Diss. ETHNo. 13620
Investigations
Hot Air Roasting
on the
of Coffee Beans
A dissertation submitted to the SWISS FEDERAL INSTITUTE OF TECHNOLOGY ZURICH
for the degree of Doctor of Technical Sciences
presented by Stefan Schenker
Dipl. Lm.-lng.
ETH
born October 01, 1968 citizen of Däniken SO
accepted
on
the recommendation of
Prof. Dr. F. Escher, examiner PD Dr. G.
Ziegleder, co-examiner
Dr. A. J. Wilson, co-examiner
Zurich 2000
Ackno Acknowl wled edge geme ment nts s
I am most I
group
grateful
Dr F Escher for f or giving
to Prof
always appreciated
have
constructive evaluation
Further,
his I
me
the
opportunity
to work
far-sighted guidance, continued support
am much
indebted to Mr
his
in
and
Mayer-Potschak
K
and Mr G. Ludwig of G W Barth Ludwigsburg GmbH & Co., D-Freiberg/Neckar,
for their unlimited confidence and the t he generous funding of this work by G W Barth I
also
gratefully acknowledge the support of Kerne
(CH-Aarau),
Haco
(CH-Birsfelden)
AG
and
(CH-Gumligen), Migros
conducting industrial roasting trials Di Sano-Hofmann, Dr I
am
Betriebe Birsfelden
Migros Genossenschaftsbund (CH-Zunch)
to Mr Bruno Meier for his extensive
is addressed
A
AG
special thanks
help
and
project support
in
and also to Mr D Baumann and and Mrs Mrs Tabea
Fredy Nager,
Mr S GloorandtoMr T Kerne
Ziegleder (Fraunhofer Institute
G
very grateful to PD Dr
AG
Engineering
Food
Engineering and Packaging, IVV, D-Freismg)
for Process
and to Dr A J Wilson
(Centre for
Cell and Tissue Research, University of York, GB-York) for engaging themselves to be my co-examiners, for their critical
discussions
welcoming
I
also thank
us in York
My special thanks Perren,
who
by
and
extended to my
Dr Martin Müller
Dr Beat
Pompizzi
and his collaborators for
and
is
and supervisor Dr
supported
Rain Rainer er
my approach to the
directed at Stephan Handschin for his untiring
project contribution
(Laboratory
also
of Electron
Support
in
electron
Microscopy
1,
Institute for Forest, Snow
greatly acknowledged
for his kind cooperation and aid in
warmly
support
predecessor
Frey (Swiss Federal
Barbara Wunderli for her help I
are
and and substantial
Landscape Research, WSL) Dr Roberto
J Wilson
of my dissertation and fruitful
and his microscopic help and
Many thanks
work k microscopic wor
Zurich),
A
enthusiastically accompanied
world of roasting
microscopy
is
Dr
review
I
in aroma
am
ETH
and
grateful
analysis
to
and to
sensory analysis and data evaluation
would like to thank Dr. G Kahr
(Institute of Geotechnical Engineering,
ETH
Zurich) for his help with mercury-porosimetry and for putting the porosimeter at our disposal Many thanks also turn on Mr R Kunzli (DMP AG, CH-Hegnau) for his invaluable
services
well
Peter Bigler for his excellent mechanical work and solutions
as on
Many thanks also
during industrial measurements
to the
and
throughout the project
as
diploma students Stefan Hess, Angela Birchler, Cornelia
Heinemann, Matthias Huber
as well as
to all other students and collaborators who
contributed to this project Finally, my most sincere thanks go to all my colleagues who made
a
great, cheerful and inspiring work environment during the years
I
Table of contents
I
Table of contents
!
li
Abbreviations
III
Summary
IV
Zusammenfassung
1
Introduction
2
Literature review
V
VII
IX
1 ,
3 .
.
_____
3
2.1 Coffee in perspective 2.1.1
2.2
Taxonomy, appearance, cultivation
and
post-harvest processing
3
2.1.2 Historical, socio-cultural and economical aspects of coffee
5
2.1.3 Chemical composition of green and roasted coffee beans
6 9
Roasting technology 2.2.1 General considerations
on
9
roasting
9
2.2.2 Coffee roasting
2.2.3
Dehydration and
2.2.4
Appearance
and
chemical chemical reactions induced by b y roasting
general properties
of roasted coffee beans
2.3 Structural properties of the coffee bean
2.3.1 Morphology of the green coffee bean 2.3.2 2.3.3
Changes
of macrostructurc
Changes of cell and
13
15
15
during roasting
17
pore structure during roasting
18
2.4 Flavor profile of green and roasted coffee 2.4.1
12
Analysis Analysis of coffee flavor
20 20
2.4.2 Flavor of green coffee beans
22
2.4.3 Flavor profiles of roasted coffee
23
2.4.4
Staling
of roast coffee
27
11
Table ot contents
3
Experimental 3.1
3.2
___
_ _ _ _ _
29
Raw material
30
Roasting 3.2.1
29
30
Laboratory roasting trials
3.2.2 Industrial roasting trials
36
3.3 General analytical methods
38
3.3.1 Roast loss
38
3.3.2 Color
39
3.3.3 Water content
39
3.3.4 Extraction yield
40
3.3.5 Surface oil
40
3.3.6 Antioxidative potential
40
3.4 Characterization of structural and physical properties of coffee
41
beans
3.4.1
Volumetry
41
3.4.2
Mercury porosimetry
41
3.4.3
Dynamic mechanical thermal analysis (DMTA)
42 43
3.4.4 Electron microscopy
3.5 3.6
Gas desorption measurement and gas analysis
Analysis
of coffee aroma
compounds
and flavor
47
3.6.1 General methodological considerations
47
3.6.2 Isolation of the volatile fraction
47
3.6.3
Gas chromatography FID (GC-FID)
3.6.4
Gas
3.6.5
48
49 chromatography mass spectrometry (GC-MS) Aroma extract dilution analysis by gas chromatography olfacto¬ 50
metry (GC-O) 3.6.6
4
45
51
Sensory evaluation
53
Results and discussion
,
4.1 Characterization of process dynamics 4.1.1 Heat transfer and development of bean temperature
4.1.2
Dehydration and loss
of organic matter
53 53
.59
4.1.3
Development of bean color
66
4.1.4
Gas formation
70
4.1.5 Extraction yield
77
Table of contents
4.2
Changes
4.2.2 Volume increase
86
4.2.3
96
during roasting Structural changes during roasting Changes in porosity
Development 4.3.2 4.3.3 4.3.4 4.3.5
5
of aroma
compounds profile
HI
and flavor
120
120 Aspects of methodology Character impact compounds 124 Formation of aroma compounds during roasting 128 Influence of roasting parameters on aroma profiles 131 Influence of roasting time and temperature on sensory quality of the coffee beverage 137
of the roasted product during storage
Changes
141
4.4.1
Gas
141
4.4.2
Oil
144
4.4.3
Staling
dcsorption migration
148
Conclusions ,
6
79 79
4.3.1
4.4
of bean structure
4.2.1 Tissue structure of the green coffee bean
4.2.4 4.3
III
157
5.1 Critical process factors
157
5.2 Process optimizations
160
References
163
IV
**•
\
t"41%*
.
I
I
l
v
..
'
C \
"
v»
,
I it'
V
Il
Abbreviations
Air-to-bean ratio
ABR
Aroma extract dilution analysis
AEDA
Aroma
ANOVA C.
impact compound Analysis of variance
AIC
Cojfea
CH
Switzerland
Combined hedonic aroma response measurements
CHARM
Characteristic ion
CI CIE
Comission Internationale d'Eclairage
C02
Carbon dioxide
cryo-SEM Ü
Germany dry basis Dynamic mechanical thermal analysis Digital pressure and flow control Swiss Federal Institute of Technology
db
DMTA
DPFC
ETH F
France
FD-factor GB
Cryo scanning electron microscopy
Flavor dilution factor Great Britain
GC
Gas chromatography
GC-FID
Gas chromatography with flame ionization detector
GC-MS
Gas chromatography mass spectrometry
GC-0
Gas chromatography olfactometry
HL
Temperature profile high /low
FITST
High temperature short time
Internal standard
TStd
LHC
Temperature profile low
CSD
Least
LTLT MS
Low
02
significant difference test
temperature long time
Mass spectrometry
MTMT NMR
to high continuous increase
Medium
temperature medium time
Nuclear magnetic resonance
Oxygen
VI
Abbreviations
PC
Organic roast loss
ORL
Personal computer / workstation / notebook
incorporated proportional, differential and integrational parts PHL Temperature profile pre-heating / high / low PLHC Temperature profile pre-heating / low to high continuous increase prep-FIPLC Preparative high performance liquid chromatography PT-100 Electrical resistance temperature probe
PDI
RI
Retention index
RE
Reconstructed ion current (GC-MS)
RIC
Roast loss
O jls \2i
Simultaneous distillation/extraction
SEM
Scanning electron microscopy Stable isotope dilution assay Tertiary butyl methyl ether Transmission electron microscopy
SIDA
t-BME
TEM
Glass transition temperature
T to
t-test
Student's t-test
vpm
Volume per million
wb
wet basis
VII
III
Summary
Coffee is
one
of the most
the
harvesting
ripe coffee "cherries"
producer countries. operation from th e
important internationally traded food commodities. After processed
are
to
dry green coffee beans in
In the consumer countries, roasting is the most
the
important unit
in converting green beans into roast coffee with its specific flavor.
Apart
primary process objective of flavor development, it is important to generate
favorable bean
properties
project contributes influence
on
the
to
for
preservation
of quality during storage. The
the identification of
product properties
as
Roasting trials were mainly carried
a base
important process factors
present
and their
for process optimization.
out with
a
fluidized-bed hot air laboratory
roaster, allowing for coffee roasting under well-defined process conditions. The hot air temperature profile and the air
velocity were carefully controlled and, in addition
to batch pile temperatures, the bean
core
temperature
was
measured. Humid air
roasting and water quench cooling were operated optionally. with
sightglasses combined with
enabled trials
an
optical online observation of
on an
optical setup including a
single bean in
industrial scale were carried
the bean
laboratory trials.
a
roasting chamber
stereo
microscope
process. Measurements and
out in order to receive information on
dustrial roasting conditions, which served standard for the
A
starting point
as a
continuous
The structural, physical and chemical
changes of
as
a
and
during roasting were followed by volumetry, porosimetry, microscopy,
thermal and chemical
in¬
analysis. Instrumental aroma analysis
was
and
complemented
with sensory analysis. Green bean
quality and initial
\\ atcr content
in
particular have
a
major impact on the
process development and the resulting product properties. The temperature profile is the
most
formation affect
crucial parameter in the process
as
well
dehydration
as
design.
It determines both flavor
structural product properties. Different temperature profiles
and the chemical reaction conditions in the bean which control
Summary
gas formation,
expansion
browning
as well as
and flavor
VIII
development.
the structure resistance
opposed
A
to it
are
a
larger cumulated
pore volume and
bean
again temperature and
dehydration related factors. High temperature roasted beans exhibit volume,
for
force
driving
a
greater bean
larger cell wall micropores than low
temperature roasted coffee of identical degree of roast. These properties assumed to increase the undesired
mass
transfer and to accelerate the
arc
staling
process. Hot air
humidity must
be considered
as
yet another important process parameter
which influences the heat transfer rate and may affect various water content related
developments. to be
The amount of hot air in relation to the coffee batch size turned out
critical for roaster
coffee of
operation.
Low air-to-bcan ratios resulted in
superior cup-quality, whereas excessive
bland, dull and flat aroma
and
design
sensory properties. A lower ratio is assumed to prevent physical
stripping and excessive contact with
"microclimate"
air streams led to products of
enclosing
the beans. These
oxygen and may create
findings also stress
the
a
favorable
important role of
oxidation processes during roasting and storage.
Process
optimization requires specification of
because not all desirable
product properties
can
compromising target quality
a
be maximized at the same time.
High aroma quality is achieved with moderate roasting processes temperatures. Provided there is a
a
low air-to-bean ratio,
medium degree of roast should be 6 min
or
an
a
on
aroma
profile and result in excessive
porous bean structure which is
should operate with
a
for
the target flavor
very stable product during
lack of aroma strength. High temperature conditions
unfavorable
generally cause
gas formation and
a
very
impairing quality retention during storage. Roasters
fairly high proportion of conductive heat transfer and
air-to-bean ratios. For the most part, there may be
oxygen-free coffee technology. stage
a
medium
optimal roasting time
longer, depending
profile. Restrictive low temperature conditions yield storage, but
an
at
no
On the other hand,
may be worth to consider for
at
low
requirement for completely
an
oxygen-free final roasting
further im estigations.
IX
IV
Zusammenfassung
Kaffee ist eine der
wichtigsten international gehandelten Rohwaren.
Früchte des Kaffeebaumes werden noch in den Anbauländern
grünen Kaffeebohnen verarbeitet. In den Konsumentenländern
wichtigste Verarbeitungsschritt,
wobei Grünkaffee
zu
Die reifen
lagerfähigen
ist das Rösten der
ausgeprägt duftenden,
in
geschmackvollen Röstkaffee verwandelt wird. Neben dieser primären ProzcssZielsetzung ist die Erzeugung
von
günstigen Produkteigenschaften wichtig,
drohenden Qualitätszerfall während der
Lagerung entgegenwirken. Die vorlie¬
genden Untersuchungen leisten einen Beilrag Prozessfaktoren und deren Einfluss auf das von
die dem
zur
Identifikation
Endprodukt
als Basis
von
zur
wichtigen
Optimierung
Röstprozessen.
Röstversuchc wurden
vorwiegend
massstab unter exakt definierten
mit einem Heissluft-Fliessbettröster im Labor¬
Prozessbedingungen durchgeführt.
turprofil und die Luftzufuhr wurden
Haufentemperaturen wurde auch
die
genau
Das
Tempera¬
gesteuert. Neben den gebräuchlichen
Kerntemperatur der Bohnen erfasst.
Es konnte
wahlweise mit trockener oder feuchter Luft geröstet oder zusätzlich mit Wasser-
quenche gekühlt werden. Eine Sichtglas-Röstkammer kombiniert mit einem Stereo¬
mikroskop
erlaubte
Röstprozess.
Die
optische online-Beobachtungen
Messungen
und Versuche
an
einzelner Bohnen
Tndustrieröstcrn ergaben Daten
im zu
den industriellen Röstbedmgungen, welche als Ausgangspunkt und Massstab für die Laborversuche dienten.
Die strukturellen,
physikalischen und chemischen Verände¬
rungen der Bohnen wurden mit Volumetrie,
mischer und chemischer
Analyse verfolgt.
Porosimctrie, Mikroskopie,
Die instrumcntclle
ther¬
Aroma-Analyse
wurde durch sensorische Prüfungen ergänzt. Die
Rohstoffqualität
und insbesondere der Ausgangswassergehalt beeinflussen den
Prozessverlauf und die
Produkteigenschaften wesentlich.
gische Bedeutung kommt jedoch
dem
Temperaturprofil
Die
zu.
grossie technolo¬
Die
Rösttemperatur
Zusammenfassung
A
bestimmt die Aromabildung und die strukturellen dendem Ausmass. Sie beeinflusst den
Veränderungen
in entschei¬
Trocknungsprozess und bestimmt
fischen chemischen Reaktionsbedingungen,
die Bildung
von welchen
Bräunungsprodukten und Aromastoffen stark abhängig ist.
die
spezi¬
von Gasen,
Die treibende Kraft
zur
Volumenzunahme und der entgegengesetzte Strukturwiderstand sind ebenfalls
temperatur-
und
trocknungsabhängige Faktoren. Hochtemperatur-geröstete Bohnen
weisen im Vergleich
Tieftcmperatur-gerösteten Kaffees verstärkte Expansion,
zu
grösseres kumuliertes Porenvolumen und grössere Zcllwand-Mikroporen auf. Vermutlich fordern diese Fagenschaften einen unerwünschten Stofftransport bei der
Lagerung
und wirken sich
auf den
negativ
Alterungsprozcss aus.
Feuchtigkeit darf ebenfalls nicht vernachlässigt werden, und sich vermutlich
beeinflusst
auswirkt. Das Verhältnis
von
auf
LBV
Luftströme
ergab
Heissluftmenge
Produkte
generell
zu
von
hoher
Chargcngrösse (Luft-zu-Bohnen-
zu
klima"
und
um
und betriebliche Grösse. Ein
Aromaqualität, während übermässige
Kaffees mit flacher und aromaschwachcr sensorischcr
Charakteristik führten. Ein tiefes LBV schützt
Austrag
da sie den Wärmeübergang
wassergehaltsabhängige Röstvorgänge
Verhältnis, LBV) erwies sich als wichtige konstruktive tiefes
Die Heissluft-
vor
physikalischem Aromastoff-
übermässigem Sauerstoffkontakt und schafft
ein vorteilhaftes "Mikro¬
die Bohnen. Die Ergebnisse belegen die herausragende Rolle oxidativer
Prozesse während der Rostung und der Lagerung.
Prozess-Optimierungen erfordern eine kompromissbereite Festlegung
der Zielqua¬
lität, weil sich nicht alle im Produkt erwünschten Eigenschaften gleichzeitig maximieren lassen. Eine hohe Aromaqualitat wird durch moderate Prozesse mit
mittelhoher Tcmperaturfiihrung erzielt. Bei tiefem LBV soll die Röstzeit für einen mittleren
Röstgrad
6 min oder mehr
Bedingungen ergeben
Hochtcmperafur-Röstung
ein
zwar
stabiles, jedoch aromaschwaches Produkt.
bewirkt ein starkes, aber unvorteilhaftes Aroma, eine
übermässige Gasentwicklung sollten einen mittleren
betragen. Ausschliessliche Tieftemperatur-
und eine sehr
bis hohen Anteil
aufweisen und mit tiefem LBV
operieren.
poröse Bohnen struktur. Röstanlagen an
konduktivem
Ein vollständiger Ausschluss
stoff in der gesamten Herstellungstechnologie ist
Sauerstoff-freier letzter Rostabschnitt fur weitere gezogen werden.
Wärmeübergang von Sauer¬
unnötig. Hingegen sollte
Untersuchungen
ein
in Betracht
1
1
Introduction
Coffee presents
of the world's most favorite
one
for its delightful smell and flavor
While the
beverage
is consumed
as well as
beverages. It is greatly appreciated
for the stimulating effects of caffeine.
mainly in Europe, North
coffee plant grows at ele\ ated altitudes in
and Central America, the
tropical and subtropical regions
the world. More than 5 million tons of green coffee beans
worldwide.
Among
unique position with people earn
a
accomplished shipping.
internationally traded food commodities, coffee holds
the
greatest trade volume in financial terms. Some 20 million
living directly from coffee production. Post-harvest processing in the
producer countries, resulting
produce roast coffee
color as a
and
a
a
is
the most
important unit operation
traditional thermal process. Its
dry brittle texture.
The bean that is
in
composition
distinct quality
primary objective
exposed to roasting
natural complex "bioreactor" in which drying takes
chemical
is
in green coffee beans ready for
of the desired taste and aroma, but also to generate
and extensive chemical reactions
are
concerning
immediately after roasting
and
prevention of excessive
oil
dark
be regarded
place, water is redistributed
Roasting results
in
a
and flavor, texture, extraction
aroma
product
can
a
induced, causing profound changes of both
and bean microstructure.
appearance. Moreover, the
the
roasting
a
manufacturing.
Hot air roasting of coffee beans is is to
annually produced
all
In the consumer countries
roast coffee
are
all around
is
subject
yield
and
to substantial quality changes from
during storage. Therefore, migration
product of
the
protection
of aroma,
and the control of gas desorption during
storage presents another challenge in coffee technology. The behavior of products
during roasting and the resulting product properties important process parameters, such
as roaster
of the heat transfer media, cooling and water and interactions
occurring
in the bean
are influenced
by
a
series of
design, heat transfer, characteristics quenching. Since
during roasting
are
the
developments
inadequately understood
2
Introduction
the
roasting process
in
practice is still designed and operated mainly
on an
empirical
base. The present the coffee
investigation intended
roasting process. They
factors and their influence becomes as
to contribute to
possible
on
on a rational
fundamental
aim at the identification of
product properties
base.
so
important process
that process optimization
starting point and continuous standard. Consequently, some effort
scale
laboratory scale
roasting under well-defined
roasting equipment used in
(Perren, 1995)
was
adapted
cooling unit, allowing tural, physical in simulation
the
to coffee
was devoted
was carried
process conditions. The
preceding research project roasting, in particular with
on
to
out with
laboratory
nut
roasting
the addition of
a
quench cooling. Struc¬
changes were followed during laboratory roasting and
experiments using
the relations between
a
investigations
for efficient fluidized-bed and water
and chemical
insight into
Thereby, industrial roasting conditions served
monitor industrial roasters. The main part of th e
on
a more
the
technique of thermal analysis, thus establishing
roasting conditions and the resulting product properties. Based
preceding project
on nut
roasting (Perren, 1995) initial emphasis
was
put
on
coffee bean microstructure, using volumetry. porosimctry and microscopy. Investi¬
gations
on
coffee aroma, which is the most
outstanding product property of roast
coffee, were then introduced. Marked interactions between structure and physicochemical
developments during roasting
evaluated for process optimization.
and
storage could be established and
3
2
Literature review
2.1
Coffee in perspective
Taxonomy, appearance, cultivation processing
2.1.1
and post-harvest
The genus Coffea belongs to the botanical family oï. Rubiaceae and than 70 different and C.
liberica
species. However, only the three species of commercial
are
techniques some hybrids
of C. arabica and C.
duced with success. Since Coffea mid
importance.
was
first
are
Robusta for C.
C.
Robusta. The
canephora. Therefore,
a
C, arabica. C.
canephora
result of modem breeding
canephora have recently been intro¬
correctly described by Linnaeus in the
eighteenth century, botanists have failed
system. The most widespread varieties
As
comprises more
to
agree
Typica
precise classification
on a
and Bourbon for C. arabica and
canephora
is often
simply referred
to
as
geographical gene center of Coffea lies in the Abyssinian highlands
of
Ethiopia. The coffee
plant grows in tropical
and
subtropical regions of Central
and South
America, Africa and South East Asia, preferably in temperate and humid climates at altitudes between
of 2.5 to 4.5
m
600 and 2500
(C arabica) and
m. It
Each flower
on
a
4.5 to 6.5
growth conditions. Cultivated plants green leaves grow
is
are
shrub
m
(C. robusta). depending
a small
ellipsoidical stone fruit
green to red. The
wrapped
in
a
height
variety and
thin silverskin and
and
usually
(1988) and Clifford and Willson (1985).
approximately
15
mm
red exocarp (skin),
a
thick,
two seeds (coffee beans). Each
protected by
and fruit morphology has been described in detail
of
7 to 11 months, whereby its color
ripe cherry consists of a
gelatinous-pectic mesocarp (pulp)
seed is
on
a
generally kept at lower height. Oval shaped
length, called "cherry". The cherry ripens within
sweet
may grow to
the lateral branches together with clusters of white flowers.
develops into
changes from
or a tree that
a
parchment hull. This plant
by Illy and Viani (1995), Wriglcy
4
Literature review
The
two
C arabica and C.
species
canephora differ considerably
in their botanical,
genetic, agronomic, chemical and morphological characteristics. varieties
generally produce
(the central cut)
an oval convex
seed with
the flat side. C. canephora seeds
on
C. arabica
S-shaped longitudinal slit
an
arc more round
with
straight
a
central cut. C. arabica usually grows at higher altitudes than C. canephora and is
generally regarded
of
as
superior quality.
On the other hand,
resistant to pests and diseases. Illy and Viani ( 1995) provide
characteristics of the
Harvesting selective
two
or
Macrae, 1987 and process and is
the
the
separation of
wet
and
the beans from the
inexpensive.
on tiled or concrete terraces
layers
are
raked over at
outer
survey
on the
and
exposed
regular intervals
to
the
to
a
by
or
superior
subsequent crop
and is carried
out
by
1995, Clarke and
presents the most traditional
sun
are
spread
and air for
in small
drying.
The
prevent fermentation, and occasionally or
low temperatures. Fermentation in
optionally be included. After some four weeks
the cherries
are
dry
and the
shell has become dark brown and brittle. The husk is finally broken up in
dchullers and the beans
The
pulp
The harvested cherries
have to be covered to protect them from rain can
collected. The
process (Illy and Viani, 1995. Thorn,
layers
heaps
are
many other authors). The dry method
simple
is more
by non-selective stripping of whole branches
product quality because only ripe cherries
dry
a detailed
The latter is very labour intensive, but results in
hand-picking.
either the
canephora
species.
is carried out
processing includes
C.
wet process
arc
then stored in silos.
requires greater investment and more care, but
to better preserve the intrinsic
quality. In contrast
the
to
qualities of the bean and
dry method, during
from the bean prior to drying. As
machine, ideally within
parchment hull
are
is
parchment operation
the first 12 h after
are
completed is removed
is followed
sun
dried
harvesting.
the
a
superior coffee
pulp
is removed
is removed in
The
a
to
a
fermentation for
mechanically dried. At this stage, are
known
only before export by
by polishing, grading
a
as
"parchment
hulling
pulping
separated beans in their
essentially subjected
or
and the beans
pulp
generally believed
produce
processing
first step, the
washed and then
12 to 48 h. Then, they process
a
wet
to
is
or
the
wet
coffee". The
peeling step. This
and sorting, marketing and
shipping.
Literature review
2.1.2
5
Historical, socio-cultural and economical aspects of coffee
As mentioned in section 2.1.1, the coffee
Ethiopia, where the a
it still grows wild
discovery of coffee
hot beverage
as
and its
early
as AD
today. There
roasting
in the highlands of
plant originated are numerous
myths and legends
and brewing. Coffee is said to have become
1000. However, it
was
in Yemen, formerly called
Arabia, where spreading and horticultural propagation of coffee began in those
days, Yemen
Mocha,
was
was one
of the busiest
the centre (Thorn, 1995).
By
places in
and
AD 575.
the 13th century coffee
was an established was
from here that
began its great journey around the world. Via Mecca it first arrived
Constantinople (Istanbul), from where travellers brought
early 17th century. German, French, Italian
In
the world and its main port,
component of daily life and culture in Arabia (Heise, 1996). It coffee
on
at Cairo
it to Europe.
By
the
and Dutch traders introduced coffee to
their overseas colonies.
Coffee is
one
of the most
important internationally traded commodities
to have
the second largest trade volume
million
people worldwide obtain their income directly from
The annual coffee
production
1989 42.0 % of the world
Africa, 18.5 %
m
and is said
financial terms directly after oil. Some 20 the coffee
production.
is between 5 and 6 million tons of green beans. In
production were produced
in South America, 20.4 % in
in Asia and 17.9 % in North and Central America (D'Amicis and
Viani, 1993). Major
C. arabica
producer
countries
in 1993
were
Brazil
(1,275,000 t), Colombia (1.080.000 t), Mexico (184,000 t). Ethiopia (180,000 t), Guatemala (177,000 t), El Salvador (165.000 t). Costa Rica (148,000 t) and Honduras (121,000 t). Major producer countries that were
Indonesia
(169,000 t),
the
mamly cultivate
canephora
(441.000 t), Ivory Coast (200,000 t). Uganda (177.000 t) India
Philippines (111.000 t) and Cameroon (50.000 t) (Rehm
1996). Brazil mainly applies dry processing, whereas
processed
C.
Colombia
and
Espig,
produces
wet
C. arabica coffees.
Colombia is known
as
More than any other
the
largest producer of washed quality coffees in the world.
producer,
the country has been concerned to
develop
and
promote its coffee product and industry. This effort, together with favorable
geographical and climatic factors,
has
given Colombian coffee
its
reputation
for
Literature review
6
high quality and flavor. Colombian coffees generally provide good "body" acidity, rich flavor, and The
are
and
superbly balanced (Thorn, 1995).
highest coffee consumptions
are
coffee consumers in the world with
found in Europe. The Finnish an
are
the
biggest
annual per capita consumption of 12.6 kg
(D'Amicis and Viani, 1993). The coffee consumptions of the other Nordic countries are
also well above 10.0 kg p^1 yr l. while the figures
4.7
kg
for the United States and 4.4
kg
are
for Italy. Coffee
7.5
kg for Switzerland.
imports that
actually
are
consumed in Switzerland come to more than 56.000 tons, and the annual average
coffee
consumption amounts
2.1.3 Table 1
Chemical
to around 1000
composition of
provides a general survey
on
cups per person.
green and roasted coffee beans
the chemical
coffee beans (Illy and Viani, 1995). Other coffee components
Maier (1993). The
provided by Clarke
are
two
contains more caffeine and to
comprehensive data
a
and reviews
and Macrae (1985), Viani (1993)
species C. arabica and C. canephora
sition. Arabica beans contain more lipids,
during roasting lead
composition of green and roasted
sucrose
chlorogenic acids.
are
on
and
different in compo¬
and trigonelline, while robusta
The complex chemical reactions
totally altered composition of the roasted bean. The compo¬
sition in roasted coffee is highly dependent
on
the
roasting conditions and the degree
of roast in particular.
Lipids account
for 15 to
mainly triglycerides, C16.0 (25
...
the
I8g/100g(db)
of arabica beans. Coffee oil contains
principal fatty acids being C181 (40
...
45 g /100 g
35 g/100 g db). The lipid fraction also includes
a
relatively large
unsaponifiable fraction that is rich in free diterpencs (mainly cafestol The
nitrogen fraction
stimulating effects,
but is also
appreciated by consumers
subject to discussions
on health
has been devoted to the alkaloid caffeine. The acids of coffee
appreciable in quantity which is of chemical
Among them, its
risks,
high concentration in The
a lot
present
a
for
of work
fraction
and sensory interest (Maier, 1987).
the group of chlorogenic acids is the most remarkable
protective effect.
and kahweol).
of coffee includes caffeine, trigonelline, nicotinic acid, free
amino acids and proteins. Since coffee is very much its
db) and
one because
of
green coffee, and because of i ts antioxidative and cancer-
sensorially perceived acidity
is determined
mainly by acetic
7
Literature review
and citric acid. Melanoidins in roast coffee
1993). They constitute
a
are
poorly characterized
major heterogeneous group of brown
material that is formed at roasting. In contrast,
plished
so
to black
precursors in green beans and aroma
(Viani,
polymeric
lot of research has been accom¬
a
the volatile fraction of roast coffee. A literature review
on
far
compounds
in roasted coffee is
on aroma
provided
in
chapter 2.4. Oligosaccharides and Polysaccharides constitute about one half of the The polysaccharides present the
matter (Viani, 1993).
elements of the cell. Therefore, their for the
development
studied
extensively
composition
in the 1960s
by Thaler
dry
principal structure building
and fate
of bean microstructure. Coffee
raw bean
during roasting
is crucial
polysaccharides have been
and Arneth (1968a.
1968b, 1969) and
Thaler (1975), and other authors. Thaler's group found four different fractions in green beans, composed of mannan, cellulose, galactan and araban. More recently,
Bradbury
and
Halliday (1990), using high resolution GC-MS, identified cellulose,
mannan
and
galactan
was described as
arabinogalactan
short side chains linked
as
the
principle polysaccharides
principally ß( at
C6 to
1
—>
3) linked galactan chain with frequent
galactose residues
arabinosc residues. Mannan has been defined with only about 1 one-residue structure models were
(1999),
who
employed
Arabinogalactan coffee. Mannan
small
amounts
polysaccharides
partially criticized in
and
as a linear
at C6 per a
1
3 linked to terminal
->
ß(i
-->
4) linked mannan
100 mannose residues. These
more recent
study by Navarini
and Mannan were isolated from hot water extracts of dark roasted
was described as a branched
of
galactose
are
ß( 1
and arabinose
structurally related
to
—>
(an arabinogalactomannan). Both
those originally present in the green
arabinogalactan appears
Halliday, 1990). Yet,
4) -D-mannan substituted with
to
be more altered
it is not clear if the
two
by roasting
polysaccharides
in the
isolate
are
form
complex assembly. Under the latter hypothesis, proteinaceous material
play
a
an
individual components of
important role (Na\ armi
roasting considerably reduces
galactomannans
et al.
NMR spectroscopy in combination with classical methods.
coffee beans, even if the
(Bradbury
galactose stub
in coffee. Arabino¬
a
et al..
physical mixture,
if they
are
associated to may
1999). Leloup and Liardon (1993) found that
the molecular
in coffee cell walls.
or
weight range
of
arabinogalactans
and
8
Literature review
Tab, 1: Chemical composition of raw and roasted coffees in g /100 g db (Illy and Viani,
1995).
Component
Arabica coffee
green
Polysaccharides
49.8
Robusta coffee
roasted
green
38.0
54.4
roasted
42.0
Sucrose
8.0
0
4.0
0
Reducing sugars
O.l
0.3
0.4
0.3
Other sugars
1.0
data
2.0
Lipids
no
16.2
17.0
no
fO.O
data 11.0
Proteins
9.8
7.5
9.5
7.5
Amino acids
0.5
0
0.8
0
Aliphatic acids
1.1
1.6
1.2
1.6
Quinic acids
0.4
0.8
0.4
1.0
Chlorogenic acids
6.5
2 5
10.0
3.8
Caffeine
1.2
1.3
2.2
2.4
1.0
1.0
0.7
0.7
4.2
4.5
4.4
4.7
traces
0.1
traces
0.1
Trigonelline (including roasted
by-products) Minerals (as oxide ash) Volatile aroma
Water
8
12
to
0to5
8 to 12
0to5
Caramelization and condensa¬ tion products
(by difference)
25.4
25.9
Literature review
Roasting technology
2.2 2.2.1
9
General considerations
Roasting
is generally defined
as a
generate roast aroma compounds, texture. These intentional
roasting
on
dry heat treatment
develop color,
to
product alterations make
roasting and simple drying (Perren, 1995). Heat
goods by different modes. of coffee beans is as
hot air
Roasting
In differentiation to
mostly regarded
as to
frying
a
crispy
explicit difference between
the
be transferred
can
be carried
and often to create
roasting nuts
or
out in
a
to
the
roasting
in oil. roasting
gaseous atmosphere such
or steam.
is
applied
and other oil
to
a
number of foodstuffs, such seeds. It is
containing
a
usually involves dehydration, reaction
peptides with free monodenaturation and
2.2.2
of foods with the intention to
as
cocoa, nuts, chicory, coffee
time-temperature controlled
process that
of free amino acids and short-chained
and disaccharides during nonenzymatic browning, protein
subsequent changes
in texture (Perren. 1995).
Coffee roasting
Process
Roasting
is the most
important unit operation
flavor-full roast coffee. The
in
primary objective
converting green coffee beans into
of the process is to
produce
taste and aroma. Furthermore, coffee is roasted to generate a dark color
and brittle texture that makes
grinding
and extraction
1987, lohannessen, 1992). For coffee roasting than 190 °C
are
summary table
Types
required (Dalla Rosa
on
the
m
possible (Clarke
a
desired
and
a
dry
and Macrae,
particular, temperatures higher
et al., 1980). Illy and Viani (1995) provide
macroscopic effects
of
roasting
on
a
the coffee bean.
of roasters
The various
principles
of
roasting systems
can
be
grouped regarding different
criteria: Product flow Coffee beans
can
be roasted in batch, usually with industrial batch sizes of some
hundreds of kilograms,
or
in continuous systems. Continuous roasters
are
generally
10
Literature review
designed for large hourly capacities, whereas batch roasters provide more flexibility in process layout and control.
Mechanical principle The most commonly used systems
are
found to be the horizontal rotating drum, the
vertical fixed dram with rotating mixing elements, the vertical rotating bowl and the fluidized-bed. The
mam
task is to provide means for sufficient mixing of the beans
in order to achieve homogeneous roasting and to prevent scorching of beans. Clarke
and Macrae ( 1987) provide
summery of different industrial roasters.
an illustrated
Heat transfer
FIcat
can
be transferred to the beans
metal surfaces,
by free
or
by heat conduction
forced convection due to
a
at direct contact with
streaming media (hot air),
by radiation. Roasters generally make use of all three types relative contribution to the overall heat transfer may infrared roasting has been
reported (Kino
hot or
of heat transfer, but their
greatly differ. Although
and Takagi, 1995), this method is very
unusual for coffee. Since coffee is exclusively hot air roasted in industrial practice, it makes
sense
to
limit distinction to systems with
transfer and systems with
prevailing conductive heat
prevailing convective heat transfer.
also very useful to consider the
In this respect, it is
operating air-to-bean ratio.
Air-to-bean ratio The amount of hot air used in
beans is defined
a
roasting process in relation to the batch size
by Mahlmann (1986)
as
air-to-bean ratio
coffee). This ratio is a characteristic parameter in for
a
given degree
a
air per
roasting process,
of roast. According to Mahlmann,
typical "conventional"
(kg
figures
can
of coffee
kg green
but only applies
range from 1 in
a
process up to 150 in fully fluidized-bed systems.
Process factors of major importance The
quantity
of the time
of heat transferred to the beans presents the most
roasting process.
(Illy
It
can be determined from
and Viani, 1995). According to
in the product is correlated
a
important parameter
the bean temperature and roasting
widespread opinion,
the
degree of roast
to the final roasting temperature (Sivetz. 1991, Illy
Viani, 1995). During the last decade, the
and
time/temperature profile has been the most
11
Literature review
extensively discussed issue was
in coffee roasting. Early traditional industrial roasting
carried out with conductive type equipment, applying slow heat transfer with
long roasting times of more than 20 min. roasters enabled direct contact of beans
The introduction of gas fuel
operated
with combustion gases and allowed for
much faster heat transfer and fluidized-bed roasters (Illy and Viani, 1995, Sivetz,
1975). During the 1970s and 1980s, there
roasting times
cut down to less than 90
benefits, since this process to give
a
was
regarded
was even a trend to ultrafast s.
Inventors claimed process and
as more
considerably from that al.
in
out
1975 and 1991, Hubbard et al.,
1979, Stefanucci and Protomastro, 1982. Small and Horrell. 1993,
However, low density coffee did also cause
product
efficient, economic, and turned
low-density high-yield product (Sivetz.
The entire microstructure of low
roasting with
a series
and others).
of troubles and reservations.
density high yield coffee beans
was
found
to differ
"regular" coffees (Kazi and Clifford, 1985, Puhlmann
1986). Greater volume increase and more intense
gas formation created
et a
packaging problem (Radtke, 1975). Moreover, fast roasted coffees exhibited greater oil
sweating which
was
regarded
addition, these products have are more
a
to
be
a
sensory risk (Puhlmann
et
al, 1986). In
somewhat higher final water content. Hence, they
by oxidation and staling during storage (Radtke, 1979, Radtke-
affected
Granzer, 1982. Hinman, 1991). Last, but most important, high yield roasting has been
optimized organoleptically (Illy
and Viani,
1995). High yield coffees
not
gave
infusions that were bitter, burnt and astringent (Kazi and Clifford, 1985, Illy and
Viani, 1995). For all these reasons, ultrafast roasting has been widely abandoned in industrial practice in recent years. Roasting times of more than 4 min
are
commonly
applied again today. Still, empirically optimized temperature/time profiles vary
considerably from manufacturer
to manufacturer and
questions must be further investigated more fundamental
The
scientific
roasting process must
1995). This water
is
sprayed
on
well
kept secret. These
process development
can
be
stopped by rapid cooling
put
on a
of the beans (Illy and Viani.
air and/or
a
precise amount of
the hot beans (water quench cooling). The water is
on
be
understanding.
generally achieved by excess cold
fully evaporate content. This
so that
are
the bean surface rather than to
cooling process makes
use
of the
greatly influence
supposed
to
the bean water
high evaporation enthalpy
of water.
12
Literature review
2.2.3
Dehydration and chemical reactions induced by roasting to Illy and Viani (1995) the
According
drying phase,
a
roasting phase, where
a
take place, and
roasting process a
roughly divided into
be
can
number of complex chemical reactions
final cooling phase. During roasting the beans loose weight.
a
14 and 20
generally between
%, depending
on
the green bean quality, the process
conditions and the target degree of roast (Clarke and Macrae, 1987). A of this weight loss is due to
5
to
as
8 % for
CO2.
a
medium
dehydration, whereas another substantial part (some
also result in the formation of
widely regarded claimed
a
as
development
dehydration
organic matter into
and
considerable amount of water that is then
a
are
water
and 30 % carbon dioxide. Dehydration is
100
°C,
a
final stage of maximal
dehydration rates
a
first stage of
due to microstructural
are endothermic, whereas a number
roasting stage.
as
Viani at
a
a
bean temperature of 160 °C. whereas Streuli
Although Raemy
Nicoli et al., 1997) do
not show an increase
et
and Lambelet (1982)
reported
in
al, 1991. Illy and Viani, 1995,
in the final
roasting stages that
place during roasting have
not
can
be
yet been completely
elucidated, the reasons for this being great difficulty in reproducing the reactions that take can
to
to exothermic reactions.
The chemical reactions that take
information
150 °C.
(1973) reported
the beans, the few temperature curves
literature (Puhlmann and Meister, 1989. Da Porto
clearly attributed
to become exothermic at
(1993) claim that the process changes from endothermic
self-heating effect in
an
temperature of 140 °C. Illy and Viani (1995)
exothermic reactions to start at 190 °C. a
of authors state
On the basis of calorimetric measurements, Baltes
and Lambelet (1982) found
exothermic
claimed
and Meister (1989)
stage of accelerated but migration-limited
(1977) reported the n et result of reactions in coffee
well
again lost
of the bean.
exothermic final
as
gaseous products
of water release in three stages. They found
At first, chemical reactions
Raemy
loss of dry matter, primarily
steady process. However, Puhlmann
a
dehydration below
changes
a
a
vapor (Clarke and Macrae. 1987). Illy and Viani (1995) reported that 70 %
of the degradation products
slow
of roast) is caused by
degree
The chemical reactions that convert
as water
major part
or
simulating
all
place inside a bean in the laboratory. Nevertheless, significant
be obtained
by comparing
the
compositions
of green and roasted
13
Literature review
coffee
(Illy
and Viani. 1995, Clarke Macrae.
1985, Viani, 1993). Some of the more
extensive and complex chemical reactions during roasting affect the of green beans and include Maillard reaction, Strecker
carbohydrates
degradation, pyrolysis.
caramelization, mainly resulting in aroma, flavor and color compounds. Roasting leads to protein denaturation and
proteins with free amino
degradation.
groups react with reducing sugars to form glycosylamines
and/or aminoaldoses and/or aminoketones
by condensation. Amino acids react with
a-dicarbonyls during Strecker degradation 1995).
On
roasting there
Free amino acids, peptides and
is
a
and form aminoketones
(Illy
and Viani,
reduction in the amount of citric and malic acid and
an
increase of many of the other acids, in particular quinic acid and volatile acids
(Maier, 1987). Chlorogenic acids loss is about
proportional
coffee. Caffeine is
during aroma
to
2.2.4
strongly degraded (Leloup
the degree of roast and
is discussed
Appearance
are
little affected
separately
in
by roasting.
The formation of
and general properties of roasted coffee beans
distinguish themselves by a certain
proceeded by roasting, there
criteria and définitions for the roast loss may serve
as an
degree of roast.
indicator for the
qualitatively determined
more
partially degraded
of roast". While it basically means the extent of
which the beans have
The
is
chapter 2.4.
In contrast to green coffee beans, roasted beans
"degree
et al., 1995). The
80 % in dark roasted
can reach
thermally quite stable, whereas trigonelline
the process. Triglycerides
compounds
are
suitable in industrial
or
are
roasting,
several different possible
The overall
degree
and the state into
weight loss
of roast for
a
visually assessed external color
given
or
the
organic
raw material.
of the beans is even
practice (Clarke and Macrae. 1987). Color changes
progressively during roasting from greenish-grey
to
a
marked
brown, dark brown and almost black. Moreover it is said
to
yellow, orange,
be correlated with the
bitter/acid ratio in the cup (Illy and Viani, 1995). Also for scientific purpose the instrumental color measurement is measure
commonly regarded
of the degree of roast. However, color is
of ultrafast
the most
as an
indicator for the
degree
some
appropriate
less reliable indicator in the case
roasting, since the interior of the bean is less roasted than
and Viani. 1995). On the other hand,
properties
a
as
authors also
of roast, such
as
the
the outside (Illy
suggested chemical
methylpyrazine ratio
14
Literature review
(Hashim and Chaveron, 1996), isomers
of quinic acid (Scholz-Böttcher and Maier,
1991) and the ratio of certain amino acid enantiomers (Nehrig and Maier, 1992). The ability to retain the gases formed during roasting presents remarkable
properties of coffee beans.
of the
one
It is well known, that roasted whole beans
contain large quantities of entrapped carbon dioxide that is only released more
during
than 4 months of storage (Clarke and Macrae, 1987, Radtke, 1975). The
amount of
gas development is
dependent on the degree of roast. According to Sivetz
and Desrosier (1979), about half of the total whole bean. Even
though, measured
pressure), whole beans contain
a
quantity
C02 must
roasted bean, which for
typical case
a
is retained in the roasted
C02 generated
at standard conditions NTP
and Macrae. 1987). This
to
most
of
approximately
(20 °C, 101.3
2 to 5 ml
kPa
C02 (Clarke
be held under considerable pressure within was
a
calculated by Clarke and Macrae ( 1987)
be 6.4 at (648 kPa). Radtke (1975) calculated even higher pressures of 800, 570
and 550 kPa, respectively, for 3 different fully roasted coffees in the cold state. A
substantial part of the
entrapped
gas is
only lightly bound in
the bean, since it is
easily released during grinding. The gas desorption process during storage is often
accompanied by migration
coffee oil to the bean surface. The extent of oil migration is bean beans
quality and possible pre-treatments, such
as
are known to be more delicate to roast, since
(Lee, 1999).
dependent
on
of
the green
dec a ffei nation. Decaffeinated
they tend
to more "oil-sweating"
On the other hand it is also known that oil migration in decaffeinated
beans is controllable by the roasting conditions and the target degree of roast. Darker
roasted beans tend
roasting
is
to
regarded
a more severe
as
oil
migration. Applying intensive heat during
migration promoting
and detrimental to the roast coffee
quality. The mechanisms of this mass transfer have not been
extensively investigated
so
are
poorly understood, since they
far.
The amount of dry matter that is transferred into the coffee on a series
of parameters, such
of roast and the
as
variety
roasting temperature,
procedures (Clarke
and
origin
as well as
raw
the conditions
and Macrae. 1987. Nicoli
Extraction yields greater than 50 %
of th e
beverage
el al..
is
dependant
material, the degree
during
1990. Hinz
the extraction et al.,
are achieved in industrial extraction
1997).
technology
by applying high pressures and temperatures. Conventional home-brewing leads
to
15
Literature review
an extraction
complex and
yield below so
30 % (Peters, 1991). The extraction mechanics
far authors have failed
to agree on
a
are
very
commonly accepted model of
this process.
2.3 2.3.1
Structural properties of the coffee bean of the green coffee bean
Morphology
Green coffee beans do
not exhibit
strated in
specific folding, recognizable
Figure I,
a
a
uniform and homogenous morphology. As illu¬
itself, creates the typical shape with
the central cut
In the middle part of
The cytoplasm of the
hydrates
and
is embedded the small
being folded
one
appreciable
amounts
(Bürgin, 1969,
single layer of epidermal
distinguish a thin layer
of mucila¬
embryo.
are
of caffeine,
distributed
stored within numerous oil bodies with
al., 1997). These oleosomes
are
a
chlorogenic
acids and minerals
homogeneously throughout
located close to the plasmalemma. forming
a
diameter range of 0.2 to 0.3
remarkably stable
oleosins. The stability of oleosomes
important role
seems to
cytoplasm is free of lipids described some sort of
a
play
an
um
parenchymatous cell walls
or
layer of proteins, called
and seed imbibition. The
in the
center
and contains proteins and carbohydrates. Dentan
vacuole in the
are
(Wilson
and do not aggregate a
physiology during lipid biosynthesis
the bean and
layer of variable thickness. They
coalesce (Huang, 1996). Their surface is shielded by
The
upon
parenchyma cells essentially contains lipids, proteins, carbo¬
(Dentan, 1985). The lipids
et
slice
parenchymatous storage cells (Dentan, 1985).
a transverse section one can
ginous material in which
a
the flat side
on
Dentan, 1985). At the periphery of the seed, there is cells. The main bean part consists of
as
plant
of the
(1985)
cytoplasm filled with carbohydrates.
of ripe coffee beans
are
particularly thick and do
not
enclose any intercellular spaces (Dentan, 1985). Reinforcement rings give them
a
nodular appearance in cross sections. The bulk of the full grown cell wall consist of the
secondary wall (Dentan, 1977).
In certain areas the cell wall is crossed by many
plasmodesmata (Dentan, 1985). Wilson SEM and found green beans.
no evidence
et al.
of additional
(1997) analyzed freeze-fractures by
pre-existing channels within
They observed cellulose microfibrils
and described them
the wall of
as
organized
16
Literature review
in
polarized orientation by FF/TEM.
general model concept of the organization
network of polysaccharide microfibrils that is stabi¬
of the
plant cell wall suggests
lized
by proteinatious cross-links
a
The
and embedded in
a
gel of pectic-cellulosic
material (Nultsch, 1996, Wilson and Fry, 1986). This complex cell wall architecture has been remarkably visualized with light micrographs of the onion primary cell
wall in
a
study by McCann
polymers from
et al. (1990). They sequentially extracted
the native wall and analyzed the remaining structure in the microscope. Both above
mentioned studies do useful hints
on
the
not directly apply to coffee beans.
Nevertheless, they give
general structural architecture of plant cell walls.
Fig. 1: Schematic transverse section of a coffee bean (Dentan, 1985).
folding gives the bean
its typical shape with
The bulk of the bean consists of
a central
cut
parenchymatous cells.
on
A specific
the flat side.
17
Literature review
2.3.2
Changes of macrostructure during roasting
The volume increase presents the most obvious structure
during roasting. Clarke
and Macrae (1987) described bean
occurring progressively, but including decrease of density. It is
not
decrease i s
a
a
a
"popping phase", leading
expansion
to sounds
accompanying
the
degree of roast,
and Macrae, 1987). Dalla Rosa et al.
but also of the
expansion,
due to
information
on
a
initial bean water content
together with these data. Guyot case
of rapid roasting
degree
different
a
60 % at
including
was
Guyot
et al.
a
very
steady and continuous
positively correlated a
weight loss
to
weight
of 18 %. No
was
provided
greater expansion in the of the same
of the roasting temperature
clearly confirmed in
a
on
the
comprehensive
C. arabica and C. canephora beans from six at
temperatures of 220, 235, 250, 265,
identical degree of roast. Values of relative volume increase
ranged for Colombian Arabica coffee from 1.59
Uganda from 1.37
a
roasting temperature
or
origins. These coffees were roasted an
to
significant influence
et al. (1998)
280 and 295 °C to
be
et al. (1985) reported even
development. This relationship
study by Ortolâ
to
by
vapor and gas within the
high temperatures. Comparing final products
at
of roast, he found
volume
development
of the beans of 40
swelling
if the
speed of roasting (Clarke
bean (Illy and Viani, 1995). Illy and Viani (1995) reported
loss. They noted
or
(J 980) found that the resulting bean volume is
rapid formation of water
volume increase and found the
the term
and density
correlated with the final roasting temperature. Bean expansion is caused
rapid pressure build-up
as
to considerable
phase of instantaneous expansion. Volume increase
function of the
of the bean
quite clear from this statement whether
"popping phase" applies only authors suggest
macroscopic change
to
1.84, and for Robusta from
to 1.55.
(1985) regarded
the maximization of bean
expansion
as
beneficial for
quality. Also Small and Horrell (1993) aimed for maximum volume increase in order to
produce high yield coffee. They reported that fast roasting (1
...
3 min) of
pre-dried (<
5 g /100 g
beans with
high extraction yield. Their physico-chemical model concept of bean
a
wb) coffee beans leads
expansion features the chlorogenic detected
a
sharp decomposition
temperature range close
to
acids
as
to
greatly expanded
key-components, since
of these acids with substantial
or
"puffed"
the authors
C02 evolution
in
a
the glass transition temperature Ttt of the bean, tnstanta-
Literature review
neous
18
pressure build-up
during
a
softened stage of the bean would result in
"puffing" effect. Small
and Horrell (1993) also realized the
the bean water content.
They suggested
in order to allow for more
Concerning actual values cation
by Brandlein
around 216 °C.
theory
significant influence of
drying step outside
aggressive fast roasting conditions within
of Ts, they refer to
et al. (1988).
The
They described
a
value mentioned in
a
the roaster
the roaster.
patent appli¬
patent authors stated T2 of coffee beans being
the softening effect of water in glass transition
and attributed greater expansion of fast roasted beans to the
contents retained
2.3.3
to move the
a
higher water
during high temperature roasting.
Changes of cell
and
pore structure during roasting
Chemical reactions, dehydration and the large volume increase
accompanied by profound structural changes of green beans
are
of both the cell wall and the cytoplasm
of the green bean. Wilson et al. (1997) reported the
cytoplasmatic matrix
during roasting
proteinaceous/polysaccharide
starts to denature after
roasting. Oil droplets coalesce and finally form
a
the initial stages of
layer that "flows" around the inner
surface of the cell wall. A further scries of publications dealt with the microstructure in
fully roasted beans. Roasted bean tissue presents excavated cells with the,
glance, unaltered cell walls building
a
at first
framework. This structure has been exten¬
sively described using light microscopy, scanning electron microscopy (SEM) and
image analysis (Bürgin. 1986, Massini
1997).
et
J 969, Dentan, 1977, Dentan and Illy, 1985, Puhlmann et al.,
al, 1990. Gutierrez
The voids of excavated cells
possible major tissue cracks, make
et al., 1993, can
be
Uly
regarded
up for the main
and Viani, 1995, Wilson et al., as
macropores and, apart from
part of the bean porosity. Radtke
(1975) reported porosity values in roasted beans ranging from 0.38
depending on different
(21
...
23
the
to 0.49 um
origin and pretreatment of coffee. Kazi and Clifford (1985) found
average cell sizes
for
"high yield" (34...
pm) coffees, respectively. Massini
hensive investigation
interpret. Gutierrez
on coffee bean
and
"regular"
development
reported the entire bean surface
be cracked after 10 min of roasting. However, their to
pm)
et a l. (1990) described the
of pores in the course of roasting using SEM and
surface seem to be difficult
40
to
micrographs of the roasted bean
et al.
(1993) presented
porosity. Various physical methods
SEM and image analysis were used to determine the
porosity
a
compre¬
as
well
as
of coffees roasted at
19
Literature review
different
coffee
temperatures
to
the same
found to have
was
low temperature roasted
a
products.
7«/cropores
et al., 1997).
in the cell wall
as
measurements
the
porosity of the cell wall (Salceb,
al, 1990. Gutierrez
et al.,
by roasting conditions.
a
that the macropores of roasted beans
nm.
two different
accessible
radius) which
quality.
types of micropores of
in pore structure have
The pore structure controls
may determine the
high
mass
allow the mobilized coffee oil
1995, Wilson
staling process
are
a
average radius of 50
major impact on
nm
et
the final
product
transfer phenomena during storage and
gas adsorption capacity and the
(Saleeb, 1975, Radtke, 1975, Massini
and Viani,
an
respectively.
Roasting-induced changes
the
nm
are
so-called ink-bottle structure. In contrast, Wilson et al. (1997), using electron
microscopy, found and 5
The fate of the
is unknown. Saleeb (1975) concluded from gas
through very narrow micropores of molecular magnitude (2.8 form
1993, Illy and
So far, very little is known about the formation of
affected
pfasmodesmata during roasting adsorption
roasting alters
et al.. 1986. Massini et
Viani, 1995, Wilson
of roast. Again, high temperature roasted
statistically significantly greater macropore area than
A number of authors assume that
1975, Puhlmann
degree
et al.
gasdesorption properties
1990). Fine micropores
are
assumed to
to migrate to the bean surface (Puhlmann, 1986,
Illy
al, 1997). Moreover, the loss of aroma compounds and
probably related
to microstructurc (see chapter 2.4.4).
20
Literature review
2.4 2.4.1
Flavor profile of green and roasted coffee
Analysis
Since the
of coffee flavor
of roasted coffee is based
aroma
compounds that occur only in traces and
methodology
is
on
are
on
an
on
coffee
aroma.
a
of organic
sophisticated
Although instru¬
incredible pace during the last three
decades, the investigation of complex food
Generally, it involves
complex mixture
volatile by nature,
required for qualified research
mental analysis has been advancing
a
aroma remains
a
demanding task.
the following steps (Marsili, 1997):
Isolation and concentration of volatiles
Separation Identification
Quantification
Investigation of
sensory properties and impact
(Validation of analytical results with
on human aroma
the help of models)
The methods used for isolation of food flavor compounds
result of
an aroma
factors, such
as
analysis. Sample preparation
is
high complexity
of aroma
Fleadspace analysis and distillation techniques
are most
critical for the
complicated by
low concentration levels, variation of
volatile interactions and the
perception
are
a
number of
volatility, instability, matrix-
composition (Marsili, 1997). the
most
suitable isolation
methods (Clarke and Macrae. 1985). Sohent extraction and vacuum distillation
are
commonly used distillation techniques. However, each method implies preferential isolation of some
recommended to
compounds and discrimination of others. Therefore, use at
it is strongly
least two different isolation methods in order to be able to
compare the results (Marsili, 1997).
The simultaneous distillation/extraction (SDE) according to Likens and Nickerson
(1964) is
one
of the most
widely used and valuable solvent extraction techniques
roast coffee. The SDE apparatus provides for the simultaneous condensation steam distillate
and
recycled, and thus
an
immiscible organic solvent. Both liquids
the steam distillablc. sohent soluble
from the aqueous phase
successfully applied
to
the solvent (Marsili,
to coffee
in
a
arc
compounds
for
of the
continuously
arc
transferred
1997). This method
has been
series of investigations by various authors (e.g.
Literature review
21
Bade-Wegner et al, 1993 is
and 1997, Holscher et al., 1990, Vitzthum et al.
1990).
It
convenient, requires simple handling, gives good recovery and limits time
consumption (Holscher method
was found
and Steinhart,
to be the
1991). One of
relatively great heat impact
the
disadvantages of the
on
the
sample that might
generate artefacts.
Vacuum
distillation,
accurately described
more
as
direct solvent extraction with
subsequent high vacuum transfer, presents another widespread isolation technique
applied
on
coffee (e.g. Clarke and Macrae. 1985. Blank et al.. 1991 and 1992,
Holscher et al, 1990 and 1991). A solvent extract is obtained from The aroma fraction is then a
high vacuum transfer
separated from
to
a
the non-volatile
series of cryo-traps. It
yields
ground coffee.
compounds by means of an aroma isolate with an
odor that resembles very much the odor of the original sample. The main advantages of this method lie in the
isolation of polar and
relatively
low heat
impact
hydrophilic volatiles, since
to the
sample
no water
and
improved
is in contact with the
sample.
The
complex composition
all volatile
compounds
in
of coffee aroma one
usually makes it impossible
gas-chromatographic run.
generally required, which in most cases
A
procedures
Holscher et al.
can
be found in
is carried out by column
(1990), Vitzthum
chromatography (GC) requires reasons as outlined above
two different
Usually,
a
Bade-Wegner et
al.
separate
pre-fractionation
is
chromatography or
preparative high performance liquid chromatography (prep-HPLC). of these
to
A
description
ct al. (1993), Blank et al. (1992),
(1990). Subsequent separation by gas
high performance capillary column.
For the same
for isolation techniques, it is recommended to
use
at least
types of these stationary phases (Marsili, 1997).
the flame ionization detector (FID) is the preferred device for quantifi¬
cation of GC
separated compounds. Accurate quantification
certain flavor
compounds which occur frequently
levels (Grosch
technique
et
m
can
extremely
be difficult for
low concentration
al., 1990). Stable isotope dilution assay (SIDA) is
to overcome this
problem. Generally, the identification
performed by gas chromatography mass spectrometry (GC-MS).
of
a
suitable
compounds
is
22
Literature review
Gas is
chromatography olfactometry (GC-O), sometimes referred to
an
of
important analytical tool
"GC-sniffing",
in aroma research because it characterizes the odors
single compounds emerging from
1997). Here, the human nose acts the GC column. Extract dilution or
as
as
the
sniffing port
of the instrument (Marsili,
the detector used for evaluating the effluent of
techniques, such
as
CFIARM (Acree et al., 1984)
AEDA (Grosch, 1993), provide means to even evaluate the relevance and impact
of
the entire
single compound within
a
dilution of the extract and which the aroma
compound
can
are based on
be detected
the
principle, that the higher the dilution
by GC-O, the greater
of the food. However. GC-O also
implies
(1997) describes the "out of context effect", the and an
profile. They involve stepwise
aroma
a
its contribution to the
series of limitations. Marsili
"contrast effect", human limitations
systematic limitations imposed b y the test design. Therefore, imposed by analysis
at
should be checked and confirmed
the result of such
by sensory analysis
of models
(Marsili, 1997, Grosch. 1995).
2.4.2
Flavor of green coffee beans
Although Altho ugh green coffee beans as
having
no
pleasant aroma
gated, since they do 55
new
possess
a
generally regarded
as
such and
large number
generally regarded
of volatiles of volatiles (Clarke and Macrae, 1985).
to 52 volatiles already known
the authors identified even 13
as
arc
flavor, the volatiles of green beans were investi¬
or
compounds were added
(1975). Surprisingly,
not consumed
are
by Vitzthum
et al.
pyrazines, although these
are
products resulting from heat treatment. They found that the
odor of green coffee beans is
mainly caused by methoxypyrazines. According
recent literature
and Steinhart
by
Hol scher
volatiles have been identified
so
actually have
on
an aroma
impact
far.
the
Steinhart (1995) also added some 30 mental mental work. work. They found that
carbonyl function
and
are
a
a
than 200 green coffee
small number of these
compounds
green coffee. Holscher and
newly identified volatiles from their experi¬ of all identified
known breakdown
and ketones (e.g. ethanat.
more
typical flavor of
majority
dation of lipids. The list includes
aldehydes
Only
(1995)
to
compounds possess
a
products generated during autoxi-
hydrocarbons (e.g. ethane, i-pentanc, etc.), propanal. n-butanal, 2-butanone, 2,3-butane-
dione, 2,3-pentanedionc etc.). acids, esters, lactones, nitrogen compounds, sulfur
compounds (e.g. methional) ethers, halogens, phenols
and furans (e.g. 2-methyl-
23
Literature review
furan, furfural, etc.). Among the newly identified compounds were
for instance
hexanal, (E)2-nonenal, (E.Z)2.4-decadien-al. (E,E)2,4-decadicnal, linalool.
ß-damascenonc, 3-methyl-2-buten-l-ol Moreover, most of these compounds 2.4.3
are
and
2~
as
well
as
3-methylbutyrate.
afso found in roast coffee.
Flavor profiles of roasted coffee
The chemical reactions that
are
by roasting produce
induced
a
different volatiles. So far, more than 800 different volatiles from chemical classes have been identified in roast coffee
1989). Investigations
a
of
wide range of
et al., 1996, Flament,
the Maillard reaction and the volatile fraction of roast
on
been re\iewed among others coffee have been
(1990),
(Nijssen
vast amount
by Clarke and Macrae (1985), Clarke
Ho et al. (1993) and Reineccius ( 1995).
The reaction pathways of roast aroma formation have been reviewed by Holscher
and Steinhart (1994). As
they
are
of
a
very complex nature,
a
number of studies has
been devoted to aroma formation in model systems. Stahl and Parliment (1993)
reported found
the
on
generation of 2-furfurylthiol
in
cysteine-ribose model systems
and
increasing quantities with increasing temperatures and roast time. Also
Hofmann and Schicberlc (1998a) investigated the formation of 2-furfurylthiol in various precursor systems. They suggested that different formation pathways for
2-furfurylthiol
may
run
in
parallel during food processing. The authors also found
the formation of various pyrazines. 2-acetyl- and
cysteine
and
carbohydrates
to
be
dependent
and Schieberle, 1998b). Heat treatment in
favored pyrazine formation.
on
the system water content (Hofmann
dry systems
Bohnenstengel
known and newly identified volatiles
2-propionyl-2-thiazoline from
and
increasing temperatures
and Baltes (1992) reported
on
well
resulting from asparaginc/glucose and aspartic
acid/glucose mixtures under roasting conditions. So far. very little information is available
compounds
in coffee beans
on
during roasting
the formation
on this
with Robusta coffees. Coffee samples were roasted a
constant
of aroma
and the influence of different
conditions. Silwar and Liillmann (1993) reported
different temperatures for
development
subject in
on a
an
roasting
investigation
laboratory scale roaster
at
length of time of 5 min. resulting in products
of various degrees of roast. The authors stated from cup testing that aroma formation
24
Literature review
starts around
170
°C, when
a
peanut-like
roast note
can
be
perceived.
At
180 to 190 °C coffee-like flavor arose, whereas the "real" flavor of roasted coffee
only appeared at
220 to 230 °C. After
passing this point,
the the flavor
was
slightly over-roasted (240 °C) and typically over-roasted (250
study
did also demonstrate
a
...
judged to be
260 °C). This
continuous increase of the total amount of volatiles
with increasing temperature up to 250 °C, followed by decreasing quantities beyond
this temperature. Similar developments were described for furans Furans and caramel
compounds were found
2-furfurylthiol continued
generally reached to be
the
a
be fully developed at 230 to 240 °C.
to be formed up to 260 °C. The formation formation of pyrazines
maximum at 250 °C.
Beyond this
incorporated in melanoidins. Still, the group
respective compounds were found
Another recent
study by Mayer
and the degree of roast
arabica. For
a
series of
on
dark)
had the
on
temperature they
assumed
of pyrazines is heterogeneous and
to react individually.
concentrations of aroma
the
are
et al. (1999) dealt with the influence of coffee origin
compounds,
concentration depending and
to
pyrazines.
and
on
in blends of C.
the authors found considerable differences in
origin of blend.
greatest impact
compounds
The degree of roast (light, medium
propanal. 2(5)-ethyl-4-hydroxy-5(2)-methy 2(5)-ethyl-4-hydroxy-5(2)-methyll-
3(2H)-furanone. guaiacol, 4-cthyl-guaiacoL 2-furfurylthiol, 3-methyl-2-buten1-thiol and methanethiol. In blends of Colombia and Kenya coffees guaiacol and
2-furfurylthiol developed unhindered and were greatly increased with increasing
degree
of roast. Other
compounds such
f or developed to a maximum for
a medium
as
2,3-butanedionc and
degree
2,3-pentanedione
of roast and exhibited lower concen¬
trations in dark roasted coffees. In recent years,
more
research has been addressed to the sensory relevance of
volatile compounds and the identification of
investigations revealed
an
in coffee. Olfactometric
impressive variety of different aroma
However, they also showed that only
actually dominate
key odorants
the sensory
a
qualities
small number of potent aroma
perception (Holschcr
et al..
1990).
in coffee.
compounds
These most
important aroma contributors were termed aroma impact compounds, aroma key
compounds, character impact odorants selected
or
just potent odorants.
frequently cited aroma impact compounds
is
An overview
provided in Table
2.
on
Literature review
Tab, 2: Selection of
25
frequently cited aroma impact compounds in roasted
Arabica coffee.
Compound
References
(incomplete)
2,3-Butanedione (= Diacetyl)
Blank (1992), Grosch (95, 96), Semmelroch (1995a, 96)
ß-Damascenone (= 2,6,6-Trimethyl-
Holschcr (1990). Blank ( 1991, 1992), Grosch ( 1995,
1,3-cyclohexadienyl) 2,3-Diethyl-5-methyl pyrazine
1996). Semmelroch (1995a. 1995b, 1996) Blank (1901,
1992). Grosch (1995, 1996), Scmmclroch
(1995a, 1995b, 1996)
2-Ethyl-3,5-dimethyl pyrazine
Blank (1991, 1992). Grosch (1995, 1996), Scmmclroch
(1995a. 1995b. 1996)
4-Ethyl guaiacol
Blank (1991,
1992). Grosch (1995, 1996), Semmelroch
(1095a, 1905 b. 1996)
5-Ethyl-3-hydroxy-4-methvl2|5H]-furanone (= Abhexon)
Blank (1991, 1992), Grosch (1996), Semmelroch (1995b,
2-Furfurylthiol (= Furfuryl-mercaptan) (= 2-Furanmethanthiol)
Holscher (1990), Blank (1991, 1992), Grosch (1995,
Guaiacol
Holschcr (1990), Blank (1991, 1992), Grosch (1995,
1096)
1996), Scmmclroch (1995a, 1995b, 1996)
1996). Scmmclroch (1995a, 1995b, 1996)
3-Hydroxy-4,5-dimcthyl-2[5H]-
Blank (1991. 1992), Grosch
furanone (= Sotolon)
4-Hydroxy-2.5-dimcthyl-3[2H]furanone
(= Furaneol)
Holschcr (1990), Blank (1991. 1992), Grosch (1996), Semmelroch (1995b. 1996)
2-Tsobutyl-3-methoxy pyrazine
Holschcr (1990), Grosch (1996)
3-Isobutyl-2-methoxy pyrazine
Blank
Linalool
Blank (1991. 1992)
3 -Mercapto-3 -methylbuty I formiate
Holschcr (1990, 1991). Blank (1991, 1992), Grosch
(1991)
(1995. 1996), Scmmclroch (1095a. 1995b, 1996)
Methional (=
3-Methylthio-lpropanal) (= 3-Mcthylmercapto-propionalclehyde)
Holschcr (1990). Blank (1901. 1002), Grosch (1995,
1996), Semmelroch (1995a. 1995b. 1996)
2-/3-Methyl butanal
Grosch (1995,
3-Methyl-2-buten-1 -thiol
Holschcr (1990, 1991), Blank
1996), Semmelroch (1995a, 1996)
(1991, 1992), Grosch
(1995, 1996). Semmelroch (1995a)
2-/3-Methyl butyric acid
Holscher (1990), Blank (1992)
2-Methyl-3-furanthiol (= 3Mercapto-2-methylfuran
Holscher (1990), Blank (1991, 1992), Grosch (1995,
2,3-Pentanedionc
Blank (1991.
1996)
1992). Grosch (1995, 1996), Scmmclroch
(1995a. 1996)
2.3.5-Trimethyl pyrazine
Blank (1001. 1902), Grosch
Vanillin
Blank ( 1001, 1092), Semmelroch (1995a, 1995b. 1996)
4-Vinylguaiacol
Holschen 1990). Blank (1991. 1992). Grosch (1995.
1996). SemmchoclH 1995a. 1995b, 1996)
26
Literature review
Tressl and Silwar (1981) investigated sulfur-containing
determined the threshold of low
as
0.01 to 0.5
From 1 to 10
ppb
possessed
the authors stated that
compound or
and
on
sulfury note. Thus,
a
may be considered either
as
as
perceived like freshly roasted coffee.
compound, depending
al. (1990) regard 2-methyl isoborneol aroma character
was
the aroma of stale coffee with
2-furfurylthiol
an off-flavor
compounds
2-furfurylthiol. They found that in concentrations
ppb 2-furfurylthiol
it
aroma
impact
as an aroma
the concentration. Vitzthum et
responsible for the harsh, earthy
and
moldy
of Robusta coffees. Holscher et al. (1990) determined the aroma
impact compounds
of roasted Colombian coffee. As the most
important compounds,
they listed 2-methyl-3-furanthiol, 2-furfurylthiol, methional, 3-mcrcapto-3-methylbutylformatc, 2-isobutyl-3-mcthoxy pyrazine, 2-mctylbutyrate, ß-damascenonc and furaneol. Three of the
animal-like, catty smelling sulfur-containing
impact compounds were further described by Holscher Another extensive list of aroma
impact compounds
was
and Steinhart
aroma
(1991).
provided by Blank
et al.
(1991 and 1992). 3.5-dimethyl-2~ethyl pvrazine, ß-damascenone, 3-mercapto-
3-methylbutylformate most
and
2-ethyl-3.5-dimethyl pyrazine turned
be the three
powerful aroma contributors of ground coffee in this study. However, these
investigations also displayed that differ from the
one
in the
the situation in
ground coffee
beverage. Semmelroch et
comprehensive list of aroma impact compounds
(1995b)
determined from
headspace analysis
the
isotope
impact compounds differed
significantly between Arabica and Robusta coffees. Grosch another
may considerably
al. (1995b), using stable
dilution assays, showed that the quantities of 14 aroma
al.
out to
et al. (1995) provided
in coffee. Semmelroch
following key odorants
et
in
ground coffee powder; 2,3-butaneclione, 2,3-pentanedione, 3-methyl-2-butenthiol, methional, 2-furfurylthiol and 3-mercapto-3-mcthylbutyllbrmiate. A subsequent
investigation by Semmelroch
and Grosch (1996) with stable
and sensory experiments yielded yet another list of aroma coffee brews. A summery
given by Grosch various aroma of
et al.
on
studies
(1996). Finally,
impact compounds
2-furfurylthiol
and
concerning
on
a
recent
isotope dilution
impact compounds
the aroma of roasted coffee
investigation
on
sensory perception indicated
4-vinylguaiacol.
assays of
was
the influence of a
great influence
27
Literature review
Concerning
the
aroma
composition, some interesting parallels
found in other roasted foodstuffs. 1 -octen-3-one and aroma
on
of roasted
as
be
can
important contributors
to
the
and Cadwallader, 1998). Ziegleder (1991) reported
the aroma fraction of roasted cocoa. About 20 aroma
for the first time. The
coffee
2-cthyl-3,5-dimethyIpyrazine, 2,3-butanedione.
3-methylbutanal were identified
chicory (Baek
to
compounds were identified
listing also includes major contributors
to coffee aroma, such
as4-hydroxy-2.5-dimethyl-3[2H]-furanonc (furaneol), 2-/3-methyl butyrate, guaia¬ col, 2,3-butanedione and linalool. Another comprehensive source for comparison of
coffee, cocoa and
tea is
Sensory perception
of coffee
but also
compounds,
provided by Flament (1989). beverages
is not exclusively determined
by other important flavor compounds, such
like acetic acid and citric acid, and bitter
solids in the
beverage contributes
to the "body"
of the
beverage
"cup quality"
of
a
coffee
Staling
Since the
unprotected aroma of freshly roasted coffee
during storage, aroma freshness becomes
realized
a
and therefore affects detailed description
beverage.
of roast coffee
2.4.4
losses
organic acids
components. In addition, the content of
the sensory product properties. Illy and Viani ( 1995) provide of the factors that constitute the
as
by aroma
early that adequate packaging
a
is
crucial
subject
to
quality
severe
quality parameter.
It
was
significantly extend shelf-life of roast
can
coffee. On the other hand, it is not easy to measure the freshness of coffee. Vitzthum
and Werkhoff (1979)
suggested
indicator substances, such relation to
as
the
use
2-methylfuran in relation
methylfuran. They showed
storage temperatures,
as
well
of certain
as
quantity ratios to 2-butanone
the accelerated
staling process of ground coffee
(1979) confirmed
developed vent packaging materials
important role
of
(1989) used
combined
a
headspace analysis,
approach
as
compared
to
light
the beneficial effects of newly
on shelf-life. Tressl et
furfurylmercaptan
as
the same measure for
freshness and found greater staling rates in dark roasted coffees Arackal and Lehmann
methanol in
the promotion of staling due to elevated
compared to whole beans. Kwasny and Werkhoff (f 979) used
roasts.
or
of selected
al.
(1979) pointed
out the
in the staling process. Spadone and Liardon for the
multivariate statistics
reported significant qualitative changes
investigation and
sensory
of staling,
including
analysis. The authors
of roast coffee, even when stored under the
28
Literature review
best
and elevated temperatures were
possible conditions. High product humidity
found
to
be the most detrimental storage parameters. Oo had
an influence on coffee
samples stored in cans. Similar extent of aroma modification
sample series gassed
and 3 %
at levels of I %
was
detected for
02, whereas maximum staling
was
found in coffee samples packed in air. Rather unexpectedly, the author found both
02 dependent Kallio
et al.
and independent chemical reactions involved in the staling process.
(1990) investigated
development
the
of
headspace volatiles during
storage of ground coffee in air tight packages filled with C02 and air, respectively.
Surprisingly, they reported similar rates of alteration of most
analyzed
for both
of the volatiles
storage conditions, though conceded methodological limitations.
Steinhart and Holscher ( 1991 ) suggested that coffee freshness is constituted by low-
boiling components, such and
cc-dicarbonyls.
as
low-molecular sulfur compounds, Strecker-aldehydes
The authors
regarded methane thiol
as
the most
indicator of coffee freshness. Leino et al. (1991) characterized the
stored C. arabica and C. canephora coffees and the sensory
respective beverages.
The ratios of 2-methylfuran/2-butanone,
2-methylfuran/propanal were used
and
coffee for 18 months
at room
compounds profile, whereas
as indicators of
temperature led
to
In
ageing process, a
but
are
headspace
of
properties of the
acetone/propanal
coffee freshness. Storing the
several
changes
in the
aroma
the perceived odor intensities did not change during
storage. Hence, the concluded that certain compound ratios the
important
inadequate
suitable to monitor
to predict the sensory quality of the beverage.
further study these ratios were used to
commercial Finnish coffee blends
are
(Leino
investigate et al.,
the
staling process of
1992). Flolscher
two
and Steinhart
(1992a) used headspace cryo-focusing analysis, GC-olfactometry and statistical discriminant analysis. As
reported earlier, they
found
again great correlation
between the loss of methanethiol and the loss of coffee freshness during storage. In an
additional study Flolscher and Steinhart (1992b) formulated
concept of staling in roast coffee. They stated that
physico-chemical processes that lead characterized
products.
to
a
a
a two
step model
first step is determined
by
decrease of \olatiles. A second step is
by oxidative reactions, resulting in aroma-relevant oxidation
29
3
Experimental Raw material
3.1
Raw material
quality over were
selection
a
was
basically targeted
to maximum
continuity of coffee
long-term period. Green beans of defined varieties
used in order to minimize
from the same
supplier
and single origin
product inhomogenity. Still, using different lots
but from different crop years, the coffee
considerable, but acceptable range. Coffees were obtained from
quality varied two Swiss
in
a
import
companies. Main In
experiments
general,
if not
Colombia with
specified otherwise,
a water
of 10
content
ments were carried out with a
a
wet-processed
to
C arabica Linn, variety from
11 g / 100 g (wb)
wet-processed
was
used. Some
experi¬
C. arabica Linn, variety from Costa
Rica.
Comparison For trials with the intention of both
species
beans
a
comparison of different raw materials, coffees from
C. arabica and C. canephora were roasted.
Wet-processed
C. arabica
originated in Colombia, Costa Rica and Guatemala, dry-processed Santos was
imported from Brazil. Beans
of C.
canephora originated
in Uganda.
Blends hi trial series
involving industrial scale roasting
C. arabica beans
purchased
was
for color
used. Furthermore,
comparison.
a
a
commercial blend of 100 %
number of roast coffee brands
was
Experimental
30
3.2
Roasting
3.2.1
Laboratory roasting trials
Fluidized-bed hot air
laboratory roaster
Roasting experiments were carried roaster in batches
&
out
with
fluidized-bed hot air
a
of 100 g green beans. The roaster was built by G.W. Barth GmbFI
Co., D-Freiberg/Neckar. for a research project for coffee
adapted
laboratory
roasting.
on nut
It allowed for coffee
roasting (Perren, 1995)
and
roasting under well-defined
process conditions with accurate control of hot air temperature, air velocity and bean core
temperature. Fluidized-bed roasting and cooling were performed in separate
sections. Steam injection into the hot air inlet and water spray injection into the
cooling
air provided options for humid
atmosphere roasting
and water quench
cooling, respectively. A schematic drawing of the roaster is
provided
given
in Figure 2. and technical data
in Table 3.
Roasting section:
Air of ambient temperature
(Elektror, D-Esslingen/Neckar). Air velocity
knecht, CH-Gossau/ZH). The air parallel electrical heaters S10000
was
sucked in by
was
controlled by means of
heated
Hot air temperature
an airflow meter
a
flap
(Schilt¬
roasting temperatures by
to
laboratory roaster. 20
300 °C
...
temperature
(isothermal processes)
velocity
Hot air flow rate
Capacity Cooling
by
a
two
8D8 (Leister, CH-Kagiswil). Optionally, satura-
Tab, 3: Characteristic technical data of the
Max. deviation of hot air
radial fan RD2
was
valve in the inlet stream in front of the fan and measured
Hot air
are
±
1 °C
1.0
...
0.47
3.0 ms-1
.. .
1.41
nV mitr1
100 g green beans air flow rate
Cooling time for 100 g beans to achieve Tbean < 40 °C (without water quenching)
0
...
60
s
2.8 m^ mitr1
Experimental
31
A
exhaustst
exhaust
an
A
an
5>
qreen coffee
toast coffee
O
oof
X
l~i
compressed
I
o
air
*-P4~
*
HXE
'O
f
colci s*\ O
steam
conti ol
nn.i/oy
panel/
diyttal-
data
conveCe»
air
PDI
to
control ler
and
aquisition
amplifia
Fig. 2: Fluidized-bed hot velocity. 2: Inlet 5:
watet
air
laboratory roaster.
air flap valve. 3: Inlet air radial fan. 4: Electrical heaters.
Optional steam injection.
probe PT100
1: Airflow meter for inlet air
for
6: Static
air
mixing element.
7: Temperature
airin(COntrouei) temperature. 8: Thermocouple for airm
temperature. 9: Thermocouple for recording coffee pile temperature. 10:
Roasting chamber.
valve.
13:
Cooling
air
11; Cooling chamber.
radial fan.
12:
14: Water
Cooling
air inlet
flap
quench spray Injection.
15; Pressurized water container.
ted steam
brated
was
by
a
fed to the hot
static
CH-Winterthur). steel tube of 10 inlet and
mixing
air stream
element
(176 g nA air). The air stream
diameter and
a
height of 24
a removable wire mesh cover on
roasting were collected
at
the
air outlet
equili¬
ME SMV-X DN100 (Sulzer Chemtech,
The roasting chamber for batch roasting consisted of
cm
was
by
cm
with
a
stainless
a wire mesh bottom
the top. Silver skins
a \ aeuum suck
coming
off
for air
during
in system. Hot air tempe-
Experimental
rature
was
32
measured
by
a
PT100 temperature
probe right before
the
roasting
chamber and used to control the heater's power.
Cooling section:
The roasted beans were transferred
by removing
roasting chamber and pouring the beans into
the
An air stream of ambient
temperature ensured
a
a
to the
cooling section
the cooling chamber.
fast fluidized-bed cooling of the
beans in the cylmdric cooling chamber. For water
sprayed through
manually
quench cooling cold water
was
hollow cone nozzle 212.054.17.AC (Lechler, D-Metzingcn) into
the air stream before the chamber.
The control and data
acquisition system consisted of
(Philips, D-Kassel),
an
CH-Hegnau-Volketswil)
a
PDT-controllcr KS 4580
analog-to-digital converter/amplifier and
a
MIDAS
PC with the software FLOWCHART
(DMP,
(CoraTec,
D-Jülich).
Measurement of bean core
temperature
For determination of bean core
temperature 2
coffee were
prepared for placement
were drilled into
the holes in to
a
folding gap.
mm
the
point
a
diameter. A thermo¬
lay
and in
inserted into was
paid
in the bean core tissue and not in the
patent by Perren
a
was
Figure 3a. Special attention
could be lead into the cylindric
thermocouples were placed
mm
diameter (Thermocoax, F-Surèsncs)
thermocouples were installed
by Perren (1995)
thermocouples
in the bean core. Fine holes
hand drill of 0.3
of measurement
The mounted
3 beans per batch of 100 g green
thermocouples
barb arrangement as illustrated in
ensure that
described
the bean tissue using
K with 0.25
couple type
of
or
in
a
et al.
special fixation device (1994), by which the
roasting chamber. Additional
in the vicinity of the beans in order to measure pile
temperatures. The batch of green beans
was
added to the chamber before trans¬
ferring the entire setup into the pre-heated laboratory roaster. This arrangement allowed for
partially free motion of
fluidized batch without
data
the
loosing them.
acquisition system of
thermocouple-equipped beans within
All
the roaster
thermocouples were connected with and temperatures
recorded online. At least 10 temperature
averaged
curves
in order to overcome bean inhomogemties.
were
the
the
monitored and
from individual beans
were
Experimental
33
Single bean roasting A
and
roasting chamber with sightglass and
scope
was
developed
sistant sightglasses
for
cm.
to
One green coffee bean
was
chamber in order
nearly square
the
(Olympus. CH-Volketswil) glasses.
a
plane
cylindric stainless steel
in the
glass part and back
to circular
as
illustrated in
by Perren (1995)
was
Figure
thermocouples inside. A stereo microscope was
placed
in
a
and fixed
Again,
the
SZ 6045TR
horizontal position in front of the sight-
color 3CCD video camera KY-F55B (JVC, CFI-Oberwil) for
3b.
again.
integrated in the roasting
For bright illumination four cold light sources were focused
microscope
and thermoré¬
prepared for core temperature measurement
described
to lead
parallel in
Two
a stereo micro¬
height: 27 cm) changing the shape of cross section
tightened thermocouples inserted
special fixation device
optical setup including
optical online process recording.
gradually from circular
two
an
12 em) were installed
(7x
roasting chamber (0 10
by
optical online process recording
image acquisition. Pre-heating
on
was attached
of the roaster
was
the bean. A to
the stereo
only partially
possible.
Fig.
3: Fixation of ture.
thermocouples in
3a: Scheme for
one
the bean for measuring bean core tempera¬
thermocouple (t), allowing
for partially free
motion of the bean (b). 3b: Scheme with two thightened thermocouples and
t2)
to
keep
the bean
(b)
in a
fixed
position
for optical observation.
(t1
Experimental
Isothermal
34
roasting processes
Isothermal processes in the
roasting process. For
roasted in either or
in
are suitable to
a
a
the
majority of experiments green coffee beans were
high-temperature short-time roasting process (HTST)
low-temperature long-time process (LTLT)
process characteristics
given
as
rature medium-time process
roasting time
of 300
processes roasting
final
investigate the general influence of temperature
s was
was
in Table 4. In some
(MTMT) with
applied.
targeted
to
a
Roasting parameters
experiments,
the same
260 °C,
according
°C
the
to
medium-tempe¬
a
hot air temperature of 240 °C and
In order to be able to compare the
degree
product color. Typical product properties
Tab. 4:
at 220
at
of roast, based
are also
a
two main
on roast loss
presented in Table
and
4.
for the HTST and LTLT process and typical prop¬
erties of roasted products.
HTST
LTLT
roasting
roasting
Process parameters: Flot air temperature
260 °C
Hot air flow rate Hot air
1.08 m3 min
velocity
2.3
Roasting time Cooling
220 °C
m
air flow rate
1.08 m3 min"
s~l
155... 180
l
2.3
540
s
1.41 m3 miir
m
...
'
s"1
720
s
1.41 m3 min"1
'
Product properties (typical values): Color (L*/a*/b*)
Roast loss (RV) Water content
24.06/9.26/ 11.33
24.02/9.27/ 11.17
15.33%
15.81 %
2.68 g /100 g (wb)
2.15g/l00g(wb)
Roasting processes with temperature profile In industrial
practice coffee
the effects of
is not roasted under isothermal conditions. Therefore,
pre-heating, continuous temperature increase
in the final stage of roasting
on
or
reduced temperature
product properties were studied by developing four
temperature profile processes (Table 5). (a) high temperature with stage (HL), (b) continuous temperature increase from low
to
a
reduced final
high (LHC), (c) pre-
Experimental
35
heating temperature with subsequent LHC process (PLHC). (d) pre-heating temper¬ ature, high temperature
at medium
stage and reduced temperature
at
final stage
(PHL).
Tab. 5:
Temperature-time profiles for non-isothermal roasting processes (Set values).
Process Temperature
Time
Total
roasting time HL
240 °C
150
220 °C
LHC
210s 270
s
240 °C
55
s
150 °C
180
s
270
s
240 °C
50
s
150 °C
180
s
90
s
240 °C
140
s
220 °C
210
s
continuous increase 150
PLHC
continuous increase 150
PHL
no
Roasting
s
—>
240 °C
—> 240
°C
hot air flow (technical)
360
s
325
s
500
s
620
s
of beans with adjusted initial water content
Trial series dedicated to the influence of initial green bean water content
on
properties
of 11.1 g /
were
carried out by
100 g (wb) of a C. arabica
achieved
adjusting
the
original water content
variety from Costa Rica. Reduction
roasting
of water content
was
by vacuum freeze drying and resulted in products with water contents
of
7.3, 5.5, 5.0 and 3.2 g /100 g (wb) bean, respectively. An increase of water content was
accomplished by exposing
activity of aw
=
0.90
at
a
the beans to
a
humid
atmosphere with
a
water
temperature of 37 °C for variable periods of time. Green
beans with water contents of 14.4. 15.9 and 18.2 g /100 g (wb) bean were obtained.
Experimental
3.2.2
36
Industrial
Roasting trials different
roasting trials
and measurements
a
industrial scale were carried out in three
roasting systems. Commercial roasting conditions were recorded
Probat RZ 3500Y and the Gothot
system
on
series of
new
Rapido-Nova systems. With
the Barth CR-1250
roasting processes were tested.
of the roasting conditions in
Recording
in the
a Probat
RZ 3500Y roaster
Description of the system The RZ 3500Y (Probat, D-Emmerich) is
bowl type (Clarke and Macrae. 1987)
capacity
of
320
a
fixed
surrounded
periphery Water
by
shown in Figure 4 and
vertical shaft and
through centrifugal force assisted by a
batch roasting system of the rotating-
kg green beans. Coffee beans
horizontal bowl with
reaching
as
a
are
and fall clown into the
quenching
is
the beans
roasting,
the
roasting chamber,
chamber. Exhaust air is partially recirculated
a
at
a
rotating
the bowl
the bowl bottom. On
to the center in
spiral-shaped circuits
are
cooling bowl that works
applied first in
periphery of
entering from
multi-plate ring, they fall back
operated
fed into the center of
carried to the
hot air
hot air. At the end of the
are
was
discharged over
on
similar
the
principles.
and then in the
cooling
to the burner.
Measurements A shaft cover
on
top of the roaster provided access
this cover, 4 thin stainless steel tubes were
inserted into the
type K of 1.0
(MP2
in
mm
(0 2.5
mm) of various lengths (0.5
roasting chamber, each of them leading
diameter (Thermocoax. F-Suresnes) to the
Figure 4). Additional thermocouples
pipework into
the hot air
chamber
and MP^ in
(MPj
roasting chamber. Through
to the
supplies and
the air
were
the roasting
Figure 4). Thermocouples were connected to
a converter/
placed
humidity.
roasting time removed at
thermocouple
to
acquisition software. air
m)
through existing
discharge stream near
and
measure
1.2
points of measuring
inserted
amplifier MIDAS (DMP. CH-Hegnau-Volketswil) A
one
...
dew-point hygrometer
The coffee
of approx. 270
s.
was
was
roasted in
a
a
PC-notebook with data at
MP| (Figure 4)
3-stage process with
a
to
total
In order to record roasting dynamics, samples were
regular intervals during roasting, cooled immediately and analyzed.
37
Experimental
Fig.
4: Probat RZ 3500Y industrial
roasting system with
(operation manual, Probat, D-Emmerich). blower. 3: 6: Air
Rotating cooling bowl.
discharge cyclone.
MP; Points
7: Air
4;
1: Burner.
discharge.
2: Hot
Rotating roasting bowl.
discharge blower.
of temperature measurements,
chamber. 3: Air
capacity
a
8:
of 320 air
kg
supply
5: Bowl drive.
Cooling
1: Air inlet.
air
supply.
2: Roasting
Experimental
Recording The
38
of the
roasting conditions
in
a Gothot
Rapido Nova roaster
Rapido-Nova roaster (Gothot, D-Emmerich) consisted
coffee batch of 400 transfer. A
kg
and
set
a
couples were inserted into
a
a vertical
drum for
a
of rotating paddles to increase the rate of heat
of this kind has been described and illustrated
standard-type roaster
Clarke and Macrae (1987). In
of
similar way
by
for the Probat RZ 3500Y, 4 thermo¬
as
the roasting chamber and measurements were carried out
in the same manner. A two-stage process with
a
total
roasting time of approx.
354 s
was used.
Roasting trials with
a Barth CR-1250 roaster
The CR-1250 roaster (G.W. Barth is
a
50
kg batch roaster based
on
roaster NR (Barth), described
Ludwigsburg GmbH
the
& Co.,
D-Freiberg/Neckar)
design principles of the partially fluidizing nut
by Perren (1997). Different roasting processes with
various temperature-time profiles were carried out. targeted to different degrees of roast.
3.3
3.3.1
General analytical methods
Roast loss
The overall
weight difference between
immediately after roasting
RL
=
I00
•
and
cooling
-££££!!
the green coffee batch and the roasted batch was defined as roast loss
I°£l1
(%)
(RL):
(equation 1)
m
green
where:
ni
mtoast
we*»nt °f green coffee beans (g)
'
:
wei§nt
°f roasted coffee beans (g)
The loss of organic
dry matter
calculated
the water content of green and roasted beans and the roast loss
by taking
into account:
was
defined
as
organic roast loss (ORL) and
was
Experimental
39
dm,. ORL= 100-
(100-RL)
roast
(%)
dm,.
(equation 2)
i re en
where:
RL:
roast loss
dm,. srccn
dm
3.3.2 Color
:
(%)
dry matter of
green beans
wb)
wb)
g,
Color
a
ground finely
household two-disk coffee
in
CH-Rapperswil). to form
a
The
ground coffee
an even
was
Samples
transferred into
C ':=
was defined
Va
of coffee beans were
grinder Espresso
surface. Color values were
color space. Chromaticity
Chroma Meter CR-310 (Minolta,
reflection area of 19.6 cm2.
CH-Dietikon) with
3.3.3
/ 100 g,
dry matter of roasted beans (g /100
roast
was measured with a tristimulus colorimeter
pressed
(g
+
a
E20
petri dish
depicted in
(Turmix,
and gently
the CIEZA/*/?*
b".
Water content
Roasted coffee beans of roasted beans were
Samples Espresso
E20 (Turmix.
coffee water content
ground finely
CH-Rapperswil).
was carried
out
in
a
household two-disk coffee
grinder
Gravimetrical determination of roast
using either oven dehydration
or an
infrared
dehydration apparatus. Oven method: The determination
was
carried
out
according
to the
Swiss Food
Manual (1973). 5 g ground coffee were dried at 103 °C for 5 h.
Infrared
dehydration: Approximately
1 g
ground
roast
accurately into the loading tray of an infrared dryer LP The coffee
was dried at
coffee
was
weighed
16 (Mettler, CH-Greifensee).
120 °C for 10 min.
Green coffee beans Green coffee beans a
are
hard and
tough
and not suitable for grinding. However, after
first dehydration step, the beans become
more
brittle and
ready
for
grinding.
Experimental
Therefore,
a
40
two-step dehydration procedure
was used
Manual (1973). A pile of 100.0 g green beans the
weight loss recorded.
0.63
The beans
103 °C. The total water content
was oven
dried for 2 h
was oven
was calculated
dried
combining
was carried out
103 °C and
again during 5
h at
the two weight losses.
according
Manual (1973). A sieving fraction of ground coffee with
and 0.63
mm.
accounting
specified conditions.
a
particle size between 0.25
was
25 mL of the filtered extract were
on coffee
was used
extracted with 200 mL water under
carefully evaporated
water bath and oven dried for 3 h at 103 °C. The extraction
dry extract based
to the Swiss Food
for more than 30 Vc of the total coffee weight,
for the analysis. A quantity of 10.0 g coffee
at
Extraction yield
The determination of extraction yield
3.3.5
to the Swiss Food
then ground to particles smaller than
A quantity of 5 g ground coffee
mm.
3.3.4
were
according
yield
is given
as
on a
percent
dr\ matter.
Surface oil
Batches of roasted beans
were
stored in 500 mL septum flasks
desorption analysis. Oily beans were spread
on a soft
as
used for gas
plate and thoroughly blotted
off from the surface oil with absorbent Kleenex paper. Coffee oil residues flask were removed
by ethanol.
The
amount
of surface oil
was
on
the
then determined
gravimetrically. 3.3.6
Antioxidative potential
The antioxidative
potential of roast coffee due
compounds during roasting oxidation induction was
ground finely
period
was
as described
a
and water
by Hadorn a
and Zürcher (1974). Roast coffee
particle size
was
heated and kept
at
Headspace exhausts were lead into
conductivity
<
500 pm were put into
679 Rancimat (Metrohm. CH-Herisau) together with 5.0 g
soy oil. The reaction vessel the oil.
formation of antioxidant
by modifying a method for measuring the
and 200.0 mg thereof with
the reaction vessel of
through
estimated
to the
was
a
continuously measured.
tivity indicates the end of the induction period,
100 °C and
an
water-filled
measuring vessel
A
air stream led
sharp increase
in conduc¬
Experimental
3.4
41
Characterization of structural and
physical
properties of coffee beans 3.4.1
Volumetry
A displacement method based
on a
system described by Mohsenin (1986)
to determine bean volume and bean
peanut oil (density 910 kg nr3 roasted beans
was
weighed into
beside the balance. 1ne basket for 15
s
carried
at 25 a
density.
°C) and placed
on a balance.
wire basket which
was immersed into
in order to release air bubbles out
A small container
was
was
was
used
filled with
A lot of 30 g of
suspended
on a
support
the oil and moved up and down
trapped between
likewise with the empty basket. From the
the beans. Immersion
was
weight difference of
the
immersed basket with and without coffee beans and using oil density the bean volume
was
calculated. Bean
density
was
and bean volume. Relative bean volume
taking into account 3.4.2
the
computed
was
based
as ratio
on
of bean
weight in
air
the volume of green beans,
weight loss during roasting.
Mercury porosimetry
Mercury porosimetry makes use of the properties of mercury as and is based
on
a
non-wetting liquid
the measure of mercury intrusion into the pores of
a
sample
at
various pressures. Higher mercury pressures allow for the intrusion of increasingly smaller pores. The
inversely proportional relationship
intrudable circular pore and the Washburn
between the size of
applied mercury pressure
is described
by the
equation (Adamson, 1990):
YHct r
C0Sd|JJa
(equation 3)
P
where:
pore radius (m)
r:
surface tension of mercury (yHcr
Ypicr:
0pja jr.
:
contact
angle of mercury (0,,
pressure (Pa)
=
=
0.5Nm
130°)
)
an
Experimental
A
42
mercury-porosimeter Carlo
with
a macro-
and
a
Erb a 2030 (Carlo Erba Strumcntazione, I-Rodano),
micropore unit
halves) were placed in
a
was used.
About 0.4 g roast coffee (4
6 bean
-
dilatometer and evacuated in the macropore unit for some
15 min. The dilatometer
was
carefully filled with mercury and the sample
was
checked for macropores by increasing the pressure gradually to 100 kPa ( 1 bar). The
dilatometer
was
then transferred
gradually increased
to
micropore unit, where
to the
400 MPa
(4000 bar) during
the pressure
was
45 min and the volume of
intruded mercury recorded. Pressure values were converted into values of "equi¬ valent pore radius"
angle r
=
of 130° and
a
the Washburn
using
micropores from 6428
to the maximum
maximum dV/d
a
mercury contact
surface tension of 0.5 Nm-1, measured macropores ranged from
50 pm to 6.43 pm and
corresponding
equation. Assuming
log (r/io). where
nm
to 1.61 nm.
in the distribution function
was
The pore radius defined rmajn
V is the cumulated pore volume and iq
(normalization, dimensionless exponent).
The 1-50 value describes the
=
=
Ira
equivalent
pore radius at which 50 % of the total cumulated pore volume is filled with mercury.
Porosity
e
was
defined
as
the ratio of absolute volume per g bean and absolute
volume of intruded mercury per g bean.
In
applying high pressures
to sensitive structures
accomplished for
microscopy
and
to foods a
and careful
thorough evaluation of potential artefacts
interpretation of the results
roast coffee beans
by
means
of
is required. This
cryo-scanning
energy-dispersive X-ray microanalysis (Schenker
was
electron
al., 1998).
et
Further successful applications of mercury porosimetry to food have been
for roasted nuts (Perren, 1995). air dried
due
vegetables (Karathanos. 1996)
reported
and other
foods.
3.4.3
Dynamic mechanical thermal analysis (DMTA)
DMTA was used
to
investigate softening phenomena of coffee bean structure during
heating. Green coffee beans were manually ground with sandpaper
approximately was
3
clamped in
mm
a
thickness and with
two
sample while carrying
out
slices of
parallel plain sides. One slice
plate plate measuring geometry of
(Rheometrics, NJ-Piscataway, USA).
to
A constant
a
Solids
pre-load
dynamic testing (oscillation)
at
of 0.5 a
Analyzer
kg
was
frequency
per
run
RSA II
kept
on
the
of 1 Hz and
Experimental
a
strain
43
amplitude of
1 % of the slice thickness. The temperature
increased from ambient to 260 °C with
operated with Firmware version 5.0.0
continuously
°C min"1. The RS AII
heating rate of 5
a
was
and Rliios software version 4.2.2.
was
Storage
modulus G' and loss modulus G" were calculated online.
3.4.4
Electron microscopy
Cryoscanning electron microscopy (Cryo-SEM) The bean tissue structure
investigated by cryo-SEM, with
was
Netherlands) equipped with
scope (Philips. The
(Bal-Tec, FL-Balzers). Pieces a
scalpel and transferred
exposed
were
with 15
130 °C at
The
an
515 micro¬
Philips
SEM cryo unit SCLJ020
a
of beans were frozen in
liquid nitrogen, fractured by
the cold stage of the preparation chamber. The
°C for 10 min under p
to -80
platinum.
nm
to
a
<
2
•
10~4
Pa and
specimens were examined
at
a
samples
eryo-sputter-coated
temperature below
-
accelerating voltage of 12 kV.
Energy-dispersive X-ray microanalysis Elemental
Northern
analysis
0.1 60
The
microscope
working distance
window
in
Philips SEM. equipped with
a
mapping.
of 12
The
was
mm.
operated
at an
Elemental mapping
spatial resolution
was
(live time) and
a
dead time of 24
s
Small bean
buffer
pieces were fixed by
(pH 7.4)
tetroxide in
at
dehydrated
in
Gatan
4 7c
a
at 4 °C and
64 a
of
by Frey
of 25 kV and
carried out by energy
pixels with
a
dwell time of
magnification of
10 000 for
to 13 keV.
chemically fixed specimens
glutaraldchyde solution same
Tracor®
in 0.1
cacodylate
buffer, post-fixed by 1 % Osmium
again rinsed in the same buffer. Specimens
graded series of ethanol. transferred into water free acetone
point dried. They were sputter-coated with
analyzed with a field emission a
a
4 °C. rinsed in the
cacodylate buffer
and critical
with
x
was
a
in detail
(40 c/c) in the energy range of 0
Scanning electron microscopy (SEM)
were
64
as described
accelerating voltage
All spectra in the spot mode were acquired at
s. s
performed
energy-dispersive X-ray analysis system and
et al. (1996). a
was
50
nm
of Au/Pd
and
SEM Hitachi S-700. Images were recorded digitally
Digi-Scan interface
at an
accelerating voltage
of 10 kV.
44
Experimental
Analysis of the microfibril network in an in-lens field emission
bars were fixed at
by
a
the cell walls
was
carried out likewise using
scanning electron microscope Hitachi S-900. Small tissue
3 %
glutaraldehyde solution
4 °C, rinsed in the same buffer, dehydrated in
a
in 0.1 cacodylate buffer
(pH 7.4)
graded series of ethanol, defatted
in diethylether, transferred into acetone and subjected to critical point drying. The
specimen was glued with conducting carbon colloid in a holder, placed into fracture device BAF 300 and fractured at room ration
was
carried out with 200 Hz of
100 Hz of carbon
into the
a freeze-
temperature. Electron beam evapo¬
platinum carbon from
perpendicular with rotating sample.
The
an
angle
sample
of 45
°
and
was transferred
microscope. Micrographs were recorded digitally using the BSE-image with
a Gatan
Digi-Scan interface
at an
accelerating voltage of
10 kV.
Transmission electron microscopy (TEM)
Specimen preparation according
to
for TEM analysis
carried out with
a
modified
procedure
Angermüller and Fahimi (1982). Small bean pieces were fixed in
half-strength Karnovsky*s fixative 0.1 M
was
at room
phosphate buffer postfixation
was
temperature for 1 hour. After washing in
done for 1 h in 1 % osmium tetroxide in
0.1 M imidazole buffer. Samples were then rinsed in distilled water and dehydrated in
a
graded series of ethanol with stepwise embedding
Ultrathin sections were cut using
uranyl acetate electron scan
a
CCD camera.
100
kV.
Epon/Araldite resin.
Reichert-Jung ultramicrotome
and lead citrate, before
microscope at
in
examining
in
a
and stained with
Hitachi H-600 transmission
Images were recorded digitally using a Gatan slow-
45
Experimental
3.5
Gas
desorption measurement
and gas analysis
Sample preparation in 500 mL
Batches of 100 g green beans were
placed
after roasting. Each flask
tight with
was closed
thickness and then evacuated in the vacuum
septum flasks immediately
special rubber septum of
a
headspace analysis system with
pump (Leybold, D-Köln) during 2 min. The flasks
were
a
12
mm
Trivac D8B
stored at room
temperature in darkness while bean gas desorption took place.
Headspace analysis Sampling
of
out with the
was carried
headspace
equipment
and method described
by Bertoli (1989) and Margadant (1991). Total headspace pressure Two
gas chromatographs were
operated
in
parallel
to monitor
hand, and short-chain hydrocarbons and other gases
Details
on
Tab. 6:
are
given
recorded.
O2, N2, CO2 and
on one
analytical conditions
was
on
CO
the other hand.
in Tables 6 and 7.
Analytical conditions for determination
of O2,
N2, CO2,
CO and Ar in
coffee headspace. Gas
Chromatograph
Fisons GC 8340 (Brechbuehler, CH-Schlieren)
Porapak Q 80/100 mesh; 3 m x 2 mm glass Molecular sieve 5Â 60/80 mesh; 3 m x 2 mm glass
Packed column 1 (right)
Packed column 2 (left) Hot wire detector
Body temperature 150 °C; filament temperature 240 °C; attenuation 1; gain lOx
Oven temperature
60 °C. isothermal
Injector temperature
100 °C
Carrier
Helium 5.0
Carrier flow column 1
24.0 mL min"1 (DPFC flow mode)
Carrier flow column 2
75.0 mL min"1 (DPFC flow mode)
Polarity
Sample injection
Polarity change after elution of column
1
peaks
measuring valve with
2
sample
Electro-activated
loops Software
of 250
pL volume each
Chrom-Card, version 1.17
46
Experimental
Tab. 7:
Analytical conditions
for determination of short-chain hydrocarbons and
other gases in coffee headspace. Gas
Fi sons 8330
Chromatograph
(Brechbuehler. CH-Schlieren) Packed column
Alumina Fl, 60/80 mesh; 3
FID-detector
310 °C: Range 1; attenuation 0
Oven
Iso-stage 1: 90 °C, 1 Rate I: lS^nmr1
temperature programming
Iso-stage
m
x
2
mm
glass
min
2: 300 °C
Cooling Injector temperature
250 °C
Carrier
Helium 5.0
Carrier flow
DPFC flow mode, 20.0 mL raiir1
Detector gases
Air 120 kPa; Hydrogen 60 kPa; no make
up gas
2 mL sample loop
Sample injection
Measuring valve with
Software
Chrom-Card. version 1.17
The total
amount
of gases released from the beans
headspace pressure. An external standard
nCsH^
and nC^H^ (2 vpm each,
tative determinations of
m
was
calculated from the
gas mixture of CH4, C2FL5, C3H8, nC^io,
Ni: Garbagas, CPI-Zurich)
hydrocarbons. Identification of peaks
comparison of retention times from reference substances.
was
was
used for quanti¬
accomplished by
47
Experimental
3.6
of coffee aroma compounds and flavor
Analysis
3.6.1
General methodological considerations
Due to the
technological nature
analytical effort in
the
analysis of coffee volatiles
research would have been
present work intended
analysis and.
of the present project
a
full
project on
employ
to
applied in order
flavor analysis. A
a
Quantification
important elements
two
different
otherwise flavor
of
qualified aroma
for isolation of
the most critical step of
single high resolution capillary column catered
compound was
was
performed
no
for maximum
prc-fractionation of the isolates. Aroma
evaluated using gas chromatography olfactometry. in
a
relative
manner
by comparing between diffe¬
rently roasted products. Stable isotope dilution analysis 3.6.2
as
techniques
to avoid potential artefacts in
separation performance, however, with relevance of
restricted,
time-consumption by making minor metho¬
dological concessions. Nevertheless, volatiles were
coffee roasting, the
its own. The methodology chosen in the
the most
at the same time, to limit
was
on
was not used.
Isolation of the volatile fraction
Simultaneous distillation/extraction
(Likens-Nickerson)
Simultaneous distillation/extraction (SDE)
was carried
out with
a
Likens-Nickerson
apparatus (Likens and Nickcrson. 1964, reviewed by Marsili. 1997). The procedure
appropriate
for coffee has been described
30 g ground coffee
was
by Holscher
(pentane
/
(10
cm
was
an
of
internal
extracted with 50 mL
diethyl ether mixture 1:1) for 2 h. After drying with anhydrous
sodium sulfate the extract
column
(1990). A portion
combined with 500 mL distilled water and
standard of 2-Butanol (Fluka 19025, CFI-Buchs) and solvent
et al.
height,
was concentrated to
less than 1 mL
by means of a Vigreux
0 1 cm).
Vacuum distillation A modified
apparatus according
Holscher et al. (1990) connected with
(pentane
/
was used
a vacuum
to Schieberle
and Grosch (1983) and described by
for vacuum distillation, comprising 3
pump Trrwic 4/8B
(Leybold, D-Köln).
diethyl ether mixture 1:1) were added
to
cryo-traps
100 mL solvent
30 g ground coffee and to
internal standard of 2-Butanol (Fluka 19025, CH-Buchs). The mixture
was
an
frozen
48
Experimental
with liquid nitrogen and
(p
< 0.005
mbar) and
Dehydration
3.6.3
exposed in
a
to vacuum distillation at room
second
and concentration of the isolate
Gas
chromatography
FID
of
an
extensive series
separation
are
a
was
< 0.008
mbar).
used for separation and semi-quanti¬
of reference substances.
as well as
The
for characterization
analytical conditions for
8. Peaks in the chromatograms were characterized by
retention indices (RI) calculated
relative amount of
(p
was carried out as described above.
compounds from isolates
given in Table
for 2 h
(GC-FID)
A GC with flame ionization detector (FID) tative evaluation of aroma
70 °C
at
step
temperature for 3 h
compound
according
X
was
to Van den Dool and Kratz (1963). The
defined
as:
Ax
QFTD
=
x
where:
-r-t— AIStd
(-)
Qi-tdv relative amount to
Ax : AISul
Tab. 8: Gas
(equation 4)
:
of compound X the internal standard
peak area
of
peak area
of internal standard
compound
Analytical conditions for GC-FID analysis
Chromatograph
as
compared
X
of coffee volatiles.
Hewlett Packard GC 5890 series II
(Hewlett Packard. CH-Basel)
Capillary column
Supelcowax 10, 60 m, ID 320 pm, film thickness 0.25 pm (Supelco, CH-Buchs)
Detector
FID. 250 °C
Injector temperature
220 °C
Oven temperature programming
Iso-stage
1: 46 °C. 3 min
Rate 1: 4 Tmhr1
Iso-stage
2: 240 °C, 5 min
Carrier
Helium 5.0
Carrier flow
90 kPa column head pressure
Injection volume
1
Injection mode
Split 1:12
Software
Chemstation, \ersion A.03.34
pL
49
Experimental
Gas chromatography mass spectrometry (GC-MS)
3.6.4
for GC-MS measurements (Table 9) were
Analytical conditions applied
(RIC), corresponding
semi-quantitative characteristic identified
Tab. 9:
semi-quantitative evaluation
and Kratz (1963). In alternative to
GC-FID, relative amounts of
Gas
to those
in GC-FID analysis. Peak retention indices (RI) were calculated according
to Van den Dool
areas
kept close
to
evaluation
ion of the
a
few
via
compounds were calculated using GC-MS peak
equation 4.
according
compound
In some cases of co-eluted compounds
to
in
equation
5
was
applied, based
question. Compounds
were
on
Chromatograph
of coffee volatiles
Fisons 8065 (Brechbühler. CH-Schlieren)
Mass spectrometer
SSQ 710 (Finnigan MAT, CA-San Jose, USA)
Capillary column
Supelcowax ness 0.25
220 °C
Injector temperature Oven temperature program¬
ming
Iso-stage
I: 46
°C, 3 min
Rate 1: 4 °C miir1
Iso-stage 2: Carrier
10, 60 m, ID 320 pm, film thick¬
pm (Supelco, CH-Buchs)
240
°C, 5 min
Helium 5.0
Carrier flow
90 kPa column head pressure
Injection volume
0.5
Ionization potential
70 eV
Interface
240 °C
heating
pL
Mass range
40
Software
IC1S. version 7
...
300 amu
a
generally
by comparison of mass spectra and Rl with reference substances.
Analytical conditions for GC-MS analysis
a
Experimental
50
QMS(CI) where:
=
-j-^ -'100 A(MS)IStd
(equation 5)
Relative amount of
Qms
pared
to
peak area share MS
Aroma extract dilution
on
ion
peak area (RIC) of compound co-eluted compound
A.(MSUStd :
as com¬
MS
:
SCI :
3.6.5
X
the internal standard, based
characteristic
/V(MS)X
compound
of characteristic ion (%)
peak area (RIC)
analysis by
X and
of internal standard
gas chromatography
olfactometry (GC-O) The GC-FID system was
equipped with a column
end split, leading to
for olfactometry (Marsili, 1997). Aroma extract dilution analysis
a
sniffing port
was carried
out
with un-diluted isolates and dilutions 1:4. 1:16. 1:32, 1:64, 1:128, 1:256, 1:512
and
J : 1024. Each sequence of GC effluents
marked onset and end odor
quality
series were
point
to an assistant
of
a
was
sniffed
person. The online
processed into CHARM
the greatest dilution at which this a measure
a
FD-factor
with
a
FD-factor of 1024
(AIC) for
the
acquired data
a
of
a
compound was still perceivable in
more were
respective roast coffee.
and indicated
comptete difution
specific compound
higher than 256 were regarded or
persons. They
response chromatograms (Acree et al, 1984;
for the aroma relevance of
with
at least two
perceivable odor by pressing a button
reviewed in Marsili, 1997). The FD-factor for
It represents
by
a
the GC effluent.
compound. Aroma compounds
as
specified
was defined as
key compounds. Compounds as
"aroma impact compound"
51
Experimental
3.6.6
Sensory evaluation
Sensory evaluation by
an
expert panel
A quantity of 12 g ground roast coffee
0.3 L
boiling water
was
was
poured over
allowed to cool down to
Three expert coffee tasters
it.
approximately sipped
the
placed in
a
The coffee
porcelain drinking bowl,
suspension
was
and
stirred and
50 °C, while coffee particles deposited.
beverage using spoons.
Flavor profiling Flavor profiling
by
a
trained industrial panel of 10
standard sensory room and with
prepared immediately before
a
panelists
was
carried out in
professional support service. Filter coffee
a
was
sensory analysis, using 55 g ground coffee per liter
water. Profiling of samples according to selected sensory attributes was carried out
using
a
10
cm
provided with
line scale ranging from "attribute
a mark
not marked" to
"very marked",
f or t he reference. Data were evaluated statistically
by analysis
of variance (ANOVA), least significant difference test (LSD) and Student's t-test.
52
i
i
I
!
Serk?
i
„eo!
Blank leaf
/
!
!
Do
4
Results and discussion
Characterization of process dynamics
4.1
4.1.1
Heat transfer and development of bean temperature
The heat impact and the temperature required to roast coffee to other
roasted food
must exceed
products, such
190 °C for
a minimal
as nuts, malt or
length of time
chicory.
as
compared
a
sufficiently reactive
a relevant
process temperature
be measured to describe the overall thermal impact. Usually, the temperature
of the bean pile is recorded for practical reasons bean
high
In general, temperature
to provide
roast environment. Therefore, the residence time and must
arc
core
although
the measurement of the
temperature would be preferable for more precise description of the
roasting process. The
development
of pile and bean core temperature during isothermal roasting in
the laboratory roaster is shown in
rapid convective heat transfer. discontinuities such
Figure
5. The high air to bean ratio resulted in
The temperature increase
as described
was
steady without
for hazelnuts (Perren. 1995). After
a
a
any
similar, short
initial heating stage in all three processes, heating rates were found to be dependent on
the hot air inlet temperature. The bean core temperature
was
exceeded by the
batch pile temperature in each process. Neither of them ever reached hot air inlet temperature. Even constant
during excessive roasting beyond usual degrees
temperature difference
remained. This difference
between
point rather
to a
roast
the hot air inlet and the bean
a
core
disappeared completely when bean models made from
aluminium were heated. Therefore, it seems off through the very thin
of
thermocouple
unlikely that
an
undesirable heat flow
has affected the measurements. The results
particular situation regarding
the
proportions of heat conduction,
convection of heat and radiation due to the small batch size in the laboratory roaster.
54
Results and discussion
Figure
6 shows the bean temperatures
compared
to
during roasting in industrial roasters
the temperature development in the
industrial roasting systems
use
ratures that exceed
When roasting
laboratory roaster. Generally,
much lower air to bean ratios
roaster, resulting in lower heat transfer
to
as
as
the laboratory
the beans. Nevertheless, product tempe¬
225 °C were achieved in the final stages of industrial roasting.
was carried
out to the same
degree
of roast, both industrial roasting
times were found to be between those of the HTST and LTLT laboratory processes.
A series of industrial roasting trials with the Barth CR-1250 roaster and experiments with various temperature profiles demonstrated (data not shown here), that final
bean temperatures
arc
generally
not related directly to
the degree of roast. Coffee
batches of identical degree of roast may originate from roasting processes with
different end temperatures. Therefore, data many authors
arc
merely
on bean final
temperatures supplied by
of relative value because they only apply for
a
given
raw
matcriai and given process conditions. Internal heat generation due to exothermic chemical reactions in the beans has been
suggested by various authors (Baltes. 1977. Raemy Viani, 1995). However,
a
Uly and
substantial additional temperature increase in the final
roasting stages caused by such reactions the industrial
and Lambelet, 1982.
roasting processes.
was neither found
In the
in the laboratory
nor
laboratory roaster, the expression of
in an
exothermic stage in temperature curves
during excessive roasting might have been
suppressed by radiation phenomena
by superior inverse heat transfer from the
or
beans to the air. In the industrial trials the target degree of roast may have been too
low and the process terminated before proceeding into any exothermic final stage. In fact, only the observed unhindered temperature increase in
spite of reduced air
flow rate and heat transfer in the final stage may suggest the existence of
exothermic stage. The heat generation
might or
be too moderate to
as
measured
by differential thermal analysis
greatly influence bean temperatures in
a
roasting process,
substantial influence occurs only with high degrees of roast. The
arc
consistent with the few literature data
al.,
1991; Severini
et al.. 1991;
additional temperature increase
on
present results
temperature development (Da Porto
Illy and Viani, 1995), where also
in
an
the final roasting stage is shown.
no
et
significant
Results and discussion
A more detailed
analysis
Probat RZ system is
depending
on
the
was lower than
55
of the temperature development during roasting with the
provided in Figure
position
in the
7. Pile temperatures differed
rotating bowl.
The actual bean core temperature
the pile temperatures. Therefore, literature data
temperature development must be interpreted with due care, termed From
as bean
Figure
considerably
concerning product
they
as
are most
often
temperature, while in fact they generally represent pile temperatures.
7 it may also be noted that with the Probat RZ roaster
heat transfer in roasting stages 2 and 3 of the
3-stages process
a
reduction of
is achieved
by
reduction of air flow rate instead of hot air inlet temperature.
260240 : 220-
200 :
Ü 180-
t 16°: 3
140-
2?
120 :
80:
E-
60-
:
! > '
!
'
ii
i
;
1
'
1
* >
1
1
i
i
M l\ Vv s^_
40 :
20
; 1
1
1»
CD j r\n q. 100-
i
1
'
;
111
;
,,,; \\
toasting
o i
MTMT
LTLT
:
roasting \ roasting \
'."*
i
i
\ *
A,
-
n
u
~
1
0
-r^
100
i
l
i
i
i
i
200
300
400
500
600
Time
Fig.
5:
Temperature isothermal 220 °C
•
700 800
(LTLT),
and
at 260 CC
(HTST),
subsequent cooling. Curves
n=6, MTMT: n=10, LTLT: n=10).
l
'
l
900 1000
(s)
of bean pile (thick curves) and bean core
laboratory roasting
'
i
i
(thin curves) during
240 °C (MTMT) and are
averaged (HTST:
56
Results and discussion
ndustrial roasting: Probat RZ3500Y (320
kg)
Gothot Rap. Nova (400
kg)
Laboratory roasting (100 g): HTST: Bean HTST: Pile
LTLT: Bean - LTLT: Pile
—\
200
'
1
'
300
1
400
Time
Fig.
6: Temperature roasters
development
during roasting
in the
1—
1
500
—i
1
600
800
(s)
laboratory roaster
to the same
700
—
degree
of roast.
and in industrial
(Industrial roasting;
blend of 100 % C. arabica, laboratory roasting: C. arabica, Colombia).
Results and discussion
57
320 280-
MP, 240-
?"
i
•*""^s
C) 200-
o
CD i.
13
160-
MP2, pile (n=5):
crt
position
1.
CD Q.
120-
position B
£ CD
A
position C
80-
-
-
position
D
MP2, bean core (n=2)
40
—— position
B
0 50
150
100
Time
200
250
300
(s)
o
Fig.
7: Industrial
100 %C.
roasting with the roaster Proba t RZ3500 Y (Commercial blend of
arabica).
different positions section of the C and D.
7a: Hot in
air inlet
temperature (MP-j)
and temperature at
the roasting bowl (MP2) during roasting. 7b: Cross
roasting bowl with the locations of measuring position A, B,
58
Results and discussion
The
properties
temperature
of green coffee, in
particular the initial water content, affects the
Figure 8 presents
curves.
the
temperature
during
curves
HTST
laboratory roasting of coffee beans with different initial water content. For these measurements, coffee with humidified
as
described.
a
water content
of 11.1 g /100 g (wb)
was
dried
or
Higher initial water contents result in slower heat transfer.
in industrial practice, this fact makes far-reaching standardization o f the green bean water content mandatory. Small
green beans before
entering
and Horrell (1993) even
suggested pre-drying of
the roasting process in order to
improve the temperature
development for achieving high yield coffee.
250
200 Ü y ./
CD
150-
Z5
Initial water content
'I /'/
_t_j
ii'
cri i
X=5.5g/100g(wb)
;
i
j
CD Q.
ii
100-
/ i
£ O
X=7.3g/100g(wb)
j.'i
X=11.1 g/100g(wb)
i
P
h-
i
X=15.9g/100g(wb) X=18.2g/100g(wb)
50—-
0
~i——i—•—i—•—i—•—i—'—i—•—i—'—i—>—i——i—•—r
0
30
60
90
120
150
Time
Fig.
180
210
X
=
HTST
270
300
(s)
8: Influence of Initial water content of green beans
during isothermal
240
on
the pile temperature
laboratory roasting. Original
11.1 g/100 g wb (C. arabica, Costa Rica).
water content
59
Results and discussion
Dehydration
4.1.2
the
During
organic matter
the water in the green beans is
roasting process,
partially transformed into volatiles. Moreover,
matter is
water is
and loss of
generated as
a result
of chemical reactions and
a
vaporized,
substantial
and
dry
amount
of
again vaporized. Generally,
coffee beans loose between 14 and 20 % of their weight during roasting, depending on
green bean
The
quality, roasting conditions
degree
of roast.
relationship between roast loss (RL), organic roast loss (ORL) and water
content for HTST and LTLT
loss
and the
was found
in the
loss of organic
is shown in
Figure
was
initiated later during
found to have
dehydration. Dehydration, RL
and ORL
a
seems
to
achieving
Low
ORL \ allies
The results confirm the conclusions made
of dry matter is much more controlled
suggested by Puhlmann the result of their
on
bean
weight loss and
as
high
as
et al.
in HTST roasting.
(1980) that
may be lost
due
to
in
a
steady
and continuous
different dehydration mechanisms
and Meister (1989)
was
not observed
particular roasting conditions.
and may take on some water from
during grinding. Moreover, gases
water content, to
the loss
by temperature than by residence time.
as
and must have been
It should be noted that accurate
determination of water content in coffee beans is difficult. Roasted beans
hygroscopic
The
temperature processing even
by Dalla Rosa
Dehydration during isothermal roasting took place Stepwise dehydration
progressive roasting.
proceeded faster and more extensive during
roasting than during LTLT processing. be incapable of ever
more
major impact
HTST
manner.
9. The greatest rate of roast
early process stages, mainly caused by dehydration, whereas
matter was
roasting temperature
roasting
arc
very
surrounding air, while some water
are
also removed during analysis of
and thus, affecting the result. Therefore, data
on water content have
be interpreted with due care.
Figure
10 compares dehydration in the
The decrease in bean water content
compared
to
was
laboratory roasting, which
development. during
laboratory
In industrial
and industrial roasting processes.
delayed
in both industrial processes
is primarily due to
a
different temperature
roasting, the major part of water
the second half of the process. Both industrial final
identical colors, but
a
Rapido Nova process.
slightly lower water content
was
as
was
removed
only
products exhibited
achieved with the Gothot
60
Results and discussion
The effect of
roasting time
identical color
was
the final water content and RL of
on
confirmed with
Longer roasting times resulted greater RL with data
as
on
final water content
The
provided by Kazi
a diffusion-limited
findings
in moisture
nor
in ORL.
a certain
on
different initial water contents.
consumption required
to
The
humidity
roasting behavior. a
13 demonstrates
laboratory roasting in dry
lag phase.
not differ
The additional a
only
energy
delay in temperature
important parameter influencing
It
can
humidity. Even dehydration runs slightly
be assumed that the heat transfer in humid air
capacity. The effect of
vapor pressure gradient between the beans and the humid air may be
temperature increase and therefore cause
a faster
found in the hot air of industrial
roasting systems.
As
a
a
compensated
major part
of water
may accumulate and
Further
investigations
air
generate
humidity
a
humid
atmosphere in
and its effects
on
the
can
be
of the hot air is
recirculated for economical reasons, water from the beans and from water
cooling
smaller
progress of roasting.
humidity revealed, that a substantial amount
on
the
significant differences during isothermal
is more efficient due to its greater specific heat
Measurements of air
A high
and humid hot air at the same temperature. RL and ORL
progress slightly faster at elevated air
by a faster
by
faster increase of RL.
the additional water causes
of hot air presents another
an initial
and
higher initial water content (Figure 8).
roasting process. Figure
faster after
1 1.
agreement
RL. ORL seems to be
slightly affected by
increase in beans with
are in
time, and the final products did
They did, however, differ in
vaporize
Figure
and Clifford (1985) and
initial water content lead to greater dehydration rates and
converged after
shown in
dehydration process.
12 shows the influence of initial water content
Water contents
as
products of slightly lower water content
compared to short time roasted samples.
Hinman (1991), and indicate
Figure
in
laboratory roasting trials
products with
quench
roasting chamber.
product quality
are
required.
Results and discussion
61
200
400
600
800
600
800
Time (s)
400
Time
(s)
12
10
HTST
8 m
LTLT
\
6
roasting
roasting
4 2
'•••••
0 0
200
400
Time
Fig.
9:
600
800
(s)
Development of roast loss (RL), organic roast loss (ORL) content
(X) during isothermal
(C. arabica, Costa Rica).
HTST and LTLT
and bean water
laboratory roasting
Results and discussion
62
12 *w ---
Industrial roasting:
Probat RZ3500Y (320 kg)
10-
--•- Gothot R.N.
(400 kg)
o
Laboratory roasting: • HTST (100 g)
8-
LTLT (100
X c
after
6-
CD
*-\ \
c
cooling
after
>
\\
o o
g)
cooling \
2
J cooling
q
step 1
4-
I step
after
O 0
crj ^r
-|
1
0
1
1
100
1
200
1
1
1
300
400
Time
Fig.
10:
Dehydration
of coffee beans
roasting (Commercial blend of
1
1
1
1
500
A
1
1
600
1
700
1
800
(s)
during laboratory 100 % C.
arabica).
and industrial scale
63
Results and discussion
17.5 • HTST roasting
LTLT roasting
17.0
P' ----
_~j
16.5
Trend curves •
GC o
CO
O
16.0-1
CO °
15.5
15.0 -i
27
26
'
r
25
Lightness
L*
(-)
F/g. 11: Relation between roast loss and lightness of HTST coffees in
a
range of medium degree of roast.
and LTLT roasted
64
Results and discussion
0
200
400
600
800
0
200
400
600
800
Time
Initial water content
—•—X
_._^._X
Fig.
=
=
(s)
(g /100 g wb):
11.1
(original)
5.0
--*--X
X =
7.3
"--
12: Influence of initial water content of green beans roast loss
=
3.2
X=14.4
on roast loss
(RL), organic
(ORV) and water content (X) during isothermal HTST laboratory
roasting (C. arabica, Costa Rica).
65
Results and discussion
,-C*
16:
„-"
.
®
£12cr
8~
''.'
---»----dry
:8r-
4-
^*'r"
humid
--o--
0- •"""" —r
i
I
0
40
1
!
80
^
1
120
200
160
(s)
Time Of)
'
!
—,
-dry
16-
^ 12J
-- o-
-
humid " -
-
Lr
8-
o
4: 0-
„.o-"°:-:8---'
„O ®
r8-9 '
I
I
i
i
i
0
40
80
120
160
(s)
Time "o" -S—<
19
I £_
--»---dry
•-.--
CD o o t-
^
3 x
-
8-
*
,x
--o--
6 -
4
^;'cV-.
-
"--'ft-----
-
-»-—
o~ i
I
0
40
I
73; Influence of
roasting
'
air
organic roast loss (ORL)
laboratory roasting.
'
I
80
Time
HTST
humid
"#,-
~
vj
F/g.
200
— #
I
120
160
'
2()0
(s)
humidity (dry
vs.
humid)
and bean water content
on
roast loss
(RL),
(X) during isothermal
Results and discussion
4.1.3
66
Development
During roasting,
of bean color
the color of coffee beans change from pale green-grey to
orange, brown and finally dark brown and black. This color
in
Figure
14 for the HTST and the LTLT
color space
as
well
in the 7,*C*
as
yellow,
development is shown
laboratory processes in the CTE
L*a*b*
The color changed faster with
plane.
higher
temperatures, but followed the identical pathways regardless of the type of process.
Lightness decreased continuously, whereas chromaticity first increased in
an
early
stage of roasting (yellow phase), and then decreased again continuously. The decrease of lightness in isothermal processes seems to follow
of reaction. It is
a
was
found
to be
suitable indicator of the
roasting trials with beans between RL and color
Figure 15 illustrates
can
the
a first order
type
highly correlated with
RL and ORL. Therefore, color
degree of roast for
given
from
vary i n
different a
a
origin
raw
material. However,
revealed that the
wide range depending
relationship
green been
on the
development of bean color in laboratory
quality.
and industrial scale
roasting. Browning rates leveled off during laboratory roasting, although the highest bean temperatures
are
achieved in the final
roasting stages. Apart from temperature
and concentration of reactants the presence of water seems to be
non-enzymatic browning
in coffee. From
a certain
rate of
non-enzymatic browning gradually falls,
glassy
state
(cf.
chapter 4.2.2).
enzymatic browning
The effect of
key factor for
point of dehydration onwards
as
the bean enters
glass transition
a more on rates
the
and more of non-
is known for other food systems (Karmas et al., 1992). In turn,
the influence of temperature
on color
organic roast loss. Temperature seems that cannot be
a
overcome
development to set
a
is in
parallel
to
the influence
on
limit of maximum color development
by longer residence time. Higher temperatures led
to
a
greater browning potential and potentially lower L'k values. Roasting below 190 °C allows only for moderate color
development and incomplete roasting (Dalla Rosa et
al., 1980). As
from
laboratory
s
expected
the different temperature
and industrial roasting, the color
greatly delayed. 180
was
In the Gothot roaster, the color change
of roasting, but
half of roasting.
development
a
high rate
of lightness decrease
development
in industrial roasting was
in
was
not initiated before
was found
during
the second
67
Results and discussion
While using the
same
color descnbing system, color values foi
product measured by different authors
and devices
can
given coffee
a
vaiy considerably. The coloi
characteristics of 18 dilfcient roast coilee brands (coimriercially available, mainly
from the Swiss market) I,
'
are
given
values in the present thesis
products. Within
a narrow
lightness and chromaticity
in
Figure
to usual
16
The data
degrees
range of degrees ol
are
intended to relate
of roast found loast
the
commercial
m
t elation
ship between
is almost hneai.
SO-
co CO
CD
50
"
c +-•
40
•"
O)
_,
10
!
15
,
j
20
,
j
25
Chromaticity Fig.
14:
Development
of bean color
14a: Presentation
L* C*plane
in
during HTST
,
(
30
r
,
35
C* (-)
and LTLT laboratory roasting.
the CIE L*a*b* color space. 14b- Presentation
in
the
68
Results and discussion
75 70-
Industrial
%.;r'A
-
--
-
65-—
60
3
55
Il
so
| §
roasting:
Probat RZ3500Y (320
Gothot R.N. (400
kg)
kg)
Laboratory roasting: •-
HTST (100
g)
—a—LTLT (100 g)
a
45 40
target degree
U)
of roast
3530-
25 20
K I
'
0
100
'
1
1
200
300
1
1
'
1
400
500
Time
Fig.
15: Decrease of
lightness
(Commercial blend of
L*
'
1
600
'
1
700
r-
800
(s)
during laboratory
100 % C. arabica).
and industrial scale roasting
69
Results and discussion
24
Paulig Original 22igros
Exquisito
Coop. Jubilor classico
20-
x x
^
Eduscho Gala Nr 1 #
5
18-
1 16-1
Migros
Colombia
La
/•* Migros Festkaffee
-®-
Caffe Chicco d'oro Chicco d oro.
-'
»
Onko Gold
^
Linda-*
^
Medaille d'or
Coop Jubilor El Sol-•%—Jacobs Lavazza Qualita Migros Mocca»^4 ~~^
oro
Golden Bean No1
E
'
2 14
•
Q
'
Migros •^ \ Migros Gastronom
Migros Caruso Espresso
•
lly, Tnest Amici Caffe
12
'
progressing ^Migros Espresso classico « Starbucks House Blend
10 (
19
i
•
20
i
i
21
22
•
'
I
23
24
I
25
'
I
degree '
26
Lightness Fig.
I
27
L*
'
I
28
of roast
T"1
'
29
30
I
31
'
I
32
'
33
(-)
16: Relationship between lightness L* and chromaticity C* within the range of
commercial degrees of roast coffee brands
and color characteristics of 18 different roast
Results and discussion
70
Gas formation
4.1.4
During roasting,
a substantial amount
Maillard reaction.
Figure
of gases is formed
17 illustrates gas formation
isothermal laboratory roasting, measured
as
as a result
of pyrolysis and
during HTST
and LTLT
headspace pressure after storing bean
samples for 4 months. Any gas loss during roasting itself was not taken into account. As
expected from
formed
only
the
of dry matter loss, the
development
in the second half of the process. For
gas formation
was
highly dependent
roasting temperature
on
the
a
major part
given
of gases
raw material
roasting conditions,
as
was
the rate of the
higher
in HTST led to greater rates than the lower temperature in
LTLT.
Figure
18 presents the influence of LTLT, MTMT and HTST
the same
degree of roast (color)
on
laboratory roasting
the amount of different gases released
to
during
storage. The increase of total gas in final products with increasing degree of roast and
higher roasting temperatures observed here
was also
reported by Radtke ( 1975)
and Meister and Puhlmann ( 1989). CCb is the most dominant component in coffee gas. CO and
N2 present further major components.
meters on COo formation follows
the trend observed for total gas formation. In
contrast, CO and N2 quantities seem
to be
Clarke and Macrae (1987) have provided case,
the
The influence of process para¬
a
independent
of the roasting temperature.
percentage value for C02 of 87 %. In
the percentage of CO2 in the total coffee gas
was
shown to be
dependent
on
applied roasting temperature.
Figure 19 shows with the
the
development
development
of short-chain
LTLT
roasting were formed,
During HTST
and pentane
presence might have indicated
products
was
are
compounds
on
as
during
exception of pentane
secondary products of lipid
progressing
However, the influence of roasting temperature these oxidation
the same
but in greater amounts with the
(Figure 20). Since methane, ethane oxidation, their
hydrocarbons, that went in parallel
of the amount of total gas. High formation rates were found
in the second half of the process.
of oxidation reactions.
headspace concentrations
not consistent. Hence, most
probably they cannot
accounted for lipid oxidation. Munari et al. ( 1997) reported an
our
a
of be
coinciding picture of
overall increase of minor volatiles formation during roastinc.
Although their
71
Results and discussion
work shows that the development of
propionaldehyde, methylfurane or general trend.
subject of
The
single aroma compounds such
as
2 methyl
2.3 butandione may differ considerably from the
compounds formation
aroma
in
is covered
chapter 4.3.3. Tertiary butyl methyl ether (t-BME)
(Figure 20) and found
to
was
identified in the roast coffee headspace
be formed during roasting (Figure 19). A similar trend in
formation of this unexpected compound unexpected compound
was found
in each case with several coffee
bean varieties from different origin. Artefacts due to the
procedure were carefully ruled out. t-BME chemistry, as a roast
is volatile and has
is known
a characteristic
coffee component, except for
as a
headspace analysis
widely used solvent in
wet
odor. So far, it has not been described
one
study by Wang
et al.
t-BME has been mentioned in the context of contaminants. Ethers in
(1983) where
general
are
not
well known to contribute to the volatile fraction of foods. However, since all functional groups to
produce t-BME
roast coffee, chemical formation
Still,
the formation
pathway
Considering
be found in other
compounds compounds occurring
of t-BME in coffee does
way
as
the
on
the
majority of other minor coffee
gas
dependent
may contribute to the aroma of roast coffee.
the fact that the
major part
of gases formed during roasting remains
within the bean and is only released during storage, the great amount of gases must cause
headspace
an
gas pressures
at
are
related to the free volume within the beans and the
are taken into account, a model
developed as shown
The model The model goes in
entrapped
extensive pressure build-up inside inside the bean. If the measured
roasting temperatures be
in
not seem unreasonable.
of t-BME is unknown. t-BME is
roasting conditions in the same components. It
can
in
Figure
parallel with
21. Gases lost
of bean pressure build-up
can
during roasting during roasting were not considered.
the gas formation, except for
a
temporary stagnation
the stase of greatest volume increase. The model model suggests that the bean pressure
may easily exceed 10 are
bar (1000 (1000 kPa), and it confirms that the highest bean pressures
found in the final
roasting stages.
At the end of excessive
roasting, bean pressures of more than 20 bar
can
high temperature
be assumed. Radtke (1975) calcu¬
lated bean pressures of three different three different fully roasted coffees in the cold the cold state to be 8.0.
5.7 and 5.5 bar. Assuming at this
temperature
are
a final
process bean temperature of 230 °C, the pressures
13.5, 9.62 and 9.28 bar, respectively. Thus, they
are
exactly
72
Results and discussion
within the same pressure range
as
it is conceivable that the internal
in the model outlined above. On the other hand,
gas is
only partially pressure-effective, since
substantial part of the gas may be present in
together with water
vapor
are
the
an absorbed state.
driving force
for bean
a
At any rate, the gases
expansion during expansion during roasting.
i_
CO
E c
o
co
E
CD 13 Ü3 CO
en
v.„
i_
o
a.
1200 -•-— HTST
1000
--
800
-
roasting
LTLT roasting
f*
1'I"-"-
CD
co
CD
o
600
CO o (I)
o
CO CD SZ
400
.
i
medium degree
200 ..-*'-
0
*
¥
of roast _iii'ii^
2
4
8
10 12 12
Roast loss RL
Fig.
i
14
16
18
20
(%)
17: Gas formation during HTST and LTLT laboratory roasting expressed
headspace pressure after roasting
as
expressed by
immediately after roasting roasting
4 months
are unconsidered.
storage, and related
as
to progressive
RL. Data represent released gases from
to complete gasdesorption. Gas losses during
73
Results and discussion
^ 1200-
] Total
CO
]
£ 1000
3C02 |N2
CD =5 CO
800
gases
CO and others
D O,
m
and Ar
o
CD OD
CO CD v_
CL
600«3
CD Ü
CO
400-
Q. CO "O
CO CD
200-
X
MTMT
LTLT
HTST
Roasting process Fig. 18: Quantities of major gases (expressed
during LTLT, MTMT color. Percent
and HTST
distribution
as
headspace pressure) formed
laboratory roasting
(mean
and standard deviation
calculated from
headspace partial pressures.
released
4 months
during
unconsidered.
02
to the identical roast
storage.
Data
Gas losses
and Ar were not separated.
s,
n=4)
as
represent gases
during roasting
are
74
Results and discussion
400000
Ck
—D— Total gas formation 350000 -
2>
-
650000
UJ
250000
-
-
2
-•-
-
.
-
-
-a-
-
450000
---
--•-•• -•-+---
CO
!p CL
propane
50000
-800
3 CO CO CD O
-600
pentane
#
hexane
-
-400
--o-- t-BME
A--
•
150
Time
19: Development
of minor
-200 .
100
50
0
CD O CO Q. CO -a
-D-
Fig.
CD
butane
-250000
100000
o
-1000
ethane
\~
150000-
to
E
methane
—---
CO
200000 -
,o-- M200
Minor gas components:
O
CO CD
14UU L-.
(Headspace pressure)
300000
-
CO CD
I
200
(s)
components
gas formation
during
high
temperature laboratory roasting (HTST). Headspace concentrations expressed
as
GC
peak areas.
The development of total gas formation
(headspace pressure) is given forcompahson. are
not included.
Gas losses during roasting
Results and discussion
75
> 3000
LU -
HTST
CD c
CO
CO
CD
g> 2000
CD
CD
c
C
CO
CO
X
CD
c 0
g LL
1000
H T
0
>
3000
6
8
12
10
14
16
LTLT
-
75 c O)
2000
—
'co
Q 1000
-
II I
0 >
J I
'
'
I
4
6
I
1
8
1
12
1
14
i
16
Headspace (unmodified) Headspace with
3000
1 uL t-BME
2000
-
1000
-
added
9 LL
i
5000
CO
"en
i
10
4000 c
i
r\
.A-
0
4
6
8
10
12
X 14
16
Retention time (min) Fig. 20: FlD-chromatograms of headspace samples from HTST (top) and LTLT
(middle) roasted coffee beans major peaks
was identified as
of identical
degree
of roast. One of the
tertiary butyl methyl ether (t-BME) by GC-
MS, by comparison of retention time, and by in-situ adding of reference t-BME
(bottom)
Results and discussion
76
-1.8
CD
E J=5 -1.6
o > c
crj CD
-1.4
SZi
CD
.> -1.2
_cü CD
-1.0
150
100
Time
Fig. 21: Model of
gas pressure
build-up
250
(s) inside
temperature roasting (HTST), based
a
on
temperature, volume increase and porosity.
coffee bean
during high
measured gas formation,
77
Results and discussion
4.1.5
Figure
22 shows the
LTLT
Extraction
a
of extraction yield
development
as
influenced by HTST and
laboratory roasting. During LTLT roasting the relatively high extraction yield
for green beans of
yield
medium
continuously decreased of roast. The
degree
followed more
or less
and reached 23 % for LTLT roasted beans
development
of yield in HTST roasting initially
the same trend, but started to deviate after
a roast loss
of 6 %.
The extraction yield increased up to 30 %, before it started to fall again during the
final
roasting stages.
medium
degree
The extraction yield of
of roast
rable LTLT roasted
high temperature roasted coffee of
29 % and thus, much
was around
product. This
extraction yields achieved
a
a
greater than of the compa¬
result confirms earlier reports
of greater
by high temperature roasted low-density coffees (e.g.
Dalla Rosa et al., 1980, Kazi and Clifford, 1985, Maier, 1985, Small and Horrell,
1993). Thaler and Arneth (1968a) and Thaler (1975) reported that green bean polysaccharides
are water soluble.
a
substantial part of
Moreover, since the applied method
for the determination of yield includes all types of dry matter, other soluble consti¬ tuents
of the green bean
(oligosaccharides, sucrose, various
sugars, minerals, acids,
etc.) result in considerably high extraction yields of green and moderately roasted beans. Most of these chemical reactions
compounds. and
a
non-polysaccharide soluble
during roasting
In turn, roasting induces
and
are
green bean constituents enter
comcrtcd into volatile
major changes
on
the
or
insoluble
polysaccharide fraction
substantial part of the initially insoluble cell wall polysaccharides is trans¬
formed into soluble matter and contributes increasingly
Apparently,
the net result of these
two counter-current
trend to lower yield with the continuation of observed
during LTLT roasting. However,
extractable solids. The extraction
yield
to the extraction yield.
developments
roasting. This trend
it may only reflect the
is also affected
is
a
general
was
clearly
potential of
by structural properties of the
roasted coffee bean tissue. A more porous microstructure and greater surface area for mass transfer in HTST roasted of
decreasing extraction yield.
samples
may super-compensate the
general trend
Results and discussion
78
31 30 29
32
28
\
•-•-
s
CD
">>
27
c
g
26
o
CO -t—»
25
X
LU
24 23
HTST
LTLT
22
roasting roasting -]
T
4
8
1
1
10
1
12
Roast loss RL
Fig. 22: Development of
extraction
laboratory roasting
1
1
1
1
14
!
16
I
1
18
1
[-
20
(%)
yield during high
and
low
temperature
of C. arabica beans from Costa Rica. A medium
degree of roast is achieved at
a
roast loss of 15 %.
Results and discussion
4.2
Changes of bean structure
4.2.1
Tissue structure of the green coffee bean
Figures 23, present are
79
a
24 and 25 show the tissue structure of the green coffee bean. The cells
compact and dense structure with
of spherical shape
or
radially stretched
no
intercellular spaces. In general, they
to
ellipsoids depending
the location
on
within the bean, but also vary considerably. The cell walls of coffee beans
unusually thick
give them
the
as
compared
to tissues
typical nodular
Moreover, it complicates
of other plant seeds. Reinforcement rings
appearance in the cross sectional view
The cell wall material causes the
are
exceptional hardness
and limitâtes
and
(Figure 25).
toughness of
the seed.
microscopic specimen preparations
all
involving embedding techniques. A
single large gas-filled bubble
(Figures
23 and 24). Most
found in the
was
probably, they originate from
procedures during post-harvest processing of described similar structures visible in
stained
specimens
analysis they do
dry matter
a
be bound
dehydration
of
a
Coffee oil
superior technique
organization of
was found
occur in other oil
to
the
layer alongside
in all examined
(1985), Wilson
the coffee cherries. Dentan
light micrographs
by
a
biomembrane
to preserve not show
cytoplasm
containing seeds such
of the cell wall
samples
et al.
of
and
or
include any deposits of
and monitor native cell structures. any other details of the
was
cytoplasm.
only revealed by TEM-analysis.
was
to
as nuts
(Perren, 1995). In coffee beans these
be of about 0.5 pm diameter and to be located in
(Figure 26).
In general, the subcellular arrangement
very similar to the structures described
by Dentan
(1997) and other authors.
(Figure 27). Regardless
principle does compare
a
complex three-
of the different kinds of
rides involved in coffee beans, the structure and
network in
(J985)
chemically fixed
The cell wall polysaccharide microfibrils seem to be arranged in dimensional network
dehydration
be organized in numerous oleosomes (oil bodies), that also
spherical organelles were found a
the severe
solids-containing cell liquor. Cryo-SEM very
However, cryo-SEM micrographs did The structural
of numerous cells
vacuoles. However, in the present freeze-fracture SEM
not seem to
due to
often presents
as
cytoplasm
to the situation
polysaccha¬
complexity of this microfibril
in the
primary cell wall of onions
80
Results and discussion
as
shown
by McCann
et al. (1990).
As stated
by Dentan (1985),
the cell walls
are
crossed by plasmodesmatae channels in certain areas, providing cell-to-cell connec¬ tions between
protoplasts (Figure 26).
No
directly visual evidence was found for th e
existence of additional channels within the walls. The cell compartmentalization. the storage of
cell walls do not
only have
a
lipids within oleosomes and the thick
physiological function
in nature, but also
excellent stability properties during storage of the green coffee bean.
explain
the
81
Results and discussion
Fig. 23: Cryo-SEM micrograph of the green coffee bean tissue structure.
It
clearly
illustrates the dense and compact structure in the green bean. Numerous cells show
a
single large bubble (B) in
the cytoplasm. Most
probably they
stem from the dehydration procedures during the post-harvest processing
of coffee cherries.
(Image:
B.
Frey,
S. Handschin).
82
Results and discussion
Fig. 24: Cryo-SEM micrograph
of 3
or
4
adjacent cells
The cytoplasm of each cell (CP)
wall material (CW) Bubbles seem
to be gas-filled
(B)
in
is
in
surrounded
the green coffee bean
by strong frames
the cytoplasm have
no membrane
and
They display changes during severe dehydration
post-harvest processing of coffee cherries (ImageHandschin)
of cell
B
Frey,
in
S
83
Results and discussion
Fig.
25: SEM micrograph
of the green coffee bean tissue structure from
chemically fixed specimen.
The cytoplasm (CP) is visible in some cells,
whereas it is removed in others due to
preparation. Surrounding cell walls (CW) show
a
striking
and
a
fractioning during specimen
are
of remarkable thickness and
typical structuring with characteristic reinforcement
rings. The detachment of the cytoplasm from the cell walls displays artefact caused by the fixation procedures. (Image: S. Handschin).
an
84
Results and discussion
is*i§
tit lltl^fciiīfe*i3M*il
1 Fig.
26: TEM micrograph of a cell wall in
line is formed
a
green coffee bean. The continuous dark
by the middle lamella (ML), that lies between the thick cell
walls (CW) and the
perpendicular
to
cytoplasm (CP)
the middle lamella
of two are
channels (P) through the wall. Coffee oil is within
S.
the
cytoplasm
Handschin).
and located
adjacent cells.
Dark lines
parts of plasmodesmatae
organized
along
in oleosomes
the cell wall.
(O)
(Image:
85
Results and discussion
Fig.
27: SEM
micrograph
of a cell wall cross section of a
and fractured specimen from
a
green coffee bean. The structure of the
fracture surface indicates the presence of network of
chemically fixed, de-oiled
a
complex three-dimensional
polysaccharide microfibrils. (Image:
S. Handschin).
Results and discussion
4.2.2
86
Volume increase
Influence of time
during roasting
temperature profile
and initial water content
The decrease of bean density during high and low is
presented in Figure
the a
28. As
certain residence time onwards, on
time and
low-density coffee
was
was much
29 and 30 show the
Figures
the
temperature,
However, density decrease a
of the fast heat transfer in the
a result
highest rates of density decrease were found
Depending
temperature laboratory roasting
a
limited
in the first half of
curves
high
or
would
laboratory roaster processing. From
low density coffee
by temperature,
level off.
continuously
and the
was obtained.
potential
to achieve
higher in HTST processes than in LTLT processing.
development
of bean volume in laboratory scale
processes. Although volume increase not necessarily has to go inversely in parallel
with the
density decrease,
Density decrease processes,
as no
as
well
a
as
corresponding pattern volume increase
instantaneous expansion
was
found for coffee beans.
present
a
in all
steady change
was observed that would
lead to
a
discon¬
tinuity in the curves. This kind of expansion was confirmed by optical online obser¬ vation of coffee beans
during roasting
at various temperatures. High temperature
conditions resulted in much higher expansion rates conditions
(Figure 29). Figure
roasting
at
a
In industrial scale
delayed
trial
a
as a
function
low
tempe¬
medium degree of roast.
roasting, the volume increase
highest bean temperatures
roasting process
levels,
low temperature
to
large difference between high and
as
due to much slower heat transfer because of
31 ). Since the
compared
30 compares the relative bean volume
of roast loss and clearly shows the rature
as
are
well
as
the dehydration were
large batches
generally found
of beans
at the end of
and the water content of beans may still be at
(Figure
an indus¬
sufficiently high
major part of the overall bean expansion is produced only during
the second
half of processing. As different initial water
lopment
of the green beans result in
of bean temperature, water
expansion. shown in
contents
The influence of initial
Figure
content
also has
water content
a
a
different deve¬
major impact
on
bean
during laboratory roasting
32. Lower initial water contents result in
an accelerated
is
and greater
volume increase. Products with identical ORL exhibited different volumes. Small
87
Results and discussion
and Florrell (1993) reported similar findings and beans in order to
suggested pre-drying
of the green
produce low-density coffees.
A general relationship between the time temperature program and the density and volume
produced in
the beans
was
found to
apply
for both the laboratory
as well as
industrial scale roasting processes. Beans roasted at higher temperatures exhibited
greater bean volume with
and lower density than beans roasted at lower
longer roasting times. Therefore,
degree of roast with
a
a
the total roasting time to achieve
given raw material
and volume properties of
a
roasted
is
a
reliable indicator to
to
impose
a
for loss of
a
to
the statements
are
by various
or
browning,
predict the density
authors (Dalla Rosa et
the
by Dalla Rosa
a
and
roasting temperature
expansion that cannot
be
et al. (1980),
greater potential of bean expansion. However,
the final temperature achieved in
during
dry matter
limit of maximum
residence time. Similar ratures led to
given
et al. 1985, Kazi and Clifford, 1985. Scverini et al., 1991. Small
Horrell, 1993). Like seems
a
product. This relationship has been described in
series of investigations with different objectives
al., 1980, Guyot
temperatures
as
overcome
by
higher tempe¬
outlined above, not
process, but the total thermal energy transferred
the entire process presents the critical factor. Finally, different bean volumes
obviously related
to different average cell sizes (Kazi and Clifford, 1985).
88
Results and discussion
1300
-K LTLT roasted samples
\
1200
V\
1100
•
HTST roasted samples
-
1000 >^
\
CD
~~m
9
900
cö c
k
"n
\
800
T3 Hi
,
C
CO
•—
__
nog
'
700
CD
CÛ
600 500 400
excessive roasting I
0
'
i
200
'
i
'
400
i
I
600
800
—i—'
Time (s)
Fig.
28: Decrease of bean density during high- and low-temperature
roasting. Dotted lines represent trend medium degree of roast.
curves
laboratory
and arrows indicate
a
89
Results and discussion
1.8-
t—'—r—•—r
'
'
'
'•'
.
i—i—i—i—i—i—i—i—
1.8
LTLT roasted
samples
®
V
-1.6'
I
jg
ö
CD
Trend curve
'
CD >
1.6
•
#
1.4'
>
roasted
r
é
CO Si
HTST-
»'
Trend
.
_
__...,
0
,
î -i—i—>—i—1
r._T
1.2
1.04! 0
100 150 200 250
50
Time
Fig. 29: Development of
_JB--H-"
.
curve
1.0-
_
.
samples
1.2
1.4
(
200
400
(s)
,
r—t
1
600
1—
800
Time (s)
bean volume increase
during high (left)
and
low
temperature (right) laboratory roasting (C. arabica, Costa Rica).
1.9 1.8 CD
1.7
••--
HTST process
-»---
LTLT process
E 1.6 o > c
1.5
CO CD sz>
CD
.> as CD
.-
1.4
i
4 I
1.3
!
medium 1.2
degree
0C
1.1
of roast i i
1.0
t i
6
8
'
10
Roast loss
Fig.
i
i
i
i
12
14
16
18
of roast
20
(%)
30: Characteristic development of bean volume increase
degree
^
(roast loss) during high and
low
roasting (C arabica, Costa Rica, identical with
as a
function of the
temperature laboratory
raw material in
Figure 29).
Results and discussion
90
1800o
O) .*: m
E
1700-
f
1600:
step
£ o >
/
"o CD
(J)
/
«
1200-
/
'
'
y
y
V
i
•
i
100
_
•
1
i
200
'
I
300
31: Increase of
'
I
400
Time
Fig.
Probat RZ3500Y
(320 kg) N. (400 kg)
Laboratory roasting: •••- HTST (100 g) —a—LTLT (100 g)
800700-
_B_
_..+_... Gothot R.
>*
jg
_AA
Industrial roasting: _
1000-
900-
D
__A--~A
m
y
2
.A-r
JA"
1300-
1100-
step
^A"--A"'
_
Vf—
cooling
a
1400-
9
1
'
i
» •
CD
after
cooling
»
1500-
after
"
I
500
'
!
600
'
I
700
'
800
(s)
specific bean volume during laboratory and industrial scale
roasting. Open symbols indicate samples
compared to
of
a
higher degree
of roast
as
the industrial end products. An identical commercial blend of
100 % C. arabica beans
was used
for each trial.
91
Results and discussion
2.2
2.0-
-•-
1.8-
1.6-
,-""""' Initial water content of green beans:
1.4-
-X
=
-—X
=
-•—
X
=
--
X
=
•
-^
1.2-
1.0-
4
6
8
5.0g/100g(wb) 7.3g/100g(wb) 11.1 g/100g (wb) 14.4g/100g(wb) 12
10
14
Organic roast loss (%) Fig.
32: Influence
of initial water content
laboratory roasting. medium
An
on
bean
organic roast loss of
expansion during
ORL
degree of roast (C. arabica, Costa Rica).
=
HTST
7.0 corresponds to
a
Results and discussion
92
Model of coffee bean
Bearing
expansion
in mind that the gas pressure in the bean reaches its
final stage of roasting, it is in
an
not obvious
early stage of roasting in
why
highest level during
the highest expansion rates
the case of laboratory roasting processes
stage of dehydration in the case of industrial roasting. Likewise bean
expansion
is limited to low
beyond usual degrees
expansion rates
of roast. Finally the
higher expansion rates
are found
in the final
question needs
are
the
found
early
or an
it is not clear
why
roasting stages
and
to be answered why much
during high temperature roasting
as
compared
to
low temperature processes. The volume increase of coffee beans
of gas and water vapor
opposed
as
the
during roasting
driving force,
it due to the hard and
to
Furthermore, polysaccharides in
but limited
by structural resistance
tough cell wall material
an
amorphous
foodstuffs may undergo glass transitions, content, which in turn
promoted by development
is
semi-crystalline
or
depending
in coffee beans.
on
temperature and
change the physical properties completely (Slade
1991). Glass transition phenomena
may play
an
important role
in
state
water
and Levine.
in structural resis¬
tance of coffee tissue.
Figure
33 shows the assumed principle state diagram of coffee bean
linking T„ is
a
the glass transition temperature T„ to the water content of the beans. Since
material property attributed to
transition from as
one state into
a
particular polysaccharide, there
the other in foodstuffs with
a
composition
in coffee beans. Hence, several different glass transitions
tures
are
to
Roasting implies
dehydration of the
changing
bean. In
the bean from
finally back into
a
a more
large rise
Figure
33 the
in temperature
roasting curves
hard and glassy initial state into
glassy state.
exceed T„ in stage 2. the more
The
a
is as
at different
be expected and softening phenomena in foods may be of
characteristic.
bean
polysaccharides
as
well
may
a more
no
complex tempera¬
a more
as
cross
sharp
fuzzy
extensive
Ta twice,
rubbery state
and
The more the bean temperature T[-,ean will
pronounced the rubbery state will be, allowing
for
expansion. heating stage during laboratory roasting
rubbery state
of the bean with
is
passed quickly, which leads
high expansion rates
in a
early stage
to
a
of roasting. The
Results and discussion
rubbery state
93
may be more
pronounced with HTST roasting, resulting in greater
volume increase than in LTLT
final
roasting stage
increase. The
may cause
roasting.
The return to the
high structure resistance
heating stage before exceeding Ts
is
glassy state during
the
and hinder further volume
considerably prolonged
in indus¬
trial processes. The rubbery state of the bean will be reached only during the second
half of industrial roasting, The
hypothesis
of bean
as
the water content is still
expansion outlined above
on a
was
sufficiently high level.
supported by experimental
data obtained from dynamic mechanical thermal analysis (DMTA) measurements,
simulating
a
be related to 210 °C
slow
roasting process (Figure 34).
glass transitions in
(Tg2), respectively.
At least
two
softening events could
the temperature ranges of around 130 °C (Tgj) and
As the
heating rate
in these
experiments
5 °C min-1, dehydration and temperature development differed
real
roasting conditions. Hence. Tgi
Nevertheless,
130 °C
was
also
and Tgi must be
reported
to
be
a
was as
low
as
considerably from
interpreted with due care.
critical temperature in nut
roasting
by Perren (1995). Small and Horrell (1993) suggested acids with
subsequent
an instantaneous
COo formation in
sible for extensive bean
analytical data that show
decomposition
high temperature processes
to be respon¬
expansion. This theory cannot be upheld in view a
steady
and continuous decrease of these acids
connection between initial content of chlorogenic acids and bean In conclusion, gas formation,
dehydration, bean temperature
present the most important parameters affecting
during roasting.
of chlorogenic
and
of no
expansion.
and
roasting time
the volume increase of coffee beans
The shift in the balance between force and resistance due to these
parameters controls the steady and continuous increase of bean volume. It is influ¬ enced
by
particular.
the
roasting conditions
in
general, and by
the
roasting temperature in
Results and discussion
94
HTST roasting end
LTLT
0
\
\\
roasting
:
CO
range of glass transition
CD
temperature Tg
E CD
**^«*
start
~i—•—i—p—r
-i
r
1
Water content
Fig.
33: Hypothetical state
diagram of coffee bean cell wall polysaccharides with
Tg range (strictly qualitative assumption) development
for HTST and LTLT
glassy state of
the bean. 2: Tjjean
roasting. >
and
1:
temperature-moisture
Heating stage, T^ean
Tg, more rubbery state allowing
volume increase, stage of greatest expansion rates. 3; Tbean more
<
glassy state with high structure resistance.
<
Tg, for
Tg, again
95
Results and discussion
Temperature (°C) Fig.
34: Dynamic mechanical thermal analysis
clamped between (oscillation) with
a
a
of coffee bean slices
plate-plate measuring geometry. Dynamic testing
heating rate
of 5 °Cmim1.
G": Loss modulus. The ratio G'/G" is
phenomena during heating. 2: First glass transition
dehydration.
(DMTA)
a suitable mean
1: Moderate
(Tgp).
G': Storage modulus. to monitor softening
general softening due
3: Trend to moderate
4: Second glass transition
(Tg2).
to
heating.
hardening
5: Trend to
to progressive dehydration and subsequent increase of
due to
hardening
Tq.
due
96
Results and discussion
4.2.3
Structural
changes during roasting
Hot air roasting of coffee beans involves
microstructural changes. As off and
come
some
a
series of substantial macroscopic and
of bean
a result
cracks appear
expansion, remainings
of silver skins
the bean surface. In HTST and LTLT
on
laboratory roasting most silver skins came
off within the first
or
the first
two
minutes, respectively, without producing any sounds. For this reason, this event is not related
in any
way with popping sounds,
bean surface. Major surface cracks and near
the
to
the
generation
hair, probably accompanied by
ously enlarged
roasting
during
as
bean
preferably
on the flat side
a
major crack starts with
sharp popping sound. The
on
the
of the bean,
of beans during a
crack
as fine as
The crack is then continu¬
typical popping sounds in coffee
by escaping gas. The sounds become remarkably frequent
the final roasting stage.
dehydration
and chemical reactions
profound microstractural change
a
large void occupying
immediately after subjecting it is caused
by built
the bean
to
to
a
cytoplasm of the green
of structural change. The most
is the formation of excavated cells with the
towards the walls and
probably
during roasting lead
of both the cell wall and the
Figures 35-45 illustrate this dynamic process
striking appearance
The
a
of
expansion proceeds.
may be caused
Volume increase,
bean.
created
poles in particular. Optical online observations
roasting revealed, that a
are
but with increasing shear stress
cytoplasm pressed
the cell centre. This state is entered
high temperature (Figure 35).
Most
up pressure due to water vapor and gas formation.
layer of modified cytoplasm becomes thinner on continuation of roasting, since and more cell mass is converted into gases and water vapors and cell sizes
more
increased. It also seems to undergo
irregular surface (Figure 38)
more
one cell wall side to
structures in
found
to
most
a
and to filament-like structures stretching from
opposite (Figure 39). Occurrence of filament-like cytoplasm
frequent
and
typical with higher degrees
of roast. The numerous
shape of burst bubbles embedded in the cytoplasm layer (Figure 38)
likely connected with
zation of coffee oil. In
(Figure 41)
viscosity increase during roasting, leading to
large numbers (Figure 40) was observed in some tissue regions and was
be more
voids in the
the
a
are
was
the break up of oleosomes and the
are
subsequent mobili¬
general, the structure of high temperature roasted beans
comparable
to the
one
in low temperature roasted coffee. It may
97
Results and discussion
appear slightly more
disorganized. However, since inhomogenities from cell
within the same bean were far more
ently roasted beans,
a distinction on
pronounced than possible variations the basis of different
to cell
of differ¬
roasting conditions would
be unreasonable. The TEM
micrographs
in Figures 44 and 45 demonstrate that the
original cytoplasm structure in
green beans is subject to
roasting too. Although oleosomes are reported are
completely
oil to fuse and form as
new
well
coalesced oil
as
droplet diameter falling in
to be very stable (Huang, 1996), they
upon the
droplets. Numerous oil droplets
cytoplasm layer (Figure 45).
the range of 0.5 to 1.0 pm, much
than 6 pm diameter were found in some cases and also
(1997).
profound changes during
partially destroyed in the roasting process, allowing the mobilized
or
observed within
well-organized
In contrast,
numerous
smaller
droplets but
were
found in SEM analysis of chemically fixed
The
findings
indicate the presence of
mobilized lipid phase in
a matrix
a
no
The cell wall frame work appears
as
The average
larger droplets of more
reported by Wilson
et al.
droplets larger than 1.0
specimens (Figures
disorganized, rearranged
of other denatured
were
pm
42 and 43). and
highly
cytoplasmic constituents.
the most stable structure
part during roasting.
Nevertheless, closer investigations by SEM analysis suggest fundamental changes in the microfibril network of cell walls in roasted beans surface
points
to
a more
(Figure 46).
The fraction
muddled three-dimensional network made of denatured
nucrofibrils with shorter chain lengths
compared to
green bean cell walls. Similar
microscopic findings have been described by Wilson
et al. (1997). Moreover, this
visual
polysaccharides reported by
impression
is
supported by analytical data
Thaler and Arneth (1968a,
(1990), Navarini
as
on
1968b, 1969), Thaler (1975). Bradbury and Halliday
et al. (1999).
and
fundamental changes in chemical
Leloup
and Liardon (1993). These authors stated
composition of the polysaccharide fraction during
roasting. Leloup and Liardon found
that
roasting considerably
reduces
the
molecular weiaht ranee of arabinoaalactans and galactomannans in cell walls.
Considering
the fact of coffee oil and gas transport across the bean tissue during
storage, the existence of must
a cell
wall
micropore network allowing
for mass transfer
be assumed. Plasmodesmatae channels present in the green bean were also
found in the roasted state in
some
spots of the cell walls (Figures 44 and 45).
Results and discussion
However, these channels
98
are
cell-to-ccll connections
and
do
not
the bean surface. Moreover, it is unclear, whether these channels transfer
or
seem to
play
that the
roasting process alters
Illy
congested by denatured proteins. Thus, they a
key role in mass transfer. the
provide access are
to
free for mass
do contribute but do not
On the other hand, several authors assume
porosity
of the cell wall (Gutierrez et al., 1993;
and Viani, 1995; Massini et al., 1990; Puhlmann et al., 1986; Saleeb, 1975;
Wilson et al., 1997). The
microscopic investigations favor
the cell wall microchannels microfibril network.
are
embodied
by
a
model concept where
the individual meshes
of the
99
Results and discussion
âlïlÉl
m
llllîlll ï^^sSillllpfP
öftllitef
ISpl ^ïïïy&M&Mmmmyim
|^lil|l||^^^^|i^^^^;
;\jr:
^p^pflliS
glll
Fig.
35: Cryo-SEM micrograph of 60
s
a coffee
cell in
an
early stage of roasting after
of low temperature roasting (LTLT-process). The
already already rearranged rearranged
and forms
a
cytoplasm (CP)
thick layer along the cell walls (CW). A
large void occupies the cell centre. Smaller voids (V) of structure appear in the
is
cytoplasmic layer. (Image: B. Frey,
a
burst-bubble
S. Handschin).
100
Results and discussion
Üil^^^ft^>
KU ^^li||^||^^Ä|j
yÊÊSÊÈ
fr;4Éo§Mlillti
Mi
HiiBIIIBi^ÄPir:
Fig.
36:
Cryo-SEM micrograph Cryo-SEM micrograph
'.£: ""SIÜ^Äf^^lSII
of the tissue structure of
a coffee bean roasted
180 s at 220 °C (LTLT). Each cell exhibits changes of the the
for f or
cytoplasm.
Irregular layers of modified cytoplasm (CP) stretch along the cell walls
(CW).
A number of smaller voids (V),
within these layers. (Image: B.
but of various sizes,
Frey, S. Handschin).
are embedded
101
Results and discussion
Fig.
37: Cryo-SEM micrograph of the tissue structure of bean after 360
microstructural
s
of roasting. With
changes developed
cytoplasm appears somewhat thinner
preparation
the
B.
Frey,
LTLT roasted coffee of
roasting the
layer
of modified
proceeding
further.
The
and more irregular. Cracks
artefacts and may indicate
cytoplasmic layer. (Image:
a
a
high
(C)
content of oil in
S. Handschin).
are
the
102
Results and discussion
Fig.
38:
Cryo-SEM micrograph
of a cell
in a
fully roasted coffee bean
at
a
medium
degree of roast (600 s, LTLT processed) The remaining layer of modified
cytoplasm (CP) along
the cell wall (CW)
is
only thin It exhibits
burst-bubble structure with numerous embedded voids S
Handschin)
(Image
a marked
B
Frey,
103
Results and discussion
Fig.
39:
Cryo-SEM micrograph roasted
for
600
s.
It
of
a
was
cell in
a
fully roasted coffee bean,
obtained from
the
same
LTLT
specimen
as
micrograph 38 and illustrates the large discrepancies of appearance between different neighboring cells from the same
cytoplasmic structures (CP) stretching from opposite were found
sample. Filament-like
one cell wall
in numerous cells. (Image: B.
Frey,
(CW) side S.
to the
Handschin).
Results and discussion
Fig.
40:
104
Cryo-SEM micrograph for 780 s
(LTLT process)
occurrence
regions roast
of the cell structure
of filament-like
was found
(Image
B
It shows
a
in a bean
excessively roasted
tissue region with cumulated
cytoplasmic structures
The presence of such
to be more frequent and typical with
Frey,
S
Handschin)
higher degrees of
105
Results and discussion
Fig.
41:
Cryo-SEM micrograph
of the tissue structure
roasted coffee bean (120 s, HTST process)
comparable
to the
than B
possible variations
Frey,
S
Handschin)
were
a
high temperature
In general, the structure
one in LTLT roasted beans
to cell within the same bean
in
is
Inhomogentties from cell
found to be much more pronounced
due to different roasting conditions
(Image
106
Results and discussion
Fig.
42: SEM
micrograph
of
a
chemically fixed specimen from
an
excessively
roasted coffee bean (220 s, HTST process). Layers of modified cytoplasm
(CP) spread along
the cell wall
technique provides
a
numerous embedded
(CW) framework.
different
image of
the
The applied preparation
layer structure, showing
droplets. (Image: S. Handschin).
107
Results and discussion
Fig. 43: SEM micrograph of
a
specimen from
chemically fixed
roasted bean (identical specimen from 3 cells, separated
by
a
as
in
cell wall
micrograph 42)
(CW) junction
behavior of the middle lamella (ML) caused
the walls
or mobilized
The
droplets
It shows parts
a marked stair within
may be either more
and coalesced oil droplets
excessively
Different fraction
wall Modified cytoplasm (CP) with numerous embedded
along
an
(Image
droplets (O) lies
or less intact
S
the cell
Handschin)
oil bodies
108
Results and discussion
ML CW
Fig.
44: TEM micrograph of
a
cell wall
in a
partially roasted coffee bean (80
HTST process). The middle lamella (ML) forms
a
s,
continuous black line
and separates the cell walls (CW) and the layers of modified cytoplasm
(CP)
of the
channels (P)
S. Handschin)
two are
adjacent cells. Parts
of modified plasmodesmatae
visible perpendicular to the middle lamella.
(Image:
109
Results and discussion
CW
ML
s
Fig.
45: TEM
micrograph
specimen
as
in
of
a cell wall in a
micrograph 44).
partially roasted coffee bean (identical The middle lamella
presumably modified plasmodesmatae channels (P) Oil
droplets (O)
cytoplasm (CP)
(ML) are
of various sizes lie embedded in the or
alongside
to it.
(Image:
S.
and
clearly visible.
layer
Handschin).
parts of
of modified
Results and discussion
Fig.
46: SEM
micrograph
110
of a cell wall cross section of a
and fractured specimen from
a
fully roasted coffee bean.
the fraction surface suggests fundamental network of roasted cell walls
S.
Handschin).
chemically fixed, de-oiled
as
compared
changes
The structure of
in the microfibril
to the green bean.
(Image:
111
Results and discussion
in
Changes
4.2.4
Characteristics of
porosity
porosimetric curves
and model of pore structure
and mercury intrusion
The pore structure of green and roasted coffee beans
47 shows
porosimetry. Figure
typical porosimetric
beans. The intruded pore volume is related to 2
nm
teristic
investigated by mercury
curves
from roasted coffee
micropore sizes from
to the
10 pm down
radius. In general, curves obtained from coffee were of consistently charac¬
shape. They exhibited only minor mercury intrusion over
possible pore sizes
and
were
diameter range of 20 to 50 a
was
single peak
The model in
Access for
by
then dominated
nm. This
a
a
range of
wide
sharp increase
in
a narrow
pattern of narrow-ranged pore sizes resulted in
in the pore size distribution function.
Figure
48
explains
the origin and generation of this
mercury to the excavated cell lumina
the cell walls, forming
a
is
shape
of curve.
provided by small micropores
so-called "ink bottle" pore system. Therefore, only
pressure corresponding to the small size of the entrance pores allows for
penetration for the
filling
of the cell lumina.
micropores of
Consequently, high values
for
a
in
high
mercury
apparent pore volume
the cell wall were obtained, while this corresponded to the
of the cell lumina. Hence, the pore size at the maximum of the distribution
function (rmam) represents the size of the cell wall cumulated pore volume at the end of analysis (2
nm
micropores.
The value for
pore radius) corresponds to the
overall bean porosity. The model is
supported by
SEM analysis of mercury intruded coffee beans after
porosimetric analysis (Schenker et al, 1998). revealed
a
picture
No artefacts such
The micrographs in
a
as structure
collapse due
to
high pressure during porosimetry were
mapping
of mercury
freeze fracture across the cells clearly confirms, that mercury does enter
extent during porosimetry (Figure 50), A weak signal
walls, indicating that
mercury must have
intrude the cell lumina. Mercury rization
and 50
of still intact cell wall structure and mercury-filled cell lumina.
observed. Cell lumina were filled with spheres. The elemental in
Figures 49
was
passed
a
was even
cell wall
to full
detected in the cell
micropore network
then partially withdrawn during the
procedure after porosimetric analysis. This, together with
to
depressu-
the contamination
112
Results and discussion
of mercury with cell constituents, resulted in the formation of stabilized small
spheres. As has already been mentioned in the experimental part, there may be limitations in
applying mercury porosimetry nation for
Moreover, is based
due to pressure sensitivity of some foods. Exami¬
potential artefacts and careful interpretation of it must
on
a
always
be
kept
the results
arc
in mind that the concept of mercury porosimetry
series of idealizing assumptions, such
as
cylindrical shape
intruded pore. The rate of mercury intrusion in coffee beans is low and
low rate o f pressure increase
necessary.
during analysis. Nevertheless,
the
of the
requires
a
present results show
that the method is suitable for roasted coffee beans and successful in describing the bean pore
structure.
porosimetry
can
stability of coffee
be attributed
The model concept of
shapes and sizes, consistent with
The
an
mainly
a
unique type
exposed
to
mercury
unusually thick cell walls.
of pore opening of
findings made by Saleeb (1975). From a
the
a
very narrow size is
shape of
gas adsorption
very narrow pore size distribution in coffee beans. He
suggested multilayer adsorption nm
tissue
"ink-bottle" pore structure with large cavities of different
but with
isotherms he concluded
range of 2
to the
bean
and
capillary condensation
in
micropores
radius being responsible for the ability of massive CO2
in the
uptake in roast
coffee. For various types of wheat cells Chesson et al. (1997) reported cell wall
micropores
in the size of 3 to 6
adsorption measurements.
nm diameter. These data were also obtained
by
gas
113
Results and discussion
°)
900-
|
800
-35
Cumulated pore volume
c
cd 700-
B
--•-••
Pore radius
distribution function
600
-30
-25
—o
20
o>
o >
CD
500
[averaged curves, n=2]
O
O o-
400
-o •o
o
-15
r
P >
300
E
»10
-a
£ 200-I -5
100=3
Q
0
-i
ni
it]
1
1—i
r'i
ji'ff
'
***
•
i
T 'i*i'
~
100
10
i
i'n
0
'
[*
1000
10000
Equivalent pore radius (nm) Fig.
47: Typical porosimetric data obtained from roasted coffee beans
processed sample,
(HTST
160 s).
jt\yji4- p°uusu A AS/4 pressure cell lumina
eJnt 3~* mercury* ~
Fig. 48: Model coffee bean pore system surrounded by pressurized liquid mercury
during porosimetric analysis. Access small micropores in the cell walls.
to the cell lumina is provided
Only
a
high pressure corresponding
by to
the small size of these entrance pores will allow for mercury penetration
of the
large cell lumina.
Results and discussion
Fig.
49:
114
Cryo-SEM micrograph
of cells in
a
roasted bean intruded with mercury
during porosimetric analysis. The tissue shows structure with
penetrated It
was then
an
intact cell wall
mercury-filled cell lumina. Mercury (Hg)
a cell wall
micropore system before intruding
(CW)
must have
the ceil lumina.
partially withdrawn during the depressurization procedure after
the porosimetric analysis. (Image: B.
Frey S. Handschin).
115
Results and discussion
Fig. 50: Cells of a roasted coffee bean intruded with mercury during porosimetric
analysis
50a
walls 50b
Cryo-SEM micrograph
of 3 adjacent cells with integer cell
Mercury mapping obtained by X-ray microanalysis from
same location as in
50a
Bright spots
are
the
generated by great mercury net
counts and indicate the presence of mercury
(B Frey,
S
Handschin)
116
Results and discussion
Influence of roasting
Figure 51 shows
the
on
pore structure of cumulated intruded pore volume for green
development
coffee and beans of various degrees of roast. It documents the influence of HTST
roasting
on
porosimetric curves.
volume to
caused
a
A slight but continuous increase of cumulated pore
final value of 100 mirPg-1
was
observed for green beans. It may be
partially by micropores and partially represent
compression of coffee volume increase, bean
oil at
are
as well.
to
Greater values for final
observed with progressing roasting. Moreover,
shift to greater rmam and r^o values with increasing degree of roast
(Figure 52). These
due
high pressures. Since roasting involves substantial
porosity gradually increases
cumulated pore volume
artefact
an
data indicate that cell wall
micropores
are
a
slight
was observed
formed and/or
enlarged during roasting. At equal degree of roast, curves of cumulated pore volume were influenced
roasting conditions (Figure 53).
As
by
expected from greater volume increase, high
temperature roasted samples showed substantially greater overall porosity
compared
to low
the
as
temperature roasted beans. Further, they exhibited significantly
greater rmain values, meaning that HTST roasted samples developed wider cell wall
micropores than LTLT roasted beans. of HTST and LTLT roasted beans is
A survey of volumetric and porosimetric data
given
in Table 10.
Overall porosity values were in the same order
as found
by Radtke (1975). Values
for rraam fall between the two cell wall micropore sizes obtained from electron
microscopy by Wilson
et al.
(1997). They
are
porosimetric values proposed by Saleeb ( 1975), reported by Chesson
et
al.
(1997)
considerably higher than
but coincide with porosimetric data
for wheat cell walls. The
relationship between overall porosity and process temperature Ortolà et al. (1998). Kazi
and Clifford
are
(1985) and Puhlmann
findings
on bean
the size of cell wall
to
micropores
micropores
are assumed to
phenomena during storage.
be
porosity. So far,
dependent
cm
the
on
the
in agreement with et al.
contrast with conclusions of Gutierrez et al. (1993). who did not find
influence of roasting conditions
the gas
no other
(1986), but a
study
significant has shown
roasting conditions. These
be of great importance since they control mass transfer
Results and discussion
117
In conclusion, mercury porosimetry showed the existence of
network that is enlarged during roasting and
Origin
and structure of this system
are
not
dependent
on
a
cell wall
the process conditions.
yet elucidated satisfactorily. It is still
unclear, whether it consists of countable microchannels rather than of network. However,
dimensional case,
microscopic
and
micropore
a
complex
porosimetric results support a model of a three-
permeable wad-like network of polysaccharide microfibrils.
increased polysaccharide degradation at
larger cell wall micropores found
in
higher temperatures
In this
may cause the
high temperature roasted coffee beans.
118
Results and discussion
°>
900
co
£
800 700
cd
F
.5
600
o
averaged curves] 500
cd
green bean (unroasted) [n=4]
i~
400
Q.
CD
nnn
t>
300
=3
c
200-
=3
o
Ü
s
[n=4]
roasted for 80
s
[n=4]
roasted for 120
s
roasted for 160
s
[n=2] of roast) [n=2]
(medium degree
ioo H
£
roasted for 40
-i—i—ii 1111
10
1
-i—i—i—i 11111
1—i—i—i
100
i
ri i
-i—i
1111
10000
1000
Equivalent pore radius (nm) Fig.
51: Influence of HTST
roasting
increasing degrees
of roast.
on
porosimetric curves
of coffee beans with
I O.Ü
h
I 13-2~
+-
L--''"
linear fit
*
C
^'^ ^-
.3 12-8~
15 >
T3
"g.
2
LU
averaged r50 values
**~~
'
*~
**
12.4-
2
•"
O Q.
.-""
.-"
12.0'
m 11
I
K
I.On
i
i
0
20
•
i
l
40
60
-
•
i
i
80
100
Time
Fig.
*
i
l
120 140
i
|
160
""i
180
(s)
52: Relationship between degree of roast (roasting
time) and the averaged
pore size at which 50 % of total pore volume is mercury penetrated (r50).
119
Results and discussion
vO)
900-
CO
E 800o
700-
.5 600o
o
HTST roasting [n=3]
500-
r
*~
o a.
= .
main
13.4
nm
400-
S
LTLT roasting [n=2]
300 o
r
=
mam
200
11.2
nm
100
£ i
Ü
10
100
nil
10000
1000
Equivalent pore radius (nm) Fig.
53: Influence of HTST and LTLT
with identical
degree
roasting
of roast and
Tab. 10: Influence of roasting conditions
on
on
on
porosimetric curves of beans
rma;n.
volume and pore characteristics of
coffee beans at equal degree of roast.
Roast loss
(kg ur3)
Bean volume V^ (mm3 g~')
Hg-intruded volume Vjjs (mm3 g-1) Bean unam
porosity
(nm)
LTLT
roasting
roasting
14.95
(%)
Bean density
HTST
e
=
Vh/Vtj (-)
747
622
1609
15.01
1350
640
850 0.528
0.474
13.4
11.2
Results and discussion
4.3
120
Development
of aroma
compounds profile
and
flavor Aspects of methodology
4.3.1
The isolation
technique
in
aroma
investigation. Figure 54 shows isolates obtained
two
by
analysis
is critical for the result of
the differences in aroma
a
particular
compounds profiles from
different methods. Isolates obtained from simultaneous
distillation/extraction (SDE) were for example, considerably
generally of higher concentration
2.3-butanedione,
more
and
2,3-pentanedione
displayed,
and guaiacol
than isolates from vacuum distillation (VD). 4-\ inylguaiacol exhibited the largest
discrepancy.
It
was
not found
in VD-isolates of LTLT roasted
samples,
but present
in substantial amount in SDE-isolates. In general, greater artefacts due heat influence with the SDE
technique
In contrast, polar and hydrophil compounds
isolate. Acetic acid is to have
a
10-fold
and
are
an extreme
likely
are
therefore
to
technique.
be retained in the water
under-represented
representative
higher content in
greater
may be assumed. On the other hand, the
group of pyrazines seems to be widely unaffected by the type of isolation
phase during SDE-isolation
to
in the
respective
of this group because it
was found
VD than in SDE-isolates, relative to the internal
standard. Isolation
by
imposes
VD
lower heat influence
troublesome to handle than the SDE
technique.
suitable method for coffee, but may sensory relevance of bases of at least
tageous and aroma
an aroma
two isolation
on
SDE
cause some
compound within
proved
sample, to
be
a
but is
convenient and
the
profile must be assessed the SDE
on the
technique is advan¬
purely relative (semi-quantitative) evaluation
of
compounds.
Aroma isolates were
"on column" GC
exposed
to further heat strain
injection. They were subjected
during conventional instead
separation performance
of the
capillary column
superior separation performance compared
to
a
30
m
of the 60
m
of
to GC-FID analysis without prc-
fractionation, accepting incomplete separation of compounds. Therefore,
as
more
artefacts. In consequence, the
techniques. Nevertheless,
may be sufficient for
the
was essential.
Figure
the
high
55 shows the
polar column (Supelcowax 10)
unpolar column (DB 5). With
a
in
use
polar column peaks were
121
Results and discussion
evenly distributed over
the entire
peaks were overlaid within
a
analysis time, whereas with
identification and
to
be
unpolar column
compressed time/temperature window
of the analysis. However, regardless of
compounds had
the
in the first part
high separation performance, co-elution
of
accepted in some cases, with subsequent restrictions for
quantification.
122
Results and discussion
FID signal (mV) Fig.
54: GC-FID
chromatograms from
beans. Isolates
according
were
FID signal (mV) aroma
obtained
isolates of LTLT roasted coffee
by simultaneous distillation/extraction
to Likens and Nickerson, 1964 (left) and by vacuum distillation
(right). Open arrows point to examples
chromatograms.
of inconsistencies between the two
123
Results and discussion
i K
CD
-T^ FID signal (mV)
FID signal (mV)
Fig.
55: GC-FID chromatograms of beans
a
SDE
Chromatographic separation
a 60 m
aroma
of compounds
polar capillary column Supelcowax
capillary column
DB 5
(right)
isolate from roasted coffee
10
(left)
was performed using
and a 30
m
unpolar
Results and discussion
4.3.2
Table 11
124
Character impact compounds
gives
a
survey
on
selected identified
qualities and sensory relevance
as
analyzed by GC-olfactomctry. FD-factors show,
that the degree of contribution to the overall aroma
compound
to
compounds, their aroma
aroma
perception varies widely from
compound. So far, more than 800 compounds have been identified in
the volatile fraction of roast coffee, but the aroma may be dominated only number of so-called aroma
impact compounds (Holschcr
et al.,
by
a
small
1990). A widely
used synonym for the latter is character impact odorants. In the present investi¬
gation, listed compounds with
an
FD-factor 512
or
1024
are
considered
as
aroma
impact compounds (AIC) with high sensory relevance. A group of 11
AIC
was
identified for high temperature laboratory roasted
Colombian coffee, whereas 6 AIC out of it made up the
for low
respective group
temperature roasted coffee. The majority of these compounds is well-known in lite¬ rature to contribute to the
Czemy
et
group of AIC! (Blank
al., 1999, Grosch, 1995. Grosch
Semmelroch and Grosch.
et al..
et al.,
1991, Blank
1996, Holscher
et al.,
et
1992,
al., 1990,
1995a, Semmelroch and Grosch, 1996, and others).
2,3-butanedione. 2-furfurylthiol. methional. 2-ethyl-3,5-dimethyl pyrazine, methyl
butyrate. guaiacol and 4-hydroxy-2.5-dimethyl-3[2H]-furanone belong
to
this
category. However, 3 compounds ha\e
as
AIC
of coffee.
not
yet been described in literature
2-hydroxy-3-methyl-2-cyclopenten- 1-one, 3-methyl-mercapto-3-methyl
butylformiate
and
propyl pyrazine appeared
specific coffee provenience
in
to
In the
use.
be
exclusively characteristic
case
compound, namely dimethyl-propyl pyrazine,
of
propyl pyrazine
a
for the
similar
is described in literature. 2,3-butane¬
dione, propyl pyrazine. 2-hydroxy-3-methyl-2-cyclopcnten-l-one, 4-hydroxy-
2,5-dimethyl-3[2H]- furanone reach
an
FD-factor 512
or
and the unknown
compound with
RI
=
2329 did
not
greater in LTLT roasted products. Hence, they were
exclusive AIC of HTST roasted beans. In turn, guaiacol (AIC) and ß-damascenone were
important aroma contributors
2-furfurylthiol
is
characteristic for low temperature roasted beans.
generally regarded
as
coffee. However, in high concentrations than
as an
AIC, since it
is
reported
to
one it
of the most
may be more considered
change
concentration (Tressl and Silwar. 1981).
important
its aroma
AIC in roast
as an
off-flavor
quality depending
on
the
Results and discussion
Aroma
125
compounds with highest sensitivity to
the roasting conditions exhibited large
FD-factor deviations between HTST and LTLT
roasting. They
may
serve
"process indicator" aroma compounds. Most typical representatives of this are
2.3-butanedione.
arc
considered
group
2,3-pentanedionc. propyl pyrazine. linalool, 2-hydroxy-
3-methyl-2-cyclopenten-l-one AIC
as
as
the most
and again the unknown
compound with
RI
=
2329.
important aroma contributors. However, evaluation
of sensory relevance concluded from FD-factors imply methodological limitations. For reasons outlined above, aroma isolates
only partially representative for
are
roast coffee they were obtained from. In addition, odor may considerably differ from real conditions, as aroma aroma
qualities depending
perceived profile in
the final coffee
profiles of roasted beans, and the aroma
aroma
the
on their concentration
as
GC effluents
compounds can change their
(Tressl
et al.,
1981). Moreover, the
is different from analytical aroma
beverage
the extraction
perceptions in
the
procedure,
the different matrix (water)
complexity of human odor perception mechanisms have
a
major impact
on
compounds. Recent investigations have shown large discrepancies between impact compounds profile
beverage (Czerny
et al.,
in roast coffee and the sensory relevant profile in
1999).
In conclusion, the spectrum of aroma impact compounds is assumed to be deter¬ mined
by the
whole aroma the
raw
material, whereas the degree of expression of each
compounds profile
quality of
AIC within the
is subject to roasting conditions. In other words,
the green bean determines the
aroma
profile potential, whereas
roasting technology determines the specific part of this potential that realization. Some AIC
essential to
embody
produce
the different
the
are
found
ubiquitous
is
brought
to
in coffee and therefore seem to be
general odor perception "coffee", whereas others do more
potential due
to different
origin.
126
Results and discussion
Tab. 11:
Alphabetical listing of selected identified aroma compounds from high and low temperature laboratory roasted Colombian coffee beans and their sensory relevance.
Compound
No.
Acetic acid
1
2
p-Anis aldehyde (= 4-Mcthoxybenzaldehyde
3
2,3-Butanedione (= Diacetyl)
4
ß-Damascenone (= 2,6.6-Trimethyl-
RP Aroma qualityb
I46l 2070
FD factor0
present study /
HTST LTLT
(literature)
toasting roasting
(pungent) grass,
-
-
hay (sweet,
-
-
mint) 908 butter (butter)
1851 fruits, flowers,
1024
256
16
128
(honey, fruity, tea)
1,3-cyclohexadienyl) (= (E)-2-buten-1 -one) 5
2,3-Diethyl-5-methyl pyrazine
1505 (earthy, roasty)
6
2,3-Dimethyl pyrazine
1334
7
2,5-Dimethyl pyrazine
1304
8
0
10
n.a.d
-
roasty, nuts
2,6-Dimethyl pyrazine
13 11 sulfur-like, nuts
2-Ethcnyl-5-methyl pyrazine
1493
2-Ethyl-3.5-dimethyl pyrazine
1468 (earthy, roasty.
musty, burnt
n.a.
-
4
4
4
4
64
4
1024 1024
potatoes) 3-Etfiyl-2,5-dimethyl pyrazine
1443
11
12
4-Ethyl guaiacol
13
2-Ethyl-3-methyl pyrazine
1403
roasty, nuts
4
16
14
2-Ethyl-5-methyl pyrazine
1388
caraway
4
32
2-Ethyl-6-methyl pyrazine
1380 cheese, caraway
4
1
-
16
Ethyl pyrazine
1320
17
1440 bouillon, potatoes
2-Furfurylthiol (= Furfuryl-mercaptan)
Guaiacol
-
-
-
1024
1024
512
1024
(roasty. sulfur-like, coffee-like)
(= 2-Furanmethanthiol) 18
4
2025 flowers (spicy)
15
-
1889 medical, adhesive
(smoky, phenolic, aromatic, spicy) 19
Hexanal
20
4-Hydroxy-2.5-dimeth\l-3[2IIlfuranone (= Furaneol0)
1016
1
grass
2058 roasty. sweet,
-
1024
256
1024
32
(caramel)
(= 2.5-dimethyl-4-hydroxy-3[2H> furanone) 21
2-Hydroxy-3-methyl-2-c\ clopenten1 -one (= 3-Meth> 1-1,2-c\ do pentanedione
1851 (spices, celeriac)
22
2-Lsobutyl-3-methoxy pyrazine
1525 herbes, smoky
(earth), paprika)
64
127
Results and discussion
Tab. 11: Alphabetical listing of selected identified aroma
compounds from high
and low temperature laboratory roasted Colombian coffee beans and
their sensory relevance.
Compound
No.
RF Aroma qualityb piesent study
FD factor0
HTST
/
(literature) 23
Kahweofuran
24
Linalool
toasting loasling
1769 coffee-like, smoky
1555
grass, vegetables
4
1
256
4
1024
1024
128
16
256
64
(flowers)
25
Methional (= 3-Methylthio-l-
1462 cooked potatoes
propanal) (= 3-Methyl-mercaptopropionaldehydc) 26
2-Methyl butanal
(sweet) 857 caramel, nuts
(malt) 27
3-Methyl-2-buten-1 -thiol
1042
vegetables, green (sulfur-like, foxy, amin-ltke)
28
3-Methyl butyric acid
1680
1024
sweaty, pungent
1024
(fermented) 29
Methyl dihydro cyclopcnta pyrazine
30
2-Methyl-3-furanlliiol (= 3-Mcrcapto2-methylfuran
31
3-Methyl mercapto-3-methyl but\l
(roasty, sweet)
1304 sulfur-like (toasty,
n.a.
n.a.
32
32
1024
1024
16
51
128
4
1024
64
16
32
512
256
1024
4
_
-
256
256
meat-like) 1525 hcibes
formiate 32
l-Octen-3-one
33
2.3-Pentanedionc
34
1274
fungi, hay
989 butter (butter)
Propyl pyrazine
1418
potatoes,
vegetables 35
2,3,5-Trimethyl pyrazine
1402 herbes, bouillon
(roasty, earthy) 36
unknown
1625
roasty, nuts
37
unknown
1667
spicy, bouillon
38
unknown
2093 muck
39
unknown
2139 herbes, smoky
40
unknown
2^29 sweet, medicine
41
Vanillin
42
4-Vinyl guaiacol (= 4-Vinyl-2-metfioxy phenol) a
kl
b
Perception
c
d e
on
Supelcowax 10. at
bum
smiling pott
Flavoi dilution lac toi n a
Not
analyzed
Tuadename ol 1'umenich SA
(sweet, vanilla) 2245 sweet, flowers
(spicy-phenolic)
128
Results and discussion
4.3.3
Formation of aroma compounds during roasting
Figures
56 to 59 show the formation of selected
different roasting stages. Since chemical
characteristics of formation
However, the three selected
development isothermal
can
pathways
in the bean
vary considerably from
AIC in
Figures
of important compounds. It
laboratory processes
important aroma compounds during are
compound
56 and 57 may give was
very complex, the
a
to
compound.
typical picture
similar for the HTST and LTLT
and is characterized
by
low formation rates in the
first third of roasting time, followed by rapid formation in the second third. the final
roasting stage
pyrazine
and
formation
the
the concentrations of
was
already superimposed by
high temperatures.
2,3-pentancdione
as
A group of
shown in
of this kind. In contrast,
above described pattern of
example
the
spicy
AIC
the
spicy
It is clear from
nated
at
a
Figure
as
due to
well
as
remarkably consistent behavior did
not follow
the
in the final process stage. For AIC
guaiacol,
the
2-furfurylthiol (Figure 59)
at
buttery
AIC
as well as
the
high final temperatures.
57 that with both laboratory processes aroma
important compounds already decreased when
medium
order to achieve
a
degree
part of their study
(light, medium
the process
high aroma level. But other flavor compounds, such must
be considered
may greatly contribute to the aroma
results cannot be
quantities of
compared directly on
was
a
termi¬
of roast. This fact may require to stop roasting in time, in
acids and bitter components,
rylthiol
compounds
important compounds
spicy smelling
and roast) AIC
of
2-hydroxy-3-methyl-2-cyclopenten-l-one continuously increased even
during excessive roasting
number of
and
decay
shown in Figure 58
superimposed decay
smoky, aromatic
2,3-butanedione,
as
59 exhibited
group of other
a
again, indicating that aroma
an accelerated
pyrazines
Figure
During
2-ethyl-3,5-dimethylpyrazine, propyl-
to decrease
3-methylbutyrate were found
of
to those
the influence of the
as
of dark roasted coffees. The present
presented by Mayer
can
et al. (1999). The
is limited to
and dark), whereas the present results cover the
consistent formation trends
organic
well. Guaiacol and 2-furfu¬
degree of roast
the green to roasted beans be\ond usual
as
a narrow
development from
degrees of roast. Nevertheless,
be found. 2,3-butanedione,
guaiacol and 2-furfurylthiol in Colombia coffee
and also
pyrazine in Kenya coffee exhibit similar developments.
range
some
2.3-pentanedione,
2-ethyl-3.5-dimethyl
129
Results and discussion
0.35
2.8 -
3
0.30-
-
2-ethyl-3,5-dimethyl pyrazine
A
GO
— •—
propyl pyrazine
---A---
3-methyl butyric
< 0.25"*-
0.20
.
acid
2.4
X
2.0
J
-1.6
H
A.
0.15-
1.2
i »
A
0.10-
-0.8 D
0.05
4r
-0.4 O
o-
0.00
0.0 200
400
Time
F/g.
56; Quantitative
during
*
800
600
(s)
development of three selected aroma impact compounds
HTST
(solid symbols) and
roasting. Sampling took place
roasting time
to achieve
at
a medium
LTLT
(open symbols) laboratory
1/3, 2/3, 3/3 and 4/3 of the normal
degree
of roast.
0.25
2-ethyl-3,5-dimethyl pyrazine «
medium
-
2.4
degree
0.20-1 "'•"' Propyl pyrazine —a—
|
A
D',--A
of roast
3-methyl butyric acid -'
0.15-
i
v.
b2.0
h
1.6
I >
\
-1.2
0.10-
\
A
è
..Z\ .--•r
-0.8
0.05
0.00
-0.4 A"
-Of
JÊz
-a
70
i
50
Lightness Fig.
57: Quantitative
0.0 —T
60
40
L*
t
O
30
'
r-
20
(-
development of three selected aroma impact compounds in
relation to the degree of roast (color). Solid symbols: HTST roasting, open
symbols: LTLT roasting. Medium degree dicular line.
of roast is marked
by
a
perpen¬
130
Results and and discussion
X
8
6
Roast loss
Fig.
10
(%)
58: Quantitative development of pyrazines during high temperature laboratory
roasting related medium
A roast los loss s of 13 %
to roast loss.
corresponds
to
a
degree of roast.
r-0.12 0.7 0.6
-—
-•-
-
-A---
0.5
<
guaiacol 2,3-butanedione 2,3-pentanedione
•—2-furfurylthiol
0.4 A
0.3 0.2
0.1 0.0
i—'—i—•—r
6
10
8
Roast loss
Fig.
59: Development of relative
quantities
during HTST laboratory roasting. medium degree of roast.
12
14 14
16
(%)
of various important aroma compounds A roast loss loss of 13 % corresponds to
a
131
Results and discussion
In summary, the first roasting stage
result result in
not
large
aroma
precursors. The greatest
proceeds
to
at
a
still high water content of the beans does
quantities, but
aroma increase
a water content
the final
by high temperatures, whereas
Hence, there
are
rates
are
important
found
to
as soon
below 5 g /100 g (wb). Some
quantities decrease again during caused
may be
roasting stage
other
shifts in the aroma considerable shifts
develop aroma dehydration
as
compound
aroma
compounds decay
due to
compounds develop unhindered.
compounds profile during
the last
roasting stages. 4.3.4 At
a
Influence of
given coffee
raw
roasting parameters material
profiles
on aroma
chemical reactions in the coffee bean, which
may
be considered
revealed that the Roasting trials with ground raw coffee beans revealed that "bioreactor"
is essential
producing
compound profiles
compares aroma beans at identical formed
in
degree of roast.
process. Table 12
provides
roasting processes
rature conditions
of the few
not
on
and
high
low
"bioreactor".
presence of this intact
coffee
Figure 60
aroma.
temperature roasted coffee
As with AIC, the same aroma
compounds were
the
a
profile
semi-quantitatfve survey
on
are
specific for
compounds
was
a
certain
the influence of different
generation of aroma compounds relevant
exceptions,
nor
the
and the
found to be
to coffee flavor.
dependent on
the
tempe¬
during roasting. With 2-hydroxy-3-methyl-cyclopenten-l-one
necessarily
compounds
bond
the
the response of the of the formation of to
a
chemical classes. Neither
pyrazines
responded rather individually The
acceptable
compound within
The formation of most aroma
is
of
an
as a
during HTST and LTLT laboratory roasting, although the quantities
relative importance of each
one
the conditions of
control l roasting parameters contro
showed
a common
compound t o the sulfur
trend and
a
as
process
containing
compounds
aroma
to varying roasting processes.
majority of aroma compounds were formed
to
a
greater extent with greater
process heat impact. Roasting temperatures below 220 °C resulted in roast coffee of weak aroma strength. LTLT roasted coffees exhibited lowest values for most of the
compounds.
As
an
exception. ß-Damascenone
temperature conditions (Table 12). severe
is formed
On the other hand. HTST
temperature profile and the shortest roasting time did
o f aroma compounds. quantities of
The
preferentially
roasting with
not
develop
high final bean temperature in this
at
low
the most
the greatest
process may
132
Results and discussion
have induced
a more
extensive
decay
of aroma
with lower final bean temperatures. achieved with compounds were achieved
the other processes
compounds than
The greatest overall
quantities of aroma
the LHCI temperature profile in which the tempe¬
rature was continuously increased up to 240 °C and held there for the final roasting
stage. However, maximum quantities of aroma compounds must
positively related
to
superior sensory quality of
a
not
not
generally
effective
generation of higher
in
a
pre-drying stage
concentrations of
aroma
compounds. Most compounds were even slightly more pronounced without and
drying stage. 2,3-butanedione to be
compounds Therefore,
no
pre-drying
2,3-pentanedione were
the
significantly increased with the application of
general benefit from enhanced formation of be
can
be
coffee aroma.
A comparison between LHC and PLHC processes revealed that was
necessarily
a
pre-
only important
pre-drying stage.
a
aroma-precursors during
expected. Even during short time roasting processes, there
obviously sufficient time for
the reactants to form precursors and final
is
aroma
compounds. Including
temperature reduced final stage in the roasting process (PHL versus
a
PLHC) did
not affect the overall aroma concentrations,
Reduced final
but caused
a
shift in profile.
temperatures enhanced 2.3-diethyl-5-methyl pyrazine
and
lowered 3-methyl-mercapto-3-methyl butyl formiate, 2,3,5-trimethyl pyrazine and lowered
2,3-butanedione, 2,3-pentanedione and ß-damascenon.
It may be concluded that
reduced final
for
compounds, A
but
comparison
during
process temperature
disadvantageous disadvantageous to
of HL and LHC
a
temperature sensitive
the formation of more stable
compounds.
roasting shows that high temperature exclusively
the initial roasting stage
Therefore,
is beneficial
a
was
not
efficient in
producing aroma strength.
sufficiently high temperature during the medium
stages is required. HL and LTLT processes did
not follow this
or
final
roasting
requirement
and
consequently yielded only weak aroma strength. Distinct temperature profiles resulted in coffee
compounds profile Figure
61 shows
an
products of individual
and may therefore influence the sensory
attempt
to \ isualize sensory qualities on
comparison of aroma compounds quantities
aroma
aroma
perception.
the bases of relative
and their related sensory aroma quality.
HTST roasted coffee appears superior to LTLT roasted beans in all sensory groups.
133
Results and discussion
Processes with
or
without
a
final stage of reduced temperature caused marked
differences in the group of earthy, roasty. smoky
buttery notes.
A pre-drying step resulted in weaker
for the buttery note, which is due
2,3-pentanedione. Figure
61 must not be overinterpreted,
without statistical treatment. Moreover,
quality does
In conclusion, development of aroma
profile and
a wide
from the same
raw
conditions such
as
affect the aroma
be
subjected
(Vitzthum
to
range of distinct
of 2,3-butanedione and
as
it
comprises
relative and normalized
on
the
and in the group
development of all notes, except
to higher concentrations
of systematic limitations. It is based only
classes of sensory aroma
compounds
can
simple grouping of compounds
not air
of aroma compounds
can
staling
Steinhart, 1992a and 1992b. Leino
to
humidity
mechanisms
Spadone et al.,
be obtained
only temperature, but also other process and contact with oxygen may
quality. Moreover, the aroma fraction of roast coffee
and Werkhoff, 1979,
quantities
be controlled mainly by the temperature
profiles
the air to bean ratio,
and
number
not take different FD-factors into account.
material. Of course,
changes
a
and
1992)
immediately
is known to
after
roasting
Liardon. 1989, Holscher
and
134
Results and discussion
0)
_r?
FID signal
FID -f""iT—"I
Fig.
"1
!
signal
r~~T—T—T
Î"
T""
60: GC-FID chromatograms of SDE aroma isolates from HTST roasted (left) and LTLT roasted (right) coffee samples of identical
degree
of roast.
135
Results and discussion
Tab. 12: Influence of different temperature profiles3 on
as defined
in Tables 4 and 5
quantities of important aroma compounds
the relative
in
laboratory
roasted coffees of identical degree of roast.
Compound
Ax/Apstd. (-)
LTLT
HTST
HL
LHC
PLHC
PHL
0.196 0.130
2,3-Butanedione
0.130 0.204 0.110 0.171
ß-Damascenone
0.019 0.015 0.016 0.016 0.016
0.019
2,3~Diethyl-5-methyl-pyrazine
2.695
2-Furfurylthiol Guaiacol
2-Hydroxy-3 -methyl 2~cy clopenten-1-one
2.273
3.308 3.638 3.572 4.033
0.019 0.035 0.027 0.029 0.022
0.024
0.107 0.144 0.131 0.160 0.141
0.148
-
0.042 0.041 0.041 0.050 0.042 0.042
Linalool Methional
0.025
3-Methyl butyric acid
0.017 0.023 0.022 0.018 0.022 0.193 0.178 0.170
0.098 0.154
0.143
0.567
0.653
0.777 1.001 0.750 0.816
0.011
0 .015
0.014 0.014 0.013 0.015
Methyl-dihydro cyclopcnta pyrazine
3-Methylmercapto-3~ methyl butyl formiate
2.3 -Pentanedionc
Propyl pyrazine p-Vinyl guaiacol
0.125 0.180 0.127 0.183 0.175 0.143 0.404
0.512 0.416 0.533 0.551 0.478
0.015 0.019 0.014 0.020
0.018
0.016
0.304 0.614 0.445 0 .740
0.554
0.446
a.
Temperature profiles: LTLT:
Low
temperature long time. HTST: High temperature short time. HL:
High temperature and temperature reduced final stage. LHC: Continuous temperature increase from low to high. PLHC: Pre-hcatmg with subsequent LHC process. PHL: Pre-heating, high temperature, reduced final stase.
136
Results and discussion
all selec;ted
compounds (13 compounds)
100-]
herbes
(2) /
*\
^X
/ '
i
/ /;
!
\
\0-
/
//.
1
W.
\
/
n
\
/
'
,•
/
''' \
\
'1
_,*-
"**\
—
<
-
' <
\/C?^. -^^
/'
\\
,/; / sweet ;
/
with
\
caramel
,
!/ / 't
and
notes
(3)
x7
compounds with spicy, bouillon
•* ,
.
-""
Os •""T^
(5)
group of
2^V^
"
notes
\
\
>
/ /
V\ \ compounds
\ /
/
group of
(4)
V
\
bv
notes
>
1
//
/
40-
v
^/
-
i\
/ /
/ /.',
V
earthy, roasty and smoky
\ \;-v
d)
with
1
sweaty / note
"^v,. "*<•..^
> /
/
«-*""'"" ÏPfiZr"
group of compounds
group ot
compounds
compounds with
with butter
notes
(2)
potatoe notes (4)
roasting type: LTLT
Fig.
HTST
HL
LHC
PLH c
PHL
61: Influence of laboratory temperature profiles, leading to identical degree of
roast,
on aroma
compounds grouped according
to sensory properties.
Normalized presentation with the highest quantity of an aroma compound
receiving
the value 100, values added up in each group and divided
number of compounds.
by the
137
Results and discussion
4.3.5
Influence of roasting time and temperature
on
sensory quality
of the coffee beverage the development of the sensory profile during LTLT roasting
Figure 62 illustrates to
a
medium
and beyond and shows how sensory properties
degree of roast
shifted with continuing of roasting. The "green"
"roasty" note
or even a marked
note decreases
arc
in favour of
a
"burnt" note in overroasted products. The bitter taste
is increased continuously during roasting, whereas overall acidity is decreased at
least in the initial
with the instrumental analytical data
and marked increase of aroma is followed by
a
stagnation
development
The
roasting stages.
presented
pleasant
is visible in
development
in aroma
"floral"
a number
priate termination point
Figure 63 provides
as formation
The low sensory
notes
for 200
s
and 400
s
and decay of aroma of the overroasted
was rated lowest
and exhibited
The data emphasize again the
in the
score
of attributes. For example, it
and "citrus-like"
pronounced "burnt" note.
in Figure 56. A highly significant
strength between samples roasted
compounds compete with each other. product
of "aroma intensity" coincided
in the
highly significant
importance of
a
the appro¬
roasting process.
the sensory flavor
properties of isothermally high and low
temperature laboratory roasted coffees with identical degree of roast. Deviations were
statistically significant for
the attributes "bitterness",
"green" note, "burnt"
note, "roasty" note and "aroma intensity". An apparent contrast in the "floral" note was
just
not significant, whereas
presumably due
to difficulties in
no difference could
defining
and
rature roasted coffees turned out to
be more
intensified unpleasant notes, such
as
sensory score of these
discrepancy
two
in extraction
beverages yield
different solids content in the
Expert panel tasting
of
roasted beans (results
between the
insights
on
length
of
additional
a
partially be attributed to the marked The influence of
may not be restricted to the attribute "body".
beverages from laboratory
confirmed the trends
roasting time roasting
in aroma, but also to comprise
respective coffee beans.
series of coffee
not shown)
powerful
may also
beverage
components.
distinguishing this note. High tempe¬
"burnt" and "bitter". However, differences in
of the
but also affect flavor
be seen in the "spicy" note,
and industrial
concerning
the relation
and the sensory profile. Moreover, it provided
factors influencing the
aroma
quality. A marked
138
Results and discussion
general divergence
was found between coffees roasted in industrial scale with a low
air to bean ratio and coffees roasted in laboratory scale in full fluidized-bed with
an
air to bean ratio that is several orders of magnitudes greater than in industry.
Laboratory roasted beans usually tasted more bland, dull industrial
products
roasted under
equivalent temperature
oxidation clue to intensified contact with oxygen
stripping by
the hot air stream may
difference. The rather
or
due to
tively, would have relevant consequences
on
roaster
design roasters that operate
allowing
for the creation of
bean
in
to
the
conditions. Greater
physical aroma
air to bean ratios
design and
It may be necessary to
accurately limited
to
are
finding, if confirmed quantita¬
lopment.
a
compared
provide reasonable explanations for this
of roast coffee. This
quality
as
an actual
preliminary result suggests that high
detrimental to the flavor
contact
and flat
air to bean ratios,
on low
enclosing "microclimate",
process deve¬
and with oxygen
required dose. This point will be further discussed
chapter 4.4.3.
The
comparison between instrumental
there is aroma
no
proportional
or otherwise
impact compounds
progress in
and sensory analytical data revealed that
simple relationship between
the
quantities of
and the sensory quality of the beverage. Substantial
understanding this relationship
has been achieved
recently by Czerny
lot more research is needed
et
on these
al. (1999) and other authors. Nevertheless,
a
complex connections between instrumental
and sensory data of coffee aroma.
139
Results and discussion
acidity 100
roasty
aroma
intensity
***
note
***
floral
burnt note
**
note
bitterness
green note
citrus-like note -•-
200
s n
•*--400s
11
=
** =
*-
-•—
Fig.
600 800
s s
(fully roasted)
p<0.01
*** =
p
<
0.001
(excessively roasted)
62: Sensory profiles of coffee beverages from LTLT laboratory roasted beans
of increasing degree of roast. The
corresponded
to
a
medium
degree
product with of roast.
a
roasting time
of 600
s
Results and discussion
140
acidity 100
T
aroma
bitterness
intensity
roasty
green
*
note
note
burnt note
=
p
** =
Fig.
p
floral note
spicy note
n^8 *
*
< 0.05
-*-HTST
< 0.01 -*-
63: Influence of HTST and LTLT roast
on
laboratory roasting
the sensory profiles of the
LTLT
to the same
degree
respective coffee beverages.
of
Results and discussion
4.4
Changes
4.4.1
Gas
At the end of
141
of the roasted product during storage
desorption
roasting,
major part
a
only released during storage. very slowly, yet. it is be
Gas
of this gas is
desorption in whole beans
strates the gas desorption
and Puhlmann.
of various whole bean
curves
storage period. Since the amount of gas within
during roasting, identical degree different roasting conditions
different had an
two
the bean and is
is known to
greatly accelerated by grinding and storage in
ground coffee (Radtke, 1975, Meister
visible when the
entrapped within
on
a
products during
a
2 month
continuously increased
comparison
of the effects of
gas desorption during storage. This fact is
clearly
desorption curves of LTLT roasted beans with slightly
degrees of roast
compared. However,
are
equally important influence
by HTST than by LTLT roasting
on
the process temperature profile
gas desorption. More extensive gas formation
was
expressed once more
pressure at the end of storage. Moreover, HTST roasted
in
greater headspace
samples showed much
greater desorption rates. Green beans of identical origin, but subjected nation
the form of
1989). Figure 64 demon¬
bean is
of roast is crucial for
proceed
to decaffei-
by cthylacetate, formed roughly equal amounts of gases during HTST
roasting. This result must
not be generalized,
as
it
depends
on
the
applied decaf fein-
ation technique. However, regardless of equal gas quantities, decaffeinated beans
exhibited greater initial Different
desorption rates.
desorption rates
the microstructure may
location of may be
entrapped
regarded
the gas is
as
are
mainly
play
an
due to different pressure gradients. Tn addition,
important role for desorption properties.
gas in the cells is
not
yet clear. Cell lumina of roasted beans
pressurized gas-filled containers.
substantial amount of gases
can
cytoplasmic layer
and in the
micropore network
gas location most
likely
The size of cell wall
opposed
The fact that
easily released during grinding supports this theory.
to mass
is
a
a
major part
of
On the other hand,
a
be assumed to be located adsorbed to the modified of the cell walls. The true nature of
combination of the two
micropores would have
transfer. Since
micropores (see 4.2.4).
The
a
assumptions.
major influence
on
resistance
high temperature roasted beans develop larger
the structural differences may contribute
considerably
to
Results and discussion
142
greater desorption rates of these products. This theory of structure-related influence is
supported by
equal
the greater desorption rates of decaffeinated roast coffee. Due to
gas formation, the pressure gradient and therefore the
feinated beans may be considered the
hysteresis between
two
equal
to that
desorption curves
in the untreated
is most
by structural differences. Greater desorption rates
may be
changes
This
assumption concurs with additional observations
during both
Packaging problems with roast coffee been
solved
long
ago
Moreover, bean
as
with
are
to full extent
roasting step.
in industrial practice.
in air tight bags due to gas desorption have
by introducing vent packaging materials (Radtke, 1975).
importance, but may also affect staling. compounds
the
due to more severe struc¬
the decaffeination and the
However, the gas formation and desorption behavior
aroma
in decaf¬
sample. Hence,
probably caused
tural
of the cell walls
driving force
partially swept
is not
only
of
technological
It would not be unreasonable to assume that
away
together with
the
escaping gases.
major component gases, diffusion of aroma compounds
is subject to the same conformity of physical laws
out
of the
imposed by microstructure.
143
Results and discussion
1100
0
5
10
15
20 25
30
35
40
45 50
55
60
Storage time (days) Fig.
64: Gas desorption during storage of differently roasted coffees with identical
degree of roast and of equally roasted coffees degree
of roast.
with
slightly different
144
Results and discussion
Oil migration
4.4.2
Roasted coffee beans exhibit
Figure 65 provides of the
The
a
occasionally
droplet distribution
was
oil
exposed
to
already in amounts
a
given
mainly influenced by
tend to more
severe
less
severe
oil
droplets appeared
not restricted to
the entire surface. Thereafter,
become visual by eye. For process is
or
microscopic view of the phenomenon. During
migration process, numerous small
evenly over
a more
the
raw
the bean surface.
specific surface areas, but spread
they coalesce
to
larger droplets and
of roast (Table 13). Darker roasted beans
degree
oil
the initial stages
material, the extent of this oil migration
migration. Online
roasting revealed that
on
"oil sweating".
process observations of coffee beans
migration
can even
develop
to visible extent
the roaster. With excessive roasting beyond usual degrees of roast large
of oil suddenly emerged
on
the bean surface. Starting from certain spots of
the surface it soon covered the whole bean with
an
local injury of the bean surface also known
"tipping", where small bits of bean
tissue
are
as
oil film. It may have been due to
burst off.
Provided the same
degree
of roast, roasting conditions govern the
subsequent
oil
migration process (Figure 66). High temperature roasted coffees developed much more
surface oil than low temperature roasted
products. Migration ended after
a
storage period of approximately 2 months. Structural
changes
organization
in the coffee bean tissue during roasting destroy the native cell
and mobilize the coffee oil (see 4.2.3). The high gas pressure gradient
between the bean core, the outer bean parts and the exterior may drive the oil the bean.
Additionally,
chapters 4.2.3
the flow may be assisted
and 4.2.4. the oil transport
can
by capillarity.
uniform distribution of oil
droplets displayed in Figure
generally permeable, three-dimensional,
Accordingly, rence
oil
droplets
can
narrow
size of free ways for oil
modified process.
cytoplasmic matrix,
to
use
of
by roasting.
an
The
65 supports the model of
a
wad-like network of polysaccharides.
emerge everywhere
is not restricted to the openings of
of
As outlined in
be assumed to make
extensive micropore network developed in the cell walls of beans
out
on
the cell surface and their occur¬
major cracks
in the surface. Flowever, the
pass, together with
may make up for the
a
high viscosity
of the
slowness of the oil migration
145
Results and discussion
Similar
to gas desorption,
the oil migration is determined by gas pressure and micro-
structural factors. The gas pressure may
Therefore,
a
greater driving force
beans. Also the structural
can
act as the
driving force
for oil migration.
be expected for high temperature roasted
pre-conditions
in HTST roasted coffees favor oil
migration, since HTST coffees develop larger cell wall micropores. Consequently, minimal oil migration
can
and light degrees of roast.
be achieved employing low temperature roasting profiles
Results and discussion
Fig. 65: Cryo-SEM micrographs
146
of t he surface of
a
high temperature dark roasted
coffee bean, illustrating the initial stage of the oil migration process. 65a:
Immediately after roasting. Smooth epidermal cell surfaces. 65b: After day
of
storage. Numerous very small oil droplets (Images: B. Frey, S. Handschin).
cover
1
the surface
147
Results and discussion
Tab. 13: Influence of the degree of roast
on
the extent of oil migration
during stor¬
age of high temperature roasted coffee beans.
Roast loss
Surface oil after 33 d storage
(%)
(g oil/100 g bean)
Linear regression
14.28
0.064
regression coefficient
15.45
0.325
r
16.17
=
0.992
0.441
16.86
0.628
17.35
0.646
-^ 1.8-1 CO
E °>
1-6 1.4
o
9. o CD
1.2
LOH
Storage time:
0.8
Z]42d
0.6 H
H64d
CD Ü
03
00
0.4 0.2 H
0.0
HTST
Fig.
66: Surface oil identical
on
LTLT
HTST and LTLT
degree
of roast after
laboratory roasted coffee beans with
storage.
148
Results and discussion
Staling
4.4.3 The
unprotected aroma of fresh roast coffee starts
Oxidation is assumed to
bean is
play
an
to deteriorate soon after roasting.
important role in this staling process.
apparently well protected against oxidation by
(4.2.3) and by antioxidative constituents, such
as
foods
are
the native cell organization
chlorogenic acids (Morishita and
Kido, 1995). However, these protective capacities
during roasting.
The green
destroyed
are
large extent
to
On the other hand, some Mai Hard-products of thermally processed
well-known to exert antioxidant effects (Nienaber
and Eichner,
1995,
Severini et al., 1994 and numerous other authors).
Roasting-induced antioxidant capacity Figure time
67 shows the effect of roast coffee
as determined with
the oil caused
green beans on
powder
in soy bean oil
by roast coffee powder indicates
an extension
an
of the time of induction
antioxidant activity. Even with
was observed.
capacity during roasting. Nicoli
While they found reaches
a
an
maximum,
roast. These
our data
present
a
suggests
a
enhancement of the
to
the
a
similar deve¬
degree
may be
on
than LTLT coffee
on
of roast.
in which the antioxidant capacity
continuous increase up to dark
roasting conditions
on induction time
Increasing effects
found
due to different
the
the same
a
degrees of
roasting conditions.
development
is shown in Table 14. HTST roasted beans exhibited
capacity
more
et al. (1997)
optimal degree of roast
contrasting developments
The influence of the
effect
an
of antioxidant properties of coffee brews in relation
lopment
the induction
the Rancimat© method. The increase of induction time of
induction time with darker degrees of roast suggest
antioxidant
on
of the antioxidant
substantially greater
degree of roast. This result
superior antioxidant potential in high temperature roasted coffee
due to
intense formation of protective Maillard-prodncts. It is in accordance with
various
previously described differences
in the formation of chemical
compounds
during roasting. However, concerning oxidation processes, this superior antioxidant potential
in HTST beans may probably not be effective
considerable disadvantages resulting from access
a more
for oxygen. A similar interaction and
competitive factors have also been reported
enough
to make up for
open microstructure with
process-dependency
for hazelnut
the
greater
of these
roasting by Severini
two
et al.
149
Results and discussion
(1994) and by Perren (1995) treatment
as
well
as
for model
systems during high temperature
by Severini and Lerici (1995).
Nevertheless, comparative results between differently roasted products obtained with the Rancimat® method must be
interpreted with
due care and attention.
Rancimat© method is subject to fundamental limitations, increases of water
conductivity (Sandmeier
and
as
it measures merely
Ziegleder. 1985). Moreover, with
differently roasted coffees other factors than antioxidant activity
Although very finely ground, differences desorption
may be involved.
in the particle size distribution and in gas
of these particles may influence the diffusion of antioxidant volatiles into
the soja oil and
potential.
The
cause
different
grades of realization
of the
present antioxidant
150
Results and discussion
4.0-
__...•
3.9-
.•-•'""' 3.80
,•'"' 3.7-
c o
3.6-
O
3.5-
/
..--#/
T3
3.4-
3.3-
Reference
s
< 3.2'"
i
'
'
I
0
40
I
'
i
80
120
'
'
i
160
I
200
Roasting time (s) Fig.
67: Influence of
roasting time
on
the antioxidant properties of ground HTST
laboratory roasted coffee, expressed
suspension determined with
induction time of
as
the Rancimat method.
reference (3.22 h). Medium degree of roast a t 160
Pure
soybean
oil
oil
as
soybean
s.
Tab. 14: Incremental effect of roast coffee powder from differently roasted beans on
the induction time in reference to pure soybean oil.
Samples were
roasted to the same degree of roast. Increment of induction time (h)
(max. deviation)
Reference: Pure soja oil Green coffee
0
0.28
-
(0.05)
LTLT roasted coffee
0.32
(0.01)
HI Si roasted codec
0.68
(0.05)
Results and discussion
151
Oxidative reactions Tlie
staling process of roasted coffee beans during storage
is accelerated with more
intense exposure to oxygen. For this reason, oxidation reactions
play a key role in
the
are
most
likely
to
staling process. Recent studies showed an extensive formation
of free radicals during the final stages of roasting and
a
subsequent decrease of these
radicals during storage (Santanilla et al.. 1981, Baesso et al., 1990. Hofmann et al..
1999a and 1999b). These radicals
are
known to induce oxidation reactions.
However, the lipid fraction of roast coffee turned oxidation.
Headspace analysis
as
found
only small quantities
methane, ethane and pentane.
already present in minor quantities in freshly roasted beans (see
chapter 4.1.4) and slightly increased during storage. was even
be rather resistant to
of stored coffee beans showed
of typical secondary lipid oxidation products, such These alkanes were
out to
during forced oxidation
A slow formation of pentane
of extracted oil from roasted coffee beans
and confirmed the relative stability of the lipid fraction. It may be caused
protective action
of MaiUard reaction
coffee beans
also
oxidation is
was
et al.
(1993). The results suggest that
mainly affecting compounds other than lipids.
The considerable oxygen
consumption immediately after roasting reported by Hinman (1991) among others, for the oxidation of sensitive flavor The results from were
may be used,
compounds.
headspace analysis from differently roasted beans during storage
difficult to interpret,
it
as
was
unclear, whether the measured oxidation
products were already present after roasting and desorbed later result of oxidation
during storage. Nevertheless,
structure and gas desorption
or
if
they were
step staling process with
of differently roasted beans had
a first
by oxidation.
properties have
a
an
impact
on
major impact
likely also
rates
of oxidation for low-density coffee beans
wall
to
a
during
the second
physico-chemical processes,
a
but
step. Hmman (1991) found greater
as
compared
substantial influence of t he bean pore structure
micropores
two
assumed, that structural product
at least on these
most
points
a
the
step determined by physiochemical processes and
It may therefore be
on oxidation
the
it seems that differences in micro-
oxidation processes during storage. Holscher and Steinhart (1992b) proposed
second step
the
The slow process of oxidation of
products.
reported by Nicoli
by
to
regular coffees. This
on oxidation.
Larger cell
and increased area of internal surface (4.2.4) may provide easier
152
Results and discussion
access
in
for oxygen and more extensive exposure of sensitive compounds to oxidation
high temperature roasted low-density coffees. Consequently, may lead to
roasting processes
a more
low temperature
stable product against oxidation and staling.
For foodstuffs in general and for roast coffee in particular, oxidation is
being exclusively detrimental
to the
regarded
as
product quality by most authors (Radtke, 1982,
Hinman, 1991, and other authors). As
consequence to the extreme,
a
roasting,
grinding and packaging processes should employ completely oxygen-free techno¬ logies. In general, this
may be indeed valid for the most
long-term situation during storage
of the
part and
product. However,
it is worth considering
that oxygen may be needed during aroma formation and that oxidation eventually could
unpleasant compounds hypothesis
is
improve
such
supported by
a
as
the
limited extent of
sulfur-containing compounds. This
sensitive
frequently claimed observation that freshly roasted
aspect, there
to this
a
compounds profile by oxidizing
aroma
coffee comes to maximum flavor quality only after
According
may apply to the
may be
no
a
few hours of exposure to air.
requirement
for the
development
of
oxygen-free operating conveying, storage bin and grinding equipment. Nevertheless, the situation within the roaster may again be different, since the
elevated temperatures cause final stage of the
greatly accelerated chemical reactions.
roasting process could
be
An
oxygen-free
perfectly effective in preventing
the
beans from excessive oxidation and might be beneficial for the product aroma. So
far, the oxidation processes during roasting work
on
are
inadequately understood.
the formation of free radicals and oxidation during roasting is
Changes
of the aroma
The different aroma
at first with
required.
compounds profile
compounds of coffee beans exposed
rature behaved very individually during storage (Table
tration
Further
to air and ambient
15). An increase of concen¬
subsequent decrease later during storage, such
furfurylmercaptan by Tressl
et
al.
(1979),
was
tempe¬
as described
not observed. Acetic
for
acid and
2,3-butanedione exhibited only minor losses during storage. Still, all other
compounds listed were subject decrease
was found
to a
more
or
less
severe
loss
or
decay.
A
large
for 2-ethenyl-5-melhyl pyrazine, linalool and propyl pyrazine.
The average percentage of aroma
compound loss
for HTST and LTLT roasted beans
Results and discussion
was around
aroma
153
57 % and differed only slightly for the
compounds quantities
compared
to LTLT
concentration
was
products.
after storage
In contrast, since the average initial aroma
quantities
of numerous
compounds
products,
compounds
the
showed substantial
2-ethenyl-5-mcthyl pyrazine,
2-methyl butanal experienced substantially higher
relative losses in HTST roasted beans. In contrast, ß-damascenone
compound to experience considerably greater losses relative losses of 2-furfurylthiol
compound
compounds
the percentage of loss did not differ greatly
differences for HTST and LTLT roasted beans. linalool and
as
68.
Figure
between different processes. However, certain
2-furfurylthiol,
HTST roasted
HTST
higher for the high temperature roasted
was much
beans. This finding is clearly visible in
case
processes. Absolute
slightly greater in
were
substantially higher in freshly
average loss of absolute
In
two different
has been described
seem
as a
to
was
the only
in LTLT roasted beans. Greater
be particularly meaningful, since this
key-role player
in the
staling process (Tressl
et
al., 1979). From various proposed staling indicators only the butanedione/2-methyl furan ratio (Leino et al., 1992)
was
applicable
to our data. This ratio increased from
2.10 to 2.80, and from 0.93 to 1.22 in HTST and LTLT roasted beans, respectively. A further
developed staling process
these figures,
or
that the
storage conditions but
not
Unprotected coffee beans
in the HTST coffee may be concluded from
concept of staling ratios is only applicable for different for the distinction of different roasting conditions.
subject
are
to extensive
profile during storage. Aroma compounds
undergo further chemical reactions, such close
relationship between
(Nicoli
ct
are
as
changes
in the aroma
either lost due to diffusion,
oxidation. Some authors
comprehensive view
may indicate
a
shift in the
not
only loose aroma compounds, but also experiences
proportion
of the
products.
a
a
a
compounds of
result, roast
considerable
compounds, since each compound reacts individually
to the various influences. The results
process in
they
play together
the three factors micro structure, gas desorption and oxidation. As
coffee does
or
suggested
the gas desorption and the losses of aroma
al. 1993). A more
compounds
high temperature
provide evidence for
a more severe
staling
roasted coffees than in low temperature roasted
154
Results and discussion
Tab. 15: Loss of selected aroma compounds during storage of high and low tem¬
perature roasted coffee beans for 81 days
fresh
Ax/ Aistd (-) Acetic acid
LTLT roasted coffee
HTST roasted coffee
Compound
at 25 °C and exposed to air.
stored loss
Ax/ AlStd (-)
(%)
fresh
stored loss
Ax/ Aistd
Ax/ Aistd
(%)
(-)
0.56
0.42
25
0.26
0.26
0
2,3-Butancdione
0.21
0.14
33
0.14
0.11
21
ß-Damascenone
0.04
0.02
50
0.05
0.01
80
2,3-Dimethyl pyrazine
0.50
0.19
62
0.33
0.13
61
2,5-Dimethyl pyrazine
1.28
0.46
64
0.94
0.35
63
2,6-Dimethyl pyrazine
1.33
0.49
63
0.96
0.36
62
0.11
0.03
72
0.07
0.03
57
0.05
66
0.13
0.04
69
2-Ethenyl-5-methyl pyra¬ zine
2-Ethyl-3,5-dimethyl pyra¬ 0.15
zine
3-Ethyl-2,5-dimethyl pyra¬ zine
0.34
0.13
65
0.28
0.09
68
2-Ethyl-3-methyl pyrazine
0.15
0.06
60
0.12
0.05
58
2-Ethyl-5-methyl pyrazine
0.31
0.10
68
0.25
0.09
64
2-Ethyl-6-methyl pyrazme 0.44
0.14
68
0.36
0.12
67
0.23
68
0.47
0.17
64
0.94
62
0.89
0.45
49
0.34
0.14
59
0.03
0.01
67
Ethyl pyrazine
0.71
2.50
2-Fur fury] thiol
Guaiacol
2-Hydroxy-3-methyl2-cyclopenten-1 -one Kahwcofuran Linalool
2-Methyl butanal
3-Methyl butyric acid
2,3-Pentanedione
0.28
56
0.04
0.04
0
0.25
0.09
64
0.15
0.05
67
0.06
0.01
83
0.05
0.02
60
0.46
0.23
50
0.29
25
2.00
1.J4
43
1.27
0.73
43
0.63
0.25
60
0.38
0.19
50
0.02
72
0.04
0.01
75
0.07
Propyl pyrazine 2.3.5-Trimethyl pyra/ine
p-Vinyl guaiacol
0.63
0.46
0.37
0.13
65
0.29
0.10
66
2.21
1.09
51
0.98
0.46
53
155
Results and discussion
c
o
1 .
le
1
-<
~a
t
'n
Ü
ça
T3
CO
Ü
CO
>,
CL
CCS
">>
CD
.9
u_
<
o
CD
"3 JO
<
c
o
©
T5
©
0.1
-
in
© c
cô
©
c
© C
"n
T3
TD
co" <
c
JZ
E
o
©
"3 CQ
a
CO
">* a
t
CM
CD
0_
o
x:
a.
+-•
t
CO
LU
csi
CM
,
,,
MWHtt
J^V
~
I
r
roasting process F/g.
68: Loss of aroma
compounds during
a
storage period of 81 days for HTST
and LTLT roasted coffee beans. The beans ambient temperature during storage
were
exposed
to air and
156
t.j
*U
4
r
-vi»
J
*»
f
,
V
.
i&1 "i
\
/
%i.^
i
157
5
5.1
Conclusions
Hot air
Critical process factors of coffee beans is
roasting
a traditional
thermal process which in spite of its
great importance in practice is still designed and operated mainly basis. The
coffee
principal objective and
aroma
a
subject
Instability of compounds in any kind of coffee
development is to
empirical
of the roasting process is to create the desired roast
flavor-full cup quality. The
roasted coffee beans is
on an
to
rapid
unprotected aroma fraction
and substantial deterioration after roasting.
the aroma profiles during storage is
product. Therefore,
of
the
most
a
critical factor for
demanding challenge of process in the bean
to achieve favorable chemical and structural conditions
oppose staling. In order to achieve these process objectives the
following roasting
factors and transformations during roasting must be taken into account:
Quality of green coffee beans The botanical variety, the
the
origin and
a
major
The initial
water
the processing of the green beans have
and the final
product quality.
impact
on
content
of the green beans is of particular technological importance,
may
roasting process
be controlled
dehydration.
by
a more
strictly specified procedure
of
as
this factor
post-harvest
The water content influences the bean temperature, the development of
the bean structure and all chemical reactions.
Process temperature The
development
of bean
core
temperature presents the most
important roasting
parameter and influences flavor formation and structural product properties
to
a
great extent. Different temperature profiles affect dehydration, which in turn deter¬ mines the specific conditions for chemical reactions in the bean. This is reflected
obviously in realization of
the formation a
of C(K browning
distinct profile of aroma
and flavor
compounds
out
development.
of the aroma
The
potential of
158
Conclusions
on
these reaction conditions. Out of the
hundreds of volatiles, it is
a small number
of temperature dependent aroma impact
compounds that dominate
the aroma of roast coffee. Low
the green beans is
highly dependent
temperature conditions
result in inadequate formation of aroma compounds. The highest rates of aroma
compound formation and
at
observed at
a
bean water content below 5 g /lOO g (wb)
temperatures exceeding 200 °C. On the other hand, aroma formation is super¬
imposed by
during
arc
an accelerated
decay
of some aroma
compounds
at
high temperatures
the final roasting stage.
Structural changes of the bean
for bean expansion
driving force factors that
are
are
again related
to
as well as
three states
structure resistance.
As
bean volume,
a
the temperature profile. The
the structural resistance
temperature and dehydration.
phenomena related
a
equally affected by
opposed
A
to
it
are
glass transition
involving development presumably
result, high temperature roasted beans exhibit
controls
a
greater
higher cumulated pore volume and larger cell wall micropores than
low temperature roasted coffees of the
same
degree of roast.
Hot air humidity
The
humidity of
the hot air must be considered
as
another important process
parameter. Industrial roasters using air recirculation systems from the beans and from water quench cooling
roasting atmosphere
may
so that a
it is assumed that some reactions and
accumulate water
significant humidity in the
be generated. Elevated humidities
specific heat capacity of the hot air and result in addition,
can
a more
cause
an
increased
efficient heat transfer. In
changes that depend on water content
are also affected.
Air-to-bean ratio The amount of hot air in relation to the batch size turned out to be feature of roaster design and
large batch, the application of
a
product
very important
operation. Provided adequate mechanical mixing in a
low air to bean ratio results in
cup-quality. In contrast, excessive to
a
air streams such
as
in
a
a
coffee of superior
fully fluidized-bed lead
of bland, dull and flat sensory properties. A lower ratio is assumed to
prevent physical aroma stripping and excessive contact with oxygen and favorable "microclimate"
enclosing
to create
a
the beans. Conventional conductive type
159
Conclusions
roasting systems of industrial size generally operate with reasonably
low air to bean
ratios, mainly for economical reasons. Gas formation The
large amount of internal gases formed during roasting
changes,
but also
not only acts as
the
important role concerning
driving force
for structural
mass transfer
and staling during storage. The loss of aroma compounds appears to
be
closely related to
plays
an
gas desorption.
Transformation of structure The structural organization of the native coffee seed, even after drying, provides far-
reaching protection against
compartialization.
adverse external
The
impacts.
sophisticated cell
the storage of lipids within oleosomes, and the unusually thick
cell walls obviously fulfil
specific physiological tasks.
completely changed during roasting.
The native structure is
The cell compartialization is destroyed, coffee
oil is mobilized, and the cell walls become increasingly porous and new structural
as
properties
of roasted coffee beans
depend
on
permeable.
The
the roasting conditions,
outlined above for different temperature profiles. In addition, the present investi¬
gations show
a
strong interaction between bean microstructurc and mass transfer
involving chemical and physico-chemical processes during storage.
A more porous
microstructure seems to disadvantageously favor mass transfer and to accelerate the
staling process. Greater pore volume and larger micropores in high temperature roasted beans promote faster gas desorption and oil migration, and may enhance access
for oxygen, resulting in accelerated loss and decay of aroma compounds. A
considerably more stable bean at the
is achieved at low temperature conditions, although
expense of aroma "strength".
Oxidation
Sensitive
aroma
Oxidation rates and
compounds and lipids
are
determined by
a
are
the target of oxidative processes.
complex interaction
inhibiting factors. Native antioxidants
are
of
a
scries of promoting
destroyed, but replaced by
a
roasting-
induced antioxidant capacity of MaiUard type products. On the other hand, roasting is known to form
reactions.
a
substantial amount of free radicals that induce oxidative
Availability
of oxygen
can
be regarded
as
the limiting factor for the
Conclusions
160
progression of oxidation
and staling. It is evident that this factor is determined by
the structural properties of the roasted beans.
5.2
Process
optimizations
Different coffee manufacturers put individual priorities
properties, mainly depending coffees
or
on
reactions and
the desirable
product
whether they produce roasted beans and ground
soluble coffee. The present investigations
desirable product properties
on
clearly show that
be maximized at the same time, because
can
changes are reacting in
not
all
not
all
the same direction to changes in process condi¬
tions. Therefore, process optimization requires specification of
a
compromise in
target quality.
Roasting technology cannot make for
a
up for poor quality of the
given type of green coffee blend, roasting
aroma
quality
6 min,
a low
air
is the main flavor determinant. High
depending
to bean ratio, an
on
profile. A roasting phase
generating sufficient aroma strength.
other hand, high temperature conditions
profile and should
optimal roasting time should be
the target flavor
medium temperature is essential in
on
However,
is achieved with moderate, non-extreme processes of medium
temperatures. Provided
longer than
raw material.
be avoided. A final
generally cause
phase
at reduced
an
at
On the
unfavorable aroma
temperature has
an
impact
the aroma profile. The target degree of roast should not be set too dark, and the
process must be terminated in time because decay of aroma
final
roasting stage. Only green coffee
of
high quality
compounds during
the
may withstand more severe
roasting conditions. The
highest porosity in bean structure
is achieved
by high temperature conditions
and leads to maximum extraction yield. Nevertheless, such
believed to
provide
an
Oxygen contact should
unfavorable structure
be limited
precisely
to
to the
a
low density product is
oppose oxidation and staling.
required level by
two measures.
low air to bean ratio reduces the amount of air that is in contact with the
First,
a
beans
during roasting and creates
Therefore,
a roaster
of a favorable "microclimate"
design operating with
heat transfer may be
a
enclosing the beans.
fairly high proportion of conductive
advantageous. Secondly,
the
implementation
of moderate
161
Conclusions
temperature profiles assures the generation of favorable structural pre-settings in the bean for storage. For the most part, there is
conveying aroma
and grinding
no
requirement
for
completely oxygen-free roasting,
technology. Some oxygen
may actually
be needed for
formation. On the other hand, the accelerated chemical reactions at elevated
temperatures during the final roasting stages might play
a
key role in the subsequent
staling process. Oxygen-free roasting during the final roasting stages be worth of consideration for further investigations.
may therefore
162
J
^
*S>.
X
»..H
»
>r"
v
^
,.
Si
,
»*U i
U »i r*
i Ve.* la,«
l)
(Î
163
6
References
Acree, T. E., Barnard, J. and Cunningham, D. G. 1984. A procedure for the sensory analysis of gas chromatographic effluents. Food Chemistry, 14: 273286.
Adamson, Wiley
A. W. 1990. Physical Chemistry of Surfaces, 5th ed. p. 567. John & Sons Inc.. New York.
Angermüller,
S. and Fahimi, H.D. 1982. Imidazole-buffered osmium tetroxidc:
excellent stain for visualisation of lipids in transmission electron
Histology Journal,
an
microscopy.
14: 823-835.
Arackal, T. and Lehmann, G.1979. Messung des Quotienten 2-Methylfuran/ 2-Butanon
ungemahlenem Röstkaffee während der Lagerung Luftausschluss. Chemie Mikrobiologie Technologie der Lebensmittel, von
unter
6 (2):
43-47.
Bade-Wegner, H., Holscher, W. and Vitzthum, O. G. 1993. Quantification of 2-methyhsoborneol in roasted coffee by GC-MS. Proceedings 15th ASIC Colloquium. Paris, p. 537-544. Bade-Wegner, H., Bendig, I., Holscher, W. and Wollmann, R. 1997. Volatile compounds associated with the over-fermented flavour defect. Proceedings 17th ASIC Colloquium, Paris, p. 176-182.
Baek, H. H. and Cadwallader, K. R. 1998. Roasted chicory aroma evaluation by gas Chromatography/mass spectrometry/olfactometry. Journal of Food Science, 63 (2): 234-237.
Baesso,
M.
L.,
Da Suva, E.
C, Vargas, H., Cortez, J. G. and Pelzl, J. 1990. Use
of electron spin resonance for the determination of staling of roast coffee in
polyethylene bag packs. -Forschung, 191:24-27. Baltes,
Zeitschrift
W. 1977. Rosteffekte auf die
ASIC
Colloquium. Paris,
Institute
Lebensmittel-Untersuchung
Kaffeezusammensetzung. Proceedings
von
8th
Dosenkonserven Entwicklung einer
Analysenmethode. Ph.D. thesis No. 8952, of Technology (ETH). Zurich. Switzerland. und
und
p. 85-96.
Bertoli, C. 1989. Die Kopfraumanah se
Apparatur
fur
Swiss Federal
164
References
Blank, 1., Sen, A., Grosch, Wr. 1991. Aroma impact compounds of Arabica and Robusta coffee. Qualitative and
quantitative investigations. Proceedings 14th
ASIC Colloquium, Paris, p. 117-129.
Blank, I., Sen, A., Grosch, brew
W. 1992. Potent odorants of the roasted
of Arabica coffee. Zeitschrift
für
powder
and
Lebensmittel-Untersuchung und
-Forschung, 195:239-245. Baltes, W. 1992. Model reactions on roast aroma formation. Zeitschrift für Lebensmittel-Untersuchung und -Forschung, 194:
Bohnenstengel,
C.
and
366-371.
Bradbury, A. G. W. and Halliday, D. J. 1990. Chemical structures of green coffee bean polysaccharides. Journal of Agricultural and Food Chemistry, 38: 389392.
Brandlein, L. S., Schechter, S. M. and Mahlmann, J. P. 1988. Coffee roasting method. United States Patent. US 4,737.376.
Bürgin, E. 1969. Unbehandelter und behandelter Kaffee unter Proceedings 4th ASIC Colloquium. Paris, p. 63-73. Czerny, M., Mayer, F.
and Grosch, W. 1999. Sensory study
impact odorants of roasted arabica coffee. Journal Chemistry, 47: 695-699.
dem Mikroskop.
on
the character
of Agricultural and Food
Chesson, A., Gardner, P. T. and Wood, T. J. 1997. Cell wall porosity and available surface area of wheat straw and wheat grain fractions. Journal of the Science of Food and Agriculture, 75: 289-295.
Clarke,
R. J. and Macrae, R. (eds.) 1985. Coffee, Volume 1: Chemistry, 1st ed.
Elsevier Applied Science, London.
Clarke,
R. J. and Macrae, R. (eds.) 1987. Coffee, Volume 2: Technology. 1st cd.
Elsevier Applied Science, London.
Clarke,
R. ,T. 1990. The volatile
compounds of
roasted coffee. Italian Journal of
Food Science, (2): 79-88.
Clifford, M. N.
and Willson, K. C. (eds.) 1985. Coffee: Botany, biochemistry and
production of beans and beverage.
AVI Publishing Company Inc., Westport,
Connecticut. Dalla Rosa,
M., Lerici,
C.
R., Piva,
M. and Fini, P. 1980. Processi di trasfor-
mazionc del caffè: aspetti chimici, fisici
c
tecnologici. Nota
5. Industrie delle
bevande, (12): 466-472.
D'Amicis, A.
and Viani,
R. 1993. The
consumption
of coffee. In Caffeine, coffee,
and health. Garattini, S. (Ed.), p. 1-16. Raven Press. New York.
References
165
Da Porto, M.
Study
on
C, Nicoli, C, Severini, C, Sensidoni, A. and Lerici, C. R. 1991.
physicochemical changes
in coffee beans
during roasting. Note
2.
Italian Journal of Food Science, (3): 197-207.
Dentan, E. 1977. Structure fine quium, Paris, p. 59-64. Dentan, E.
du grain de café vert.
Proceedings
8th ASIC Collo¬
1985. The
microscopic structure of the coffee bean. In Coffee. Botany, biochemistry and production of beans and beverage. M. N. Clifford and K. C. Willson (eds.), p. 284-304. Croom Helm. London.
Dentan, E. and Illy, A. 1985. Etude microscopique de grains de café matures. immatures
et immatures
Colloquium, Paris, Flament,
fermentes Arabica Santos. Proceedings 11th ASIC
p. 341-368.
I. 1989. Coffee, cocoa, and tea. Food Reviews International, 5 (3): 317-
414.
Frey, B., Scheidegger, C, Günthardt-Georg, M. S. and Matyssek, R. 1996. The effects of ozone and nutrient supply on stomatal response in birch (Betula pendula) leaves as determined by digital image-analysis and X-ray microanalysis. New Phytologist. 132: 135-143.
Grosch,
W. 1990.
dilution assay.
Quantification of aroma compounds using a stable isotope In Flavour Science and Technology. Bessière, Y. and Thomas,
A. F. (eds.), p. 191-194. John Wiley & Sons. Chichester.
Grosch,
W. 1993. Detection of potent odorants in foods
analysis. Trends in Food Science & Technology,
by aroma extract dilution
4 (3): 68-73.
Grosch, W. 1995. Instrumental and sensory analysis Proceedings 16th ASIC Colloquium, Paris, p. 147-155.
of coffee volatiles.
Grosch, W., Czerny, M., Wagner, R. and Mayer, F. 1996. Studies
on
the aroma
of roasted coffee. In Flavour Science. Recent D. S. Motlram (eds.), p. 200-205.
developments. A. J. Taylor and The Royal Society of Chemistry, Cambridge.
Gutierrez, C, Ortola, M. D., Chiralt, A. and Fito, P. 1993. Analysis por meb la porosidad del café tostado. Proceedings 15th ASIC Colloquium, Paris, p. 661671.
Guyot, B., Teh ana, E. and Vincent, J. C. 1985. Corrélations entre charactéristiques chimiques du café torréfié et paramètres de torréfaction. Proceedings 11th ASIC Colloquium, Paris, p. 319-331. Hadorn, H.
and Zürcher, K. 1974. Zur
Bestimmung der Oxidationsstabilität von
Ölen und Fetten. Deutsche Lebensmittel-Rundschau, 70 (2): 57-65.
References
Hashim,
166
L. and Chaveron, H. 1996. Use of
methylpyrazine ratios
to monitor
the
coffee roasting. Food Research International, 28 (6): 619-623.
Heise,
U. 1996.
Hinman, on
Kaffee
und
Kaffeehaus. Gustav Kiepenheucr Verlag, Leipzig.
D. C. 1991. Rates of oxidation of roast and
shelf-life.
ground coffee
Proceedings 14th ASIC Colloquium. Paris,
Hinz, T., Steer, A., Waldmann, C, Cammenga, H. Röstkaffee-Extraktion:
Ingenieur Technik,
Einflussparameter
und
p. 165-174.
Eggers, R. 1997. Modellierung. Chemie
K. and
69 (5): 685-690.
Ho, C.-T., Hwang, H.-L, Yu, T.-H. MaiUard reactions related to
and
aroma
Zhang, J. 1993.
Colloquium, Paris,
Hofmann,
An overview of the
generation in coffee. "
ASIC
and the effect
Proceedings 15th
p. 5 19-527.
T. and Schieberle, P. 1998a. Quantitative model studies
on
the effec¬
tiveness of different precursor systems in the formation of the intense food
odorants 2-furfurylthiol and 2-methyl-3-furanthiol. Journal of Agricultural and Food
Chemistry.
Hofmann,
46: 235-241.
T. and Schieberle, P. 1998b. Identification of key aroma
generated
from
compounds
and
carbohydrates under roasting conditions. Zeitschrift fur Lebensmittel-Untersuchung und -Forschung, 207: 229-236.
Hofmann, T., Bors,
cysteine
W. and Stettmaier, K. 1999a. Studies
on radical
interme¬
diates in the early stage of the nonenzymatic browning reaction of carbohy¬
drates and amino acids. Journal of Agricultural and Food
Chemistry,
47: 379-
390.
Hofmann, T., Bors, W. and Stettmaier, K. 1999b. Radical-assisted melanoidin formation during thermal processing of foods as well as under physiological conditions. Journal of Agricultural and Food Chemistry. 47: 391-396.
Holscher, W., Vitzthum,
O. G. and Steinhart, H. 1990. Identification and
sensorial evaluation of
aroma-impact-compounds in roasted Colombian coffee.
Café Cacao Thé, XXXIV (3): 205-212.
Holscher,
W.
compounds
and Steinhart H. 1991. New
in roasted coffee.
sulfur-containing aroma-impact Proceedings 14th ASIC Colloquium, Paris,
p. 130-136.
Holscher,
W. and Steinhart H. 1992a.
Holscher,
W. and Steinhart H. 1992b.
Investigations of roasted coffee freshness with an improved headspace technique. Zeitschrift für Lebensmittel-Untersu¬ chung und -Forschung. 195: 33-38.
Aromaveränderungen beim Altern
Röstkaffee. Lebensmittelchemie, 40: 106-111.
von
167
References
Holscher, coffee
W. and Steinhart H. 1994. Formation aroma
compounds.
Morcllo, M. and McGorrin,
Holscher,
In
pathways for primary roasted Thermally generated flavors. Parliment, T.,
R. (eds.), p. 206-217. American Chemical Society.
W. and Steinhart H. 1995. Aroma compounds in green coffee. In Food
flavors: Generation, analysis and
process influence. Proceedings 8th Interna¬
tional Flavor Conference. Charalambous,
G. (ed.),
p. 785-803. Elsevier,
Amsterdam.
Huang,
A. H. C. 1996. Olcosins and oil bodies in seeds and other organs. Plant
Physiol.. 110: 1055-1061.
Hubbard, M. H., Phipps, R. R. and Thompson, R. L. 1979. roasting process. United States Patent. US 4.169.164.
Low density coffee
Illy, A. and Viani, R. 1995. Espresso coffee; the chemistry of quality. Academic Press Limited.. London.
Johannessen, S. 1992. A method for the drying and roasting of coffee and cocoa beans. International patent application WO 92/12643.
Kallio, IL, Leino, M., Koullias, K., Kallio, S. Headspace
of ground coffee
as an
indicator of
Kaitaranta, J. 1990. storage time. Food Chemistry, and
36: 135-148.
Karathanos,
V. T., Kanellopoulos, N. K. and Belessiotis, V. G. 1996. Devel¬
opment of porous structure during air drying of agricultural plant products. Journal of Food Engineering, 29: 167-183.
Karmas, R., Buera,
M. P. and Karel, M. 1992. Effect of
glass transition
on rates
of nonenzymatic browning in food systems. Journal of Agricultural and Food
Chemistry.
40: 873-879.
Kazi, T. and Clifford, teristics of
M. N. 1985. Comparison of physical and chemical charac¬
"high yield"
quium, Paris,
and
"regular" coffees. Proceedings 11th ASIC Collo¬
p. 297-308.
Kino, T. and Takagi, T. 1995. Effects of far infrared beating
roasting. Proceedings 16th ASIC Colloquium. Paris,
Kwasny, B
on
coffee beans
p. 397-405.
H. and Werkhoff, P. 1979.
von
Abhängigkeit der Aromaindex-Kennzahl M/ verschiedenen Parametern. Chemie Mikrobiologie Technologie der
Lebensmittel, 6(1): 31-32. C. M. 1999. The art of roasting decaf. Tea & Coffee Trade Journal. 171 (10): 90-93.
Lee,
References
168
Leino, M., Lapveteläinen, A., Menchero, P., Malm, H., Kaitaranta, J. and Kallio, H. 1991. Characterisation of stored Arabica and Robusta coffees by
headspace-GC
and sensory analysis. Food Quality and Preference, 3 (2): US-
OS.
Leino, M., Kaitaranta, J.
and Kallio, H. 1992.
Comparison
of
changes in headspace volatiles of some coffee blends during storage. Food Chemistry, 43: 35-40.
Leloup,
V. and Liardon, R. 1993.
drates.
Proceedings 15th ASIC
Analytical charcterization of coffee carbohy¬ Colloquium. Pans, p. 863-865.
Leloup, V., Louvrier, A. and Liardon, R. 1995. Degradation mechanisms of chlorogenic acids during roasting. Proceedings 16th ASIC Colloquium, Paris, p. 192-198.
Likens,
S. T. and Nickerson, G. B. 1964. Detection of certain
brewing products. Proceedings Annual Meeting, Brewing Chemists, Saint Paul, 5-13.
in
Macku, C. and Shibamoto,
hop
oil constituents
American
Society
of
T. 1991. Volatile antioxidants isolated from brewed
coffee. Proceedings 14th ASIC
Colloquium. Paris,
p. 146-155.
Mahlmann, J. P. 1986. Controlled coffee roasting. European patent application, EP 0 183 878 Al.
H. G. 1985. On the
Maier,
ASIC
Maier,
composition
Colloquium, Paris,
of fast roasted coffees.
Proceedings
1 lth
p. 291-296.
H. G. 1987. The acids of coffee.
Proceedings 12th ASIC Colloquium. Paris,
p. 229-237.
Maier, H. G. 1993. Status of research in the field of non-volatile coffee compo¬ nents. Proceedings 15th ASIC Colloquium, Paris, p. 567-576.
Margadant,
P. 1991.
Untersuchungen über die Oxidationsstabilität und über die Vakuumverpackung von Sterilkonserven. Ph.D. thesis No. 9385, Swiss Federal Institute of Technology (ETH), Zurich. Switzerla nd.
Marsili,
R. (ed.) 1997.
Techniques for analyzing food aroma,
1st ed.. Marcel
Dekker Inc., New York.
Massini, R., Nicoli, and
AL
C, Cassarà, A. and Lerici, C. R. 1990. Study
on
physicochemical changes of coffee beans during roasting. Note
Journal of Food Science (2): 123-130.
physical 1. Italian
References
169
Mayer, F., Czerny, M. and Grosch, W. 1999. Influence of provenance and roast degree on the composition of potent odorants in Arabica coffees. European Food Research and Technology (Zeitschrift fur Lebensmittel-Untersuchung und -Forschung), 209 (3-4): 242-250. McCann,
M.
C., Wells, B. and Roberts, K. 1990. Direct visualization of cross¬
links in the primary plant cell wall. Journal of Cell Science, 96: 323-334.
Meister, U.
and
Puhlmann, R. 1989. Entgasungsverhalten Lebensmittelindustrie, 36 (4): 155-158.
von
Röstkaffce.
Mohsenin, N. N. 1986. Physical properties of plant and animal materials, 2nd revised and updated ed., Gordon and Breach Science Publishers Inc., New York.
Morishita, H. and Kido, R. 1995. Antioxidant activities Proceedings 16th ASIC Colloquium, Paris, p. 119-124. Munari, M., Mastrocola, D., Nicoli, formazione dellc
sostanze
of chlorogenic acids.
M. C. and Severini, C. 1997. Cinetiche di
volatili durante la tostadura del caffè. Industrie
Alimentari. 36 (4): 454-459.
Navarini, L., Gilli, R., Gombac, V., Abatangelo, A., Bosco, 1999.
Polysaccharides from hot water extracts of roasted Coffea arabica beans:
isolation and characterization.
Nehrig, U.
M.
Carbohydrate Polymers, 40 (1): 71-81.
P. and Maier, H. G. 1992. Indirect determination of the degree of roast.
Zeitschrift fur
Nicoli,
M. and Toffanin, R.
Lebensmittel-Untersuchung
C, Dalla Rosa,
processing conditions
on
M.
und
-Forschung, 195: 39-42.
and Lerici, C. R. 1990. Influence of
solid-liquid extraction
some
of coffee. Lebensmittel-
Wissenschaft und-Technologic. 23 (5): 386-389.
Nicoli,
M.
C, Innocente, N., Pittia,
P. and Lerici, C. R. 1993.
coffee: Volatile release and oxidation reactions
15th ASIC
Nicoli,
M.
Colloquium. Paris,
p. 557-565.
C, Anese, M., Manzocco,
properties
L. and Lerici, C. R. 1997. Antioxidant
of coffee brews in relation to the
Wissenschaft und
Staling of roasted during storage. Proceedings
-Technologic,
roasting degree. Lebensmittel-
30: 292-297.
Nienaber, U. and Eichner, K. 1995. Die antioxidative Wirkung Maillard-Rcaktion in and
Technology.
von Produkten
Modellsystemen und gerosteten Haselnüssen.
der
Fat Science
97 (12): 435-444.
Nijssen, L. M., Visscher, C. A., Maarse, H„ Willemsens, H. (eds.) 1996. Volatile
compounds
in
L. C. and Boelens. M.
food. Qualitative and quantitative data.
7th ed.. TNO Nutrition and Food Research Institute. Zeist, The Netherlands.
References
Nultsch,
170
W. 1996. Allgemeine Botanik. Kurzes Lehrbuch
für Mediziner
und Natur¬
wissenschaf ter. lOth ed., Georg Thieme Verlag. Stuttgart.
Ortolâ, M. D., Londono, L., Gutierrez, C. L., Chiralt, A. 1998. Influence of roasting temperature on physicochemical properties of different coffees. Food Science and Technology International, 4: 59-66. Perren, R., Escher, F., Künzli, R. and Mayer-Potschak, K. 1994. Vorrichtung zur Ermittlung von Bedingungen beim Rosten von Olsaamen. German patent DE 43 14 606C1.
Perren,
R. 1995.
Untersuchungen über
No. 11390. Swiss Federal Institute
Perren,
R. 1997. Investigations
on
das Rösten
von Haselnüssen. Ph.D. thesis
of Technology (ETH), Zurich, Switzerland.
the hot air
roasting of nuts.
The
Manufacturing
Confectioner. (6): 123-127.
Peters,
A. 1991. Brewing makes the difference.
quium, Paris,
Proceedings 14th ASIC Collo¬
p. 97-106.
Puhlmann, R., Sobek, Kaffeebohne im
E. and Bartsch, G. 1986.
Strukturveränderungen
Röstprozess. Lebensmittelindustrie
der
33 (6): 278-279.
Puhlmann, R. and Meister, U. 1989. Lebensmittelindustrie, 36 (4): 159-160.
Radtke,
R. 1975. Das Problem der
COi-Desorption
von
Röstkaffee unter dem
Gesichtspunkt einer neuen Packstoffentwicklung. Proceedings 7th ASIC Collo¬ quium. Paris, p. 323-333. Radtke, R. 1979. Zur Kenntnis des Sauerstoffverbrauchs von Röstkaffee und seiner
Auswirkung auf die sensorisch ermittelte Qualität des Kaffeegetränks. Chemie Mikrobiologie Technologie der Lebensmittel. 6 (2): 36-42. Radtke-Granzer, R. 1982. Fortschritte in der Qualitatsbeurteilung Einflüsse auf die Haltbarkeit und Prufkriterien.
quium, Paris,
Raemy,
Röstkaffee,
Proceedings 10th ASIC Collo¬
p. 81-97.
A. and Lambelet, P. 1982. A calorimetnc
and chicory. Journal of Food
Reineccius, G. A. 1995. ASIC
von
Technology,
study of self-heating in coffee
17: 451-460.
The MaiUard reaction and coffee flavor.
Colloquium, Paris,
Proceedings 16th
p. 249-257.
Rehm, S. and Espig, G. 1996. Die Kulturpflanzen der Tropen und Subtropen. Ulmer, Stuttgart. Saleeb, F.
Z. 1975.
Adsorption of carbon dioxide on roast Proceedings 7th ASIC Colloquium. Paris, p. 335-339.
and
ground coffees.
171
References
Sandmeier, D. and Ziegleder,
bestimmungen mit
G. 1985. Zur Problematik
von
Induktionszeit¬
dem Rancimat. Süsswaren, (12): 487-489.
Santanilla, J. D., Fritsch, G.
Müller-Warmuth, W. 1981. Elektronenspinresonanz-Experimcnte an Roh- und Röstkaffeeproben. Zeitschrift für Lebens¬ mittel-Untersuchung und-Forschung, 172: 81-86. and
Schenker, S., Handschin, S., Frey, B., Perren, R. and Escher, F. 1998. Verifi¬ cation of mercury intrusion into coffee beans
by scanning electron microscopy and X-ray microanalysis. Scanning. The Journal of Scanning Microscopies. 20 (3): 273.
Schenker, S., Handschin, S., Frey, B., Perren, properties
of coffee beans
18th ASIC
Colloquium, Paris,
Schenker, S., Handschin,
as
influenced
R. and
Escher, F. 1999. Structural
by roasting conditions. Proceedings
p. 127-135.
S., Frey, B., Perren, R. and Escher, F. 2000. Pore
structure of coffee beans affected
by roasting conditions. Journal
of Food
Science, (in press). Schieberle, P. and Grosch, Kruste
von
W. 1983.
Identifizierung
Roggenbrot. Zeitschrift
fur
von
Aromastoffen
der
aus
Lebensmittel-Untersuchung
und
-Forschung, 177: 173-180.
Scholz-Böttcher,
B. M. and Maier, H. G. 1991. Isomers of quinic acid and
quinidcs in roasted coffee. Indicators ASIC Colloquium, Paris, p. 220-229. Semmelroch, P.
of the
and Grosch, W. 1995a.
degree of roast
-Technologic.
Semmelroch, P., Laskawy, G., Blank, Fragrance Journal,
Semmelroch,
28: 310-313.
by stable isotope dilution
on
coffee brews. Journal of Agricultural and Food
heating rate
during 648.
the
M.
on some
assays. Flavour
10: 1-7.
P. and Grosch, W. 1996. Studies
Severini, C, Nicoli,
headspace samples. Lebens¬
I. and Grosch, W. 1995b. Determination
of potent odourants in roasted coffee
and
Proceedings 14th
Analysis of roasted coffee powders and
brews by gas chromatography-olfactometry of mittel-Wissenschaft und
?
C, Mastrocola, D.
physical
character
Chemistry,
impact odorants
of
44: 537-543.
and Lerici, C. R. 1991. Influence of
and physico-chemical properties of coffee beans
roasting process. Proceedings 14th ASIC Colloquium, Paris,
p. 641-
References
172
Severini, C, Pittia, P., Nicoli,
M. C. and Pinnavaia, G. G. 1994. Interactions
between MaiUard reaction
products and lipid oxidation during the roasting process and storage of hazelnuts (Corylus avellana). In: Developments in Food Engineering. Proceedings of the 6th International Congress on Engineering and Food. Yano, T., Matsuno, R. andNakamura, K. (Eds.), p. 289-291. Blackie Academic & Professional. London.
Severini,
C. and Lerici, C. R. 1995. Interactions between the MaiUard reaction and
lipid oxidation
in model systems during high temperature treatment. Italian
Journal of Food Science, (2): 189-196.
Silwar, R.
and Lüllmann, C. 1993. Investigation of aroma formation in robusta
coffee
Sivetz,
M.
ASIC
Sivetz,
during roasting. Café Cacao Thé, 1975. Fluid bed drying and
Colloquium. Paris, and
M.
37 (2): 145-152.
roasting of coffee beans. Proceedings 7th
p. 359-366.
Desrosier, N.
W. 1979.
Coffee technology.
AVI Publishing
Company Inc., Westport, Connecticut. M. 1991. Growth in
Sivetz,
use
of automated fluid bed roasting of coffee beans.
Proceedings 14th ASIC Colloquium, Paris, Slade, an
L. and Levine alternative
p. 313-318.
H., 1991. Beyond water activity: Recent advances based
approach
to the assesment
of food
on
quality and safety. Critical
Reviews in Food Science and Nutrition. 30 (2-3): 115-360.
Small,
L. E. and Horrell, R. S. 1993. High yield coffee
15th ASIC
Spadone, J. aroma
Colloquium. Paris,
technology. Proceedings
p. 719-726.
C. and Liardon, R. 1989. Analytical study of the evolution of coffee
compounds during storage. Proceedings 13th ASIC Colloquium, Paris,
p. 145-157.
Stahl,
H. D. and Parliment, T. H. 1993. Generation of
furfuryl mercaptan
in
cysteine-ribose model systems in relation to roasted coffee aroma. Proceedings 15th ASIC Colloquium. Paris, p. 607-615.
Stefanucci, A.
and
Protomastro,
M.
G., 1982.
Low density coffee process. United
States Patent, US 4.349.573.
Steinhart,
H. and Holscher, W. 1991.
volatiles in whole coffee beans.
Storage-related changes of low-boiling Proceedings 14th ASIC Colloquium. Paris,
p. 156-164.
Streuli,
H. 1973. Der
Colloquium, Paris,
heutige Stand der Kaffeechemie. Proceedings p. 61-72.
6th ASIC
173
References
Swiss Food Manual, 1973. Eidgenössische Drucksachen- und Materialzentrale, Bern. Volume 2, chapter 35.
and Arneth, W. 1968a. Polysaccharide der grünen Bohnen
Thaler, H.
arabica. Zeitschrift für
Lebensmittel-Untersuchung
und
von Coffea
138: 26-
-Forschung,
35.
Thaler, H.
and Arneth, W. 1968b.
Kaffees. Zeitschrift für
Polysaccharide Lebensmittel-Untersuchung
des und
gerösteten Arabica-
-Forschung,
138 (3):
137-145.
Thaler,
IL and Arneth, W. 1969. Verhalten der
Polysaccharid-Komplexe
rohen Arabica-Kaffees beim Rösten. Zeitschrift
des
für Lebensmittel-Untersu¬
chung und-Forschung, 140: 101-109. Thaler,
H. 1975. Makromolekulare Strukturen in Kaffee.
Colloquium, Paris, Thorn, J. 1995.
The
Proceedings
7th ASIC
p. 175-187.
coffee companion. The connoisseur's guide
to
the world's best
brews. The apple press, London.
Tressl, R., Grünewald,
K.
G., Kamperschröer, H. and Silwar, R. 1979.
Verhalten
einiger schwererflüchtiger Aromastoffe während der Alterung von Röstkaffee. Chemie Mikrobiologie Technologie der Lebensmittel, 6 (2): 52-57. Tressl, R. and Silwar, R. 1981. Investigation roasted coffee. Journal
sulfur-containing components of Agricultural and Food Chemistry, 29: 1078-1082. on
in
Van den Dool, H. and Kratz, P. D. 1963. A
generalisation of the retention index system including linear temperature programmed gas-liquid partition chroma¬ tography. Journal of Chromatography, 11: 463-471.
Viani,
R. 1993. The
composition of coffee.
In
Caffeine, coffee, and health.
Garattini, S. (Ed.). p. 17-41. Raven Press. New York. Vitzthum, O. G., Werkhoff, P. and Ablanque, E. 1975. Flüchtige Inhaltsstoffe des Rohkaffees. Proceedings 7th ASIC
Colloquium, Paris,
p. 115-123.
Vitzthum, O. G . and Werkhoff, P. 1979. Messbare Aromaveränderungen bei Bohnenkaffee in
Technologie
Vitzthum,
sauerstoffdurchlässiger Verpackung. Chemie Mikrobiologie
der Lebensmittel. 6 (1): 25-30.
O. G., Weisemann, C, Becker, R., and Köhler, H. S. 1990. Identifi¬
cation of
an aroma
key compound
in Robusta coffees. Café Cacao Thé, 34 (1):
27-33.
Wang, T. H., Shanfield, H. and Zlatkis, A. 1983. Analysis of trace volatile organic compounds in coffee by headspace concentration and gas chromatographymass spectrometry. Chromatographia. 17 (8): 411-417.
174
References
Wilson, L. G. and Fry, J. C. 1986. Extensin
-
a
major cell wall glycoprotein. Plant,
Cell and Environment, 9: 239-260.
preliminary investigations of oil biosynthesis in the coffee fruit and its subsequent re-distribution within green and roasted beans. Proceedings 17th ASIC Colloquium, Paris, p. 92-99.
Wilson, A. J., Petracco,
Wrigley,
M. and Illy, E. 1997. Some
G. 1988. Coffee. Longman Scientific & Technical, Harlow, UK.
Ziegleder,
G. 1991. Composition of flavor extracts of
raw
and roasted
cocoas.
Zeitschrift für Lebensmittel-Untersuchung und -Forschung. 192: 521-525.
