1 Introduction

Fossil fuels are recognized as the world’s main source of energy. However, some problems such as the depletion of oil resources and environmental pollution are among the factors that have prompted scientific studies and industrial efforts in recent years to develop alternative sources [1]. Bioethanol is one of these alternative sources that can be used as a sustainable energy source thanks to its renewability, as well as its effectiveness in reducing air pollution and preventing global warming [1,2,3].

Bioethanol can be produced from different sources, among which inedible resources such as agro-industrial and lignocellulosic wastes have received more attention [1,2,3]. Starch-based waste that is widely available worldwide is one of the feedstocks that can be used for bioethanol production. Such wastes contain large amounts of starch that can be converted into fermentable sugars [4]. For example, potatoes are one of the most important food crops in the world and are widely used in all kitchens, restaurants, and food industries. As a result, large quantities of potato peel (PP) are produced as a by-product that can be used as a cheap feedstock for bioethanol production [5]. In the literature, the starch content of PP has been reported to be about 20% (in native sample) [6] and 48 to 52% (on dry basis) [5, 7]. The native PP also contains about 10% hemicellulose, 4% cellulose, and 6% lignin [6]. The starch and cellulose content of potato peel makes it a promising feedstock for bioethanol production.

In the bioethanol production from starch-based waste, a hydrolysis step must be performed before the fermentation step to produce fermentable sugars [8]. Acidic or enzymatic hydrolysis methods can be employed for this purpose. Enzymatic hydrolysis of starchy compounds occurs in two steps by applying two types of enzymes: In the first step, α-amylase enzyme is used to produce maltose and maltodextrins by hydrolyzing the α-(1-4) chemical bonds of starch (liquefaction step). In the next step, the partially hydrolyzed starch from the liquefaction step is converted to glucose in a process called saccharification by applying glucoamylase enzyme. Then, the produced glucose can be fermented to ethanol using appropriate microorganisms [7,8,9,10].

Saccharomyces cerevisiae, the most conventional strain for production of bioethanol, has been used in some research for bioethanol production from PP [5,6,7, 11, 12]. Arapoglou et al. hydrolyzed a number of batches of PP with a combination of various enzymes (amylase, glucoamylase, and cellulase) and used S. cerevisiae for fermenting the resulting sugars to ethanol. They reported production of 7.6 g/l ethanol after fermentation, demonstrating that PP (as a zero-value waste of potato industries) exhibits a high potential for bioethanol production [5]. A comparative study between acidic and enzymatic hydrolysis of PP by Khawla et al. indicated that a combination of amylase UEB-S (produced from Bacillus sp. UEB-S) and amyloglucosidase under optimal conditions resulted in production of 69 g/l reducing sugars and 21 g/l ethanol by S. cerevisiae [7]. For reducing the time, energy consumption, and capital costs, the simultaneous saccharification and fermentation method was used for bioethanol production from PP. In this process, a maximum bioethanol concentration of 22.54 g/L was obtained under optimal conditions [6]. Although in all of these studies, the high value of PP for bioethanol production has been identified, and most of them have focused on optimizing the hydrolysis process and paid less attention to the fermentation process.

Although S. cerevisiae is a traditional yeast for bioethanol production, researchers have been studying alternative strains for increasing the ethanol production yield and reducing production cost. Zymomonas mobilis is a gram-negative bacterium with interesting properties such as high sugar consumption, high ethanol yield, and low bacterial biomass production [13, 14]. Lower biomass production by this bacterium is both effective in increasing ethanol production yield and causing fewer problems in industrial equipment. In addition, Z. mobilis as a bacterium has simpler and cheaper culture media than that of the S. cerevisiae. In this regard, various studies have shown the successful use of Z. mobilis for bioethanol production from various feedstocks [15,16,17,18,19,20]. In the study of bioethanol production from agricultural wastes, Braide et al. reported the ethanol yield percentage of 5.17% from PP by Z. mobilis and 5.51% by S. cerevisiae, indicating the good ability of Z. mobilis for bioethanol production from PP [19]. Another study comparing the performance of Z. mobilis and S. cerevisiae showed that Z. mobilis had better performance in producing bioethanol from PP [20]. However, in the abovementioned studies, the effect of some parameters and optimization of the fermentation process by considering the parameters interactions were not included, and more investigations were still required.

The aim of this study was to investigate the potential of Z. mobilis for producing bioethanol from potato peel and optimize the fermentation parameters. To this end, the appropriate conditions for enzymatic hydrolysis of PP were investigated by two enzymes. Then, the fermentation process was examined based on Z. mobilis, and various factors affecting the process were studied with the optimum conditions for maximum ethanol production obtained via response surface methodology (RSM).

2 Materials and methods

2.1 Microorganism

The strain used in this study for fermenting the sugars to ethanol was Zymomonas mobilis PTCC 1718 purchased from the Persian Type Culture Collection. The inoculum was prepared in a conical flask at 120 rpm and 30 °C for 17 h in the culture medium containing 10 g peptone from meat, 10 g yeast extract, and 20 g glucose per liter of deionized water under aerobic condition. All materials were purchased from Merck, Germany.

2.2 Materials

Potato peel was obtained from the university canteen in Mahallat Institute of Higher Education, Mahallat, Iran, in November. Then, PP was dried in an oven at 50 °C for 48 h, ground into smaller pieces by a laboratory grinder, and sieved to obtain a particle size of around 1–2 mm. The ground PP particles were stored at 4 °C for further investigations.

Termamyl (α-amylase) with enzyme activity of 3000 U/ml and amyloglucosidase with enzyme activity of 260 U/ml were purchased from Sigma-Aldrich.

2.3 Enzymatic hydrolysis of PP particles

Specifically, 20 g of dried PP was mixed with 100 ml deionized water in 250-ml flasks. For liquefaction, the mixture was treated with various units of α-amylase enzyme (10–70 units) at 90 °C, pH 7, and 120 rpm for 2 h. Then, the saccharification step was performed by applying various units of amyloglucosidase (1–15 units) to the slurry at 60 °C, pH 5, and 120 rpm for 24 h. To evaluate the amount of released glucose, samples were withdrawn from the flasks at different time intervals and heated at 100 °C for 5 min. After cooling, the samples were centrifuged at 4000 rpm for 10 min. The supernatants were used for determining the content of reducing sugars via 3,5-dinitrosalicylic acid (DNS) method [21]. All the experiments were carried out in triplicate.

2.4 Ethanol fermentation of PP hydrolysates

The fermentation step was carried out in 250-ml Erlenmeyer flasks containing different volumes of PP hydrolysates. The initial sugar concentration in the fermentation medium could be adjusted according to the experimental design (Table 1) by changing the volume of PP hydrolysates. Then, the PP hydrolysates were supplemented with different amounts of peptone from meat and yeast extract (Table 1), with the pH of the medium adjusted to 5.5. The entire flask was then sterilized at 121 °C for 15 min in an autoclave. For inoculation, subsequently, an appropriate volume of Z. mobilis cells’ suspension (providing a specific amount of bacterial dry weight based on the experimental design) was added to the medium. The working volume of the flask was then adjusted to 100 ml by adding sterilized distilled water. Finally, the mixture was incubated anaerobically at 30 °C and 80 rpm for a specific time (according to the experimental design) (Table 1).

Table 1 List of fifty RSM-designed experiments of ethanol production optimization
Full size table

2.5 Analytical methods

The bacterial dry cell mass in the inoculum culture as well as the fermentation medium was calculated as gram by measuring the optical density (OD) of the mixture at 600 nm using a standard curve. For evaluating the amount of produced ethanol, samples were taken from fermentation flasks and centrifuged at 4000 rpm for 10 min. The supernatant was then used for measuring the reducing sugar content via DNS method. For ethanol determination, the supernatant was first distilled at atmospheric pressure, and subsequently, the Caputi Jr method [22] was performed.

2.6 Optimization by response surface methodology

The response surface methodology (central composite design) was used to optimize the fermentation process conditions to achieve the maximum ethanol production within the shortest time. Factors affecting the process and their higher and lower levels in the experimental design were determined according to the other researchers’ results [15, 16, 18, 19, 23] and some preliminary experiments. Accordingly, five important factors (initial sugar concentration, bacterial dry cell weight, peptone and yeast extract weight, and fermentation time) were selected, with the amount of ethanol produced considered as the response. The results of the 50 designed experiments are presented in Table 1.

3 Results and discussion

3.1 Enzymatic hydrolysis of potato peel

The enzymatic hydrolysis of PP has been investigated previously in several studies [5,6,7, 9, 12]. Nevertheless, due to the different potato cultivars used in this study, we first tried to examine the effect of the most important factors on enzymatic hydrolysis in order to obtain the highest sugar release. For this purpose, the effects of the dosage of the two enzymes were applied, and the solid PP loading were investigated. Firstly, the impact of α-amylase (Termamyl) enzyme on the PP solids was evaluated to determine the best amount of required enzyme. The appropriate temperature, pH, and time required for this stage were reported 90–95 °C, 6.5–7, and 2–3 h, respectively, in previous studies [5,6,7, 12]. Figure 1a displays the effect of different α-amylase units on the reducing sugars released in 2 h, when the solid loading was 20%. The figure exhibits that the amount of released sugar increased as α-amylase dosage was elevated. However, beyond the enzyme dose of 50 units, no significant increase in glucose production was observed. In addition, as Liu [9] reported earlier, the use of amylase alone has little effect on glucose release.

Fig. 1
figure 1

The concentration of reducing sugars released in enzymatic hydrolysis process. a Effect of amylase dose (using 20% solid loading), b effect of amyloglucosidase dose (with 50 U amylase and 20% solid loading), and c effect of PP solid loading (with 50 U amylase and 12 U amyloglucosidase)

Full size image

Figure 1b illustrates the effect of different doses of amyloglucosidase enzyme on the amount of released sugar with a fixed amylase dosage of 50 U in 24 h. In this step, the optimum temperature, pH, and time of about 60 °C, 4.5, and 24 h, respectively, were reported by other researchers [5,6,7, 12]. As can be seen in this figure, the higher the enzyme dosage, the higher the amount of released sugar was. A maximum amount of 88.4 g/l reducing sugar was obtained in the enzyme combination of 50 U amylases and 15 U amyloglucosidases. However, upon elevating the enzyme dosage from 12 to 15 units, no significant increase was observed in the amount of reducing sugar. The trend of the results is in agreement with the previous researches’ findings [7, 9]. For evaluating the best solid loading, different PP solid loadings were used in the enzyme combination of 50 U amylases and 12 U amyloglucosidases in 24 h. Figure 1c reveals that at low PP solid loadings, the amount of released reducing sugars is not significant; however, by increasing the solid loading, a higher amount of reducing sugars was released. According to this figure, the best solid loading was 20% which gave 88.2 g/l of reducing sugars. At solid loadings greater than 20%, there was little increase in the amount of reducing sugars. In addition, high solid loading can enhance the viscosity of the medium which can inversely affect the process [9]. For comparison, Khawla et al. [7] reported the released reducing sugar concentration of 63 g/l using 45 units of amylase combined with 9 units of amyloglucosidase and 15% of PP solid loading in enzymatic hydrolysis of PP wastes.

3.2 Ethanol fermentation

3.2.1 Factors considered in RSM experimental design

Following the PP hydrolysis process, the hydrolysate was used for ethanol fermentation by Z. mobilis. Since there were few previously published study on fermentation of PP by Z. mobilis, we evaluated and optimized the effective parameters in ethanol fermentation process in this section using RSM central composite design. Based on the previous studies [15, 16, 18,19,20, 23] on Z. mobilis as well as some preliminary experiments, we identified five important factors and checked their high and low levels for optimization process by RSM. The factors were initial sugar concentration, bacterial dry cell weight, peptone and yeast extract weight, and fermentation time. In this regard, Vaheed et al. [15] also identified the same factors as dominant factors by screening the effective factors using Plackett-Burman design in ethanol fermentation of carob pods by Z. mobilis. The initial glucose concentration and dried bacterial weight ranges of 4–7 g glucose, and 0.01–0.02 g per 100 ml of fermentation medium were reported in their study, respectively. The amount of peptone and yeast extract, as nitrogen sources, was considered within 0–0.43 g (0–9.67 g/l) in the fermentation time of 24–48 h. They also revealed that the effect of pH within 5–7 had no significant effect on the response, with shaking rates at low values (around 80 rpm) showing no negative effect on ethanol production [15]. In molasses fermentation with 0.02 g/l Z. mobilis in the time range of 24–48 h, the temperature of 30 °C and pH within the range of 5.0–7.0 were reported as optimum conditions [18]. Similarly, in solid-state fermentation of carob pod by Z. mobilis, the optimum temperature of 31 °C, pH 5.1, and 0.7 g of meat peptone (7 g/l) were also reported as the optimum conditions [16].

After identifying the effective factors and their ranges, fifty experiments were designed with the amount of produced ethanol as response measured for each experiment (Table 1). Then, the results were analyzed by Design Expert (dx-7) software.

3.2.2 Optimization of ethanol fermentation process

The analysis of the results obtained from the experimental design by the software showed that F value of the model was 11.04, so the resulting model is significant. The F values and significance levels (P values) of the model and its factors, resulting from the analysis of variance (ANOVA) of the model, are presented in Table 2. P values less than 0.05 indicated that the model terms are significant. In this case, the model and the model factors A (initial sugar concentration), B (bacterial dry weight), C (peptone weight), D (yeast extract weight), and their interactions AB, AC, BC, A2, and C2 were significant. By excluding the insignificant terms from the model, the following equation (Eq. 1) for the model can be derived in terms of the coded factors:

$$ \mathrm{Ethanol}=19.9+1.48\ \mathrm{A}+0.531\ \mathrm{B}+0.596\ \mathrm{C}+0.364\ \mathrm{D}+0.493\ \mathrm{A}\mathrm{B}+0.745\ \mathrm{A}\mathrm{C}-0.416\ \mathrm{B}\mathrm{C}-0.457\ {\mathrm{A}}^2-0.651\ {\mathrm{C}}^2 $$
(1)
Table 2 Analysis of variance (ANOVA) of the developed RSM model
Full size table

The R2 value of 0.844 was calculated for the simplified model, where the predicted R2 of 0.7194 is in reasonable agreement with the Adj-R2 of 0.8092. To show this better, the model predicted values versus actual results were shown in Fig. 2. The ANOVA of the simplified model including F and P values are also reported in Table 3.

Fig. 2
figure 2

Predicted vs. actual values plot

Full size image
Table 3 Analysis of variance (ANOVA) of the simplified RSM model
Full size table

The interactive effects of the process factors on ethanol concentration are shown in response surface graphs in Figs. 3, 4, and 5. Figure 3 depicts the interaction effects of the initial sugar concentration and bacterial dry weight when peptone and yeast extract weights were both 0.2 g with 30-h fermentation time. It can be observed that the amount of produced ethanol increased with elevating of the sugar content in the medium. Ethanol production also increased with raising the inoculum concentration at constant initial sugar content. The interaction effect of initial sugar concentration and peptone weight is presented in Fig.4. In this figure, the bacterial dry weight, yeast extract, and fermentation time were 0.024 g, 0.2 g, and 30 h, respectively. This figure also reveals the positive effect of initial sugar concentration on ethanol production. The maximum ethanol production of about 23.8 g/l was observed with peptone weight of 0.41 g. Thus, addition of nitrogen sources such as peptone and yeast extract can help enhance ethanol production. The effect of peptone addition is also shown in Fig. 5 demonstrating the interaction effect of peptone weight and bacterial dry weight. This figure depicts that under conditions of 60 g/l initial sugar, 0.27 g yeast extract, and 30-h fermentation time, a maximum ethanol of 23.0 g/l was produced at peptone weight of 0.28 g and bacterial dry weight of 0.03 g. The same trend was observed in ethanol fermentation of carob pods by Z. mobilis [15].

Fig. 3
figure 3

Response surface curve for the interaction effects of initial sugar concentration and bacterial dry weight on ethanol (peptone and yeast extract weight, 0.2 g; fermentation time, 30 h)

Full size image
Fig. 4
figure 4

Response surface curve for the interaction effects of initial sugar concentration and peptone weight on ethanol (bacterial dry weight, 0.024 g; yeast extract weight, 0.2 g; fermentation time, 30 h)

Full size image
Fig. 5
figure 5

Response surface curve for the interaction effects of bacterial dry weight and peptone weight on ethanol (initial sugar concentration, 60 g/l; yeast extract weight, 0.2 g; fermentation time, 30 h)

Full size image

Using the model obtained from the experimental design, the optimum condition for maximum ethanol production within the minimum fermentation time was predicted by the software as follows: initial sugar concentration of 61.3 ± 0.0098 g/l, bacterial dry weight of 0.024 g, peptone weight of 0.35 g, yeast extract weight of 0.35 g and fermentation time of 31 h. The optimum conditions predicted by the model are similar to run 39 of experimental design table (Table 1). To verify the optimum conditions, this test was carried out in triplicate, in which a maximum amount of 23.3 ± 0.015 g/l ethanol was produced. The results of reducing sugar measurements under optimum conditions revealed that the reducing sugar remaining in the medium after fermentation was 12.9 ± 0.01 g/l, indicating that 78.9% of the initial sugar was consumed. Thus, the amount of sugar consumed during the fermentation was 48.4 g/l leading to a 48.3% ethanol yield (94.7% of the theoretical yield). The summary of the results obtained in this research are reported in Table 4. In previous studies, maximum ethanol concentrations and theoretical yields of 21 g/l and 70% [7], as well as 7.6 g/l and 46% [5], respectively, were reported for bioethanol production from PP by S. cerevisiae. Also, in simultaneous saccharification and fermentation of PP by S. cerevisiae, an ethanol yield of 0.32 g/g was obtained by Chohan et al. [6]. In a recent study of bioethanol production from PP, 5.17% v/v of ethanol is produced after 72 h fermentation by Z. mobilis and 5.51% v/v by S. cerevisiae. In another report, the highest yield of ethanol production by Z. mobilis from PP was reported after 5 days, which is higher than the yield obtained by S. cerevisiae after 7 days. Hence, the results of this study are comparable with the findings obtained by other researchers from PP by S. cerevisiae, suggesting the promising ability of Z. mobilis for bioethanol production from PP wastes. Process optimization in this research has resulted in higher bioethanol yield compared with the similar previous researches.

Table 4 Summary of the results obtained in this research for bioethanol production from PP by Z. mobilis
Full size table

3.2.3 Time course of ethanol production and sugar consumption

Under the optimum conditions, we evaluated the time variations of produced ethanol and glucose concentration in the fermentation medium. Figure 6 indicates that the lag phase was about 10 h. Then, there was a rapid increase in ethanol production and sugar consumption. After 30 h, little change occurred in the ethanol production. This result is consistent with the prediction by the model where the highest amount of ethanol within the shortest fermentation time occurred at 31-h fermentation time. Thus, a maximum ethanol of around 23–24 g/l was achieved between 30 and 40 h of fermentation time.

Fig. 6
figure 6

Time courses of produced ethanol and glucose concentrations in fermentation medium

Full size image

4 Conclusions

The results of this study demonstrated the good potential of Z. mobilis for bioethanol production from PP in comparison with previous studies using traditional yeast, S. cerevisiae. For enzymatic hydrolysis process, the combination of 50 U amylases and 12 U amyloglucosidase enzymes could effectively release the fermentable sugars from PP solids. Examination of the effect of different factors on ethanol production in the fermentation process via RSM revealed that the maximum amount of ethanol (23.3 g/l) was obtained under the following optimal conditions: 61.3 g/l initial sugar, 0.24 g/l bacterial dry cell, 3.5 g/l meat peptone, 3.5 g/l yeast extract, and 31-h fermentation time. The final concentration of ethanol obtained is not high enough for industrial scales because it may cause high energy consumption in purification step. Thus, more study is required for industrializing this process.

Availability of data and material

Not applicable.

Code availability

Design Expert software trial version