Anda di halaman 1dari 6

Appl Microbiol Biotechnol (2000) 54: 162±167 Ó Springer-Verlag 2000


A. M. LoÂpez-Contreras á P. A. M. Claassen
H. Mooibroek á W. M. De Vos

Utilisation of saccharides in extruded domestic organic waste

by Clostridium acetobutylicum ATCC 824 for production of acetone,
butanol and ethanol

Received: 5 November 1999 / Received revision: 21 February 2000 / Accepted: 25 February 2000

Abstract Domestic organic waste (DOW) collected in

The Netherlands was analysed and used as substrate for
acetone, butanol and ethanol (ABE) production. Two
The production of acetone, butanol and ethanol (ABE)
di€erent samples of DOW, referred to as fresh DOW
by solventogenic Clostridia is a well-known process.
and dried DOW, were treated by extrusion in order to
During the ®rst part of this century it was, after the
expand the polymer ®bres present and to obtain a ho-
production of ethanol, the second largest biotechnolog-
mogeneous mixture. The extruded material was analysed
ical process in the world. Following World War II, the
with respect to solvent and hot water extractives, uronic
ABE fermentation process was no longer economically
acids, lignin, sugars and ash. The total sugar content in
viable as a result of the development of the petrochem-
the polymeric fractions of the materials varied from
ical industry, which made the production of solvents
27.7% to 39.3% (w/w), in which glucose represented the
from petroleum derivatives cheaper (Jones and Woods
18.4 and 25.1% of the materials, for fresh and dried
1986; DuÈrre 1998). During the past 20 years there has
DOW, respectively. The extruded fresh DOW was used
been a revival of interest in the ABE fermentation, since
as substrate for the ABE fermentation by the solvento-
renewable resources have become possible alternative
genic strain Clostridium acetobutylicum ATCC 824. This
substrates for the production of chemicals and fuels,
strain was grown on a suspension of 10% (w/v) DOW in
instead of oil-derived feedstocks (Claassen et al. 1999).
demineralised water without further nutrient supple-
At present, one of the major bottlenecks hampering its
ment. This strain produced 4 g ABE/100 g extruded
economic viability is the cost of substrates, accounting
DOW. When C. acetobutylicum ATCC 824 was grown
for up to 60% of the total production costs.
on a suspension of 10% (w/v) DOW hydrolysed by a
Since solventogenic Clostridia are able to utilise a wide
combination of commercial cellulases and b-glucosid-
range of carbohydrate substrates (Mitchell 1998), con-
ases, the yield of solvents increased to 7.5 g ABE/100 g
siderable research into the use of substrates cheaper than
extruded DOW. The utilisation of sugar polymers in
molasses (the traditional substrate for ABE production)
both hydrolysed and non-hydrolysed DOW was deter-
has been done. From the compounds that have been in-
mined, showing that only a small proportion of the
vestigated as possible substrates for ABE fermentation,
polymers had been consumed by the bacteria. These
waste products from the dairy industry (i.e. whey) and
results indicate that growth and ABE production on
agricultural wastes seem to be the most interesting sub-
DOW is mainly supported by soluble saccharides in the
strates (Jones and Woods 1986; Claassen et al. 1998;
DuÈrre 1998). The agricultural wastes are mainly com-
posed of lignocellulose, which is considered to have great
potential as a substrate for fermentation, provided that
A. M. LoÂpez-Contreras á P. A. M. Claassen (&) á H. Mooibroek its cellulosic and hemicellulosic fractions can be de-
Agrotechnological Research Institute (ATO-DLO), polymerised and utilised eciently. Although many sol-
Bornsesteeg 59, 6708 PD Wageningen, The Netherlands ventogenic Clostridia can utilise all sugars present in
e-mail: wood cellulose and hemicellulose hydrolysates and can
Tel.: +31-317-475325
Fax: +31-317-475347 degrade polymers such as starch or xylan, they are not
able to grow on cellulose (Lee et al. 1985; Mitchell 1998).
W. M. De Vos
Laboratory of Microbiology,
For the use of lignocellulosic substrates, several strategies
Hesselink van Suchtelenweg 4, 6703 CT Wageningen, have been studied, such as the use of hydrolysates, co-
The Netherlands cultures with true cellulolytic organisms, or the addition

of cellulases to the fermentation medium (Jones and Analytical methods

Woods 1986). For the production of hydrolysates, the
The composition of DOW was determined as follows. First, sam-
ligno-cellulosic material is ®rst subjected to a pre-treat- ples were extracted according to the Technical Association of the
ment, such as steam-explosion or extrusion, in order to Pulp and Paper Industry method T 264 om-88 (TAPPI 1999a). For
expand the polymer ®bers and facilitate their hydrolysis. this purpose, 1 g of freeze-dried sample was extracted in a Soxtec
The hydrolysis can be done chemically by acid treat- System HT6 (Tenacor, HoÈganaÈs, Sweden) with, successively, eth-
anol/toluene (2:1) and ethanol. Thereafter, the samples were ex-
ment, or enzymatically. Both types of hydrolysates have tracted with hot water for 1 h at 100 °C. Extractive free samples
been tested as media for fermentation (Maddox and were dried at 60 °C for 24 h and hydrolysed with 12 M H2SO4 at
Murray 1983; Claassen et al. 1998). In this study, the 30 °C for 1 h followed by 1 M H2SO4 at 100 °C for 3 h. In the
application of extruded domestic organic waste (DOW) hydrolysate, after reduction and acetylation, neutral sugars were
as substrate for ABE fermentation is described. determined by gas liquid chromatography in a Hewlett Packard
model 5890A series II gas chromatograph (Hewlett-Packard,
Avondale, USA) equipped with a CP-SIL 88 WCOT fused silica
column (Chrompack, Middelburg, The Netherlands) and a ¯ame
Materials and methods ionisation detector. Helium was the carrier gas. Injector and de-
tector temperatures were 300 °C, and column temperature was
220 °C. Analysis of the chromatograms was carried out using
ChemStation software (Hewlett-Packard, Avondale, USA). Acid-
soluble lignin was determined spectrophotometrically at 205 nm
Clostridium acetobutylicum ATCC 824 was kindly provided by Dr.
following method T UM 250 (TAPPI 1991). Uronic acids were
Soucaille (Insa, Toulouse, France). Stock cultures were maintained
measured using the method of Blumenkrantz and Asboe Hansen
as spore suspensions in sterile 10% (v/v) glycerol at )20 °C. Spore
(1973). Acid-insoluble lignin was determined gravimetrically. Ashes
suspensions were heat-shocked for 10 min at 70±80 °C in a water
were determined by combustion of the samples at 575 °C as de-
bath prior to inoculation. For the production of pre-cultures,
scribed in method T 211 om-93 (TAPPI 1999b).
vegetative cells were grown in a semi-synthetic medium described
Sugars, saccharides, solvents, acids and furfural compounds
previously (Nimcevic et al. 1998) for 24 h at 37 °C. All experiments
were determined in clear supernatants of centrifuged culture sam-
were performed anaerobically, in a Coy anaerobic chamber (Coy
ples. Glucose was determined enzymatically using the Biotrol
Laboratory Products, USA) under an atmosphere of 20% CO2, 4%
Glucose Enzymatique Color kit (Merck, Darmstadt, Germany).
H2, 76% N2, unless indicated otherwise.
Other mono- or disaccharides were determined by high-pressure
liquid chromatography (HPLC). After passing the samples through
Sep-Pak C18 cartridges (Waters, USA) previously equilibrated
Substrates, pre-treatment and fermentation dropwise with 2 ml of methanol and 4 ml of milliQ water, sugars
were separated on a CHO-682 column (InterAction, USA) at
Two di€erent samples of DOW were used in this study. The ®rst, 85 °C. MilliQ water was used as eluent at a ¯ow rate of 0.4 ml/min.
referred to as fresh DOW, was collected in The Netherlands during Solvents, acids and furfural compounds were analysed by HPLC as
the summer season of 1998 and was stored at 4 °C for 15 days. The described previously (Gosselink et al. 1995). As internal standards,
second, referred to as dried DOW, was kindly provided by the propionic acid was used in solvent and acid analyses, while phen-
composting company VAM (The Netherlands) after collection oxyacetic acid was used in furfural and 5-hydroxymethyl-2-fural-
during the spring season. This sample was washed twice with water dehyde (HMF) determination. Furfural compounds were extracted
and dried to 85% (w/w) dry matter content. Both samples were from the samples by passing 2 ml of sample through a Sep-Pak C18
treated in a Clextral BC45 extruder (Clextral, Firminy, France) cartridge (Waters, Milford, USA) previously equilibrated as de-
under the following conditions: screw speed 100 rpm, production scribed above and eluted from the cartridge with 2 ml of methanol.
rate 12.8 kg/h for DOW and 9.9 kg/h for VAM-DOW, temperature Separation was carried out using a Shodex KC-311 column (Sho-
120 °C, reversed screw con®guration RSE-25H10/RSE-25H6. Ex- dex, Japan) at 80 °C with 3 mM H2SO4 as eluent at a ¯ow rate of
truded samples were freeze-dried, milled to a size particle of 0.5 mm 1 ml/min. A refractive index (RI) detector (Waters 410, Millipore,
in a Retsch mill and stored at room temperature until further use. Milford, USA) and an UV absorbance detector (model VWM2141,
Freeze-dried DOW was suspended in demineralised water at Pharmacia, Uppsala, Sweden) were used in series. Concentrations
10% (w/v) in serum ¯asks, closed with a rubber top and metallic of all mentioned metabolites were determined from the RI chro-
cap. The solutions were ¯ushed with N2 to remove O2 and then matograms, except furfural and HMF, which were determined
sterilised at 121 °C for 15 min. Inoculation was done with 2% (v/v) from the UV chromatograms at 280 nm.
overnight pre-cultures. A device composed of a needle, a 0.22-lm
sterile ®lter and a bicycle tyre valve were used to prevent the de-
velopment of pressure in the ¯asks. Cultures were incubated at
Scanning electron microscopy analysis
37 °C without agitation.
For measuring growth, 100 ll of culture were taken and diluted
Samples of cultures were ®ltered using sintered glass ®lters. A small
ten-fold in sterile holding bu€er (1 mM MgSO4, 25 mM KHPO4,
portion of the material on the ®lter was analysed by cryo-scanning
pH 7). The number of total cells was determined in a Neubauer
electron microscopy (cryo-SEM), using a Philips 515 microscope
counting chamber.
equipped with an Oxford cryo-unit.

The commercially available enzyme preparations Celluclast 1.5L
(cellulases) and Novozym N188 (b-glucosidase) were a kind gift
from Novo Nordisk (Bagsvaerd, Denmark). The ratio of these Extrusion and hydrolysis of DOW
enzyme preparations used for hydrolysis was 1:0.2 (Celluclast 1.5L:
Novozym N188), as advised by the manufacturer. Concentrations After storing the freshly collected DOW for 15 days at
of enzyme preparations are given as grams of enzyme preparation/
100 g of substrate. Enzyme preparations were ®lter-sterilised prior
4 °C, no signs of deterioration by microbial growth were
to addition to the substrate. The hydrolysis mixtures were incu- observed. The collected material appeared to be very
bated at 45 °C in a shaking water bath. heterogeneous, with a high content of vegetables and

fruits. The dried DOW, on the other hand, contained

mainly garden waste. After storage for 12 months in a
closed container, this dried DOW did not show micro-
bial degradation either, probably due to its low moisture
content (15%; w/w). The composition of these samples
was determined after extrusion and freeze-drying and is
shown in Table 1.
The fresh DOW showed approximately three times
more solvent and hot water extractives than the dried
DOW. These extractives represent the solvent and water-
soluble compounds present in the samples (proteins, free
sugars, etc.) and, in the case of the dried DOW, these
were partially removed by the washing treatment
applied. Also, the dried DOW had a higher content of
acid-insoluble lignin, ashes and sugars. Fig. 1 Hydrolysis of sterile 10% (w/v) fresh domestic organic waste
In order to determine optimal conditions for the en- (DOW) by di€erent concentrations of Celluclast 1.5L (C) and
zymatic hydrolysis of the DOW, di€erent concentrations Novozyme N188 (N). Ð Blank, h C 1%, j C 1% + N 0.2%, s
and combinations of cellulases and b-glucosidases were C 3%, d C 3% + N 0.6%, n C 4%, m C 4% + N 0.8%. Con-
centrations of C and N given as % g preparation/g substrate. Values
tested for the generation of glucose. The increase in are averages of duplicate determinations
glucose concentration in fresh DOW suspension during
hydrolysis is shown in Fig. 1. When only Celluclast 1.5L
was used, the highest concentration of glucose obtained respectively, were used. Only fresh DOW was used as
was 5.2 g/l after 72 h of hydrolysis; and this did not substrate for fermentation, since the concentration of
increase by increasing the concentration of Celluclast glucose obtained during hydrolysis was higher than that
1.5L from 3% to 4% (w/w). However, an increase in obtained from dried DOW.
glucose concentration was observed in the soluble frac-
tion of the DOW suspension. It rose to 18 g/l using a
combination of Celluclast 1.5L at 3% (w/w) and Fermentation of DOW by C. acetobutylicum
Novozyme N188 at 0.6% (w/w). The addition of ATCC 824
b-glucosidases (Novozyme N188) results in hydrolysis of
cellobiose to glucose and has been shown to prevent the Since the DOW suspensions were sterilised by autocla-
inhibition of cellulases, making the hydrolysis of the ving prior to fermentation, its e€ect on the composition
substrate signi®cantly more ecient (Breuil et al. 1992; of the samples was analysed. For sterilisation of the
Gregg and Saddler 1996). samples, extruded fresh DOW was resuspended at 10%
With the dried DOW, similar results were obtained, (w/v) in demineralised water and this suspension was
with maximum concentrations of glucose after 72 h of autoclaved at 121 °C for 15 min. The resulting solution
hydrolysis of 5.5 g/l when only Celluclast 1.5L at 3% was then freeze-dried, homogenised and used for anal-
(w/w) was added, and 9.6 g/l when Celluclast 1.5L at 3% ysis. The di€erences in composition between the sterile
(w/w) and Novozyme N188 at 0.6% (w/w) were added. and non-sterile samples were minimal (results not
For further experiments, concentrations of 3% and shown). During treatment of lignocellulosic materials at
0.6% (w/w) of Celluclast 1.5L and Novozyme N188, high temperature and low pH, furfural compounds can
be formed as a result of the degradation of sugars
(Larsson et al. 1999). Since furfural compounds at cer-
Table 1 Composition (w/w % in dry matter) of fresh domestic
organic waste (DOW) and dried DOW. Values are means ‹SD tain levels could inhibit bacterial growth, their forma-
(n = 3) tion should be avoided when the material is intended to
be used for fermentation (Larsson et al. 1999; Taher-
Fraction Fresh DOW Dried DOW zadeh et al. 1999). The furfural and HMF present in the
Solvent extractives 19.1 ‹ 0.8 5.9 ‹ 0.9 soluble fraction of 10% (w/v) DOW suspensions were
Hot water extractives 24.4 ‹ 2.5 7.2 ‹ 0.6 measured before and after sterilisation. No furfural or
Acid soluble lignin 1.5 ‹ 0.05 1.2 ‹ 0.03 HMF was detected before sterilisation and only 3.5 mg
Acid insoluble lignin 1.5 ‹ 0.25 10.5 ‹ 0.8 of HMF/l were found in sterile medium.
Uronic acids 8.0 ‹ 0.41 4.9 ‹ 1.6
C. acetobutylicum ATCC 824 was grown on 10% (w/v)
Sugars fresh DOW and 10% (w/v) hydrolysed fresh DOW in
Glucose 18.4 ‹ 0.6 25.1 ‹ 1.7
Xylose 4.0 ‹ 0.2 8.4 ‹ 0.7
demineralised water without further addition of nutri-
Arabinose 1.5 ‹ 0.03 2.3 ‹ 0.03 ents. These media contained insoluble material and thus
Mannose 2.0 ‹ 0.3 1.6 ‹ 0.1 it was impossible to determine cell growth by measuring
Galactose 1.6 ‹ 0.1 1.6 ‹ 0.1 the optical density of the cultures. Therefore, growth
Rhamnose 0.2 ‹ 0.01 0.3 ‹ 0.01 was followed by total cell counting using phase contrast
Ashes 9.1 ‹ 0.2 15.9 ‹ 1.1
microscopy. The growth curves obtained on both media

were very similar and showed a short lag phase followed

by an exponential growth phase, resulting in a cell
concentration of approximately 109 total cells/ml
(Fig. 2). To determine whether bacterial cells would be
attached to the substrate particles, ®ltrated samples of
fermentation cultures were analysed by cryo-SEM. In
the cryo-SEM photographs, no substrate-attached cells
were observed (results not shown), which is in agreement
with the previously observed lack of attachment of
C. acetobutylicum ATCC 824 to cellulose (Gelhaye et al.
The acids and solvents produced and the glucose
consumed during fermentation of 10% (w/v) fresh
DOW and 10% (w/v) hydrolysed fresh DOW in demi-
neralised water were determined (Fig. 3). During expo-
nential growth of the bacteria, acetic and butyric acids
were produced; and during the stationary growth phase,
solvent production (ABE) and partial re-assimilation of
the acids were observed. After 120 h of fermentation,
3 g of butanol/l were produced from fresh DOW and no
acetone or ethanol were found in the culture medium;
and butyric acid was the main product (3.5 g/l). In
contrast, when hydrolysed fresh DOW was the sub-
strate, after 120 h of fermentation the main product was
butanol, with a concentration of approximately 4.2 g/l;
and 1.5 g of acetone/l and 0.4 g of ethanol/l were found.
In the fermentation of hydrolysed fresh DOW, almost
no change in the concentrations of acids and solvents in
the medium after 48 h of fermentation was observed,
while during fermentation of fresh DOW there was a Fig. 3 Acids and solvents production by C. acetobutylicum ATCC
continuous increase in product formation. 824 on A 10% fresh DOW and B 10% hydrolysed fresh DOW. h
Butanol, j butyric acid, n acetone, m acetic acid, e ethanol,
* glucose. Values are averages of duplicate experiments
Saccharide utilisation from DOW
by C. acetobutylicum ATCC 824 during fermentation was followed and, after 4 days all
of them, except for arabinose, had been consumed. A
Table 2 shows the mono- and disaccharides in the sol- small amount of unidenti®ed material was detected in
uble fraction of sterile 10% (w/v) fresh DOW and 10% the RI chromatograms during HPLC saccharide analy-
(w/v) hydrolysed fresh DOW. The concentration of sis of the soluble fraction of 10% (w/v) fresh DOW
these saccharides in the soluble fraction of the media suspension. This material largely disappears during fer-
mentation; and hence it could represent other mono-, di-
or oligosaccharides that were utilised by the bacteria
during growth. This material did not appear in samples
from 10% (w/v) hydrolysed DOW, indicating that it
may correspond to oligosaccharides that are hydrolysed
by the cellulolytic enzymes.

Table 2 Mono- and disaccharides present in the soluble fractions

of 10% (w/v) fresh DOW and 10% (w/v) hydrolysed fresh DOW in
demineralised water. Values (g/l) are the average of at least three
di€erent measurements ‹SD. ND Not detected

10% DOW 10% hydrolysed DOW

Glucose 1.2 ‹ 0.2 18.1 ‹ 1.8

Fructose 2.5 ‹ 0.3 2.9 ‹ 0.3
Galactose ND 0.2 ‹ 0.04
Fig. 2 Growth curve of Clostridium acetobutylicum ATCC 824 in Arabinose 0.1 ‹ 0.004 0.3 ‹ 0.06
10% (w/v) DOW and 10% (w/v) hydrolysed DOW in demineralised Cellobiose 0.6 ‹ 0.1 ND
water. j Total cells DOW, r total cells hydrolysed DOW. Values are Maltose 0.2 ‹ 0.1 ND
averages of duplicate determinations

The insoluble fractions of fresh DOW and hydrolysed washing of the dried DOW prior to extrusion a€ected
fresh DOW remaining after fermentation were analysed the amount of extractives in the sample (Table 1).
in order to determine whether the (hemi)cellulosic Therefore, the percentages in the composition of both
polymers had been utilised by the bacteria during samples cannot be compared directly. Even if the ex-
growth. All polymeric sugars present in the DOW were tractives are not taken into account, the percentages of
only partially utilised. After fermentation of the hydro- glucose and lignin in the polymeric fractions of the fresh
lysed fresh DOW, 73% (w/w) of the initial amount of DOW and the dried DOW were very di€erent (61% and
polymeric glucose was consumed, while after fermenta- 49% for glucose, respectively, and 10% and 23% for
tion of fresh DOW, the polymeric glucose consumed was lignin, respectively). This indicates that when the DOW
12% (w/w) of the initial amount. From these data we had a higher content of food and fruit wastes (fresh
calculated that during enzymatic hydrolysis of the DOW DOW), the polymeric fraction showed a higher content
about 60% of the total polymeric glucose was hydroly- in glucan and a much lower content in lignin than when
sed to soluble glucose. the DOW was richer in garden waste (dried DOW).
The soluble fraction of 10% (w/v) fresh DOW con- As the extruded DOW was intended to be used as
tained 1.2 g of glucose/l, while in the soluble fraction of substrate for fermentation, the temperature during ex-
10% hydrolysed fresh DOW the concentration of glu- trusion was kept at 120 °C to prevent degradation of
cose was 18.1 g/l (Table 2). In the hydrolysed DOW sugars and formation of furfural compounds. In the case
solution, some of the glucose in the soluble fraction of the medium used in this study, the concentration of
could be derived from the hydrolysis of soluble disac- HMF was lower than that found in hydrolysates made
charides like maltose or cellobiose, amounting to 0.8 g/l from similar, steam exploded lignocellulosic material
(Table 2). Table 3 shows that the amount of glucose (Dr J. Ballesteros, CIEMAT, Madrid, Spain, personal
from the polymeric fraction of DOW is only 12.2 g/l. communication).
After correction of the total amount of soluble glucose After enzymatic hydrolysis followed by fermentation
in hydrolysed DOW (18.1 g/l) with the amount of glu- of the DOW, 27% of the glucose in the polymeric
cose coming from disaccharides, the polymeric fraction fraction was still present (Table 3). The incomplete en-
of the DOW and the soluble glucose already present zymatic hydrolysis of cellulose in lignocellulosic mate-
before hydrolysis, the residual glucose is 3.8 g/l. This rials could be due to a number of factors, including the
glucose could be a result of hydrolysis by the cellulases presence of inhibitors for the enzymes (Excoer et al.
of soluble oligosaccharides present in the DOW sus- 1991), product inhibition phenomena, inaccessibility due
pension. These oligosaccharides may correspond to the to a high degree of crystallinity in the cellulose, or the
observed unidenti®ed material that appears in the RI presence of other components, such as hemicellulose or
chromatograms from samples of the soluble fraction of lignin (Gregg and Saddler 1996).
10% (w/v) DOW. Since these peaks disappear during In the soluble fraction of 10% (w/v) fresh DOW in
fermentation, it seems that C. acetobutylicum ATCC 824 demineralised water, there were mono- and disaccha-
utilises these oligosaccharides during growth. rides (Table 2) and, most probably, oligosaccharides as
well. These substrates were enough to support growth
up to 109 total cells/ml, but not enough to support the
Discussion shift from the acidogenic to solventogenic phase; and
thus butyric acid is the main product formed during
The composition of DOW varies with the season. In fermentation (Fig. 3). Butanol is the main solvent pro-
spring and autumn the percentage of garden waste (rich duced and no acetone was found after 120 h of fer-
in woods) in the DOW is higher than in winter or mentation. In the case of hydrolysed DOW, the solvent
summer (Faaij et al. 1997). In this study, DOW collected production is higher and after 48 h of fermentation the
in two di€erent seasons and subjected to di€erent pre- solvent production was complete. The total amount of
treatments was analysed. It is very likely that the solvents (ABE) produced from hydrolysed fresh DOW
after 48 h of fermentation is 4.3 times higher than that
from fresh DOW during the same period of fermenta-
Table 3 Composition (in g/l) of the polymeric fraction of the tion: 6.8 g ABE/100 g extruded fresh DOW and
media 10% (w/v) fresh DOW and 10% (w/v) hydrolysed fresh 1.6 g ABE/100 g extruded fresh DOW produced from
DOW after fermentation by Clostridium acetobutylicum ATCC
824. Values are means of triplicate samples ‹SD
fresh DOW. The amount of solvents produced per gram
of soluble substrate consumed after 120 h of fermenta-
10% DOW 10% DOW 10% hydrolysed tion, calculated taking into account the sugars in
t=0h t = 96 h DOW t = 96 h Table 2 plus 2.8 g of oligosaccharides/l in the case of
Glucose 16.8 ‹ 0.7 14.7 ‹ 0.2 4.6 ‹ 0.3
fresh DOW, is very similar in both DOW and hydroly-
Xylose 4.8 ‹ 0.2 4 ‹ 0.3 2.3 ‹ 0.02 sed DOW. However, in the case of hydrolysed fresh
Arabinose 1.3 ‹ 0.1 0.5 ‹ 0.01 0.3 ‹ 0.02 DOW, the fermentation was complete after 48 h and the
Mannose 2 ‹ 0.3 1.8 ‹ 0.2 0.7 ‹ 0.02 concentration of solvents in the medium was higher than
Galactose 1.4 ‹ 0.1 1 ‹ 0.1 0.5 ‹ 0.02 that from fresh DOW, making this fermentation eco-
Rhamnose 0.2 ‹ 0.01 0.2 ‹ 0.02 0.2 ‹ 0.02
nomically more ecient. As observed before (Jones and

Woods 1986; Girbal and Soucaille 1998), these data Chartier P, Ferrero GL (eds) 10th Eur Conf ``Biomass for
show that the level of sugars in the medium is a very Energy and Industry'', WuÈrzburg, Germany, pp 138±141
Claassen PAM, Lier JB van, LoÂpez-Contreras AM, Niel EWJ van,
important factor for triggering the shift from the acid- Sijtsma L, Stams AJM, Vries SS de, Weusthuis RA (1999)
ogenic to solventogenic phase in ABE fermentation and Utilisation of biomass for the supply of energy carriers. Appl
thus the success of ABE fermentation. Microbiol Biotech 52: 741±755
DOW represents an interesting substrate for ABE DuÈrre P (1998) New insights and novel developments in Clostridial
acetone/butanol/isopropanol fermentation. Appl Microbiol
fermentation due to its low price and a sugar composi- Biotechnol 49: 639±648
tion that can sustain growth and solvent production by Excoer G, Toussaint B, Vignon MR (1991) Sacchari®cation of
Clostridia if an ecient hydrolysis is performed. The steam-exploded poplar wood. Biotech Bioeng 38: 1308±1317
inability of solventogenic Clostridia to utilise cellulose Faaij A, Doorn Jv, Curvers T, Waldheim L, Olsson E, Wijk Av,
makes the hydrolysis of lignocellulosic substrates a Daey-Ouwens C (1997) Characteristics and aviability of bio-
mass waste and residues in The Netherlands for gasi®cation.
necessary step prior to fermentation. However, the hy- Biomass Bioenergy 12: 225±240
drolysis, either chemical or enzymatic, results in an in- Gelhaye E, Petitdemange H, Gay R (1993) Adhesion and growth
crease in the costs of the process. Some solventogenic rate of Clostridium cellulolyticum ATCC 35319 on crystalline
Clostridial strains, including C. acetobutylicum ATCC cellulose. J Bacteriol 175: 3452±3458
Genome Therapeutics (1999) Internet address: http://www.
824, show endoglucanase and b-glucosidase activities
(Allcock and Woods 1981; Lee et al. 1985; Kim et al. Girbal L, Soucaille P (1998) Regulation of solvent production in
1994). These activities are involved in cellulose utilisa- Clostridium acetobutylicum. Trends Biotechnol 16: 11±16
tion, but the exo-glucanase activity necessary for cellu- Gosselink RJA, Dam JEG van, Zomers FHA (1995) Combined
HPLC analysis of organic acids and furans formed during or-
lose degradation seems to be lacking. The advances in ganosolv pulping of ®ber hemp. J Wood Chem Technol 15:
our knowledge of the genetics and physiology of sol- 1±25
ventogenic Clostridia, which include the sequencing of Gregg DJ, Saddler JN (1996) Factors a€ecting cellulose hydrolysis
the genome of the strain C. acetobutylicum ATCC 824 and the potential of enzyme recycle to enhance the eciency
by Genome Therapeutics (1999), would allow the con- of an integrated wood to ethanol process. Biotech Bioeng 51:
struction of strains with a wider substrate range (i.e. able Jones DT, Woods DR (1986) Acetone-Butanol fermentation re-
to degrade cellulose) and thus make ABE fermentation visited. Microbiol Rev 50: 484±524
an economically viable process again. Kim AY, Attwood GT, Holt SM, White BA, Blaschek HP (1994)
Heterologous expression of endo- b-1,4-D glucanase from
Acknowledgements The authors wish to thank J.C. van der Putten Clostridium cellulovorans in Clostridium acetobutylicum ATCC
and R.J.A. Gosselink for their help during the DOW analysis and 824 following transformation of the engB gene. Appl Environ
furfural determination, J. Donkers for help in SEM analyses, Microbiol 60: 337±340
and M.A.W. Budde for technical assistance. The work described in Larsson S, Palmqvist E, Hahn Hagerdal B, Tengborg C, Stenberg
this paper was supported by an EU Madam Curie Fellowship K, Zacchi G, Nilvebrant NO (1999) The generation of fer-
(Grant number FAIR-CT96-5047) to A.M. LoÂpez-Contreras. The mentation inhibitors during dilute acid hydrolysis of softwood.
authors guarantee that the experiments described in this manu- Enzyme Microb Technol 24: 151±159
script comply with the current laws of The Netherlands, where they Lee SF, Forsberg CW, Gibbins LN (1985) Cellulolytic activity of
where carried out. Clostridium acetobutylicum. Appl Environ Microbiol 50: 220±
Maddox IS, Murray AE (1983) Production of n-butanol by fer-
mentation of wood hydrolysate. Biotechnol Lett 5: 175±178
References Mitchell WJ (1998) Physiology of carbohydrate to solvent con-
version by Clostridia. Adv Microb Physiol 39: 31±130
Allcock ER, Woods DR (1981) Carboxymethyl cellulase and cell- Nimcevic D, Schuster M, Gapes JR (1998) Solvent production by
obiase production by Clostridium acetobutylicum in an indus- Clostridium beijerinckii NRRL B592 growing on di€erent po-
trial fermentation medium. Appl Environ Microbiol 41: tato media. Appl Microbiol Biotechnol 50: 426±428
539±541 Taherzadeh MJ, Niklasson C, LideÂn G (1999) Conversion of di-
Blumenkrantz N, Asboe Hansen G (1973) New method for quanti- lute-acid hydrolyzates of spruce and birch to ethanol by fed-
tative determination of uronic acids. Anal Biochem 54: 484±489 batch fermentation. Bioresour Technol 69: 59±66
Breuil C, Chan M, Gilbert M, Saddler JN (1992) In¯uence of b- TAPPI (1991) T UM 250. Acid-soluble lignin in wood and pulp.
glucosidase on the ®lter paper activity and hydrolysis of Useful methods. TAPPI Press, Atlanta
lignocellulosic substrates. Bioresour Technol 39: 139±142 TAPPI (1999a) T 264 om-88. Preparation of wood for chemical
Claassen PAM, Budde MAW, Buitelaar RM, Tan GBN (1998) analysis. Test methods 1998±1999. TAPPI Press, Atlanta
Production of acetone butanol and ethanol (ABE) from agri- TAPPI (1999b) T 211 om-93. Ash in wood, pulp, paper and pap-
cultural residues or Domestic Organic Waste (DOW) and long- erboard: combustion at 575 °C. Test methods 1998±1999.
term fermentation on glucose. In: Kopetz H, Weber T, Palz W, TAPPI Press, Atlanta