Abstract

VARIATION OF THE ETHANOL YIELD DURING VERY RAPID BATCH FERMENTATION OF SUGAR-CANE BLACKSTRAP MOLASSES

W. Borzani^** To whom correspondence should be addressed. To whom correspondence should be addressed. and C.H. Jurkiewicz

lnstituto Mauá de Tecnologia, Estrada das Lágrimas 2035,

09580-900, São Caetano do Sul, SP, Brazil.

Phone: 011-741 3119 - Fax : 011- 741 3131

(Received: January 25, 1998; Accepted: April 15, 1998)

Abstract - During rapid ethanol fermentation (2-3 h) of sugar-cane blackstrap molasses, a significant increase in the ethanol yield was frequently observed as fermentation proceeded, eventually leading to yields higher than the theoretical value when the end of the process was approached. In order to explain the above facts, three assumptions were examined: 1. temporary ethanol accumulation within the yeast cells; 2. variation of the dry matter content and/or of the microorganism density during the fermentation; 3. transformation of sugars into undetectable extra-cellular fermentable compounds at the initial stages of the process. Based on the experimental results presented here, the third of the above assumptions seems to explain the observed increase in the ethanol yield.

Keywords: Rapid ethanol fermentation, ethanol yield variation, ethanol fermentation of molasses.

INTRODUCTION

Many experiments have been reported showing that the ethanol yield is practically constant during batch fermentation tests carried out under usual conditions. Equation (1), for example, shows the result obtained from a typical normal test (Gómez and Borzani, 1988).

(1)

(r = 0.9994)

where M_pe and M_cs are, respectively, the masses of produced ethanol and of consumed sugars, and r is the correlation coefficient.

When, however, the yeast cells concentration was sufficiently high (60 to 70 g.L^-1, dry matter) leading to a very rapid fermentation (2 to 3 h), the ethanol yield significantly increased during the test. Similar results were also observed in rapid fed-batch tests (Borzani, 1996).

The main purpose of this paper is to present some experimental results and also to discuss some assumptions that could be considered in order to explain the cited increase in the ethanol yield.

EXPERIMENTAL

The inocula and the mashes were prepared as previously described (Carvalho et al., 1990), except 2.2 kg of pressed yeast (Saccharomyces cerevisiae) were used. The tests were carried out in 14-L fermenters under the following conditions: volume of inoculated medium, 10.0 L; total reducing sugars (TRS) initial concentration, 180-210 g.L^-1, temperature, 32.0± 0.5°C; pH, 4.7; the pH was not controlled since it practically did not vary during the test; impeller speed, 200 mim^-1; no air was supplied. Small samples (50-60 mL) of the fermenting medium were rapidly (1-3 min) cooled to 4-6°C and then used for analytical determinations. Dry matter cell concentrations (X) were measured by filtering 3.0-5.0 mL of fermenting medium through a Millipore membrane (pore diameter, 1.2 m m). After washing with 100 mL of distilled water, the material was dried at 105-110°C until it reached constant weight (2-3 h). TRS concentrations (S) were measured in the centrifuged (1,440 x g; 30 min) sample by the Somogyi (Somogyi, 1952) or by the DNS (Miller, 1959) methods. Ethanol concentrations (E) were measured in the centrifuged sample by the dichromate method (Joslyn, 1970). The concentrations of non-volatile substances in the centrifuged medium (R) were measured by drying 25 mL samples at 55-60°C until it reached constant weight. The carbon content (C) of the obtained non-volatile substances was determined in a Perkin-Elmer analyzer (model 2400 CHN).

RESULTS

Figure 1 shows the results obtained in a typical test (Test 1) (Carvalho, 1994). The following equations were used to calculate the masses of microrganism (M_x), TRS (M_s) and ethanol (M_e) during the experiment:

(2)

(3)

(4)

(5)

(6)

where V is the volume of fermenting medium, M_so, is the initial value of M_s, M_eo is the initial value of M_e, s is the dry matter content of the pressed yeast used as inoculum, and r is the yeast cell density. In Test 1, s = 0.292 and r = 1.12 x 10³g.L^-1 (Borzani and Vairo, 1983).

Since S and E were measured in the centrifuged medium, equations (3) and (5) considered the actual total cell volume (V.X/s . r ) in order to calculate M_s and M_e in the aqueous phase.

The ethanol yield (Y), from the beginning of the fermentation until a given moment t, was calculated by:

(7)

Table 1 presents the values of M_x, M_s, M_cs, M_e, M_pe and Y calculated from the curves of Figure 1.

Figure 2 shows the variation of Y and the curve M_pe vs. M_cs during Test 1.

Table 2 presents the values of Y calculated at time intervals D t = 0.5 h during Test 1.

Ethanol yields much greater than the theoretical value (0.511) were obtained.

Figure 3 shows the results obtained in a test (Test 2) carried out with the purpose of determining the masses of carbon in the non-volatile substances of the centrifuged medium. The values of M_x, M_cs, M_pe, M_r (mass of non-volatile substances in the centrifuged medium), M_c (mass of carbon of the non-volatile substances cited above), M_cc (mass of consumed carbon), Y and Y_c (ethanol yield calculated on the basis of consumed carbon) are presented in Table 3. In this case the following equations were also used.

(8)

(9)

(10)

(11)

where M_co is the initial value of M_c. In Test 2, s = 0.295 and r = 1,12 x 10³ g.L^-1.

Figure 4 shows the values of Y and Y_c calculated at time intervals D t = 0.5 h during Test 2.

Table 1: Ethanol yields (Y) and masses of dry yeast cells (M_x), TRS (M_s), consumed TRS (M_cs), ethanol (M_e) and produced ethanol (M_pe) calculated from the curves of Figure 1

Table 3: Ethanol yields (Y and Y_c) and masses of dry yeast cells (M_x), consumed TRS (M_cs), produced ethanol (M_pe), non-volatile substances (M_r), carbon (M_c) and consumed carbon (M_cc) calculated from the curves of Figure 3 (Test 2)

Figure 4: Variation of the ethanol yields (Y and Y_c) calculated at time intervals D t = 0.5 h during Test 2.

DISCUSSION

The values of Y presented in Table 2 clearly show that M_cs and/or M_pe do not represent the actual masses of consumed sugar and/or produced ethanol during the experiment.

With the purpose of explaining the results obtained in Test 1 we will assume, first of all, that the actual value of the ethanol yield was constant during the fermentation, and secondly that the variation of Y (see Figure 2) could be due to one of the following possibilities:

1. The ethanol temporarily accumulated within the cells and later diffused to the medium through the cell envelopes (Navarro and Durand, 1978);

2. The value of s r , assumed as constant (327.0 g.L^-1) when we used equations (3) and (5), varied during the test;

3. As was earlier proposed (Borzani et al., 1977), part of the sugars was transformed into extra-cellular fermentable compounds, undetectable by the adopted analytical method which were later fermented producing ethanol.

The ethanol yield, assumed as constant during the experiment, was calculated at the end of the fermentation considering both tire mass of produced ethanol in the centrifuged medium and the mass of intracellular ethanol (E_f.a .V.X_f /s .r , where a .V.X_f /s .r , is the volume of intracellular aqueous phase; 0<a <1):

(12)

where E_f, X_f and M_csf are, respectively, the values of E, X and M_cs at the end of the fermentation, and a is the relative volume of the intracellular aqueous phase.

ln Test 1 (E_f = 64.7 g.L^-1; V = 10.0 L; X_f = 70.3 g.L^-1; s r = 327.0 g.L^-1; M_csf = 1,661 g) equation (12) is shown as:

(13)

Let us consider now the first possibility cited above, that is, there was a temporary ethanol accumulation within the yeast cells. In this case, as Y* was assumed as constant during the entire process, the intracellular ethanol concentration (E_i), at any moment, may be calculated by:

(14)

Figure 5, which presents the calculated values of E_i during the Test 1 when a = 0.5 and a = 0,8, shows, for instance, E_i - E > 120 g.L^-1 at t = 0.5 h (see also Figure 1). Considering that the ability of ethanol to accumulate within the yeast cells has been seriously questioned (Pamment and Dasari, 1988) in light of more accurate measurements of intracellular ethanol concentrations (Dasari et al., 1983; Dasari et al., 1985; Dombeck and Ingram, 1986; Guijarro and Lagunas, 1984; Loureiro and Ferreira, 1983) and the existence of artifacts in the analytical procedures adopted by earlier investigators (Dasari et al., 1983), values of E_i - E as high as those calculated can not represent the

actual situation. Consequently, the assumption of intracellular ethanol accumulation can not be accepted to explain the results of Test 1.

We may examine now the second possibility cited above, that is, the variation of Y could be a consequence of the variation of s r during Test 1. Equation (15) permits the calculation of s r at each moment in order to have Y* constant:

(15)

Table 4, presenting the calculated values of s r by applying equation (15) to the experimental results of Test 1, clearly shows that the increase in Y can not be explained as a consequence of the variation of s r during the fermentation.

Figure 5: Concentrations of ethanol within the yeast cells (E_i) during Test 1, calculated by equation (14), assuming two values of the relative volume of the intracellular aqueous phase (a ).

Let us then consider the third possibility cited above, that is, the transformation of sugars into fermentable compounds (undetectable by the adopted analytical methods) that temporarily accumulated in the medium and were later fermented producing ethanol could explain the observed Y variation. The above assumption was based on results obtained by studying the oscillatory phenomena in the continuous cultivation of Saccharomyces cerevisiae (Borzani et al., 1977). The total mass of produced ethanol (M_te), including the intracellular ethanol, is given by:

(16)

We may then calculate the mass of TRS (M'_cs) which is actually consumed to produce M_te, that is:

(17)

The mass of TRS (M_ts) transformed into the assumed undetectable compounds will then be

(18)

and its concentration in the centrifuged medium (S_t) may be calculated by:

(19)

Figure 6 shows that the values of S_t, which were calculated by applying equation (19) to the results of Test 1, were practically not affected by a .

Otherwise, the results of Test 2 (Figure 4) show that while the ethanol yields (Y) which were calculated on the basis of consumed TRS increased during the experiment, the ethanol yields (Y_c) which were calculated on the basis of the consumed carbon of the non-volatile substances of the centrifuged medium did not increase. In fact, excluding the value Y_c = 0.880 (probably due to experimental errors), an average Y_c = 0.978 (standard deviation = 0.007) was obtained.

Another value that must be considered is the ratio between Y_c and Y calculated during the last 30 minutes of Test 2, when practically no more undetectable fermentable compounds existed in the medium: 0.980/0.389 = 2.52.

Figure 6: Concentrations of TRS transformed into undetectable compounds (S_t) during Test 1, calculated by equation (19), assuming three values of the relative volume of the intracellular aqueous phase (a ).

Assuming that the carbon consumed during the fermentation was the carbon of the glucose, the theoretical value of the above ratio would be 180.16/72.06 = 2.50.

We may then conclude that the transformation of TRS into extra-cellular fermentable compound which were undetectable by the analytical procedures used in this work could explain the increase in Y.

ACKNOWLEDGMENTS

The authors are grateful to Renato Piplovic for technical assistance.

NOMENCLATURE

C Carbon content of the non-volatile substances of the centrifuged medium.

E Ethanol concentration in the centrifuged medium, g.L^-1.

E_f Value of E at the end of the fermentation, g.L^-1.

E_i Intracellular ethanol concentration, g.L^-1.

M_c Mass of carbon of the non-volatile substances of the centrifuged medium, g.

M_co Initial value of Mc, g.

M_cc Mass of consumed carbon, g.

M_cs Mass of consumed TRS, g.

M_csf Value of Mcs at the end of the fermentation, g.

M'_cs Mass of consumed TRS assuming Y = Y*, g.

M_e Mass of ethanol in the centrifuged medium, g.

M_eo Initial value of Me, g.

M_pe Mass of produced ethanol, g.

M_r Mass of non-volatile substances in the centrifuged medium, g.

M_s Mass of TRS in the centrifuged medium, g.

M_so Initial value of Ms, g.

M_te Total mass of ethanol including the intracellular ethanol, g.

M_ts Mass of TRS transformed into undetectable fermentable compounds, g.

M_x Mass of yeast cells (dry matter) in the fermenting medium, g.

r Correlation coefficient.

R Concentration of non-volatile substances in the centrifuged medium, g.L^-1.

S TRS concentration in the centrifuged medium, g.L^-1.

S_t Concentration of TRS transformed into undetectable fermentable compounds in the centrifuged medium, g.L^-1.

t Time, h.

TRS Total reducing sugars calculated as glucose.

V Volume of the fermenting medium, L.

X Yeast cells concentration (dry matter) in the fermenting medium, g.L^-1.

Xf Value of X at the end of the fermentation, g.L^-1.

Y Ethanol yield calculated on the basis of consumed TRS.

Y* Value of Y at the end of the fermentation calculated considering also the intracellular ethanol.

Y_c Ethanol yield calculated on the basis of consumed carbon.

a Relative volume of the intracellular aqueous phase.

D t Time interval, h.

r Yeast cells density, g.L^-1.

s Dry matter content of the yeast cells.

Publication Dates

History