Food Chemistry 114 (2009) 1503–1509
Contents lists available at ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Analytical Methods
Influence of toasting of oak chips on red wine maturation from sensory
and gas chromatographic headspace analysis
E. Koussissi a,*, V.G. Dourtoglou a, G. Ageloussis a, Y. Paraskevopoulos a, T. Dourtoglou a,
A. Paterson b, A. Chatzilazarou a
a
b
Department of Oenology and Beverage Technology, Technological Educational Institute (T.E.I.) of Athens, Ag. Spiridonos Street, 122 10 Egaleo, Athens, Greece
Centre for Food Quality, SIPBS, University of Strathclyde, 204 George Street, Glasgow G1 1XW, Scotland, UK
a r t i c l e
i n f o
Article history:
Received 1 February 2007
Received in revised form 22 September
2008
Accepted 3 November 2008
Keywords:
Ageing
Oak chips
Toasting
Sensory analyses
Red wine
a b s t r a c t
Red wine from the Greek grape cultivar Aghiorghitiko, was aged in stainless steel with and without wood
chips toasted to different degrees (heavily, medium, lightly) and in a 225L American standard oak barrel
for 32 days. Headspace concentrations of four wood-derived congeners (furfural, guaiacol, cis and trans
oak lactones) were determined over this period after which the wines were bottled, stored and subjected
to sensory descriptive analyses. Of the 36 attributes for appearance, aroma and oral characters, 14 were
significant in univariate analyses and differentiated wines in a multivariate (principal component analysis) product space explaining 72% variance in two factors. Wines from medium toast wood chips scored
highest for woody, vegetative and smoky aromas and flavours but also for bitter taste and astringent
mouthfeel and after 14 days had the highest headspace concentrations of furfural and cis oak lactone.
Wood-related notes were ranked from heavily and lightly toasted chips, barrel and steel control.
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
Traditionally quality red wines are wood-matured in oak casks
and barrels which evoke changes in wine flavour (Diaz-Plaza,
Reyero, Pardo, Alonso, & Salinas, 2002; Escalona, Birkmyre, Piggott,
& Paterson, 2002; Garde-Cerdan & Ancin-Azpilicueta, 2006; Singleton, 1995). Cask maturations produce distinctive flavours but are
slow processes- taking from a few months to several years (Arapitsas, Antonopoulos, Stefanou, & Dourtoglou, 2003; Perez-Coello,
Gonzalez-Vinas, Garcia-Romero, Cabezudo, & Sanz, 2000). Wood
maturation factors include oak geographic origin and species, forest origin, individual tree and stave variation, and barrel volume
and age: all influence final wine character (Pérez-Prieto, Hera-Orts,
Lopez-Roca, Fernandez-Fernandez, & Gomez-Plaza, 2003a, 2003b;
Sauvageot & Feuillat 1999). Currently wine producers seek to
accelerate wood maturation by extraction of chips maximising surface area (Arapitsas et al., 2003; Gutierrez Afonso, 2002; Gutierrez
Afonso, 2003; Perez-Coello et al., 2000).
Wood maturations proceed by depolymerisation and extraction of lignin and hemicellulose components yielding low molecular weight products including: phenolic aldehydes (vanillin,
syringaldehyde, and coniferaldehyde), phenolic alcohols and ethylphenols (guaiacol, 4-methylguaiacol, 4-ethylguaiacol, 4-propyl* Corresponding author. Tel.: +31 645515032.
E-mail address: [email protected] (E. Koussissi).
0308-8146/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodchem.2008.11.003
guaiacol, and eugenol), oak lactones (trans and cis b-methyl-coctalactone) and furanic compounds (furfural, 5-hydroxymethyl
furfural, 5-methyl furfural) (Diaz-Maroto, Sanchez-Palomo, &
Perez-Coello, 2004; Garde-Cerdan & Ancin-Azpilicueta, 2006; Perez-Coello et al., 2000). Quantitative variations are observed in
wines and wood extracts (Cerdán, Gon~i, & Azpilicueta, 2004;
Chatonnet, 1999; Chatonnet, Cutzach, Pons, & Dubourdieu, 1999;
Pollnitz, Jones, & Sefton, 1999). Oxygen passage through staves
plays a role in these processes. Derived compounds vary in sensory impact (Boidron, Chatonnet, & Pons, 1988) and liquid matrix
effects of less volatile components influence headspace partitioning of aroma-active volatile congeners and thus orthonasal and
retronasal perceptions contributing to the character.
Reports on wood maturation of wines concentrate on compositional changes with few examples, in the literature, on sensory
character (Guchu, Diaz-Maroto, Perez-Coello, Gonzalez-Vinas, &
Cabezudo Ibanez, 2006; Gutierrez Afonso, 2002; Gutierrez Afonso,
2003; Perez-Coello et al., 2000; Pérez-Prieto et al., 2003a, 2003b).
Pérez-Prieto and co-workers (2003a, 2003b) explored effects of
specific cask characteristics on the sensory properties of matured
wines. Fermentations in the presence of oak chips were compared
with cask maturations in another three studies (Gutierrez Afonso,
2002; Gutierrez Afonso, 2003; Perez-Coello et al., 2000), establishing compositional and sensory differences. Guchu and co-workers
(2006) compared sensory characters of Chardonnay wines treated
with American and Hungarian oak chips.
1504
E. Koussissi et al. / Food Chemistry 114 (2009) 1503–1509
In this present study, red wine from an experimental winery
and the indigenous Greek variety Aghiorghitiko, was aged in stainless steel without wood or with oak chips varying in toast level
then compared with that aged of an American oak barrel, where
oxygen is passed through staves. Sensory characters of wines were
established by descriptive profiling and related to headspace
quantitations of four key wood-derived congeners: furfural, guaiacol, and cis and trans oak lactones. The aim was to compare the
maturation effects.
Table 1
Standards used for aroma assessor training and corresponding attributes employed in
the final profiling vocabulary.
Attribute
Reference
Aroma training standard
Floral
1a (Rose)
1b (Violet)
1 mg l 1 aqueous solution of
2-phenyl ethanol
Standarda
Fruity (berries)
2a (Strawberry)
2b (Raspberry)
2c (Blackberries)
Crushed strawberries
Crushed raspberries
Crushed blackberries
Vegetative (fresh)
3a (Bell Pepper)
3b (Grassy)
Standarda
Fresh cut green grass
Vegetative (dried)
4a (Tobacco)
4b (Pipe tobacco)
4c (Tea)
Standarda
Fresh tobacco for pipe
Standarda
Caramelised
5a (Caramel)
5b (Honey)
5c (Chocolate)
Standarda
Honey
Chocolate powder
Nutty
6a (Almond)
6b (Hazelnut)
7a (Oak extract)
2. Materials and methods
2.1. Wines and wood maturations
Wine was vintage 2004 from the Greek variety: ‘Aghiorghitiko’
(Koussissi, Paterson, & Paraskevopoulos, 2008; Koussissi, Paterson,
& Piggott, 2003; Koussissi, Paterson, & Piggott, 2007; Manessis,
2000) from Nemea (Peloponnese). During the crushing stage of
vinification, untoasted American wood powder was added (2kg
per tonne grapes). Oenological parameters of resulting wines were:
ethanol – 11.8% (v/v); total acidity: 4.5 g/l expressed as tartaric
acid, volatile acidity: 0.23 g/l expressed as acetic acid; pH: 3.86;
21.8 and 24.3 ppm of free and bound sulphur dioxide, respectively.
An American standard barrel (ASB; 225 l) of white American oak
(Quercus alba), was purchased from ‘Tonnellerie Du Monde World
Cooperage’ as an extra fine grain. Staves were of 24 years-old
wood, heat bent with a medium toasting. Three wooden staves
(heavy, medium and light toast) were purchased from the same
company, as they were of similar wood and age, and chipped into
1 1 2.5 cm pieces, oven dried at 170 °C for 2 h and immersed in
wine.
Five experimental conditions were created: a 32-day barrel
maturation (ASB) and control in stainless steel, and stainless steel
with chips from lightly, medium and heavily toasted chips. Containers with wood chips were opened and stirred manually for
each of the first 14 days and matured at an average 15 °C. After
32 days wines were bottled and stored for 82 days before descriptive sensory analysis. For ASB samples wine was taken directly
from the barrel for sensory analysis after 113 days in wood.
7b (Vanilla)
Crushed almonds
Crushed hazelnuts
Oak wood chips in 10% ethanol-water
solution
Commercial cooking vanilla powder
Spicy
8a (Pepper)
8b (Cinnamon)
8c (Clove)
8d (Nutmeg)
Black pepper
Cinnamon
Cloves in powder
Nutmeg powder
Green/black olive
Earthy
9 (Black olives)
10 (Mushroom)
Black olive paste
Standarda
Woody
a
Aroma standard from: ‘‘Wine discovery”, L’ Atelier du Vin, Paris.
Wines were assessed simultaneously for all attributes in duplicate (two sessions of five products) using a Latin square presentation design (Macfie, Bratchell, Greenhoff, & Vallis, 1989), with
30 mL aliquots in standard wine glasses (ISO 3591: 1977) in booths
with white lighting. Assessors scoring began with appearance then
proceeded to evaluate aroma, taste, flavour by mouth, mouthfeel
and finally aftertaste by using an 11-point scale from 0 (attribute
absent) to 10 (attribute at maximum intensity), on ballots.
2.3. Headspace quantitation of congeners by Solid-phase micro
extraction GC
2.2. Sensory analyses
Fifteen (15) assessors (eight males and seven females, from 21
to 28 years old), were self-selected oenology students of Technological Educational Institute of Athens. All had completed a class
in wine sensory evaluation and 10 were regular red wine consumers (at least once per week).
Assessor training and vocabulary development (5 days) was as
follows: 2 days with 23 labelled aroma standards derived from previous red wine sensory studies (Aiken & Noble, 1984; Koussissi et
al., 2002; Koussissi et al., 2003) (Table 1); and a further 2 days for
basic tastes: sweet (1, 2, 3, 4, g/l sucrose), bitter (0%, 50% and 100%
tonic water), sour (0, 0.5, 1, 2, 3 g/l tartaric acid) and astringent
mouthfeel (1, 2, 4 g/l oenological tannins). The training procedure
used ranking in stimulus intensity (Koussissi et al., 2003). A final
profiling vocabulary was achieved by giving each assessor all wine
samples and asking for a lexicon of attributes. The consensus was:
5 attributes for appearance: clarity, red, purple, brown, and blue; 14
for aroma: floral, fruity (berries), fruity (other), vegetative (fresh),
vegetative (dried), smoky, caramelised, nutty, coconut, woody, spicy,
green/black olive, earthy and chemical; 3 primary tastes: sweet, bitter
and sour; 11 for flavour by mouth: fruity (berries), fruity (other), caramelized, vegetative, smoky, nutty, coconut, woody, spicy, olive and
earthy; 2 for mouthfeel: astringent and body and aftertaste intensity (irrespective of character).
Headspace solid-phase micro extraction capillary gas chromatography (HS-SPME-CGC) was performed on aliquots (13 ml) of
wine, supplemented with 2 g NaCl, in 25 ml conical Supelco vials,
containing a (20 mm) magnetic stirrer bar. Vials were sealed with
PTFE lined silicone septa in plastic screw caps, and headspaces
were equilibrated for 30 min at 30 °C stirred at 300 rpm, and a
50/30 lm Stableflex Divinylbenzene/Carboxen/Polydimethylsiloxane (DVB/CAR/PDMS) fibre in a Supelco fibre holder was inserted
into the headspace. Volatile congeners were adsorbed for 30 min
then the fibre phase was immediately desorbed at 230 °C for
5 min in a splitless injector (equipped with a SPME liner) of a Hewlett Packard, HP 6890 Series GC system equipped with a FID. Congeners were separated in a DB-WAX coated capillary column (30 m
length 0.251 mm i.d. 0.25 lm film thickness; J&W Scientific,
USA) with injector and detector at 230 and 250 °C, respectively.
A temperature programme was used: 54 °C (4 min); to 195 °C at
6 °C min 1; held for 3 min; to 230 °C at 2 °C min 1, held for
10 min (total run time 58 min). Carrier gas was Nitrogen at
1.5 ml min 1. Chromatographic data were recorded on HP Gas
Chromatography Chemstation Rev.06.03 software. Vials were used
once without introduction of an internal standard and each wine
sample was analysed in duplicate.
Quantification of four compounds (furfural, guaiacol, cis and
trans isomers of b-methyl-c-octalactone) was achieved using
1505
E. Koussissi et al. / Food Chemistry 114 (2009) 1503–1509
calibration curves for each sample over the range 40–500 lg/l. in
10% v/v aqueous ethanol.
Table 3
Summary of one-way ANOVA analyses of appearance and oral attributes data.
Attribute
2.4. Statistical analyses
Sensory data were initially analysed with one-way analysis of
variance (ANOVA: Minitab v 13) to determine significance of attributes in product differentiations, and of individual assessor data
(Koussissi et al., 2003). Data were subsequently averaged across
the panel for: (a) all 15 assessors (b) the residual (10 assessor) panel after data of five non-discriminative assessors were removed.
Principal component analysis (PCA; Unscrambler v 7.6) was effected on averaged data from both initial and residual panels with:
all attributes; attributes with P < 0.05; attributes with P < 0.09. Significance of principal components was examined with one-way
ANOVA and averaging of replicate scores were carried out as previously (Koussissi, Paterson, & Cristovam, 2002; Koussissi etal., 2003,
2007). In all PCA weights, all variables were set to 1/SD and the full
cross-validation option was used.
3. Results and discussion
3.1. Sensory analyses: significance of attributes in product
differentiation
One-way ANOVA of the initial raw sensory data (from all assessors), revealed 10 out of the 36 attributes were significant at
P < 0.05 in product differentiations – clarity, red, purple in appearance, vegetative (dried), woody, and chemical aromas, sweet taste,
woody and earthy flavours and astringent mouthfeel. A further 3
attributes were significant at 0.05 < P < 0.09 – fruity (berries) aroma, and vegetative and smoky flavours (Tables 2 and 3). Averaging
data across the panel gave 12 significant attributes (clarity, purple
and brown from appearance, vegetative (dried), caramelised, woody,
spicy and chemical aromas, sweet and bitter tastes woody and earthy
flavours) and 2 (nutty aroma and vegetative flavour) at
0.05 < P < 0.09. From ANOVA five assessors used only 2 or less attributes significantly in discriminations and data from these assessors were removed from the matrix. The residual (10 member)
panel data were again subjected to ANOVA yielding 11 attributes
significant (P < 0.05) in wine discrimination – clarity, red, purple
from appearance, fruity (berries), vegetative (dried), woody, and
Table 2
Summary of one-way ANOVA analyses of aroma sensory data.
Attribute
Initial
data
Data averaged
across initial panel
Data of
residual
panel
Data averaged
across residual
panel
Floral
Fruity
(berries)
Fruity
(other)
Vegetative
(fresh)
Vegetative
(dried)
Smoky
Caramelised
Nutty
Coconut
Woody
Spicy
Green/black
olive
Earthy
Chemical
0.284
0.066
0.129
0.116
0.178
0.027*
0.025*
0.128
0.199
0.285
0.078
0.171
0.720
0.703
0.697
0.723
0.002**
0.003**
0.001**
0.001**
0.472
0.355
0.681
0.296
0.003**
0.208
0.989
0.116
0.029*
0.056
0.311
0.023*
0.020*
0.968
0.150
0.562
0.509
0.401
0.001**
0.146
0.877
0.047*
0.351
0.060
0.282
0.055
0.093
0.699
0.642
0.000***
0.613
0.021*
0.775
0.000***
0.812
0.029*
*
P < 0.05;
**
P < 0.01;
***
P < 0.001.
Appearance
Clarity
Red
Purple
Brown
Blue
Taste
Sweet
Bitter
Sour
Initial
data
Data averaged
across initial panel
Data of
residual
panel
Data averaged
across residual
panel
0.000***
0.011*
0.000***
0.511
0.878
0.005**
0.105
0.012**
0.000***
0.629
0.000***
0.024*
0.006**
0.306
0.621
0.011*
0.219
0.001**
0.005**
0.331
0.032*
0.285
0.537
0.011*
0.006**
0.329
0.051
0.079
0.656
0.001**
0.003**
0.322
0.116
0.270
0.042*
0.655
0.411
0.187
0.554
0.065
0.129
0.903
0.473
0.022*
0.290
0.528
0.281
0.231
0.044*
0.901
0.561
0.002**
0.553
0.695
0.134
0.008**
0.029*
0.909
0.429
0.026*
0.337
0.275
0.001**
0.029*
0.041*
0.044*
0.504
0.241
0.115
0.013*
0.307
0.283
0.110
0.188
0.251
0.160
0.182
Flavour by mouth
Fruity
0.108
(berries)
Fruity
0.455
(other)
Caramelised 0.798
Vegetative
0.081
Smoky
0.073
Nutty
0.859
Coconut
0.595
Woody
0.028*
Spicy
0.632
Green/black 0.959
olive
Earthy
0.001**
Mouthfeel
Astringent
Body
Aftertaste
Aftertaste
intensity
*
P < 0.05;
**
P < 0.01;
***
P < 0.001.
chemical aromas, smoky, woody and earthy flavours by mouth and
astringent mouthfeel. A further three attributes – aroma: fruity
(other), sweet and bitter tastes – were significant at 0.05 <
P < 0.09. Finally, when data were averaged across the residual
(10) panel 14 attributes were significant at P < 0.05: clarity, purple,
brown from appearance; floral, vegetative (dried), smoky, and chemical aromas; sweet and bitter tastes; fruity (berries), vegetative, smoky, woody and earthy flavours by mouth. A further 3 attributes
showed 0.05 < P < 0.09 nutty, woody and spicy aromas (Tables 2
and 3).
In a previous study (Koussissi et al., 2002) of 41 attributes for
appearance, aroma and oral attributes in single session, only nine
were significant in wine discriminations. Guchu and co-workers
(2006) used 23 flavour attributes to describe sensory character in
Chardonnay wines treated with American and Hungarian oak chips
but did not report the levels of significance in product differentiation. In other studies on sensory profiles of wines treated with oak
(Gutierrez Afonso, 2002; Gutierrez Afonso, 2003; Perez-Coello et
al., 2000; Pérez-Prieto et al., 2003a, 2003b), significance was observed with restricted vocabularies (less than eight terms including aroma and oral attributes), thus a potential (Carlucci &
Monteleone, 2001; Lawless, 1999) discrimination difficulty due
to extended vocabularies was not a problem.
3.2. The product space of sensory data
Principal component analysis of data from all assessors and all
36 attributes, followed by ANOVA, firstly gave a highly significant
component (PC1, P = 0.001) and a second component (PC2) with
P = 0.055 explaining 56% variance. Scores in the product space
(Fig. 1) showed good replication and clear differentiation of the five
1506
E. Koussissi et al. / Food Chemistry 114 (2009) 1503–1509
experimental wines. In PC1 (43% of variance), wines are clearly differentiated into two groups: (a) medium and heavily toasted chips
and (b) barrel and unoaked (stainless steel). Wines from appreciably (medium, high) toasted chips were clustered with high scores
on PC1, related to earthy, nutty, vegetative (fresh), vegetative (dried),
chemical, smoky, spicy, and woody aromas; earthy, vegetative, spicy,
smoky and woody flavours; bitter taste; high astringency and body
as mouthfeel; brown hues and strong aftertaste. Wines aged in barrels and unoaked in steel were also clustered with negative PC1
scores and related to: clarity and red appearance, floral, fruity (berries), fruity (other), and caramelised aromas; fruity (berries), fruity
(other), caramelised and green/black olive flavours; and sweet taste.
Wine from lightly toasted chips was scored between the two clusters on PC1 (Fig. 1).
In PC2 wine from lightly toasted chips was clearly differentiated
from all other wines, with negative scores and a close relationship
to coconut aroma, sweet taste, and blue and purple hues in appearance. Stainless steel and barrel conditions were also resolved on
PC2 with the former having a higher correlation with loadings
for red hues and clarity and also for green/black olive flavour, the
latter being more related to fruity aromas and flavours. Finally,
wines from toasted chips were also differentiated in PC2 with medium toasting yielding higher correlations with: brown colour, spicy,
woody and smoky aromas and flavours, vegetative (dried) aroma,
body in mouthfeel and higher scores in aftertaste. In contrast, wines
of heavy toasted chips were related to: earthy, nutty and vegetative
(fresh) in aromas, earthy and vegetative in flavour and bitter in taste
(Fig. 1).
PC2
Earthy (a)
CO
2
ClarityRed
Sour
Green/black olive (m)
Floral
Caramelised
(m)
Fruity (berries)
(m)
Green/black olive (a)
Nutty (m)
Coconut (m)
Fruity other (m)
Fruity (berries) (a)
Fruity (other) (a)
Caramelised (a)
BA
0
Nutty (a)
Vegetative (fresh) (a)
Earthy (m)
HT
Vegetative
(m)
Bitter
Chemical (a)
Astringent
Smoky (a)
Brown
Vegetative (dried)
(a) M T
Woody (m)
Aftertaste
Spicy (a) Smoky (m)
Spicy (m) Woody (a)
Body
Coconut (a)
-2
Sweet
Blue
Purple
LT
PC1
-4
-5
(PC_01, 43%, PC_02, 13%)
0
5
Fig. 1. Principal component analyses (PC1 versus PC2) for sensory data on five different maturation conditions: CO: control; BA: barrel maturation; LT: light toasted chips;
MT: medium toasted chips; HT: heavy toasted chips. Data from all assessors and attributes.
PC2
3
LT
Purple
Sweet
2
Spicy (a)
Vegetative (dried) (a)
Smoky (m)
Woody (a)
Brown
Woody (m)
AstringentChemical
(a)
MT
Vegetative (m) Smoky (a)
Bitter
1
Caramelised (a)
Fruity (berries) (m)
Fruity (berries) (a)
Fruity (other) (a)
0
Floral (a)
-1
BA
-6
-4
(PC_01: 59%,PC_02: 13%)
Clarity
Red
-2
Earthy (m)
CO
HT
Nutty (a)
0
2
PC1
4
Fig. 2. Principal component analyses (PC1 versus PC2) for sensory data on five different maturation conditions: CO: control; BA: barrel maturation; LT: light toasted chips;
MT: medium toasted chips; HT: heavy toasted chips. Data from residual panel, and significant in differentiation attributes.
1507
E. Koussissi et al. / Food Chemistry 114 (2009) 1503–1509
In order to enhance the information obtained from the product
space (improve principal component significance and variance explained), the PCA of the data was repeated with the data from the
discriminating 10 assessors, after removal of 14 non-significant
attributes (appearance: blue; aroma: vegetative (fresh), coconut,
green/black olive, earthy; taste: sour; flavour: fruity (other), caramelised, nutty, coconut, spicy, green/black olive; mouthfeel: body, and
overall aftertaste (Tables 2 and 3). Improvements were observed
in the significance of the first two PCs explaining the 72% variance
(PC1: 59%, P < 0.001; PC2: 13%, P = 0.005) (Fig. 2). In previous reports of sensory descriptive profiling of wines treated with oak,
data was reported from ANOVA as mean ratings or average attribute scores (Guchu et al., 2006; Gutierrez Afonso, 2002; Gutierrez
Afonso, 2003; Perez-Coello et al., 2000; Pérez-Prieto et al., 2003a,
2003b). In this study PCA was employed with many (36) attributes
and variance described above compared well with those of other
wine sensory studies (Koussissi et al., 2002, 2003; Lee & Noble,
2003). Lee and Noble (2003) in profiling Chardonnay wines
a
achieved a two-component PCA space, explaining an 80.5% variance using only eight aroma attributes. As noise increases in parallel with the number of variables ( Wold, 1989), a 72% variance
shown with 22 attributes was considered optimal.
In this final product space replicates were scored similarly (data
not shown) and the experimental wines were differentiated as follows: In PC1 (59% variance explained) the main separation was between wines treated with medium and heavily toasted chips and
the ones from the stainless steel and the barrel conditions. Wines
from medium and heavily toasted wood chips were clustered with
high scores on PC1 and related to: brown colour, woody, smoky and
vegetative aromas and flavours in the mouth, but also spicy, nutty
and chemical aromas, earthy flavour in the mouth, bitter taste and
astringent mouthfeel. On the contrary, stainless steel and barrel
aged products were negatively scored on PC1, linked to loadings
for clarity and red appearance, floral, fruity (berries), fruity (other),
caramelised aromas and fruity (berries) flavour. Wines treated with
lightly toasted chips were in the middle of the two clusters on PC1,
8000
7000
Furfural C (µg / L)
6000
5000
4000
3000
2000
1000
0
0
1
2
3
4
5
6
7
8
9 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 29 30 31 32
Days
Control
b
Barrel
Light Toasting
Medium Toasting
Heavy Toasting
1400
Cis oak lactone C (µg / L)
1200
1000
800
600
400
200
0
0
1
2
3
4
5
6
7
8
9 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 29 30 31 32
Days
Control
Barrel
Light Toasting
Medium Toasting
Heavy Toasting
Fig. 3. Evolution of concentration of furfural and cis oak lactone (lg/L) in wines during the first 32 days: Five different maturation conditions: (a) control, (b) barrel, (c) light
toasted wood chips, (d) medium toasted wood chips, (e) heavy toasted wood chips.
1508
E. Koussissi et al. / Food Chemistry 114 (2009) 1503–1509
but closer to the control and barrel wines (Fig. 2). In PC2, wines
from the light toasted chips were clearly separated from all the
rest, linked to loadings for purple colour and sweet taste. In that
component a separation between the wines from the medium
and the heavily toasted chips conditions was achieved: with the later being higher in nutty aromas and earthy flavours in the mouth
and the former being more smoky, woody, vegetative (dried) and
spicy (Fig. 2).
The control (stainless steel) wines, were linked to attributes
typical of the variety Aghiorghitiko (Koussissi et al., 2002, 2003,
2007, 2008), despite addition of wood powder during the initial
step of vinification. Barrel wine maturation can take up to several
years (Garde-Cerdan & Ancin-Azpilicueta, 2006; Singleton, 1995).
In a study of wine volatile compounds during ageing in oak, differences in extracting capacity were found primarily after four
months in the barrel (Ortega-Heras, Gonzalez-Huerta, Herrera, &
Gonzalez-Sanjose, 2004). In this study descriptive profiling was effected before this point thus similarity of stainless steel and barrel
aged wines could be explained by limited contact time between
wine and barrel. Contact of wine to medium toasted wood chips
led to an increase correlations with brown colour, spicy, vegetative
(dry), woody, smoky and chemical aromas, vegetative, woody and
smoky flavours, bitter taste and astringent mouthfeel, but when
toasting of chips increased further, correlation with those loadings
was slightly reduced and linkage was enhanced to nutty aroma and
earthy flavour. This is in agreement with Perez-Coello et al. (2000)
and Guchu et al. (2006). Thermal treatment of oak used in wine
maturations has proven a key factor influencing the quantity of
volatile phenols and furanic compounds transferred to wines
(Chatonnet, 1998; Chatonnet, 1999; Chatonnet et al., 1999) responsible for smoky, spicy and toasty aroma notes. Chatonnet (1998) reported higher toasty and vanilla aromas on contact with medium
toast. The above is in accordance with the findings of this study,
in which the medium toasted chip wines scored highest for bitter
taste and astringent mouthfeel, possibly due to increased ellagitannins extracted at this toast intensity.
3.3. Extraction of oak-derived congeners
Furfural, guaiacol, cis and trans oak lactones in wine headspaces,
were quantified by GC in each of the initial 14 days and days 21, 29
and 32 (day before removal of the oak chips) and on the initial day
of sensory profiling (day 114). Headspace concentration of each increased to a maximum at day 14 as in other studies (Arapitsas et
al., 2003) and showed a decrease in concentration on days 21 and
29, then maintained until 32nd day (Fig. 3). As reported by Chatonnet et al. (1999), furfural and cis oak lactone were at highest headspace concentration with medium followed by heavily then lightly
toasted chips and lowest with the barrel aged wine (Fig. 3). The
highest concentration of those compounds in the medium toasted
condition could be possibly related to the high scores of this condition in smoky and woody attributes. Light toasting was related to
high score for coconut aroma (Fig. 1) and showed highest initial
extraction rate for the trans isomer of oak lactone (data not shown).
A coconut note in wines has previously been correlated with cis and
trans oak lactone contents (Gutierrez Afonso, 2002). Furfural and
cis oak lactone in wine aged in stainless steel can be related to
the vinification addition of wood powder with residual extraction
from particulate matter. Headspace concentration of furfural increased until day 14 in the control (steel-aged) wine (Fig. 3).
4. Conclusions
In this study, the effect of different technologies/conditions on
red wine maturation was investigated using the tools of sensory
analyses coupled by quantitative measurement of certain headspace volatiles known to originate from the contact of wine with
wood. Sensory characters of the resulting wines were well distinguished and specified for all examined conditions. Of those, the
ones of medium and heavily toasted chips had the most significant
sensory effect, making wines more woody, vegetative and smoky
but also more bitter and astringent. Wines aged with medium toast
chips also showed highest headspace concentrations of furfural
and cis oak lactone followed by those aged with heavily and lightly
toasted chips, in the barrel and in stainless steel (ranked). Wines
from heavily toasted chips scored highest in nutty aromas and
earthy flavours and those of lightly toasted highest in sweet taste.
These results indicate that the use of oak chips is not only an
effective alternative to barrel use in red wine maturation, but also
–through toasting control- a tool to affect specific sensory characters of the resulting wines.
Acknowledgement
The Project is co-funded by the European Social Fund and National Resources – (EPEAEK II) ARHIMIDES.
References
Aiken, J. W., & Noble, A. C. (1984). Composition and sensory properties of Cabernet
Sauvignon wine aged in French versus American oak barrels. Vitis, 23, 27–36.
Arapitsas, P., Antonopoulos, A., Stefanou, E., & Dourtoglou, V. G. (2003). Artificial
aging of wines using oak chips. Food Chemistry, 86, 563–570.
Boidron, J. N., Chatonnet, P., & Pons, M. (1988). Effects of wood on aroma
compounds of wine. Connaissance de la Vigne et du Vin, 22, 275–294.
Carlucci, A., & Monteleone, E. (2001). Statistical validation of sensory data: A study
on wine. Journal of the Science of Food and Agriculture, 81, 751–758.
Cerdán, T. G., Gon~i, D. T., & Azpilicueta, C. A. (2004). Accumulation of volatile
compounds during ageing of two red wines with different composition. Journal
of Food Engineering, 65, 349–356.
Chatonnet, P. (1998). Volatile and odoriferous compounds in barrel-aged wines:
Impact of cooperage techniques and aging conditions. In A. L. Waterhouse & S. E.
Ebeler (Eds.), Chemistry of wine flavor (pp. 180–207). Washington, DC: American
Chemical Society.
Chatonnet, P. (1999). Discrimination and control of toasting intensity and quality of
oak wood barrels. American Journal of Enology and Viticulture, 50, 479–494.
Chatonnet, P., Cutzach, I., Pons, M., & Dubourdieu, D. (1999). Monitoring toasting
intensity of barrels by chromatographic analysis of volatile compounds from
toasted oak wood. Journal of Agricultural and Food Chemistry, 47, 4310–4318.
Diaz-Maroto, M. C., Sanchez-Palomo, E., & Perez-Coello, M. S. (2004). Fast screening
method for volatile compounds of oak wood used for aging wines by headspace
SPME–GC–MS (SIM). Journal of Agricultural and Food Chemistry, 52, 6857–6861.
Diaz-Plaza, E. M., Reyero, J. R., Pardo, F., Alonso, G. L., & Salinas, M. R. (2002).
Influence of oak wood on the aromatic composition and quality of wines with
different tannin contents. Journal of Agricultural and Food Chemistry, 50,
2622–2626.
Escalona, H., Birkmyre, L., Piggott, J. R., & Paterson, A. (2002). Effect of maturation in
small oak casks on the volatility of red wine aroma compounds. Analytica
Chimica Acta, 458, 45–54.
Garde-Cerdan, T., & Ancin-Azpilicueta, C. (2006). Review of quality factors on wine
ageing in oak barrels. Trends in Food Science and Technology, 17, 438–447.
Guchu, E., Diaz-Maroto, M. C., Perez-Coello, M. S., Gonzalez-Vinas, M. A., & Cabezudo
Ibanez, M. D. (2006). Volatile composition and sensory characteristics of
Chardonnay wines treated with American and Hungarian oak chips. Food
Chemistry, 99, 350–359.
Gutierrez Afonso, V. L. (2002). Sensory descriptive analysis between white wines
fermented with oak chips and in barrels. Journal of Food Science, 67(6),
2415–2419.
Gutierrez Afonso, V. L. (2003). Sensory descriptive analysis of red wines undergoing
malolactic fermentation with oak chips. Journal of Food Science, 68(3),
1075–1079.
Koussissi, E., Paterson, A., & Cristovam, E. (2002). Sensory discrimination of dry red
wines from Greece. Journal of Wine Research, 13, 165–179.
Koussissi, E., Paterson, A., & Piggott, J. R. (2003). Sensory flavour discrimination of
Greek dry red wines. Journal of the Science of Food and Agriculture, 83, 797–808.
Koussissi, E., Paterson, A., & Piggott, J. R. (2007). Sensory profiling of aroma in Greek
dry red wines using rank-rating and monadic scoring related to headspace
composition. European Food Research and Technology, 225(5–6), 749–756.
Koussissi, E., Paterson, A., & Paraskevopoulos, Y. (2008). Factors influencing sensory
quality in red wines of the variety Aghiorghitiko (Vitis vinifera L.) from Nemea.
European Food Research and Technology, 226(4), 745–753.
Lawless, H. T. (1999). Descriptive analysis of complex odors: Reality, model or
illusion? Food Quality and Preference, 10, 325–332.
E. Koussissi et al. / Food Chemistry 114 (2009) 1503–1509
Lee, S.-J., & Noble, A. C. (2003). Characterization of odor-active compounds in
Californian Chardonnay wines using GC–olfactometry and GC–mass
spectrometry. Journal of Agricultural and Food Chemistry, 51, 8036–
8044.
Macfie, H. J., Bratchell, N., Greenhoff, K., & Vallis, V. (1989). Designs to balance the
effect of order of presentation and first-order carry-over effects in hall tests.
Journal of Sensory Studies, 4, 128–149.
Manessis, N. (2000). The illustrated Greek wine book. Corfu, Greece: Olive Press.
Ortega-Heras, M., Gonzalez-Huerta, C., Herrera, P., & Gonzalez-Sanjose, M. L. (2004).
Changes in wine volatile compounds of varietal wines during ageing in wood
barrels. Analytica Chimica Acta, 513, 341–350.
Pérez-Prieto, L. J., Hera-Orts, M. L., Lopez-Roca, J. M., Fernandez-Fernandez, J. I., &
Gomez-Plaza, E. (2003a). Oak-matured wines: Influence of the characteristics of
the barrel on wine colour and sensory characteristics. Journal of the Science of
Food and Agriculture, 83, 1445–1450.
Pérez-Prieto, L. J., López-Roca, J. M., Martínez-Cutillas, A., Pardo-Mínguez, F., &
Gómez-Plaza, E. (2003b). Extraction and formation dynamic of oak – related
1509
volatile compounds from different volume barrels to wine and their behavior
during bottle storage. Journal of Agricultural and Food Chemistry, 51, 5444–5449.
Perez-Coello, M. S., Gonzalez-Vinas, A., Garcia-Romero, E., Cabezudo, M. D., & Sanz, J.
(2000). Chemical and sensory changes in white wines fermented in the
presence of oak chips. International Journal of Food Science and Technology, 35,
23–32.
Pollnitz, A. P., Jones, G. P., & Sefton, M. A. (1999). Determination of oak lactones in
barrel-aged wines and in oak extracts by stable isotope dilution analyses.
Journal of Chromatography A, 857(1–2), 239–246.
Sauvageot, F., & Feuillat, F. (1999). The influence of Oak Wood (Quercus robur L., Q.
petraea Liebl.) on the Flavor of Burgundy Pinot Noir. An examination of
variation among individual trees. American Journal of Enology and Viticulture,
50(4), 447–455.
Singleton, V. L. (1995). Maturation of wines and spirits: Comparisons, facts, and
hypotheses. American Journal of Enology and Viticulture, 46(1), 98–115.
Wold, S. (1989). Multivariate data analysis: Converting chemical data tables to
plots. Intelligent Instruments and Computers, 9(10), 197–216.