ARTICLE IN PRESS

BIOSYSTEMS ENGINEERING 99 (2008) 62 – 66

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/issn/15375110

Research Paper: PH—Postharvest Technology

Ultrastructural analysis of drying damage in parchment

Arabica coffee endosperm cells

F.M. Boréma,, E.R. Marquesb, E. Alvesc

a
Department of Engineering, University Federal of Lavras, P.O. Box 37, 37200-000 Lavras, Minas Gerias, Brazil
b
Department of Food Science, University Federal of Lavras, P.O. Box 37, 37200-000 Lavras, Minas Gerais, Brazil
c
Department of Plant Pathology, University Federal of Lavras, P.O. Box 37, 37200-000 Lavras, Minas Gerias, Brazil

ar t ic l e i n f o The objective of this work was to evaluate and compare the alterations in the structure of
coffee seed endosperm subjected to different temperatures and drying conditions. The
Article history: seeds were dried at 40, 50 and 60, with an airflow of 0.33 m3 s 1
m 2. After drying, 10 seeds
Received 14 September 2006 were randomly selected and prepared for the histochemical tests with Sudan IV and
Received in revised form scanning and transmission electron microscopy, according to the laboratory’s routine
4 August 2007 techniques. The histochemical results showed that, for the coffee parchment beans dried
Accepted 28 September 2007 at 40 1C, there was no change in the cellular integrity of the plasma membrane and vesicles.
Available online 19 November 2007 In contrast, in the endosperm of parchment coffee beans dried at 60 1C, fused oil bodies
that gave rise to large droplets in the intercellular space were observed, indicating a rupture
of the vesicles and plasma membrane. Scanning electron microscopy showed that, for the
parchment Arabica coffee beans dried at 40 1C, the internal cellular content remained intact
and full of cellular material and the space between the plasma membrane and the cell wall
was empty. However, in seeds dried at 60 1C, a rupture of the cells was observed,
represented by occluded intercellular spaces, indicating a leaking of part of the protoplasm.
The results from the transmission electron microscopy corroborated the undamaged and
the damaged structure of the coffee parchment beans dried at 40 and 60 1C, respectively.
& 2007 IAgrE. Published by Elsevier Ltd. All rights reserved.

1. Introduction attention, and several studies describe the impact of the wet
and dry processing on the physiology and quality of coffee
The drying of coffee is one of the most important processing (Bytof et al., 2000). Nevertheless, ultrastructural analyses
stages, and involves the removal of water from the coffee which occur upon drying are not well understood.
parchment beans into the surrounding environment until the Inappropriate handling of coffee cherry before and after
beans reach an equilibrium moisture content. Several re- picking, such as the use of high temperatures and drying
searches (Berbert et al., 1994; De Grandi et al., 2000; Freire, rates, can lead to a degeneration of the plasma membrane.
1998; Guimarães et al., 1998) have evaluated the drying The alterations in the structure of the plasma membrane and
system, the reduction of the energy consumption and the its deteriorating capacity to act as a semipermeable barrier
dryer’s efficiency. Recently, the importance of post-harvest are the main factors responsible for the decrease of quality in
treatments for coffee bean quality has received increasing coffee (Amorim, 1978; Salazar et al., 1994). Studies have shown

Corresponding author. Department of Engineering, University Federal of Lavras, P.O. Box 37, 37200-000 Lavras, Minas Gerias, Brazil.
Tel.:+ 55 35 38291488.
E-mail address: [email protected] (F.M. Borém).
1537-5110/$ - see front matter & 2007 IAgrE. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.biosystemseng.2007.09.027
 ARTICLE IN PRESS
BIOSYSTEMS ENGINEERING 99 (2008) 62– 66 63

Table 1 – Summary description of the coffee cup days with two pickings of approximately 0.9 m3 of coffee
classification system fruits for each repetition. The coffee was then separated
according to the density of the fruits using a mechanical
Flavour Classification characteristics washer. The cherry fruits were peeled to obtain the parch-
ment coffee which was carried to a concrete ground on the
Strictly Low acidity, mellow sweetness, pleasant mouth-feel evening it was picked and spread in layers of approximately
soft
1 cm depth. The coffee was separated into 12 equal portions,
Soft Same characteristics as strictly soft, only less six of which remained on the ground for 1 day and the
accentuated remaining six for 3 days to reduce the moisture content. The
Softish Same characteristics as soft, only less accentuated samples were stirred during the day every 30 min. Thus, the
parchment coffee remained on the ground for 1 and 3 days
Hard Lacks sweetness and softness
before artificial drying to obtain two levels of moisture
Rioysh Iodine, medicine-like inky flavour from microbe- content. After solar drying, the moisture content was
tainted beans determined at the start of the drying process in the dryers.
Rio Same characteristics as rioysh, only more This was carried out by the oven method at 10573 1C for 24 h
accentuated (Brasil, 1992). The coffee was then dried in three dryers of
fixed layers 0.13 m thick, using an airflow of 0.33 m3 s 1 m 2
(Agullo & Marenya, 2005) and three average temperatures (40,
50 and 60 1C). The temperature and relative humidity of the
ambient air were monitored using a thermohygrograph.
that, after desiccation, the plasma membrane is one of the
first points of damage. Ultrastructural analysis of the 2.2. Experimental dryer
endosperm tissues is essential to verify these works.
In a study of the sensitivity of coffee embryo to desiccation, The apparatus used in this study is shown in Fig. 1. It
Brandão Júnior (2000) found that greater ultrastructural consisted of a fan (A), an air duct (B), a plenum chamber (C)
damage, such as coalescence of lipids and probable rupture and a drying chamber (D) measuring 0.61 m by 0.61 m by
of the membrane, occurred in unripe than semi-ripe coffee 0.61 m. The plenum chamber contained a group of 3400 kW
berries. The degree of the damage was also influenced by the electrical circuits to heat the air. The drying chamber, which
coffee species. The study found that the species Coffea received four of the 12 samples dried on the ground, was
canephora was more sensitive to desiccation than the Coffea composed of four removable sections (Fig. 1E). Each section
arabica and its cells presented advanced deterioration of the received an average of 0.01 m3 of coffee.
membrane structures, even after reaching maturity. On the When the final moisture content of 11% (w.b.) was achieved,
other hand, the C. arabica seeds became more tolerant to the coffee was cooled with ambient air to interrupt the water-
desiccation during the maturation process. removing process. After cooling, samples were taken to
The beverage and presence of defects are the most determine the coffee’s final moisture content (Brasil, 1992).
important criteria to evaluate the quality of coffee. In Brazil, The dried coffee was then stored in polythene bags until
coffees are officially categorised by reference to a flavour scale cleaning and the histochemical and ultrastructural analyses.
(Brasil, 2003) as presented in Table 1. In this study, the defective grains were removed before the
The concentration of lipids in the endosperm tissue is analyses were carried out so that they would not interfere
related to the quality of the beverage. Studies have shown a with the results.
greater concentration of lipids in the coffee soft beverage with
well-defined droplets inside the protoplasts in the seeds’ 2.3. Histochemical and ultrastructural analyses
outer rims. Nevertheless, the lipids were observed to be
homogeneously distributed throughout the whole tissue For each drying temperature and solar drying period, samples
surface of the hard and rioysh coffee beverage (Goulart, composed of 10 parchment coffee beans were randomly
2002). In these types of coffee, the lipids fill the intercellular removed from the bulk coffee.
spaces without forming well-defined droplets. Despite these
works, ultrastructural analyses of the endosperm membrane
submitted to different drying temperatures are still lacking.
The objective of this study was to evaluate the effect of
different drying air temperatures on the resultant structure of
the endosperm cells of parchment Arabica coffee.

2. Material and methods

2.1. Experimental design Fig. 1 – Schematic of the experimental apparatus used to dry
the parchment coffee: (A) fan; (B) air duct; (C) plenum
The Arabica coffee, variety Catucai, was harvested manually chamber; (D) drying chamber and (E) removable sections of
using the stripping system. Four replications were done in 45 drying chamber.
 ARTICLE IN PRESS
64 BIOSYSTEMS ENGINEERING 99 (2008) 62 – 66

The samples were prepared and observed in the Cytology set up permanently in a Permalt environment. The ultra-thin
Laboratory in the Biology Department/UFLA (Federal Univer- sections were taken in golden slot grids and dried on
sity of Lavras). For histochemical analysis, the coffee beans aluminium racks covered with formvar. The sections were
were submerged in distilled water for 24 h at room tempera- post contrasted in uranyl acetate, followed by lead citrate for
ture. Cuttings of fresh tissue, obtained with a hand micro- three minutes, and then examined with transmission elec-
tome and kept in the open air, were treated for three minutes tron microscopes (TEM) Zeiss Mod. EM-109.
with the Sudan IV reagent in an ethanol solution at 80% to
visualise the lipids (Jensen, 1962).
3. Results and discussion
2.4. Preparation of the samples for scanning electron
microscopy (SEM) The ambient temperature varied from 17.2 to 21.5 1C and the
relative humidity from 48.5% to 81.5%, because drying was
Samples of parchment coffee were cut lengthways and carried out over 15 h beginning at 7:00 am. As a result, the
immersed in a modified Karnovisky solution (glutaraldehyde relative humidity of the heated air for the temperature of
2.5%, paraformaldehyde 2.5% in sodium cacodylate buffer 40 1C varied from 21.8% to 28.2%; for the temperature of 50 1C,
0.05 M, pH 7.2, CaCl2 0.001 M) and stored in a cold chamber it varied from 14.1% to 26.2%; and for the temperature of 60 1C,
until analysis. They were then infiltrated with a cryoprotector, it varied from 14.5% to 22.4%. The relative humidity for each
an aqueous solution consisting of 30% glycerol, for 30 min and level of heated air varied over a fairly narrow range and this is
transversally sectioned in liquid nitrogen using a scalpel not considered to have had a substantial effect on the results
blade. The cuts obtained were transferred to a 1% aqueous of this work (Borém et al., 1992; Afonso Júnior et al., 2004).
solution of osmium tetroxide for 1 h and were subsequently The results of the histochemical test with Sudan IV are
dehydrated for 10 min in a series of acetone solutions presented in Fig. 2. The scanning electron micrographs of
(25%, 50%, 75%, 90% and 100% three times) before being parchment coffee are presented in Fig. 3 and the transmission
taken to the critical point apparatus (Baltec CPD 030). The electron micrographs are presented in Fig. 4. These latter
specimens were placed on aluminium support stubs placed results not only confirm the light micrograph observations
over a film of aluminium foil using a carbon tape, sputter but add further relevant information. As the SEM results were
covered with gold (Baltec SCD 050) and observed in a LEO EVO carried out on dry seeds, any doubts as to the methodology
40 XVP scanning electron microscope (Leo Electron Micro- used in the preparation of the fresh cuts, such as the
scopy). Images were digitally generated and registered at possibility of the membrane ruptures occurring during the
varying magnifications in a set up to 20 kV and a working steeping process, are eliminated.
distance of 9 mm. The images were processed using the Corel There was a greater concentration of oils in a globular
Draw 9 Photopaint Software, where they were selected and shape inside the membrane of Arabica coffee parchment
arranged. beans dried at 40 1C (Fig. 2a). At this temperature, the oils were
preserved in this shape due to the integrity of the membrane
2.5. Preparation of the samples for transmission electron (Fig. 3a) and the vesicles remained undamaged (Fig. 4a). It is
microscopy (TEM) also possible to observe that the content of the cells remained
intact and the intercellular spaces remained empty. These
Samples of parchment coffee were cut lengthways and results show that there was no ultrastrucutral damage in the
immersed in a fixative solution (Karnovisky’s modified), pH endosperm cells of the parchment Arabica coffee dried with
7.2, and stored in a cold chamber until analysis. They were air at 40 1C.
then washed in cacodylate buffer 0.05 M, pH 7.2 (three times Fig. 2(b) shows some cells with oils in the extremities of the
for 10 min), post-fixed in 1% aqueous osmium tetroxide endosperm tissue while, in others, endosperm cells were full
solution for 1 h, washed twice for 15 min in distilled water, of cellular material, indicating a rupture of the vesicles and of
transferred to a 0.5% uranyl acetate solution for 12 h at 4 1C the plasma membrane when the parchment coffee was dried
and then washed once more in distilled water and dehydrated at 50 1C. At this temperature (Fig. 3b), in some cells the cellular
in a series of acetone solutions (25%, 50%, 75%, 90% and 100% content remained intact, while in others there was a rupture
three times). The dehydrated tissue was gradually infiltrated of the membrane, partial occlusion of the intercellular spaces
with spur/acetone, 30% for 8 h, 70% for 12 h and 100% twice and some vesicles were ruptured (Fig. 4b). This indicates that,
for 24 h each. The specimens obtained were set in moulds and at 50 1C, an initial disorganisation of the membrane system
polymerised at 70 1C for 48 h forming blocks that were used in occurred, leading to a loss of quality in Arabica parchment
ultramicrotomy. coffee.
The drying of parchment coffee at 60 1C (Fig. 2c) results in
2.6. Ultramicrotomy major ultrastructural damage. The vesicles and the mem-
brane were ruptured and the cellular material was spread
The blocks were trimmed. Thin (4100 nm) and ultra thin throughout the whole endosperm tissue surface. The oils
sections (o100 nm) were obtained using a diamond knife in a were not as well defined as they were at 40 1C, forming large
Reichrt-jung ultramicrotome. The thin sections were selected droplets in the intercellular spaces. At 60 1C (Fig. 3c), there
using a gold ring, placed in glass microscope slides, coloured was a complete occlusion of the intercellular spaces following
with toluidine blue (1 g of toluidine blue, 1 g of sodium borate the rupture of the cellular membranes. Inside the cells, a
and 100 ml of water), filtered in a Millipore filter (0.2 mm) and coalescence and rupture of the vesicles was observed (Fig. 4c).
 ARTICLE IN PRESS
BIOSYSTEMS ENGINEERING 99 (2008) 62– 66 65

Fig. 2 – Light micrographs of coffee seeds dried at 40 1C, (arrow) oils in globular shape inside the cells (a); coffee seeds dried at
50 1C (arrow) oils in the extremity of the endosperm tissue, occlusion of the intercellular spaces (b); coffee seeds dried at 60 1C
(arrow) oils spread throughout the cellular surface forming large droplets in the intercellular spaces (c).

Fig. 3 – Scanning electron micrographs of coffee seeds dried at 40 1C (arrow) empty intercellular spaces (a); coffee seeds dried
at 50 1C (arrow) partially occluded intercellular spaces (b); coffee seeds dried at 60 1C (arrow ) intercellular spaces completely
occluded (c).

Fig. 4 – Transmission electron micrographs of coffee seeds dried at 40 1C (a), at 50 1C (b) and at 60 1C (c).

These observations are essential to the preservation of the the putative oil bodies uniformly distributed in the
quality of coffee, as the rupture of the membranes can expose cell.
the oils to oxidation and rankness, which in turn leads to the (2) At 50 1C, an initial disorganisation of the membrane
formation of undesirable compounds that alter the coffee’s system occurred, leading to a loss of quality in Arabica
aroma and flavour. These results confirm the results of fatty parchment coffee.
acid analysis obtained by Marques (2006), who observed (3) The drying of parchment coffee at 60 1C leads to a
higher fatty acid values in coffee seeds submitted to higher coalescence and a rupture of the vesicles and of the
drying temperatures. membranes, causing occlusion of the intercellular spaces.
In this work, although different moisture contents were Independent of the temperature, the pre-drying period
used in the beginning of the artificial drying, no ultrastructur- does not interfere with the integrity of the cellular
al differences were observed in the coffee endosperm. membrane of coffee seeds.

4. Conclusions Acknowledgements

(1) The drying parchment coffee at 40 1C preserves This work was partially supported by grants from FAPEMIG,
the integrity of the plasma membrane, keeping CNPQ and Federal University of Lavras, Brazil.
 ARTICLE IN PRESS
66 BIOSYSTEMS ENGINEERING 99 (2008) 62 – 66

R E F E R E N C E S Bytof G; Selmar D; Schleberle P (2000). New aspects of coffee

processing: how do the different post harvest treatments
influence the formation of potential flavor precursors? Journal
of Applied Botany, 74, 131–136
Agullo J O; Marenya M O (2005). Airflow resistance of
De Grandi A M; Melo E de C; Berbert P A; Machado MC; Silva J de S
parchment Arabica coffee. Biosystems Engineering, 91(2),
E (2000). Eficiência da secagem de café (Coffea arabica L.), em
149–156
secador de camada fixa vertical com revolvimento mecânico.
Afonso Júnior P C; Corrêa P C; Goneli A L D; Vilarinho M C; Franc-a
In: Simpósio de Pesquisa dos Cafés do Brasil, Poc- os de caldas,
A S (2004). Dinc-âmica da contrac-ão volumétrica do fruto de
Anaisy Poc-os de caldas, Vol. 2, pp. 1085–1089
café durante o processo de secagem. Revista Brasileira de
Freire A T (1998). Projeto e avaliac-ão de um sistema para secagem
Armazenamento, 8, 29–35
combinada de café (Coffea arabica L.) despolpado. Dissertac-ão
Amorim H V (1978). Aspectos bioquı́micos e histoquı́micos do
(Mestrado em Engenharia Agrı́cola), 76p, Universidade Federal
grão de café verde relacionados com a deteriorac-ão da
de Vic-osa, Vic- osa, MG
qualidade. Tese (Livre Docência em Bioquı́mica), 85p,
Goulart P de F P (2002). Purificac-ão da polifenol oxidase e avaliac-
Escola Superior de Agricultura Luiz de Queiroz, Piracicaba,
ão de métodos bioquı́micos para aferir a qualidade da bebida
SP
do café. Tese (Doutorado em Ciência dos Alimentos), 80p,
Berbert P A; Queiroz D M; Silva J S; Pinheiro Filho J B (1994). Drying
Universidade Federal de Lavras, Lavras, MG
of coffee (Coffea arabica L.) in a fixed bed with a simulated
Guimarães A C; Berbert P A; Silva J S (1998). Ambient-air drying of
periodic airflow reversal. Journal of Agricultural Engineering
pré-treated coffee (Coffea arabica L.). Journal of Agricultural
Research, 59, 195–202
Engineering Research, 69, 53–62
Borém F M; Hara T; Silva R F (1992). Efeito da temperatura e da
Jensen W A (1962). Histochemical techniques. In: Botanical
umidade relativa do ar de secagem sobre o trincamento e a
Histochemistry, pp 206–256. Freeman and Company, San
suscetibilidade à quebra de sementes de milho, hı́brido AG
Francisco, USA
303. Engenharia na Agricultura, 1(2), 1–11
Marques E R (2006). Alterac- ões quı́micas, sensoriais e micro-
Brandão Júnior D da S (2000). Marcadores da tolerância à dessecac-
scópicas do café cereja descascado em func- ão da taxa de
ão de sementes de cafeeiro. Dissertac-ão (Doutorado em
remoc- ão de água. Dissertac-ão (Mestrado em Ciência dos
Fitotecnia), 144p, Universidade Federal de Lavras, Lavras, MG
Alimentos), 85p, Universidade Federal de Lavras, Lavras, MG
Brasil (1992). Ministério da Agricultura, Pecuária e Abastecimen-
Salazar G M R; Riano H N M; Arcila P J; Ponced C A (1994). Studio
to-MAPA. Regra para análises de sementes
morfológico, anatômico y ultraestrutural Del fruto de café
Brasil (2003). Ministério da Agricultura, Pecuária e Abastecimen-
Coffea arabica L. Cenicafé, Caldas, 45(3), 93–105
to-MAPA. Instruc-ão normativa n. 8 de 11 de junho