- Email: [email protected]
Recovery and separation of surfactin from pretreated fermentation broths by physical and chemical extraction
Biochemical Engineering Journal 38 (2008) 39–46
Recovery and separation of surfactin from pretreated fermentation broths by physical and chemical extraction Huei-Li Chen, Ruey-Shin Juang ∗ Department of Chemical Engineering and Materials Science, Yuan Ze University, Chung-Li 32003, Taiwan Received 25 April 2007; received in revised form 5 June 2007; accepted 8 June 2007
Abstract The recovery of surfactin from fermentation broths with the culture of Bacillus subtilis ATCC 21332 by physical and chemical extraction was studied, in which the broths were pretreated by acid precipitation and, if necessary, the precipitate was further dissolved in NaOH solution. The physical solid–liquid and liquid–liquid extractions were performed with different organic solvents (ethyl acetate, n-hexane) and at different times of extraction run. It was shown that better extraction was obtained with ethyl acetate than n-hexane. The extraction could be improved by increasing the times of extraction run under a given volume of the organic solution. For chemical liquid–liquid extraction of surfactin with Aliquat 336 (5–200 mM) in n-hexane, the efficiency was improved. The amounts of inorganic salt in the strip solution were also optimized. Surfactin recovered after physical or chemical extraction was finally characterized by mass and NMR spectroscopies. The results of chemical extraction showed that surfactin would readily bind with the quaternary ammonium cations of Aliquat 336. © 2007 Elsevier B.V. All rights reserved. Keywords: Surfactin; Recovery; Fermentation broths; Extraction; Ethyl acetate; n-Hexane; Aliquat 336
1. Introduction Biological molecules exhibiting particularly high surface activity are classified as biosurfactants. They generally include a wide variety of chemical structures such as glycolipids, lipopeptides, polysaccharide–protein complexes, phospholipids, fatty acids, and neutral lipids [1,2]. They also have a strong effect on interfacial rheological phenomena and mass transfer [3]. Biosurfactants have several advantages, such as low critical micelle concentration and high biodegradability, over chemical synthetic surfactants and, thus, are especially well suited for environmental applications such as bioremediation and the dispersion of oil spills [4,5]. In addition, biosurfactants are thought to be potential candidates to replace chemically synthetic surfactants in the future, particularly in food, cosmetic, and health care industries, as well as in industrial cleaning of products and in agricultural chemicals [6,7]. Surfactin, a cyclic lipopeptide produced by several strains of Bacillus subtilis, is one of the very powerful biosurfactant [7,8]. It has a cyclic lipopeptide with -hydroxy fatty acids linked to
∗
Corresponding author. Tel.: +886 3 4638800; fax: +886 3 4559373. E-mail address: [email protected] (R.-S. Juang).
1369-703X/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bej.2007.06.003
a heptapeptide (l-Glu-l-Leu-d-Leu-l-Val-l-Asp-d-Leu-l-Leu) [3,9]. Generally speaking, surfactin has an exceptional surfaceactive power because it lowers the surface tension of water from 72 to 27 mN/m at a concentration even as low as 20 M [8]. It is recognized that downstream processing in many biotechnological processes is responsible for up to 60% of the total production cost [10]. From the economic point of view, most biosurfactants would have to involve whole-cell spent culture broths or other crude preparations [1]. The most commonly used and cheap methods for the recovery of biosurfactants from fermentation broths are acid precipitation, foam separation, or the combination of both [1]. However, they usually lead to relatively low biosurfactant purity (<50%). An alternative method for biosurfactant recovery is solvent extraction [10]. In practice, a wide variety of organic solvents including methanol, ethanol, butanol, diethyl ether, n-pentane, acetone, acetic acid, chloroform, and dichloromethane have been used, either single or in combination, for this purpose [1]. The most effective solvent would be the mixture of chloroform and methanol in various ratios, which facilitates adjustment of the polarity of extraction solvents to the target extractable material [11]. However, chloroform is highly toxic compound regarded as harmful to the environment and for human health [2]. Thus, cheap and low-toxicity solvents are highly desired for biosurfactant extraction from the view-
40
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
Nomenclature A volume of aqueous solution (mL) Aliquat 336 trioctylmethylammonium chloride LLE liquid–liquid extraction O volume of organic phase (mL) SLE solid–liquid extraction
point of industrial applications. In recent years, ethyl acetate and n-hexane, which are the representatives of polar and nonpolar solvents, have been used to recover biosurfactants produced by B. subtilis [12–14]. Such types of low cost, less toxic, and easily available solvents could be employed to cut the recovery cost and minimize the environmental hazards. In this work, the fermentation broth was pre-precipitated and the precipitate was obtained by centrifuging and ovendrying. In most cases, the precipitate was further dissolved in alkaline solution. Surfactin was then recovered by solid–liquid extraction (SLE) and liquid–liquid extraction (LLE) by the solvents n-hexane and ethyl acetate (physical extraction). Because physical extraction is sometimes not selective enough and has low recovery yield [15], LLE using a commercial basic extractant (chemical extraction) was also carried out for comparison. Higher recovery and purity were simultaneously targeted because the possibility of using such solvent extraction for practical applications was examined. Factors affecting extraction performance including types of organic solvent (ethyl acetate, n-hexane), times of extraction run, extractant concentration (0.5–200 mM) were investigated. The effect of added concentration of sodium chloride or ammonium sulfate (0.5–15 g/50 mL) on the stripping in chemical extraction system was also studied. Finally, the characteristics of the final product obtained after physical and chemical extraction were identified and compared. 2. Materials and methods 2.1. Microorganisms and culture condition Here, B. subtilis ATCC 21332 was selected to produce surfactin. The nutrient broth (NB) medium contained 3 g/L beef extract, 5 g/L peptone, and the mineral salt (MS) medium at pH 7. The MS medium consisted of 40 g/L glucose, 50 mM NH4 NO3 , 30 mM KH2 PO4 , 40 mM Na2 HPO4 , 7 M CaCl2 , 0.8 mM MgSO4 , 4 M FeSO4 , and 4 M tetrasodium salt of EDTA [16]. The pH was regulated at 7 by adding 0.1 M HCl or NaOH. Prior to use, the MS medium and deionized water (Millipore, Milli-Q) were sterilized in autoclave at 121 ◦ C for 15 min. All inorganic chemicals were offered from Merck Co. as analytical reagent grade. Culture of B. subtilis ATCC 21332 was taken from −80 ◦ C frozen stock and transferred onto agar medium for pre-culture. The Bacillus subtilis culture (1 mL) was inoculated into 250-mL flask containing 100 mL of NB medium at 30 ◦ C with 200 rpm of agitation. After growing up to late exponential phase (near 14 h), the NB medium containing B. subtilis cells was inoculated and
fermented in 5-L fermenter (Firstek Scientific Co., Taiwan) with 4-L working volume at 30 ◦ C and 200 rpm for another 4 days. The aeration was controlled at a rate of 1.0 vvm. The dissolved oxygen concentration maintained 80% during the fermentation (2–4 days). The medium pH was maintained at 7.0 by automatic addition of 1 N NaOH or HCl solution to the fermenter according to the signal received from the pH electrode. The fermentation broth was pretreated as follows. First, the liquor inside the fermenter was centrifuged at 10,000 × g to remove biomass impurities; the remaining supernatant was further treated and precipitated by adding 1 M HCl to a pH of around 4. The yellowish precipitate (crude powder) was obtained by centrifuging at 10,000 × g for 15 min and oven-drying at 37 ◦ C for 2 days. In most situations, the dried precipitate was further dissolved in NaOH solution at pH 11 in order to yield a homogeneous solution. It was found that the crude precipitate had a surfactin purity of about 53% according to the method described below. 2.2. Assay surfactin concentration Culture samples were taken after centrifuging at 12,000 × g for 15 min to remove the biomass, and surfactin concentration in the clarified supernatant was measured with HPLC equipped a reverse phase C18 column (5 m, Merck) at 30 ◦ C [17]. The samples were subjected to filtration through a Millipore filter (0.45 m) before analysis. A mixture of 3.8 mM trifluoroacetic acid (20% v/v) and acetonitrile was used as the mobile phase, and the flow rate was 1.0 mL/min. An aliquot of the sample (20 L) was injected and analyzed using an UV detector (Jasco 975, Japan). The wavelength was set at 205 nm [18]. Each concentration analysis was at least duplicated under identical conditions. The reproducibility is mostly within 5%. Surfactin powder purchased from Sigma Co. served as the standard (98% purity as per label claim). The purity of surfactin in the dried sample was calculated by amount of surfactin determined by HPLC
purity(%) =
weight of dried sample powder dissolved in the solution × 98
(1)
The purities of surfactin in the recovered product and in the fermentation broth would be used to calculate the recovery yield of surfactin. 2.3. Physical extraction In SLE, 0.1 g of the dried precipitate was contacted with 25 mL of the organic solvent, n-hexane or ethyl acetate (Mallinckrodt Co.). The solution mixture was agitated using a magnetic stirrer at 250 rpm for 24 h at 30 ◦ C. Preliminary experiments had shown that the extraction studied was complete within 18 h. Then, it was centrifuged at 6000 × g for 10 min and the solvent was collected. For a given volume of the organic solution, the extraction was performed several times by fresh organic solution when organic solution was divided into the corresponding parts. After evaporation of the organic solution, surfactin solid
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
41
was dissolved in NaOH solution at pH 11 and its concentration was analyzed. The extraction efficiency was obtained by amount of surfactin in the organic solution
extraction(%) =
initial amount of surfactin in the feed sample
× 100
(2) In LLE, the pH of aqueous solution was first adjusted to be near 8.0 by the addition of 0.1 M HCl. Equal volumes of the organic solution and aqueous solution (50 mL) were mixed in a tightly stoppered 250-mL flask, and were agitated at 250 rpm for 24 h in water bath of 30 ◦ C. After phase separation, the solvent was removed with a rotary evaporator under reduced pressure conditions to obtain off-white powders. Similarly, the powder was dissolved in NaOH at pH 11 and its concentration was analyzed. Moreover, the fresh organic solution was divided into several parts to examine the effect of repeated extraction runs. 2.4. Chemical extraction Chemical extraction was tested in LLE systems only. The experimental procedures were essentially the same as those described in Section 2.3, except that the organic solution contained various amounts of Aliquat 336 (trioctylmethylammonium chloride, Aldrich Co.) in n-hexane. After phase separation, the upper organic phase was gathered for the followed stripping experiments. The concentration of surfactin in the lower aqueous phase was determined with HPLC, and the extraction efficiency of surfactin was similarly calculated according to Eq. (2). In the stripping experiments, equal volumes of the loaded organic solution and strip solution (50 mL) were mixed, in which the strip solution contained ethanol and sodium chloride or ammonium sulfate. The stripping test was also performed at a stirring speed of 250 rpm for 24 h at 30 ◦ C. 2.5. Characterization of surfactin product The final surfactin powder dissolved in chloroform (CD3 Cl) was detected by NMR spectroscopy (Bruker, Rheinstetler, Germany) to identify the structure of surfactin. FAB-MS (Fast-atom-bombardment mass spectra, model JMS-700 JEOL, Tokyo) was also used to determine the molecular weight of surfactin. 3. Results and discussion 3.1. Physical extraction for the pretreated broth It was reported that the raw fermentation broth (before acid precipitation) with B. subtilis ATCC 21332 contains macromolecules (surfactin micelles, polysaccharides, peptides, proteins, etc.), mid-molecules, and small molecules (MS medium, alcohols, phthalic acid, amino acids, glycine, serine, threonine, alanine, etc.) [19]. It should be reminded in Section 2.1 that the surfactin purity of 53% only could be achieved after acid precipitation alone. Further improvement in purity would be aided by solvent extraction. For simplicity, the total extraction (recovery) of surfactin is selected as the key parameter to evaluate the performance of each type of solvent extraction.
Fig. 1. Effect of the types of organic solvent on the extraction efficiency in LLE system at different initial concentrations of surfactin (A and O refers to the volumes of the aqueous and organic solutions, respectively).
Fig. 1 shows the effect of initial surfactin concentration on the efficiency of LLE using different organic solvents. Here, the volumes of the organic (O) and aqueous solutions (A) are identical. The extraction is much effective with ethyl acetate (>95%) than n-hexane (<21%) probably because n-hexane is a low-polarity solvent. However, the concentration effect on the extraction is both weak under the ranges studied. Rau et al. [20] have used n-hexane to extract mannosylerythritol lipid produced by Pseudozyma aphidis, and also found that the extraction is 8% only. It was also reported previously that the extraction of bioproducts with considerably high polarity by ethyl acetate solvent is rather efficient [21,22]. On the other hand, the recovered surfactin with either nhexane or ethyl acetate has a comparable purity (58–60%), although ethyl acetate gives much higher extraction than nhexane. Surfactin can be soluble in organic solvents due to the presence of hydrophobic end. The higher extraction with ethyl acetate indicates that surfactin and, at least, other protein macromolecules in the broths have highly polar hydrophobic groups. The present results likely suggest that the small molecule impurities in the broth can be first removed by extraction with n-hexane (they are left in the aqueous raffinate), and then surfactin is recovered with a highly polar solvent [12]. Due to the formation of emulsion after LLE, the SLE system was hence tested to avoid such behavior. Fig. 2(a) shows effect of the times of extraction run on the extraction and purity with ethyl acetate in SLE system. It is found that the maximum extraction of 78% is achieved when extraction runs were conducted three times. In this case, the purity of surfactin reaches a plateau of 84%. Fig. 2(b) shows the SLE results with n-hexane. It is also observed that the maximum extraction of 62% is achieved when extraction runs were made three times. On the other hand, the purity of surfactin is about 60%. The different phenomena with these two solvents are still a result of their different miscibility between solvents and the hydrophilic groups of surfactin. Ethyl acetate has an acceptable solubility in water (9.7% v/v), but nhexane is nearly insolubility in water. Therefore, more surfactin molecules are dissolved in ethyl acetate than in n-hexane.
42
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
Fig. 2. (a) Effect of the times of extraction run with ethyl acetate on the extraction efficiency and purity in SLE system. (b) Effect of the times of extraction run with n-hexane on the extraction efficiency and purity in SLE system.
As indicated above, the extraction is low with n-hexane in LLE system although the purity obtained is comparable to the case of ethyl acetate. There is hence a possibility to improve the extraction with n-hexane by increasing the time of extraction runs. Fig. 3 compares the effects of the times of extraction run on the extraction and purity with n-hexane between SLE and LLE systems. It is evident that the extraction in SLE system (64%) is more effective than that in LLE system (59%) at such a surfactin concentration (2000 mg/L). According to these results, it appears that SLE is preferred due to its easy-to-operate nature although SLE gives a slightly lower purity. However, the extraction with n-hexane, in SLE or LLE, is still uneconomical and impractical. 3.2. Chemical extraction for the pretreated broth
Fig. 3. Effect of the times of extraction run with n-hexane on the extraction efficiency and purity in LLE and SLE systems.
The recovery method for obtaining acceptable extraction and purity at the same time is highly desired. Aliquat 336 is basically an anionic (basic) liquid exchanger widely used for the extraction of amino acids [15,23].
It is selected because the zeta potential of the pretreated broth remains negative in the pH range 6.5–10 [24]. In fact, some cationic (acidic) exchangers such as di(2ethylhexyl)phosphoric acid have also been employed here, but
Fig. 4. (a) Effect of Aliquat 336 concentration in n-hexane on the extraction of surfactin. (b) Effect of initial aqueous pH on the extraction of surfactin with Aliquat 336 in n-hexane.
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
Fig. 5. Effect of added salt concentration on the stripping of surfactin.
43
they did not effectively extract surfactin from the pretreated broths. Fig. 4(a) shows the effect of Aliquat 336 concentration in n-hexane on the extraction of surfactin. It is found that 5 mM of Aliquat 336 is sufficient to extract more than 92% of initially 3000 mg/L of surfactin into the organic phase. Further increase in Aliquat 336 cannot significantly improve the extraction efficiency. Fig. 4(b) shows the effect of initial aqueous pH on the extraction of surfactin by Aliquat 336 in n-hexane. It was experimentally observed that some macromolecular impurities and/or surfactin started to precipitate when the pH decreased up to 6.0, so the pH was controlled in the range 7–10. The slight dependence of pH on the extraction confirms the results of zeta potential that surfactin molecules are negatively charged in the broths. Unlike physical extraction in LLE systems, surfactin transferred from an aqueous phase into an organic phase in chemical
Fig. 6. (a) FAB mass spectra of surfactin after physical extraction with n-hexane. (b) FAB mass spectra of surfactin after chemical extraction with Aliquat 336 in n-hexane.
44
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
extraction should be generally recovered via a stripping step. Literature survey indicates that the mixtures of alcohols and salt/water are often served as the strip solution to recover proteins and enzymes from the loaded organic solutions [25–28]. For example, Chang and Chen [25] have employed the mixture of n-butanol and potassium chloride/water to strip ␣-amylase and protease from the loaded Aliquat 336/isooctane solution. The mixture of isopropanol and potassium chloride/water has also been applied to strip nattokinase from the loaded AOT/isooctane solution [26]. In fact, the related mechanisms and systems leading to such effects have been recently reviewed [28]. Fig. 5 shows the effect of added amount of salt in salt solution on the stripping of surfactin. Here, the strip solution was
prepared by mixing equal volumes of ethanol (100%) and salt solution (g/50 mL water). It is found that the type of salts has a large effect on the stripping of surfactin from the loaded organic solution. In the case of ammonium sulfate, a relatively high stripping efficiency for surfactin (88%) is obtained at high salt concentrations (15 g/50 mL). Ammonium sulfate in this system plays a role of salting out. As a sufficient amount of ammonium sulfate is added, a large number of water molecules bind to divalent sulfate anions, thereby reducing the amount of water available to bind to proteins. If protein molecules are not hydrated by binding to water molecules, they precipitate [29]. In the case of sodium chloride, nevertheless, a high stripping efficiency (about 90%) is achieved at low salt concentrations
Fig. 7. (a) 1 H NMR spectrograms for surfactin extract after physical extraction with n-hexane. (b) 1 H NMR spectrograms for surfactin extract after chemical extraction with Aliquat 336 in n-hexane.
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
(1.25 g/50 mL). The decreased stripping at high ionic strengths is due to the fact that the negative ions in strip solution compete with the amino acid groups of surfactin molecules. Thus, it is more effective to separate surfactin at lower ionic strengths [23]. It is noticed that the purity of the surfactin powder recovered by chemical extraction with Aliquat 336 can reach up to about 70%. Further improvement in the purity is required from the viewpoint of industrial applications although the extraction (recovery yield) is acceptable. More selective extraction or stripping for surfactin from the mixtures might be achieved by choosing the type of extractant (basic or neutral) with a suitable basicity as well as the diluent (solvent) with a suitable polarity [30]. 3.3. Characterization of the recovered surfactin products An attempt was made to compare the final extract by FAB mass spectrometry and NMR spectroscopy. Fig. 6(a) shows the FAB mass spectra of surfactin after physical extraction with nhexane. The FAB mass spectra display three [M + H]+ peaks at m/z = 1008, 1022, and 1036. They correspond to the well known heptapeptide moiety of surfactin, Glu-Leu-Leu-Val-Asp-LeuLeu, linked to C13 , C14 , and C15 hydroxy fatty acids, respectively [31]. Fig. 6(b) shows the FAB mass spectra of surfactin after chemical extraction. The intensity of the peaks performs lower than those observed in Fig. 6(a), which is likely due to the fact that the quaternary ammonium salt is difficult to ionize on operating mass spectra. The [M + H]+ peaks at m/z = 1022 and 1036 in Fig. 6(a) and (b) are accompanied by the corresponding [M + Na]+ peaks appearing m/z = 1044 and 1058. Both mass spectra are very similar to those reported in the previous studies [31,32]. Fig. 7(a) illustrates 1 H NMR spectrograms of surfactin extract obtained after physical extraction with n-hexane. Three main regions of these spectra correspond to resonance of amide protons, ␣-carbon protons, and side-chain protons [31]. The present NMR spectrograms of surfactin are similar to the standard used in this work (Sigma Co.) [33]. That is, the structure of surfactin is not virtually destroyed or damaged after physical extraction with the organic solvents used. Fig. 7(b) shows 1 H NMR spectrograms of surfactin extracts obtained after chemical extraction. The differences of the peaks in comparison with Fig. 7(a) are at chemical shifts of 3.2–3.5 ppm (i.e., 3.20, 3.30, and 3.45 ppm). This is possibly because some surfactin molecules are bound with the trioctylmethyl ammonium cations of Aliquat 336 [34]. Moreover, its NMR structure is less defined in this spectrogram. Further information such as 12 C NMR spectra is necessary to entirely define the structure of surfactin after chemical extraction. 4. Conclusions After acid precipitation of the fermentation broth of Bacillus subtilis ATCC 21332 culture, surfactin was recovered more than 97% but had a purity of 55% only. The purity of surfactin was thus tried to be improved using either solid–liquid extraction or liquid–liquid extraction after the precipitate was dissolved in
45
NaOH. For LLE with ethyl acetate, an extraction efficiency of 99% was obtained when the feed surfactin concentration was 2 g/L. In the repeated SLE with ethyl acetate, the maximum extraction was 78% and the purity was 84%. The performance in physical SLE and LLE with n-hexane was comparable; for example, the repeated extraction with n-hexane had maximum extraction of 64% and 59%, respectively, and maximum purity of 58% and 63%, respectively. In the pH range 7–10, 5 mM of Aliquat 336 in n-hexane was enough to extract about 92% of 3 g/L of surfactin. Adequate addition of sodium chloride or ammonium sulfate in ethanol/water could recover 90% and 88% of surfactin from the loaded organic solution. The structure of surfactin was not destroyed after physical extraction with organic solvents. However, 1 H NMR spectrograms of the surfactin extract obtained after chemical extraction revealed the existence of the surfactin–Aliquat 336 complex. Acknowledgement Support for this work by the National Science Council, ROC, under Grant NSC95-2214-E-155-057 is gratefully appreciated. References [1] J.D. Desai, I.M. Banat, Microbial production of surfactants and their commercial potential, Microbiol. Mol. Biol. Rev. 61 (1997) 47–64. [2] S. Mukherjee, P. Das, R. Sen, Towards commercial production of microbial surfactants, Trends Biotechnol. 24 (2006) 509–515. [3] S.C. Lin, Biosurfactants: recent advances, J. Chem. Technol. Biotechnol. 66 (1996) 109–120. [4] A. Tiehm, Degradation of polycyclic aromatic hydrocarbons in the presence of synthetic surfactants, Appl. Environ. Microbiol. 60 (1994) 258–263. [5] J.C. Bertrand, P. Bonin, M. Goutx, M. Gauthier, G. Mille, The potential application of biosurfactants in combating hydrocarbon pollution in marine environment, Res. Microbiol. 145 (1994) 53–56. [6] I.M. Banat, R.S. Makkar, S.S. Cameotra, Potential applications of microbial surfactants, Appl. Microbiol. Biotechnol. 53 (2000) 495–508. [7] C. Sandrin, F. Peypoux, G. Michel, Coproduction of surfactin and iturin A, lipopeptides with surfactant and antifungal properties, by Bacillus subtilis, Biotechnol. Appl. Biochem. 12 (1990) 370–375. [8] F. Peypoux, J.M. Bonmatin, J. Wallach, Recent trends in the biochemistry of surfactin, Appl. Microbiol. Biotechnol. 51 (1999) 553–563. [9] D.G. Cooper, B.G. Goldenberg, Surface-active agents from two Bacillus species, Appl. Environ. Microbiol. 53 (1987) 224–229. [10] K. Keller, T. Friedmann, X. Boxman, The bioseparations need for tomorrow, Trends Biotechnol. 19 (2001) 438–441. [11] M.S. Kuyukina, I.B. Ivshina, J.C. Philp, N. Christofi, S.A. Dunbar, M.I. Ritchkova, Recovery of Rhodococcus biosurfactants using methyl tertiarybutyl ether extraction, J. Microbiol. Methods 46 (2001) 149–156. [12] K. Arima, G. Tamura, A. Kakinuma, Surfactin, US Patent 3,687,926 (1972). [13] I. Tadayuki, S. Shoji, Biosurfactant cyclopeptide compound produced by culturing a specific Arthrobacter microorganism, US Patent 5,344,913 (1994). [14] I. Tadayuki, S. Masaaki, Oil-and-fat-degradative Bacillus and effluent treatment, JP Patent 2001061468 (2001). [15] S. Yang, S.A. White, S. Hsu, Extraction of carboxylic acid with tertiary and quaternary amines: effect of pH, Ind. Eng. Chem. Res. 30 (1991) 1335–1342. [16] Y.H. Wei, L.F. Wang, J.S. Chang, S.S. Kung, Identification of induced acidification in iron-enriched cultures of Bacillus subtilis during biosurfactant fermentation, J. Biosci. Bioeng. 96 (2003) 174–178.
46
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
[17] Y.H. Wei, I.M. Chu, Mn2+ improves surfactin production by Bacillus subtilis, Biotechnol. Lett. 24 (2002) 479–482. [18] S.C. Lin, H.J. Jiang, Recovery and purification of the lipopeptide biosurfactant of Bacillus subtilis by ultrafiltration, Biotechnol. Tech. 11 (1997) 413–416. [19] C.N. Mulligan, R.N. Yong, B.F. Gibbs, On the use of biosurfactants for the removal of heavy metals from oil-contaminated soil, Environ. Prog. 18 (1990) 50–54. [20] U. Rau, L.A. Nguyen, H. Roeper, H. Koch, S. Lang, Downstream processing of mannosylerythritol lipid produced by Pseudozyma aphidis, Eur. J. Lipid Sci. Technol. 107 (2005) 373–380. [21] H.S. Kim, B.D. Yoon, D.H. Choung, H.M. Oh, T. Katsurag, H.S. Tani, Characterization of a biosurfactant, MEL, produced from Candida sp. SY16, Appl. Microbiol. Biotechnol. 52 (1999) 713–721. [22] D. Kitamoto, S. Ghosh, O.G.Y. Nakatan, Formation of giant vesicles from diacylmannosylerythritol and their binding to concanavalin A, Chem. Commun. 10 (2000) 861–862. [23] W. Wang, M.E. Weber, J.H. Vera, Reverse micellar extraction of amino acids using dioctyldimethyl-ammonium chloride, Ind. Eng. Chem. Res. 34 (1995) 599–606. [24] Y.S. Lee, Recovery and purification of surfactin from pretreated fermentation broths by ion exchange, Master thesis, Yuan Ze University, Taiwan (2006). [25] Q.L. Chang, J.Y. Chen, Separation and purification of two enzymes from Bacillus subtilis using Aliquat 336 reversed micelles: study of the effect of cosolvent concentration, Chem. Eng. J. 59 (1995) 303–308.
[26] J.G. Liu, J.M. Xing, R. Shen, C.L. Yang, H.Z. Liu, Reverse micellar extraction of nattokinase from fermentation broth, Biochem. Eng. J. 21 (2004) 273–278. [27] C. Dennison, R. Lovrien, Three phase partitioning: concentration and purification of proteins, Protein Expr. Purif. 11 (1997) 149–161. [28] D.S. Mathew, R.S. Juang, Role of alcohols in the formation of inverse microemulsions and the back extraction of proteins/enzymes in a reverse micellar system, Sep. Purif. Technol. 53 (2007) 199–215. [29] M.C. Dawson, Biochemical Research, 2nd ed., Oxford University Press, 1969. [30] X.C. Shan, W. Qin, Y.Y. Dai, Dependence of extraction equilibrium of monocarboxylic acid from aqueous solutions on the relative basicity of extractant, Chem. Eng. Sci. 61 (2006) 2574–2581. [31] F. Peypoux, J. Bonmatin, H. Labbe, B.C. Das, M. Ptak, G. Michel, Isolation and characterization of a new variant of surfactin, the [Val7] surfactin, Eur. J. Biochem. 202 (1991) 101–106. [32] F. Peypoux, J. Bonmatin, H. Labbe, I. Grangemard, B.C. Das, M. Ptak, J. Wallach, G. Michel, [Ala4] Surfactin, a novel isoform from Bacillus subtilis studied by mass and NMR spectroscopies, Eur. J. Biochem. 224 (1994) 89–96. [33] Y.H. Wei, I.M. Chu, Enhancement of surfactin production in iron-enriched media by Bacillus subtilis ATCC 21332, Enzyme Microb. Technol. 22 (1998) 724–728. [34] J.P. Mikkola, P.S. Virtanen, Aliquat 336 – a versatile and affordable cation source for an entirely new family of hydrophobic ionic liquids, Green Chem. 8 (2006) 250–255.
Recovery and separation of surfactin from pretreated fermentation broths by physical and chemical extraction Huei-Li Chen, Ruey-Shin Juang ∗ Department of Chemical Engineering and Materials Science, Yuan Ze University, Chung-Li 32003, Taiwan Received 25 April 2007; received in revised form 5 June 2007; accepted 8 June 2007
Abstract The recovery of surfactin from fermentation broths with the culture of Bacillus subtilis ATCC 21332 by physical and chemical extraction was studied, in which the broths were pretreated by acid precipitation and, if necessary, the precipitate was further dissolved in NaOH solution. The physical solid–liquid and liquid–liquid extractions were performed with different organic solvents (ethyl acetate, n-hexane) and at different times of extraction run. It was shown that better extraction was obtained with ethyl acetate than n-hexane. The extraction could be improved by increasing the times of extraction run under a given volume of the organic solution. For chemical liquid–liquid extraction of surfactin with Aliquat 336 (5–200 mM) in n-hexane, the efficiency was improved. The amounts of inorganic salt in the strip solution were also optimized. Surfactin recovered after physical or chemical extraction was finally characterized by mass and NMR spectroscopies. The results of chemical extraction showed that surfactin would readily bind with the quaternary ammonium cations of Aliquat 336. © 2007 Elsevier B.V. All rights reserved. Keywords: Surfactin; Recovery; Fermentation broths; Extraction; Ethyl acetate; n-Hexane; Aliquat 336
1. Introduction Biological molecules exhibiting particularly high surface activity are classified as biosurfactants. They generally include a wide variety of chemical structures such as glycolipids, lipopeptides, polysaccharide–protein complexes, phospholipids, fatty acids, and neutral lipids [1,2]. They also have a strong effect on interfacial rheological phenomena and mass transfer [3]. Biosurfactants have several advantages, such as low critical micelle concentration and high biodegradability, over chemical synthetic surfactants and, thus, are especially well suited for environmental applications such as bioremediation and the dispersion of oil spills [4,5]. In addition, biosurfactants are thought to be potential candidates to replace chemically synthetic surfactants in the future, particularly in food, cosmetic, and health care industries, as well as in industrial cleaning of products and in agricultural chemicals [6,7]. Surfactin, a cyclic lipopeptide produced by several strains of Bacillus subtilis, is one of the very powerful biosurfactant [7,8]. It has a cyclic lipopeptide with -hydroxy fatty acids linked to
∗
Corresponding author. Tel.: +886 3 4638800; fax: +886 3 4559373. E-mail address: [email protected] (R.-S. Juang).
1369-703X/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bej.2007.06.003
a heptapeptide (l-Glu-l-Leu-d-Leu-l-Val-l-Asp-d-Leu-l-Leu) [3,9]. Generally speaking, surfactin has an exceptional surfaceactive power because it lowers the surface tension of water from 72 to 27 mN/m at a concentration even as low as 20 M [8]. It is recognized that downstream processing in many biotechnological processes is responsible for up to 60% of the total production cost [10]. From the economic point of view, most biosurfactants would have to involve whole-cell spent culture broths or other crude preparations [1]. The most commonly used and cheap methods for the recovery of biosurfactants from fermentation broths are acid precipitation, foam separation, or the combination of both [1]. However, they usually lead to relatively low biosurfactant purity (<50%). An alternative method for biosurfactant recovery is solvent extraction [10]. In practice, a wide variety of organic solvents including methanol, ethanol, butanol, diethyl ether, n-pentane, acetone, acetic acid, chloroform, and dichloromethane have been used, either single or in combination, for this purpose [1]. The most effective solvent would be the mixture of chloroform and methanol in various ratios, which facilitates adjustment of the polarity of extraction solvents to the target extractable material [11]. However, chloroform is highly toxic compound regarded as harmful to the environment and for human health [2]. Thus, cheap and low-toxicity solvents are highly desired for biosurfactant extraction from the view-
40
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
Nomenclature A volume of aqueous solution (mL) Aliquat 336 trioctylmethylammonium chloride LLE liquid–liquid extraction O volume of organic phase (mL) SLE solid–liquid extraction
point of industrial applications. In recent years, ethyl acetate and n-hexane, which are the representatives of polar and nonpolar solvents, have been used to recover biosurfactants produced by B. subtilis [12–14]. Such types of low cost, less toxic, and easily available solvents could be employed to cut the recovery cost and minimize the environmental hazards. In this work, the fermentation broth was pre-precipitated and the precipitate was obtained by centrifuging and ovendrying. In most cases, the precipitate was further dissolved in alkaline solution. Surfactin was then recovered by solid–liquid extraction (SLE) and liquid–liquid extraction (LLE) by the solvents n-hexane and ethyl acetate (physical extraction). Because physical extraction is sometimes not selective enough and has low recovery yield [15], LLE using a commercial basic extractant (chemical extraction) was also carried out for comparison. Higher recovery and purity were simultaneously targeted because the possibility of using such solvent extraction for practical applications was examined. Factors affecting extraction performance including types of organic solvent (ethyl acetate, n-hexane), times of extraction run, extractant concentration (0.5–200 mM) were investigated. The effect of added concentration of sodium chloride or ammonium sulfate (0.5–15 g/50 mL) on the stripping in chemical extraction system was also studied. Finally, the characteristics of the final product obtained after physical and chemical extraction were identified and compared. 2. Materials and methods 2.1. Microorganisms and culture condition Here, B. subtilis ATCC 21332 was selected to produce surfactin. The nutrient broth (NB) medium contained 3 g/L beef extract, 5 g/L peptone, and the mineral salt (MS) medium at pH 7. The MS medium consisted of 40 g/L glucose, 50 mM NH4 NO3 , 30 mM KH2 PO4 , 40 mM Na2 HPO4 , 7 M CaCl2 , 0.8 mM MgSO4 , 4 M FeSO4 , and 4 M tetrasodium salt of EDTA [16]. The pH was regulated at 7 by adding 0.1 M HCl or NaOH. Prior to use, the MS medium and deionized water (Millipore, Milli-Q) were sterilized in autoclave at 121 ◦ C for 15 min. All inorganic chemicals were offered from Merck Co. as analytical reagent grade. Culture of B. subtilis ATCC 21332 was taken from −80 ◦ C frozen stock and transferred onto agar medium for pre-culture. The Bacillus subtilis culture (1 mL) was inoculated into 250-mL flask containing 100 mL of NB medium at 30 ◦ C with 200 rpm of agitation. After growing up to late exponential phase (near 14 h), the NB medium containing B. subtilis cells was inoculated and
fermented in 5-L fermenter (Firstek Scientific Co., Taiwan) with 4-L working volume at 30 ◦ C and 200 rpm for another 4 days. The aeration was controlled at a rate of 1.0 vvm. The dissolved oxygen concentration maintained 80% during the fermentation (2–4 days). The medium pH was maintained at 7.0 by automatic addition of 1 N NaOH or HCl solution to the fermenter according to the signal received from the pH electrode. The fermentation broth was pretreated as follows. First, the liquor inside the fermenter was centrifuged at 10,000 × g to remove biomass impurities; the remaining supernatant was further treated and precipitated by adding 1 M HCl to a pH of around 4. The yellowish precipitate (crude powder) was obtained by centrifuging at 10,000 × g for 15 min and oven-drying at 37 ◦ C for 2 days. In most situations, the dried precipitate was further dissolved in NaOH solution at pH 11 in order to yield a homogeneous solution. It was found that the crude precipitate had a surfactin purity of about 53% according to the method described below. 2.2. Assay surfactin concentration Culture samples were taken after centrifuging at 12,000 × g for 15 min to remove the biomass, and surfactin concentration in the clarified supernatant was measured with HPLC equipped a reverse phase C18 column (5 m, Merck) at 30 ◦ C [17]. The samples were subjected to filtration through a Millipore filter (0.45 m) before analysis. A mixture of 3.8 mM trifluoroacetic acid (20% v/v) and acetonitrile was used as the mobile phase, and the flow rate was 1.0 mL/min. An aliquot of the sample (20 L) was injected and analyzed using an UV detector (Jasco 975, Japan). The wavelength was set at 205 nm [18]. Each concentration analysis was at least duplicated under identical conditions. The reproducibility is mostly within 5%. Surfactin powder purchased from Sigma Co. served as the standard (98% purity as per label claim). The purity of surfactin in the dried sample was calculated by amount of surfactin determined by HPLC
purity(%) =
weight of dried sample powder dissolved in the solution × 98
(1)
The purities of surfactin in the recovered product and in the fermentation broth would be used to calculate the recovery yield of surfactin. 2.3. Physical extraction In SLE, 0.1 g of the dried precipitate was contacted with 25 mL of the organic solvent, n-hexane or ethyl acetate (Mallinckrodt Co.). The solution mixture was agitated using a magnetic stirrer at 250 rpm for 24 h at 30 ◦ C. Preliminary experiments had shown that the extraction studied was complete within 18 h. Then, it was centrifuged at 6000 × g for 10 min and the solvent was collected. For a given volume of the organic solution, the extraction was performed several times by fresh organic solution when organic solution was divided into the corresponding parts. After evaporation of the organic solution, surfactin solid
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
41
was dissolved in NaOH solution at pH 11 and its concentration was analyzed. The extraction efficiency was obtained by amount of surfactin in the organic solution
extraction(%) =
initial amount of surfactin in the feed sample
× 100
(2) In LLE, the pH of aqueous solution was first adjusted to be near 8.0 by the addition of 0.1 M HCl. Equal volumes of the organic solution and aqueous solution (50 mL) were mixed in a tightly stoppered 250-mL flask, and were agitated at 250 rpm for 24 h in water bath of 30 ◦ C. After phase separation, the solvent was removed with a rotary evaporator under reduced pressure conditions to obtain off-white powders. Similarly, the powder was dissolved in NaOH at pH 11 and its concentration was analyzed. Moreover, the fresh organic solution was divided into several parts to examine the effect of repeated extraction runs. 2.4. Chemical extraction Chemical extraction was tested in LLE systems only. The experimental procedures were essentially the same as those described in Section 2.3, except that the organic solution contained various amounts of Aliquat 336 (trioctylmethylammonium chloride, Aldrich Co.) in n-hexane. After phase separation, the upper organic phase was gathered for the followed stripping experiments. The concentration of surfactin in the lower aqueous phase was determined with HPLC, and the extraction efficiency of surfactin was similarly calculated according to Eq. (2). In the stripping experiments, equal volumes of the loaded organic solution and strip solution (50 mL) were mixed, in which the strip solution contained ethanol and sodium chloride or ammonium sulfate. The stripping test was also performed at a stirring speed of 250 rpm for 24 h at 30 ◦ C. 2.5. Characterization of surfactin product The final surfactin powder dissolved in chloroform (CD3 Cl) was detected by NMR spectroscopy (Bruker, Rheinstetler, Germany) to identify the structure of surfactin. FAB-MS (Fast-atom-bombardment mass spectra, model JMS-700 JEOL, Tokyo) was also used to determine the molecular weight of surfactin. 3. Results and discussion 3.1. Physical extraction for the pretreated broth It was reported that the raw fermentation broth (before acid precipitation) with B. subtilis ATCC 21332 contains macromolecules (surfactin micelles, polysaccharides, peptides, proteins, etc.), mid-molecules, and small molecules (MS medium, alcohols, phthalic acid, amino acids, glycine, serine, threonine, alanine, etc.) [19]. It should be reminded in Section 2.1 that the surfactin purity of 53% only could be achieved after acid precipitation alone. Further improvement in purity would be aided by solvent extraction. For simplicity, the total extraction (recovery) of surfactin is selected as the key parameter to evaluate the performance of each type of solvent extraction.
Fig. 1. Effect of the types of organic solvent on the extraction efficiency in LLE system at different initial concentrations of surfactin (A and O refers to the volumes of the aqueous and organic solutions, respectively).
Fig. 1 shows the effect of initial surfactin concentration on the efficiency of LLE using different organic solvents. Here, the volumes of the organic (O) and aqueous solutions (A) are identical. The extraction is much effective with ethyl acetate (>95%) than n-hexane (<21%) probably because n-hexane is a low-polarity solvent. However, the concentration effect on the extraction is both weak under the ranges studied. Rau et al. [20] have used n-hexane to extract mannosylerythritol lipid produced by Pseudozyma aphidis, and also found that the extraction is 8% only. It was also reported previously that the extraction of bioproducts with considerably high polarity by ethyl acetate solvent is rather efficient [21,22]. On the other hand, the recovered surfactin with either nhexane or ethyl acetate has a comparable purity (58–60%), although ethyl acetate gives much higher extraction than nhexane. Surfactin can be soluble in organic solvents due to the presence of hydrophobic end. The higher extraction with ethyl acetate indicates that surfactin and, at least, other protein macromolecules in the broths have highly polar hydrophobic groups. The present results likely suggest that the small molecule impurities in the broth can be first removed by extraction with n-hexane (they are left in the aqueous raffinate), and then surfactin is recovered with a highly polar solvent [12]. Due to the formation of emulsion after LLE, the SLE system was hence tested to avoid such behavior. Fig. 2(a) shows effect of the times of extraction run on the extraction and purity with ethyl acetate in SLE system. It is found that the maximum extraction of 78% is achieved when extraction runs were conducted three times. In this case, the purity of surfactin reaches a plateau of 84%. Fig. 2(b) shows the SLE results with n-hexane. It is also observed that the maximum extraction of 62% is achieved when extraction runs were made three times. On the other hand, the purity of surfactin is about 60%. The different phenomena with these two solvents are still a result of their different miscibility between solvents and the hydrophilic groups of surfactin. Ethyl acetate has an acceptable solubility in water (9.7% v/v), but nhexane is nearly insolubility in water. Therefore, more surfactin molecules are dissolved in ethyl acetate than in n-hexane.
42
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
Fig. 2. (a) Effect of the times of extraction run with ethyl acetate on the extraction efficiency and purity in SLE system. (b) Effect of the times of extraction run with n-hexane on the extraction efficiency and purity in SLE system.
As indicated above, the extraction is low with n-hexane in LLE system although the purity obtained is comparable to the case of ethyl acetate. There is hence a possibility to improve the extraction with n-hexane by increasing the time of extraction runs. Fig. 3 compares the effects of the times of extraction run on the extraction and purity with n-hexane between SLE and LLE systems. It is evident that the extraction in SLE system (64%) is more effective than that in LLE system (59%) at such a surfactin concentration (2000 mg/L). According to these results, it appears that SLE is preferred due to its easy-to-operate nature although SLE gives a slightly lower purity. However, the extraction with n-hexane, in SLE or LLE, is still uneconomical and impractical. 3.2. Chemical extraction for the pretreated broth
Fig. 3. Effect of the times of extraction run with n-hexane on the extraction efficiency and purity in LLE and SLE systems.
The recovery method for obtaining acceptable extraction and purity at the same time is highly desired. Aliquat 336 is basically an anionic (basic) liquid exchanger widely used for the extraction of amino acids [15,23].
It is selected because the zeta potential of the pretreated broth remains negative in the pH range 6.5–10 [24]. In fact, some cationic (acidic) exchangers such as di(2ethylhexyl)phosphoric acid have also been employed here, but
Fig. 4. (a) Effect of Aliquat 336 concentration in n-hexane on the extraction of surfactin. (b) Effect of initial aqueous pH on the extraction of surfactin with Aliquat 336 in n-hexane.
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
Fig. 5. Effect of added salt concentration on the stripping of surfactin.
43
they did not effectively extract surfactin from the pretreated broths. Fig. 4(a) shows the effect of Aliquat 336 concentration in n-hexane on the extraction of surfactin. It is found that 5 mM of Aliquat 336 is sufficient to extract more than 92% of initially 3000 mg/L of surfactin into the organic phase. Further increase in Aliquat 336 cannot significantly improve the extraction efficiency. Fig. 4(b) shows the effect of initial aqueous pH on the extraction of surfactin by Aliquat 336 in n-hexane. It was experimentally observed that some macromolecular impurities and/or surfactin started to precipitate when the pH decreased up to 6.0, so the pH was controlled in the range 7–10. The slight dependence of pH on the extraction confirms the results of zeta potential that surfactin molecules are negatively charged in the broths. Unlike physical extraction in LLE systems, surfactin transferred from an aqueous phase into an organic phase in chemical
Fig. 6. (a) FAB mass spectra of surfactin after physical extraction with n-hexane. (b) FAB mass spectra of surfactin after chemical extraction with Aliquat 336 in n-hexane.
44
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
extraction should be generally recovered via a stripping step. Literature survey indicates that the mixtures of alcohols and salt/water are often served as the strip solution to recover proteins and enzymes from the loaded organic solutions [25–28]. For example, Chang and Chen [25] have employed the mixture of n-butanol and potassium chloride/water to strip ␣-amylase and protease from the loaded Aliquat 336/isooctane solution. The mixture of isopropanol and potassium chloride/water has also been applied to strip nattokinase from the loaded AOT/isooctane solution [26]. In fact, the related mechanisms and systems leading to such effects have been recently reviewed [28]. Fig. 5 shows the effect of added amount of salt in salt solution on the stripping of surfactin. Here, the strip solution was
prepared by mixing equal volumes of ethanol (100%) and salt solution (g/50 mL water). It is found that the type of salts has a large effect on the stripping of surfactin from the loaded organic solution. In the case of ammonium sulfate, a relatively high stripping efficiency for surfactin (88%) is obtained at high salt concentrations (15 g/50 mL). Ammonium sulfate in this system plays a role of salting out. As a sufficient amount of ammonium sulfate is added, a large number of water molecules bind to divalent sulfate anions, thereby reducing the amount of water available to bind to proteins. If protein molecules are not hydrated by binding to water molecules, they precipitate [29]. In the case of sodium chloride, nevertheless, a high stripping efficiency (about 90%) is achieved at low salt concentrations
Fig. 7. (a) 1 H NMR spectrograms for surfactin extract after physical extraction with n-hexane. (b) 1 H NMR spectrograms for surfactin extract after chemical extraction with Aliquat 336 in n-hexane.
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
(1.25 g/50 mL). The decreased stripping at high ionic strengths is due to the fact that the negative ions in strip solution compete with the amino acid groups of surfactin molecules. Thus, it is more effective to separate surfactin at lower ionic strengths [23]. It is noticed that the purity of the surfactin powder recovered by chemical extraction with Aliquat 336 can reach up to about 70%. Further improvement in the purity is required from the viewpoint of industrial applications although the extraction (recovery yield) is acceptable. More selective extraction or stripping for surfactin from the mixtures might be achieved by choosing the type of extractant (basic or neutral) with a suitable basicity as well as the diluent (solvent) with a suitable polarity [30]. 3.3. Characterization of the recovered surfactin products An attempt was made to compare the final extract by FAB mass spectrometry and NMR spectroscopy. Fig. 6(a) shows the FAB mass spectra of surfactin after physical extraction with nhexane. The FAB mass spectra display three [M + H]+ peaks at m/z = 1008, 1022, and 1036. They correspond to the well known heptapeptide moiety of surfactin, Glu-Leu-Leu-Val-Asp-LeuLeu, linked to C13 , C14 , and C15 hydroxy fatty acids, respectively [31]. Fig. 6(b) shows the FAB mass spectra of surfactin after chemical extraction. The intensity of the peaks performs lower than those observed in Fig. 6(a), which is likely due to the fact that the quaternary ammonium salt is difficult to ionize on operating mass spectra. The [M + H]+ peaks at m/z = 1022 and 1036 in Fig. 6(a) and (b) are accompanied by the corresponding [M + Na]+ peaks appearing m/z = 1044 and 1058. Both mass spectra are very similar to those reported in the previous studies [31,32]. Fig. 7(a) illustrates 1 H NMR spectrograms of surfactin extract obtained after physical extraction with n-hexane. Three main regions of these spectra correspond to resonance of amide protons, ␣-carbon protons, and side-chain protons [31]. The present NMR spectrograms of surfactin are similar to the standard used in this work (Sigma Co.) [33]. That is, the structure of surfactin is not virtually destroyed or damaged after physical extraction with the organic solvents used. Fig. 7(b) shows 1 H NMR spectrograms of surfactin extracts obtained after chemical extraction. The differences of the peaks in comparison with Fig. 7(a) are at chemical shifts of 3.2–3.5 ppm (i.e., 3.20, 3.30, and 3.45 ppm). This is possibly because some surfactin molecules are bound with the trioctylmethyl ammonium cations of Aliquat 336 [34]. Moreover, its NMR structure is less defined in this spectrogram. Further information such as 12 C NMR spectra is necessary to entirely define the structure of surfactin after chemical extraction. 4. Conclusions After acid precipitation of the fermentation broth of Bacillus subtilis ATCC 21332 culture, surfactin was recovered more than 97% but had a purity of 55% only. The purity of surfactin was thus tried to be improved using either solid–liquid extraction or liquid–liquid extraction after the precipitate was dissolved in
45
NaOH. For LLE with ethyl acetate, an extraction efficiency of 99% was obtained when the feed surfactin concentration was 2 g/L. In the repeated SLE with ethyl acetate, the maximum extraction was 78% and the purity was 84%. The performance in physical SLE and LLE with n-hexane was comparable; for example, the repeated extraction with n-hexane had maximum extraction of 64% and 59%, respectively, and maximum purity of 58% and 63%, respectively. In the pH range 7–10, 5 mM of Aliquat 336 in n-hexane was enough to extract about 92% of 3 g/L of surfactin. Adequate addition of sodium chloride or ammonium sulfate in ethanol/water could recover 90% and 88% of surfactin from the loaded organic solution. The structure of surfactin was not destroyed after physical extraction with organic solvents. However, 1 H NMR spectrograms of the surfactin extract obtained after chemical extraction revealed the existence of the surfactin–Aliquat 336 complex. Acknowledgement Support for this work by the National Science Council, ROC, under Grant NSC95-2214-E-155-057 is gratefully appreciated. References [1] J.D. Desai, I.M. Banat, Microbial production of surfactants and their commercial potential, Microbiol. Mol. Biol. Rev. 61 (1997) 47–64. [2] S. Mukherjee, P. Das, R. Sen, Towards commercial production of microbial surfactants, Trends Biotechnol. 24 (2006) 509–515. [3] S.C. Lin, Biosurfactants: recent advances, J. Chem. Technol. Biotechnol. 66 (1996) 109–120. [4] A. Tiehm, Degradation of polycyclic aromatic hydrocarbons in the presence of synthetic surfactants, Appl. Environ. Microbiol. 60 (1994) 258–263. [5] J.C. Bertrand, P. Bonin, M. Goutx, M. Gauthier, G. Mille, The potential application of biosurfactants in combating hydrocarbon pollution in marine environment, Res. Microbiol. 145 (1994) 53–56. [6] I.M. Banat, R.S. Makkar, S.S. Cameotra, Potential applications of microbial surfactants, Appl. Microbiol. Biotechnol. 53 (2000) 495–508. [7] C. Sandrin, F. Peypoux, G. Michel, Coproduction of surfactin and iturin A, lipopeptides with surfactant and antifungal properties, by Bacillus subtilis, Biotechnol. Appl. Biochem. 12 (1990) 370–375. [8] F. Peypoux, J.M. Bonmatin, J. Wallach, Recent trends in the biochemistry of surfactin, Appl. Microbiol. Biotechnol. 51 (1999) 553–563. [9] D.G. Cooper, B.G. Goldenberg, Surface-active agents from two Bacillus species, Appl. Environ. Microbiol. 53 (1987) 224–229. [10] K. Keller, T. Friedmann, X. Boxman, The bioseparations need for tomorrow, Trends Biotechnol. 19 (2001) 438–441. [11] M.S. Kuyukina, I.B. Ivshina, J.C. Philp, N. Christofi, S.A. Dunbar, M.I. Ritchkova, Recovery of Rhodococcus biosurfactants using methyl tertiarybutyl ether extraction, J. Microbiol. Methods 46 (2001) 149–156. [12] K. Arima, G. Tamura, A. Kakinuma, Surfactin, US Patent 3,687,926 (1972). [13] I. Tadayuki, S. Shoji, Biosurfactant cyclopeptide compound produced by culturing a specific Arthrobacter microorganism, US Patent 5,344,913 (1994). [14] I. Tadayuki, S. Masaaki, Oil-and-fat-degradative Bacillus and effluent treatment, JP Patent 2001061468 (2001). [15] S. Yang, S.A. White, S. Hsu, Extraction of carboxylic acid with tertiary and quaternary amines: effect of pH, Ind. Eng. Chem. Res. 30 (1991) 1335–1342. [16] Y.H. Wei, L.F. Wang, J.S. Chang, S.S. Kung, Identification of induced acidification in iron-enriched cultures of Bacillus subtilis during biosurfactant fermentation, J. Biosci. Bioeng. 96 (2003) 174–178.
46
H.-L. Chen, R.-S. Juang / Biochemical Engineering Journal 38 (2008) 39–46
[17] Y.H. Wei, I.M. Chu, Mn2+ improves surfactin production by Bacillus subtilis, Biotechnol. Lett. 24 (2002) 479–482. [18] S.C. Lin, H.J. Jiang, Recovery and purification of the lipopeptide biosurfactant of Bacillus subtilis by ultrafiltration, Biotechnol. Tech. 11 (1997) 413–416. [19] C.N. Mulligan, R.N. Yong, B.F. Gibbs, On the use of biosurfactants for the removal of heavy metals from oil-contaminated soil, Environ. Prog. 18 (1990) 50–54. [20] U. Rau, L.A. Nguyen, H. Roeper, H. Koch, S. Lang, Downstream processing of mannosylerythritol lipid produced by Pseudozyma aphidis, Eur. J. Lipid Sci. Technol. 107 (2005) 373–380. [21] H.S. Kim, B.D. Yoon, D.H. Choung, H.M. Oh, T. Katsurag, H.S. Tani, Characterization of a biosurfactant, MEL, produced from Candida sp. SY16, Appl. Microbiol. Biotechnol. 52 (1999) 713–721. [22] D. Kitamoto, S. Ghosh, O.G.Y. Nakatan, Formation of giant vesicles from diacylmannosylerythritol and their binding to concanavalin A, Chem. Commun. 10 (2000) 861–862. [23] W. Wang, M.E. Weber, J.H. Vera, Reverse micellar extraction of amino acids using dioctyldimethyl-ammonium chloride, Ind. Eng. Chem. Res. 34 (1995) 599–606. [24] Y.S. Lee, Recovery and purification of surfactin from pretreated fermentation broths by ion exchange, Master thesis, Yuan Ze University, Taiwan (2006). [25] Q.L. Chang, J.Y. Chen, Separation and purification of two enzymes from Bacillus subtilis using Aliquat 336 reversed micelles: study of the effect of cosolvent concentration, Chem. Eng. J. 59 (1995) 303–308.
[26] J.G. Liu, J.M. Xing, R. Shen, C.L. Yang, H.Z. Liu, Reverse micellar extraction of nattokinase from fermentation broth, Biochem. Eng. J. 21 (2004) 273–278. [27] C. Dennison, R. Lovrien, Three phase partitioning: concentration and purification of proteins, Protein Expr. Purif. 11 (1997) 149–161. [28] D.S. Mathew, R.S. Juang, Role of alcohols in the formation of inverse microemulsions and the back extraction of proteins/enzymes in a reverse micellar system, Sep. Purif. Technol. 53 (2007) 199–215. [29] M.C. Dawson, Biochemical Research, 2nd ed., Oxford University Press, 1969. [30] X.C. Shan, W. Qin, Y.Y. Dai, Dependence of extraction equilibrium of monocarboxylic acid from aqueous solutions on the relative basicity of extractant, Chem. Eng. Sci. 61 (2006) 2574–2581. [31] F. Peypoux, J. Bonmatin, H. Labbe, B.C. Das, M. Ptak, G. Michel, Isolation and characterization of a new variant of surfactin, the [Val7] surfactin, Eur. J. Biochem. 202 (1991) 101–106. [32] F. Peypoux, J. Bonmatin, H. Labbe, I. Grangemard, B.C. Das, M. Ptak, J. Wallach, G. Michel, [Ala4] Surfactin, a novel isoform from Bacillus subtilis studied by mass and NMR spectroscopies, Eur. J. Biochem. 224 (1994) 89–96. [33] Y.H. Wei, I.M. Chu, Enhancement of surfactin production in iron-enriched media by Bacillus subtilis ATCC 21332, Enzyme Microb. Technol. 22 (1998) 724–728. [34] J.P. Mikkola, P.S. Virtanen, Aliquat 336 – a versatile and affordable cation source for an entirely new family of hydrophobic ionic liquids, Green Chem. 8 (2006) 250–255.