SciELO - Scientific Electronic Library Online

Home Pagealphabetic serial listing  

Services on Demand




Related links



Print version ISSN 0006-8705On-line version ISSN 1678-4499

Bragantia vol. 57 n. 1 Campinas  1998 









The oil content was determined in seeds of several continental African species of the coffee germplasm bank of Instituto Agronômico de Campinas, State of São Paulo, Brazil. Oil was extracted from seeds with hexane in Soxhlet apparatus. Due to the economic importance, C. arabica and C. canephora have been the best studied species concerning oil content and composition, and the results obtained are in agreement with the reported in the literature. On the other hand, only one report in the literature describes the results of oil analyses in other few species of the African continent, although it does not allow comparison with our results. The oil content of most of the species varied from 9 to 15%, therefore, similar to the range observed for C. arabica and C. canephora. The exception was C. salvatrix, with 29% of oil in the seeds.

Index terms: African coffee species, Coffea, oil content.




O teor de óleo foi determinado nas sementes de várias espécies de café do continente africano, presentes no banco de germoplasma do Instituto Agronômico (IAC), as quais foram extraídas com hexano em extrator Soxhlet. Em vista de seu valor econômico, as espécies C. arabica e C. canephora têm sido as mais estudadas no que diz respeito ao conteúdo de óleo e sua composição, e nossos resultados são concordantes com aqueles da literatura. Por outro lado, apenas um trabalho publicado relata análises de óleo em outras poucas espécies do continente africano, não permitindo total comparação com nossos dados. De modo geral, os valores obtidos para a maioria das espécies situaram-se na faixa 9 e 15%, semelhante, portanto, à encontrada para C. arabica e C. canephora. A exceção foi C. salvatrix, com 29% de óleo nas sementes.

Termos de indexação: Coffea, espécies africanas, café, teor de óleo.



Most of the knowledge on coffee bean lipids is based on the species Coffea arabica (Arabica), because it is the most economically important among several others. For the same reason, considerable information is also available to Coffea canephora (Robusta), the second most world widely cultivated coffee specie.

Lipids are important components of coffee beverage and aroma, although the majority is lost with the grounds during the preparation (Folstar, 1985). Lipids are expelled with roasting to the bean surface, forming a layer which may trap volatile aromas, impairing the immediate loss of these compounds (Clifford, 1985; Arnaud, 1988). Coffee oil is also obtained from roasted coffee and from the spent grounds of roasted and ground coffee after brewing or from the instant coffee industry. The most important commercial use for this oil is the aromatization of soluble coffee (Clarke, 1985).

In general, C. arabica beans have higher lipid content than C. canephora. However, depending on the origin and variety large variations can be found (Pinto & Carvalho, 1961; Kroplien, 1963; Tango & Carvalho, 1963; Clifford, 1985; Orsl & Dicházi, 1989).

On the other hand, only two reports in the literature present results of analysis in other coffees than C. arabica and C. canephora (Wilbaux, 1956; Chassevent et al., 1973). Chassevent et al. (1973) studied six species belonging to the section Mascarocoffea of the genus Coffea, which main characteristic is to have little or no caffeine in the seeds and a narrow geographic distribution (Clifford et al., 1991). On average, they have higher lipid content than C. arabica and C. canephora. Wilbaux (1956) investigated seeds from Coffea eugenioides, C. dewevrei, C. congensis and C. abeokutae. The Agronomic Institute has several coffee species in a living collection at Campinas, State of São Paulo, Brazil. These plants have been used for breeding program purposes for many years, however, until now they were not analysed for their lipid contents, which was the aim of this report.



Ten species were studied and for some of them more than one variety was analyzed (Table 1). Except for C. racemosa, three replicates of each specie and variety were evaluated. Ripe fruits were collected from trees growing in a living collection at the Genetics Department of Agronomic Institute, at Campinas, and dry processed. The dried fruits were de-hulled by hand and the seeds were ground with a rotating knife grinder to < 1.0 mm particles.


57n1a1ft1.GIF (14599 bytes)


To estimate the total oil content, 10-15 g of ground coffee, which had been previously dried at 105oC for 24 h, were extracted with 250 mL hexane in Soxhlet apparatus for 16-18 h. After drying overnight at room temperature and 30 minutes at 105oC the bags containing the coffee ground were weighed and the final and initial weighs used to calculate the total oil content.



The Table 1 shows the results of the total oil content of cultivars and coffee species. In agreement with previous reports in the literature, C. arabica showed higher content than C. canephora.

Kroplien (1963) analysed 36 Arabica coffee samples with a mean value of 15.5%, ranging from 14.2 to 16.8%. Seven Robusta types averaged 9.1% (7.2 to 11.0%). Pinto & Carvalho (1961) studied 33 C. arabica cultivars, detecting variation from 10.51% (cv. Amarelo de Botucatu) to 17.75% (cv. Mucronata). Tango & Carvalho (1963) analysed coffee beans of 16 C. arabica cultivars. The lowest content was observed with cv. Murta (12.4%) and the highest with Mucronata (17.0%). The total oil content of African and Brazilian C. arabica and C. canephora coffees were determined by Orsl & Dicházi (1989), which observed average values of 14.2% and 8.9%, respectively. Clifford (1985) analysed 18 Arabica samples and 45 Robustas, finding on average 16.0% and 10.1%, respectively.

Except for the cultivars of C. arabica, C. racemosa, C. stenophylla, C. eugenioides, C. kapakata and C. salvatrix, all other species showed low oil content (Table 1). The highest content, 29.18%, was observed with C. salvatrix.

Chassevent et al. (1973) observed that three species of section Mascarocoffea had more than 20% of lipids in the seeds. One of them, Coffea buxifolia, presented 30%. All these species are geographically restricted to Madagascar, Comores and Mascarenhas islands.

Wilbaux (1956) analysed the oil content in coffee species from the African continent. He found 15.6-16.1% in C. eugenioides, 14.6-15.6% in C. dewevrei cv. Excelsa and 14.8-15.2% in cv. Arnoldiana, 14.3-15.4% in C. congensis, 11.1-11.9% in C. abeokutae, 10.6-12.6% in C. canephora cv. Robusta, and 11.9-12.0% in C. liberica.

Except for C. dewevrei cv. Excelsa, C. liberica and C. congensis for which we found lower oil contents, our data are in agreement with those obtained by Wilbaux (1956). However, since nowadays C. dewevrei and C. liberica are included in a wide group called Liberio-excelsioides (Clifford et al., 1991), due to their very similar morphological characteristics, these differences might be attributed to analysis in different cultivars of these group.

Our results showed certain variability in the oil content of the coffee species of the coffee germplasm bank of the Agronomic Institute. Although none breeding program has been carried out to change this characteristic, an increase of the oil content in coffee seeds might be advantageous, since most of it is lost in the spent grounds of roasted and ground coffee after brewing, from where it could extracted and used to aromatize soluble coffee (Clarke, 1985), as the Japanese coffee industry has been doing intensively (Saldaña, 1997). Although C. arabica and C. canephora have similar oil composition, differing only in yield (Clifford, 1985), the same was not observed by Chassevent (1973) in coffee species of the section Mascarocoffea. Therefore, this has to be followed also in a breeding program aiming to change the oil content of coffee seeds.



ARNAUD, M.J. The metabolism of coffee constituents. In: CLARKE, R.J. & MACRAE, R., eds. Coffee: physiology. London, Elsevier Applied Science, 1988. cap.2, p.33-55.         [ Links ]

CHASSEVENT, F.; DALGER, G.; GERWIG, S. & VINCENT, J.C. Contribution a l'étude des Mascarocoffea. In: Colloque International sur le café (Asic), 6., Bogotá, 1973. Resumés. p.147-154.

CLARKE, R.J. Green coffee processing. In: CLIFFORD, M.N. & WILSON, K.C., eds. Coffee: botany, biochemistry and production of beans and beverage. Westport, Connecticut, Avi Publishing, 1985. cap.10, p.230-250.         [ Links ]

CLIFFORD, M.N. Chemical and physical aspects of green coffee and coffee products. In: CLIFFORD, M.N. & WILLSON, K.C., eds. Coffee: botany, biochemistry and production of beans and beverage. Westport, Connecticut, Avi Publishing, 1985. cap.13, p.305-374.         [ Links ]

CLIFFORD, M.N.; GIBSON, C.L.; RAKOTOMALALA, J.-J.; CROS, E. & CHARRIER, A. Caffeine from green beans of Mascarocoffea. Phytochemistry, Oxford, 30(12):4039-4040, 1991.         [ Links ]

FOLSTAR, P. Lipids. In: CLARKE, R.J. & MACRAE, R., eds. Coffee: chemistry. London, Elsevier Applied Science, 1985. cap.6, p.203-222.         [ Links ]

KROPLIEN, U. Green and roasted coffee tastes. Hamburg, Gordian, 1963. 260p.         [ Links ]

ORSL, F. & DICHÁZI, B. Investigation of coffee lipids. In: BIACS, P.A.; GRUIZ, K. & KREMMER, T., eds. Biological role of plant lipids. New York, Plenum Publishing, 1989. p.483-488.         [ Links ]

PINTO, M.R.G. & CARVALHO, A. Observações preliminares sobre a porcentagem de óleo nas sementes de variedades e progênies selecionadas de café. Bragantia, Campinas, 20(20):579-589, 1961.         [ Links ]

SALDAÑA, M.A. Extração de cafeína, trigonelina e ácido clorogênico de café com CO2 supercrítico. Campinas, 1997. 215p. Dissertação (Mestrado em Engenharia Química) - Universidade Estadual de Campinas, 1997.         [ Links ]

TANGO, J.S. & CARVALHO, A. Teor de óleo e de cafeína em variedades de café. Bragantia, Campinas, 22(65):793-798, 1963.         [ Links ]

WILBAUX, R. Les caféiers au Congo Belge: technologie du café Arabica et Robusta. Bruxelles, Direction de l'Agriculture des Forets et de l'Élevage, 1956. 213p.         [ Links ]


(1) This work was partially granted by "Fundação de Apoio ao Ensino e Pesquisa da Unicamp". Received for publication in October 10, 1997 and accepted in April 9, 1998.

(2) Departamento de Fisiologia Vegetal, Instituto de Biologia, Unicamp, Caixa Postal 6109, 13083-970 Campinas (SP).

(3) Centro de Genética, Biologia Molecular e Fitoquímica, Instituto Agronômico (IAC), Caixa Postal 28, 13020-902 Campinas (SP).

(4) Centro de Café e Plantas Tropicais, IAC.

(5) With a research fellowship from CNPq.

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License