Steam-explosion pretreatment of olive cake

Abstract

Olive cake was processed by steam-explosion under different steam conditions, followed by fractionation to separate the main
components. In the water-soluble fraction, the main compounds were carbohydrates. Glucose represented a significant part of
the total monosaccharide content, especially under conditions of mild severity, followed by arabinose, but the solubilization
of sugars occurred predominantly in the oligomeric fraction. Mannitol was also found in significant amounts (1.5%), similar
to that in the initial material. In the ethyl acetate extract, low molecular weight phenols were identified, the most aboundant
being hydroxytyrosol, which is present in the olive pulp. Hydroxytyrosol is abundant and has great antioxidant activity, reaching
149 mg/100 g of dry olive cake. The procedure used in this study obtained all the hydroxytyrosol residual present in the by-product.
The constitutive polymers were quantified in the insoluble fraction, and the sugar composition showed that cellulose was associated
with a high proportion of xylans and other polysaccharides rich in arabinose and galactose. This cellulose was nearly amorphous,
as it was highly susceptible to hydrolytic enzymes. The extractables in dilute alkali (not true lignins) increased as steaming
became more severe; the residual “lignin” in this fraction decreased. Enzymatic hydrolysis of the insoluble fraction using
a cellulolytic complex was also studied. The slight increase in the extent of saccharification was not proportional to the
high alkaline delignification. However, when the residues were efficiently delignified with chlorite treatment, the susceptibility
to enzymatic hydrolysis greatly increased.

Full text

ABSTRACT: Olive cake was processed by steam-explosion
under different steam conditions, followed by fractionation to
separate the main components. In the water-soluble fraction,
the main compounds were carbohydrates. Glucose represented
a significant part of the total monosaccharide content, espe-
cially under conditions of mild severity, followed by arabinose,
but the solubilization of sugars occurred predominantly in the
oligomeric fraction. Mannitol was also found in significant
amounts (1.5%), similar to that in the initial material. In the
ethyl acetate extract, low molecular weight phenols were iden-
tified, the most abundant being hydroxytyrosol, which is pres-
ent in the olive pulp. Hydroxytyrosol is abundant and has great
antioxidant activity, reaching 149 mg/100 g of dry olive cake.
The procedure used in this study obtained all the hydroxyty-
rosol residual present in the by-product. The constitutive poly-
mers were quantified in the insoluble fraction, and the sugar
composition showed that cellulose was associated with a high
proportion of xylans and other polysaccharides rich in arabi-
nose and galactose. This cellulose was nearly amorphous, as it
was highly susceptible to hydrolytic enzymes. The extractables
in dilute alkali (not true lignins) increased as steaming became
more severe; the residual “lignin” in this fraction decreased. En-
zymatic hydrolysis of the insoluble fraction using a cellulolytic
complex was also studied. The slight increase in the extent of
saccharification was not proportional to the high alkaline delig-
nification. However, when the residues were efficiently deligni-
fied with chlorite treatment, the susceptibility to enzymatic hy-
drolysis greatly increased.
Paper no. J9149 in JAOCS 77, 15–22 (January 2000).
KEY WORDS: By-product, cellulose, olive cake, phenols, sac-
charification, steam-explosion, sugars.
Olive cake is a by-product derived from the olive oil extrac-
tion industry. It is obtained by pressing the fruit, leaving a
residue of seed husks (fragmented olive stones), seed, pulp
and peel (olive cake). Olive-mill wastewater and/or vegeta-
tion water is also obtained.
Olive cake has been utilized as an energy source, fertilizer,
and animal feed (1). In spite of its high fiber and protein con-
tents, this by-product has low nutritional value owing to phe-
nolic compounds inhibiting digestive enzymes. During oil ex-
traction, polymers are formed among the phenolic substances
themselves and/or protein and cellulose so that they are un-
available for ruminant digestion. Consequently, these poly-
mers constitute an integral part of the cell wall component
and show physical and chemical properties very similar to
lignin (2). By way of certain pretreatments, it is possible to
reduce the degree of lignification of the olive cake, thus en-
hancing its nutritional value.
High-pressure steaming followed by rapid decompression
is called steam-explosion. Recently, steam-explosion has been
considered an effective pretreament (3) of waste cellulosic ma-
terials for further processing, including olive stones (4,5). The
resultant material is finely divided; and the main compo-
nents—cellulose, hemicelluloses, and lignin—are separated.
Furthermore, the enzymatic hydrolysis of cellulose is en-
hanced, which could create alternative uses for the olive cake.
The aim of this study was to apply steam-explosion to
olive cake under various conditions of severity to enhance the
effective utilization of such by-products. The solubilization
of the carbohydrates and the isolation of different organic sol-
uble substances, as well as the effect of additional treatments
on the enzymatic saccharification of steam-exploded olive
cake, were also determined.
EXPERIMENTAL PROCEDURES
Materials. Nondefatted and destoned olive cake was supplied
as pellets (average size 0.5–2.5 cm) by the oil extraction fac-
tory “Oleícola El Tejar” (Córdoba, Spain). The olive cake,
which included the residual olive stone was subjected to vi-
bratory ball milling for the purpose of determining the origi-
nal chemical composition.
Steam explosion. Steam explosion was carried out using a
2-L reactor, with a maximal operating pressure of 42 kg/cm
2
,
equipped with a ball valve opening. All experiments were car-
ried out on samples corresponding to 100 g of dry weight.
The olive cake was steamed for different periods of time
and temperatures, prior to rapid decompression. The severity
of the treatment was designated by a single factor called R
o
which links the effects of time (t, min) and temperature (T,
°C) (6).
R
o
= t exp(T − 100)/14.75) [1]
Fractionation. After steam explosion, the samples were
filtered through Albet filter paper (weight, 73 g/cm
2
; Albet,
Barcelona, Spain) in a Buchner funnel using vacuum. The
residue was washed with distilled water (3 × 150 mL) for 30
Copyright © 2000 by AOCS Press 15 JAOCS, Vol. 77, no. 1 (2000)
*To whom correspondence should be addressed at Avda. Padre Garcìa
Tejero, nº 4, 41012-Seville, Spain. E-mail: ahmoreno@cica.es
Steam-Explosion Pretreatment of Olive Cake
B. Felizón, J. Fernández-Bolaños, A. Heredia*, and R. Guillén
Departamento de Biotecnología de Alimentos, Instituto de la Grasa (CSIC), 41012, Seville, Spain

min at 60°C, shaken, and then filtered. The filtrate was con-
centrated to 250–300 mL by rotary evaporation at 40°C. The
aqueous concentrate was continuously extracted for 5–6 h
with ethyl acetate (refluxed at 77°C). The aqueous and or-
ganic phases were separated, and the organic phase was ro-
tary evaporated under vacuum at 40°C for several hours to re-
move all traces of ethyl acetate. A viscous dark brown extract
was obtained. The aqueous phase was freeze-dried. The
water-insoluble material was extracted with 0.5 N NaOH (250
mL) for 15 min at room temperature, and the aqueous alkali
extract was purged with nitrogen as necessary until the ex-
tract was relatively colorless. The dissolved lignin was acidi-
fied by drop-wise addition of 5 N H
2
SO
4
to pH 2–3. The pre-
cipitate was centrifuged, washed to neutral pH, and freeze-
dried.
Analytical methods. Moisture, fat, and ash contents were
determined according to AOAC methods (7). Protein was de-
termined by the micro-Kjeldahl method using the N × 6.25
conversion factor. Free sugars and uronic acids were quanti-
fied by colorimetric methods (8,9). Mannitol content was de-
termined by high-pressure liquid chromatography (HPLC)
using an Aminex HPX-87H column (Bio-Rad Laboratories,
Richmond, CA), refractive index detection, and 0.01 N
H
2
SO
4
as an eluent. Total polyphenols were colorimetrically
determined with Folin-Denis reagent and caffeic acid as stan-
dard (10). Cellulose, lignin, and hemicelluloses were deter-
mined according to Goering and Van Soest (11).
Detection and quantification of phenolic compounds in
ethyl acetate extracts were carried out by HPLC in a Waters
600 apparatus (Milford, MA) with an ultraviolet-visible pho-
todiode array detector (Waters 996). The separation was per-
formed using a Spherisorb ODS-2 column (5 µm, 250 × 4.6
mm, Tecnokroma, Barcelona, Spain), and the flow rate was 1
mL/min. The detection wavelength was 280 nm. The mobile
phase consisted of orthophosphoric acid in water, pH 2.5, and
acetonitrile, with a gradient from 5 to 25% of acetonitrile in
30 min, maintained for 10 min, and increased to 50% after 5
min. Phenolic compounds were identified by their retention
times and absoption spectra in the 200–380 nm range. Pheno-
lic standards were purchased from Sigma Chemical Co. (St.
Louis, MO) except for the oleuropein, which was obtained
from Extrasynthese (Genay, France). Hydroxytyrosol (3,4-di-
hydroxyphenyl ethanol) was obtained from oleuropein by
acid hydrolysis.
The content of water-soluble low-molecular-weight sugars
was determined by analysis with or without trifluoracetic acid
(TFA) hydrolysis (2 N TFA at 121°C for 1 h) prior to derivatiza-
tion to alditol acetates and then gas chromatography (GC) (12).
The composition of the noncellulosic neutral sugars of the
water-and-alkali insoluble fraction was determined by acid
hydrolysis with 2 N TFA (see above). Total sugars (cellulosic
and noncellulosic) were determined by two-stage acid hydrol-
ysis using 72% H
2
SO
4
at 40°C for the first stage and posthy-
drolysis with 1 M H
2
SO
4
at 100°C for 4 h for the second
stage. Neutral sugar released by the acid hydrolysis was mea-
sured by GC as alditol acetates.
Klason lignin was determined gravimetrically (13). The α-
cellulose determination was carried out from bleached cellu-
lose, which was extracted with 17.5% NaOH, and the residue
was measured gravimetrically. The hemicelluloses were also
determined as the difference between bleached cellulose and
α-cellulose. Bleached cellulose was prepared by chlorite
delignification (14).
Crystallinity of bleached cellulose preparations was deter-
mined by X-ray diffraction on a X-ray diffractometer
(Siemens D-5000) with a position-sensitive proportional
counter. The relative crystallinity indices (RCI) were calcu-
lated by the empirical method described by Segal et al. (15).
Cellulase activity was determined by the filter-paper assay
on Whatman number 1 paper to determine the total cellu-
lolytic activity after incubation in 0.05 M sodium acetate
buffer, pH 4.8 at 50°C for 1 h (16). The released reducing
sugars (RS) were determined by the dinitrosalicylic acid
(DNS) method using
D
-glucose as standard (17). One unit of
filter-paper activity (FPU) was the amount of enzyme that re-
leased 1 µmol of glucose/min.
Enzymatic hydrolysis. Hydrolysis was carried out in 50-
mL glass flasks in a shaker at pH 4.8 (0.05 M sodium acetate
buffer) and 50°C using a commercial cellulase (Cellubrix).
Cellubrix (Novo Nordisk Ferment AL, Dittingen, Switzer-
land) is a mixture of cellulase originating from Trichoderma
reesei and β-glucosidase originating from Aspergillus niger.
A weighed amount of water-and-alkali-insoluble fibers and/or
additional treated sodium chlorite fiber was placed in a flask,
and an enzyme solution supplemented with 0.01%
Thimerosal as a biocide was then added to a final volume of
5 mL. The enzyme activity was 70 FPU/g substrate, and the
concentration of the insoluble material was 8% (wt/vol). To
follow hydrolysis, 0.1-mL aliquots of the hydrolysate were
separated from the solid residue by filtration (through glass
wool placed in Pasteur pipettes) and after 1, 8, 24, 48, and
72 h were analyzed for RS. The overall sugar yield, or per-
centage of polysaccharides utilization, was calculated in rela-
tion to the polysaccharide content of the untreated original
material (% saccharification).
RESULTS AND DISCUSSION
Composition, treatment, and fractionation. The chemical
composition (g/100 g of dry weight ± SD, n = 3) of milled
olive cake was as follows: Moisture, 5.42 ± 0.45; fat, 3.28 ±
0.15; protein, 9.80 ± 1.30; free sugars, 1.88 ± 0.17; uronic
acids, 1.15 ± 0.01; fiber detergent-acid, 47.0 ± 0.22; fiber de-
tergent-neutral, 61.6 ± 4.58; cellulose, 27.9 ± 1.49; hemicel-
luloses, 14.6 ± 4.80; lignin, 16.8 ± 1.38; α-cellulose, 17.7 ±
0.90; klason lignin, 19.5 ± 1.40; polyphenols, 1.21 ± 0.01;
ash, 7.72 ± 0.61; mannitol, 1.45 ± 0.21.
Olive cake contains a large amount of fiber, approximately
60% of the dry matter, the cell walls being mainly polysac-
charides (cellulose 18–28%, hemicellulose 14%) and lignin
(17–20%). Since the type of parenchymal cell of the olive
fruit pulp is scarcely lignified (18), the lignin data seem to be
16 B. FELIZÓN ET AL.
JAOCS, Vol. 77, no. 1 (2000)

overestimated. It is composed mostly of condensed tannin
and other polymers. The protein (10%) and ash (8%) contents
were high. The fat contents (3%) of the samples studied were
high enough to continue with oil extraction. Free sugars con-
stituted about 2% of the dry matter, a low value compared to
the free sugar content of the olive pulp (13–14%) (19). This
indicated that a large part of the free sugars was lost during
the olive oil extraction process and/or during the storage of
this by-product. Since mannitol (1.5% of the dry weight) is
an unfermentable substance, the losses regarding the manni-
tol content in the pulp (about 5%) are minor.
The low uronic acid content (1.1%), compared to cellulose
and hemicellulose contents, indicates that only a small
amount of pectic substances was present, when in fact they
were the main polysaccharides of the olive cell wall. The rea-
son for this loss is similar to the loss of free sugar.
Olive cake was processed by steam-explosion under dif-
ferent steam temperatures for 135 s prior to rapid decompres-
sion (explosion). Table 1 shows the experimental conditions
of this study, as well as yields of both the water-soluble frac-
tion recovered (in g/100 g of dry initial olive cake) and the
insoluble residue recovered by filtration after washing.
The content of olive stone fragments (>0.2 mm) that re-
mained in the insoluble residue decreased as the steam condi-
tions became more severe, declining from 13 to 2%. These
data confirm that the pretreatments were effective in autohy-
drolysis of the heavily lignified material and the hard olive
seed husk (5).
Table 1 shows the yield of ethyl acetate extracts obtained
from their soluble fractions as a function of the severity of the
pretreatment. A considerable portion of the insoluble fraction
was extractable with dilute alkali and increased severity. Al-
though the lignin recovered by alkaline extraction, followed by
acid precipitation from steam-exploded olive stones, was simi-
lar to exploded hardwood lignin (4), the “lignin” obtained by
alkaline delignification from steam-exploded olive cake was
not a true lignin. This was confirmed by ultraviolet and infrared
spectroscopy (data not shown), and supports the idea regarding
the composition of the “lignin” present in the olive cake.
Characterization of the soluble fraction. The water-solu-
ble fraction was characterized after phenol and other products
formed from the thermal degradation with ethyl acetate had
been removed (Table 2). The main compounds were carbohy-
drates (23–27%), representing from 5.6 to 6.8% of the dry
weight of the initial olive cake.
The effect of steam-explosion pretreatment on hemicellu-
lose solubilization was low compared to other lignocellulosic
materials rich in glucuronoxylans, where the hemicelluloses
became almost completely water soluble (20). This can be ex-
plained by the characteristics of the major polysaccharides of
olive pulp, which is rich in pectic polysaccharides, xylan, xy-
loglucan, arabinan, and galactoglucomannan.
The amounts of neutral sugars, released either in mono-
meric or oligomeric form, are given in Table 3. Glucose, to-
gether with arabinose, represented a significant part of the total
monosaccharide content. Although part of these sugars may
originate from hemicellulose, a certain amount of the glucose
can be due to free sugars, which decreased in quantity because
of thermal degradation occurring at high severities.
Solubilization of sugars occurred predominantly in the
oligomeric fraction (Table 3), showing a large difference in
sugar composition when steam explosion was applied at
lower or higher severity. While the content of depolymerized
hemicelluloses decreased with severity, more xylose-rich
oligomers were solubilized. These differences were more pro-
nounced between the points of treatment of log R
o
3.36 and
3.81. Since the response of olive cake to the steam-explosion
process was different from that of other plant tissues, even
olive stones (which contain much more lignin and xylan than
olive cake with a large solubilization of oligosaccharides of
STEAM-EXPLOSION PRETREATMENT OF OLIVE CAKE 17
JAOCS, Vol. 77, no. 1 (2000)
TABLE 1
Experimental Conditions for Stream-Treated Olive Cake and Yield (g/100 g of dry initial
olive cake) of Both the Water Soluble Fraction and the Insoluble Residue Recovered
Severity index (log R
o
)
a
3.04 3.36 3.81 3.95 4.10 4.25
Time (s) 135 135 135 135 135 135
Temperature (°C) 193 204 219 224 229 234
Water-soluble substances 23.3 24.0 28.0 28.6 26.5 27.6
Ethyl acetate-extracted 1.60 2.00 2.20 2.60 2.80 2.00
Water-insoluble substances 56.9 51.7 54.7 54.5 60.9 60.0
Alkaline-extracted 9.60 10.7 12.5 21.6 25.8 33.5
Olive stone fragments 12.9 7.05 4.60 4.70 4.30 2.10
a
R
o
= t exp (T − 100/14.75), where t is time (min) and T is temperature (°C).
TABLE 2
Chemical Characterization of the Water-Soluble Fraction After Ethyl
Acetate Extraction (g/100 g water-soluble fraction freeze-dried)
Severity index (log R
o
)
3.04 3.36 3.81 3.95 4.10 4.25
Total sugars 26.3 25.6 24.9 22.7 24.3 26.6
Protein 8.21 11.0 10.5 10.9 11.1 11.0
Mannitol 7.05 5.37 5.00 5.45 4.99 5.41
Ash 16.1 17.1 16.1 15.4 15.2 14.6
Polyphenols 4.39 4.69 4.00 4.75 4.42 5.30
Uronic acids 2.56 2.7 2.18 2.16 2.10 2.01
Other compounds
a
35.4 33.6 37.3 38.6 37.9 35.1
a
Quantified by difference: 100 − % known compounds.

xylose) (5,20), one can suppose that at least a certain amount
of the xylose recovered from the pretreated olive cake could
be part of residual olive stones present in this material that
decreased with severity (Table 1).
In the soluble fraction, mannitol also stands out (5–7%)
(Table 2). The amount of this compound after the milder treat-
ment (7%) was similar to the initial material before treatment,
decreasing only slightly owing to thermal decomposition.
These quantities are relatively important and could represent
an economically interesting product from olive cake.
Table 2 also shows that a substantial portion of the water-
soluble material produced during pretreatment (34–39%) was
not identified and was quantified only by its difference from
known compounds. This could be due to the high percentage
of chemical transformations and to condensation reactions be-
tween carbohydrates, proteins, and polyphenols.
In the ethyl-acetate extracted phase, HPLC analysis
showed the presence of low molecular weight products that
were identified by comparing their retention times and ab-
sorption spectra in the ultraviolet region (200–400 nm) with
those of known compounds. The quantitative analyses of all
identified compounds are shown in Figure 1.
The amounts of identified phenolic acids (vanillic and sy-
ringic acids) and aldehydes (vanillin and syringaldheyde)
were low compared to those obtained from true lignocellu-
losic material (i.e., steam-exploded olive stones) (5). The rea-
sons are (i) that these substances originated from lignin degra-
dation and (ii) that the content of true lignin in olive cake was
low, because of a parenchymal skin with a primary wall. Only
vanillic acid, which has been reported as a phenol present in
olive pulp, reached values of 8.5 mg/100 g of the dry matter,
double the amount of the other phenols. Also present were
3,4-dihydroxybenzoic acid and pyrocatechin, which has also
been detected in the olive-mill wastewater or “alpechin” (21).
Of all the low-molecular-weight phenols solubilized by the
steaming pretreatment, hydroxytyrosol and tyrosol were the
most abundant (Fig. 1). These compounds were present in the
olive pulp (21) and seed (22), respectively, where they were
part of the phenolic glucosides. Hydroxytyrosol stands out for
its abundance and great antioxidant power. The amount solu-
bilized increased as the steaming conditions increased in
severity, reaching 149 mg/100 g of the dry olive cake. Start-
ing at log R
o
of 4.10, it began to decrease. This substance is a
commercially unavailable natural antioxidant and is responsi-
ble for the stability of oils that contain it (23). Its recovery
could be of interest.
These water-soluble, noncarbohydrate compounds re-
leased or formed during steam explosion show strong in-
hibitory effects against enzymes and microorganisms (24).
Therefore, these compounds must be removed before the
other solubilized sugars or cellulose can be efficiently con-
verted by microorganisms to a wide variety of products.
Characterization of the residual insoluble material.
Table 4 shows the chemical composition of the insoluble frac-
tion obtained from olive cake after steam explosion, water
washing, and alkali extraction. The constitutive polymers
were quantified from the standard methods used for quantifi-
cation of wood polymers. These results were similar when
hemicelluloses and cellulose were determined by hydrolysis
with TFA and H
2
SO
4
, respectively, followed by quantifica-
tion of monosaccharides by GC. The residual lignin was in-
soluble in H
2
SO
4
, the so-called Klason lignin (13).
The composition of sugars (Table 4) shows that cellulose
was associated with a high proportion of xylans and other
polysaccharides rich in arabinose and galactose. The rela-
tively high amount of hemicellulosic material that remained
in the insoluble residue (almost in their totality by milder
treatments log R
o
3.04 or 3.36) contrasts with the almost com-
18 B. FELIZÓN ET AL.
JAOCS, Vol. 77, no. 1 (2000)
TABLE 3
Composition of Neutral Sugars in Monomeric and Oligomeric Soluble Fraction
(g/100 g water-soluble fraction freeze-dried)
a
Severity index (log R
o
)
3.04 3.36 3.81 3.95 4.10 4.25
Rhamnose M 0.36 0.47 0.41 0.48 0.42 0.45
O 2.75 2.62 1.64 1.4 1.71 1.54
Arabinose M 1.12 1.30 1.23 1.47 1.23 1.31
O 8.48 6.95 4.16 3.37 3.98 4.78
Xylose M 0.15 0.17 0.22 0.29 0.30 0.42
O 1.47 4.15 9.37 7.70 8.08 9.44
Mannose
b
M 0.53 0.24 0.38 0.83 0.76 0.34
O 0.39 0.56 0.61 0.60 0.59 0.68
Galactose M 0.17 0.13 0.15 0.18 0.22 0.17
O 2.10 2.28 1.84 1.70 1.76 1.80
Glucose M 3.06 2.26 1.14 1.13 1.44 1.13
O 5.74 4.42 3.80 3.56 3.77 4.57
Total sugars M 5.39 4.57 3.53 4.38 4.37 3.82
O 20.9 21.0 21.4 18.3 19.9 22.8
a
Determined without (M) and with (O) trifluoroacetic acid hydrolysis prior to gas chromatographic
derivatization.
b
Data corrected with mannitol content obtained by high-performance liquid chromatography.

plete solubilization of the hemicelluloses in other heavily lig-
nified tissue rich in glucuronoxylans, where the autohydroly-
sis reaction was effective.
Since the total amount that could be extracted with dilute
alkali increased as the steaming became more severe
(Table 1), the residual “lignin” in this fraction decreased no-
tably, while the protein and ash content remained virtually un-
changed (data not shown).
The cellulose recovered from treated olive cake was nearly
amorphous, as the relative crystallinity index (RCI) was low
(Fig. 2). This characteristic of the cellulose had a profound ef-
fect on its reactivity, for the hydroxyl groups located in the
amorphous regions were highly accessible, reacted readily, and
made the cellulose susceptible to hydrolytic enzymes (25).
The enzymatic hydrolysis of the insoluble fraction using a
cellulolytic complex Cellubrix (an endo/exo cellulase prepa-
STEAM-EXPLOSION PRETREATMENT OF OLIVE CAKE 19
JAOCS, Vol. 77, no. 1 (2000)
FIG. 1. Evolution of the content (mg/100 g dry matter) of low-molecular-weight phenols obtained from steam-exploded olive cake as a function of
treatment severity (log R
o
), where R
o
= t exp (T − 100)/14.75, and t is time (min) and T is temperature (°C). (A) ●●, 3,4-Dihydroxybenzoic acid;
, hydroxytyrosol; ■, tyrosol. (B) ✕, Pyrocatechin; ●●, vanillic acid; ▲, syringic acid; ◆, vanillin; ■, syringaldehyde.
TABLE 4
Chemical Characterization and Sugar Composition of the Water-and-Alkali-Insoluble Frac-
tion (g/100 g insoluble fraction)
Severity index (log R
o
)
3.04 3.36 3.81 3.95 4.10 4.25
α-Cellulose 27.4 32.4 43.0 47.5 56.7 59.0
Hemicelluloses
a
29.8 27.6 19.4 16.6 18.4 21.8
Klason lignin 36.6 34.9 35.1 32.5 23.6 21.7
Sugars
b
Arabinose TFA 1.59 —
c
2.29 — — 1.33
H
2
SO
4
2.03 — 1.00 — — 1.06
Xylose TFA 18.3 — 7.08 — — 6.54
H
2
SO
4
17.5 — 4.28 — — 7.20
Mannose TFA 0.13 — 0.35 — — 0.37
H
2
SO
4
0.27 — 1.29 — — 0.90
Galactose TFA 0.73 — 0.71 — — 0.71
H
2
SO
4
0.87 — 0.98 — — 0.71
Glucose TFA 0.67 — 1.33 — — 1.87
H
2
SO
4
26.0 — 32.6 — — 45.1
a
Hemicellulose was determined by difference between bleached cellulose and α-cellulose.
b
Sugars determined by 2 N trifluoroacetic acid (TFA) and H
2
SO
4
hydrolysis prior to gas chromato-
graphic derivatization.
c
Not determined.

ration without pectinase activity) was also studied. At high
enzymatic concentrations, a maximum of 41.8 and 57.7% of
the total polysaccharides present in the insoluble residue of
pretreated material at log R
o
3.36 and 4.10, respectively, were
solubilized after 72 h (Fig. 3A). The slight increase in the ex-
tent of saccharification was not proportional to the large alka-
line delignification (Table 4), which occurred when the sever-
ity of the steaming treatment was increased. When the
residues were efficiently delignificed with chlorite treatment,
however, the susceptibility to enzymatic hydrolysis increased
notably, reaching 80–92% saccharification after only 24 h of
incubation (Fig. 3B). These results indicate that the accesibil-
ity of substrate may be hindered by certain residual phenolic
compounds or “lignin”-forming linkages with the polysac-
charides, so that drastic treatment of delignification for im-
proving the enzymatic action is necessary.
The amount of sugars released by enzymatic hydrolysis
was on the order of 15 g of fermentable sugar per 100 g of dry
olive cake after 72 h of incubation for a water-insoluble frac-
tion extracted with dilute alkali, and the yield increased to
about 20 g for combining alkali and chlorite bleaching pre-
treatment after only 24 h of incubation. Therefore, in addition
to the sugars solubilized during steam explosion, up to 7
g/100 g of dry matter such as monosaccharides or oligosac-
charides was left (Tables 1 and 3). The sugar yield obtained
by hydrolysis with cellulase from insoluble material of the
olive cake, steamed and alkali/chlorite pretreated, could be
considered a good source of fermentable carbohydrate.
ACKNOWLEDGMENTS
This work has been supported by the Comisión Interministerial de
Ciencia y Tecnología, Grants OLI 96-2127. The authors are grateful
to Dr. Manuel Brenes for the HPLC analysis of phenols, and to Oleí-
cola El Tejar (Córdoba) for supply of olive cake.
REFERENCES
1. Valiente, C., E. Arrigoni, R. Esteban, and R. Amadó. Chemical
Composition of Olive By-Product and Modifications Through
Enzymatic Treatments, J. Sci. Food Agric. 69:27–32 (1995).
2. Molina, E., and A. Nefzaoui, Recycling of Olive By-Products:
Possibilities of Utilization in Animal Nutrition, International
Biodeterioration and Biodegradation Symposium:227–235
(1996).
3. Duff, S.J.B., and W.D. Murray, Bioconversion of Forest Prod-
ucts Industry Waste Cellulosic to Fuel Ethanol: A Review,
Biores. Technol. 55:1–33 (1996).
4. Fernández-Bolaños, J., B. Felizón, A. Heredia, R. Guillén, and
A. Jiménez, Characterization of the Lignin Obtained by Alka-
line Delignification and of the Cellulose Residue from Steam-
Exploded Olive Stones, Ibid. 68:1–12 (1998).
5. Fernández-Bolaños, J., B. Felizón, M. Brenes, R. Guillén, and
A. Heredia, Hydroxytyrosol and Tyrosol as the Main Com-
pounds Found in the Phenolic Fraction of Steam-Exploded
Olive Stones, J. Am. Oil Chem. Soc. 75:1–7 (1998).
6. Heitz, M., F. Carrasco, M. Rubio, A. Brown, and E. Chornet,
Physicochemical Characterization of Lignocellulosic Substrates
Pretreated via Autohydrolysis: An Application to Tropical
Woods, Biomass 13:255–273 (1987).
7. AOAC, Official Methods of Analysis, Food Composition, 15th
edn., Association of Official Analystical Chemists, Arlington,
Vol. 1, 1990, pp. 915–919.
20 B. FELIZÓN ET AL.
JAOCS, Vol. 77, no. 1 (2000)
FIG. 2. X-ray diagram of cellulose from steam-exploded olive cake (log R
o
= 3.36). The relative crystallinity index
(45.2) was calculated from X-ray diffractograms by the method of Segal et al. (15) as [(I
(002)
− I
am
)/I
(002)
] × 100,
where I
(002)
is the maximum intensity of 2 θ = 22.5° (reflection attributed to the crystalline region) and I
am
is the in-
tensity at 2 θ = 18.5° (reflection attributed to amorphous region).

8. Dische, Z., Color Reactions of Carbohydrates, in Methods in
Carbohydrates Chemistry, edited by R.L. Whistler and M.L.
Wolfram, Academic Press, New York, 1962, pp. 477–512.
9. Blumenkrantz, N., and G. Asboe-Hansen, New Method for
Quantitative Determination of Uronic Acids, Anal. Biochem. 54:
484–489 (1973).
10. Garrido, A., and M. Fernández, Estudio crítico de la determi-
nación colorimétrica de polifenoles y su aplicación a las
salmueras de aceitunas negras de mesa, Grasas Aceites 27:
107–112 (1976).
11. Goering, H.K., and P.J. Van Soest, Forage Fiber Analysis (ap-
paratus, reagents, procedures and some applications), J. Assoc.
Off. Anal. Chem. 4:1–10 (1973).
12. Englyst, H.N., and J.H. Cumming, Simplified Method for the
Measurement of Total Non-starch Polysaccharides by Gas Chro-
matography of Constituent Sugars as Alditol Acetates, Analyst
109:937–942 (1984).
13. Fengel, D., and G. Wegener, Wood—Chemistry, Ultrastructure
and Reactions, edited by W. de Gruyter, Berlin and New York,
1984, pp. 92–95.
14. Browning, B.L., Methods of Wood Chemistry, edited by Inter-
science, New York, 1967.
15. Segal, L., J.J. Greeley, J.R. Martin, and C.M. Conrad, An Em-
pirical Method for Estimating the Degree of Crystallinity of Na-
tive Cellulose Using the X-ray Diffractometer, Text. Res. J.
29:786–794 (1959).
STEAM-EXPLOSION PRETREATMENT OF OLIVE CAKE 21
JAOCS, Vol. 77, no. 1 (2000)
FIG. 3. Enzymatic hydrolysis of (A) the water-and-alkali insoluble fraction, and (B) the water-
and-alkali insoluble fraction posttreated with sodium chlorite, at severities of log R
o
3.36 (open
bars) and of log R
o
4.10 (solid bars). Experiments were done in triplicate at an enzyme con-
centration of 70 FPU/g cellulose. The values for the reducing sugars (RS) are the averages ±
SD. Values above each bar represent saccharification ratios (%) referred to as a percentage of
the theoretical yield of the maximal total sugars. The hydrolysis conditions are given in the Ex-
perimental Procedures section. FPU, filter paper activity units.

16. Mandels, M., and D. Sternberg, Recent Advances in Cellulase
Technology, J. Ferment. Technol. 54:267–286 (1976).
17. Miller, G.L., Use of Dinitrosalicylic Acid Reagent for Determi-
nation of Reducing Sugars, Anal. Biochem. 31:426–428 (1959).
18. Jiménez, A., R. Guillén, C. Sánchez, J. Fernández-Bolaños, and
A. Heredia, Changes in Texture and Cell Wall Polysaccharides
of Olive Fruit During “Spanish Green Olive” Processing, J.
Agric. Food Chem. 43:2240–2246 (1995).
19. Guillén, R., A. Heredia, B. Felizón, A. Jiménez, A. Montaño, and
J. Fernández-Bolaños, Fibre Fraction Carbohydrates in Olea eu-
ropaea (Gordal and Manzanilla var.), Food Chem. 44: 173–178
(1992).
20. Yu., V.E., H. Ding, J.M. Labavitch, and S.P. Shoemaker, Enzy-
matic Hydrolysis of Pretreated Rice Straw, Biores. Technol. 59:
109–119 (1997).
21. Limiroli, R., R. Consonni, A. Ranalli, G. Bianchi, and L.H.
Zetta, NMR Study of Phenolics in the Vegetation Water of
Three Cultivars of Olea europaea: Similarities and Differences,
J. Agric. Food Chem. 44:2040–2048 (1996).
22. Maestro-Durán, R., R. León-Cabello, V. Ruiz-Gutiérrez, and A.
Vázquez-Roncero, Glucósidos fenólicos amargos de la semilla
del olivo (Olea europaea), Grasas Aceites 45:332–335 (1994).
23. Baldioli, M., M. Servili, G. Peretti, G., and G.F. Montedoro, An-
tioxidant Activity of Tocopherols and Phenolic Compounds of
Virgin Olive Oil, J. Am. Oil Chem. Soc. 73:1589–1593 (1996).
24. Palmqvist, E., B. Han-Hägerdal, Z. Szengyel, G. Zacchi, and K.
Rèczey, Simultaneous Detoxification and Enzyme Production
of Hemicellulose Hydrolysates Obtained After Steam-Pretreat-
ment, Enzyme Microbial. Technol. 20:286–293 (1997).
25. Sjöström, E., Wood Chemistry: Fundamentals and Applications,
2nd edn., Academic Press Inc., San Diego, 1993, pp. 204–206.
[Received February 16, 1999; accepted July 26, 1999]
22 B. FELIZÓN ET AL.
JAOCS, Vol. 77, no. 1 (2000)