Ethanol production in a bioreactor with an integrated membrane distillation module

Article abstract of DOI:10.2478/s11696-011-0088-0

Astract: Ethanol production in a bioreactor with integrated membrane distillation (MD) module has been investigated. A hydrophobic capillary polypropylene membrane (Accurel PP V8/2 HF), with an external/internal diameter ratio, d _out/d _in = 8.6 mm/5.5 mm and pore size 0.2 μm, was used in these studies. The products (mainly ethanol and acetic acid) formed during the fermentation of sugar with Saccharomyces cerevisiae inhibited the process. These products were selectively removed from the fermentation broth by the MD process, which increased the efficiency of the conversion of sugar to alcohol from 0.45 g to 0.5 g EtOH per g of fermented sucrose. The bioreactor efficiency also increased by almost 30 %. Separation of alcohol by the MD generates a higher yield of ethanol in the permeate than in the broth. The enrichment coefficient amounted to 4-8, and depended on the ethanol concentration in the broth. The separated solutions did not wet the membrane in use for 2500 h of the MD experiments and the retention of inorganic solutes was close to 100 %.

Full text of DOI:10.2478/s11696-011-0088-0

Chemical Papers 66 (2) 85–91 (2012)
DOI: 10.2478/s11696-011-0088-0
ORIGINAL PAPER
Ethanol production in a bioreactor with an integrated membrane
‡
distillation module
Marta Barancewicz, Marek Gryta*
West Pomeranian University of Technology, Szczecin, Institute of Chemical and Environmental Engineering,
ul. Pu ꢀl askiego 10, 70-322 Szczecin, Poland
Received 20 May 2011; Revised 7 July 2011; Accepted 12 July 2011
Ethanol production in a bioreactor with integrated membrane distillation (MD) module has been
investigated. A hydrophobic capillary polypropylene membrane (Accurel PP V8/2 HF), with an
external/internal diameter ratio, dout/din = 8.6 mm/5.5 mm and pore size 0.2 µm, was used in
these studies. The products (mainly ethanol and acetic acid) formed during the fermentation of
sugar with Saccharomyces cerevisiae inhibited the process. These products were selectively removed
from the fermentation broth by the MD process, which increased the efficiency of the conversion of
sugar to alcohol from 0.45 g to 0.5 g EtOH per g of fermented sucrose. The bioreactor efficiency
also increased by almost 30 %. Separation of alcohol by the MD generates a higher yield of ethanol
in the permeate than in the broth. The enrichment coefficient amounted to 4–8, and depended on
the ethanol concentration in the broth. The separated solutions did not wet the membrane in use
for 2500 h of the MD experiments and the retention of inorganic solutes was close to 100 %.
ꢀ
c 2011 Institute of Chemistry, Slovak Academy of Sciences
Keywords: bioreactor, membrane distillation, bioethanol
Introduction
tillation column. Several methods have been proposed
to minimise the product inhibition by the continu-
ous removal of ethanol from the fermentation broth.
These new technologies are frequently both more envi-
ronmentally friendly and cheaper than the traditional
ones (Bai et al., 2008; Cardona & Sanchéz, 2007; Kar-
gupta et al., 1998).
Continuous fermentation is more attractive than
the batch process due to its higher productivity, better
process control and improved yields (Choi et al., 2009;
Kargupta et al., 1998; Park et al., 1999). The chal-
lenge is how to effectively remove the yeast-produced
metabolites from the broth. A selective separation of
ethanol only has the effect of increasing the concentra-
tion of other metabolites and substrates that have not
been used. It creates an unfavourable reaction envi-
ronment for yeast and, as a result, suddenly decreases
Ethanol is regarded as a source of energy and valu-
able fuel, and it is also an important material in
the food and chemical industries (Demirbas, 2007).
Ethanol fermentation from renewable feedstock using
Saccharomyces cerevisiae is a product-inhibited pro-
cess. The inhibition begins when alcohol concentration
−
3
exceeds 30 g dm and it is very significant for ethanol
−
3
concentration over 80 g dm (Bai et al., 2008). There-
fore, fermentation is traditionally conducted in batch
reactors, resulting in a solution containing 5–12 % of
ethanol. Due to bioethanol’s low concentration, its re-
covery from the culture broth by distillation is seen
to be a cost-intensive operation. Moreover, a major
problem in commercial ethanol production is the ex-
cessive amount of waste water discharged from the dis-
*
Corresponding author, e-mail: marek.gryta@zut.edu.pl
Presented at 38th International Conference of the Slovak Society of Chemical Engineering, Tatranské Matliare, 23–27
‡
May 2011.

8
6
M. Barancewicz, M. Gryta/Chemical Papers 66 (2) 85–91 (2012)
the number of yeast cells and fermentation efficiency
after a few dozen hours (Gyamerah & Glover, 1996;
Maiorella et al., 1984). In order to decrease the con-
centration of metabolites, a proportion of the broth
is removed from the bioreactor (Escobar et al., 2001;
Kaseno et al., 1998).
Several investigations have shown that the appli-
cation of membrane processes reduced many techno-
logical problems, such as the negative influence of fer-
mentation conditions on yeast, increased productivity,
and reduced production costs (Escobar et al., 2001;
Kargupta et al., 1998; Maiorella et al., 1984). Mem-
brane distillation can be successfully applied to effect
ethanol removal from the fermenting broth (Gryta et
al., 2000; Gryta, 2001; Kaseno et al., 1998).
The MD is an evaporation process of volatile feed
components through air-filled pores of a hydrophobic
membrane (non-wetted). An MD membrane separates
two aqueous solutions differing in temperature and
composition. The driving force for the mass trans-
fer through the membrane pores is the difference in
vapour pressure on both sides of the membrane, which
depends on the temperature and the solution com-
position in the layers adjacent to the membrane. By
distilling a water–ethanol solution by the MD, a flow
of both ethanol and water vapour through the mem-
brane is achieved. However, at a given temperature,
the volatility of ethanol is higher so the distillate ob-
tained would be enriched in ethanol. For diluted solu-
tions of alcohol (up to 10 %), it is possible to obtain
a distillate with a concentration of ethanol 3–8 times
higher than that in the feed (Gryta, 2001; Kaseno et
al., 1998).
The components of a fermentation broth can
quickly foul the surface of membranes (fouling). Foul-
ing is one of the main reasons limiting the industrial
application of membrane processes. In the case of the
MD, the presence of alcohol poses an additional threat
as it may accelerate the membrane wettability. Dur-
ing separation of the broth by the MD process, a
significant decrease in membrane module productiv-
ity was observed due to increased membrane wetta-
bility (Udriot et al., 1989). However, results obtained
in other studies did not confirm any such rapid foul-
ing of membranes caused by broth during the MD
Experimental
The fermentation process was carried out in the
laboratory installation described previously (Gryta et
al., 2000). The installation was made from elements
and materials produced for constructing industrial in-
3
stallations. The reactor vessel (5 dm ), the valves, and
the pipeline were made from acid-resistant steel (ASI
316L). The content of the bioreactor was stirred us-
ing an external circulation system in which the liquid
was pumped by a GN-G35JF59ETZ pump with an
impeller (Micropump, USA). A water cooler and an
electric heater were mounted in the circulation sys-
tem, controlled by a RE26 regulator (Lumel, Poland).
The regulation systems used in the installation sta-
3
−1
bilised both the fluid flow-rate (± 5 dm h ) and its
temperature (± 1.5 K).
For the fermentation studies with the MD, the
membrane module was connected to the external cir-
culation system of the bioreactor. A polypropylene
capillary membrane (Accurel PP V8/2 HF, Membrana
GmbH, Germany) was assembled inside the MD mod-
ule. The length and external and internal diameters
of the capillary amounted to 215 mm, 8.6 mm, and
5.5 mm, respectively. The membrane had pore sizes
with a nominal and maximum diameter of 0.2 µm
and 0.6 µm, respectively, and porosity of 73 %. The
2
effective membrane area amounted to 0.037 m . The
broth and distillate streams flowed co-currently from
the bottom to the upper part of the MD module. The
broth flowed alongside the membrane surface (inside
the capillary bore), whereas the distillate flowed on
the shell side. The inlet temperatures of the streams
of the broth and distillate were kept at 310 K and
293 K, respectively.
The fermentation solution was prepared by dissolv-
ing 100 g of sucrose into thrice boiled tap water. The
tap water used had the following average ionic compo-
−
3
+
+
2+
2+
sition AIC/(mg dm ): K 7, Na 29, Mg 18, Ca
−
2−
−
−
65, Br 0.15, SO4 100, NO3 1.3, Cl 55 (ion chro-
matographic analysis). The tap water was considered
to be an adequate source of mineral salts for microor-
ganisms for the 100 h period of fermentation (Gryta,
2001). A commercially available Gamma HEFE yeast
(Saccharomyces cerevisiae, AB Enzymes, Germany)
was used as the microorganism in the amount of 5
(Gryta, 2001; Kaseno et al., 1998). In addition to the
−
3
fermentation conditions, the results obtained in the
studies were most probably influenced by the type of
membranes used. Therefore, further research is needed
into the resistance of membranes used for fermentation
broth separation in the MD process.
The study presents the results of long-term investi-
gations into fermentation linked with the MD process.
The performance of hydrophobic polypropylene capil-
lary membranes was evaluated when volatile metabo-
lites were released from the bioreactor. The influence
of such broth components as sugar, yeast, and alcohol
on the magnitude of ethanol flux was determined.
g dm . The dry yeast was rehydrated for 30 min,
while the broth was agitated periodically. Next, the
broth (2 dm ) was poured into the MD bioreactor. The
fermentation process was carried out in the continuous
mode for 5 days. After each series of experiments, the
installation was rinsed with water several times and
then with a solution of NaOH (1 mass %), the residue
of which was removed by rinsing the installation 2–3
times with distilled water.
3
Samples of the broth (15 mL) were collected ev-
ery 24 h. A sample was first centrifuged (2000 min
centrifuge MPW-350R, Med-Instruments), and then
−
1
,

M. Barancewicz, M. Gryta/Chemical Papers 66 (2) 85–91 (2012)
87
1
00
50
40
30
20
10
0
filtered (membrane filter 0.45 µm, Millipore). The con-
tent of sugar and alcohol in the solution was deter-
mined on the basis of an analysis of the total organic
carbon (TOC-Analyser multi N/C, Analytic Jena).
For this analysis, a filtered sample was divided into
two portions. One portion was allowed to evaporate
to obtain an approximately 50 % reduction in the
sample volume and later diluted in the same manner
as the sample that was not evaporated. The ethanol
concentration in the broth was calculated based on
the TOC difference between the evaporated and non-
evaporated samples. This method assumes that the
main components of the broth are ethanol and sucrose.
For this assumption, as was shown earlier on the basis
of chromatographic analyses, the determination error
of ethanol and sucrose concentration did not exceed
8
0
0
0
0
6
4
2
c
c
0
0
20
40
t/h
60
80
Fig. 1. Variations in sugar and ethanol concentrations in the
broth (cS, cE) during the fermentation process. Aver-
age value of four fermentation series in steel bioreactor:
sugar ( ) and ethanol (ꢀ); three fermentation series in
3
–5 % (Barancewicz et al., 2009).
glass reactor: ethanol (+), (×), (☼) and sugar (ꢁ),
The concentrations of anions and cations were
( ), ( ).
◦
measured using the ion chromatographic method with
conductivity detector (850 Professional IC, Metrohm
Herisau, Switzerland). The separation of anions was
achieved on 1.7 mm × 3.5 mm Metrosep RP guard
column in series with a 250 mm × 4.0 mm Metrohm
A Supp5-250 analytical column. A 150 mm × 4.0 mm
Metrosep C2-150 analytical column was used for sep-
aration of cations.
The electrical conductivity of the solutions exam-
ined was measured by a 6P Ultrameter (Myron L
Company, USA). The morphology and composition
of the fouling layer was studied using scanning elec-
tron microscopy (SEM) coupled with energy disper-
sion spectrometry (EDS).
100
50
40
30
20
10
0
80
60
40
c
c
2
0
0
0
20
40
60
80
100
t/h
Fig. 2. Variations in sugar and ethanol concentrations in the
broth (cS, cE) during the fermentation process with
MD: sugar ( ) and ethanol (ꢀ); without MD: sugar ( )
Results and discussion
Materials used in the construction of industrial in-
stallations are subject to corrosion and moderate dis-
solution, which may influence the rate of growth of
microorganisms and lead to the formation of a scaling
layer on the membrane surfaces (Gryta, 2007). The
laboratory installation was made from the same ma-
terials as were used in the design of large-scale in-
dustrial installations. Those elements that came into
contact with the broth (e.g. reactor vessel, pipelines)
were made from acid-resistant steel. This type of steel
contains heavy metals the presence of which can have
a negative influence on the yeast growth. For this rea-
son, at the first stage of the investigations, the pro-
cess of sugar fermentation in the bioreactor was re-
peated several times. The results of the experiments
in some series of measurements show discrepancies in
the course of the concentration changes in sugar and
ethanol. However, a similar discrepancy was obtained
for the fermentation conducted for comparison pur-
poses in a thermostated glass vessel (Fig. 1). There-
fore, a conclusion can be drawn that the ASI 316L
steel used for the installation construction did not af-
fect the course of fermentation, and the differences in
and ethanol ( ).
◦
the rate of yeast growth in the early stages of fermen-
tation (lag-phase) were probably responsible for the
concentration changes recorded.
Most of the fermentation experiments presented
here were conducted at a constant rate of rotation of
−
1
the pump impeller (approximately 700 min ). A ro-
tating impeller is capable of tearing yeast cells apart,
which reduces the concentration of active yeast. In
order to assess the intensity of the process, fermenta-
tion tests were carried out in which the speed of ro-
−
1
tation was increased up to 1600 min . However, the
results of ethanol and sugar concentrations thus ob-
tained were similar to those presented in Fig. 1, which
−
1
were obtained at 700 min , suggesting that even if
the rotating impeller destroyed some yeast cells, it had
no decisive influence on the results of the experiments
performed.
In the next stage of the fermentation experiments,
the evaporation of volatile compounds from the broth
in the MD process was used. In the initial stage of fer-

8
8
M. Barancewicz, M. Gryta/Chemical Papers 66 (2) 85–91 (2012)
0.55
30
500
2
2
1
5
0
5
4
3
2
1
0
00
00
00
00
0
.50
0.45
E
0
.40
10
5
c
N
0.35
0
0
10
20
30
40
–
50
60
0
.30
3
0
100
200
300
t /h
400
500
cB/(g dm )
O
Fig. 5. Batch fermentation with MD process: The influence of
ethanol concentration in the broth (cB) on the distil-
late flux (N ) ( ) and on the ethanol concentration in
Fig. 3. Efficiency (E, g of ethanol per g of sugar) of batch fer-
mentation (consecutive 400 h of bioreactor operation
time) with MD (ꢁ) and without MD ( ).
distillate (cD) ( ).
◦
6
0
creased when most of the sugar was fermented (Fig. 4).
The changes in ethanol concentration in the broth
had no significant influence on the rate of permeate
flux, but they did affect the concentration of the dis-
tillate obtained (Fig. 5). The amount of separated
ethanol increased when its concentration in the broth
also increased and, after several hours of fermentation,
it was possible to obtain a distillate containing more
50
40
30
c
20
3
than 200 g of ethanol per dm . For the results pre-
sented here, the enrichment factor obtained amounted
to 4–10. The enrichment of the distillate in ethanol
is an important benefit of the MD process because,
besides increasing the fermentation efficiency (separa-
tion of volatile metabolites from the broth), a concen-
trated solution of ethanol devoid of yeast and salts
is also obtained. In the case of dilute solutions, an
increase in distillate concentration, e.g. in the range
1
0
0
0
200 400 600 800 1000 1200 1400
/h
t
O
Fig. 4. Batch fermentation with MD process: The changes in
ethanol concentration in the broth (cE) during opera-
tion time (tO).
3
of 50–100 g of ethanol per dm requires the evapo-
3
3
ration of 500 dm of water from 1 m of the feeding
solution. Due to this, pre-concentration of the ethanol
solution by the MD process affords significant energy
saving during further alcohol pre-concentration by tra-
ditional distillation.
mentation with the MD, the rate of sugar to ethanol
conversion increased slightly and its concentration in
the broth decreased due to steady evaporation of
ethanol through the membranes (Fig. 2).
The calculations showed that application of the
MD process increased the efficiency of fermentation
from 0.45 g to 0.5 g of ethanol per gram of sucrose
In order to limit the amount of waste water gen-
erated during ethanol production, it is desirable to
increase the final concentration of alcohol through
the fermentation of concentrated solutions of sugar.
This, however, leads to a high inhibition of the prod-
uct formation. The problem may be resolved by us-
ing a membrane bioreactor with a continuous dosage
of sugar. The results obtained for continuous fermen-
tation linked with the MD process are presented in
Fig. 6. A continuous dosage of sugar solution into the
bioreactor allowed stabilisation of the alcohol concen-
(
Fig. 3).
In addition, productivity of the bioreactor with
the MD connected increased by almost 30 %. The re-
sults obtained are consistent with the earlier reports,
which highlighted the benefits of combining fermenta-
tion with the MD process (Gryta, 2001; Gryta et al.,
000).
The long-term experiments confirmed the consis-
tently high efficiency of the MD process used for
the separation of ethanol from the broth during the
batch fermentation. Due to the steady evaporation of
ethanol through the membranes, its concentration de-
2
−
3
tration at 40 g dm and, as a result, increase of the
3
−2 −1
d . During sugar
permeate flux up to 4.5 dm m
fermentation, ethanol and other by-products such as
acetaldehyde, formic acid, lactic acid, acetic acid, 1-
propanol, 2-methyl-1-butanol, and 2,3-butanediol are

M. Barancewicz, M. Gryta/Chemical Papers 66 (2) 85–91 (2012)
89
5
1
00
formed (Gryta, 2001). Most of these compounds are
volatile so they can be separated by the MD pro-
cess. However, the MD process does not separate non-
volatile metabolites and their concentration increases
steadily. Fig. 7 presents changes in the concentra-
tion of organic acids in the broth, the amounts of
which increased when both the fermentation time and
the amount of sugar used in the experiment were in-
creased. The data show that, where continuous fer-
mentation is concerned, it is necessary to limit the in-
crease in non-volatile metabolites, e.g. through a par-
tial broth removal (bleeding) in addition to the MD
process.
8
0
0
0
0
4
3
2
1
0
6
4
2
N
0
0
20
40
60
80
100
t/h
Fig. 6. Changes in broth composition (c) and permeate flux
N ) over the fermentation time (t): with continuous
The MD module assembled in the installation was
in operation, with just a few weeks’ suspension, for al-
most one year. The degree of wettability of the mem-
branes used in the study was assessed by periodic ex-
amination of changes in the maximum flux (distilled
(
sugar dosage: sugar concentration ( ), ethanol con-
centration (ꢁ), permeate flux (ꢀ); without continuous
sugar dosage: sugar concentration ( ), ethanol concen-
tration (
◦
), permeate flux (☼).
3
water with 10 g of NaCl per dm as a feed). The val-
ues measured of permeate flux obtained for water and
broth are presented in Fig. 8. The efficiency of the MD
module decreased significantly during the first 500 h
of its operation, which was also observed in the case of
other hydrophobic membranes (Gryta & Barancewicz,
2
500
100
80
60
40
20
0
2
000
2
010). The MD process efficiency was stable for the
1
500
subsequent 2000 h. Despite the presence of salt in the
−
1
feed (electrical conductivity of 5000 µS cm ), the
1
000
c
c
electrical conductivity of the distillate obtained did
−
1
not exceed 5 µS cm . Such a low value for the dis-
tillate’s electrical conductivity indicated that, despite
being in contact with water and fermentation broth
for almost a year, the membrane did not become wet-
ted.
5
00
0
0
20
40
t/h
60
80
The Accurel PP V8 HF membranes have a spongy
structure and their pore size is below 1 µm. The SEM
investigations revealed that the surface area of the
membrane walls also contains larger pores, several mi-
crometers in diameter (Fig. 9). These large pores can
become wetted, resulting in an increase in tempera-
ture polarisation. As a result, a decrease is observed
in the MD module efficiency.
Fig. 7. Changes in organic acids concentration (c) in the broth
and MD distillate over the fermentation time (t). Broth:
acetic acid ( ), succinic acid (ꢂ), formic acid (ꢀ); dis-
tillate: acetic acid ( ).
The decrease in efficiency observed in Fig. 8 may
also be associated with biofouling of the membrane.
However, the SEM investigations showed that, despite
long-term use of the MD module for broth separation,
no yeast cells were observed on the membrane sur-
faces (Fig. 9D). Therefore, it may be concluded that
biofouling will not occur in the MD modules applied to
ethanol separation from the fermentation broth. This
result is consistent with previous observations (Gryta,
2
001).
After drying the membrane which was extracted
from the module, gas permeability investigations were
conducted. The changes in the air permeability of the
membrane extracted from the bioreactor were simi-
lar to the changes in permeability observed in a new
membrane. This confirms the conclusion that a small
amount of precipitate visible on the membrane surface
Fig. 8. Changes in the permeate flux (N ) and distillate electri-
cal conductivity (κ) during the MD module operation
time (t): permeate flux ( ), electrical conductivity (ꢂ);
feed: distilled water with NaCl ( ), broth ( ).

9
0
M. Barancewicz, M. Gryta/Chemical Papers 66 (2) 85–91 (2012)
Fig. 9. SEM images of polypropylene capillary Accurel PP V8/2 HF membrane: internal surface (A), cross-section (B), external
surface (C), and internal surface after one year of operation in MD bioreactor (D).
3
000
500
000
500
000
Conclusions
2
The results confirmed that a combining the fer-
mentation with the MD process effectively separated
ethanol and other volatile metabolites from the broth.
As a result, efficiency in the sugar to ethanol con-
version was increased, as was the productivity of the
bioreactor. No intensive biofouling development was
detected and the polypropylene capillary membranes
used maintained their high efficiency of separation
throughout the period of the MD experiments.
No decline in permeability of the polypropylene
membranes used was observed over the course of
the fermentation experiments. This indicates that the
presence of cells in the broth did not lead to their
deposition on the membrane surface. The SEM obser-
vations of the membrane samples confirmed that the
surface of membranes was clean and practically free
of yeast cells.
2
1
1
J
5
00
0
0
200 400 600 800 1000 1200 1400
∆
P/Pa
Fig. 10. Influence of pressure difference (∆P) on air perme-
ability (J) of Accurel PP V8/2 HF membrane: new
membrane ( ) and membrane after one year of oper-
ation in MD bioreactor ( ).
◦
Acknowledgements. The Polish State Committee for Sci-
entific Research is acknowledged as providing support for this
work (2009–2012).
(
Fig. 9D) did not hinder its permeability (Fig. 10).
The result obtained also indicates that the decrease
observed in the MD process efficiency (Fig. 8) was
largely due to a certain proportion of the larger pores
on the membranes surface becoming wetted.
References
Bai, F. W., Anderson, W. A., & Moo-Young, M. (2008). Ethanol

M. Barancewicz, M. Gryta/Chemical Papers 66 (2) 85–91 (2012)
91
fermentation technologies from sugar and starch feedstocks.
Biotechnology Advances, 26, 89–105. DOI: 10.1016/j.biotech
adv.2007.09.002.
Barancewicz, M., Sasim, M., & Gryta, M. (2009). Zastosowanie
chromatografii jonowej do badania przebiegu fermentacji.
Prace i Studia, 77, 89–101.
Cardona, C. A., & Sanchéz, Ó. J. (2007). Fuel ethanol pro-
duction: Process design trends and integration opportunities.
Bioresource Technology, 98, 2415–2457. DOI: 10.1016/j.bior
tech.2007.01.002.
Choi, G.-W., Kang, H.-W., & Moon, S.-K. (2009). Repeated-
batch fermentation using flocculent hybrid, Saccharomyces
cerevisiae CHFY0321 for efficient production of bioethanol.
Applied Microbiology and Biotechnology, 84, 261–269. DOI:
Gryta, M., Morawski, A. W., & Tomaszewska, M. (2000).
Ethanol production in membrane distillation bioreactor.
Catalysis Today, 56, 159–165. DOI: 10.1016/S0920-5861(99)
00272-2.
Gyamerah, M., & Glover, J. (1996). Production of ethanol by
continuous fermentation and liquid–liquid extraction. Jour-
nal of Chemical Technology and Biotechnology, 66, 145–
152. DOI: 10.1002/(SICI)1097-4660(199606)66:2<145::AID-
JCTB484>3.0.CO;2-2.
Kargupta, K., Datta, S., & Sanyal, S. K. (1998). Analysis
of the performance of a continuous membrane bioreactor
with cell recycling during ethanol fermentation. Biochem-
ical Engineering Journal, 1, 31–37. DOI: 10.1016/S1369-
703X(97)00006-5.
1
0.1007/s00253-009-1946-3.
Kaseno, Miyazawa, I., & Kokugan, T. (1998). Effect of product
removal by a pervaporation on ethanol fermentation. Jour-
nal of Fermentation and Bioengineering, 86, 488–493. DOI:
10.1016/S0922-338X(98)80157-8.
Demirbas, A. (2007). Progress and recent trends in biofuels.
Progress in Energy and Combustion Science, 33, 1–18. DOI:
10.1016/j.pecs.2006.06.001.
Escobar, J. M., Rane, K. D., & Cheryan, M. (2001). Ethanol
production in a membrane bioreactor. Applied Biochemistry
and Biotechnology, 91–93, 283–296. DOI: 10.1385/ABAB:91-
Maiorella, B. L., Blanch, H. W., & Wilke, C. R. (1984). Eco-
nomic evaluation of alternative ethanol fermentation pro-
cesses. Biotechnology and Bioengineering, 26, 1003–1025.
DOI: 10.1002/bit.260260902.
93:1-9:283.
Gryta, M. (2007). Effect of iron oxides scaling on the MD pro-
cess performance. Desalination, 216, 88–102. DOI: 10.1016/
j.desal.2007.01.002.
Gryta, M. (2001). The fermentation process integrated with
membrane distillation. Separation and Purification Technol-
ogy, 24, 283–296. DOI: 10.1016/S1383-5866(01)00132-0.
Gryta, M., & Barancewicz, M. (2010). Influence of morphol-
ogy of PVDF capillary membranes on the performance of
direct contact membrane distillation. Journal of Membrane
Science, 358, 158–167. DOI: 10.1016/j.memsci.2010.04.044.
Park, B. G., Lee, W. G., Chang, Y. K., & Chang, H. N. (1999).
Long-term operation of continuous high cell density culture
of Saccharomyces cerevisiae with membrane filtration and
on-line cell concentration monitoring. Bioprocess Engineer-
ing, 21, 97–100. DOI: 10.1007/PL00009070.
Udriot, H., Ampuero, S., Marison, I. W., & von Stockar,
U. (1989). Extractive fermentation of ethanol using mem-
brane distillation. Biotechnology Letters, 11, 509–514. DOI:
10.1007/BF01026651.