Experimental optimization of enzymatic and thermochemical pretreatments of bread waste by central composite design study for bioethanol production Open Access

The decrease in oil resources and the use of fossil fuels, which are regarded as one of the main pollution problems, are driving a renewed interest in alternative energy sources. Biofuels are used in various applications such as: transportation, heating, cooling, manufacturing and electricity generation. It is mainly used as a fuel additive and rarely in pure form for transportation. These fuels are biodegradable, free of sulfur and aromatics and less toxic, their combustion products are less reactive with sunlight, but an improvement in some of their properties is often required and desirable, such as octane number of gasoline (Azhar et al. 2017). Greater attention has been focused over recent years on all techniques of biofuels production. They can be derived from biological raw material and hence are a renewable, sustainable alternative to fossil fuels, they may help to develop more ecologically benign lifestyles, fighting both the problems of climate change and increasing global energy demands.

Currently, the two most prevalent and viable biofuels are bioethanol and biodiesel. Both are suitable for use in conventional gasoline or diesel engines, either neat or in blends in any proportion with conventional fuels.

Biodiesel is derived from different sources, and can be obtained by the transesterification of plant oil and animal oil, animal fats and even soap stock.

Bioethanol can be produced by the fermentation of a wide variety of feed stocks if they contain sugar, starch or cellulose. The latter includes solid waste derived from agriculture or food industry, liquid waste from industrial wastewater or a dedicated energy crop. Besides, two feed stocks can be used at the least costly way and at a large scale : maize in the United States and sugarcane in Brazil.

The choice of raw material depends on its high availability and its chemical and physical properties and the processes that lead to a high percentage of bioethanol in the long term. Most importantly, these wastes will not compete with human food products. They are referred to as second-generation renewable fuels.

The choice of an appropriate method to convert biomass into biofuel remains an essential task based on the substrate characteristics.

There are different pretreatment methods, which can be divided into three main categories: chemical, physical and enzymatic. These pretreatment methods can be combined to improve the degradability of residues.

Physical treatment includes fractionation biomass structure by using mechanical means to reduce particle size with increase in porosity (accessible surface area) as well as the hydrolysis process improvement (Hafid et al. 2017a, 2017b).

The pretreatment step is to break down the complex biomass structure into soluble compounds and release monomeric sugars into hydrolysate for effective bioethanol production. The pretreatment process consists of maximizing the carbohydrates release while minimizing mineralization reactions, chemicals, and energy requirements. (Subhedar & Gogate 2013). Further, it is important for any efficient pretreatment technique performed to limit the degradation of carbohydrates and the formation of inhibitors that slow down the processes of hydrolysis and fermentation. (Singh et al. 2015).

The cell structure of biomass is complex and difficult to pass through, so fractionation and breaks down lignin-carbohydrates bonds require further pretreatment processes such as chemical methods. Liquid hot water pretreatment process, also known as hydrothermal pretreatment, is one of the physical pretreatment processes that convert biomass into its digested form. Hydrothermal pretreatment consists of disrupting the structural of insoluble fraction to enhance biodegradation rate and produce more bioethanol. The latter is viewed as a cost-effective and eco-friendly way for treating biomass, because it does not require any acids or chemicals and is carried out at temperatures ranging from 60 to 170 °C. (Hafid et al. 2017a, 2017b).

Chemical pretreatment process is one of the most often used strategies for hydrolysis or saccharification of organic substrates (Hafid et al. 2017a, 2017b). During this pretreatment process, degradability takes place in the solid phase, and the solid separation can be performed easily.

Chemical pretreatments involving oxidizing agents, acids, alkalis and salts can be used to remove lignin, hemicelluloses and cellulose from lignocellulosic biomasses. Hydrochloric acid (HCl), sulfuric acid (H₂SO₄), hydrogen peroxide (H₂O₂), and acetic acid (CH₃COOH) are commonly used in the acid pretreatment process, while the alkaline pretreatment process employs various compounds such as: aqueous ammonia (NH₃, H₂O), sodium hydroxide (NaOH), potassium hydroxide (KOH), calcium hydroxide (Ca(OH)₂), and magnesium hydroxide (Mg(OH)₂). These chemicals are used to improve traditional productivity processes, such as for instance anaerobic digestion (Ding et al. 2017; Ma et al. 2018).

However, these chemical pretreatment methods require a lot of energy, corrosive materials and harsh operating conditions generating toxic byproducts. As compared to physical and chemical pretreatments, enzymatic pretreatment is regarded as an alternative technique for sustainable green energy generation (Banu et al. 2020). Enzymatic hydrolysis increases sugars yield compared to those solubilized by acids; therefore, it is preferred to physical and/or chemical pretreatment process. Furthermore, the enzymes are extremely highly efficient to degrade the target substrates and may operate at temperatures ranging from 40 to 50 °C. Therefore, hydrolysis requires less energy, and the produced hydrolysate contains less toxic substances (Behera et al. 2014). However, due to a slower reaction rates, high cost and the need to ensure a careful control over reactions conditions. Enzymatic hydrolysis can be viewed as an exclusive pretreatment method that presents technological barriers and this approach is less desirable at commercial scale (Behera et al. 2014; Vohra et al. 2014).

To improve the release of monomeric sugars the pretreatment techniques are often combined with respect to a single pretreatment method. This increases the digestibility of a substrate in a cost-effective manner and reduces the pretreatment intensity of single pretreatment method and thus reduces the pretreatment input energy costs (Atelge et al. 2020).

Responses surface methodology (RSM) is a mathematical and statistical technique for designing experiments, model building, assessing the effect produced by several factors, and reducing the number of experiments. Bioethanol process optimization using RSM models has been previously reported (Sükrü Demirci et al. 2017; Mihajlovski et al. 2021). More fermentable sugars can be achieved though enzymatic hydrolysis from waste bread and by applying RSM model (Sükrü Demirci et al.2017). Some researchers have optimized various fermentation parameters such as: the time of fermentation and yeast inoculum by using the RSM model to achieve an optimal bioethanol yield (Mihajlovski et al. 2021).

The aim of this study is to improve the pretreatment process of bread waste by using RSM. A series of successive pretreatment methods (chemical, thermal and enzymatic) have been performed to enhance bioethanol production as well as to optimize various process parameters such as: waste bread ratio, alpha-amylase concentration, temperature and pH. Additionally, the kinetic of bioethanol production has been modeled by using the modified Gompertz model.

Baguette white bread (BWB) is the Algeria's most heavily consumed product. Baguette white bread waste (BWBW) was collected from the local bakery, stored between 4 and 6 °C and showed no signs of mould growing. To improve the hydrolysis processes, the raw material was cut into small pieces, ground (MODEL CB16E, WARING COMMERCIAL, USA) and homogenized and sieved using a screening machine (CISA model BA-200N) to reach the required particle size in diameter, upper than 50 μm. The characteristics of BWBW are given in Table 1.

The moisture content of BWBW was determined according to the EN 1097-5 test technique. Samples were weighed before and after they have been dried in an muffle furnace at 95 °C for 24 hours (Nabertherm GmbH, Germany). Total sugars characterization was performed using the phenol sulfuric acid method (Dubois et al. 1956).

Pretreatments are needed to improve the digestibility of solid content. They significantly affect the further steps of global conversion scheme, particularly, the organic matter hydrolysis. In such a case, simultaneous thermal, chemical and enzymatic pretreatment processes have been performed to maximize the yield of reducing sugars (RS) from BWBW.

The hydrolysis process was carried out using a lab glass bioreactor with a total volume of 500 mL. To heat and monitor the operating temperature during the hydrolysis, the bioreactor was equipped with an integrated heating element and a thermostat bath (Witeg Labortechnik GmbH; MSH-20D, Germany). Note that the combined pretreatment processes were carried out for 80 min. The experiments were performed for various pH values, which were adjusted by adding NaOH (1N) or HCl (1N). Enzymatic pretreatement process of BWBW was carried out using alpha-amylase derived from Aspergillus oryzae (A8220-50ML, Sigma-Aldrich, Denmark). The amounts and the ratios selected for thermal, chemical and enzymatic pretreatment processes are given in Tables 2 and 3.

where y is the predicted response (Total sugars yield, %), x is each factor (substrate loading, alpha-amylase enzyme concentration, temperature and pH), was the average value of the response at the central point of the experimental design, ⁠, ⁠, are, respectively, the coefficient of the linear, quadratic and interaction effects.

The development of an efficient process requires the selection of suitable microorganisms. In the case of ethanolic fermentation, Saccharomyces cerevisiae remains the most commonly used yeast due to its ability to ferment a range of different sugars into ethanol and to enhance ethanol yield (Bai et al. 2008). A strain of VdH2 of S. cerevisiae was selected for bioethanol fermentation. S. cerevisiae was activated in yeast peptone dextrose (YPD) (80 mL) medium containing (g/L): yeast extract, 10; peptone, 20; and glucose, 20. The yeast was stored in an incubator at a temperature of 35 °C and speed agitation of 120 rpm for 16 h prior to use for bioethanol production.

The fermentation medium (g/L) contained yeast extract, 6; KH₂PO₄, 5; MgSO4, 1; (NH₄)₂SO₄, 2. The pre-cultivated cultures were transferred to the fermentation medium at an inoculum size of 10% (v/v) for ethanol production. All media and flasks were sterilized by using an autoclave at 121 °C for 15 min.

The optimized BWBW hydrolysate was used for ethanol fermentation. Experiments were carried out in a cylindrical stainless steel batch reactor of 16 cm height and 13.5 cm diameter with a working volume of 1.6 L. The reactor cover was equipped with 2 ports of 0.5 cm in diameter. One serves to collect the samples and the other one for gas output as shown in Figure 1. The mixing inside the bioreactor is ensured using the anchor impeller of 2 cm diameter, which was located at a distance about 1/3 above the bottom. This impeller rotates at 150 rpm, inducing tangential flow inside the fermenter. The ratio of bioreactor diameter to impeller one was about 6.75 cm.

80 mL of inoculum and medium were added into the reactor. The temperature was maintained at 35 °C, the agitation speed was set to 150 rpm, and the fermentation process was carried out for 72 hours. The samples were withdrawn to measure the RS, cell viability and ethanol concentration. The tests were performed in triplicates and the results provided were the average values.

Total sugars were measured by the phenol sulfuric acid method using a UV-VIS Spectrometer (PG Instruments limited, MODEL: T60UV) (Dubois et al. 1956). Total Kjeldahl nitrogen (TKN) and TSS were measured according to Standard Methods (APHA 1998). The pH value was measured using an Inolab (Prolabmas, Murni Swadaya, Jakarta, Indonesia) multi-parameter 720 device. The bioethanol composition was measured using a gas chromatograph (GC) (SHIMADZU GC-2014). The GC was fitted with an Rtx column (1–30 m, 0.25 mm ID, 0.25 μmdf). Nitrogen was used as carrier gas at a flow rate of 1 mL/min. The oven temperature was set to 62 °C with a rate increase of 25 °C/min until the temperature of 120 °C was reached. The injector and detector temperatures were set to 220 and 300 °C, respectively.

The validity of the adjusted model was assessed through analysis of variance (ANOVA). Regression analysis indicates that the model is highly significant, which is clearly indicated by Fisher's F-test (5.98) with a very low probability value (p model = 0.001) (Table 4). The high F values and low p-value (<0.005) trumpeted as significant (Montgomery 2017). R² was evaluated as 0.86, which mean a good regression model performance and a good degree of correlation between the experimental and calculated values.

Evaluating the importance of individual parameters was performed by calculating the p-value (Table 4). It can be found that the bread ratio takes the value of (0.008), alpha-amylase is about (0.033) and pH is about (0.005), which significantly affect the RS concentration. Conversely, the temperature has not a significant effect on the response, while the interaction effect of temperature and pH significantly impact the sugar concentration (0.069).

The interaction effects of the selected response (total sugars) were studied by plotting three-dimensional 3D response surface and contour plots. The interaction effects between two variables on the RS yield are given in Figure 2.

Figure 2(a) shows the interaction effects between waste bread ratio and alpha-amylase concentration on the RS yield, while the pH and temperature are kept constant at 6 and 60 °C, respectively. WB ratio has an effect on WB conversion into RS (p value = 0.008). When the BWBW ratio ranged from 0.03 to 0.08 g/mL, and alpha-amylase concentration was around 0.3 mL/L, the RS yield varied between 72 and 89%. However, after a new increase of substrate ratio to 0.1 g/mL, the RS yield decreased significantly (68%). In fact, the increase of substrate loading may cause less enzyme availability per substrate unit, non-specific enzyme binding increases medium viscosity, reduces both mass transfer and enzyme transport onto the surface of the starch granules (Demirci et al. 2017). Similar results have been reported previously by (Demicri et al. 2017), who observed that the increase of substrate led to the significant decrease of glucose yield. (Mihajlovski et al. 2020) reported that the increase of WB increases proportional to the RS to reach the maximum, further increases in WB concentration reduced sugars yield.

Besides, alpha-amylase concentration effect on RS yield is not significant as WB ratio (p value = 0.033), it is clear that the higher enzyme volume leads to the higher RS release from WB. Nevertheless, the concentration (0.2 mL/L) provided the maximum reducing sugars yield; therefore it is more interesting in view of the economic cost. However, some researchers (Fuji & Kawamura 1985) has outlined that without adding alpha-amylase during fermentation process, the RS yield is reduced. Similar enzyme volume was recommended by (Han et al. 2019), they suggested the use of 0.1 mL/L of enzyme volume for waste cake hydrolysis ever in practice.

Figure 2(b) shows the interaction effects between WB ratio and pH on RS yield while alpha-amylase concentration and temperature are kept constant at 0.6 ml/L and 60 °C respectively. It can be seen that the RS reached their maximum yield (<80%) when the bread ratio was around 0.3 g/mL and the pH ranged between 4 and 7. Furthermore increases or decreases of pH may caused decreased RS yield (Figure 2(b)).

It is can be seen that at pH 2 and pH 10, the RS yield decreased to (68 and 67%) respectively, this means that alkaline or acidic pretreatment conditions are not suitable for hydrolyzing bread waste. A good agreement was reached between our results and those found by (Ariunbaatar et al. 2014), these authors are emphasised that involving low or high pH values during chemical pretreatment of biomasses remains not suitable, due to the rapid substrates biodegradation that contain high amounts of carbohydrates, such as bread waste.

Figure 2(c) shows the interaction effects between pH and alpha-amylase concentration on RS yield while bread ratio and temperature are kept constant at 0.06 g/ml and 60 °C respectively. When the pH value of the medium was 6 and alpha-amylase concentration was around 0.3 ml/L, RS reached their maximum yield (Figure 2(c)) (Table 2). These findings are similar to those found by Hudečková et al. (2017). In their case the optimal conditions for alpha-amylase are at a pH value of 6 and temperature of 80 °C. RS yield reached was about 85.4% at (100 °C) (Table 2). The purpose of the thermal pretreatment is to modify the insoluble fraction structure to make it more biodegradable. Moreover, the thermal pretreatment process does not require acids or chemical compounds and it is carried out at temperatures ranging from 60 to 170 °C. It seems to be a cost-effective process and more environmentally friendly for converting biomass into biofuels (Hafid et al. 2017a, 2017b).

Validation experiments at the optimum conditions were conducted to maximize the sugars concentration. The optimum level of each factor was determined by means of response optimizer tool via MINITAB, it was found that the optimal bread ratio was 0.03 g/mL, in fact, while the WB ratio was between 0.03 and 0.08 g/mL, the RS yield varied from 72% to 89%. However, with an increase of the bread ratio to the value of 0.1 g/mL, the RS yield dropped significantly to 68%. Similar results were achieved by Demirci et al. (2017) and Mihajlovski et al. (2020), who found that the increase of WB concentration increased the RS yield until the maximum was reached. Furthermore, increases in WB concentration resulted in a fall in the RS yield. The optimal conditions of temperature and pH were at 100 °C and 5.3 respectively. These conditions are similar to those reached by Hudečková et al. (2017), they found that optimal conditions for alpha-amylase activity are pH 6; 80 °C. Han et al. (2019) recommended the use of alpha-amylase in a temperature range of 60–100 °C. The optimal concentration of alpha-amylase was about 0.2 mL/L. Similar enzyme volume was achieved by Han et al. (2019), who suggested the use of 0.1 mL/L of enzyme volume for waste cake hydrolysis in practice. The RS yield reached 90% under these conditions.

The experimental and simulated profile of the bioethanol concentration are depicted in Figure 3. The experiments were carried out at initial pH 4.5 and under agitation speed of 150 rpm at 30 °C for 72 h. These conditions promote S. cerevisiae growth (Mansouri et al. 2016). It can be observed that the bioethanol production started immediately during the first hours of fermentation. In such case, the metabolic activity of microorganisms increases too, leading to a faster conversion of BW into bioethanol.

After 30 h of ethanolic fermentation, the ethanol concentration decreases due to glucose depletion, ethanol oxidation and volatile fatty acids (VFA) accumulation (Shafaghat et al. 2010). Furthermore, the ethanol accumulation in the medium may have resulted in the deactivation of essential ethanol-producing enzymes such as pyruvate decarboxylase and alcohol dehydrogenase (Chohan et al. 2020). The ethanol concentration reached was about 85 g/L after a 72 h fermentation process. In such case, the bioethanol yield is close to the value of 0.40 g/g. These results are in agreement with those previously reported by Datta et al. (2018). These authors found that the bioethanol concentration produced from WB was about 54 g/L, with bioethanol yield about 0.37 g/g. However, bioethanol produced from the fermentation of enzymatic pretreatment process of bread waste resulted in the highest bioethanol concentration of 128 g/L (Pietrzak & Kawa-Rygielska 2015).

The simulated curve was estimated from a parameter-fitting process using the experimental data and Equation (2), which represents the modified Gompertz model. A good agreement is observed between this model and the experimental data, giving a good fit with R² > 0.97, and the RMSE value < 0.09. The kinetic parameters of the Gompertz model are given in Table 5.

It was found that the P-value obtained is about 81.1 g/L. It is worth mentioning that the P-value represents the bioethanol concentration that can be potentially produced after a given pretreatment process. Moreover, the estimated R_max (maximum bioethanol production rate) indicated that 2.1 g/L of ethanol was produced for every hour. Furthermore, a short lag phase of 2 h was observed for bioethanol production. This suggested that 1 h was required for the S. cerevisiae cells to adapt to the medium. Duration of lag phase observed in this study was required to hydrolyze starch into glucose (Chohan et al. 2020).

Total sugars provide a consistent source of energy to keep cells alive (Figure 4). The total sugars uptake has evolved exponentially within the 30 hours. This decrease in sugars matched well with the increase of bioethanol production. Then, the uptake of total sugars rates gradually decreased and dropped to low levels after 72 hours of fermentation. It is important to note that the total sugars have not been fully assimilated by the yeast, in such case the yeast growth was stopped due to the accumulation of toxic substances (Meyer et al. 1988).

To maximize RS yield from bread waste for bioethanol production, a combination of thermo-chemical pretreatments and enzymatic hydrolysis were investigated. The developed RSM models gave high coefficient of determination values (>0.86). Results showed that the optimal conditions for highest RS yield from waste bread were obtained at: bread ratio of 0.03 g/mL, alpha-amylase concentration of 0.2 mL/L, temperature of 100 °C and pH 5.3; the RS yielded a value of 90% under these conditions, leading to a bioethanol concentration of 85.8 g/L. These combined pretreatment processes use low energy and are cost effective, without sacrificing pretreatment efficiency. Furthermore, the modified Gompertz model provides a maximum bioethanol production rate, a maximum potential bioethanol concentration and a lag phase duration to the values of 2.1 g/L/h, 81.1 g/L and 2.2 h respectively. Experimental data obtained from this study provides substantial knowledge about the bioethanol production from food waste containing high cellulose and hemicellulose sugars.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.