: the carotenes, which are devoid of oxygen
function (like b-carotene) and the xanthophylls, which
contain at least one oxygen function (like lutein). The
chemical structure clearly indicates that carotenoids are
lipophilic compounds, and, therefore, water-insoluble,
except when very strongly polar groups, such as poly-
saccharides, esterify the carotenoid backbone. A good
example is crocetin glycosyl esters, pigments that give
the colour to saffron strigmata and gardenia fruits
(Pfander & Witter, 1975; Pfister, Meyer, Steck, & Pfander,
1996). For a comprehensive description of the chemical
properties of carotenoids, the reader is refered to recent
reviews (e.g. Britton, Liaaen-Jensen, & Pfander, 1995).
Substituted carbon–carbon double bonds can exist in
two different configurations depending on the relative
arrangement of the four substituents. In the case of the
carotenoids, a double bond is normally designated as cis
Fig. 1. Structure of an acyclic carotenoid (lycopene) and of a dicyclic carotene (b-carotene).
or trans on the basis of the disposition of the two
heaviest substituents, i.e. those that are the continuing
parts of the main polyene chain. As illustrated in Fig. 2,
when the two substituents R1 and R2 are located on the
opposite sides of the axis of the double bond, the
configuration is trans; when they are on the same side of
the axis, the configuration is cis.
The absorption of light by carotenoids is due to the
presence of conjugated carbon-carbon double bonds in
which the electrons can be considered as belonging to
this system as a whole rather than to an individual
atom. This series of conjugated double bonds is called
the chromophore. When light is absorbed by a mole-
cule, the whole energy of the photon is transfered to it.
The absorbing molecule is lifted from its normal state of
lowest energy (ground state) to an energy-rich state
(excited state). According to Bohr’s theory of molecular
structure, a molecule can exist only in a series of discrete
states of electronic energies that correspond to the
absorbance bands of the absorbance spectrum (for a full
description, see Rabinovich, 1956; Goedheer, 1966;
Rabinovich & Govindjee, 1969). A typical absorbance
spectrum of a carotenoid contains three bands, the
maxima of which are functions of the chromophore
lengths (Fig. 3; see Britton & Young, 1993; Britton,
1995). In certain cases, the first and/or third peaks are
difficult to observe. From the comparison of the absor-
bance spectrum obtained with different carotenoids, it
was concluded that every modification in the circuit of
electrons along the conjugated bonds is reflected in the
absorbance spectrum of the pigment and, therefore, can
be detected by using spectroscopic methods. In contrast,
the chemical modifications that do not affect the chro-
mophore are not seen in the absorbance spectrum of the
pigment.
Molecules in an excited state can de-excite through
several ways. If the molecule is well protected from
interaction with other molecules, the simpliest pathway
for it to loose its excitation energy is to emit a photon.
This is called fluorescence. This de-excitation pathway is
 B. Schoefs / Trends in Food Science & Technology 13 (2002) 361–371
Fig. 2. Scheme of cis- and trans-double bonds.
Fig. 3. Changes in the absorbance spectrum of carotenoids as a
function of the chromophore length (C: x-carotene, N: neuro-
sporene, and L: lycopene).
in competition with other pathways, i.e. internal con-
version and transition to the triplet state. In the case of
carotenoids, the fluorescence intensity is weak (Koyama
& Hashimoto, 1993) due to the fact that the main de-
excitation pathway occurs through the transition to the
triplet state, which is nonfluorescent (see Rabinovich &
Govindjee, 1969).
Chlorophylls and related pigments
The name ‘chlorophyll’ was initially given to the green
pigment involved in photosynthesis in 1818 (Pelletier &
Caventou, 1818) but the correct chemical formulas for
the chlorophylls, describing them as magnesium com-
plexes devoid of iron and phosphorus, were only
obtained by Willstätter and Stöll (1913). The detailed
understanding of the structure of chlorophyll resulted
from the studies performed by Fisher and collaborators
(Fischer & Orth, 1937; Fischer & Stern, 1940) who were
363
the first to delineate the structure of the porphyrin ring.
Since these pioneer works, more than 50 different
chlorophyll-related molecules have been reported
(reviewed by Scheer, 1996). Here I will focus our atten-
tion to chlorophyll and some related molecules, which
potentially may be found in foodstuffs.
Basically chlorophyll molecules are conjugated tetra-
pyrroles, to which a cyclopentanone ring, conjoint with
ring III, has been added (=a phorbin, see Fig. 4). The
macrocycle is planar. All naturally occurring chlor-
ophylls have a propionic acid residue at position 17.
The position 173 is generally esterified with a long chain
alcohol, usually phytol. Chlorophyll b differs from
chlorophyll a by the presence of an aldehyde residue
instead of a methyl group at position 7 (Fig. 4). Chlor-
ophyll a and chlorophyll b are the most abundant pig-
ments in terrestrial plants and in green algae.
Chlorophyll b can be considered as typical from these
organisms. The inner seed coat from Cucurbitaceae,
although green, does not contain chlorophyll molecules
but protochlorophyll. This pigment differs from chlor-
ophyll because the pyrrole IV is not reduced (Fig. 4), i.e.
protochlorophyll belongs to the porphyrin group of
compounds. Brown algae and diatoms contain pigments
similar to protochlorophyll, named chlorophyll c, and
do not belong to the chlorophyll-group of molecules. In
addition the nonreduced ring IV of chlorophyll c is
esterified at position 17 by an acrylic residue — instead
of a propionic side group, the terminal carboxylic group
being generally not esterified (structure not presented).
From the structures, it is clear that a chlorophyll mole-
cule is made up of a hydrophilic part, the macrocycle,
and by a hydrophobic part, the phytol chain. The most
hydrophilic segment of the macrocycle are the cyclo-
pentanone ring and the propionic ester group (position
17). Therefore, nonesterified macrocycles are much
more polar than the esterified ones. The nature of the
other side groups as well as their configuration also
modify the polarity of the molecules (reviewed by Seely,
1966).
As with carotenoid, chlorophylls and related molecules
present a closed circuit of conjugated double bonds, which
allows them to absorb light (Fig. 4). The absorbance
spectra of chlorophyll a in an organic solvent is presented
in Fig. 5. The spectrum indicates the presence of two
distinct bands: one in the red, corresponding to the first
excited state and another one in the blue, corresponding
to the second excited state. A similar spectrum is
obtained with chlorophyll b and protochlorophyll but
differs in the relative (to the intensity in the blue region)
absorbance intensity in the red and by the position of
the bands. This varies not only from pigment to pig-
ment but also as a function of the environment.
In contrast to carotenoid, chlorophyll and chlor-
ophyll-related molecules can de-excite by emitting
fluorescence and only from the first excited state. Here
364 B. Schoefs / Trends in Food Science & Technology 13 (2002) 361–371
Fig. 4. Structure of chlorophyll a, chlorophyll b and protochlorophyll.
the higher singlet excited state de-excites initially by
internal conversion to the first excitation state which, in
turn, de-excites by emitting fluorescence. Consequently,
a fluorescence spectrum contains only one band (Fig. 5).
It should be stressed again that chemical modifications
that do not affect the chromophore, like trans-ester-
ification at position 173, are not reflected in either
absorbance or fluorescence spectra.
It is important to recognize that plant tissues show the
colour of the predominant pigment but may contain
also many other coloured molecules. These will only
appear when the dominating pigment disappears, such
as seen in autumn; for example the degradation of
chlorophyll molecules (Henry, Houghton, & Brown,
1987) unmasks the carotenoids.
Regardless of their origin — in situ or extracted —
chlorophyll and carotenoid are fragile molecules and
easily modified. These modifications can significantly
alter their colour, the commercial value and also the
nutritive quality. A change in colour may elicit a nega-
tive impression. Grapefruit juice during processing can
become brick red, a colour associated by consumers,
even if wrongly (Arena, Fallico, & Maccarone, 2000),
with pigment degradation, an observation that results in
low product acceptance. Conversely, in certain cases, a
change in colour is not only desirable but could even
enhance acceptance of the foodstuffs. The most striking
example of a positive change in colour is that which
occurs when lobsters are cooked. During heating, the
astaxanthin (red) is released from the naturally occuring
astaxanthin–protein complexes of the lobster’s original
blue shell.
Methods of analysis: a brief overview
An array of methods can be used for chlorophyll and
carotenoid analysis. The choice is usually guided by the
type of information needed. Generally, the pigment
must be extracted before analysis, sometimes from a
complex matrix. Therefore, efficient extraction and
analytical protocols are requested.
 B. Schoefs / Trends in Food Science & Technology 13 (2002) 361–371
Fig. 5. Room temperature (A) absorbance and (F) fluorescence
spectra of chlorophyll a in acetone.
General considerations
As both chlorophyll and carotenoid molecules have a
long chain of conjugated double bonds they react easily
with acid, base, oxygen and light (Anguelova & War-
thesen, 2000). The pigment solution containing chlor-
ophyll reacts easily with oxygen upon illumination
resulting in the formation of activated oxygen species.
These molecules are very reactive and able to oxidize
other organic molecules, including lipids and proteins.
For this reason, care should be taken during extraction
Fig. 6. Room temperature absorbance spectrum of olive oil.
365
and analysis. The production of singlet oxygen by
chlorophyll is not restricted to aqueous pigment extracts
but also occurs in less polar solvents like oil (Anguelova
& Warthesen, 2000; Haila, Nielsen, Heinonen, & Skibs-
ted, 1997). For this reason, oil containing a high
amount of chlorophyll or related compounds should be
stored in the dark and at reduced temperature (Jung
& Min, 1991). Because carotenoids are active oxygen
species quenchers (Goulson & Warthesen, 1999), special
care should be taken with products having a low con-
tent in carotenoids e.g. pumpkin seed oil (Schoefs,
2001a).
Spectroscopic methods
The absorbance spectrum reflects the organization of
the conjugated double bond system and constitutes the
fingerprint of pigments (see above). Absorbance spec-
troscopy appears to be the simpliest way to identify
major pigments present in a mixture (Schoefs, 2000).
Once identified, it is possible to use a set of equations to
estimate their respective concentration (reviewed by
Bertrand & Schoefs, 1999). This is illustrated in Fig. 6,
which displays the absorbance spectrum of olive oil.
The comparison of this spectrum with those of pure
pigments allows the identification of the pigment famil-
ies composing the oil i.e. carotenoids and pheophytin.
Because the environment of the pigment (e.g. solvent,
temperature, ligation to protein, etc.) strongly influence
the position and the shape of the spectrum, a crude
absorbance spectrum is useless for direct measurements
of the pigment concentration. Measurements require
pigment extraction with a solvent, for which equations
366 B. Schoefs / Trends in Food Science & Technology 13 (2002) 361–371
have been established or, at least, in which specific (or
molar) absorbance coefficients have been determined.
With these data, it is always possible to establish a new
equation set, adapted to the particular situation. The
precision of the method depends on the type of device
used, the ability to determine with precision the absor-
bance maxima and, of course, on the accuracy of the
absorption coefficient used for the calculation. It is
therefore advisable to check regularly the literature for
new values and equation sets. Identification of pigments
on the sole basis of an absorbance spectrum, however,
presents strong limitations.
The overlapping of the absorbance bands of the pig-
ments present complicates the estimation of individual
pigment concentration and is much less efficient espe-
cially if the number of pigments is higher than three.
The absorbance spectrum of a full extract is of little
use when there is a dominance of one pigment or if
several compounds have similar spectra. Fluorescence
spectroscopy can be more diagnostically helpful due to
the selective excitation of a number of chlorophylls and
chlorophyll-related pigments but useless in case of car-
otenoids, since their fluorescence is very weak (see
above). The limitation of the spectroscopic methods is
illustrated by the example presented in Fig. 7, which
displays the room temperature absorption spectrum of a
green beverage. The product ingredient label indicated
nothing about the green colour; one would expect that
the green colour to be regular chlorophyll a. Several
maxima could be observed. The peaks at 413 and 634
nm could belong to chlorophyllin, pheophytin and pro-
tochlorophyll(ide), whereas the one at 437 nm could
reflect the presence of few amounts of chlorophyll a.
Therefore additional analyses are required to precise the
nature of the green pigment(s) (see below).
For completeness, it is necessary to add that the cis-
carotenoid isomers can be recognized because the
absorbance spectrum presents an additional band in the
Fig. 7. Room temperature absorbance spectrum of a green
beverage.
UV-region. The position of the cis-double bound is
reflected in the ratio Acis/AMAX.
In conclusion, spectroscopic methods usually permit a
crude identification of the pigments present in an
extract, but specific composition remains obscure. To
resolve a pigment mixture, chromatographic methods
should be used.
Chromatographic separation
Open column
Various phases like powdered sucrose, DEAE-
sepharose, cellulose or MgO/Hyflosupercel have been
used to achieve chlorophyll and carotenoid separation
(Wasley, Scott, & Holt, 1970; Strain, Cope, & Svec,
1971; Rodriguez-Amaya, Raymundo, Lee, Simpson, &
Chichester, 1976; Omata & Murata 1983). The all-trans-
b-carotene can be separated from 9-cis- and 13-cis-iso-
mers using an open column filled up with calcium
hydroxide.
Thin-layer chromatography
Thin-layer chromatography (TLC) and high-perfor-
mance TLC (HPTLC) are often used for the separation
and isolation of individual classes of molecules as TLC
is fast, effective and relatively cheap. Special care should
be taken when the separation is done on silica gel: it is
necessary to previously neutralize the acidity of silica in
advance in order to avoid epoxide-furanoxide rearran-
gement in carotenoids (Schiedt & Liaaen-Jensen, 1995)
and chlorophyll-pheophytinization. Unfortunately,
separation of compounds with a similar structure is
usually rather difficult. This is well illustrated by the
comparison of the chromatograms obtained of pumpkin
seed oil pigments separated by TLC and by HPLC (Fig.
8). The TLC method allows the separation of 4 bands,
whereas 14 peaks were obtained with HPLC. One can
see that TLC methodology can be supplemented by
more efficient separation techniques, such as HPLC
coupled with UV–vis and/or fluorometric detection.
High-performance liquid chromatography
Chlorophylls and carotenoids have a clear hydro-
phobic character and analysis by C18 reversed-phase
(RP) columns is prefered. More recently, a C30-RP
came on the market and is particularly suitable for the
separation of carotenoids as the interactions of the pig-
ments and the stationary phase are maximized by the
similar size. The efficiency of such a column is very high,
allowing the discrimination of several cis-isomers of the
same carotenoid (Emenhiser, Sander, & Schwartz,
1999). It has been successfully applied to determination
of saponified carotenoids in orange juice (Rousseff,
Raley, & Hofsommer, 1996). However, when a mixture
is complex, coelutions become rapidly limiting. A less
selective stationary phase, like the C18-RP, is preferably
used for less detailed analysis. The mobile phase is
 B. Schoefs / Trends in Food Science & Technology 13 (2002) 361–371
Fig. 8. Chromatogram of the pigments from pumpkin seed oil. Upper
pannel: TLC — (petroleum ether/acetone 80/20 v:v). Bottom panel:
HPLC — [methanol, acetonitrile, methylene chloride, see (35)].
usually made up of organic solvents, except when very
polar molecules like glycosyl esters of carotenoid have
to be separated. In this later case, a polar organic sol-
vent, mixed with a small amount of water is recom-
mended (Dufresne et al., 1999).
To improve pigment separation, heating the column is
sometimes proposed. This procedure is not however
recommended because it can trigger carotenoid iso-
merization, as well as epimerization and allomerization of
chlorophyll molecules. These modifications are not
detected using usual HPLC methods (Hyvärinen &
Hynninen, 1999). Care should be taken to employ proper
chromatography conditions to ensure that pigments do
not escape analyses. This is illustrated when the analysis
of pumpkin seed oil using two different reversed-phase
HPLC methods, both of which are recommended for the
analysis of plant pigments, were compared (Schoefs,
1999).
Supplementing HPLC with a diode-array detector
together with the availability of powerful computers
has especially increased the analytical power of HPLC.
Using such detectors it is now possible to follow simul-
taneously the elution on the full UV–vis range (190–
367
800 nm). This guarantees that each pigment can be
followed at its absorbance maximum, i.e. with max-
imum sensitivity.
Supercritical and subcritical HPLC
Supercritical fluid extraction has been suggested as an
alternative method for selective isolation of carotenoids
in one step and conditions in which pigment degrada-
tions are avoided see Pfander & Niggli (1995). For
instance, Favati, King, Friedrich, and Eskins (1988)
isolated b-carotene and lutein from leaf concentrates at
40
C and pressure ranging between 29 and 70 MPa.
From the analytical point of view, supercritical fluid
extraction is compatible with supercritical fluid chro-
matography since the two techniques can share the
mobile phase and same devices, favoring the develop-
ment of extraction and separation methodologies.
Because the elution strength of pure CO2 in respect of
the carotenoids is rather weak, it is necessary to add
cosolvents to increase carotenoid solubility in CO2. The
solubility parameter theory predicts that the maximum
solubility of a compound is reached when the solubility
parameter of the solvents equals that of the solute. For
instance, b-carotene can be extracted at 43
C and
approximately 70 MPa (Favati et al., 1988). However,
some types of equipment do not allow elution at this
elevated pressure. New conditions of pressure and tem-
perature should be found using a diagram representing
the variation of pressure versus density for different
temperatures (Smith, 1988). Through a lower tempera-
ture, densities similar to those reached at higher tem-
peratures, can be obtained for lower pressures, the
densities corresponding to subcritical conditions. Using
this technique, Ibanez, Lopez-Sebastian, Tabera, and
Reglero (1998) successfully separated b-carotene from
lycopene in less than 10 min. The best separation was
obtained at 10
C under a pressure of approximately 35
MPa, with a home-packed capillary column containing
deactivated silonal groups with an octamethyl-cyclo-
tetrasiloxane reagent. The structure is broken at high
temperature, forming O–Si–CH3 bonds with the silica
phase.
None of the methods described above is entirely sui-
table for an exhaustive separation of a complex mixture
of pigments. The tetrapyrrole fingerprint is often suffi-
cient to identify its chromophore, it does not contain
enough information to determine the complete structure
of the pigment (e.g. Schoefs, 2000). This observation
can be partly deduced from chromatographic behaviour
and from the comparison of the obtained retention data
with literature. Other methods like mass spectrometry
(MS) will have to be applied to characterize fully the
structure of the molecule (Schoefs, 2001b). Recent
developments in MS that allow the analysis of sub-
nanogram quantities, have made this technique attrac-
tive for foodstuff research.
368 B. Schoefs / Trends in Food Science & Technology 13 (2002) 361–371
Analysis of tetrapyrroles by MS
Chlorophylls have long presented special analytical
challenges to MS, because of their high mass, low vola-
tility and thermal instability. Thus, chlorophyll a and
chlorophyll b have given only pyrrolytic fragments in
electron impact (EI) MS (Waller, 1974). This could be
explained by the exceptional stability of the p-electron
system of porphyrin rings and the strong magnesium-
oxygen coordination. Similarly, pheophytin has only
one weak molecular ion signal in EI-MS. Introduction
of desorption-ionization MS in the 1970s and 1980s,
such as chemical ionization, secondary ion MS, fast-
atom bombardment, field, plasma and recently matrix
assisted laser desorption, opened ways to molecular ion
detection and, thus, to direct molecular weight determi-
nation. In all these methods, the necessity of sample
vaporization prior to ionization is avoided. Rather, the
desorption and ionization processes occur directly from
the condensed phase. For details on desorption ioniza-
tion techniques, the interested reader is referred to the
comprehensive review by Hunt and Michalski (1991).
When the matrix contains salts, addition of alkali
metal(s) can also be observed. Occurrence of additional
signals in the lower mass region reflects the presence of
chlorophyll degradation products, like 10-OH chlor-
ophyll a (m/z 908) or pheophytin a (m/z 870) and signals
at m/z 482, m/z 556 and m/z 615. However, it is difficult
to decide whether they were present in the sample before
the analysis or if they are a result of sample preparation
and/or analysis techniques. Regardless of the origin of
these degradation products, the determination of their
molecular mass can be used to build up a fragmentation
scheme of the target compound. It is obvious that such
schemes can be very useful for the determination of the
structure of chlorophyll, especially degradation pro-
ducts, such as change in the esterification by fatty acids
at the C172 position and chlorophyll halogenation at the
C20 position (see Fig. 4) (Hunt & Michalski, 1991).
Most recent progress in MS analysis of tetrapyrroles
has been obtained with the development of atmospheric
ionization methods, i.e. atmospheric pressure chemical
ionization (APCI) and electrospray ionization (ESI).
The APCI technique, in combination with RP-HPLC,
has proved to be efficient in detecting chlorophyll a and
its nine degradation products at low nanogram levels.
The procedure is approximately 1000 times more sensi-
tive than thermospray ionization (Eckhardt, Kelly, &
Maxwell, 1991).
To demonstrate the potential of electrospray MS in
chlorophyll research, an ESI interface was employed
with an ion trap mass spectrometer as mass analyser
(Schoefs, 2001b). ESI is a mild ionization technology
feature, allowing the detection of chlorophyll a-proto-
nated molecular ion (M+H)+, m/z 893.5, which dom-
inates and is accompanied only partly by a molecular
radical ion (M+. m/z=892.5). It is remarkable that the
spectrum was obtained with approximately 2 pmol of
the compound. Further structural information has been
obtained with the unique MSN capability of the ion trap
analyser, allowing consecutive dissociation of side chain
functional groups from the selected precursor ion (up to
8 steps; Schoefs, 2001b). The MSN procedure is highly
selective and enables consecutive cleavage of at least
seven functional groups around the porphyrin, provid-
ing valuable structural information about the tetra-
pyrrole. Using MS methods, chlorophyll allomers and
their derivatives, produced during fruit and vegetable
processing (Minguez-Mosquera, Gallardo-Guerrero,
Homero-Mendez, & Garrido-Fernandez, 1995), have
been isolated and their structure elucidated (Hyvärinen
& Hynninen, 1999).
All the methods described above are invasive techni-
ques that consume sample, time and money. Alter-
native, noninvasive procedures of analysis have been
developed to follow pigment content in food products
and also to estimate its impact on product perception.
Noninvasive methods
Evaluation of colour and its perception
The visual impression is mainly a function of the pig-
ments present but is also affected by morphological
factors like epidermal hair or cuticular waxes, the shape
and the orientation of the cells in the epidermis and the
sub-epidermis. In fact, pigments and surface topo-
graphy selectively absorb, reflect and refract the incident
visible light sensed by the eye. The signals, generated at
the retina level, are transduced through the optic nerve
to the brain and interpreted as colour. When light
strikes the eye, it is detected by one of the three colour
sensors of the retina : a red, a green, or a blue. It is
known that the information is not sent as individual
colour but as a red/green signal, a yellow/blue signal, or
a black/white signal. When all the wavelengths are
reflected by an object, the eyes see it as white, whereas
when all are absorbed, the object appears black.
It is out of the scope of this review to explain in detail
the basis of colour measurement (reviewed by Wyszecki
& Stiles, 1987). In brief, each colour can be described by
a set of three parameters i.e. hue, the dominant shade;
saturation (or chroma), how much colour is present,
and lightness, the degree of darkness of a particular
colour. Therefore, it is necessary to describe a different
hue, saturation and lightness for each unique set of
illuminant or observed conditions. By using a standard
illuminant and a standard observer, the amount of light
reflected by an object can be converted into the hue,
saturation and lightness values. Additionally, a sample
can be compared to any standard with these three attri-
butes. In 1976 the CIE adopted a standard method of
calculating colour attributes, known as the CIE Lab
Colour Space. The lightness coefficient L* ranges from
black (=0) to white (=100) while the coordinates a*
 B. Schoefs / Trends in Food Science & Technology 13 (2002) 361–371
and b* designate the colour on a rectangular-coordinate
grid perpendicular to the L* axis. The colour at the grid
origin is achromatic (i.e. grey; a*=0, b*=0). On the
horizontal axis, positive and negative a values indicate
the hue of redness (positive values) and greenness
(negative values), whereas on the vertical axis, b* indi-
cates yellowness (positive values) and blueness (negative
values). For instance, change in colour was used to
characterize the modifications of the orange-red colour
of fish flesh fed with various diets (Bjerkens et al., 1997;
Akhtar, Gray, Cooper, Garling, & Booren, 1999) or
to follow the yellowness of semolina during milling
(Borrelli, Troccoli, Di Fonzo, & Fares, 1999). In both
cases, the decrease in colour was correlated to a mod-
ification of the carotenoid content. There are conflicting
reports in the literature on the correlation between
colour measurement and pigment composition. This is
especially true when more than one pigment has to be
monitored. Several mathematical combinations of the
parameters are used to predict pigment modifications
occuring during food processing (e.g. Steet & Tong,
1996; Gomez, Pardo, Navarro, & Varon, 1998; Ma &
Shimokawa, 1998). Some authors also use the hue
angle (Wyszecki & Stiles, 1987) to characterize colour
modifications (Akhtar et al., 1999; De Ell, van Kooten,
Prange, & Murr, 1999). Other methods, based on
reflected and scattered light, were used to derive
estimated pigment content and to predict colour
(McClements, Chantopromchal, & Clydesdale, 1998).
Determination of the freshness of vegetables
The freshness of vegetables can be assessed using
chlorophyll fluorescence kinetic measurements, which
reflect photosynthetic activity (Kautsky, Appel, &
Amann, 1960). It is out of the scope of this review to
explain in detail the background of the development of
this method and the interested reader is referred to
recent reviews (Rohacek & Bartak, 1999). In brief, the
chlorophyll fluorescence yield depends on the redox
state of the photosystem II primary acceptor — QA.
When all the QA are oxidized, for instance after a period
of complete darkness, the level of fluorescence, denoted
F0, is minimum. In constrast, when all the QA are
reduced, for example during a saturating light pulse, the
level of fluorescence, denoted Fm, is maximum. The
Table 1. Plants containing relatively high amount of particular non commercially available carotenoids
Pigment Latin name
Cucurbitaxanthin Cucurbita maxima Pumpkin
30 ,40 -cis-Glycolic and trans-isomers of idoxanthin – –
Violaxanthin, lutein-5,6-epoxide
Neoxanthin, violaxanthin and luteoxanthin Urtica sp.
Flavoxanthin Chrysanthemum sp. Chrysanthemum flowers
Neochrome
369
state of the photosynthetic apparatus can be estimated
using the Fv/Fm ratio with Fv=FmF0. Healthy plants
exhibit a Fv/Fm ratio of approximately 0.8. When the
photosynthetic process is inactivated, such as in senes-
cent plants or in cold-stored plants, the ratio decreases
(De Ell et al., 1999; De Ell & Toivonen, 1999). Tian,
Woolf, Bowen, and Fergusson (1996) showed that the
ratio could be a sensitive indicator of responses of
broccoli to hot water treatment, even before visual
changes were noted. Other parameters can be obtained
from chlorophyll fluorescence kinetic to characterize
the physiological state of the plants (Lichtenthaler &
Rinderle, 1988).
Pigment identification and quantification
Definitive pigment identification and quantification
require usually the development of standards, which are
critical to food pigment analysis. Those standards that
are not commercially available need to be prepared
from scratch. Table 1 lists some interesting sources for
such a purpose.
Schiedt et al. (1995) have defined the minimum cri-
teria for the identification of carotenoids:
 the absorbance spectra in the UV-Vis region,
obtained in at least two different solvents, should
be in agreement with the chromophore suggested;
 the chromatographic properties must be identical
in TLC (Rf) and HPLC (t0 R), and be co-eluted
with authentic samples ;
 a mass spectrum should be obtained that allows
at least confirmation of the molecular mass.
Although such ‘rules’ were not specified for the iden-
tification of chlorophyll molecules and chlorophyll-
related compounds, similar criteria are suggested.
Commonly the retention values of pigments before
and after treatment are compared and at times are
found to exceed 100% after treatment. Such a
paradox, which was frequently reported in blanched
and cooked vegetables and fruits (Stahl & Sies, 1992),
has been attributed to the greater extractability of the
pigments after cooking. It is believed that heating
triggers cell-wall rupture facilitating release of the
pigments.
Common name Reference
Matsuno, Tani, Maoka, and Komori (1986)
Aas, Bjerkeng, Halten, and Storebakken (1987)
Mango fruit Razungles, Babic, Sapis, and Bayonove (1996)
Nettles Razungles et al., 1996
Razungles et al., 1996
Green grapes Razungles et al., 1996
370 B. Schoefs / Trends in Food Science & Technology 13 (2002) 361–371
Future trends
In food analysis, continued study of various techni-
ques offers the best possibility of finding the most
powerful method for monitoring pigment separation
for high quality and safety of food products. For
instance, the versatility of HPLC and MS as analytical
tool offers an almost endless number of applications
for food analysis. The versatility of HPLC techniques
also make it an ideal procedure for use by analytical
quality control and research and development labora-
tories in the food and beverage industry. Objective,
physicochemical methods to determine the quality of
food products are often destructive and time-consum-
ing. Simultaneous development of rapid and non-
destructive indicators of quality must be continued. In
establishing these tools, it is necessary to correlate phy-
sical parameters e.g. firmness, soluble solids, dry matter,
colour with physiological data (e.g. respiration and
photosynthesis). These methods could involve those
techniques such as delayed luminescence (Triglia, La
Malfa, Musumeci, Leonardi, & Scordino, 1998), near
infrared spectroscopy (Slaughter, Barett, & Boersig,
1996), and etc. which were, in the past, restricted to
basic research.
Acknowledgements
The author thanks Dr. Lauro (The Natural Color
Resource Center, California State Polytechnic Uni-
versity) for its critical reading of the text. BS is Research
Associate at the Centre National de la Recherche
Scientifique (France).
References
Aas, G. H., Bjerkeng, B., Halten, B., & Storebakken, T. (1987). Idox-
anthin, a major carotenoid in the flesh of Artis charr (Salvelinus
alpinus) fed diets containing astaxanthin. Aquaculture, 150, 135–
142.
Akhtar, P., Gray, J. I., Cooper, T. H., Garling, D. L., & Booren, A. M.
(1999). Dietary pigmentation and deposition of a-tocopherol
and carotenoids in rainbow trout muscle and liver tissue. Journal
of Food Science, 64, 234–239.
Anguelova, T., & Warthesen, J. (2000). Degradation of lycopene,
a-carotene, and b-carotene during lipid peroxidation. Journal of
Food Science, 65, 71–75.
Arena, E., Fallico, B., & Maccarone, E. (2000). Influence of car-
otenoids and pulps on the color modification of blood orange
juice. Food Science, 65, 458–460.
Bertrand, M., & Schoefs, B. (1999). Photosynthetic pigment
metabolism in plants under stress. In M. Pessarakli (Ed.),
Handbook of plant and crop stress (pp. 527–543). Marcel
Dekker.
Bjerkens, B, Refstie, S., Fjalestad, K. T., Storebakken, T., Rodbotten,
M., & Roem, A. J. (1997). Quality parameters of the flesh of
Atlantic salmon (Salmon salar) as affected by dietary fat content
and full-fat soybean meal as partial substitute for fish meal in
the diet. Aquaculture, 157, 297–309.
Borrelli, G. M., Troccoli, A., Di Fonzo, N., & Fares, C. (1999).
Durum wheat lipoxygenase activity and other quality
parameters that affect pasta colo. Cereal Chemistry, 76,
335–340.
Britton, G. (1995). UV/visible spectroscopy. In G. Britton, S. Liaaen-
Jensen, & H. Pfander (Eds.), Carotenoids, Vol 1B: Spectroscopy
(pp. 13–62). Birkhäuser.
Britton, G., Liaaen-Jensen, S., & Pfander, H. (1995). Carotenoids. Vol
1A: Isolation and Analysis. Birkhäuser
Britton, G., & Young, A. J. (1993). Methods for isolation and analysis
of carotenoids. In G. Britton, & A. J. Young (Eds.), Carotenoids in
photosynthesis (pp. 407–457). Chapman & Hall.
Burton, G. W., & Ingold, K. U. (1984). b-carotene: an usual type of
lipid antioxidant. Science, 244, 569–573.
De Ell, J. R., & Toivonen, P. M. A. (1999). Application of chlorophyll
fluorescence as an indicator of physiological changes in cold-
stored broccoli after transfer to room temperature. Journal of
Food Science, 64, 501–503.
De Ell, J. R., van Kooten, O., Prange, R. K., & Murr, D. P. (1999).
Application of chlorophyll fluorescence techniques in post-
harvest physiology. Horticultural Revue, 23, 69–107.
Dufresne, C., Cormier, F., Dorion, S., Niggli, U. A., Pfister, S., &
Pfander, H. (1999). Glycosylation of encapsulated crocetin by a
crocetin by a Crocus sativus L. Cell culture. Enzyme and Micro-
bial Technology, 24, 453–462.
Eckardt, C. B., Keely, B. J., & Maxwell, J. R. (1991). Preservation of
chlorophyll-derived pigments in sedimentary organic matter.
Journal of Chromatography, 557, 271–278.
Emenhiser, C., Sander, L. C., & Schwartz, S. J. (1999). Capability of
polymeric C30 stationary phase to resolve cis-trans carotenoid
isomers in reversed-phase liquid chromatography. Journal of
Chromatography, 707A, 205–216.
Favati, F., King, J. W., Friedrich, J. P., & Eskins, K. (1988). Supercritical
carbon dioxide extraction of carotene and lutein protein con-
centrates. Journal of Food Science, 53, 1532–1535.
Fisher, H., & Orth, H. (1937). Die chemie des pyrrols, Vol. 2. Part 1.
Akademie Verlagsges
Fisher, H., & Stern, A. (1940). Die chemie des pyrrols, Vol. 2. Part 2.
Akademie Verlagsges
Franceschi, S., Biddi, E., Vecchia, C., Talamini, R., D’Avanzo, B., &
Wegni, E. (1994). Tomatoes and risk of digestive-tract cancers.
International Journal of Cancer, 59, 181–184.
Goedheer, J. C. (1966). Visible absorption and fluorescence chlor-
ophyll and its aggregates in solution. In L. P. Vernon, & G. R. Seely
(Eds.), The chlorophylls (pp. 147–184). Academic Press.
Gomez, R., Pardo, J. E., Navarro, F., & Varon, R. (1998). Colour dif-
ference in paprika papper varieties (Capsicum annuum L.) culti-
vated in greenhouse and in the open air. Journal of Science and
Food Agriculture, 77, 268–272.
Goulson, M. J., & Warthesen, J. J. (1999). Stability and antioxidant
activity of b-carotene and high oleic canola oil. Journal of Food
Science, 64, 996–999.
Haila, K. M., Nielsen, B. R., Heinonen, M. I., & Skibsted, L. H. (1997).
Carotenoid reaction with free radicals in acetone and toluene at
different oxygen partial pressures. Food Research and Technol-
ogy, 204, 81–87.
Henry, G. A. F., Houghton, J. D., & Brown, S. B. (1987). The degra-
dation of chlorophyll-a biological enigma. New Phytologist, 107,
255–302.
Hunt, J. E., & Michalski, T. J. (1991). Desorption-ionization mass
spectrometry of chlorophylls. In H. Scheer (Ed.), Chlorophylls
(pp. 835–853). CRC Press.
Hyvärinen, K., & Hynninen, P. H. (1999). Liquid chromatographic
separation and mass spectrometric identification of chlorophyll
b allomers. Journal of Chromatography, 837A, 107–116.
Ibanez, E., Lopez-Sebastian, S., Tabera, J., & Reglero, G. (1998).
Separation of carotenoids by subcritical fluid chromatography
with coated, packed capillary columns and neat carbon dioxide.
Journal of Chromatography, 823A, 313–319.
 B. Schoefs / Trends in Food Science & Technology 13 (2002) 361–371
Jung, M. Y., & Min, D. B. (1991). Effects of quenching mechanisms of
carotenoids on the photosensitized oxidation of soybean oil.
Journal of the American Oil Chemists Society, 68, 653–658.
Kautsky, H. A., Appel, W., & Amann, H. (1960). Chlorophyll-
fluoreszenz und kohlensureassimilation XIII. Mitteilung Die fluor-
escenzkurve und die photochemie der pflanzen. Biochemische
Zeitschrift, 332, 277–292.
Koyama, Y., & Hashimoto, H. (1993). Spectroscopic studies of car-
otenoids in photosynthetic systems. In G. Britton, & A. J. Young
(Eds.), Carotenoids in photosynthesis (pp. 327–408). Chapman &
Hall.
Lichtenthaler, H. K., & Rinderle, U. (1988). Chlorophyll fluorescence
signatures as vitality indicator in forest decline research. In H.
K. Lichtenthaler (Ed.), Application of chlorophyll fluorescence in
photosynthesis research, stress physiology, hydrobiology and
remote sensing (pp. 143–151). Kluwer Academic Publishers.
Ma, X-W., & Shimokawa, K. (1998). In vitro studies in ethylene-
enhanced chlorophyll degrading peroxidase and its reaction
products in banana (Musa sapientum L.) fruits. Journal of the
Japanese Society for Horticultural Science, 67, 3–15.
Matsuno, T., Tani, Y., Maoka, T., Matsuo, K., & Komori, T. (1986).
Isolation and structural elucidation of cucurbitaxanthin a and b
from pumpkin Cucurbita maxima. Phytochemistry, 25, 2837–
2840.
Mayne, S. T., Janerich, D. T., Greenwald, P., Chorost, S., Tucci, C.,
Zaman, M. B., Melamed, R., Kiely, M., & McKneally, M. F. (1996).
Dietary b-carotene and lung cancer risk in U.S. nonsmokers.
Journal of the National Cancer Institute, 86, 33–38.
McClements, D. J., Chantopromchal, W., & Clydesdale, F. (1998).
Prediction of food emulsion color using light scattering theory.
Journal of Food Science, 63, 935–939.
Mercandante, A. Z., Britton, G., & Rodriguez-Amaya, D. B. (1998).
Carotenoids from yellow passion fruits (Passiflora eludis). Journal
of Agricultural and Food Chemistry, 46, 4102–4106.
Minguez-Mosquera, M. I., Gallardo-Guerrero, L., Homero-Mendez,
D., & Garrido-Fernandez, J. (1995). Involvement of copper and
zinc ions in green staining of table olives of the variety Gordal.
Journal of Food Protein, 58, 564–569.
Minguez-Mosquera, M. I., Homero-Mendez, D., & Gandul-Rojas, B.
(1995). Detection of bixin, lycopene, canthaxanthin, and
b-apo-80 -carotenal in product. JAOAC International, 78, 491–
496.
Omata, T., & Murata, N. (1983). A rapid and efficient method to
prepare chlorophyll A and B from leaves. Plant and Cell Physiol-
ogy, 22, 1093–1100.
Palozza, P., & Krinsky, N. I. (1991). The inhibition of radical-initiated
peroxidation of microsomal lipids by both a-tocopherol and b-
carotene. Free Radical Biology and Medecine, 11, 407–414.
Pelletier, F., & Caventou, J. B. (1818). Sur la matière verte des
feuilles. Annales de Chimie et de Physique, 9, 194–196.
Pfander, H., & Niggli, U. (1995). Chromatography. Part V: Super-
critical-fluid chromatography. In G. Britton, S. Liaaen-Jensen, &
H. Pfander (Eds.), Carotenoids, vol 1A: isolation and analysis
(pp. 191–198). Birkhäuser.
Pfander, H., & Witter, F. (1975). Carotenoid glycosides. 2. Carotenoid
content of safran. Helvetica Chimica Acta, 58, 1608–1620.
Pfister, S., Meyer, P., Steck, A., & Pfander, H. (1996). Isolation and
structure elucidation of carotenoid-glycosyl esters in gardenia
fruits (Gardenia jasminoides Ellis) and saffron (Crocus sativus
Linné). Journal of Agricultural and Food Chemistry, 44, 2612–
2615.
Rabinovich, E. I. (1956). Photosynthesis and related processes, Vol. 2,
Part 2. Interscience Publishers.
Rabinovich, E. I., & Govindjee, ? (1969). Photosynthesis. John Wiley
& Sons.
Razungles, A. J., Babic, I., Sapis, J. C., & Bayonove, C. L. (1996). Par-
ticular behavior of epoxyxanthophylls during veraison and
371
maturation of grape. Journal of Agricultural and Food Chemistry,
44, 3821–3825.
Rodriguez-Amaya, D. B., Raymundo, L. C., Lee, T. C., Simpson, K. L.,
& Chichester, C. O. (1976). Carotenoid pigment changes in
ripening Monordica charantia fruits. Annals of Botany, 40, 615–
624.
Rohacek, K., & Bartak, M. (1999). Technique of the modulated
chlorophyll fluorescence: basic concepts, useful parameters, and
some applications. Photosynthetica, 37, 339–363.
Rousseff, R., Raley, L., & Hofsommer, H-J. (1996). Application of
diode array detection with a C30 reversed phase column for the
separation and identification of saponified orange juice car-
otenoids. Journal of Agricultural and Food Chemistry, 44, 2176–
2181.
Scheer, H. (1996). Tetrapyrroles in photosynthesis. In B. Schoefs, F.
Franck, & J. Aghion (Eds.), Bulletin de la Société Royale des Sci-
ences de Liège, 65, 310–321.
Schiedt, K., & Liaaen-Jensen, S. (1995). Isolation and analysis. In
G. Britton, S. Liaaen-Jensen, & H. Pfander (Eds.), Carotenoids.; Vol
1A: isolation and analysis (pp. 81–108). Birkhäuser.
Schoefs, B. (1999). The importance of spectrography and
chromatography for food quality conservation. In: Proceedings
of the 5th European Symposium on Food Authenticity.
pp. 1–21.
Schoefs, B. (2000). The pigment composition of pumpkin seed oil.
In J. Francis (Ed.), Proceedings of the 4th international symposium
on natural colourants (pp. 85–92). SCI Publishers.
Schoefs, B. (2001a). Analyse der Farbstoffe in Kürbiskernöl. Lebens-
mittel and Biotechnologie, 2001, 8–11.
Schoefs, B. (2001b). Determination of tetrapyrrole pigments in
foodstuffs. International Laboratory, 33, 34–39.
Schoefs, B. (2003). Chlorophyll and carotenoid analysis in food
products. A case by case. Trends in Analytical Chemistry (in
press).
Seely, G. R. (1966). The structure and chemistry of functional
groups. In L. P. Vernon, & G. R. Seely (Eds.), The chlorophylls
(pp. 77–109). Academic Press.
Slaughter, D. C., Barrett, D., & Boersig, M. (1996). Nondestructive
determination of soluble solids in tomatoes using infrared spec-
troscopy. Journal of Food Science, 61, 695–697.
Smith, R. M. (1988). Supercritical fluid chromatography. London: RSC
Chromatography Monographs.
Stahl, W., & Sies, H. (1992). Uptake of lycopene and its geome-
trical isomers is greater from heat-processed that from
unprocessed tomato juice in humans. Journal of Nutrition, 122,
2161–2166.
Steet, J. A., & Tong, C. H. (1996). Degradation kintics of green color
and chlorophylls in peas by colorimetry and HPLC. Journal of
Food Science, 61, 924–931.
Strain, H. H., Cope, B. T., & Svec, W. A. (1971). Analytical proce-
dures for the isolation, identification, estimation, and identifi-
cation of the chlorophylls. Methods in Enzymology, 23A, 452–
476.
Tian, M. S., Woolf, A. B., Bowen, J. H., & Fergusson, I. B. (1996).
Changes in colour and chlorophyll fluorescence of broccoli
florets following hot water treatment. Journal of the American
Society for Horticultural Science, 121, 310–313.
Triglia, A., La Malfa, G., Musumeci, F., Leonardi, C., & Scordino, A.
(1998). Delayed luminescence as an indicator of tomato fruit
quality. Journal of Food Science, 63, 512–515.
Waller, G. R. (1974). Biochemical Application in mass spectrometry.
Wiley.
Wasley, J. W. F., Scott, W. T., & Holt, S. (1970). Chlorophyllides c.
Canadian Journal of Botany, 24, 376–383.
Willstätter, R., & Stöll, A. (1913). Untersuchungen über Chlorophyll.
Springer Verlag.
Wyszecki, G., & Stiles, W. S. (1987). Colour science. Wiley & Sons.