Innovative Food Science and Emerging Technologies 8 (2007) 390 – 394

www.elsevier.com/locate/ifset

Comparative study on the carotenoid composition of the peel

and the pulp of different citrus species
Attila Agócs ⁎, Veronika Nagy, Zoltán Szabó, László Márk, Róbert Ohmacht, József Deli
Department of Biochemistry and Medical Chemistry, University of Pécs, Medical School, Szigeti u. 12, H-7624, Pécs, Hungary

Abstract

The carotenoid compositions of the peel and the pulp of various citrus fruits were compared with HPLC methods using C18 and C30 columns.
The extracts usually contain β-cryptoxanthin and lutein in considerable amounts and in all species except lime, the red apocarotenoid β-citraurin
as well. In case of lime and mandarin the carotenoid compositions of peel and pulp show a good coincidence while in orange, clementine,
grapefruit, lemon and kumquat there are a lot more differences. Lime extracts contain practically only two carotenoids: β-carotene and lutein. The
carotenoid components of the saponified extracts of kumquat were separated on calcium carbonate columns and were investigated in detail. The
components were identified with HPLC–DAD and HPLC–MS.
© 2007 Elsevier Ltd. All rights reserved.

Keywords: Carotenoids; Citrus fruits; Kumquat; HPLC–MS

Industrial relevance: Citrus fruits are important starting materials for juice production. Their carotenoid fingerprint shows differences not only in different species but
the proportion of certain pigments can be different in the same fruit according to where the plants were grown and how they were processed. Comparison of the
carotenoid content of different fruit products (e.g. juices) can give us useful hints about the quality of the product and about the amount of these important natural
antioxidants.

1. Introduction sinensis), mandarin (Citrus reticulate), clementine (Citrus

Dietary carotenoid antioxidants from fruits and vegetables have
long been known to play an important role in human health. For
their implication in cancer prevention there is a growing interest in
the antioxidant composition of various fruits and vegetables. Citrus
fruits contain a considerable amount of carotenoids and form part
of the daily nutrition of the people. That is why in the recent years a
lot of articles have focused on the investigation of these fruits.
However, these studies focused only on the major carotenoid
components and were restricted to various orange species and
orange juices (Curl & Bailey, 1956; Dhuique-Mayer, Caris-
Veytrat, Ollitrault, Curk, & Amiot, 2005; Gil-Izquierdo, Gil, &
Ferreres, 2002; Lee & Castle, 2001; Meléndez-Martínez, Britton,
Vicario, & Heredía, 2005; Mouly, Gaydou, Lapierre, & Corsetti,
1999; Schlatterer, & Breithaupt, 2005).
Our present paper reports on the HPLC investigation of the
carotenoid composition of fresh citrus fruits (orange (Citrus
reticulata), kumquat (Citrus fortunella), grapefruit (Citrus para-
dise), lime (Citrus aurantifolia) and lemon (Citrus limon)) focus-
ing on the up till now not investigated kumquat. The peel and the
pulp of the fruits were processed and investigated separately.
2. Materials and methods
2.1. Reagents and standards
During our work, analytical grade chemicals were used.
All solvents used in high-performance liquid chromatography
(HPLC) were of HPLC grade. Authentic samples (zeaxanthin,
lutein, β-cryptoxanthin, β-citraurin, β-carotene, α-carotene,
violaxanthin, (9Z)-violaxanthin, mutatoxanthin) were taken
from our collection.
2.2. Plant materials

⁎ Corresponding author. Tel.: +36 72 536 001/1864; fax: +36 72 536 225. The citrus fruits originated from different countries and were
E-mail address: attila.agocs@aok.pte.hu (A. Agócs). purchased in a Tesco and a Spar supermarket in Pécs (Hungary).
1466-8564/$ - see front matter © 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ifset.2007.03.012
 A. Agócs et al. / Innovative Food Science and Emerging Technologies 8 (2007) 390–394 391

Table 1
Carotenoid composition (%) of different citrus peels
Peak Carotenoid Orange Mandarin Clementine Kumquat Grapefruit Lemon
1 Unidentified 0.6 0.4 0.2 0.8 1.7 0.5
2 Unidentified 1.4 0.3 0.8 0.8 1.0 0.5
3 Unidentified 2.2 0.6 0.6 0.2 1.1 0.3
4 Unidentified 1.2 0.4 0.7 0.2 5.8 2.5
5 Neochrome 1.6 3.1 5.7 1.5 2.2 4.3
6 Violaxanthin + furanoid 13.6 8.7 1.4 9.8 8.9 0.2
7 Unidentified – 2.1 0.7 – – –
8 β-Citraurin 10.8 13.3 28.0 16.6 9.3 4.0
9 (Z)-β-citraurin 0.9 1.0 2.1 1.4 1.5 0.1
10 (Z)-β-citraurin 1.0 1.2 2.5 1.0 0.7 0.1
11 Luteoxanthin 2.9 1.6 6.1 3.7 2.1 0.5
12 (9Z)-violaxanthin 33.8 18.0 7.9 16.9 6.4 1.6
13 (13Z)-violaxanthin 1.4 0.3 0.2 1.7 1.7 0.1
14 Lutein 6.6 5.4 4.1 5.5 7.6 8.3
15 Unidentified – – – 2.1 – –
16 Unidentified 0.8 0.2 0.6 0.8 1.3 1.3
17 Unidentified 0.8 0.3 1.0 0.8 1.3 1.0
18 Monofuranoid like (λmax = 450, 423 nm) 0.9 0.6 0.4 1.1 0.7 1.4
19 Cryptochrome 3.5 2.8 2.8 5.7 4.3 1.4
20 Diepoxide like 0 1.4 1.9 0 0 7.4
21 Unidentified 0.4 0.4 1.6 0.8 1.2 10.9
22 Monoepoxide like (λmax = 475, 450 nm) – – – 1.2 1.5 3.1
23 Monofuranoid like (λmax = 450, 423 nm) 2.1 4.5 3.7 3.8 2.1 7.9
24 α-Cryptoxanthin 0.3 0.3 0.2 0.2 1.4 1.5
25 β-Cryptoxanthin 3.5 23.4 13.4 11.4 11.3 19.9
26 (Z)-β-cryptoxanthin 1.2 3.9 4.0 2.8 1.6 3.6
27 ξ-Carotene 1.2 1.8 0.3 0.9 7.5 7.2

2.3. Pigment extraction
To obtain reliable samples 300–500 g (fresh weight) of fruits
were used for extraction. Peel and juice sacs were separated and
cut into small pieces. The samples were extracted three times
with methanol and then twice with diethylether. The extracts
were combined and evaporated and saponified in ether with
30% KOH/MeOH at room temperature. The saponified
pigments were stored in benzene solution at − 20 °C, under
nitrogen and under exclusion of light until preparation of HPLC
samples. General methods including sample taking, extraction,
workup and quantitative determination of chlorophyll were
described in detail in a previous study of yellow paprika (Matus,
Deli, & Szabolcs, 1991). To avoid pigment decomposition, cis-
trans isomerisation and epoxide–furanoid oxide rearrangement,
the isolation of carotenoids was carried out under nitrogen in
darkness using methanol for dehydration at low temperatures
(4–23 °C).
2.4. High-performance liquid chromatography
HPLC system I for C18 column: Dionex pump model 580,
Hewlett Packard 1050 detector with HP ChemStation software
and Waters-991, photodiode array detector; column 250 ×
4.6 mm i.d., Licrospher C18 (Merck), 5 μm endcapped; gradient
elution with A (12% (v/v) H2O in methanol), B (methanol), C
(30% (v/v) dichloromethane in methanol), i.e. 0–2 min: 100%
A; 2–10 min: to 80% A/20% B; 10–18 min: to 50% A/50% B;
18–25 min: to 100% B; 25–27 min: 100% B; 27–34: to 100%
C; 34–41 min 100% C (linear steps), flow rate 1.25 ml/min.
HPLC system II for C30 column: Dionex pump model P 580
NDG, Dionex 340S diode array detector and Chromeleon soft-
ware; column: 250 × 4.6 mm i.d. Kovasil C30, 5 μm, endcapped;
eluent: A: 10% (v/v) H2O in methanol, B: tert-butyl methyl
ether. Gradient program: 0–35 min 8% B–25% B, 35–50 min
25% B–50% B, 50–60 min 50% B–70% B, 60–70 min 70%B,
70–75 min 70% B–8% B (regeneration), flow rate: 1 ml/min.
MS: Finnigan AQA, APCI + polarity without splitting;
Crown U: 4 kV, Temp: 400 °C, full scan mode, data acquisition
rate: 0.2 scans/s; AQA max voltage: 30 V, mass filter: 10, 20,
30 V.
2.4.1. Identification of the peaks
The total carotenoid contents were determined by UV–VIS
methods. Carotenoids were identified on the basis of the same
retention times, same spectral characteristics with standards and
co-chromatography with authentic samples. The peaks in a
chromatogram were identified with the help of authentic
carotenoid samples, various chemical tests (Baranyai, Matus,
& Szabolcs, 1982; Matus et al., 1991) and the UV–VIS spectra
of the individual peaks. Photodiode array measurements of
spectral properties for the individual peaks (from 300 to
510 nm) were made at the upslope, apex and downslope of the
curves. The matching of the three spectra indicated the degree of
peak purity.
2.4.2. Quantification
The chromatograms were evaluated quantitatively by
relating the heights of the individual carotenoids to that of
canthaxanthin using an internal standard (Matus et al., 1991).
392 A. Agócs et al. / Innovative Food Science and Emerging Technologies 8 (2007) 390–394

Table 2
Carotenoid composition (%) of the pulp of citrus fruits
Peak Carotenoid Orange Tangerine Clementine Kumquat Grapefruit Lemon
1 Unidentified 1.6 0.2 0.5 0.8 – –
2 Unidentified 1.2 0.2 0.8 0.8 – –
3 Unidentified 1.1 0.3 0.6 0.5 – –
4 Unidentified 0.3 0.2 0.9 0.4 – –
5 Unidentified 3.8 0.3 0.6 2.4 – –
6 Unidentified 0.4 0.5 1.2 1.1 – –
7 Neochrome 2.0 0.8 2.4 0.2 – –
8 β-Citraurin 2.0 2.1 2.0 7.1 – –
9 Auroxanthin 0.8 0.4 0.3 0.7 – –
10 Auroxanthin 0.5 0.4 0 1.1 – –
11 Luteoxanthin 11.2 3.6 3.2 0.6 – –
12 Luteoxanthin + (9Z)-violaxanthin 14.2 8.2 7.2 0.6 – –
13 Mutatoxanthin 3.4 2.0 1.4 5.7 – –
14 Mutatoxanthin 5.1 2.3 1.7 4.5 – –
15 Unidentified 1.0 0.4 0.1 1.3 – –
16 Lutein 14.9 11.6 8.3 10.6 12.0 5.7
17 Monofuranoid like (λmax = 450, 423 nm) 0.5 1.6 2.0 1.0 – –
18 Monofuranoid like (λmax = 450, 423 nm) 0.6 1.1 2.7 3.3 – –
19 Monofuranoid like (λmax = 450, 423 nm) 0.8 1.8 2.6 3.4 – –
20 α-Cryptoxanthin 1.5 0.8 0.9 0.6 – –
21 β-Cryptoxanthin 13.1 30.6 28.4 26.4 4.3 36.0
22 (Z)-β-cryptoxanthin 2.4 5.4 7.5 5.2 1.5 8.7
23 ξ-Carotene 1.3 3.0 5.7 3.3 – –
24 Unidentified – – – – 26.7 4.6
25 α-Carotene – – – – 4.8 1.2
26 β-Carotene – – – – 19.5 11.1

2.5. Open column chromatography (CC) In the pulp of citrus fruits, the main carotenoid is β-
cryptoxanthin (∼ 30%). Lutein and (9Z)-violaxanthin were also
For open column chromatography on calcium carbonate detected in larger amounts (Fig. 2). Lime extract contains
(Biogal, Hungary) hexane:toluene (5–20% toluene) mixtures practically only two main carotenoids: β-carotene (36.8%) and
were used as eluent. lutein (29.7%). As minor components β-cryptoxanthin (5.3%),
α-carotene (4.2%) and (Z)-luteins (3.2%) could also be
3. Results and discussion identified.
There are a lot of publications about orange species and
3.1. Comparative study of carotenoid composition of citrus different juices (Bonnacorsi, Verzera, Trozzi, & Zappalá, 2003;
species on C18 column Cortés, Frígola, & Torregosa, 2004; Curl & Bailey, 1956;
Dhuique-Mayer et al., 2005; Gil-Izquierdo et al., 2002; Lee &
The extracts of the peels and the pulps usually contain
β-cryptoxanthin, lutein and (9Z)-violaxanthin in considerable
amounts and in all species except lime the red apocarotenoid
β-citraurin as well. In case of lime and mandarin the carot-
enoid compositions of peel and pulp show a good coincidence
while in orange, clementine, grapefruit, lemon and kumquat
there are a lot more differences (Tables 1 and 2, Fig. 1).
In the peels, the main carotenoids are (9Z)-violaxanthin
(8–33%), β-citraurin (11–28%), and β-cryptoxanthin (3–23%)
(Fig. 3). The lutein content in the peel of all citrus species is
4–8%. Some furanoids (neochrome, luteoxanthin, auroxanthin)
and Z-isomers were also detected. In the apolar region of chro-
matograms we found some peaks (peak 20–22) which could
be identified tentatively. The UV–VIS spectra of these peaks
and the retention times suggest that these compounds may be the
5,6- and 5,8-epoxides of β-cryptoxanthin. The extract of lime
Fig. 1. Structure of some carotenoids. Antheraxanthin: R_e, Q_c;
peel, whose colour is green, contains mainly chloroplast pig-
mutatoxanthin: R_f, Q_c; auroxanthin: R_Q_f; neochrome: R_g, Q_f;
ments: lutein (46%), (9Z/9′Z)-lutein (3.4%), (13Z/13′Z)-lu- α-carotene: R_a, Q_b; neoxanthin: R_g, Q_e; β-carotene: R_Q_a;
tein (1.8%), β-cryptoxanthin (3.4%), (Z)-β-cryptoxanthin violaxanthin: R_Q_e; β-cryptoxanthin: R_c, Q_a; zeaxanthin: R_Q_c;
(0.9%), α-carotene (8.5%) and β-carotene (24.2%). β-citraurin: R_c, Q_CHO; zeacarotene: R_a, Q_h; lutein: R_c, Q_d.
 A. Agócs et al. / Innovative Food Science and Emerging Technologies 8 (2007) 390–394 393

Fig. 2. HPLC separation of carotenoids in orange pulp (a) and kumquat pulp (b) on C18 column. For peak numbers see Table 2.

Castle, 2001; Meléndez-Martínez, Vicario, & Heredía, 2003;
Meléndez-Martínez et al., 2005; Mouly et al., 1999; Schlatterer
& Breithaupt, 2005). In our case the carotenoid fingerprints
were similar to those published in these works although we
could always detect β-citraurin (2–7%) in the citrus pulps
(Table 2) in significant amount.
3.2. Detailed investigation of kumquat
The kumquat is quite widespread in the Mediterranean area
and its carotenoid content has not been investigated up till now
so we set ourselves the aim to investigate it and compare its
carotenoid composition with other citrus fruits. Calcium car-
bonate column chromatography and HPLC–MS on a C18 and a
C30 column were used to isolate and identify the extracts. The
peel extract is available in larger amount, the pulp contains
carotenoids only in a very small quantity. Thus, only the peel
extract was processed further. The HPLC study of kumquat
peel extract on C18 phase shows that β-citraurin (16.6%), (9Z)-
violaxanthin (= violeoxanthin) (16.9%), β-cryptoxanthin
(11.4%) and violaxanthin (9.8%) are the main components.
Cryptochromes (5.7%), lutein (5.5%), luteoxanthin (3.7%),

Fig. 3. HPLC separation of carotenoids in orange peel (a) and kumquat peel (b) on C18 column. For peak numbers see Table 1.
394 A. Agócs et al. / Innovative Food Science and Emerging Technologies 8 (2007) 390–394
(Z)-β-cryptoxanthin (2.8%), (Z)-β-citraurins (2.4%), (13Z)-
violaxanthin (1.7%), neochrome (1.5%), ξ-carotene (0.9%)
and α-cryptoxanthin (0.2%) were detected as minor compo-
nents (Fig. 3).
For the detailed investigation, first the peel extract of
kumquat was partitioned between hexane (epiphase) and 10%
water in methanol (hypophase). The epiphase contains volatile
oils and β-cryptoxanthin (37%) beside some minor carotenoid
components. The epiphase of kumquat peel was submitted to
CC (CaCO3, toluene/hexane) and the fractions were afterwards
submitted to repeated CC. β-Cryptoxanthin was isolated and
identified by its retention time, UV–VIS spectrum and by co-
chromatography with authentic sample.
Polar carotenoids were enriched in the hypophase which were
separated by successive calcium carbonate column chromatog-
raphy. During the first chromatography eight fractions were
separated (fraction 1–8). The fractions were investigated on C30
phase by HPLC–DAD and HPLC–MS techniques. In fraction 1
(13Z)-violaxanthin (λmax: 458, 432, 411, 332 nm; [M]+: 600),
neochrome I (λmax: 445, 419, 398 nm; [M]+: 600), (9Z)-
violaxanthin (λmax: 461, 434, 411 nm; [M]+: 600), luteoxanthin
(λmax: 441, 415, 394 nm; [M]+: 600) and (9Z)-antheraxanthin
(λmax: 465, 439 nm; [M]+: 584), were detected by HPLC.
Fraction 2 mainly contained violaxanthin (λmax: 466, 418,
414 nm; [M]+: 600) and some minor components so this fraction
was separated by further chromatography. Fraction 3 is a mixture
of antheraxanthin (λmax: 470, 443 nm; [M]+: 584), β-citraurin
(λmax: 460 nm; [M]+: 432), cis-β-citraurins (λmax: 445, 419 nm;
[M]+: 432), β-citraurin-epoxide (λmax: 460 nm; [M]+: 448) and
lutein (λmax: 470, 443 nm; [M]+: 568). Fraction 4 contained pure
lutein (94% purity). Fraction 5 had two main components, lutein
and zeaxanthin, this fraction was purified further by chroma-
tography. Fraction 6 gave pure β-citraurin (90% purity), fraction
7 contained mainly β-cryptoxanthin as well as fraction 8 where
cis-β-cryptoxanthins could also be identified. From fraction 2
pure violaxanthin (98% purity, λmax: 483, 453, 427 nm in
benzene) and pure (9Z)-violaxanthin (97% purity, λmax: 481,
451, 424 nm in benzene) were isolated in two successive
chromatographic steps. From fraction 5 lutein (92% purity, λmax:
487, 457, 433 nm in benzene) and zeaxanthin (96% purity, λmax:
493, 463 nm in benzene) could be isolated in a single
chromatographic step. By repeated open column chromatogra-
phy violaxanthin, lutein, zeaxanthin and β-citraurin could be
isolated in 90–99% purity.
4. Conclusions
Our results demonstrate that all citrus species except lime
have similar carotenoid profile although there are some dif-
ferences in the proportion of the pigments. The apocarotenoid
β-citraurin (and its isomers) proved to be a characteristic
carotenoid in citrus species and could be isolated in pure form
among other more widespread pigments. Kumquat being sim-
ilar to orange has the same pigment constituents, but while in
oranges (9Z)-violaxanthin predominates its major component
is β-cryptoxanthin.
Acknowledgements
We thank Mrs. Andrea Bognár and Mr. Norbert Götz for
their skillful assistance. This study was supported by the grants
OTKA K 60121 and T 46476 and by DSM Nutritional Products
(Basel).
References
Baranyai, M., Matus, Z., & Szabolc, J. (1982). Determination, by HPLC, of
carotenoids in paprika products. Acta Alimentaria, 11, 309−323.
Bonnacorsi, I., Verzera, A., Trozzi, A., Zappalá, M., Dugo, P., & Mondell, L.
(2003). Carotenoid profile of sweet orange and mandarin essential oils.
Italian Journal of Food Sciences, 15, 133−139.
Cortés, C., Esteve, M. J., Frígola, A., & Torregos, F. (2004). Identification and
quantification of carotenoids including geometrical isomers in fruit and
vegetable juices by liquid chromatography with ultraviolet-diode array
detection. Journal of Agricultural and Food Chemistry, 52, 2203−2212.
Curl, L., & Bailey, G. F. (1956). Orange carotenoids. Part I. Comparison of
carotenoids of Valencia orange peel and pulp. Journal of Agricultural and
Food Chemistry, 4, 156−162.
Dhuique-Mayer, C., Caris-Veytrat, C., Ollitrault, P., Curk, F., & Amiot, M.
(2005). Varietal and interspecific influence on micronutrient contents in
citrus from the Mediterranean area. Journal of Agricultural and Food
Chemistry, 53, 2140−2145.
Gil-Izquierdo, A., Gil, M. I., & Ferreres, F. (2002). Effect of processing
techniques at industrial scale on orange juice antioxidant and beneficial
health compounds. Journal of Agricultural and Food Chemistry, 50,
5107−5114.
Lee, H. S., & Castle, W. S. (2001). Seasonal changes of carotenoid pigments and
color in Hamlin, Earlygold, and Budd Blood orange juices. Journal of
Agricultural and Food Chemistry, 49, 877−882.
Matus, Z., Deli, J., & Szabolc, J. (1991). Carotenoid composition of yellow
pepper during ripening: Isolation of α-cryptoxanthin-5,6-epoxide. Journal
of Agricultural and Food Chemistry, 39, 1907−1914.
Meléndez-Martínez, A. J., Britton, G., Vicario, I. M., & Heredía, F. J. (2005).
Color and carotenoid profile of Spanish Valencia late ultrafrozen orange
juices. Food Research International, 38(8–9), 931−936.
Meléndez-Martínez, A. J., Vicario, I. M., & Heredía, F. J. (2003). A routine
high-performance liquid chromatography method for carotenoid determi-
nation in ultrafrozen orange juice. Journal of Agricultural and Food
Chemistry, 51, 4219−4224.
Mouly, P. P., Gaydou, E. M., Lapierre, L., & Corsetti, J. (1999). Differentiation
of several geographical origins in single-strength Valencia orange juices
using quantitative comparison of carotenoid profiles. Journal of Agricultural
and Food Chemistry, 47, 4038−4045.
Schlatterer, J., & Breithaup, D. E. (2005). Cryptoxanthin structural isomers in
oranges, orange juice, and other fruits. Journal of Agricultural and Food
Chemistry, 53, 6355−6361.