Food Chemistry 118 (2010) 851855

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Chlorogenic acids and other relevant compounds in Brazilian coffees processed

by semi-dry and wet post-harvesting methods
Giselle S. Duarte a, Antnio A. Pereira b, Adriana Farah a,*
a

Laboratrio de Bioqumica Nutricional e de Alimentos, Departamento de Bioqumica, Instituto de Qumica, Universidade Federal do Rio de Janeiro, Ilha do Fundo,
RJ 21941-909, Brazil
b
Empresa de Pesquisa Agropecuria de Minas Gerais, Centro Tecnolgico da Zona da Mata (CTZM/EPAMIG), Viosa, Minas Gerais 36571-000, Brazil

a r t i c l e

i n f o

Article history:
Received 5 December 2008
Received in revised form 5 May 2009
Accepted 13 May 2009

Keywords:
Coffee
Post-harvesting
Semi-dry post-harvesting method
Wet post-harvesting method
Chlorogenic acid
Trigonelline
Sucrose

a b s t r a c t
The levels of nine chlorogenic acids, caffeine, trigonelline and sucrose were determined by HPLC-UV and
HPLC-RI systems in wet and semi-dry post-harvested coffee seeds from 17 Brazilian Arabica cultivars and
progenies. Coffees processed by wet method showed higher contents of chlorogenic acids (p = 0.02) and
trigonelline (p < 0.01), and lower content of sucrose (p = 0.02) compared to those produced by a semi-dry
method. Regarding caffeine, no difference was observed between both methods. The implications of the
differences observed in the chemical composition of coffee seeds treated by wet and semi-dry methods
on cup quality deserve investigation.
2009 Elsevier Ltd. All rights reserved.

1. Introduction
Coffee is the most consumed food product in the world. Brazil is
the main producer and exporter of green coffee seeds, being
responsible for 35% of world production. Brazilian coffee is present
in most blends sold around the world. Moreover, Brazilian internal
coffee consumption has been growing signicantly in recent years.
Today, Brazilian consumption represents 14% of world coffee demand and corresponds to 32% of the whole European coffee market
(ABIC, 2008).
After harvesting of the fruits, green coffee seeds are obtained by
one of three different methods known as dry, wet and semi-dry
processing (Teixeira, Brando, Thomaziello, & Teixiera, 1995).
Although all methods aim at removing the fruit esh of coffee cherry, they do it in different ways (Trugo & Macrae, 1984). In the dry
method, the whole cherry (bean, mucilage and pulp) is dried under
the sun or in a mechanical dryer, followed by mechanical removal
of the dried outer parts (Rothfos, 1980). In this process, the drying
operation is the most important step because it affects the nal
quality of coffee. While the over-dried fruits become brittle and
produce too many broken (defective) seeds during hulling, on the
other hand, fruits that are not sufciently dried become more vulnerable to deterioration caused by fungus and bacteria attacks
* Corresponding author. Tel./fax: +55 21 2562 8213.
E-mail address: afarah@iq.ufrj.br (A. Farah).
0308-8146/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodchem.2009.05.042

(Rothfos, 1980). The nal product resulting from the dry method
is an unwashed or natural coffee (Smith, 1985).
In the wet method, which requires the use of specic equipment and substantial amounts of water, the pulp is eliminated
by a pulper, followed by mucilage removal. This is carried out by
chemical products or by natural fermentation for varying lengths
of time, which will depend on climatic conditions and fruit ripeness (Teixeira et al., 1995). At the end of fermentation, the wet processed seeds are washed and dried (Leloup, Cancel, Liardon, Rytz, &
Pithon, 2004). The nal product is a washed or parchment coffee (Smith, 1985). The key differences between the dry and the wet
methods are the pulp removal in the wet method the fruit pulp is
separated from the seeds before drying stage and the fermentation step, which only happens in the wet method.
The semi-dry or pulped natural method is an intermediate between dry and wet processes that started to be used in Brazil in
the early 1990s. In this method, the cherries are pulped and the
seeds dried while surrounded by the mucilage, without the fermentation step for mucilage removal (Teixeira et al., 1995). Coffee
seeds resulting from this method are called pulped natural coffees.
Regarding cup quality, washed coffees are known to present better quality, less body, higher acidity and more aroma than the unwashed coffees (Mazzafera & Padilha-Purcino, 2004), while pulped
natural coffees present an intermediate body between washed and
unwashed ones. Moreover, pulped natural coffees are also strongly
appreciated in blends for espresso coffee (Teixeira et al., 1995).

852

G.S. Duarte et al. / Food Chemistry 118 (2010) 851855

It is well known that post-harvest processing has pronounced

effects on the chemical composition of coffee seeds, specially in
water-soluble components like sugars, caffeine, trigonelline and
chlorogenic acids (Smith, 1985). Despite the considerable availability of studies comparing coffees processed by dry and wet methods
in the literature (Balylaya & Clifford, 1995; Clifford & Ramirez-Martinez, 1991; Guyot, Guele, Assemar, Tchana, & Pomathios, 1995;
Knopp, Bytof, & Selmar, 2006; Leloup et al., 2004; Mazzafera & Padilha-Purcino, 2004), data on the chemical composition of coffees
processed by semi-dry methods are still scarce.
Chlorogenic acids (CGA), a major family of phenolic compounds
that represent 612% of coffee constituents in mass (Farah, DePaulis, Trugo, & Martin, 2005), are known to be responsible for coffee
pigmentation and astringency (Farah & Donangelo, 2006). The
thermal degradation of chlorogenic acids during roasting will result in formation of phenolic substances that contribute to bitterness (Clifford, 1985a) and aromatic compounds like phenols,
which are undesirable to cup quality (Toci & Farah, 2008). Additionally, CGA participates in colour formation through its incorporation into the backbone of melanoidins (Farah & Donangelo,
2006). The major CGA classes in coffee are the caffeoylquinic acids
(CQA), feruloylquinic acids (FQA) and dicaffeoylquinic acids (diCQA), with at least three isomers per class (Clifford, 2000). Caffeine
is a xanthine derivative that gives a bitter characteristic taste to
coffee (Farah, DePaulis, Moreira, Trugo, & Martin, 2006). Trigonelline is a pyridine derivative known to contribute indirectly to the
formation of desirable and undesirable aroma compounds during
roasting (Macrae, 1985; Moreira, Trugo, & Maria, 2000). Trigonelline content has also been correlated to good cup quality (Farah,
Monteiro, Calado, Franca, & Trugo, 2006). Sugars, particularly sucrose, which is the most abundant simple carbohydrate in coffee,
act as aroma precursor, originating several volatile substances during roasting such as furans, aldehydes and carboxylic acids (Farah,
DePaulis et al., 2006; Farah, Monteiro et al., 2006; Teixeira et al.,
1995). In the last few years, CGA, caffeine and trigonelline have
also been the subject of several investigations in view of their
potentially positive biological effects in humans.
The aim of this study was to compare the composition of the
nine main chlorogenic acids, caffeine, trigonelline and sucrose in
Brazilian green coffee seeds produced by wet and semi-dry postharvesting methods.

2. Materials and methods

2.1. Samples
Seventeen coffee samples were used in this study, being four
Arabica cultivars (Red Catua; Rubi; Yellow Bourbon and Topzio)
and 11 hybrids, 10 from the crossing of C. arabica cv. Yellow Catua
with Timor Hybrid (numbers 111) and two from the crossing of
cv. Red Caturra with Timor Hybrid (numbers 12 and 13). The coffee
fruits were produced in a farm located in Paula Cndido, Minas
Gerais, Brazil. After harvesting, fruits were processed by wet and
semi-dry methods.
2.1.1. Post-harvesting treatments
The ripe coffee cherries were harvested by hand picking, followed by wet or semi-dry processing. For wet processing, coffee
fruits were quickly washed, manually pulped, followed by 8 h
soaking for fermentation and mucilage removal. After soaking,
parchment coffees were washed and de-hulled. The semi-dry processed seeds, were quickly washed and manually pulped, followed
by washing and de-hulling, without mucilage removal. After wet or
semi-dry processing, coffee seeds were stored in plastic bags under
20 C until chemical analysis.

2.2. Extractions for chromatographic analyses

Prior to extraction, all seeds were milled to pass a 0.46 mm
sieve. CGA were extracted with aqueous methanol (40%) and claried with Carrez solutions according to Trugo and Macrae (1984),
also described in Farah et al. (2005). Caffeine and trigonelline were
extracted with 95 C distilled water and claried with aqueous
plumb acetate (20%) (Farah, DePaulis et al., 2006; Farah, Monteiro
et al., 2006; Trugo & Macrae, 1984). For sucrose evaluation, green
coffee seeds were extracted with aqueous methanol (40%) and claried with acetonitrile (Trugo, Farah, & Cabral, 1995). Extractions
were performed in duplicates.
2.2.1. HPLC analyses
2.2.1.1. Chlorogenic acids. CGA were analysed, using a HPLC gradient system with a high-precision pump (Shimadzu model LC-10AD, Kyoto, Japan), a UV detector (Shimadzu model LC-10-AD) operating at 325 nm, and a ODS-C18 column (Rexchrom: 5 lm,
250  4.6 nm, Regis Technologies, Morton Grove, IL) coupled with
a guard column (Rexchrom: 5 lm, 10  3 nm, Regis Technologies).
Chromatographic data was recorded and integrated in the Class VP
computer software (Shimadzu). A gradient using 10 mM citric acid
solution and methanol was performed (Farah et al., 2005). The
identity of CGA was conrmed by LCMS analyses as described
by Farah, DePaulis et al. (2006) and Farah, Monteiro et al. (2006).
In this work, the authors used IUPAC numbering system for CGA.
When citing other authors, their numbering has been changed
when necessary for consistency.
2.2.1.2. Caffeine and trigonelline. Caffeine and trigonelline analyses
were performed according to Trugo and Macrae (1984), also described in Farah, DePaulis et al. (2006) and Farah, Monteiro et al.
(2006). Analyses were performed by an isocratic system consisting
of pump (Knauer, Model 64, Germany), integrator (Shimadzu,
Model C-R6A) and UV detector (Shimadzu, model SPD-1-AV) operating at 272 nm for caffeine and 264 nm for trigonelline. A reversal
phase column (ODS C-18 Merck, Darmstadt, Germany) was used
with 40% methanol as mobile phase for caffeine and with 5% methanol for trigonelline. Flow rate was 1 mL/min.
2.2.1.3. Sucrose. Sucrose analyses were performed according to Trugo et al. (1995), adapted. A pump (Knauer, model 64) with a 20 lL
Rheodyne xed loop injector (Oak Harbor, WA, USA), Spherisorb
NH2 columm (Supelco, Bellefonte, PA, USA), RI detector (Waters, Milford, MA, USA) and integrator (Hewllet Packard, model) were used. A
mobile phase with acetonitrile 80%, at 1 mL/min was used.
2.3. Water content
In order to express the results in dry matter basis (dm), the
water content of all samples was determined according to AOAC
procedures (2000).
2.4. Statistical analysis
HPLC results were tested for differences by GraphPad Prism
software, version 4.0 (San Diego, CA, USA), using paired t-test
method and considered signicant when p 6 0.05.
3. Results and discussion
3.1. Chlorogenic acids
The levels of CQA, FQA and diCQA in green coffee seeds processed by wet and semi-dry post-harvesting methods are

853

G.S. Duarte et al. / Food Chemistry 118 (2010) 851855

presented in Tables 1 and 2, respectively. For the rst time, the distribution and content of nine CGA (3-CQA; 4-CQA; 5-CQA; 3-FQA;
4-FQA; 5-FQA; 3,4-diCQA; 3,5-diCQA; 4,5-diCQA) are compared in
Brazilian coffees processed by wet and semi-dry methods. The CGA
contents observed in this work are consistent with previous data
for coffee in general (Farah & Donangelo, 2006). In the wet method,
the average contents of CQA, di-CQA and FQA were, respectively,
4.94 0.41; 0.94 0.05 and 0.32 0.02 g/100 g (dm), representing
81.3%, 13.3% and 5.3% of total CGA content (6.08 0.44 g/100 g,
dm). In this group, the sample with highest level of CGA was hybrid
1 (7.4 0.21 g/100 g, dm), while the one with the lowest level was
C. arabica cv. Topzio (5.2 0.57 g/100 g, dm).
In the semi-dry method, the mean levels of CQA were
4.70 0.37 g/100 g (dm), followed by diCQAs with 0.98 0.02 g/
100 g (dm) and FQA with 0.30 0.01 g/100 g (dm). These classes
accounted for 81.0%, 14.0% and 5.1% of total CGA levels
(5.8 0.38 g/100 g, dm), respectively. The highest content of total
CGA was observed in hybrid 2 (6.7 0.58 g/100 g, dm), mainly
due to higher contents of 5-CQA and, to a lesser extent, 3,5-diCQA.
The lowest content was observed in hybrid 3 (5.4 0.25 g/100 g,
dm). The highest and lowest CGA contents present in hybrids 2
and 3, respectively, may result from different genetic inheritance

of C. arabica and C. canephora, since Timor hybrid samples were

produced by crossing of these two species. The higher CGA content
in C. canephora comparing to C. arabica has been extensively reported (Clifford & Ramirez-Martinez, 1991; Farah & Donangelo,
2006).
Although the distribution prole of CGA classes in samples treated by the semi-dry method was similar to those treated by the
wet method, on average, coffees processed by the wet method presented signicantly higher total CGA content (p = 0.02) than those
processed by the semi-dry method. This difference was observed in
all CGA classes, except for the diCQA.
No data on CGA composition of semi-dry method is available in
the literature. However, literature has reported a higher CGA content in wet processed Robusta seeds than in dry processed seeds.
This difference in CGA contents has been attributed to losses of
other water-soluble components by lixiviation and fermentation
in the wet method (Wootton, 1974), in addition to slight degradation of such components in dry processed seeds exposed to sun
heat. In this work, it is possible to observe that the CGA composition of semi-dry seeds are intermediate to those of dry and wet
processed seeds, being more comparable to results obtained for
dry processed seeds. This result is coherent, since semi-dry pro-

Table 1
Chlorogenic acids contents in coffee samples processed by semi-dry post-harvesting method.
Samples

3-CQA

4-CQA

5-CQA

3-FQA

4-FQA + 5-FQA

3,4-diCQA

3,5-diCQA

4,5-diCQA

Hybrid 1
Hybrid 2
Hybrid 3
Hybrid 4
Hybrid 5
Hybrid 6
Hybrid 7
Hybrid 8
Hybrid 9
Hybrid 10
Hybrid 11
Hybrid 12
Hybrid 13
Yellow Bourbon
Red Catua
Rubi
Topzio

460.0 30.2
476.1 18.1
391.2 25.3
455.5 15.4
465.5 22.4
502.8 20.0
438.0 14.6
472.1 26.5
440.7 15.2
479.2 22.6
480.1 38.9
509.1 19.2
510.5 52.3
403.9 41.2
437.9 40.5
445.9 17.1
536.0 15.3

615.1 10.5
666.3 24.3
534.3 65.8
590.7 89.0
657.3 58.1
657.4 20.5
615.1 32.0
647.4 16.7
624.8 37.6
660.3 41.9
662.3 18.8
676.2 22.5
680.3 20.9
582.6 74.0
605.8 30.2
617.9 24.6
689.4 28.2

4200.4 100.1
4116.4 98.5
3362.2 52.7
3210.1 74.3
4317.5 101.5
3562.2 78.4
4090.3 70.5
3907.9 35.2
3723.5 40.1
3905.2 98.2
3920.3 84.1
2967.4 14.6
3005.2 65.2
3305.0 74.0
3226.3 41.2
3191.5 54.2
3251.2 35.4

28.0 0.5
29.3 2.5
27.5 3.4
30.7 1.2
28.3 1.3
28.6 2.4
29.4 5.6
27.9 2.3
26.9 6.0
28.2 2.4
31.3 1.7
39.0 1.9
41.2 2.8
27.1 6.5
32.8 7.1
28.0 1.3
28.6 1.3

258.2 20.1
264.6 14.5
246.1 17.2
309.2 14.3
299.8 15.7
316.9 16.8
285.4 35.3
259.9 17.1
290.0 24.3
270.0 25.2
271.3 36.4
253.9 30.5
267.0 22.7
221.4 50.3
247.1 47.8
206.9 12.7
257.9 19.9

158.3 15.3
203.7 12.8
151.9 19.7
186.7 9.8
154.1 17.6
216.6 18.2
149.9 19.1
168.8 11.1
191.3 21.7
194.3 17.3
196.3 16.2
253.5 28.3
249.3 31.4
220.6 14.4
214.5 21.7
238.0 8.5
238.3 9.7

367.3 10.1
459.4 19.3
328.1 21.2
353.6 12.8
343.9 16.7
414.4 19.1
312.4 21.1
313.2 17.6
387.8 11.0
380.2 10.0
380.3 10.0
431.6 15.0
432.6 10.0
462.1 8.5
440.4 13.3
442.3 12.5
423.2 12.2

181.1 7.1
249.7 14.6
201.6 15.0
189.1 4.5
190.6 9.1
204.4 16.0
197.4 12.2
170.2 6.8
240.5 9.7
219.0 12.4
220.3 18.4
243.2 18.2
250.1 14.3
125.0 17.0
229.4 4.5
248.5 13.0
225.8 10.0

CQA = caffeoylquinic acids; FQA = feruloylquinic acids; diCQA = dicaffeoylquinic acids. Results are shown as mean of duplicate extraction standard deviation, expressed as
mg/100 g of coffee in dry weight basis.

Table 2
Chlorogenic acids contents in coffee samples processed by wet post-harvesting method.
Samples

3-CQA

4-CQA

5-CQA

3-FQA

4-FQA + 5-FQA

3,4-diCQA

3,5-diCQA

4,5-diCQA

Hybrid 1
Hybrid 2
Hybrid 3
Hybrid 4
Hybrid 5
Hybrid 6
Hybrid 7
Hybrid 8
Hybrid 9
Hybrid 10
Hybrid 11
Hybrid 12
Hybrid 13
Yellow Bourbon
Red Catua
Rubi
Topzio

530.2 20.1
504.0 41.5
487.6 15.6
416.0 48.9
479.8 16.6
478.9 17.2
464.7 18.1
467.2 33.5
540.4 74.0
535.6 50.1
512.4 20.6
506.3 13.5
480.8 10.2
501.5 17.0
528.1 23.3
421.1 21.7
384.0 14.6

675.6 52.1
716.3 21.3
677.3 17.1
547.3 18.4
701.0 7.4
609.0 10.0
619.4 28.0
646.3 25.2
727.9 21.3
705.2 17.1
732.3 19.8
662.0 30.8
627.9 20.9
605.2 15.6
685.3 24.3
571.8 21.2
503.2 17.1

4845.1 84.2
4351.7 13.2
3901.7 5.4
3395.8 98.3
4396.5 11.5
3239.6 76.6
3862.2 65.3
4574.7 90.3
4017 30.2
4125.8 21.3
4125.8 20.8
3527.8 15.2
2852.0 15.4
3410.2 17.5
3256.8 41.6
3682.6 38.8
3162.8 19.0

30.2 10.3
31.3 8.5
29.7 7.1
32.4 0.3
35.1 9.5
30.7 5.1
29.8 4.0
27.7 3.4
31.4 8.2
32.1 9.9
35.6 10.0
30.9 8.9
38.5 0.9
34.1 0.1
34.9 7.1
26.6 7.0
20.5 0.5

298.2 21.8
264.6 40.1
263.1 35.1
330.7 1.2
379.2 9.1
305.6 10.1
366.7 18.7
317.1 52.2
327.8 0.20.0
328.7 41.8
315.8 4.6
236.9 31.1
252.6 20.0
269.6 15.8
283.4 19.7
228.7 21.5
209.4 17.0

196.2 10.1
212.1 14.6
178.8 2.8
160.9 17.5
184.1 24.6
234.3 18.1
153.2 17.1
207.3 11.0
189.5 18.3
205.7 13.3
223.5 19.0
203.5 16.8
216.3 17.0
275.6 21.7
280.1 22.8
180.8 19.0
189.2 9.0

410.2 30.5
401.7 2.8
366.0 14.4
314.1 10.4
401.1 9.7
419.8 10.5
351.0 15.4
402.5 32.5
342.2 38.0
325.0 7.5
402.5 18.6
383.4 16.1
387.1 18.0
510.1 17.0
270.5 22.9
205.2 28.0
363.2 7.5

220.3 10.2
292.1 15.4
266.4 17.1
206.9 9.5
247.5 10.4
273.5 11.6
102.4 10.7
290.9 9.8
218.7 10.4
215.8 3.2
265.7 17.2
247.6 14.3
260.1 12.0
198.4 16.1
274.8 17.0
268.5 18.1
238.0 18.1

CQA = caffeoylquinic acids; FQA = feruloylquinic acids; diCQA = dicaffeoylquinic acids. Results are shown as mean of duplicate extraction standard deviation, expressed as
mg/100 g of coffee in dry weight basis.

854

G.S. Duarte et al. / Food Chemistry 118 (2010) 851855

cessed seeds do not undergo soaking and fermentation like wet

processed seeds. Moreover, it is noteworthy to point that this difference between the chemical composition of semi-dry and wet
processed seeds may be considerably amplied if soaking time is
increased in the wet method. While in this work, seeds have been
soaked for 8 h, soaking time may be increased to up to 48 h (Teixeira et al., 1995).
Considering the similarity of dry and semi-dry methods, the
present results are in agreement with results from Balylaya and
Clifford (1995), using three dry and wet-processed C. canephora
cv. Robusta samples and Leloup et al. (2004) for six Robusta samples. On the other hand, in Balylaya and Clifford (1995), wet-processed Arabica seeds (n = 3) showed lower levels of total
chlorogenic acids compared to dry processed seeds. In this work,
when only Arabica seeds (n = 4) were compared statistically, no
signicant difference was observed between wet and semi-dry
methods. Therefore, the difference observed between both methods (p = 0.02) is attributed to the hybrid samples (n = 13), which
presented a similar behaviour to Robusta samples. This difference
between the results obtained for Arabica and Robusta/hybrids
samples may be attributed to differences in their cell-wall constitutions (Toci & Farah, 2008).
3.2. Caffeine
In the coffee seeds processed by the wet method, the contents
ranged from 1.05 0.02 to 1.53 0.01 g/100 g (dm) while in those
seeds processed by semi-dry method, caffeine contents ranged
from 1.12 0.02 to 1.54 0.02 g/100 g (dm) of coffee (Table 3).
These results are in agreement with literature reports (Farah, DePaulis et al., 2006; Farah, Monteiro et al., 2006; Macrae, 1985). No
signicant difference between both methods was observed regarding caffeine contents. This is in agreement with results from Balylaya and Clifford (1995) and Leloup et al. (2004) for dry and wet
processed seeds, and may be possibly due to the thermo-stability
and the lower contents of caffeine in coffee compared to CGA.
3.3. Trigonelline
Trigonelline contents in seeds processed by the wet method
ranged from 0.80 0.02 to 1.40 0.01 g/100 g (dm) with average
of 0.90 0.01 g/100 g (dm), while in seeds processed by semi-dry
method, trigonelline contents ranged from 0.64 0.01 to
0.92 0.01 g/100 g (dm) with average of 0.81 0.02 g/100 g (dm)

(Table 3). These contents are in agreement with general coffee data
(Farah, DePaulis et al., 2006; Farah; Martn, Pablo, & Gonzalz,
1998; Monteiro et al., 2006). As with CGA, the relative increase
in trigonelline content of coffee seeds processed by wet method
compared to semi-dry method (p < 0.0001) may be due to loss of
other components with higher water solubility by lixiviation and
thermal degradation.
3.4. Sucrose
The content of sucrose in coffee treated by the wet method ranged from 8.0 0.02 to 10.8 0.05 g/100 g (dm), with average of
9.0 0.03 g/100 g (dm) (Table 3). Sucrose content in seeds treated
by the semi-dry method ranged from 7.10 0.02 g/100 g to
14.5 0.05 g/100 g (dm) with average of 12.3 0.07 g/100 g (dm)
(Table 3). These contents are generally in agreement with reports
by Farah, DePaulis et al. (2006), Farah, Monteiro et al. (2006) and
Knopp et al. (2006). Sucrose content in coffee seeds processed by
the semi-dry method was signicantly higher than in those processed by the wet method (p = 0.02) when including hybrids and
Arabica samples. However, when evaluating only Arabica samples,
no differences were found between semi-dry and wet methods.
These results are also in agreement with those by Knopp et al.
(2006).
In contrast with our results, Leloup et al. (2004) did not nd differences when comparing sugar contents in dry and wet Robusta
samples. However, our results are comparable to results from
Guyot et al. (1995) who evaluated samples of C. canephora cv. Robusta treated by dry and wet methods.
As stated above, differently from the wet method, in the semidry method the mucilage is kept on the bean. According to Avallone, Guiraud, Guyot, Olguin, and Brillouet (2001), the mucilage
is composed mostly by cell well polysaccharides, pectic substances
and monosaccharides. Sucrose is not present in a dominant proportion in the mucilage when compared with other sugars. The
authors state that this disaccharide is present mostly inside the
coffee bean. Thus, it would not be possible to easily conclude that
coffees produced by semi-dry methods have more sucrose than
those produced by the wet method. Therefore, sucrose losses observed in our wet processed hybrid seeds are possibly mostly
due to the high water solubility of this compound during washing
and soaking and due to its use as a substrate for fermentation
(Wootton, 1974). As with the CGA, the difference between the results obtained for Arabica and hybrids from Robusta samples might

Table 3
Caffeine, trigonelline and sucrose contents in coffee samples processed by semi-dry and wet post-harvesting methods.
Samples

Hybrid 1
Hybrid 2
Hybrid 3
Hybrid 4
Hybrid 5
Hybrid 6
Hybrid 7
Hybrid 8
Hybrid 9
Hybrid 10
Hybrid 11
Hybrid 12
Hybrid 13
Yellow Bourbon
Red Catua
Rubi
Topzio

Semi-dry method

Wet method

Caffeine

Trigonelline

Sucrose

Caffeine

Trigonelline

Sucrose

1.54 0.02
1.38 0.02
1.40 0.01
1.30 0.02
1.34 0.02
1.37 0.03
1.32 0.01
1.35 0.01
1.25 0.01
1.20 0.01
1.11 0.02
1.19 0.01
1.12 0.03
1.13 0.02
1.13 0.02
1.21 0.01
1.43 0.01

0.83 0.03
0.81 0.03
0.64 0.02
0.84 0.02
0.76 0.02
0.77 0.02
0.79 0.03
0.78 0.03
0.85 0.03
0.90 0.03
0.89 0.03
0.76 0.03
0.77 0.02
0.90 0.01
0.78 0.04
0.74 0.01
0.92 0.01

11.21 0.02
11.6 0.01
7.13 0.02
10.39 0.01
10.10 0.01
9.82 0.01
9.25 0.02
9.50 0.03
10.31 0.05
9.01 0.01
8.88 0.01
14.54 0.02
13.41 0.02
10.05 0.01
11.68 0.01
11.98 0.02
12.66 0.02

1.53 0.01
1.42 0.01
1.45 0.01
1.15 0.01
1.46 0.01
1.19 0.03
1.39 0.01
1.47 0.02
1.29 0.01
1.05 0.01
1.14 0.01
1.23 0.01
1.22 0.01
1.05 0.02
1.26 0.02
1.19 0.02
1.16 0.01

0.87 0.01
0.96 0.01
0.88 0.01
0.93 0.01
0.93 0.01
0.85 0.01
0.86 0.01
0.86 0.01
1.03 0.02
0.92 0.01
1.12 0.01
0.80 0.02
0.83 0.01
0.89 0.01
0.92 0.01
0.80 0.01
0.85 0.02

10.41 0.01
10.56 0.01
8.78 0.01
9.73 0.01
9.89 0.01
10.49 0.05
8.53 0.05
9.64 0.05
9.34 0.05
8.05 0.04
9.61 0.04
8.98 0.01
10.03 0.01
10.88 0.04
9.85 0.03
9.16 0.05
8.49 0.04

Results are shown as mean of duplicate extraction standard deviation, expressed as mg/100 g of coffee in dry weight basis.

G.S. Duarte et al. / Food Chemistry 118 (2010) 851855

be attributed to differences in their cell-wall constitutions (Toci &

Farah, 2008) but should be thoroughly investigated.
4. Conclusions
For the rst time, the contents of nine chlorogenic acids, caffeine, trigonelline and sucrose were evaluated together in Brazilian
Arabica coffee cultivars and hybrids treated by wet and semi-dry
post-harvesting methods. The wet method produced an increase
in chlorogenic acids and trigonelline contents and a small loss in
sucrose contents comparing to semi-dry method. However, since
the soaking period in the wet method is largely variable, more dramatic changes may be observed in these coffee components when
seeds are soaked for longer periods of time. Sensorial analysis
should be carried out in order to correlate the chemical differences
observed between both post-harvesting methods with cup quality.
Acknowledgements
The authors would like to thank Henrique Theodoro Martins
Peixoto, Dartanhan Augusto Fernandes Lima and Jos Carlos Silva
Guimares for assistance with coffee production, harvesting and
processing; and to Carolina Castro Borges for technical assistance
with trigonelline analysis. The nancial support of CBPandD/
Caf-EMBRAPA; CNPq and FAPERJ is greatly acknowledged.
References
ABIC. (2008). Notcias. <http://www.abic.com.br/exportao.html> Accessed
20.04.2008.
AOAC (2000). Ofcial methods of analysis of the association of ofcial analytical
chemists (17th ed.). Gaithersburg, MD, USA: Association of Ofcial Analytical
Chemists.
Avallone, S., Guiraud, J. P., Guyot, B., Olguin, E., & Brillouet, J. M. (2001). Fate of
mucilage cell wall polysaccharides during coffee fermentation. Journal of
Agricultural and Food Chemistry, 49, 55565559.
Balylaya, K. J., & Clifford, M. N. (1995). Chlorogenic acids and caffeine contents of
monsooned Indian Arabica and robusta coffees compared with wet and dry
processed coffees from the same geographic area. In ASIC proceedings of 16th
colloque coffee (pp. 316324). Kyoto.
Clifford, M. N. (1985a). Chlorogenic acids. In R. J. Clarke & R. Macrae (Eds.), Coffee
chemistry (pp. 153202). Elsevier, Applied Science: Amsterdam.

855

Clifford, M. N. (2000). Chlorogenic acids and others cinnamates-nature, occurrence,

dietary burden, absorption and metabolism. Journal of Science and Food
Agricultural, 80, 10331043.
Clifford, M. N., & Ramirez-Martinez, J. R. (1991). Phenols and caffeine in wetprocessed coffee seeds and coffee pulp. Food Chemistry, 40, 3542.
Farah, A., DePaulis, T., Moreira, D. P., Trugo, L. C., & Martin, P. R. (2006). Chlorogenic
acids and lactones in regular and water-decaffeinated Arabica coffees. Journal of
Agricultural and Food Chemistry, 54, 374381.
Farah, A., DePaulis, T., Trugo, L. C., & Martin, P. R. (2005). Formation of chlorogenic
acids lactones in roasted coffee. Journal of Agricultural and Food Chemistry, 53,
11051113.
Farah, A., & Donangelo, C. M. (2006). Phenolic compounds in coffee. Brazilian Journal
Plant Physiology, 18(1), 2336.
Farah, A., Monteiro, M. C., Calado, V., Franca, A. S., & Trugo, L. C. (2006). Correlation
between cup quality and chemical attributes of Brazilian coffee. Food Chemistry,
98, 373380.
Guyot, B., Guele, D., Assemar, S., Tchana, E., & Pomathios, L. (1995). Inuence du
mode de prparation du caf vert robusta sur as composition chimique et ses
qualits organoleptiques. In ASIC proceedings of the 16th colloque scientique
international sur le Caf (pp. 267277). Paris.
Knopp, S.-E., Bytof, G., & Selmar, D. (2006). Inuenced of processing on the content
of sugars in green Arabica coffee seeds. European Food Research and Technology,
223, 195201.
Leloup, V., Cancel, C., Liardon, R., Rytz, A., & Pithon, A. (2004). Impact of wet and dry
process on green composition and sensory characteristics. In ASIC proceedings of
20th colloque coffee (pp. 93100). Bangalore, India.
Macrae, R. (1985). Nitrogenous compounds. In R. J. Clarke & R. Macrae (Eds.), Coffee
chemistry (pp. 115152). Elsevier, Applied Science: Amsterdam.
Martn, M. J., Pablo, F., & Gonzalz, A. G. (1998). Discrimination between arabica and
robusta green coffee varieties according to their chemical composition. Talanta,
46, 12591264.
Mazzafera, P., & Padilha-Purcino, R. (2004). Post harvest processing methods and
alterations in coffee fruit. In ASIC proceedings of 20th colloque coffee. Bangalore,
India.
Moreira, R. F. A., Trugo, L. C., & Maria, C. A. B. (2000). Compostos volteis do caf
torrado. Parte II. Compostos alifticos, alicclicos e aromticos. Quimica Nova, 23,
195201.
Rothfos, B. (1980). Coffee production. Hamburg: Gordian-Max-Rieck.
Smith, A. W. (1985). Agricultural practices. In R. J. Clarke & R. Macrae (Eds.), Coffee
chemistry (pp. 1823). Elsevier, Applied Science: Amsterdam.
Teixeira, A. A., Brando, C. H. J., Thomaziello, R. A., & Teixiera, R. (1995). In A. Illy & R.
Viani (Eds.), Espresso coffee: The science of quality (pp. 9195). Italy: Elsevier
Academic Press. 197, 198.
Toci, A. T., & Farah, A. (2008). Volatile compounds as potential defective coffee
seeds markers. Food Chemistry, 108, 11331141.
Trugo, L. C., Farah, A., & Cabral, L. (1995). Oligosaccharides distribution in Brazilian
soybean cultivars. Food Chemistry, 52(4), 385387.
Trugo, L. C., & Macrae, R. A. (1984). Chlorogenic acid composition of instant coffee.
Analyst, 109, 263266.
Wootton, A. E. (1974). The dry matter loss from parchment coffee during eld
processing. In ASIC proceedings of the 16th Colloque Scientique International sur
la Chimie des Cafs Verts (pp. 316324). Paris.