A study on chemical constituents and sugars extraction from spent coffee grounds

Article abstract of DOI:10.1016/j.carbpol.2010.07.063

Spent coffee grounds (SCG), the residual materials obtained during the processing of raw coffee powder to prepare instant coffee, are the main coffee industry residues. In the present work, this material was chemically characterized and subsequently submitted to a dilute acid hydrolysis aiming to recover the hemicellulose sugars. Reactions were performed according to experimental designs to verify the effects of the variables H₂SO ₄ concentration, liquid-to-solid ratio, temperature, and reaction time, on the efficiency of hydrolysis. SCG was found to be rich in sugars (45.3%, w/w), among of which hemicellulose (constituted by mannose, galactose, and arabinose) and cellulose (glucose homopolymer) correspond to 36.7% (w/w) and 8.6% (w/w), respectively. Optimal conditions for hemicellulose sugars extraction consisted in using 100 mg acid/g dry matter, 10 g liquid/g solid, at 163 °C for 45 min. Under these conditions, hydrolysis efficiencies of 100%, 77.4%, and 89.5% may be achieved for galactan, mannan, and arabinan, respectively, corresponding to a hemicellulose hydrolysis efficiency of 87.4%.

Full text of DOI:10.1016/j.carbpol.2010.07.063

Carbohydrate Polymers 83 (2011) 368–374
Contents lists available at ScienceDirect
Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol
A study on chemical constituents and sugars extraction from spent
coffee grounds
a,∗
b
b
b
a
Solange I. Mussatto , Livia M. Carneiro , João P.A. Silva , Inês C. Roberto , José A. Teixeira
a
IBB - Institute for Biotechnology and Bioengineering, Centre of Biological Engineering, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal
Department of Biotechnology, Engineering College of Lorena, University of São Paulo, Estrada Municipal do Campinho s/n, 12602-810 Lorena/SP, Brazil
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 12 July 2010
Received in revised form 26 July 2010
Accepted 27 July 2010
Available online 6 August 2010
Spent coffee grounds (SCG), the residual materials obtained during the processing of raw coffee powder
to prepare instant coffee, are the main coffee industry residues. In the present work, this material was
chemically characterized and subsequently submitted to a dilute acid hydrolysis aiming to recover the
hemicellulose sugars. Reactions were performed according to experimental designs to verify the effects of
the variables H2SO4 concentration, liquid-to-solid ratio, temperature, and reaction time, on the efficiency
of hydrolysis. SCG was found to be rich in sugars (45.3%, w/w), among of which hemicellulose (constituted
by mannose, galactose, and arabinose) and cellulose (glucose homopolymer) correspond to 36.7% (w/w)
Keywords:
Spent coffee grounds
Chemical composition
Hemicellulose
and 8.6% (w/w), respectively. Optimal conditions for hemicellulose sugars extraction consisted in using
◦
1
00 mg acid/g dry matter, 10 g liquid/g solid, at 163 C for 45 min. Under these conditions, hydrolysis
Dilute acid hydrolysis
Experimental design
efficiencies of 100%, 77.4%, and 89.5% may be achieved for galactan, mannan, and arabinan, respectively,
corresponding to a hemicellulose hydrolysis efficiency of 87.4%.
©
2010 Elsevier Ltd. All rights reserved.
1
. Introduction
tion problems due to the toxic nature (presence of caffeine, tannins,
and polyphenols) (Leifa, Pandey, & Soccol, 2000). Nowadays, there
is great political and social pressure to reduce the pollution aris-
ing from industrial activities. In this sense, conversion of SCG to
value-added compounds is of environmental and economical inter-
est.
Coffee is one of the world’s most widely consumed bever-
ages, and spent coffee grounds (SCG), the solid residues obtained
from the treatment of coffee powder with hot water to prepare
instant coffee, are the main coffee industry residues with a world-
wide annual generation of 6 million tons (Tokimoto, Kawasaki,
Nakamura, Akutagawa, & Tanada, 2005). Considering this huge
amount of coffee residue produced all over the world, the reuti-
lization of this material is a relevant subject. Some attempts for
reutilization of SCG have been made, using it as fuel in industrial
boilers of the same industry due to its high calorific power of about
Hemicelluloses, the second most common polysaccharides in
nature, are heterogeneous polymers of pentoses (xylose and arabi-
nose), hexoses (mannose, galactose, glucose), and sugar acids. In
recent years, bioconversion of hemicellulose has received much
attention because of its practical applications in various industrial
processes, such as for the production of fuels and chemicals (Saha,
2003). Hemicelluloses are usually found in the nature in association
with other polymeric fractions, namely the cellulose and lignin. To
be efficiently used in bioconversion processes, the hemicellulose
polysaccharide needs to be separated from these other structures.
Different processes may be used for this purpose, among of which,
dilute acid hydrolysis stands out as one of the most efficient
to selectively release hemicellulose sugars (Mussatto & Roberto,
2004). The major problem of acid hydrolysis is that the decomposi-
tion of monomeric sugars produced during the reaction takes place
simultaneously with the hydrolysis of polysaccharides. To prevent
sugars decomposition, it is very important to conduct the pro-
cess under adequate reaction conditions. The experimental design
statistical methodology is a useful tool to define such conditions
performing a minimal number of experiments. This methodology
has been employed in several works to maximize the sugars recov-
ery from agro-industrial residues through the establishment of the
5
000 kcal/kg (Silva, Nebra, Silva, & Sanchez, 1998), as an antiox-
idant material source (Yen, Wang, Chang, & Duh, 2005), or as a
source of polysaccharide with immunostimulatory activity (Simões
et al., 2009). Kondamudi, Mohapatra, and Misra (2008) demon-
strated that SCG can be used for the production of biodiesel and fuel
pellets. SCG was also considered an inexpensive and easily available
adsorbent for the removal of cationic dyes in wastewater treat-
ments (Franca, Oliveira, & Ferreira, 2009). However, none of these
strategies have yet been routinely implemented, and most of these
residues remain unutilized, being discharged to the environment
where they cause severe contamination and environmental pollu-
∗ Corresponding author. Tel.: +351 253 604 424; fax: +351 253 678 986.
E-mail addresses: solange@deb.uminho.pt, solangemussatto@hotmail.com
(
S.I. Mussatto).
0
144-8617/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carbpol.2010.07.063

S.I. Mussatto et al. / Carbohydrate Polymers 83 (2011) 368–374
369
Table 1
4
2
experimental design with the real and coded values of the variables used for dilute acid hydrolysis of SCG hemicellulose, and responses obtained to each experimental
condition.
Assay
Variables
Responses
Real values (coded values)a
Sugars (g/l)b
Efficiencies (%)c
X1
X2
X3
X4
Cel
Glu
Arab
Gal
Man
ꢀGal
ꢀMan
ꢀArab
ꢀHemi
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
10 (−1)
14 (+1)
10 (−1)
14 (+1)
10 (−1)
14 (+1)
10 (−1)
14 (+1)
10 (−1)
14 (+1)
10 (−1)
14 (+1)
10 (−1)
14 (+1)
10 (−1)
14 (+1)
12 (0)
100 (−1)
100 (−1)
140 (+1)
140 (+1)
100 (−1)
100 (−1)
140 (+1)
140 (+1)
100 (−1)
100 (−1)
140 (+1)
140 (+1)
100 (−1)
100 (−1)
140 (+1)
140 (+1)
120 (0)
15 (−1)
15 (−1)
15 (−1)
15 (−1)
45 (+1)
45 (+1)
45 (+1)
45 (+1)
15 (−1)
15 (−1)
15 (−1)
15 (−1)
45 (+1)
45 (+1)
45 (+1)
45 (+1)
30 (0)
100 (−1)
100 (−1)
100 (−1)
100 (−1)
100 (−1)
100 (−1)
100 (−1)
100 (−1)
140 (+1)
140 (+1)
140 (+1)
140 (+1)
140 (+1)
140 (+1)
140 (+1)
140 (+1)
120 (0)
0
0.03
0
0.05
0.01
0
0.05
0
0.14
0
0.10
0.02
0
0.21
0
0.11
0.10
0.09
0.08
0
0
0
0
0.06
0
0.11
0
0.27
0
0.09
0
0.28
1.08
0.25
0.94
0.13
0.09
0.08
0.08
0.06
0.13
0.05
1.32
1.03
1.85
1.04
1.96
0.87
1.91
1.11
2.12
1.79
2.54
2.06
1.79
1.67
1.64
0
0
0
0
0
0
0
0
0
0
0
0
1.2
0.8
2.1
0.8
5.3
1.2
4.0
0
0
0
0
4.8
4.8
7.8
4.2
79.8
87.4
100.0
88.2
100.0
73.8
100.0
94.2
100.0
100.0
100.0
100.0
100.0
100.0
100.0
0.2
0.2
0.4
0.2
4.6
5.9
6.0
4.9
7.7
4.5
7.0
5.8
74.9
44.8
83.9
61.6
7.6
0.15
0.11
0.34
0.13
0.09
0.30
0.11
0.24
0.13
11.24
4.68
14.31
4.74
0.21
0.19
0.18
0.6
2.4
0.7
0.6
1.5
0.8
1.2
0.9
56.6
33.1
72.1
33.5
1.3
1.2
1.1
0.07
0.27
0.07
0.68
0.11
0.51
0.18
14.26
5.12
13.64
9.45
0.59
0.43
0.43
1
1
1
1
1
1
1
1
1
1
2.0
100.0
55.8
100.0
100.0
5.5
12 (0)
12 (0)
120 (0)
120 (0)
30 (0)
30 (0)
120 (0)
120 (0)
4.0
4.0
7.0
6.9
a
b
c
◦
X1: liquid-to-solid ratio (g/g); X2: acid concentration (mg/g); X3: reaction time (min); X4: temperature ( C).
Cel: cellobiose; Glu: glucose; Arab: arabinose; Gal: galactose; Man: mannose.
ꢀ: efficiency of hydrolysis; Hemi: hemicellulose.
best hydrolysis operational conditions (Mussatto & Roberto, 2005;
Neureiter et al., 2004; Roberto, Mussatto, & Rodrigues, 2003).
In view of the aforementioned, the present work evalu-
ated the sugars extraction from SCG hemicellulose, as a first
step to explore the use of this material in fermentative pro-
cesses. Initially, the material was chemically characterized, and
2.2. Dilute acid hydrolysis
Dried SCG was submitted to hydrolysis reactions under differ-
ent conditions of H SO concentration (100–140 mg/g dry matter),
2
4
◦
liquid-to-solid ratio (10–14 g/g), temperature (100–180 C), and
reaction time (15–75 min), which were combined as proposed in
experimental designs. For the experiments, SCG and the required
amount of acid solution were placed in 200-ml stainless steel batch
cylindrical reactors. Under the desired temperature, the dully cov-
ered reactors were introduced into a silicone oil bath where they
were maintained during the necessary time. At the end of each
reaction, the reactors were immediately cooled in ice bath, and
the resulting solid material was separated by filtration. The filtrates
its hemicellulose content was determined. In
a subsequent
stage, extraction reactions were performed by using dilute acid
under different operational conditions (liquid-to-solid ratio, acid
concentration, reaction time and temperature), which were pro-
posed according to experimental designs. The condition able to
maximize the extraction results was established by statistical
analysis.
(
hemicellulosic hydrolysates) were analyzed for sugars (cellobiose,
2
. Material and methods
glucose, arabinose, mannose, galactose, and xylose), degradation
products (furfural, hydroxymethylfurfural, and total phenols), and
acetic acid determination.
2.1. Raw material characterization
Spent coffee grounds (SCG) were supplied by NovaDelta-
2
.3. Experimental designs
Comércio e Indústria de Cafés, S.A. (Campo Maior, Portugal). The
◦
material with around 80% humidity was dried at 60 ± 5 C to 10%
Initially, a 24 full-factorial design with three levels leading to
9 sets of experiments was made to evaluate the effect of four
moisture content, being thus stored until be required for processing
or analysis.
1
variables, namely the liquid-to-solid ratio (X ), acid concentration
For chemical characterization, dried SCG was subjected to a
quantitative acid hydrolysis with 72% (w/w) sulfuric acid. In this
1
(X ), reaction time (X ), and temperature (X ), on the hydrolysis
2 3 4
of SCG hemicellulose. For statistical analysis, the variables were
coded according to Eq. (1), where each independent variable is rep-
resented by x (coded value), X (real value), X (real value at the
method, 2 g of sample were first added to 10 ml 72% H SO and
2
4
◦
maintained at 50 C for 7 min. After this pre-treatment, distilled
water was added to the mixture to dilute the H SO to 1N, and incu-
i
i
0
2
4
◦
center point), and ꢁX (step change value). The levels of the vari-
bated at 121 C for 45 min. The monosaccharides and acetic acid
contained in hydrolysates were determined by HPLC in order to
estimate the contents of samples in cellulose (as glucan), hemicel-
lulose (as mannan + galactan + arabinan + xylan) and acetyl groups.
Protein content was estimated by the Kjeldahl nitrogen method,
and a factor of 6.25 was used to convert nitrogen into protein. Ashes
were determined by weight difference before and after incinera-
i
ables investigated in this study are given in Table 1. Three assays
in the center point were carried out to estimate the random error
needed for the analysis of variance, as well as to examine the pres-
ence of curvature in the response surfaces. Hydrolysis efficiency
of galactan, mannan, arabinan, and hemicellulose were taken as
responses of the design experiments.
◦
tion of the SCG sample in a muffle furnace at 550 C for 4 h. Before
X − X
i
0
weighing, samples were placed in a desiccator for 50 min. The min-
eral content was determined by Inductively Coupled Plasma Atomic
Emission Spectrometry (ICP-AES). All determinations were carried
out in triplicate.
xi =
(1)
ꢁ
X
i
4
2
Based on the results obtained in the 2 design, a 2 central com-
posite design was proposed to maximize the sugars recovery from

3
70
S.I. Mussatto et al. / Carbohydrate Polymers 83 (2011) 368–374
Table 2
2
2
experimental design with the real and coded values of the variables used for dilute acid hydrolysis of SCG hemicellulose, and responses obtained to each experimental
condition.
Assay
Variables
Responses
Sugars (g/l)b
Cel
Realvalues(coded values)a
Efficiencies (%)c
X3
X4
Glu
Arab
Gal
Man
ꢀGal
ꢀMan
ꢀArab
ꢀHemi
1
2
3
4
5
6
7
8
9
45 (−1)
75 (+1)
45 (−1)
75 (+1)
60 (0)
140 (−1)
140 (−1)
180 (+1)
180 (+1)
160 (0)
160 (0)
160 (0)
160 (0)
0
0
0
0
0
0
0
0
0
0
0.29
0.25
0.96
1.51
0.43
0.39
0.38
0.61
0.22
1.18
1.68
1.93
1.32
0.50
1.67
1.99
1.67
0.82
1.90
0.80
12.56
14.77
12.78
9.00
15.18
14.47
15.91
14.32
14.14
12.19
8.58
94.9
100.0
96.3
41.9
66.2
65.0
31.0
79.3
81.7
76.5
63.6
48.2
46.8
98.2
100.0
77.7
29.4
98.2
100.0
98.2
48.2
100.0
47.1
64.5
80.5
77.3
44.7
88.0
89.5
86.4
76.5
70.1
63.7
13.55
13.30
6.34
16.23
16.73
15.66
13.02
9.87
67.8
100.0
100.0
100.0
100.0
100.0
91.8
60 (0)
45 (−1)
75 (+1)
60 (0)
140 (−1)
1
0
60 (0)
180 (+1)
9.58
a
b
c
◦
X3: reaction time (min); X4: temperature ( C).
Cel: cellobiose; Glu: glucose; Arab: arabinose; Gal: galactose; Man: mannose.
: efficiency of hydrolysis; Hemi: hemicellulose.
ꢀ
SCG hemicellulose. This new design, leading to additional 10 sets
of experiments, was made to evaluate the effect of the previously
selected variables: temperature and reaction time, in the range of
values given in Table 2. Acid concentration and liquid-to-solid ratio
were fixed in these assays in 100 mg/g, and 14 g/g, respectively.
The experimental results were fitted with second-order poly-
nomial equations by multiple regression analyses. The quadratic
280 nm after dilution’s correction, CF and CHMF are the concen-
trations (g/l) of furfural and hydroxymethylfurfural determined
by HPLC, and εF and εHMF are the extinction coefficients (l/g cm)
of furfural (146.85) and HMF (114.00) previously determined by
ultraviolet spectroscopy at 280 nm.
TP = 4.187 × 10−2(ALIG280 − APD280) − 3.279 × 10−4
(3)
(4)
models were expressed according to Eq. (2), where yˆ represents
i
HMF
APD280 = [(C ε ) + (C εHMF )]
F
F
the response variable, b0 is the interception coefficient, b , b and
i
ii
The recovered sugar yield (Y , in gram of substance that can be
bij are the regression coefficients, n is the number of studied vari-
ables, and X and X represent the independent variables. Where
S
achieved from 100 g of SCG dry matter), and the hydrolysis effi-
ciency (ꢀ, %) were calculated using Eqs. (5) and (6), respectively,
where C is the concentration of the component in the liquid phase
i
j
possible, the models were simplified by elimination of statistically
insignificant terms.
(g/l); M is the amount of SCG (dry matter) employed in the experi-
n
n
n−1
n
ꢀ
ꢀ
ꢀꢀ
ment (g); V is the volume of liquid solution employed (l) and Ymax
is the maximum yield in recovered sugars that can be attained (g
per 100 g dry matter).
2
yˆ = b +
b X +
b Xi
ii
+
b X X
ij i j
(2)
i
0
i
i
i=1
i=1
i=1 j=i+1
ꢁ
ꢂ
V
Statistical significance of the regression coefficients was deter-
mined by Student’s t-test, and the proportion of variance explained
YS
=
C ×
× 100
(5)
(6)
M
ꢁ
ꢂ
by the models were given by the multiple coefficient of determi-
YS
2
ꢀ =
× 100
nation, R . Results were analyzed by using the softwares Statistica
Ymax
version 5.0, and Design Expert version 5.0.
3
. Results and discussion
2.4. Analytical methodology
3.1. Chemical composition of SCG
Glucose, arabinose, mannose, galactose, xylose, cellobiose,
acetic acid, furfural and hydroxymethylfurfural concentrations
were determined by high-performance liquid chromatography
Carbohydrates are the most important constituents in coffee
beans (Arya & Rao, 2007), and similarly in SCG. As can be seen
in Table 3, SCG is a residue rich in sugars polymerized into cellu-
lose and hemicellulose structures, which correspond to almost half
of material dry weight (45.3%, w/w). Hemicellulose is composed
by three sugars, mannose being the most abundant, followed by
galactose and few quantities of arabinose (mannan:galactan and
galactan:arabinan ratios of 1.54 and 8.12, respectively). Xylose was
not found in SCG composition. These results are in agreement with
observations made by other authors that polysaccharides in coffee
cell wall are constituted by mannose, galactose, arabinose and glu-
cose sugars, forming mainly galactomannan, arabinogalactan and
cellulose structures (Arya & Rao, 2007; Fischer, Reimann, Trovato,
& Redgwell, 2001; Oosterveld, Harmsen, Voragen, & Schols, 2003;
Redgwell, Curti, Fischer, Nicolas, & Fay, 2002). Sugars composition
in SCG was found to be 46.8% mannose, followed by 30.4% galactose,
19.0% glucose, and 3.8% arabinose, revealing mannans as the major
polysaccharides in this residue. These values are a bit different from
those reported by Simões et al. (2009), who reported the presence
of mannose (57%), followed by galactose (26%), glucose (11%) and
(
HPLC) on a Jasco chromatograph. For sugars determination, a
refractive index detector and a Varian column Metacarb 87P
◦
(
300 mm × 7.8 mm) at 80 C were used. Ultrapure water was used
as eluent at a flow rate of 0.4 ml/min. Acetic acid concentration was
also determined in a refractive index detector, but using a Bio-Rad
◦
Aminex HPX-87H (300 mm × 7.8 mm) column at 60 C, and sulfuric
acid 0.005 M as eluent at a flow rate of 0.7 ml/min. Furfural and
hydroxymethylfurfural were determined with a UV detector (at
2
80 nm) and a Nucleosil 120-5 C18 5 ␮m (4.6 mm × 250 mm) col-
umn at room temperature, using acetonitrile/water (1/8 with 10 g/l
acetic acid) as the eluent at a flow rate of 0.8 ml/min. In all the cases,
a sample volume of 20 ␮l was injected.
For total phenols determination, the pH of the hydrolysate sam-
ples was raised to 12.0 with NaOH 6.0 M and the resulting solution
was diluted with distilled water in order to obtain an absorbance
reading not exceeding 0.5. The phenols concentration was then
calculated according to Eqs. (3) and (4), where TP is the total
phenols concentration (g/l), ALIG280 is the absorbance reading at

S.I. Mussatto et al. / Carbohydrate Polymers 83 (2011) 368–374
371
Table 3
Chemical composition of spent coffee grounds.
Components
Dry weight (g/100 g)
Cellulose (glucan)
Hemicellulose
Arabinan
8.6
36.7
1.7
Galactan
Mannan
Proteins (N × 6.25)
Acetyl groups
Ashes
13.8
21.2
13.6
2.2
1.6
Minerals
(mg/kg)
Fig. 1. Chromatogram profile of sugars solubilized from spent coffee grounds acid
hydrolysis.
Potassium
Phosphorus
Magnesium
Calcium
Aluminum
Iron
Manganese
Copper
Zinc
3549.0
1475.1
1293.3
777.4
279.3
118.7
40.1
32.3
15.1
nd
nd
extracted in all the used reaction conditions, but the concentra-
tion values of them varied to each assay. The chromatogram profile
shown in Fig. 1 reveals glucose, arabinose, mannose and galactose
as the only sugars present in SCG hydrolysate that is in agreement
with the chemical composition previously presented. Cellobiose (a
glucose dimer connected by a glycosidic bond) and glucose were
liberated in low amounts in all the assays, demonstrating that
the used reaction conditions were not suitable for the cellulose
hydrolysis, but only for the hemicellulose hydrolysis, which was the
objective of the present work. In fact, cellulose hydrolysis requires
the use of more severe conditions of acid concentration, tempera-
ture and reaction time than those here used (Mussatto & Roberto,
2004). For the conditions that promoted the highest hemicellulose
hydrolysis efficiencies (>60%, Table 1), galactose and mannose were
the most abundant sugars present in the produced hydrolysates,
that is in accordance to the chemical composition previously deter-
mined, which revealed that galactan and mannan are the main
sugars polymers present in SCG.
As a consequence of the differences in the sugars extraction, dif-
ferent hydrolysis efficiency values were achieved, varying between
0% and 100% for galactan, 0 and 72.1% for mannan, 4.2 and 100% for
arabinan and 0.2 and 83.9% for the overall hydrolysis of hemicellu-
lose (Table 1). Arabinose was the sugar more easily released from
the hemicellulose structure, since it was totally extracted in 10 of
the 19 evaluated hydrolysis conditions. Even in the mildest reac-
tion condition (assay 1), arabinose was detected in the produced
hydrolysate, and no other sugar was found. Galactose was totally
released in 3 of the studied conditions, suggesting that galactan in
SCG has a major susceptibility to hydrolysis than the mannan struc-
ture, which was not totally hydrolysed in any of the performed
assays. In fact, arabinogalactans dissolve better than linear man-
nans, and one of the reasons for this is the association of linear
mannans to form crystalline regions (Bradbury & Atkins, 1997).
In addition, in a polymeric structure, side chains are more easily
released than the main chains. Arabinogalactans in SCG consist of
a main chain of 1 → 3 linked galactose branched at C-6, with side
chains containing arabinose and galactose. Galactomannans con-
sist of a main chain of 1 → 4 linked mannan with galactose unit side
chains linked at C-6, and different degrees of branching (Bradbury
& Halliday, 1990; Navarini et al. (1999); Nunes, Reis, Domingues,
& Coimbra, 2006). Such fact explains the major difficult in extract
mannose than galactose or arabinose from the SCG hemicellulose.
A statistical analysis of the efficiency results was then per-
formed. The Pareto charts shown in Fig. 2 represent the estimated
effects of the variables used during the dilute acid hydrolysis on
the efficiency responses. The length of each bar was proportional
to the standardized effect. Bars extending beyond the vertical line
corresponded to effects statistically significant at 95% confidence
level. This analysis revealed positive and significant effects of tem-
perature and reaction time, as well as the interaction between
these variables, in all the evaluated responses. Such effects suggest
that sugars extraction was higher when the temperature and reac-
Sulfur
Chromium
nd: not detected.
arabinose (6%). In fact, differences in the chemical composition of
SCG may occur according to the variety of beans utilized, and the
roasting and extraction processes that they were submitted.
In addition to polysaccharides, SCG also contains significant pro-
tein content (13.6% w/w) (Table 3). According to Arya and Rao
(
2007), roasted coffee contains on average 3.1% (w/w) protein. Dur-
ing the process with water for the instant coffee prepare, many of
the other grain components are extracted, and as a consequence,
the non-extracted components have their proportion concentrated
in the residual solid material. Therefore, the protein content is
higher in SCG than in the coffee grains. It must be emphasized that
as the protein content in SCG was calculated from the total nitro-
gen content of the samples, it may have been overestimated due
to the presence of other nitrogen-containing substances (caffeine,
trigonelline, free amines and amino acids) (Delgado, Vignoli, Siika-
aho, & Franco, 2008). Even so, the value here found was close to
the value reported by Ravindranath, Yousuf Ali Khan, Oby Reddy,
Thirumala Rao, & Reddy (1972), which was of about 14%, but was
higher than the value reported by Lago, Antoniassi, & Freitas (2001),
which varied between 6.7% and 9.9%.
Protein-bound amino acids were not quantified in the present
work, but some authors have reported that SCG contains the follow-
ing amino acids (range of values, mg/100 g): aspartic acid (10–137),
glutamic acid (618–987), serine (46–85), histidine (7–378), glycine
(
131–567), threonine (18–160), alanine(260–388), arginine(8–13),
tyrosine (152–288), cystine (281–362), valine (324–488), methio-
nine (53–136), phenylalanine (29–477), isoleucine (275–377),
leucine (567–779), lysine (100–164), and proline (169–338) (Lago
et al., 2001).
SCG also contains ashes, which, according to the ICP-AES
analysis, are constituted by several minerals. Potassium is the
most abundant element, followed by phosphorus and magnesium
(
Table 3). Potassium is also the most abundant element in coffee
beans, corresponding to 40% of the oxide ash (Arya & Rao, 2007).
After extraction with hot water for the instant coffee prepare, most
of the mineral elements are easily extracted (Arya & Rao, 2007), but
SCG still contains potassium as predominant.
3
.2. Hemicellulose sugars extraction from SCG
Table 1 shows the experimental results obtained according to
the 24 experimental design. It can be noted that sugars were

3
72
S.I. Mussatto et al. / Carbohydrate Polymers 83 (2011) 368–374
Fig. 2. Pareto chart for the effects of liquid-to-solid ratio (X1), acid concentration (X2), reaction time (X3) and temperature (X4), and their interactions, during the dilute acid
hydrolysis of spent coffee grounds, on the efficiency of hydrolysis of galactan (A), mannan (B), arabinan (C), and hemicellulose (D).
tion time used for dilute acid hydrolysis were increased. The other
two variables, acid concentration and liquid-to-solid ratio, did not
show significant influence on the responses at 95% confidence
level.
cellulose structure, but it was probably released and subsequently
degraded.
In fact, when analyzing the concentration of toxic compounds
in the hydrolysates, it can be noted that elevated values of furfural
were obtained according to the used reaction conditions (Table 4).
Furfural is a by-product generated from the degradation of pentose
sugars, like arabinose (Mussatto & Roberto, 2004). In the first exper-
imental design, the highest furfural concentration obtained was
0.06 g/l, while in the second experimental design, a value of 3.93 g/l
was found in the hydrolysate produced under the highest temper-
Since mannose, the main sugar present in SCG was not fully
extracted under the evaluated conditions; additional hydrolysis
assays were performed aiming to improve the extraction of this
2
sugar. In this step, a 2 full-factorial design was used to determine
the best conditions of temperature and reaction time to be used on
SCG hydrolysis. Acid concentration and liquid-to-solid ratio were
fixed at the lowest levels used in the first design. The experimen-
tal results obtained in these assays are shown in Table 2. When
comparing these results with those achieved in the 24 experimen-
tal design (Table 1) it is evident that the sugars extractions and
the hydrolysis efficiencies results were improved due to the use
of this higher range of values for temperature and reaction time.
The efficiency of hydrolysis of galactan, for example, attained the
maximum value for 6 of the 10 evaluated conditions. Efficiency of
hydrolysis of mannan was increased to 81.7% and resulted in a total
efficiency of hemicellulose hydrolysis of about 90% (assay 6). Oth-
erwise, the efficiency of hydrolysis of arabinan, which had attained
the maximum value in most of the experiments of the first design,
was only maximum in three of the evaluated conditions (Table 2).
Probably, some degradation of the released arabinose occurred in
these experiments, giving thus efficiency results lower than 100%.
Therefore, the arabinose efficiency results lower than 100% does
not mean that this sugar was not fully released from the hemi-
Table 4
Concentration of toxic compounds in the hydrolysates produced from SCG under
2
different operational conditions, according to the 2 full-factorial design.
Assay
Toxic compounds (g/l)
Acetic acid
Furfural
HMFa
Total phenols
1
2
3
4
5
6
7
8
9
0.74
0.79
0.93
1.36
0.83
0.82
0.81
0.90
0.76
1.03
0.06
0.14
2.81
3.93
0.91
0.81
0.69
2.42
0.10
3.56
0.01
0.04
0.37
0.49
0.14
0.13
0.11
0.35
0.02
0.55
4.08
4.68
5.48
1.42
5.24
5.13
3.51
3.45
4.15
1.82
1
0
a
HMF: hydroxymethylfurfural.

S.I. Mussatto et al. / Carbohydrate Polymers 83 (2011) 368–374
373
Fig. 3. Response surfaces described by the models representing the efficiencies of hydrolysis of spent coffee grounds galactan (A), mannan (B), arabinan (C), and hemicellulose
D).
(
ature and reaction time (assay 4, Table 4). As a consequence of this
high arabinose degradation, very low efficiency of hydrolysis of ara-
binan was obtained, corresponding to 29.4% only (assay 4, Table 2).
Besides furfural, other toxic compounds including hydroxymethyl-
furfural (HMF), acetic acid, and phenols were also found in SCG
hydrolysates. Presence of these compounds is common in hemicel-
lulosic hydrolysates produced by dilute acid hydrolysis (Mussatto
efficiency of hydrolysis of galactan (ꢀGal), mannan (ꢀMan), arabinan
(ꢀArab), and hemicellulose (ꢀHemi), as a function of the reaction time
(X ) and temperature (X ).
3
4
2
ꢀGal(%) = 101.76 − 3.90X − 3.51X3 − 6.50X4
3
7.61X2 − 8.40X X4 (R2 = 0.90)
−
(7)
(8)
4
3
&
Roberto, 2004).
The experimental results in Table 2 were used to estimate the
2
ꢀ
Man(%) = 75.28 − 7.53X − 50.85X4
3
main effects of the variables and their interactions on the effi-
ciency of hydrolysis of SCG hemicellulose. The statistical analysis
revealed a negative main effect (p < 0.05) of both, reaction time and
temperature in this response. This means that the hemicellulose
hydrolysis efficiency was increased when the reaction time and
temperature were decreased. However, this behavior was not lin-
ear for the temperature, since the quadratic effect of this variable
was also significant for this response at 95% confidence level. This
means that the lowest temperature is not the most suitable for an
−29.15X X4 (R2 = 0.93)
3
2
ꢀArab(%) = 90.91 − 16.08X − 9.51X3 − 24.00X4
3
−
9.16X2 − 12.53X X4 (R2 = 0.89)
(9)
4
3
2
efficient hemicellulose hydrolysis, but there is a region between
ꢀHemi(%) = 85.10 − 8.83X3 − 9.80X4 − 36.60X4
◦
1
40 and 180 C where this response is maximized. A second-order
−
24.30X X4 (R2 = 0.96)
(10)
3
polynomial is thus the most suitable to describe the variations of
this response as a function of the temperature and reaction time.
Regression analyses were performed to fit the responses with
the experimental data. When it was possible, the variables that
were not significant (even at 90% confidence level) were excluded
from the models. In other cases, the non-significant variables were
kept in the models to minimize the error determination. Eqs.
The obtained models did not show lack-of-fit and presented high
2
determination coefficients (R ≥ 0.89), explaining more than 89% of
the variability in the responses. Three-dimensional response sur-
faces described by the models were plotted (Fig. 3). Note that the
fitted surfaces have a non-linear region where the responses values
are maximized (dark region). Based on the four obtained models,
(
7)–(10), where the variables take their coded values, represent the

3
74
S.I. Mussatto et al. / Carbohydrate Polymers 83 (2011) 368–374
Acknowledgements
NovaDelta-Comércio e Indústria de Cafés, S.A. (Campo Maior,
Portugal).
References
Arya, M., & Rao, L. J. M. (2007). An impression of coffee carbohydrates. Critical Reviews
in Food Science and Nutrition, 47, 51–67.
Bradbury, A. G. W., & Atkins, E. D. T. (1997). Factors affecting mannan solubility in
roast coffee extracts. In Proceedings of ASIC conference, 17th colloquium Nairobi.
Bradbury, A. G. W., & Halliday, D. J. (1990). Chemical structures of green coffee bean
polysaccharides. Journal of Agricultural and Food Chemistry, 38, 389–392.
Delgado, P. A., Vignoli, J. A., Siika-aho, M., & Franco, T. T. (2008). Sediments in coffee
extracts: Composition and control by enzymatic hydrolysis. Food Chemistry, 110,
168–176.
Fischer, M., Reimann, S., Trovato, V., & Redgwell, R. J. (2001). Polysaccharides of green
arabica and robusta coffee beans. Carbohydrate Research, 330, 93–101.
Franca, A. S., Oliveira, L. S., & Ferreira, M. E. (2009). Kinetics and equilibrium studies
of methylene blue adsorption by spent coffee grounds. Desalination, 249, 267–
272.
Kondamudi, N., Mohapatra, S. K., & Misra, M. (2008). Spent coffee grounds as a
versatile source of green energy. Journal of Agricultural and Food Chemistry, 56,
1
1757–11760.
Fig. 4. Optimum region by overlay plots of the four responses (efficiency of hydrol-
ysis of galactan – ꢀgal, mannan – ꢀman, arabinan – ꢀarab, and hemicellulose – ꢀhemi)
as a function of the temperature and reaction time used for dilute acid hydrolysis of
spent coffee grounds.
Lago, R. C. A., Antoniassi, R., & Freitas, S. C. (2001). Composi c¸ ão centesimal e de
aminoacidos de café verde, torrado e de borra de café soluvel. In 2 Simposio de
Pesquisa dos Cafés do Brasil Vitoria, ES. Resumos, (p. 104).
Leifa, F., Pandey, A., & Soccol, C. R. (2000). Solid state cultivation—An efficient method
to use toxic agro-industrial residues. Journal of Basic Microbiology, 40, 187–
1
97.
a graphical optimization was conducted using the ‘Design expert’
program. The method basically consists of overlaying the curves
of all the models according to the criteria imposed. The optimal
working conditions were defined to attain maximum hydrolysis
efficiencies for galactan, mannan, arabinan, and hemicellulose. The
criteria adopted were (a) galactan hydrolysis efficiency greater than
Mussatto, S. I., & Roberto, I. C. (2004). Alternatives for detoxification of diluted
acid lignocellulosic hydrolyzates for use in fermentative processes: A review.
Bioresource Technology, 93, 1–10.
Mussatto, S. I., & Roberto, I. C. (2005). Acid hydrolysis and fermentation of brewer’s
spent grain to produce xylitol. Journal of the Science of Food and Agriculture, 85,
2453–2460.
Navarini, L., Gilli, R., Gombac, V., Abatangelo, A., Bosco, M., & Toffanin, R. (1999).
Polysaccharides from hot water extracts of roasted Coffea arabica beans: Isola-
tion and characterization. Carbohydrate Polymers, 40, 71–81.
9
5%, (b) mannan hydrolysis efficiency superior to 76%, (c) arabinan
hydrolysis efficiency higher than 94%, and (d) hemicellulose hydrol-
ysis efficiency greater than 84%. The overlaying plot attained (Fig. 4)
shows a dark area where all the criteria imposed were satisfied. A
point was thus chosen in this area (marked by the square) where
all the responses were maximum. This point was assigned as opti-
Neureiter, M., Danner, H., Frühauf, S., Kromus, S., Thomasser, C., Braun, R., et al.
(2004). Dilute acid hydrolysis of presscakes from silage and grass to recover
hemicellulose-derived sugars. Bioresource Technology, 92, 21–29.
Nunes, F. M., Reis, A., Domingues, M. R. M., & Coimbra, M. A. (2006). Characterization
of galactomannan derivatives in roasted coffee beverages. Journal of Agricultural
and Food Chemistry, 54, 3428–3439.
Oosterveld, A., Harmsen, J. S., Voragen, A. G. J., & Schols, H. A. (2003). Extraction
and characterization of polysaccharides from and roasted Coffea arabica beans.
Carbohydrate Polymers, 52, 285–296.
Ravindranath, R., Yousuf Ali Khan, R., Oby Reddy, T., Thirumala Rao, S. D., & Reddy, B.
R. (1972). Composition and characteristics of Indian coffee bean, spent ground
and oil. Journal of the Science of Food and Agriculture, 23, 307–310.
Redgwell, R. J, Curti, D., Fischer, M., Nicolas, P., & Fay, L. B. (2002). Coffee bean arabino-
galactans: Acidic polymers covalently linked to protein. Carbohydrate Research,
◦
mum point and corresponded to the use of temperature of 163 C,
and reaction time of 45 min. Under these conditions, the model pre-
dicted hydrolysis efficiencies of 100%, 77.4%, 89.5%, and 87.4% for
galactan, mannan, arabinan, and hemicellulose, respectively.
4
. Conclusions
3
37, 239–253.
Roberto, I. C., Mussatto, S. I., & Rodrigues, R. C. L. B. (2003). Dilute-acid hydrolysis for
optimization of xylose recovery from rice straw in a semi-pilot reactor. Industrial
Crops and Products, 17, 171–176.
Saha, B. C. (2003). Hemicellulose bioconversion. Journal of Industrial Microbiology
and Biotechnology, 30, 279–291.
Silva, M. A., Nebra, S. A., Silva, M. J. M., & Sanchez, C. G. (1998). The use of biomass
residues in the Brazilian soluble coffee industry. Biomass and Bioenergy, 14,
457–467.
Simões, J., Madureira, P., Nunes, F. M., Domingues, M. R., Vilanova, M., & Coimbra, M.
A. (2009). Immunostimulatory properties of coffee mannans. Molecular Nutrition
& Food Research, 53, 1036–1043.
SCG is an agro-industrial residue composed in the major-
ity by carbohydrates, being mannose, galactose, and arabinose
from hemicellulose) and glucose (from cellulose), the main sug-
(
ars present. Optimal conditions for hemicellulose sugars extraction
by dilute acid hydrolysis were established, and consisted in using
◦
1
4
7
00 mg acid/g dry matter, 10 g/g liquid-to-solid ratio, at 163 C for
5 min. Under these conditions, hydrolysis efficiencies of 100%,
7.4%, 89.5%, and 87.4% can be achieved for SCG galactan, man-
nan, arabinan and hemicellulose, respectively. The hydrolysate
produced under these optimal conditions is a promising source of
sugars, which could be used in the production of several chemi-
cal compounds by chemical or fermentation processes. Mannose,
for example, the main sugar obtained from SCG hydrolysis, could be
used for the production of mannitol, a chemical with a wide variety
of uses in the food industry.
Tokimoto, T., Kawasaki, N., Nakamura, T., Akutagawa, J., & Tanada, S. (2005). Removal
of lead ions in drinking water by coffee grounds as vegetable biomass. Journal
of Colloid and Interface Science, 281, 56–61.
Yen, W. J., Wang, B. S., Chang, L. W., & Duh, P. D. (2005). Antioxidant proper-
ties of roasted coffee residues. Journal of Agricultural and Food Chemistry, 53,
2658–2663.