Modelling Based Analysis and Optimization of Simultaneous Saccharication and Fermentation for the Production of Lignocellulosic-Based Xylitol

. 13 Simultaneous saccharification and fermentation (SSF) configuration offers an efficient used of the reactor. In this 14 configuration, both the hydrolysis and fermentation processes are conducted simultaneously in a single bioreactor 15 and the overall process may be accelerated. Problems may arise if both processes have different optimum 16 conditions, and therefore process optimization is required. This paper presents the development of mathematical 17 model over SSF strategy implementation for producing xylitol from hemicellulose component of lignocellulosic 18 materials. The model comprises of the hydrolysis of hemicellulose and the fermentation of hydrolysate into 19 xylitol. The model was simulated for various process temperature, prior hydrolysis time, and inoculum 20 concentration. Simulation of the developed kinetics model shows that the optimum SSF temperature is 36 o C, 21 whereas conducting a prior hydrolysis at its optimum hydrolysis temperature will further shorten the processing 22 time and increase the xylitol productivity. On the other hand, increasing the inoculum size will shorten the 23 processing time further. For an initial xylan concentration of 100 g/L, the best condition is obtained by performing 24 21-hour prior hydrolysis at 60 o C, followed by SSF at 36 o C by adding 2.0 g/L inoculum, giving 46.27 g/L xylitol within 77 hours of total processing time.

4 Xylitol can be used as a sweeter; it is commonly used in food industries and is categorized as safe for diabetics 5 [6][7][8][9]. Other than that, xylitol is used as a building block for ethylene glycol and propylene glycol formation using 6 Ruthenium or copper as the catalyst in hydrogenolysis process [10,11] and as a building block for 2,3,4,5 7 tetrahydroxypentanoic acid and xylonic acid using diperiodatoargentate (III) and Ru (III) as the catalyst through 8 oxidation [12,13].

10
Lignocellulose-based materials are pretreated and hydrolyzed using dilute sulfuric acid, after which the 11 hydrolyzate is purified using chromatography to obtain xylose. Pure xylose solution is then catalytically 12 hydrogenated using Raney-Nickel or Ru/C (Ruthenium-carbon) as the catalysts so that it becomes xylitol [14,15].

13
This conventional xylitol production method has several disadvantages: the process uses a lot of energy as it is 14 conducted at relatively high pressure and temperature (50-60 bar and 140-200 o C); the process requires delicate 15 purification of the hydrolysate to obtain pure xylose; and need more investment in types of equipment, 16 considerable intermediate purification, product recovery, catalyst deactivation, and recycling process [15][16][17]. An 17 alternative method of producing xylitol from lignocellulose-based materials involves a bioprocessing system that 18 includes enzymatic hydrolysis using xylanase to obtain xylose containing hydrolysate, and microbial fermentation 19 to convert xylose in hydrolysate into xylitol [16,18,19]. The hydrolysis and fermentation processes are normally 20 conducted in different reactors or better known as separate hydrolysis and fermentation (SHF) because they have 21 different operating conditions. As a result, the fermentative sugar production process requires a long processing 22 time and is still considered uneconomical.

23
Alternatively, the bioprocess route for xylitol production, that is the hydrolysis and the fermentation, can be 24 conducted simultaneously or better known as simultaneous saccharification and fermentation (SSF). In principle,

25
both the hydrolysis and fermentation processes take place simultaneously in the same reactor, providing direct 26 utilization of hydrolysis product, that is the monomeric sugar, as the carbon source for the fermenting agentform 27 the desired product [19][20][21]. Consequently, both processes are conducted at the same operating condition.

28
Moreover, the application SSF process from lignocellulose-based materials minimizes the potential substrate inhibition on the fermentation process as well as the potential product inhibition on the enzymatic hydrolysis 1 process, increases the yield and productivity [21][22][23].
2 Previous studies showed that the use of SSF combined with prior hydrolysis in ethanol production was able to 3 increase the ethanol yield than the SSF only [24,25]. Burhan et al. [19] reported the implementation of SSF for 4 xylitol production from oil palm empty fruit bunch (OPEFB). In his research, the duration of the prior hydrolysis 5 process was varied to achieve the optimum results. Overall, at the same total processing time, up to 4-fold increase 6 of the xylitol yield from OPEFB when compared with that of SHF was obtained [19]. Prior hydrolysis is necessary 7 to provide sufficient substrate for initializing the fermentation. However, the optimum temperature for conducting 8 both hydrolysis and fermentation simultaneously has been overlooked, the SSF was conducted at the optimum 9 fermentation temperature that led to low enzymatic-hydrolysis activity. The optimum temperature of xylan 10 hydrolysis using xylanase is reported in the range of 40-70 o C [26,27], whereas the optimum temperature for 11 xylitol-producing yeast fermentation is reported in the range of 10-44 o C [28][29][30][31]. Conducting the SSF at the 12 optimum temperature for both hydrolysis and fermentation would significantly increase the process performance.

13
Followingly the temperature setting for SSF as well as the duration of the prior hydrolysis process can be 14 optimized further to higher xylitol productivity.

15
Modelling may serve as a useful tool to explore the interaction of between parameters. In particular kinetic 16 modelling can be applied to search for the optimal SSF configuration, that is the operating temperature, the 17 duration of the prior hydrolysis or the initialization of the fermentation based on certain initial cell concentration.

18
This paper presents the development of a kinetic model for lignocellulosic material-based xylitol production using 19 SSF. The model was further used for thoroughly studying the effect of process temperature, prior hydrolysis and 20 switching time (the start of SSF) as well as the inoculum size to estimate xylitol concentration and its productivity 21 so that the results could feasibly be applied on a laboratory level and eventually developed on both a pilot and an  The model was built by assuming that the process took place in a single batch reactor. The xylan-based 26 hemicellulose in lignocellulosic material was pretreated before being put into the reactor for enzymatic hydrolysis 27 and fermentation. Xylitol production through the bioprocessing pathway is shown in Fig 1. During the hydrolysis 28 process, xylan was hydrolysed into xylose. During the fermentation process, xylose was further utilized by the yeast as the carbon sources for biomass growth and xylitol formation. In this model it was assumed that xylitol 1 was the only metabolite product that was produced during the fermentation and that a decrease in cell 2 concentration was neglected. Overall, the mass balances describing the process are presented in Eq. 1 -4.
Like other chemical reactions, a higher temperature can increase the rate of enzymatic reactions. However, higher 10 temperature also raises the rate of thermal denaturation and the loss of the biocatalyst activity [32,33]. Within the 11 range of 40-60C, however, the overall rate of xylan hydrolysis using xylanase is still increasing with temperature

12
[26] and the overall effect of temperature on this enzymatic reaction can be modelled following the Arrhenius 13 equation as is presented in Eq. 7.
0 is the concentration of the initial concentration of enzyme used (g/L), ℎ is the catalytic constant of 15 hydrolysis (ℎ −1 ) , ℎ is the Arrhenius constant for the hydrolysis reaction (ℎ −1 ), ,ℎ is the activation energy 16 of the hydrolysis reaction (kJ/mol), is the universal gas constant (kJ/mol.K), and ℎ is the temperature of

Biomass growth
1 The rate of biomass growth is defined following the first order reaction kinetics with respect to the biomass 2 concentration ( , g/L) (Eq. 8), whereas the biomass specific growth rate (µ, ℎ −1 ) is defined following the Monod 3 equation that correlates the the specific growth rate with substrate (xylose) concentration ( , g/L).
In which , is the growth saturation constant on xylose (g/L) and , is the maximum specific growth 5 rate of the fermentation process (ℎ −1 ). Further, the effect of temperature of biomass growth can be modelled 6 following Sanchéz et al.
[29] as: Where , is the cell activation energy for growth (kJ/mol), , is the deactivation energy when the cell has 8 entered the death phase (kJ/mol), is the cell activation coefficient (ℎ −1 ), is the cell inactivation 9 coefficient (ℎ −1 ). Other effects of microenvironment conditions, such as the oxygen concentration in the 10 fermentation broth or the acidity level of the media, were not considered in this model. Combination of equations

14
Xylitol production rate is modelled using the growth-associated product. The equation is approached as follows [32]: Combination of equations (9), (10), and (12) are given as follows.
Where / is the yield of xylitol formed from biomass activity.

Determination of cell and xylitol productivity
1 After the xylose was completely converted, we could measure how much cell and xylitol productivity obtained in 2 each configuration process. The cell productivity is determined as follows (Eq. 14).
Where and are cell productivity (g/(L.h)) and total processing time (h), respectively. Also, xylitol 4 productivity is defined as follows: Where is the xylitol productivity (g/(L.h)).

8
The simulation started with the hydrolysis of xylan to xylose followed by the fermentation of the xylose to xylitol.

9
The initial concentration of xylan was set at 100 g/L. Unless stated otherwise, the initial biomass concentration of 10 0.5 g/L was introduced at the beginning of the fermentation process. The maximal total process time was set to 11 be 200 hours.

12
All simulation processes were stopped when the xylitol reached the maximum concentration in each process and 13 the other compounds were considered not to interfere with the process.

19
Xylitol production via the SHF system was simulated as the reference. The SHF was conducted at the optimum 20 temperature for each process, as have been calculated in the previous section (Fig 2). During this simulation, the 21 hydrolysis was set to proceed at 60 o C whereas the following fermentation was set to proceed at 34 o C. The

22
temperature switch was assumed to occur instantly. Fermentation was initiated shortly after the completion of 23 xylan hydrolysis, that was at 99% xylan conversion. The results of the simulation of xylitol production using the 24 SHF method are shown in  were 21.85 g/L and 46.07 g/L, successively, which were achieved at the 6 th days or 128 hours of processing time.
4 Dominguez et al. [38] reported that the xylitol production using 120 g/L synthetic substrate by using 1.2 g/L initial 5 yeast concentration gave xylitol and yeast concentration near to 80 g/L and 5 g/L, respectively, after 72-hour 6 fermentation. The fermentation was conducted at the optimal fermentation temperature and all xylose were 7 utilized within the observed fermentation time. This reported fermentation time was comparable to the time 8 required for consuming all xylose in the fermentation simulation, 80 hours (Fig 3).

11
During SSF, all components of the process: xylan as the substrate, the xylanolytic enzyme, and the biomass 12 inoculum (as the fermenting agent) were present in the bioreactor, such that the hydrolysis and fermentation 13 occured simultaneously. Two distinct strategies were evaluated: conducting SSF at the optimum hydrolysis

16
At the optimum hydrolysis temperature of 60 o C (Fig 4a), xylan was converted into xylose resulting an increasing product formation is observed (Fig 2). By the end of this simulation, only as much as 115.31 g/L of xylose was 21 formed.

22
At the optimum fermentation temperature of 34 o C (Fig 4b), hydrolysis proceeded slowly. Xylan was slowly 23 hydrolyzed and would be completely hydrolyzed at 110 hours. Although the biomass inoculum was already 24 present from the start of SSF, the low xylose concentration led to slow biomass growth. Significant biomass 25 concentration was only observed after 43 hours and the biomass reached stationary phase when the substrate 26 exhausted. Xylose was completely utilized at 109 hours, resulting in xylitol and biomass concentration of 46.06 27 g/L and 21.83 g/L, respectively. Nevertheless, in comparison to the SHF process, the SSF method proceeded faster 28 to achieve the same xylitol concentration. Applying the fermentation optimum temperature is preferable than the 29 optimum hydrolysis temperature in SSF.

Simultaneous hydrolysis-fermentation at optimum SSF temperature 2
In determining the optimum temperature for SSF, simulations were conducted within the temperature range of 3 30-43 o C using Eq. 7, Eq. 11, and Eq. 13 for temperature-dependent hydrolysis, biomass growth, and xylitol 4 formation, respectively. Fig 5 showed the contour plot between temperature, the processing time, and the resulting 5 xylitol productivity. The lowest to the highest xylitol productivity was represented by the dark blue to dark red 6 contour area, respectively.

7
The maximum xylitol concentration can always be achieved at various SSF temperature. However, the processing 8 time required to achieve the maximum xylitol concentration varied with temperature, leading to a variation in the 9 xylitol productivity. The low enzymatic activity at the range of temperature of 30-43 o C resulted in slow xylose 10 accumulation. The low xylose concentration led to slow biomass growth and xylitol formation. For total 11 processing time under 60 hours, low xylitol productivity (< 0,15 g/(L.h)) was observed at the simulated 12 temperature range (Fig 5). An increasing trend of xylitol productivity was observed at processing time 60-100 13 hours. However, the xylitol productivity slowly decreased after 100 hours of processing time. The optimum xylitol

18
Previous research on SSF for xylitol production from OPEFB that was conducted at 30 o C with initial xylan 19 concentration of 10 g/L (assuming the average xylan composition in OPEFB was 22.5% [41]) for 120 hours, gave 20 xylitol productivity of 0.047 g/L [19]. Compared to this, this ideal simulation showed faster processing time and 21 resulted in higher xylitol productivity.

24
The performance of SSF may be improved by conducting a prior hydrolysis process, at the optimized temperature 25 for hydrolysis, before the initiation of SSF. The overall process, the combination of the prior hydrolysis process 26 and the SSF process is called as semi-simultaneous hydrolysis and fermentation (semi-SSF). In practice the 27 initiation of SSF can be set by the addition of biomass inoculum on various concentration. This event will be 28 referred as the switching time, in the remaining discussion.
The determination of the optimum switching time was conducted by varying the duration of prior hydrolysis, 1 ranging between 0 to 48 hours, at certain initial cell concentration. The results are shown in Fig 6, with the dark 2 blue to dark red colour denotes the lowest to highest xylitol concentration, respectively.

3
In general, an increase in the switching time resulted in longer total processing time required to achieve the 4 maximum xylitol concentration (Fig 6). The best configuration was obtained by prior hydrolysis time of 8 hours 5 led to total processing time of 93 hours to achieve the maximum xylitol concentration (Fig 6a). The later the 6 switching time, the lower the ability of the biomass to ferment so that productivity decreases. The obtained results 7 are consistent with previous study conducted by Burhan et al. [15], in which prior hydrolysis resulted in higher 8 xylitol concentration and productivity.

9
The overall processing time could be further improved by increasing the initial biomass concentration for the 10 fermentation, or in other words increasing the inoculum size added to the system (Fig 6b-c). Table 2 shows the 11 effects of initial cell concentrations on cell and xylitol productivities. In addition, the increase in the initial cell 12 concentrations shortened the total processing time despite of longer prior hydrolysis time. These results showed 13 that the fermentation process was the limiting factor of xylitol production. Increasing concentration of cell 14 inoculum is thus recommended to increase xylitol productivity and shorten the total process time.

17
Various process configurations for xylitol production have been simulated. We summarized the effect of process 18 configuration and process temperature to overall processing time, cell and xylitol productivities at the same initial 19 xylan and cell concentration (100 g/L and 0.5 g/L). The summary of cell and xylitol productivity and the time 20 required to achieve the maximum xylitol concentration are shown in Table 3. SSF at optimum condition 21 temperature provided better cell and xylitol productivities than SSF at the optimum fermentation temperature.

22
SSF with high initial cell concentration increased the cell and xylitol productivities further. The best SSF 23 configuration obtained in the simulations, that was 21 hours prior hydrolysis at the optimum hydrolysis 24 temperature followed by SSF at 36°C by adding cell inoculum up to 2 g/L, shortened the overall processing time 25 to achieve the maximum xylitol concentration to 77 hours, or 39.84%, when compared with the SHF. The overall 26 processing time for SHF was 128 hours whereas the overall processing time for the best SSF configuration was 27 77 hours.

28
The obtained results confirmed previous results of Burhan et al. [19] and Ӧhgren et al. [42] which produced xylitol 1 and ethanol using the SHF and SSF methods and obtained higher productivity results when using the SSF 2 configuration.

3
The results of SSF simulation showed that the optimum temperature, intermediate xylose formation-reduction, 4 xylitol formed, and biomass growth could be predicted adequately. Experimental validation through testing the 5 SSF temperature and the pre-hydrolysis time could be conducted further to confirm the accuracy of model 6 parameters that were used.

7
Overall, simulation of the developed kinetic model has been applied to design the configuration and the 8 operational of SSF process. The model could be further improved by the incorporation of non-ideal conditions 9 such as the inhibitory term to the hydrolysis process [43] or the inhibitory term to xylitol fermentation process [7, 10 44, 45] to give a more accurate estimation of the SSF process of lignocellulosic material. Details kinetics of the 11 related process as well as the estimated concentration of inhibitory substance in a specific process, for example 12 SSF of OPEFB, needed to be defined. Nonetheless, this paper showed that SSF or semi SSF is an alternative 13 process configuration that led to higher product (xylitol) productivity.

16
The kinetic model describing the SSF for xylitol production from hemicellulose of lignocellulosic material has 17 been successfully developed and simulated. Our simulation showed that the performance of SSF process was 18 affected by the process temperature, the length of prior hydrolysis or the switching time, and the biomass initial 19 concentration. Overall, it was concluded that the SSF configuration led to higher xylitol productivity than the 20 SHF. The best SSF configuration was combination of prior hydrolysis at the optimum hydrolysis temperature for 21 21 hour (semi SSF), SSF temperature of 36 o C, and initial biomass concentration of 2 g/L, which the led to an

Consent for publication
All authors confirm and consent to publish this manuscript.

3
The simulation data that support the findings of this study are available from the corresponding author upon 4 reasonable request.

7
The authors declare that they have no conflict of interest.     Table Captions 14 Table 1. Supporting data for modeling and simulation of xylitol production 15 Table 2. The effect of initial cell concentration in the inoculum to SSF for xylitol production 16 Table 3. Results of all configurations of processes of xylitol production Xylitol production by using the SHF method, both the hydroylysis and the succeeding fermentation were conducted at their optimum temperatures Xylitol production via SSF method conducted at (a) the optimum hydrolysis temperature and (b) the optimum fermentation temperature Figure 5 The results of SSF simulation for contour map showing temperature, process time, and xylitol productivity Figure 6 Contour map showing the effect of switching time to xylitol concentration at initial cell concentration of (a) 0.5 g/L; (b) 1 g/L; and (c) 2 g/L.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.