Enhanced Biosurfactant Production by Bacillus pumilus 2IR in Fed-Batch Fermentation Using 5-L Bioreactor

The enhancement of the production of lipopeptide biosurfactant by Bacillus pumilus 2IR as an oil field-isolated bacterium was investigated in the fed-batch fermentation using a 5-L bioreactor. The initial study on the culture medium composition used for biosurfactant synthesis revealed that biosurfactant production was greatly affected by glucose, crude oil, potassium nitrate and ammonium sulfate as pivotal carbon and nitrogen sources which were utilized for biosurfactant synthesis in the bioreactor. The results obtained from the batch fermentation process in the bioreactor showed that the quantities of biomass and biosurfactant produced were 4.15 and 0.98 g/L, respectively. Similar results revealed that fed-batch fermentation led to the production of 5.71 g/L biomass and 1.06 g/L biosurfactant in the bioreactor.


Introduction
The term ''biosurfactant'' has been referred to an isolated or non-isolated compound obtained from microbial cells to affect surfaces and interfaces and reduce required efforts to overcome surface tension. This process allows one system to disperse into another (Amodu et al. 2014;Mulligan 2004;Sari et al. 2014). It is of industrial interest to develop the biosurfactant production in which the cost of raw materials and the process become minimal. Only if this requirement is met and the price of the biosurfactant becomes lower than that of chemical ones, the biosurfactant will have a chance to be used in a large scale for the industrial application. One of the most important strategies for increasing biosurfactant production by microorganisms is the modification of the fermentation process. For the development of any fermentation process, two main areas are normally considered, those associated with the strain development and those associated with the process development. The productivity and yield of the fermentation can be improved through the strain selection and development program (Avili et al. 2012;Cheniker et al. 2010). Improvement in the microbial medium and culture conditions is another important aspect to enhance the fermentation process. Another area associated with the process development includes the mode of reactor operation and fermentation technique. Since the industrial goal of study on the fermentation process is the cost-effective production of bioproducts, the fed-batch culture has been widely employed for the improvement in the production of various primary and secondary microbial metabolites, proteins and other biopolymers compared to the batch fermentation (Sivapathasekaran and Sen 2013;Oh et al. 2002).
From a biotechnological viewpoint, the production of the biosurfactants is important owning to their potential application in medicine, food, cosmetic pharmaceutical, agricultural and petrochemical industries (Joshi-Navare and Prabhune 2013;Ławniczak et al. 2013;Nitschke and Pastore 2006;Patel et al. 2015;Sharma et al. 2014). Generally, due to the commercial consideration and expanding industrial demands, the development of the biosurfactant application is focused mainly on the biosurfactant production quantity and the synthesis of the highly active biosurfactant for specific applications (Mukherjee et al. 2008).
In this view, the biosurfactant has been found an important application in the recovery of oil in the oil reservoirs. Many attempts have been made to use the microbial biosurfactant process for improving the oil production. For this aim, microorganisms with the ability to tolerate high salinity, temperature, pH and rock permeability are needed since salinity and temperature are high in the depth of the most oil reservoirs. These environmental variables in the oil reservoirs usually restrict the use of bacterial strains that can be applied for the oil industry such as improving the quality of the petroleum via biological method and microbial enhanced oil recovery (MEOR) (Armstrong and Wildenschild 2015;Sarafzadeh et al. 2014;Youssef et al. 2013).
Moreover, as the oil extracted from different areas has specific compositions and properties, its biological removal expenses are very high. Thus, indigenous microorganisms are able to reduce the recovery and remediation cost through the environmentally friendly removal of oil. A major challenge to the oil industry is the extraction of the maximum amount of oil from the reservoirs. To recover entrapped oil, costly tertiary methods (including chemical and thermal process) are applied (Oh et al. 2002). MEOR is an alternative tertiary oil recovery technology where microbial metabolites and activities are used to improve the recovery of residual oil from the depleted and marginal reservoirs, thereby extending their life. This technology takes advantage of the ability of indigenous or injected microorganisms to synthesize useful products by fermenting inexpensive raw materials (Pornsunthorntawee et al. 2008;Sarafzadeh et al. 2014). The use of exogenous microorganisms in ex situ methods of MEOR entails the increase in the costs of the oil recovery technology (Ghojavand et al. 2008;Matar et al. 2009).
Indigenous microbes of interest are often stimulated with inexpensive substances to produce and release biosurfactants through in situ method. In this view, it has been indicated that the stimulation of indigenous populations of the biosurfactant-producing bacteria would be among the most low-cost MEOR methods (Gaytán et al. 2015). Hence, the indigenous microorganisms that exhibit the ability to tolerate harsh conditions in the oil reservoir such as high salinity, temperature and pH with the lack of oxygen would be advantageous for efficient MEOR (Armstrong and Wildenschild 2015). On the other hand, biodegradation in soils contaminated with oil is limited due to the low solubility of these components and the high capacity of adsorption in soil and sediments. One way to increase the solubility of the oil is the utilization of the biosurfactant as a mobilizing agent (Ebadi et al. 2017;Sivapathasekaran and Sen 2013). Several technologies have already been developed and implemented for the remediation of soils and sediments in order to decrease oil recovery costs. In situ soil treatment by the biosurfactants is a preferable method which has a lower cost and is less destructive than ex situ bioremediation (Ławniczak et al. 2013;Patel et al. 2015).
The current knowledge about the production of the biosurfactant by indigenous microbes existing in the oil reservoir for in situ utilization in oil recovery is not sufficient. Little information is available concerning the on-site application of the biosurfactants in the oil reservoirs in contradiction to the exogenous microbes. In this study, lipopeptide biosurfactant production by Bacillus pumilus 2IR which was locally isolated from oil-contaminated soil in an oil field was investigated in batch culture using shake flasks. The comparison of the growth behavior and biosurfactant synthesis pattern of the culture in the batch and the fed-batch fermentation was carried out in a 5-L bioreactor to obtain the highest biosurfactant concentration. The effects of operating process parameters on the biosurfactant synthesis were also investigated.

Microorganism and Culture Conditions
The strain B. pumilus 2IR was used as a biosurfactant-producing bacterium which was obtained from an oil field. The isolate B. pumilus 2IR was grown aerobically in a mineral salt medium, namely medium Enrichment (medium E) that contained following components (g/L): (NH 4 (Ghojavand et al. 2008(Ghojavand et al. , 2012Youssef et al. 2004). The pH of the medium was adjusted to 7 using 1 N NaOH.

Biosurfactant Production
The production of the biosurfactant was studied by transferring a single colony of the bacterial strains into a 200-ml Erlenmeyer flask containing 50 ml nutrient broth medium, followed by incubation at 30°C and 150 rpm agitation for 16 h to obtain an optical density of 0.5 at a wavelength of 600 nm using a spectrophotometer. Subsequently, 10 ml of prepared culture content was utilized to inoculate 500 ml of medium Enrichment (medium E) in an Erlenmeyer flask, followed by the incubation of the culture at 30°C on a rotary shaker at 150 rpm for 72 h. Bacterial cells were separated from the culture broth by centrifugation (Centrifuge Sorvall. RC-5B) at 8000 rpm for 10 min at 4°C. The collected supernatant was acidified with 6 N HCl to pH 2.0 and allowed to settle at 4°C overnight. A white precipitate containing the biosurfactant was then recovered by centrifugation at 10,000 rpm for 15 min. The precipitate was suspended in a minimal amount of distilled water and adjusted to pH 7.0 using 1 N NaOH. The solution was then lyophilized (LABCONCO) and weighed to determine acidprecipitated biosurfactant (Ghojavand et al. 2008;Joshi et al. 2008;Vaz et al. 2012).

Biomass Concentration
The biomass analysis was carried by measuring the dry cell weight of the bacterial cells. The culture broth was centrifuged (10,000 rpm for 15 min), and supernatant was separated from biomass. Thereafter, the cell pellet was washed with distilled water twice and dried by heating at 95°C until constant weight was attained (Joshi et al. 2008;Makkar and Cameotra 2002).

Surface Tension Measurement
The surface tension measurements of the culture broth supernatant were taken with a tensiometer (KRUESS K1OT model, Germany) according to the Du-Nouy's ring method at room temperature (Vaz et al. 2012;White et al. 2013).

Effect of Different Carbon Source, Nitrogen Source and C/N Ratios on Biosurfactant Production
The effect of various sources of carbon and nitrogen in the culture medium on the biosurfactant production by the strain 2IR was studied by performing a series of experiments in which one variable changed at a time, while other variables kept in fixed conditions. To study the effect of different carbon sources on the biosurfactant production by B. pumilus 2IR, carbon sucrose in the basal medium E was replaced with the carbon sources under study. The tested carbon source was added at an equivalent concentration (10 g/L) except for hydrocarbons where they were used at 1% (v/v). The tested carbon sources were carbohydrate sources (glucose, sucrose, lactose, glycerol, molasses, starch and cheese whey) and hydrocarbons (hexadecane, crude oil and paraffin oil) (Abdel-Mawgoud et al. 2008;Abouseoud et al. 2008). Similarly, the effect of the addition of different nitrogen sources to the culture medium on the biosurfactant production was studied. In order to evaluate the various nitrogen sources, organic nitrogen sources (urea, yeast extract and peptone) and inorganic sources (potassium nitrate, ammonium chloride and ammonium sulfate) were replaced by an equivalent amount (1 g/L). Finally, the C/N ratio was studied as the fundamental aspect to improve the biosurfactant production, while the concentration of nitrogen kept constant. To test the effect of the C/N ratios on the biosurfactant production, B. pumilus 2IR was grown in the medium E at four different C/N ratios of 4:1, 8:1, 12:1 and 16:1 using sucrose (different concentrations) and potassium nitrate (fixed concentration of 1 g/L) as carbon and nitrogen sources, respectively (Kiran et al. 2010;Makkar and Cameotra 2002).

Effect of Different Environmental Parameters on Biosurfactant Production
The effect of pH, temperature and sodium chloride (NaCl) concentration on the production of biosurfactant by B. pumilus 2IR was studied. In order to evaluate the effect of the initial pH value of medium on the biosurfactant production, the pH value of the culture medium was adjusted in the range of 4.0-10.0 using 1 N NaOH and 1 N HCL. Furthermore, the effect of the incubation temperature was studied by setting the culture temperature at 25, 30, 35, 40 and 45°C. To examine the effect of the salinity on the biosurfactant production in the culture medium E, sodium chloride was added to the medium to obtain final concentrations of 1-15% (w/v) (Haddad et al. 2009;Khopade et al. 2012a;Sarafin et al. 2014). All these experiments were carried out in 500-ml shake flasks containing 200 ml medium E, on a rotary shaker at 150 rpm. Surface tension activity, dry cell weight and acid-precipitated biosurfactant production were the monitored parameters.

Batch and Fed-batch Fermentation for Biosurfactant Production in the Bioreactor
Batch culture of the isolate B. pumilus 2IR for biosurfactant synthesis was carried out in a 5-L stirred tank bioreactor which integrated with accessories and automatic systems for dissolved oxygen (DO), pH, impeller speed, aeration rate and temperature (INFORS-Minifors-HI). The stirred bioreactor was supplied with two six-bladed turbine impellers which were used for agitation. A steam-sterilizable glass pH electrode was used to monitor the culture pH.
Air was sparged through a pipe sparger placed below the bottom impeller. The gas flow rate was 1.5 vvm. Impeller speed was 150 rpm, and temperature was maintained at 30°C. A volume of 2 L of basal medium E was added to the bioreactor, which had been autoclaved at 121°C for 20 min. After sterilization, the integrated bioreactor was cooled to 25°C and then 20 ml of the sterilized trace elements solution was added to 2 L of the medium E. The bioreactor was then inoculated and samples were taken for chemical analysis. The fed-batch fermentation was started after 24-30 h of the batch cultivation as the culture entered stationary phase. The fed-batch mode was allowed to continue up to 144 h. The fed-batch experiments were performed with the same setting as in the batch fermentation (agitation 150 rpm, temperature of 30°C and aeration of 1.5 vvm). A volume of 400 ml of the sterilized medium E with a double concentration of components was added by intermittently mode at a 1-day interval for 5 days manually. The carbon and nitrogen sources utilized in the medium E for the batch mode in the bioreactor were glucose, crude oil, potassium nitrate and ammonium sulfate with a concentration of 30.31 g/L, 0.8% (v/v), 2.88 and 2.4 g/L, respectively (Fooladi et al. 2016). The carbon and nitrogen sources utilized for the fed-batch mode were similar to those used for the batch mode in the bioreactor using a double concentration of them. The samples were collected, before and after each cycle to determine the biomass concentration, acid-precipitated biosurfactant production and glucose consumption.

Analytical Methods
Samples of the cultures were collected at regular time intervals and centrifuged at 10,000 rpm for 10 min at 4°C. The supernatant was used for measuring acid-precipitated biosurfactant (Haddad et al. 2009;Najafi et al. 2011). The pellet was washed with distilled water twice and dried by heating at 95°C until constant weight was attained (Abouseoud et al. 2008;Lotfabad et al. 2009). The concentration of glucose was determined directly using glucose assay kit (Cobas-Glu-Roche).

Screening of Important Medium
Components for Biosurfactant Production Figure 1 depicts biosurfactant production by B. pumilus 2IR using different carbon and nitrogen sources. As shown in Fig. 1a, among sugar-based carbon sources tested, glucose, sucrose and cheese whey showed high biomass production with concentrations of 3.40, 2.80 and 2.56 g/L, respectively. Furthermore, high reductions in the surface tension were measured for glucose, sucrose and cheese whey with values of 31, 33 and 34 mN/m, respectively. The comparison of the biosurfactant produced from different carbon sources indicated that glucose and sucrose resulted in a high biosurfactant production with concentrations of 0.85 and 0.77 g/L, respectively, implying the fact that B. pumilus 2IR was able to better assimilate sugarbased carbon sources for enhancing the biosurfactant production. Figure 1a shows that the biosurfactant produced from crude oil (0.25 g/L) was higher than that from paraffin (0.13 g/L). Evidently, the surface tension activity obtained from the crude oil (39 mN/m) was lower than that from the hexadecane (41 mN/m). Moreover, Fig. 1a reveals that the strain 2IR exhibited a better growth on the crude oil with the higher biomass production of 1.43 g/L than that from the paraffin (1.25 g/L). Accordingly, glucose and crude oil showed better effects on the biosurfactant production. In this regard, it can be deduced that the glucose acts as a primary carbon source during the initial growth phase, while in the later stages of fermentation, bacteria possibly use the crude oil as a carbon source. Snehel et al. (2008) demonstrated that an appropriate combination of simple carbon source of mannitol (1.6 g/L) and an inducer hexadecane (63 g/L) enhanced the production of biosurfactant. Figure 1b shows the biosurfactant production using various nitrogen sources. Obviously, the strain 2IR was able to utilize all tested nitrogen sources; however, the production of biosurfactant was not considerable when urea, yeast extract, peptone and ammonium chloride were used, compared to that from ammonium sulfate and potassium nitrate. The highest biosurfactant concentrations measured were 0.69 and 0.84 g/L when ammonium sulfate and potassium nitrate were utilized, respectively. Evidently, ammonium sulfate and potassium nitrate were the most suitable nitrogen sources, resulting in a higher reduction in surface tension with the values of 32 and 31 mN/m, respectively. This finding indicated the fact that this strain had a strong capability to assimilate inorganic nitrogen compounds. The results of the study well agreed with the findings reported by Aparna and Srinikethan (2012) and Abdel-Mawgoud et al. (2008) who found that glucose was the best carbon source for the maximum biosurfactant production by B. clausii 5B and B. subtilis BS5, respectively, while Joshi et al. (2008) showed that the highest amount of biosurfactant was produced by B. licheniformis using molasses at concentrations of 5.0-7.0% (w/v). Moreover, this result was consistent with the report given by Makkar and Cameotra (2002) who worked on the biosurfactant synthesis by Bacillus subtilis MTCC 2423. They produced a higher activity for biosurfactant (30 mN/m) when potassium nitrate and sucrose were used as a nitrogen source and carbon source, respectively. In this regard, Haddad et al. (2009) achieved a maximum surface tension reduction of 27.3 mN/m by Bacillus subtilis HOB2 using ammonium sulfate.
Biosurfactant production under different C/N ratios was investigated by the utilization of glucose and potassium nitrate as the carbon and nitrogen source, respectively, which exhibited the highest positive effect on the biosurfactant production and surface tension reduction (Fig. 1c). It is obvious that an increment in C/N ratio from 4 to 12 favored biomass concentration, biosurfactant production and surface tension reduction up to 2.68, 0.76 g/L and 33 mN/m, respectively. However, a higher increase in C/N ration from 12 to 16 decreased biosurfactant concentration to 0.45 g/L, suggesting that a relatively low C/N ratio favored biosurfactant synthesis by bacterial cells so that a too high concentration of carbon source had no favorable effect on biosurfactant production by the strain 2IR. The results were in close agreement with the findings obtained by Abouseoud et al. (2008) who reported C/N ratio 10:1 was the best ratio of substrate for the maximum production of rhamnolipid biosurfactant by Pseudomonas fluorescens, whereas the biosurfactant activity and emulsification index decreased by increasing C/N ratio to 30:1 and 50:1. The study on Nocardia sp. revealed that stimulation of the biosurfactant synthesis attained a C/N ratio of 20:1 compared to ratio of 10:1 and 15:1; however, there were no significant differences between C/N ratios of 30:1 and 35:1. The results indicated that the highest biosurfactant production was related to nitrogen-limiting medium (Khopade et al. 2012a). Lotfabad et al. (2009) found that the highest production of the biosurfactant (0.84 g/L) by the strain Pseudomonas aeruginosa MR01 was achieved using a C/N ratio of 20:1. Figure 2a shows the relation between NaCl concentration and the biosurfactant production by the strain B. pumilus 2IR. As can be found, the strain 2IR has the capability to grow and produce biosurfactant in different concentrations of NaCl. Obviously, an increment in the salinity of the medium to 7% (w/v) favored the biomass concentration, the biosurfactant production and the surface tension activity up to 2.65, 0.62 g/L and 33 mN/m, respectively. The biomass concentration and the biosurfactant production were not considerable when NaCl increased more than 7% (w/v). The best response was obtained with increasing the salinity up to 7%, which implied that the composition of the culture medium with the almost moderate NaCl content favored the biosurfactant production by the strain 2IR. As can be seen, the too low salinity reduced the amount of the biomass, implying the fact that NaCl is an essential medium component to improve the biomass and the biosurfactant production. However, data from Fig. 2a show that a higher increase in the culture salinity from 7 to 15% considerably decreased the biosurfactant concentration to 0.26 g/L. This result well agreed with the findings reported by Ghojavand et al. (2008) who showed that NaCl concentration of 4-8% (w/v) was the optimal salinity for the biosurfactant production by B. subtilis PTCC 1696. In some cases, there is an inverse relation between the sodium chloride concentration and the biosurfactant production yield. Lotfabad et al. (2009) showed that a gradual addition of NaCl in the culture medium led to a decrease in the biosurfactant production by the strain P. aeruginosa. The study on the biosurfactant production by Nocardiopsis sp. B4 showed that this strain was moderately halophilic in the nature so that the maximum biosurfactant production was obtained in the presence of 3% (w/v) of NaCl and it retained almost 80% of its activity in the presence of 12% (w/v) of NaCl. Since about 90% of the salinity content of the reservoirs brine is composed of NaCl salt (Ghojavand et al. 2008), the surface activity of the biosurfactant solution in our studies was performed at different NaCl concentrations and showed a major surface tension reduction at sodium chloride concentration of 7% (w/v), indicating a remarkable characteristic of this strain for wide application in MOER by in situ production of the biosurfactant in the oil fields which have been subjected to water injection. In an attempt for the biosurfactant production, the strain V. salarius (KSA-T) was found to be halophilic, as the highest biosurfactant synthesis was obtained in the presence of 4% (w/v) of NaCl; however, in the presence of 10% (w/v) of NaCl, it lost 25% of its activity (Elazzazy et al. 2015).

Effect of pH Value
The pH value of the medium is a key factor for cell growth and the production of primary metabolites. The biosurfactant production by B. pumilus 2IR was shown to be affected by the initial pH value of the culture medium. As shown in Fig. 2b, there was no notable increase in the biomass and biosurfactant production at a low pH value (4.0-5.0), indicating that the low pH value led to unfavorable conditions for the bacterial cell growth and biosurfactant synthesis. This could be attributed to the deleterious effect of the low pH value on the metabolic activities of Bacillus cells in the production of the biosurfactant (Khopade et al. 2012a;Ghojavand et al. 2008). However, the biomass concentration and biosurfactant production positively increased at pH values above 5.0. Consequently, the surface tension of the medium decreased from 38 mN/m at pH 4.0 to the value of 33 mN/m at pH 8.0. As is evident, there were not considerable changes between pH 7.0 and 8.0. As can be seen, the optimum pH value for maximum production was between 6.0 and 8.0 (Fig. 2b). Obviously, when the initial pH value of the culture medium increased from 8.0 to 10.0, the biomass concentration and the biosurfactant production drastically decreased to 1.40 and 0.25 g/L, respectively, so that the surface tension increased up to 39 mN/m. It was found that the strain Virgibacillus salarius (KSA-T) showed the maximum biosurfactant production when pH increases to 9.0 (Elazzazy et al. 2015). Similar study was performed by Lotfabad et al. (2009) who demonstrated that pH 8.0 was the optimum pH value for the strain Pseudomonas aeruginosa MR01 to obtain the highest reduction in the surface tension and highest emulsification activity.

Effect of Culture Temperature
Temperature is one of the critical parameters that affect the biosurfactant production. Figure 2c depicts the effect of the culture temperature on the production of biosurfactant by B. pumilus 2IR in a temperature range of 25-40°C. The results obtained in the present study revealed that the surface tension of the strain 2IR reached to the lowest level of 34 mN/m when the strain was cultivated at 30°C with producing the maximum amount of the biomass (2.62 g/L) and the biosurfactant (0.53 g/L). Figure 2c shows that although the isolate 2IR was able to grow at the incubation temperature range of 35-40°C, the production of biosurfactant was considerably low compared to that from the incubation temperature range of 25-35°C. Surface tension value increased at incubation temperatures above 35°C and reached to a highest amount of 43 mN/m at 45°C where biomass concentration and biosurfactant production were 0.6 and 0.1 g/L, respectively, indicating that the strain 2IR was not able to efficiently produce biosurfactant at high temperatures. The similar results were reported by Khopade et al. (2012a) and Haddad et al. (2009) who studied on the biosurfactant synthesis by Nocardia sp. B4 and B. subtilis HOB2, respectively. Inappropriate incubation temperature caused to change in the composition of the biosurfactant produced by Arthrobacter paraffineus and Pseudomonas sp. (Khopade et al. 2012b). The similar study showed that the strain V. salarius (KSA-T) was moderately thermophilic and attained the ideal growth and maximum biosurfactant production at the temperature range of 40-45°C (Elazzazy et al. 2015).

The Profile of Biosurfactant Synthesis by B. pumilus 2IR in Batch Fermentation Mode Using a 5-L Bioreactor
The typical profile of the cell growth of the isolate 2IR and the biosurfactant production in a 5-L batch bioreactor was studied using medium E with the composition of glucose, crude oil, potassium nitrate and ammonium sulfate as pivotal carbon and nitrogen sources. As shown in Fig. 3, the strain 2IR was able to grow well and produce the biosurfactant in the bioreactor. As shown in Fig. 3, the production of biosurfactant was enhanced decisively when bacterial cells entered to the exponential growth phase. The growth of bacterial cells was rapid during the initial stages of fermentation so that culture entered to a stationary growth phase after 48 h of cultivation. Experimental results illustrated in Fig. 3 indicated that the maximum concentration of the biomass and the lipopeptide biosurfactant in the bioreactor was 4.15 and 0.98 g/L, respectively, which occurred at 60 h of the cultivation so that there was no a considerable increase in the biosurfactant concentration after 60 h. This finding pointed out that the biosurfactant production by the isolate 2IR was associated with the bacterial growth as an increase in the biomass concentration occurred simultaneously with the increment in acid-precipitated biosurfactant production at the exponential growth phase. As shown in Fig. 3, B. pumilus 2IR was able to grow well and produce biosurfactant in the batch bioreactor. These results well agreed with the findings reported by Amani et al. (2010) who studied the biosurfactant production by Bacillus subtilis. According to the authors, the lipopeptide biosurfactant was produced efficiently by the strain in the proposed integrated bioreactor. The maximum biosurfactant production in the bioreactor was about 2.5 g/L. Moreover, the strain could attain its maximum level of the biomass and the biosurfactant concentration at 40 h of fermentation time.

The Comparison of Glucose Consumption for Biosurfactant Production in Shake Flask and Batch Bioreactor
In order to compare glucose consumption by B. pumilus 2IR in the shake flask and the bioreactor, the profile of the variation in glucose concentration of the culture medium in both batch bioreactor and shake flask was studied. As shown in Fig. 4, glucose was rapidly consumed during 24 h of fermentation time in the bioreactor which was related to the early exponential growth phase (Fig. 3), so that it was fully depleted after 36-h fermentation. On the other hand, the consumption of glucose by the bacterial cells was slower in the shake flask culture so that it was consumed completely at 60-h fermentation. Amani et al. (2010) reported the substrate concentration during fermentation time in the shake flask and fermenter. The results revealed that the growth continued until 60 h where the sucrose as the carbon source was completely consumed in the shake flask while the concentration of sucrose reached to its minimum level at 40 h in the bioreactor. As mentioned previously, the lipopeptide biosurfactant production by B. pumilus 2IR was a growth-associated process and the enhanced production may be resulted from extending the active growth phase of B. pumilus 2IR. Hence, the fed-batch operation would be a suitable approach to employ the activated cells of B. pumilus 2IR for the biosurfactant synthesis by the addition of fresh nutrients as soon after they were depleted during the batch growth. In the fed-batch culture, a concentrated medium should be added at the end of the growth phase in the batch culture to stimulate growth and the consequent biosurfactant production. A concentrated medium is required since the nutrient concentrations are very low in a batch culture and consequently the substrate could be depleted. Moreover, the cell concentration is usually in a maximum level at the stationary phase of the batch culture, which means that there would be a very high demand for fresh nutrients. The concentration of the feed during the fed-batch culture is an important factor for providing the quantity of the nutrient entering to the reactor and ultimately controls the fed-batch culture process. The time course of the biosurfactant production, the biomass concentration as well as glucose consumption by B. pumilus 2IR in the fed-batch fermentation was investigated using the 5-L bioreactor. As shown in Fig. 5, biomass concentration and biosurfactant production started to increase proportionally during the first 24 h of cultivation time where the culture was at its early exponential phase. As is evident from Fig. 5, glucose was rapidly consumed during the exponential phase of the batch culture and feeding of the culture was carried out before the full depletion of glucose. Glucose consumption trend in the fed-batch culture was similar to that obtained from 5-L bioreactor. As shown in Fig. 5, glucose was depleted during the 24 h of cultivation when the culture was at the exponential growth phase.
During the fed-batch fermentation, glucose consumption increased as the appropriate concentration of the medium was applied (400 m/L of medium E, with double concentration of the composition). The findings indicated that the glucose consumption quantity was approximately equal to the glucose feed rate. Furthermore, Fig. 5 shows that the biosurfactant concentration was improved throughout the five cycles of feeding to 1.06 g/L in the fed-batch cultivation. The findings indicated that when the nutrients were limited in the stationary phase, the bacterial growth and the biosurfactant production had no further increment. Obviously, as a rise in the biomass concentration occurred with an increase in the biosurfactant production, the period of the product formation was extended (144 h in this study) by feeding strategy which led to consequent enhancement in the growth and the product formation. As shown in Fig. 5, the growth of the strain B. pumilus 2IR in the each fed-batch cycle was accompanied by increasing biosurfactant production during the fed-batch fermentation experiment time. In this view, the biosurfactant production profile was consistent with the biomass production profile, indicating the point that similar to the batch culture in the bioreactor, a rise in the biomass concentrations occurred simultaneously with the enhancement in the biosurfactant production. Cell growth of the isolate 2IR reached to its maximum level during the fed-batch experiments with the average concentration of 5.71 g/L in the bioreactor at 144 h of fermentation. Amin (2014) studied the fed-batch strategy for enhancing the biosurfactant production by Bacillus subtilis. The strain produced surfactin as a growthassociated product in a conventional batch process. Utilizing the fed-batch fermentation greatly established a cost-effective commercial surfactin production, as the highest surfactant of 36.1 g/L and cell biomass of 31.8 g/L were achieved after 12 h of the cultivation time. Glucose concentarƟon (g/L)

Fermentation time
Shake flask 5-L batch bioreactor Fig. 4 The profile of glucose consumption by B. pumilus 2IR for biosurfactant production in a 5-L batch bioreactor and shake flask

Comparison of Biomass and Biosurfactant Production in Different Modes of Fermentation
Acid-precipitated biosurfactant production and biomass concentration were compared in terms of different fermentation operation systems as given in Table 1. As is evident, the maximum biomass achieved in the bioreactor (4.15 g/L) increased by 1.3-fold compared to the highest biomass measured in the shake flask fermentation (3.40 g/ L). Similarly, the maximum lipopeptide biosurfactant concentration achieved in the 5-L batch bioreactor (0.98 g/ L) was approximately 1.16-fold higher than the highest biosurfactant attained in the shake flask (0.85 g/L). Bence (2011) showed that the growth and surfactin production by B. subtilis ATCC 21,332 in the bioreactor were greater than the results obtained in the shake flask experiments. From the results obtained, they found that the biomass and surfactin concentrations measured in the bioreactor were enhanced compared to that obtained from the shake flask studies. The author also showed that the highest production of surfactant was exhibited at 20 h of the cultivation time in the bioreactor. The utilization of the bioreactor for the production of rhamnolipid biosurfactant by Pseudomonas aeruginosa was investigated in micro-bioreactor conditions by Rahman and Gakpe (2008). The results revealed that the biomass and biosurfactant concentrations could be improved in the instrumented bioreactors compared to those in the shake flask. The reason for these changes may be due to the differences in hydrodynamic conditions and shear rate created in the shake flask. In addition, fermentation process in the shake flask is subjected to oxygen limitation. Moreover, enhanced growth and biosurfactant production can be achieved over smaller periods of time in the instrumented fermenters, thus indicating higher biomass and biosurfactant concentration. It is noteworthy that the fed-batch fermentation in the bioreactor decisively enhanced the concentration of the biomass and the biosurfactant compared to the batch mode so that production of the biomass and the biosurfactant increased up to 5.71 and 1.06 g/L, respectively.
In this study, the substrate feed was not taking out from the culture during the fed-batch fermentation and therefore lipopeptide biosurfactant concentration increased with the fermentation time (Fig. 5). Increasing biomass and biosurfactant concentration in the fed-batch system could be attributed to the higher supply of nutrients to the bacterial cells of the strain 2IR during fermentation. The highest biosurfactant production observed in the fed-batch fermentation (1.06 g/L) was improved 8.2% compared to that in the batch fermentation using the bioreactor. Similarly, the highest biomass concentration obtained in the fed-batch culture had 37.6% improvement compared to that from the batch cultivation in the bioreactor. The results were  consistent with the findings reported by Lee et al. (2004) who studied on the enhancement of the biomass and rhamnolipid production in the batch and the fed-batch fermentation. The study showed that the final bacterial cell and the biosurfactant production were 5.3 and 17.0 g/L, respectively, in the batch fermentation. Enhancement of dry cell mass up to 6.1 g/L and rhamnolipid production up to 22.7 g/L was observed when the fed-batch culture was used. The fed-batch culture resulted in a 1.2-fold increase in the dry cell mass and a 1.3-fold increase in rhamnolipid production, compared to the production of the batch culture.
The utilization of the fed-batch techniques for the production of extracellular biosurfactant is known well due to its operational advantages. In this context, little work has been published concerning the fed-batch experiments for the synthesis of the lipopeptide biosurfactant by B. pumilus. The results obtained in the present work were in close agreement with the findings reported in other studies. In the study fulfilled by Samsu et al. (2014), it was found that a higher yield of rhamnolipid biosurfactant could be obtained when a suitable feeding strategy for the fed-batch culture was implemented. In this view, 14.39 g/L of rhamnolipid biosurfactant was obtained from the fed-batch mode. The results showed a fivefold improvement in the rhamnolipid biosurfactant concentration, compared to the batch culture. Nur et al. (2012) investigated the production of glycolipid biosurfactant by P. aeruginosa in the fed-batch culture with different feeding strategies. They applied a pulse feeding strategy using diesel as a carbon source. The results indicated that higher biomass and biosurfactant concentrations of 12.6 and 3.13 g/L were produced, respectively, compared to the results obtained from the batch culture with values as high as 9.4 and 2.35 g/L, respectively. The authors also showed that the biomass concentration and glycolipid biosurfactant were improved by 34 and 33%, respectively, which were in good agreement with our results obtained in the current study. Furthermore, Sivapathasekaran and Sen (2013) mentioned 35% improvement in the biosurfactant production by marine bacteria when unsteady state of the fed-batch was used over the batch mode. Fredrico et al. (2010) found that the higher values of the rhamnolipid production (16.9 g/L) were obtained when the system was supplied by extra nitrogen and carbon during the fed-batch fermentation. The most important aspect which should be taken into account is the economic synthesis of biosurfactant.
Biosurfactant could be produced by microbial cells growing on unrefined substrates or even waste materials. By contrast, many commercially available surfactants for EOR are obtained from a petroleum origin or synthesized from petrochemical feedstocks. The potential of the utilization of the microbial biomass for the production of biosurfactant is also considered as an economically favorable approach (Das and Mukherjee 2007;Pacwa et al. 2011). On the other hand, in situ production of biosurfactant by inoculated microorganisms could enhance 10-200% of the oil recovery. The in situ growth of biosurfactant-producing microorganisms in the oil reservoir may offer many advantages including continuous production to offset adsorption and radial dilution of surfactants, an abundant production of potentially recoverable biomass and recycle in EOR (Pacwa et al. 2011;Youssef et al. 2013). Therefore, further research should focus on the various approaches to application of the biosurfactant in the oil recovery to make this method economically competitive with the current strategies.

Conclusion
Promising improvement in biosurfactant production can be achieved using the fed-batch culture operations. Irrespective of the mode of operations, whether it is the batch or the fed-batch fermentation, the pivotal factors that influence biosurfactant production are the medium composition and the physiochemical culture conditions. The highest biosurfactant produced by B. pumilus 2IR was 1.06 g/L in the fed-batch fermentation using the 5-L bioreactor. The production of biosurfactant by the strain 2IR could be notably utilized for the economically viable in situ oil recovery in harsh conditions of the oil reservoirs.