376 S. POPOV et al.: Bioethanol from Sugar Beet Juice, Food Technol. Biotechnol.

48 (3) 376–383 (2010)

ISSN 1330-9862 original scientific paper

(FTB-2433)

Bioethanol Production from Raw Juice as Intermediate of

Sugar Beet Processing: A Response Surface Methodology
Approach

Stevan Popov, Jovana Rankovi}*, Jelena Dodi}, Sini{a Dodi} and Aleksandar Joki}
University of Novi Sad, Faculty of Technology, Department of Biotechnology and Pharmaceutical
Engineering, Bulevar cara Lazara 1, RS-21000 Novi Sad, Serbia
Received: February 1, 2010
Accepted: May 4, 2010

Summary
Response surface methodology (RSM) was used for selecting optimal fermentation time
and initial sugar mass fraction in cultivation media based on raw juice from sugar beet in
order to produce ethanol. Optimal fermentation time and initial sugar mass fraction for
ethanol production in batch fermentation by free Saccharomyces cerevisiae cells under anaer-
obic conditions at the temperature of 30 °C and agitation rate of 200 rpm were estimated
to be 38 h and 12.30 % by mass, respectively. For selecting optimal conditions for indus-
trial application, further techno-economic analysis should be performed by using the ob-
tained mathematical representation of the process (second degree polynomial model). The
overall fermentation productivity of five different types of yeast was examined and there
is no significant statistical difference between them.

Key words: optimization, bioethanol, raw juice, Saccharomyces cerevisiae, response surface
methodology

Introduction simultaneous production of ethanol in refineries built as

In recent years, a new round of enthusiasm in bio- their extensions. Based on an extensive overview of ap-
mass and bioenergy has been initiated with the recog- proaching bioethanol production development in Voj-
nition that the global crude oil reserves are limited, and vodina, Serbia, it can be concluded that this region has a
their depletion is occurring much faster than previously large potential for renewable energy, especially energy
predicted. In addition, the environmental deterioration from biomass – biodiesel and bioethanol (2-4). Molasses
resulting from the overconsumption of petroleum-de- is commonly used feedstock for bioethanol production.
rived products, especially transportation fuels, is threat- In sugar beet processing, it is a by-product obtained at
ening the sustainability of human society. Ethanol, both the end of the process and the cost of its production is
renewable and environmentally friendly, is believed to considerably higher than the cost of raw juice produced
be one of the best alternatives, leading to a dramatic in- at the beginning by water extraction of sliced sugar beet.
crease in its production capacity. Currently, the global Raw juice contains about 15–20 % of dry solids. Its
ethanol supply is obtained mainly from sugar and starch purity ranges between 85 and 90 %, which means that
feedstock (1). Less expensive production of sugar from there are about 85–90 % of sugars and 10–15 % of non-
sugarcane indicates that application of sugar beet for bio- sugars in dry matter (5). The obtained raw juice can be
ethanol production has great potential. Also, increased used either directly for ethanol and sugar production dur-
yield and sugar production efficiency have led to a re- ing the sugar beet harvest season, or it can be concen-
duction in sugar beet requirements. For these reasons, trated in an evaporator and stored for several months
the majority of existing sugar plants in Europe started (6).

*Corresponding author; Fax: ++381 21 450 413; E-mail: [email protected]

S. POPOV et al.: Bioethanol from Sugar Beet Juice, Food Technol. Biotechnol. 48 (3) 376–383 (2010) 377

In statistics-based approaches, response surface meth- ing the suspension, at the temperature of 30 °C and
odology (RSM) has been extensively used in fermenta- agitation rate of 200 rpm. The samples were taken at
tion medium optimization. RSM is a group of statistical predetermined time intervals for analysis.
techniques for designing experiments, building models,
evaluating the effects of factors and searching for the op- Measurement techniques
timum conditions. In RSM, the experimental responses
to the design of experiments are fitted to quadratic func- During the fermentation, samples for analysis were
tion. The number of successful applications of RSM sug- taken at the beginning and at the predetermined time
gests that the second-order relation can reasonably ap- intervals: 4, 8, 12, 24, 30, 36 and 48 h from the moment
proximate several of the fermentation systems. of inoculation. Cell number in fermentation medium was
The aims of this study are to describe the number of determined with Neubauer Haemocytometer under 400´
yeast cells, fermentable sugar consumption and ethanol magnification using an optical microscope (Wild M20,
production employing mathematical relationships as well Heerbrugg, Gais, Switzerland). Viable cells were count-
as to find optimal fermentation time and initial sugar ed with the methylene blue staining technique (7). The
concentration for bioethanol production from raw juice free amino nitrogen content was determined by ion-se-
in batch fermentation by free Saccharomyces cerevisiae cells. lective electrode. The pH was measured directly in the
Also, the overall fermentation efficiency of different types cultivation medium by the laboratory multiparameter ana-
of yeast (extensively used as starter cultures in various lyzer Consort C863 (Consort, Turnhout, Belgium) with
branches of the fermentation industry) are compared. the glass electrode. Total dissolved salt (TDS) content and
conductivity were determined using conductivity elec-
trode. The amount of soluble ash was calculated from
Materials and Methods the following formula (8):
Microorganisms and inoculum preparation w(ash)=(0.0018[a–(b´0.9)]´20)/% dm /1/
Five different commercial types of Saccharomyces ce- where a is the conductivity of the sample (mS/cm) and b
revisiae were used throughout this research: dried distill- the conductivity of distilled water (mS/cm). Dry matter
er’s yeast – DD (Lallemand Inc, Rexdale, Ontario, Ca- (dm) was determined by the standard drying method in
nada), dried wine yeast – DW1 (Lallemand Inc, Rexdale, an oven at 105 °C to a constant mass (8).
Ontario, Canada), dried wine yeast tolerant of high etha-
nol concentration – DW2 (Lallemand Inc, Rexdale, On- The samples of the substrates and cultivation media
tario, Canada), dried baker’s yeast – DB (Alltech, Senta, were centrifuged at 4000 rpm for 15 min. Then sucrose
Serbia) and fresh baker’s yeast – FB (Alltech, Senta, Ser- and reducing sugar content (sum of glucose and fruc-
bia). In order to rehydrate dried yeasts and metabolical- tose) of the supernatant were determined (Jasco, Inc, Eas-
ly acclimatize the cells prior to fermentation, yeasts were ton, MD, USA, pump PU-980, detector RI-930, sampler
suspended in a small quantity of culture medium under AS-950, 20 mL injection loop, column sugar KS-801, elu-
aerobic conditions for 2 h (temperature of 30 °C, agita- ent: water at flow rate of 0.6 mL/min and elution time
tion rate of 200 rpm) and then introduced to the rest of 30 min). Fermentable sugar content was expressed as the
the culture medium. On the other hand, fresh baker’s sum of sucrose and reducing sugars. Sucrose, glucose,
yeast was also suspended in a small quantity of culture and fructose standards were purchased from Supelco (Bel-
medium and immediately used for inoculation. lefonte, PA, USA). All chemicals were of reagent grade
or better.
Substrates Ethanol was determined directly from the samples
of the fermentation mash by gas chromatography, using
Raw juices were obtained from sugar factories in Cr-
a HP 5890 Series II GC (Agilent Technologies Inc, Santa
venka (SF1) and Senta (SF2), Serbia, during the harvest
Clara, CA, USA) equipped with a flame ionization de-
season 2007/2008. The sugar beet was harvested in Voj-
tector, a Carbowax 20 M column at 85 °C and the carrier
vodina. Substrates were kept at –18 °C until use. Raw
gas was helium. Injector and detector temperature was
juices were diluted with distilled water to give a total
maintained at 150 °C.
sugar mass fractions of 5 and 10 % and also used without
dilution with a total sugar mass fraction of approx. 13 %,
obtained from sugar beet processing technology. The sub- Statistical analyses
strates were adjusted to pH=5.0 with 10 % sulphuric acid All the data presented are the means of three exper-
(by volume). iments repeated with substrates from two sugar facto-
ries. The results were statistically tested by analysis of
Fermentation conditions variance at the significance level of p=0.05. The adequa-
Fermentations were carried out in a 2-litre bench- cy of the model was evaluated by coefficient of determi-
-scale bioreactor with fermentation medium of 1.5 L. The nation (R2) and model p-value. Response surface meth-
bench-scale bioreactor with the substrate was sterilized odology is a useful model for studying the effect of
by autoclaving at 121 °C and pressure of 2.2 bar for 30 several factors influencing the responses by varying
min. The sterile medium was inoculated to give the ini- them simultaneously and carrying out limited number
tial yeast cell concentration of 108 (cells/mL) (approx. 3 of experiments. For the description of the responses Y
g of yeast dry solids per 1000 mL of medium). The ferm- (cell number (CFU), ethanol content (% by volume) and
entation was carried out in batch mode under anaerobic fermentable sugar uptake (% by mass), a second-degree
conditions for 48 h, including the time needed for mak- polynomial model was fitted to the data:
378 S. POPOV et al.: Bioethanol from Sugar Beet Juice, Food Technol. Biotechnol. 48 (3) 376–383 (2010)

Y=b0+SbiXi+Sbii2Xii2+SbijXiXj /2/
where b0 represents intercept, bi represents the linear, bii
quadratic and bij interaction effect of the factors. The
factor variables and their ranges are: X1 fermentation
time (0–48 h) and X2 initial concentration of fermentable
sugars in cultivation media based on raw juice (5, 10
and 13 % by mass).
Statistical analyses were performed using Statistica
software v. 9.1. (9). Response surface plots were generat-
ed with the same software and drawn by using the
function of two factors.

Results and Discussion

During this experiment three independent cultiva-
tions were carried out with raw juices from two differ-
ent domestic sugar factories as fermentation medium.
Results represent average values obtained during those
experiments. Analyses of raw materials show that the
compositions of raw juices were typical of such inter-
mediate products in domestic sugar factories (Table 1).
The observed differences are the consequences of the
technological procedure used for sugar beet processing.
Given intermediate products with their compositions
should be considered as convenient raw materials for
the preparation of the cultivation media for bioethanol
manufacturing process.

Table 1. Composition of substrates

SF1 SF2
Raw juice content
w/%
sucrose 12.85 13.81
reducing sugars 0.07 0.07
fermentable sugars 12.92 13.88
dry matter 14.70 15.80
Fig. 1. Ethanol fermentation of culture media based on raw juice
free amino nitrogen 0.13 0.11 with initial fermentable sugar mass fractions of: a) 5, b) 10 and c)
ash content 0.28 0.47 13 %

Course of fermentation of culture media based on raw these experiments, fresh baker’s yeast (FB) was used as
juice a production microorganism; this type of yeast is applied
in almost all distilleries in our region.
The course of fermentation of culture media based
The main metabolic pathway involved in the etha-
on raw juice under the applied experimental conditions
nol fermentation is glycolysis (Embden-Meyerhof-Parnas
had to be determined first. This was done by examina-
or EMP pathway), through which one molecule of glu-
tion of cell number, fermentable sugar, free amino nitro-
cose is metabolized and two molecules of pyruvate are
gen, TDS, pH value and ethanol content during fermen-
produced. Under anaerobic conditions, the pyruvate is
tation of the media based on raw juice with selected
further reduced to ethanol with the release of CO2.
initial fermentable sugar mass fractions. These factors
Theoretically, the yield is 0.511 for ethanol and 0.489 for
are regularly used in observing the biotechnological pro-
CO2 on a mass basis of metabolized glucose. Two ATPs
cess. The results of this procedure can be used for deter-
produced in the glycolysis are used to drive the biosyn-
mination of significant factors that may be included in
thesis of yeast cells, which involves a variety of energy-
modelling of selected responses.
-requiring bioreactions. Therefore, ethanol production is
Figs. 1a-c give the dependence of cell number, fer- tightly coupled with yeast cell growth, which means that
mentable sugar, free amino nitrogen, TDS, pH value and yeast must be produced as a co-product. Without the
ethanol content during the fermentation of media based continuous consumption of ATPs in the growth of yeast
on raw juice with initial fermentable sugar mass frac- cells, the glycolytic metabolism of glucose will be inter-
tions of 5, 10 and approx. 13 %, respectively. During rupted immediately, because of the intracellular accu-
S. POPOV et al.: Bioethanol from Sugar Beet Juice, Food Technol. Biotechnol. 48 (3) 376–383 (2010) 379

mulation of ATP, which inhibits phosphofructokinase During further fermentation time, ethanol volume frac-
(PFK), one of the most important regulating enzymes in tions were not significantly changed (p=0.1401). The ob-
the glycolysis (1,10). tained ethanol content at the end of fermentation of me-
From Fig. 1 it can be seen that during the first 12 h dia based on raw juice with initial fermentable sugar mass
of fermentation yeast cell number increases almost in li- fractions of 5, 10 and 13 % was 3.64, 6.75 and 7.88 % (by
near manner in all applied culture media. The exponen- volume), respectively. These results indicate that during
tial phase of yeast cell growth was affected by the resi- raw juice fermentation, the obtained ethanol content is
dual oxygen content in fermenting mash. After 24 h of similar to the content obtained during thick juice and
fermentation, yeast count reached about 1.78·10 8 , molasses fermentation (11). This fact implies that lower
2.27·108 and 2.01·108 cells/mL in culture media based on purity of raw juice does not have a negative impact on
initial sugar mass fractions of 5, 10 and approx. 13 %, ethanol production.
respectively. With further prolongation of fermentation pH value and TDS content of all applied culture me-
time, the increase of yeast cell number was insignificant dia were slowly decreased during the first 12 h of fer-
(p=0.6585). These results are in good correlation with lit- mentation and afterwards remained almost unchanged.
erature data shown for sugar beet thick juice and mo- During growth, it is important for the yeast to preserve
lasses (11). a constant intracellular pH. There are many enzymes
Results shown in Fig. 1 indicate that fermentable sug- functioning within the yeast cell during growth and
ar content significantly decreased during the fermen- metabolism. Each enzyme works best at its optimal pH,
tation (p=0.0169), coinciding with an increase in biomass which is acidic because of the acidophilic nature of the
and ethanol production. The residual sugar must be con- yeast itself. When the extracellular pH deviates from the
trolled at a very low level. For example, no more than 2 optimal level, the yeast cells need to invest the energy to
and 5 g/L are controlled for residual reducing sugar and either pump in or pump out hydrogen ions in order to
total sugar, respectively, in the ethanol production from maintain the optimal intracellular pH. If the extracellu-
starch materials. Any ethanol fermentation research, lar pH deviates too much from the optimal range, it may
which is expected to be practical, needs to fulfill these become too difficult for the cell to maintain constant in-
criteria (1). After 12 h of fermentation of culture media tracellular pH, and the enzymes may not function nor-
with initial sugar mass fraction of 5 %, the remaining mally. If the enzymes are deactivated, the yeast cell will
sugar content is negligible. For media with initial sugar not be able to grow and produce ethanol efficiently (13).
mass fraction of 10 and approx. 13 %, residual ferment- According to the results in Fig. 1, pH values remained
able sugar content is minimal (approx. 0.5 %) after 24 in the optimal range for yeast cell activities during total
and 30 h of fermentation, respectively. The presented da- fermentation time of culture media based on raw juice
ta are in good correlation with literature data for sugar with initial fermentable sugar mass fractions of 5, 10 and
beet thick juice and molasses (11). This fact implies that 13 %. It can be concluded that under the experimental
the cost-effectiveness of longer duration of process conditions applied during this research, correction of this
should be questioned from the technological and econo- parameter at some stage of cultivation is not necessary.
mical point of view.
During fermentation, yeasts assimilate simple com-
Statistical analyses of the modelled responses
pounds, like amino acids and ammonium solids, and
generate complexes. According to the obtained results The results of the statistical analyses are presented
(Fig. 1), during the first 12 h of fermentation nitrogen in Table 2. The coefficients are related to actual variables.
concentration significantly decreased in all fermenting The ANOVA results for modelled responses are reported
mashes (p=0.0218). During the exponential phase of yeast in Table 3. Relatively high values of R2, obtained for all
cell growth, assimilated nitrogen was incorporated into responses, indicate good fit of the experimental data to
biomass. With further prolongation of the fermentation Eq. 2. All polynomial models tested for the selected re-
time, the nitrogen content slowly decreased for all ap- sponses were significant at 95 % confidence level (p=
plied mashes and this decrease was not as significant as 0.05, Table 3). The model F-values of 46.62, 90.30 and
in the first 12 h. Total decrease of nitrogen content dur- 99.39 for cell number, ethanol content and fermentable
ing 48 h of fermentation in all applied mashes was 0.04– sugar uptake, respectively, imply that models for select-
0.07 % (by mass). ed responses are significant. Fig. 2 shows the parity plot
Ethanol production is closely related to the growth of the observed and predicted values for modelled re-
of yeast cells. The ethanol fermentation rate of non-grow- sponses.
ing yeast cells is at least 30-fold slower than that of the
growing ones (12). According to the obtained results
Mathematical model of raw juice fermentation
(Fig. 1), ethanol content in culture media during the first
12 h of fermentation of mashes with all applied initial The goodness of fit of the model was checked by
sugar mass fractions significantly increased. During the determination coefficient that was found to be 0.928
further fermentation up to 48 h, ethanol content in mash for the response of cell number, which indicates that less
with sugar mass fraction of 5 % was slightly changed, than 10 % of the variations could not be explained by
corresponding to minor remaining sugar content and the model. As for significance of the polynomial coef-
stagnation of yeast cell growth. With initial fermentable ficients, their p-values suggest that the most important
sugar mass fractions of 10 and 13 %, intensive increase factor influencing cell number is initial mass fraction of
of ethanol volume fraction was prolonged until 30 h. fermentable sugars. The mutual interaction between ini-
380 S. POPOV et al.: Bioethanol from Sugar Beet Juice, Food Technol. Biotechnol. 48 (3) 376–383 (2010)

Table 2. Regression equation coefficients for selected responses

Cell number Ethanol volume fraction Fermentable sugars

Effects
Coefficient p-value Coefficient p-value Coefficient p-value
intercept
b0 0.299 0.23992 –2.030 0.1324 1.142 0.4775
linear
b1 0.270 0.0002 0.722 0.0261 0.426 0.3734
b2 0.031 0.0002 0.155 0.0010 –0.255 0.0001
quadratic
b11 –0.015 0.00017 –0.038 0.0341 0.036 0.1790
b22 –0.001 <0.0001 –0.004 <0.0001 0.007 <0.0001
interaction
b12 0.002 0.0018 0.013 <0.0001 –0.027 <0.0001

Table 3. Analysis of variance (ANOVA) of the modelled responses

Source
Response Residual Model F-value p-value R2
DF SS MS DF SS MS
cell number 18 0.22 0.012 5 2.89 0.58 46.62 <0.0001 0.928
ethanol volume fraction 18 5.74 0.32 5 143.96 28.79 90.30 <0.0001 0.962
fermentable sugars 18 14.20 0.79 5 391.95 78.39 99.39 <0.0001 0.965

DF – degree of freedom, SS – sum of squares, MS – mean squares

tial mass fraction of fermentable sugars and fermenta- and engineering (14). It combines multiple responses into
tion time is significant at the level 0.05, shown in Fig. 3. one response called desirability function by choice of val-
For the response of ethanol volume fraction, coeffi- ue from 0 (one or more characteristics are unacceptable)
cient of determination was found to be 0.962, which in- to 1 (all process characteristics are on target). Each of the
dicates that only 3.8 % of the variations could not be estimated responses is transformed to an individual de-
explained by the model. Fermentation time is less sig- sirability value ranging from 0 to 1. The value of indivi-
nificant compared to the initial mass fraction of ferment- dual desirability increases as the desirability of the cor-
able sugars for linear as well as quadratic effects. The responding response increases. The overall desirability
effects of initial mass fraction of fermentable sugars and of the process is computed as geometric mean of the in-
fermentation time on ethanol volume fraction are pre- dividual desirability functions (15).
sented in Fig. 4.
The results of optimization by desirability function
The goodness of fit of the model was checked by the
approach for maximization of cell number and ethanol
determination coefficient, which was found to be 0.965 for
volume fraction are given in Fig. 6. These conditions were
the response of fermentable sugar, indicating that less
selected taking into account that the ethanol fermen-
than 5 % of the variations could not be explained by the
tation rate of growing yeast cells is considerably faster
model. As for significance of the polynomial coefficients,
than that of the non-growing ones. As it can be seen from
their p-values suggest that the most important linear fac-
Fig. 6, the highest values of desirability function are ob-
tor is initial concentration of fermentable sugars and the
tained in the region of high initial sugar mass fractions
same conclusion can be applied to quadratic effects.
and long fermentation times. The optimal values of fer-
The effects of initial mass fraction of fermentable sug- mentation time and initial sugar mass fraction are 38 h
ars on sugar uptake during fermentation time are shown and 12.30 % (by mass), respectively, for the highest val-
in Fig. 5. ue of desirability function i.e. 1. Even so, the fermenta-
tion time needed to reach desirability around 0.9 varies
Optimization of raw juice fermentation from 30 to 48 h, which is in good agreement with the
The final goal of response surface methodology is industrial practice. On the other hand, initial sugar mass
the process optimization. Thus, the developed models can fraction for the same region of desirability function var-
be used for simulation and optimization. To optimize the ies in the region between 10 % for fermentation time of
process with two or more output responses, it is helpful 40 h and 13 % for fermentation time of around 30 h. To
to use the concept of desirability function. The desirabil- select optimal fermentation time and initial sugar mass
ity function is one of the most widely used methods for fraction for industrial application, further techno-eco-
optimization of multiple response processes in science nomic analysis should be done.
S. POPOV et al.: Bioethanol from Sugar Beet Juice, Food Technol. Biotechnol. 48 (3) 376–383 (2010) 381

Fig. 3. The effects of initial mass fraction (w0) of fermentable

sugars and fermentation time on the cell number

Fig. 4. The effects of initial mass fraction (w0) of fermentable

sugars and fermentation time on ethanol volume fraction

Fig. 2. Parity plot showing observed vs. predicted values for Fig. 5. The effects of initial mass fraction (w0) of fermentable
modelled responses sugars and fermentation time on fermentable sugar
382 S. POPOV et al.: Bioethanol from Sugar Beet Juice, Food Technol. Biotechnol. 48 (3) 376–383 (2010)

Fig. 6. The overall desirability function of the fermentation process

Productivity of different yeast types strains. In the first 4 hours of fermentation, productivity
of FB was considerably higher in comparison with other
Culture media, prepared from raw juice with the types of yeast applied in dried forms, DD, DW1, DW2
degree of dilution adjusted to achieve initial mass frac- and DB. This fact can be explained with diverse lengths
tion of fermentable sugars estimated as optimal, were ino- of lag phases for different types of yeast. The length of
culated with different types of yeast in order to evaluate the lag phase depends on factors such as nutrition,
their fermentation productivity. Examined types of yeast growth conditions, inoculation density, temperature and
are commercially available and extensively used as star- growth history of the inoculum (12). According to the
ter cultures in various branches of the fermentation in- statistical analyses of illustrated results (Fig. 4), ethanol
dustry. Dried distiller’s yeast (DD) is planned for use in production was significantly increased during the first
ethanol fuel and alcohol fermentations for beverages. 12 h of fermentation for all applied yeast strains, DD,
Yeasts DW1 and DW2 are used for wine production and DW1, DW2, DB and FB (p=0.0008). During this period,
are selected because of very short lag phase and rapid maximum production of 2.10, 2.07, 2.28, 2.37 and 2.19
fermentation. Strain DW2 is also tolerant of high ethanol g/(L·h) for DD, DW1, DW2, DB and FB, respectively,
volume fractions. The same strain of commercially pro- was achieved. In the further course of fermentation up
duced baker’s yeast is applied in two different forms, as to 48 h, the productivity significantly decreased (Fig. 7)
fresh and dried yeast (DB and FB). for all examined yeast strains (p<0.0001). Experimental
Fig. 7 shows the dependence of the ethanol pro- results (Fig. 7) indicate that there is no statistical differ-
duction on the fermentation time and the applied yeast ence between productivity of the applied yeast strains

Fig. 7. Dependence of the ethanol production on different yeast strains during fermentation of raw juice (DD – dried distiller’s
yeast, DW1 – dried wine yeast, DW2 – dried wine yeast tolerant of high ethanol volume fractions, DB – dried baker’s yeast and FB –
fresh baker’s yeast)
S. POPOV et al.: Bioethanol from Sugar Beet Juice, Food Technol. Biotechnol. 48 (3) 376–383 (2010) 383

after 12 h of fermentation. This fact implies that further 3. S.N. Dodi}, S.D. Popov, J.M. Dodi}, J.A. Rankovi}, Z.Z. Za-
analyses and optimisation for selecting optimal fermen- vargo, Potential development of bioethanol production in
tation time and initial sugar mass fraction for industrial Vojvodina, Renew. Sust. Energ. Rev. 13 (2009) 2722–2727.
application can be done with one of the examined stra- 4. S.N. Dodi}, S.D. Popov, J.M. Dodi}, J.A. Rankovi}, Z.Z. Za-
vargo, Biomass energy in Vojvodina: Market conditions,
ins, meaning that the obtained responses will be valid
environment and food security, Renew. Sust. Energ. Rev. 14
for each of the five strains applied during this research, (2009) 862–867.
DD, DW1, DW2, DB and FB. 5. A. Hinková, Z. Bubník, Sugar beet as a raw material for
bioethanol production, Czech. J. Food Sci. 19 (2001) 224–234.
Conclusion 6. S. Henke, Z. Bubník, A. Hinková, V. Pour, Model of sugar
factory with bioethanol production in program SugarsTM,
J. Food Eng. 77 (2006) 416–420.
This research has confirmed that efficient ethanol
7. V.R. McDonald, Direct microscopic technique to detect
production from raw juice as intermediate product of
viable yeast cells in pasteurized orange drink, J. Food Sci.
sugar beet processing is technically possible with all 28 (1963) 135–139.
examined commercial strains of yeast Saccharomyces 8. Official Methods of Analysis, AOAC International, Gait-
cerevisiae. Under the applied experimental conditions, hersburg, MD, USA (2000).
raw juice can be used for fermentation mash prepara- 9. Statistica (Data Analyses Software System) v. 9.1, StatSoft,
tion without any addition of nutrients, with only initial Inc, Tulsa, OK, USA (2010) (www.statsoft.com).
pH correction. The regression equations obtained in this 10. N. Kosaric, F. Vardar-Sukan: Potential Source of Energy
study can be used to find the optimum conditions of and Chemical Products. In: The Biotechnology of Ethanol,
fermentation process in industrial scale from economic Classical and Future Applications, M. Roehr (Ed.), Wiley-
point of view. Extrapolation of the responses beyond the -VCH, Weinheim, Germany (2001) pp. 89–106.
range of used experimental values may not be valid. 11. S. Dodi}, S. Popov, J. Dodi}, J. Rankovi}, Z. Zavargo, R.
Jevti} Mu~ibabi}, Bioethanol production from thick juice
as intermediate of sugar beet processing, Biomass Bioen-
Acknowledgement ergy, 33 (2009) 822–827.
Financial support from the Ministry of Science of the 12. W.M. Ingledew: Alcohol Production by Saccharomyces cere-
Republic of Serbia is highly acknowledged (grant BTN- visiae: A Yeast Primer. In: The Alcohol Textbook, Nottingham
-20009). University Press, Nottingham, UK (1999).
13. N.V. Narendranath, K.C. Thomas, W.M. Ingledew, Acetic
acid and lactic acid inhibition of growth of Saccharomyces
cerevisiae by different mechanisms, J. Am. Soc. Brew. Chem.
References
59 (2001) 187–194.
1. F.W. Bai, W.A. Anderson, M. Moo-Young, Ethanol fermen- 14. M. Khayet, C. Cojocaru, G. Zakrzewska-Trznadel, Response
tation technologies from sugar and starch feedstocks, Bio- surface modelling and optimization in pervaporation, J.
technol. Adv. 26 (2008) 89–105. Membr. Sci. 321 (2008) 272–283.
2. S.N. Dodi}, S.D. Popov, J.M. Dodi}, J.A. Rankovi}, Z.Z. Za- 15. C. Cojocaru, M. Khayet, G. Zakrzewska-Trznadel, A. Ja-
vargo, Potential contribution of bioethanol fuel to the tran- worska, Modeling and multi-response optimization of per-
sport sector of Vojvodina, Renew. Sust. Energ. Rev. 13 (2009) vaporation of organic aqueous solutions using desirability
2197–2200. function approach, J. Hazard. Mater. 167 (2009) 52–63.