High-Temperature Fermentation and Its Downstream Processes for Compact-Scale Bioethanol Production

Abstract

High-temperature fermentation (HTF) of ethanol can reduce costs of cooling, sterilization, and related equipment compared to the costs of general ethanol fermentation. To realize HTF, however, there are various issues to be considered, such as the fermentation temperature upper limit for ethanol-producing thermotolerant yeast, the size of a fermenter that does not require cooling, and the effective temperature for suppressing microbial contamination. This study focused on these issues and also on downstream processes that exploit the advantages of HTF at temperatures exceeding 40 °C. The permissible size of a fermenter without cooling was estimated by simulating heat generation and heat dissipation. Fermentation productivity at high temperatures when using the thermotolerant yeast Kluyveromyces marxianus and the inhibitory effect of high temperatures on the growth of contaminant microorganisms were examined. After fermentation, the recovery and concentration of ethanol were performed by reduced-pressure distillation (RPD) and membrane separation. These experiments demonstrate that efficient HTF can reduce the amount of saccharifying enzymes in simultaneous saccharification and fermentation and can shorten the transition time from the saccharification step to the fermentation step in separate saccharification and fermentation, that RPD at fermentation temperatures enables a smooth connection to the HTF step and can be performed with a relatively weak vacuum, and that membrane separation can reduce the running cost compared to the cost of general distillation on a compact scale.

1. Introduction

As global warming progresses, urgent measures are needed to significantly reduce the use of fossil fuels such as oil and coal and replace them with biofuels in order to suppress CO₂ emissions [1,2]. Biofuels are produced from raw materials derived from plants that have fixed CO₂, so their fuel consumption is recommended as carbon neutral. Bioethanol, a type of biofuel, has been used in Brazil and the United States since the 1970s as an alternative to gasoline. Although the use of bioethanol began as a response to the oil crisis that occurred at that time, bioethanol is currently produced and consumed in large quantities primarily from the perspective of energy security. On the other hand, in many countries, food residues and agricultural residues containing large amounts of sugars are being discarded [3,4,5], and the challenge is to effectively utilize the sugar content rather than simply disposing of it by incineration or composting. Turning these residues into biofuels is expected to contribute to the reduction in CO₂ emissions [6]. However, these residues are generated in scattered locations, and it is difficult to collect them in large quantities. Converting the residues into biofuel and consuming them locally or on-site would be cost-effective. To achieve this, biofuel will need to be produced on a more compact scale than in the past, and in order to ensure profitability on that scale, new bioethanol production technology and a system for utilizing the fermentation residue are required.

Ethanol that can be used as a biofuel is generally produced through a multi-step process including pretreatment, saccharification, fermentation, solid–liquid separation, distillation, and dehydration [7,8]. There are mainly two types of the second and third steps: separate hydrolysis and fermentation (SHF) and simultaneous saccharification and fermentation (SSF) [9,10]. In SHF, saccharification and fermentation are performed separately at their respective optimal temperatures, whereas in SSF, they are performed together at a lower temperature that is far from the optimal temperatures of the saccharifying enzymes. In the presence of sufficient amounts of enzymes, the time required for SSF is shorter than that for SHF. In high-temperature fermentation (HTF), which is expected to have several advantages compared to general low-temperature fermentation [10,11], the amounts of enzymes required can be reduced because SSF is carried out at a temperature about 10 °C higher than the temperature in general fermentation [12]. Additionally, although the ethanol fermentation reaction is exothermic, cooling water circulation is not required in HTF if the size of the fermenter is designed to keep the heat transfer from the fermenter to the air larger than the generated heat by the fermentation, and thus the costs of maintaining the appropriate temperature for the fermenting yeast, such as the costs of cooling equipment and operating power, can be reduced. Furthermore, ethanol can be easily recovered by reduced-pressure distillation (RPD) [11], which can be carried out using relatively simple equipment at almost the same temperature as that for HTF immediately after fermentation. In this way, by linking HTF and RPD, it is expected that the process of saccharification, fermentation, solid–liquid separation, and simple distillation can be shortened and can proceed continuously. In addition, the further coupling of membrane separation may enable the production of biofuels in compact-scale processes up to the step of dehydration [13]. In a typical ethanol factory, a distillation facility consisting of two towers is used to recover ethanol from the fermentation liquid and concentrate it to about 95% (v/v). The first tower separates the ethanol from solids, and the second tower concentrates the ethanol. RPD and membrane separation serve as the first and second towers, respectively. Since operating the distillation facility costs more than half of the total operating cost, the introduction of membrane separation is expected to be economical [13].

Microorganisms with excellent thermoresistance and high ethanol productivity are essential for HTF [10,14,15,16,17,18]. In addition, microorganisms with a wide substrate spectrum that can utilize various types of biomass and with strong resistance to the toxic by-products present in fermentation broths are also suitable. Kluyveromyces marxianus has excellent characteristics and potential for HTF in addition to strong thermotolerance [18]. It has broad substrate specificity and a high growth rate [19,20,21], and it can utilize pentoses such as xylose, disaccharides such as sucrose and lactose, trisaccharides such as raffinose, and polysaccharides such as inulin in addition to hexoses [10,22,23]. Furthermore, there have been many studies on fermentation using K. marxianus, including adaptive evolution for stress resistance toward industrial applications [23,24,25], and its usefulness has been demonstrated. However, the application of HTF and connection to downstream processes toward compact-scale biofuel production are limited.

In this study, in order to establish an economical- and compact-scale bioethanol production process, we investigated the process from HTF using K. marxianus to membrane separation via RPD. First, to determine how high the growth rate and fermentation productivity can be maintained up to high temperatures, we examined the growth rate and ethanol productivity at high temperatures. The latter was tested by SHF and SSF using rice as a raw material. At a high temperature that has no negative effect on growth and fermentation, the size of a fermenter that does not require cooling was estimated by simulating heat generation and heat dissipation during fermentation. After fermentation, ethanol was recovered directly from the fermentation tank by RPD, and further concentration of ethanol was performed by membrane separation. Additionally, the effect of high temperatures in the range of HTF on the growth of contaminating microorganisms using model food residues was examined.

2. Materials and Methods

2.1. Strain and Media

The yeast strain used in this study was K. marxianus DMKU 3-1042 [18], which was deposited in the NITE Biological Resource Center (NBRC) under the deposit number NBRC 104275. Cells were grown in a YP (1% yeast extract and 2% peptone) medium containing 2% glucose (YPD medium) as a pre-culture. A YP-rice medium containing 10% rice (w/v, dry weight) (YPR medium) was used for SSF and SHF in a 10 L fermenter. Cell growth in the rice medium was not determined due to the strong turbidity of the medium. All media were autoclaved before use.

2.2. Determination of Specific Growth Rates at Different Temperatures

One isolated colony was obtained from a YPD agar plate, inoculated into 3 mL of a YPD liquid medium, and incubated at 30 °C and 160 rpm for 18 h. The cell culture suspension was inoculated into 30 mL of the YPD medium to adjust OD₆₆₀ to 0.1, and incubation was carried out at 30 °C and 160 rpm for 4 h. The cell culture suspension was inoculated into 30 mL of the YPD medium to adjust OD₆₆₀ to 0.1, and incubation for the determination of the specific growth rate was performed at different temperatures from 30 °C to 45 °C and 160 rpm for 4 h. The turbidity at OD₆₆₀ was measured every 1 h to determine the specific growth rate. Experiments were carried out in triplicate.

2.3. Ethanol Production by SSF and SHF with Rice as Biomass Followed by RPD

One isolated colony was obtained from a YPD agar plate, inoculated into 3 mL of YPD medium, and pre-cultured at 30 °C and 160 rpm for 8 h. Five hundred microliters of the pre-culture was inoculated into 50 mL of the YPD medium, and cultivation was conducted at 30 °C and 160 rpm for 18 h. In SHF, 5 L of the YPR medium that was sterilized by an autoclave and Uniase S containing α amylase and glucoamylase (2000 U/g rice; Yakult Pharmaceutical Industry Co., Tokyo, Japan) was added to a 10 L fermenter (NBC-10000, Mitsuwa Frontech, Osaka, Japan), and then saccharification was carried out at 50 °C and 180 rpm for 18 h. After that, 50 mL of the pre-culture was added, and then fermentation was performed at 40–45 °C and 180 rpm for 20 h. In SSF, a sterilized YPR medium, Uniase, and 50 mL of the pre-culture were added to the 10 L fermenter, and then saccharification and fermentation were performed at 40–45 °C and 180 rpm for 20 h. For distillation, a vacuum pump (V-103, Nihon BUCHI, Tokyo, Japan), a cooling unit (CCA-1112A, EYELA, Tokyo, Japan), and a glass condenser unit (Mitsuwa Frontech) were used. These parts were directly connected to the 10 L fermenter, a system similar to a system that had been used for ethanol recovery after SHF [11]. After fermentation, the vapor pressure of the fermenter was reduced to 200–250 mbar, and the temperature of the cooling unit was set at 4 °C. The fermenter was set at 45 °C and 100 rpm and ethanol solution was collected for 6 h. Experiments were carried out in triplicate.

2.4. Simulation of Heat Generation and Heat Dissipation During Fermentation

The fermenter is assumed to be a cylindrical tank, and the liquid depth is 1.3 times the tank diameter. The heat released from the fermenter is assumed to be only from its lateral surface. The fermenter capacity V (Equation (1)) and heat transfer area A (Equation (2)) are determined by the tank diameter and liquid depth:

$$\mathrm{V}={\displaystyle \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$4$}\right.}\mathsf{\pi }{\mathrm{D}}^2\times 1.3\mathrm{D}$$

(1)

$$\mathrm{A}=\mathrm{\pi }\mathrm{D}\times 1.3\mathrm{D}$$

(2)

The heat released from the fermenter Q_loss can be estimated using the following formula (Equation (3)), where U (15 kcal/m²/h/°C) is the overall heat transfer coefficient due to natural convection, T_in is the temperature inside the tank, and T_out is the ambient temperature:

$${\mathrm{Q}}_{\mathrm{l}\mathrm{o}\mathrm{s}\mathrm{s}}=\mathrm{U}\mathrm{A}({\mathrm{T}}_{\mathrm{i}\mathrm{n}}-{\mathrm{T}}_{\mathrm{o}\mathrm{u}\mathrm{t}})$$

(3)

On the other hand, assuming that the fermentation heat Q_gen is 11.95 kcal/mol-EtOH, the ethanol (MW = 46.07) concentration increases from 0% to 5% in 24 h, the fermentation rate (the ethanol production rate) is constant during that time, and the heat of fermentation per hour is calculated by the following formula (Equation (4)):

$${\mathrm{Q}}_{\mathrm{g}\mathrm{e}\mathrm{n}}=\mathrm{V}\times 1000\times 0.05/46.07\times 1000\times 11.95/24$$

(4)

2.5. Membrane Separation

MOR zeolite membranes were prepared on mullite porous tubular supports (an o.d. of 12 mm, i.d. of 10 mm, length of 100 mm, and pore size of 1.3 μm) purchased from Nikkato Co., Tokyo, Japan, by a secondary growth method. Detailed procedures can be found elsewhere [26]. The prepared membranes were characterized by X-ray diffraction (XRD, Rigaku Smartlab, Cu Kα radiation).

Membrane separation tests were performed by pervaporation (PV) with a feed solution being supplied to the outer surface of a zeolite membrane at a 66 g/min flow rate by a pump. Retentate, the fraction of the feed solution that was not permeated through the membrane, was circulated back to the feed tank. The inner side of the membrane was maintained under vacuum. The vapor that permeated through the membrane was collected by liquid nitrogen. The whole setup was heated at 60–70 °C. The concentrations of ethanol in the feed tank and in the permeate were measured by gas geometrography (GC-8A, Shimadzu Co., Kyoto, Japan). The flux through the membrane was calculated from the collected mass at the liquid nitrogen trap. As a feed solution, 130 g of bioethanol solution obtained by RPD was used.

2.6. Measurement of Glucose and Ethanol Concentrations in the Culture Liquid

Glucose and ethanol concentrations in the culture liquid were determined by High-Performance Liquid Chromatography (HPLC) (Hitachi High-Tech, Tokyo, Japan). Samples were collected from the culture liquid, passed through a 0.45 μm filter (Nihon poll, Tokyo, Japan), and subjected to HPLC analysis. The HPLC system consisted of an L-2130 Pump, L-2490 Refractive Index Detector, L-2200 Autosampler, L-2350 Column oven, and Hitachi Model D-2000 Elite HPLC System Manager, equipped with a GL-C610-S Gelpack^® column (Hitachi Chemical, Tokyo, Japan) using distilled water from an RFD240NA Water Distillation Apparatus (Aquarius, ADVANTEC^®, Tokyo, Japan) as a mobile phase at a flow rate of 0.3 mL/min.

2.7. Determination of Cell Numbers of Contaminating Microorganisms in a Non-Sterile Medium

A model kitchen refuse (MKR) medium was prepared according to a previous report [27]. Kitchen refuse was mixed with a YP medium and ground using a kitchen mixer until large particles disappeared. The resultant solution was called the MKR medium, which contained 14% (w/w) cooked meat, 40% vegetables (a peel of carrot, potato, and Chinese radish), 30% (w/w) fruits (apple, banana, and orange peel), 10% (w/w) cooked rice, and 6% (w/w) green tea residue. To determine the cell numbers of contaminated microorganisms, an MKR medium that had not been sterilized was spread on YPD plates after being appropriately diluted. The plates were incubated at different temperatures for 12 h and the number of colonies was counted. Experiments were carried out in triplicate.

3. Results

3.1. Optimal Temperature for Specific Growth Rate of K. marxianus DMKU 3-1042

K. marxianus DMKU 3-1042, which is one of the most thermotolerant yeasts [18,28], was used for this study. As mentioned above, it is more economical for HTF to be close to the optimum temperature of the saccharifying enzymes, but it must be in a temperature range in which the fermenting yeast grows well and the fermentation rate is high. In order to determine the suitable temperature range for the growth of yeast, we first compared specific growth rates at different temperatures. DMKU 3-1042 cells were grown in a YPD medium at 30–45 °C under a shaking condition at 160 rpm, and the specific growth rate at an early growth phase, for 4 h after the initiation of cultivation, was determined (Figure 1). The plots in the figure indicate that the optimal temperature for the specific growth rate is 38–40 °C under the conditions tested. Above this temperature range, the specific growth rate gradually declined. DMKU 3-1042 can grow even at 49 °C on plates [28], but it was found that the growth rate slows down when the temperature exceeds 40 °C. The specific growth rate is directly proportional to the growth. The low specific growth rates at lower temperatures may be due to a low affinity for nutrients [29], and the low specific growth rates at higher temperatures may be due to damage by the accumulation of toxic compounds including reactive oxygen species [25].

3.2. Ethanol Production at High Temperatures Using Rice as a Raw Material

Since RPD experiments as described below were conducted in a 10 L fermenter that can withstand low pressure, we conducted fermentation experiments at 40 °C, which was found to be the optimum temperature for the specific growth rate, or above 40 °C using a 10 L fermenter. SHF and SSF were carried out using rice as a raw material. In SHF, since saccharification was completed before the start of fermentation, fermentation was initiated under conditions in which approximately 7.5% of glucose was present. Ethanol concentrations reached a maximum within 20 h at all temperatures tested. Figure 2a shows ethanol concentrations in the fermentation liquid at 20 h of fermentation. Both SHF and SSF showed the highest ethanol concentrations from 40 °C to 42 °C in the temperature range tested, but the ethanol concentrations gradually decreased at temperatures above 43 °C. Ethanol concentration was slightly higher with SSF than with SHF at temperatures tested. The reason for this is not clear, but the initial concentration of glucose may have a negative effect on fermentation in SHF. From the results, it was found that temperatures up to 42 °C are suitable for HTF.

A small amount of glucose remained at 20 h, and its residual amount increased with an increase in the temperature above 42 °C in both SHF and SSF (Figure 2b). Consistent with this, the residual amount of glucose increased at temperatures above 42 °C when examined in YP containing 16% glucose [25]. It is likely that such high temperatures become a critical stress, producing toxic compounds including reactive oxygen species, to hamper the growth and fermentation ability of the yeast. In addition to ethanol as a main product, acetic acid and other organic acids are also generated as by-products, and these may become stresses [25]. The high concentration of sugar is also a stress, inhibiting yeast growth and affecting the total ethanol production [25].

3.3. Estimating the Size of a Fermenter That Does Not Require Cooling in HTF by Simulating Heat Generation and Heat Dissipation During Fermentation

One of the benefits of HTF is that it reduces cooling costs. However, as the size of the fermenter increases, the amount of heat generated during fermentation exceeds the amount of heat released from the surface of the fermenter, and cooling becomes necessary. Therefore, in order to estimate the limit of the size of a fermenter that does not require cooling, we conducted a simulation of the amount of heat generated by the fermentation reaction and the amount of heat released to the outside of the fermenter (Figure 3). Considering the upper limit of the temperature for efficient ethanol production shown in Figure 2, the generated heat and released heat for the fermenter capacity were simulated under the conditions described in the Materials and Methods with a fermentation temperature and external temperature of 42 °C and 30 °C, respectively (Figure 3a). The cross point of the two lines was 14.7 m³, which can be considered as the limit fermenter capacity that does not require cooling. We also plotted the capacity of the fermenter at the cross point when the external temperature was changed (Figure 3b). When the fermentation temperature and the external temperature were set at 45 °C and 30 °C, respectively, the cross point of the two lines was 28.7 m³ (data not shown). These data indicate that the capacity of the fermenter that does not require cooling is greatly restricted by both the fermentation temperature and the outside temperature.

3.4. Ethanol Recovery from Fermentation Liquid by RPD

When the temperature of the fermentation liquid is 40 °C or higher, distillation is possible without a strong degree of vacuum, and it can be achieved with simple equipment as described in Section 2. An RPD equipment unit that was directly connected to a 10 L fermenter was used to recover ethanol from the fermentation liquid. RPD was carried out at 45 °C for 6 h immediately after the fermentation of SSF or SHF for 20 h. As a result, the concentration of recovered ethanol was 28 ± 5% (w/v), and the ethanol recovery rate from the fermentation liquid was estimated to be 75 ± 8%.

3.5. Ethanol Concentration Using a Zeolite Membrane

RPD can remove relatively large particles in the fermentation liquid to collect a clean ethanol solution, which is suitable for membrane separation. HPLC analysis showed that non-ethanol peaks were significantly reduced in the post-RPD solution compared to the pre-RPD solution. We thus further concentrated the ethanol solution collected by RPD using an MOR zeolite membrane, which allows water to pass through but does not allow ethanol to pass through (Figure 4). Accordingly, the ethanol concentration increased with separation time and reached over 90% after 25 h. The flux through the membrane decreased with time. This is because the decline in water concentration in the feed reduces the driving force of the permeation. Since the permeated fraction through the membrane was almost water, the ethanol recovery rate was 99.7%. This high recovery rate is one of the advantages of applying dehydration membranes instead of applying ethanol-selective membranes. An air sweep can be employed instead of liquid nitrogen, which makes the process simpler [13]. The required time to concentrate ethanol can be shortened by enlarging the membrane area. However, it may not be necessary to shorten the downstream process time as the fermentation itself takes time.

No changes were observed with an MOR zeolite membrane used to concentrate the RPD solution by XRD analysis performed before and after the test, suggesting the durability of an MOR membrane in an RPD solution. The water selectivity of the membrane measured with a synthetic mixture of 90 wt% ethanol and 10 wt% water before and after concentrating the bioethanol remained nearly identical. The flux, however, declined by approximately 17%, likely due to by-products plugging the pores or other factors.

3.6. Suppression of the Growth of Contaminating Microorganisms by High Temperatures

Another advantage of HTF is that it can suppress the growth of contaminating microorganisms that have negative effects on fermentation, enabling savings in cost and time for sterilizing raw materials for fermentation. Fermentation liquid contains an abundance of fermenting yeast cells, which are difficult to distinguish from contaminating microorganisms by simple experiments such as spreading it on plates, and we therefore examined the suppression of the growth of contaminating microorganisms in an unsterilized medium without the addition of fermenting yeast. A non-sterile MKR medium containing household garbage, which was prepared as described in the Materials and Methods, was spread on YPD plates and incubated at different temperatures for 48 h (Figure 5). The number of colonies formed at 30 °C was 1.2 × 10⁷ CFU/mL and the numbers of colonies formed at 40 °C and 45 °C were 1.0 × 10⁴ CFU/mL and 1.5 × 10³ CFU/mL, respectively. These findings suggest that HTF can strongly suppress the growth of contaminating microorganisms.

4. Discussion

Since economies of scale cannot be expected in compact-scale bioethanol production processes, the introduction of new technologies that can reduce costs at each step of the process and their combination are considered essential from a profitability perspective. Additionally, it is important that the connections between the steps in the bioethanol production process are smooth and require only low-energy input.

In this study, K. marxianus DMKU 3-1042 showed a high ethanol fermentation performance up to 42 °C in both SHF and SSF (Figure 2). In the case of SHF, it is necessary for HTF to decrease the temperature of the solution after saccharification from more than 50 °C to 42 °C, but for general fermentation using thermosensitive Saccharomyces cerevisiae, it is necessary to decrease the temperature to about 30 °C. Therefore, HTF can save the time and cost required to lower the temperature. In the case of SSF, since the difference in fermentation temperatures for HTF and general fermentation is more than 10 °C, larger amounts of saccharifying enzymes are required for general fermentation that proceeds in the same period of time, making HTF more economical. It was reported that an increase of 5 °C in the fermentation process reduces the cost of enzymes by up to 50% [12]. In general fermentation using S. cerevisiae, cooling is required to avoid fermentation failure due to a temperature rise by the fermentation reaction, but as shown in the simulation, HTF does not require cooling up to a certain scale (Figure 3). The capacity of the fermenter varies depending on the outside temperature as it depends on the heat dissipation of the fermenter. External cooling and cooling equipment are not required as long as temperatures are maintained within acceptable limits. At high temperatures at which HTF can be conducted, the growth of many mesophilic microorganisms that were present in the raw materials can be greatly inhibited (Figure 5). As long as raw materials for fermentation are managed so as not to be left at room temperature or outside for a long time, ethanol productivity may not be significantly affected without sterilization, and eliminating the need for sterilization allows for reductions in sterilization equipment, costs, and time.

Distillation under reduced pressure downstream from HTF is also relatively easy and almost without time loss since it can be carried out at almost the same temperature. Considering the vapor pressures of ethanol at 45 °C and 30 °C, the pressure at 30 °C must be reduced to at least half of the pressure at 45 °C. RPD at 45 °C is thus much easier than RPD at 30 °C. The latter requires more energy to cool the condenser in addition to the equipment that can ensure lower pressure. The ethanol recovered after RPD can be further concentrated by membrane separation (Figure 4). Membranes have a modular design and can be applied on both small and large scales [30]. The permeate side vacuum can be replaced by air sweep, which makes the process simpler and less energy-consuming [13]. So far, ethanol production from starch raw materials such as cassava starch [31], sweet potato [32], and corn [33] has been reported. In this study, rice was used as the raw material, but when using other starch raw materials, the properties may differ depending on each material. For example, when using potato starch, it is difficult to stir as it contains a large amount of fiber. However, basically, it is possible to use the technology introduced in this study if the necessary and sufficient amounts of liquefying enzyme (a-amylase) and saccharifying enzyme (glucoamylase) can be added to convert starch into glucose.

A large-scale ethanol production process generally consists of pretreatment, saccharification, fermentation, solid–liquid separation, distillation, and dehydration. The fermentation is carried out at low temperatures by cooling the fermenter. Membrane separation technology using an A-type zeolite membrane is used for the final dehydration (water content < 15%) step in some cases [13]. In this study, we investigated saccharification fermentation at high temperatures and with vacuum distillation and membrane separation, which may be suitable for a compact-scale ethanol production process as mentioned above. Although it may not be easy to assess the profitability of the latter process, we believe that the following cases are worthy of consideration. When using food waste or residues, including household waste, as a raw material, it is desirable to produce ethanol locally or as close to the site as possible in order to minimize transportation costs for raw materials and products. On the other hand, the final treatment of food waste using incinerators of local governments or other agencies requires large processing costs, and the use of food waste through ethanol production can lead to reductions in processing costs. The disposal of fermentation residue is also costly, but there is a potential for profits to be made if fermentation residue can be used as fertilizer or feed. Furthermore, converting food waste into biofuels leads to a reduction in CO₂ emissions. In any case, it is necessary to consider not only the economic efficiency of the fermentation production process but also the profitability based on these points.