Rotten oranges ( Figure 1 ), by-products from agri-food distribution chains in Lomé-Togo, served as a raw material rich in fermentable sugars, offering significant

potential for bioethanol production. To valorize this waste in this green energy, their juice was mechanically extracted with a juicer after peeling and manual cutting. The resulting must was supplemented with sodium glutamate (SG) to stimulate yeast growth and improve fermentation kinetics. The conversion of sugars into ethanol was carried out by the yeast Saccharomyces cerevisiae (Saf-Levure S.I. Lesaffre), usually applied in the form of active dry yeast.

Prior to its utilization as a raw material, rotten oranges were carefully washed to remove all impurities, including plant debris, plastics, sand, and microorganisms. They were then peeled, cut, and pressed to extract the raw juice. After filtration, the collected filtrate was concentrated by evaporation through heating without reflux in a glass flask to obtain four types of musts with concentrations of 16 ˚Brix, 20 ˚Brix, 24 ˚Brix, and 28 ˚Brix. The ethanol fermentation of the prepared musts was carried out with the genus Saccharomyces cerevisiae, used as fermentative microorganisms. Once ethanol fermentation of the musts was complete, the produced bioethanol was recovered by fractional distillation through mass transfer using a simple distillation column in order to determine the ethanol content in the musts at the end of fermentation.

The principle adopted for this characterization was based on oven-drying slices of degraded orange, with an initially known mass, placed in a Petri dish at a temperature of about 103˚C ± 2˚C until complete removal of water and volatile matter. The moisture and volatile matter content (MC) was then calculated using Equation (1) [20] .

$\text{MC}=\frac{\text{MH}-\text{MS}}{\text{MH}-\text{MB}}\times 100$ (1)

where:

MH = mass of the fresh rotten orange + Petri dish before oven-drying;

MS = mass of the dried rotten orange + Petri dish after oven-drying;

MB = mass of the empty Petri dish.

Dry matter content (Si), which is the mass percentage (%) of dry matter contained in an organic substance, was expressed relative to its total mass (wet matter). In practice, this percentage is theoretically calculated using the value of the moisture and volatile matter content (MC), according to Equation (2) [20] .

Si + MC = 100% (2)

The experimental protocol used consisted of calcining oven-dried slices of rotten orange for two hours in a crucible placed in a muffle furnace, with the maximum temperature set at 550˚C. This operation was carried out with a heating rate of 10˚C/min from the beginning to the end of the experiment. After the complete destruction and elimination of the organic matter in the form of gases escaping from the crucible, the ash content (AC) was evaluated using Equation (3) [20] .

$\text{AC}=\frac{M_1-M_0}{M_2-M_0}\times 100$ (3)

where:

AC = Ash content of the rotten oranges;

M₀ = Mass of the empty crucible;

M₂ = Total mass of the crucible and the sample before calcination;

M₁ = Total mass of the crucible and the ashes after calcination.

In this study, the extraction method applied for recovering raw juice from rotten oranges was mechanical pressing, using a juicer, commonly employed by juice vendors in Lomé-Togo for preparing citrus juices. The raw juice obtained through mechanical pressing was heated to 80˚C for 60 minutes, then cooled to room temperature before being stored in a freezer at −23˚C for later use.

To minimize substrate inhibition, the effects of feeding yeast with different sugar concentrations on the ethanol production by batch and fed-batch cultures [21] in a 1-L fermentor were investigated. To achieve this purpose, the raw juice taken from the freezer was first thawed, and the resulting liquid was concentrated by heating without reflux to obtain four types of concentrated musts (16 ˚Brix, 20 ˚Brix, 24 ˚Brix, and 28 ˚Brix). These intentionally prepared musts were stored in a freezer (−23˚C) for later operations. To make a comparison, sucrose musts were also prepared to carry out fermentation in fed-batch mode. Ethanol fermentation of the sucrose musts was used as a reference for that of the rotten orange musts in order to optimize bioethanol production in this study. Indeed, the sucrose musts used as a model substrate to carry out the ethanolic fermentation in order to transpose the method to the case of orange juice because sucrose, glucose, and fructose are the sugars predominantly present in orange juice, as referenced in the work done by [16] . Moreover, the hydrolysis of sucrose leads to the formation of glucose and fructose, which are the monosaccharides whose ethanolic fermentation based on Saccharomyces cerevisiae yields the best bioethanol production rates [16] .

Pre-fermentation is the step that precedes ethanol fermentation itself. At this stage, an inoculum was prepared with 1/10 of the volume of each rotten orange must, with the addition of baker’s yeast of the genus Saccharomyces cerevisiae (1 g/L) and SG in a concentration range from 0 to 8 g/L. The resulting ethanol fermentation broths were left to ferment under anaerobic conditions at room temperature (30˚C - 32˚C) for 24 hours, allowing the fermentative microorganisms to adapt to their culture medium.

At this stage, each pre-fermented must was mixed with the remaining 9/10 of the must to initiate ethanol fermentation properly [22] . After stirring, the mixture was left to ferment at room temperature (30˚C - 32˚C) under two ethanol fermentation modes, namely batch fermentation and fed-batch fermentation.

The ethanol fermentation reaction was periodically monitored every 24 hours from the start until the exhaustion of each must by measuring the total soluble solids (TSS), expressed in degrees Brix. An Abbe/Azzota refractometer, model AR-2 [20] , was used for this measurement. At the end of fermentation, the limit attenuation (LA) of each fermented must was determined using Equation (4).

LA = [(Initial Brix − Final Brix)/Initial Brix] × 100% (4)

The alcoholic strength, or ethanol content produced from the musts at the end of fermentation, was determined using the pycnometric method according to the recommendations of AOAC (Association of Official Analytical Chemists), method 982.10.

Ethanol fermentation in batch mode and fed-batch mode were comparatively applied with the aim of maximizing ethanol yield.

The physicochemical properties of rotten oranges, namely moisture and volatile matter content (MC), dry matter content (Si), mineral fraction or ash (AC), as well as the proportion of organic matter (OM), are illustrated in the form of a two-dimensional pie chart ( Figure 2 ).

The analysis of the data from this representation indicates that the decomposed oranges used in this study have a moisture and volatile matter content of 87.10% ± 0.50%. In addition, the dry fraction corresponding to a dry matter content (Si) of 12.90% ± 0.50% is divided into an ash content (AC) of 0.67% ± 0.01% and an organic matter content (OM) of 12.23 ± 0.49%.

The variation of TSS as a function of time during batch fermentation of sucrose musts with normal density (20 ˚Brix) and very high density (28 ˚Brix), supplemented with SG at different concentrations, produced the curves illustrated in Figure 3(A) and Figure 3(B) .

A decrease in TSS over time was observed in each must. However, this decrease in TSS was noticeably more pronounced in the sucrose musts supplemented with SG.

The ethanol contents (EC) obtained through batch fermentation of sucrose musts NG (20 ˚Brix) and VHG (28 ˚Brix), with or without SG supplementation, are shown in Figure 4 .

From the results presented in Figure 4 , it can be observed that, regardless of the batch fermentation method adopted or the sucrose must concentration, whether NG (20 ˚Brix) or VHG (28 ˚Brix), the use of SG as a growth factor led to significantly higher ethanol contents compared to the control musts.

The results shown in Figure 5 present the experimental yields of ethanol production (EYP) in batch mode with sucrose musts NG (20 ˚Brix) and VHG (28 ˚Brix).

The ethanol EYPs in this study are relatively close to the theoretical Gay-Lussac yield of 100%. They range from 35.46% ± 0.67% to 87.92% ± 0.65% for NG sucrose musts and from 31.40% ± 0.11% to 66.85% ± 0.33% for VHG sucrose musts.

The curves in Figure 6 illustrate the variation of TSS over time in VHG (28 ˚Brix) sucrose musts under semi-batch fermentation.

It can be seen in this figure that the decrease in TSS is not continuous due to the periodic additions of sugars to the fermentation broths, which resulted in

cascading increases in TSS in the musts over time. This practice is particularly intended to reduce the stress that fermentative microorganisms would undergo if they were exposed to a must with an initially high sugar concentration (Crabtree effect). Thus, to lower the osmotic pressure across the yeast cell membrane surface during ethanol fermentation, it has been wisely recommended not to introduce the total amount of fermentable sugars into the culture medium at once, but rather progressively.

The analysis of these curves shows that, at the end of semi-batch fermentation, the SG concentration of 4 g/L appeared to produce the most significant decrease in TSS, as indicated by the corresponding curve reaching the lowest level at the end of fermentation. However, this decrease in TSS is practically similar to that obtained with SG concentrations of 2 g/L and 8 g/L. In contrast, in the neutral sucrose must, the decrease in TSS was comparatively much lower.

A more detailed analysis of the results presented in this section shows that the addition of SG had a very significant positive effect on sucrose consumption during fed-batch fermentation.

Table 1 records the ethanol contents (EC) obtained from fed-batch fermentation of sucrose musts (28 ˚Brix) as a function of SG concentrations.

From the results presented in Table 1 , it can be observed that the use of SG as a growth factor led to significantly higher EC values compared to the neutral must.

In this study, the increase in EC resulting from the use of SG is consistent with the results illustrated in Figure 6 .

The experimental yields of production (EYP) of bioethanol from fed-batch fermentation of sucrose musts (28 ˚Brix), as a function of SG concentrations, are presented in Table 2 . The EYPs from fermentation of sucrose musts (28 ˚Brix) ranged between 67.34% ± 0.77% and 69.66% ± 0.42% for sucrose musts supplemented with SG, whereas that of the neutral sucrose must was only 32.99% ± 0.18%.

The supplementation of sucrose musts (28 ˚Brix) with SG improved the experimental yield of bioethanol production in fed-batch mode, with values ranging from 104.12% to 111.15% compared to the neutral sucrose must.

The results presented in Table 3 indicate that the limit attenuation (LA) values of ethanol fermentation in NG sucrose musts (20 ˚Brix) were higher than those of VHG sucrose musts (28 ˚Brix). The presence of SG in the musts had a beneficial effect on LA values. Moreover, fed-batch fermentation yielded higher LA values than batch fermentation for the high-density sucrose must (28 ˚Brix).

However, it was observed that the LA values of NG sucrose must (20 ˚Brix) fermented in batch mode were significantly higher than those corresponding to the fermentation of VHG sucrose must (28 ˚Brix).

Figures 7-11 present the influence of SG on the decrease of TSS in rotten orange musts during ethanol fermentation in batch mode, with different initial must concentrations in ˚Brix: 9.5, 16.0, 20.0, 24.0, and 28.0.

The shape of these curves reveals that there are no significant differences between the pure must and the musts supplemented with SG (2 g/L). In particular, in Figure 11 a markedly less pronounced decrease in TSS is observed in the must supplemented with SG (2 g/L) compared to the neutral must. This suggests that the presence of SG in rotten orange musts slows down the consumption of fermentable sugars.

The ethanol contents (EC) obtained during batch fermentation of rotten orange musts with initial concentrations ranging from 9.5 ˚Brix to 28.0 ˚Brix are shown in the histograms of Figure 12 .

The results indicate that ethanol production increased progressively with the initial concentration of the rotten orange musts. Thus, in the present study, fermentation of rotten orange musts produced ethanol contents (% v/v) ranging from 4.55% ± 0.15% to 16.51% ± 0.21% for initial concentrations from 9.5 to 28.0 ˚Brix. The highest ethanol content (16.51% ± 0.21% v/v) was obtained with the must having the highest initial concentration, 28.0 ˚Brix, and supplemented with SG (2 g/L), while the lowest ethanol content was recorded in the neutral must with the lowest initial concentration, 9.5 ˚Brix.

The experimental yields of production (EYP) of bioethanol from rotten orange musts (28 ˚Brix) through batch fermentation are presented in Table 4 . The highest ethanol EYP, 75.15% ± 0.96%, was obtained from the rotten orange must with an initial concentration of 28 ˚Brix supplemented with SG (2 g/L).

The lowest EYP, 51.21% ± 1.90%, was recorded for the must with an initial concentration of 16 ˚Brix supplemented with SG (2 g/L). Overall, it can be observed that, with the exception of musts having initial concentrations of 28 ˚Brix and 9.5 ˚Brix, all other musts showed higher ethanol EYPs when supplemented with SG (2 g/L).

The results presented in Table 5 indicate the final capacities of consumption of fermentable sugars contained in rotten orange musts, expressed as limit attenuation LA (%), through batch fermentation using Saccharomyces cerevisiae.

The curves presented in Figure 13 show the evolution of total TSS in rotten orange musts (28 ˚Brix) over time through ethanol fermentation in fed-batch mode.

The shape of these curves shows that there is no positive improvement between the neutral must and the musts supplemented with SG.

Among the EC values ( Table 6 ) obtained through fed-batch fermentation of very high-density rotten orange musts (28 ˚Brix), the highest, 14.72% ± 0.12% (v/v), was achieved with supplementation of the must with SG (2 g/L).

The ethanol EYPs from fed-batch fermentation of rotten orange musts at 28 ˚Brix are recorded in Table 7 .

These results indicate that supplementation of the must with SG (2 g/L) produced the highest yield, with a value of 66.98% ± 0.54%, compared to 63.72% ± 0.25% for the neutral must, which gave the lowest yield.

The results presented in Table 8 show that ethanol fermentation of neutral VHG rotten orange musts yielded higher LA values than neutral VHG sucrose musts, i.e., 51.79% ± 0.00% versus 23.81% ± 0.60%. While supplementation of sucrose musts with SG had a beneficial effect on LA, which generally increased with SG concentration, supplementation of rotten orange musts, on the other hand, had a negative effect on LA, which worsened as SG concentration increased.

When the two fermentation modes of rotten orange musts are compared ( Table 9 ), it is found that batch mode achieved higher LA values.