The Influence of Process Parameters on Lactic Acid Fermentation in Laboratory Scale Fermenter

MATERIALS AND METHODS

Microorganism: Lactobacillus rhamnosus, a homo-fermentative bacterium to produce L-Lactic acid was used in this study.

Inoculum preparation: For the preparation of the starter culture, a single colony from Lactobacilli slant agar was transferred into a bijou bottle containing 10 mL culture medium and left overnight (24 h). After 24 h cultivation, 1 mL of lactobacillus liquid culture was pipetted into a bijou bottle containing 9 mL of media and incubated at 37°C for 5-6 h, the time needed for the bacteria to reach the exponential growth phase. A 10% inoculum grown in the medium was used in all fermentation processes. The 10 mL starter culture was then transferred into 250 mL shake flask containing 90 mL media. The culture was incubated in a rotary shaker set at 37°C using 150 rpm for 5-6 h. The 100 mL inoculum was then transferred aseptically into an inoculation flask for inoculation in bioreactor.

Table 1:	Reference condition for designed experiment in bioreactor
L: Low level and H: High level

Design of experiment: Optimization process had been designed by Taguchi Method using STATISTICA software (Version 6.0) where the three significant parameters (agitation speed of impeller (rpm), partial dissolved oxygen (pO₂) and pH) were set at various values. These parameters were observed to be the most significant parameters that affected the growth of Lactobacillus bacteria (Teuber, 1993). The experimental conditions were set according to Table 1.

Bioreactor operation: A 2 L Stirred Tank Reactor (STR), B-Braun fermenter, with batch mode of operation was operated at the different values of parameters using Taguchi method as shown in Table 1. The agitation speed and culture temperature were controlled at 150 rpm and 37°C. Due to the fact that the used strain is facultative anaerobic, there is no need for air-sparging. The medium of the cultures had the following composition: Peptone, 10.0 g L^-1; Yeast extract, 4.0 g L^-1; Glucose, 9.8 g L^-1; Lactose, 20.0 g L^-1; Tween 80, 1.0 mL; Potassium phosphate (K₂HPO₄), 2.0 g L^-1; Sodium acetate, 5.0 g L^-1; Ammonium sulfate ((NH₃)₂SO₄), 2.0 g L^-1; Magnesium sulfate (MgSO₄), 0.2 g L^-1 and Manganese sulfate (MnSO₄), 0.05 g L^-1, respectively. The culture pH in the fermenter was controlled at 6 by adding Sodium Hydroxide (NaOH) and Chloric acid (HCl). All the media were sterilized at 121°C for 20 min. The conditions of operating parameters were set accordingly. A cascade control method was applied in the operation where agitation was set up to 500 rpm for low level and 1000 rpm for high level. Air flow was not supplied. All of these parameter values were controlled through online once inoculation to the bioreactor takes placed. The pO₂, rpm and pH levels were monitored during the fermentation process. The fermentation process was conducted for 24 h and at different time intervals (every 30 min), sample was withdrawn and the OD, total cells number and the concentration of glucose and produced lactic acid were analyzed.

Analytical method: Optical Density Analysis (OD), Total Cell Number (TCN) and Cell Dried Weight (CDW) were analyzed using Maizirwan (2002) protocol.

Glucose and lactate analysis: One milliliter of sample was transferred into a 1.5 mL Eppendorf tube and centrifuged at 3000 rpm for 10 min. The supernatant was transferred into a cuvette and analyzed using the YSI Biochemical Analyzer.

RESULTS AND DISCUSSION

Growth kinetic study: Table 2 shows that L. rhamnosus cells grow fastest in Run 1 and 4. This is because of the suitable operational condition with acidic pH and enough pO₂ level that it may have. The growth rate has been correlated with the fastest doubling time, t_d which also occurred in Run 1 and 4. Between Run 1 and 4, production of lactic acid is much higher in Run 1. Therefore, Run 1 has showed the highest productivity which is 0.1282 g^-1 h.

Parameter influence study by Taguchi method: Taguchi method was employed as a tool for systematic experimental design. It allows several effects of parameters to be simultaneously determined effectively and efficiently. Taguchi’s approach complements two important areas: 1) Clearly defines a set of orthogonal arrays (OA) and 2) Devised a standard method for analysis of results. The combination of standard experimental design techniques and analysis method in the Taguchi approach produces consistency and reproducibility rarely found in any others statistical method (Roy, 1990).

After getting the results of cell concentration, substrate and product concentration for each run, those values were used in Taguchi method to analyze which parameters that most affect the lactic acid production. There are many types of analytical method that can be used to study the correlation of each parameter (under controlled) in bioreactor such as Signal-to-Noise (S/N) ratios, Larger-the-better, Smaller-the-Better etc. Here, S/N ratios analysis method was chosen. Analysis was done by selecting lactic acid production as dependent variable. S/N ratio is one of the analyses in Taguchi method which determined the most significant correlation between parameters. The change in the quality characteristics of a product under investigation in response to a factor introduced in the experiment design is the signal of the desired effect.

Table 2:	Summary of growth kinetics data for bioreactor fermentation

Table 3:	Effect of size for each parameter

Table 4:	Correlation between rpm and pO2

Table 5:	Correlation between rpm and pH

Table 6:	Correlation between pO2 and pH

It measures the sensitivity of the quality characteristic of the product which being investigated in a controlled manner, to the external influencing factors (noise factors) which is not under control (Ani et al., 2002). The S/N is simply the ratio of the mean to the standard deviation.

Main effect plot: As shown in Fig. 1 the pH has a bigger main effect than rpm and pO₂. That is, the line connecting the means responses for low pH and high pH, lower rpm and higher rpm have a steeper slope than the line connecting the mean responses at the low and high setting of pO₂. Although the pH appears to affect the lactic acid yield more than rpm and pO₂. Thus, the most significant parameter that affects the production of lactic acid according to this analysis is the pH (Table 3). This proven that the lower pH which was applied to Run 1 and 4 produced better production of lactic acid. In cascade control that has been applied in bioreactor fermentation, pO₂ was categorized as a master to the rpm where the rpm have to work in order to maintain the desired pO₂. Therefore, up and down the rpm level affected the pO₂.

Analysis of variance (ANOVA): ANOVA is the quantitative measure of the influence of individual factors/parameters. It is important for determining the relative importance of the various factors/parameters. Table 4-6 show that the influences of rpm to the pO₂ is about 0.097871 or 90%, rpm to pH is 83% and pO₂ to pH is 60%, respectively. This indicated that rpm and pO₂ exert the most significant effect on the production of lactic acid. The correlation between rpm and pH has low significant effect to the bioreactor operation, thus low benefit to the production of lactic acid and can be adjusted independently.

Fig. 1:	Means plot using S/N ratio

Fig. 2:	Online monitoring during Run 3 fermentation

Fig. 3:	Online monitoring during Run 4 fermentation

The combination of rpm and pO₂ has significantly affected the production of lactic acid using bioreactor. Therefore, good production of lactic acid is depending on controlling these two dependent parameters.

Bioreactor online monitoring: During the bioreactor fermentation, online monitoring using MFCS/Win has been carried out. Parameters of stirrer speed (rpm), pO₂ and pH have been monitored and recorded. Figure 2 and 3 showed the graphs of parameters monitoring for Run 3 and 4, respectively, that have been carried out for 24 h of fermentation. Between the four graphs, monitoring parameters for Run 3 and 4 showed the best. This is due to the ability of controlling the level of pO₂ at desired level without the stirrer exceeds its range which is 1-1000 rpm for both runs. For Run 3, the initial pO₂ level was set to be 10%, while pH level was maintained at 7. After certain time, when the concentration of the cells was increased, the pO₂ level was slightly dropped. At that time, stirrer was function to agitate the culture from slow to rapid agitation to support the pO₂ and maintained it at 10%. After the pO₂ level reached its initial concentration, stirrer decreased it rpm and maintained at the suitable value of pO₂ level required. This is called as cascade control where pO₂ just supplied by the rpm not from the sparger. As described earlier, Lactobacillus bacteria are species of facultative anaerobic. Therefore, pO₂ was the critical parameter that has to be maintained during the fermentation. High level of pO₂ is not suitable for this species of bacteria. As for Run 4, the function of cascade (rpm) was also shown. The stirrer speed was increased when pO₂ decreased and vice versa when it want to decrease the excess pO₂. As compared to Run 1 and 2 (figure is not shown here), a bigger fluctuation occurred to the rpm. It was exceed the range of 0-500. It can be said that pO₂ and stirrer speed relate closely to each other to maintain stable performance of bioreactor.

CONCLUSIONS

In conclusion, from the three parameters controlled during the fermentation in the bioreactor, pH with acidic condition showed the most influence factor and be the important parameter to the cells to perform best growth. The process will become more productive if under the acidic pH condition, the pO₂ supplied which is closely related to the stirrer speed maintaining the suitable dissolved oxygen to the cells is controlled under suitable level. High agitation of the impeller did not affect the cells strain because this cell is not shear sensitive and also a gram positive bacterium.

ACKNOWLEDGMENT

The authors would like to thank Research Centre of the International Islamic University Malaysia (IIUM) for funding this research under Project No. IIUM 504/022/3/LT 27.