Colloids and Surfaces B: Biointerfaces 128 (2015) 568–576

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb

Breaking oil-in-water emulsions stabilized by yeast Guilherme F. Furtado a , Carolina S.F. Picone a,b , Maria C. Cuellar c , Rosiane L. Cunha a,∗ a

Department of Food Engineering, School of Food Engineering, University of Campinas, Campinas 13083-970, SP, Brazil School of Technology, University of Campinas, Limeira 13484-332, SP, Brazil c Department of Biotechnology, Delft University of Technology, Julianalaan 67 Delft, 2628BC, The Netherlands b

a r t i c l e

i n f o

Article history: Received 26 November 2014 Received in revised form 27 February 2015 Accepted 2 March 2015 Available online 9 March 2015 Keywords: Emulsion Stability Demulsification

a b s t r a c t Several biotechnological processes can show an undesirable formation of emulsions making difficult phase separation and product recovery. The breakup of oil-in-water emulsions stabilized by yeast was studied using different physical and chemical methods. These emulsions were composed by deionized water, hexadecane and commercial yeast (Saccharomyces cerevisiae). The stability of the emulsions was evaluated varying the yeast concentration from 7.47 to 22.11% (w/w) and the phases obtained after gravity separation were evaluated on chemical composition, droplet size distribution, rheological behavior and optical microscopy. The cream phase showed kinetic stability attributed to mechanisms as electrostatic repulsion between the droplets, a possible Pickering-type stabilization and the viscoelastic properties of the concentrated emulsion. Oil recovery from cream phase was performed using gravity separation, centrifugation, heating and addition of demulsifier agents (alcohols and magnetic nanoparticles). Long centrifugation time and high centrifugal forces (2 h/150,000 × g) were necessary to obtain a complete oil recovery. The heat treatment (60 ◦ C) was not enough to promote a satisfactory oil separation. Addition of alcohols followed by centrifugation enhanced oil recovery: butanol addition allowed almost complete phase separation of the emulsion while ethanol addition resulted in 84% of oil recovery. Implementation of this method, however, would require additional steps for solvent separation. Addition of charged magnetic nanoparticles was effective by interacting electrostatically with the interface, resulting in emulsion destabilization under a magnetic field. This method reached almost 96% of oil recovery and it was potentially advantageous since no additional steps might be necessary for further purifying the recovered oil. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Several kinds of processes such as bioremediation, two-phase fermentation, aqueous extraction of edible oil from oilseeds and microbial production of diesel and jetfuel replacements can result in an undesirable formation of oil-in-water emulsions. The presence of microbial cells can make the oil separation quite complex since these particles can act as stabilizers, emulsifiers or even producing biosurfactants by the own cells [1–9]. Thus, in these processes the emulsion formation can bring difficulties associated with process yield, quality and efficiency which could lead to a high-cost process and a non-competitive product [4,5]. A good knowledge of the physical and physicochemical properties of emulsions can help to design process conditions leading to a maximum oil separation and recovery. The attractive or

∗ Corresponding author. Tel.: +55 19 35214047; fax: +55 19 35214027.. E-mail address: [email protected] (R.L. Cunha). http://dx.doi.org/10.1016/j.colsurfb.2015.03.010 0927-7765/© 2015 Elsevier B.V. All rights reserved.

repulsive interactions between droplets depend on the electrostatic and steric forces as well as on interfacial tension between the immiscible phases. Therefore, droplets coalescence depends on the emulsion composition and process conditions to obtain this emulsion [10]. Oil separation could require previous flocculation and coalescence steps, which can be achieved by physical or chemical methods, or by a combination of both [11]. The more simple method to promote the creaming of droplets is gravitational separation. However large tanks are necessary to avoid a high separation rate that could promote shear between the droplets and favor reemulsification [12]. On the other hand, a long residence time would be necessary if the creaming rate is low or when the emulsion is kinetically stable. The application of centrifugal forces is a technique widely used in the petroleum industry, allowing separation of small droplets, but with high investment costs and maintenance [13]. Demulsification can also be promoted by increasing the temperature of the emulsion effectively changing the viscosity of the phases [10,14]. Chemical methods affect the interfacial

G.F. Furtado et al. / Colloids and Surfaces B: Biointerfaces 128 (2015) 568–576

properties improving coalescence by adding a constituent which effectively changes the viscosity of the phases and/or the surface charge [15]. The combination of physical and chemical methods has been used to obtain a more effective emulsion breakup. Alcohols have been successfully used for breaking up microbial stabilized emulsions [16,17] and recent studies have shown that the use of magnetic nanoparticles coated with amphiphilic material is effective in the stabilization and destabilization of emulsions. These particles act as emulsifying agents stabilizing the interface but in the presence of a magnetic field these particles migrate rapidly towards to the applied field, destabilizing the droplets interface and thereby inducing coalescence of the droplets [18,19]. This method has shown growing interest due to the rapid and easy separation of complex multiphase systems [20]. In order to provide information leading to the maximum oil separation the use of model systems, with components having known properties, allows a better evaluation of the physical and physicochemical phases and interface properties of the emulsions. In this study, we have studied model systems of emulsions, using hexadecane as the oil phase and baker’s yeast as model microorganism. The emulsions have been evaluated to allow understanding the mechanisms involved in their stabilization, and to further develop the demulsification protocols. Finally, different chemical and physical demulsification methods were performed in order to evaluate oil recovery.

2. Materials and methods 2.1. Materials Ultrapure water from a Millipore Milli-Q system (resistivity 18.2 M/cm) was used. Commercial yeast (Saccharomyces cerevisiae, Itaiquara, Ltd., Tapiratiba, BRAZIL) was purchased in the local market. Hexadecane was supplied by Sigma–Aldrich Co. (St. Louis, USA) and magnetic nanoparticles (EMG 607) by Ferrotec Corp. (Santa Clara, USA). Iso-butanol and ethanol were of analytical grade.

569

2.2. Emulsion preparation The emulsions were prepared by mixing 30% (v/v) of hexadecane to 70% (v/v) of aqueous yeast suspension (10, 20 and 30% w/w) using a mechanical impeller RW20 (IKA, Campinas, Brazil) at 900 rpm (150 s−1 ) during 30 min. These yeast concentrations were chosen to ensure emulsions stability and they were also based on biotechnological processes. The final yeast concentration in emulsions was 7.47, 14.84 and 22.11% (w/w). Such a wide concentration range allowed understanding the yeast role on the emulsion stability. Hexadecane was used because it presents physical properties (as density and viscosity) similar to some oils obtained from fermentative processes as farnesene [21] (a renewable diesel precursor). All emulsions were prepared in triplicate. In addition, an emulsion prepared at a concentration of 7.47% (w/w) was centrifuged at 10,000 rpm (15,317 × g) until complete phase separation (without cream phase). From the hexadecane and aqueous phases recovered after centrifugation, new emulsions were prepared in the same ratios of aqueous and oil phase but with no addition of yeast.

2.3. Yeast characterization Surface hydrophobicity of yeast was determined by spectrophotometric method based on the microbial cells adhesion to hydrocarbons (as hexadecane) [22,23]. First, cells were suspended in phosphate buffer (0.1 M). Next, three milliliters of this cell suspension was added in a glass tube containing 1 ml of hexadecane. The mixture was shaken for 25 s in a vortex mixer and after 30 min the absorbance of the aqueous phase was measured. Absorbance values of samples was performed using a spectrophotometer Spectro Quest 2800 (UNICO, Dayton, USA) in 600 nm wavelength. Results were expressed as percentage of cells adhered to hexadecane in relation to pure hexadecane. The charge density of yeast cells suspended in Milli-Q water (0.005% w/w) was determined in a Nano-ZS Zetasizer equipment (Malvern Instruments, Worcestershire, UK). The interfacial tension between the aqueous yeast suspension (10, 20 and 30% w/w) and oil phase was measured by the pendant drop method using a TrackerS tensiometer (Teclis,

Fig. 1. Kinetics of interfacial tension between water and hexadecane in the presence of different concentrations of yeast (%, w/w) (a), visual appearance of the emulsions after 24 h, creaming index (CI), oil content (OC) and density () of the cream phase (b).

570

G.F. Furtado et al. / Colloids and Surfaces B: Biointerfaces 128 (2015) 568–576

Longessaigne, France). Assays were performed at 25 ± 0.1 ◦ C with the formation of a drop of the aqueous phase in the oil phase. A syringe diameter of 3 mm was used and the drop volume was 10 ␮L. 2.4. Emulsion characterization Emulsions were characterized after phase separation by the creaming index (CI), which was evaluated from the ratio between the volume of cream phase to the initial volume of the emulsion. The oil content (OC) of the separated phases was determined by Bligh and Dyer method [24]. The density of separated phases was measured using a DMA 4500 densimeter (Anton Paar, Graz, Austria) at 25 ± 0.1 ◦ C. The droplets size distribution of the cream phase was determined based on the static light scattering method using a Multi-Angle Static Light-Scattering Mastersizer (Malvern Instruments, Worcestershire, UK). Microscopy was performed on a Carl Zeiss Axio Scope A1 microscope (Zeiss, Oberkochen, Germany). Samples of cream phase were placed in slides covered with glass coverslips and visualized at 400× magnification for qualitative evaluation. Ten images of each slide were obtained. The rheological measurements of the cream phase were performed using a rheometer Physica MCR301 (Anton Paar, Graz, Austria) with a cone-plate

geometry (50 mm, 2◦ angle, truncation 208 ␮m). Flow curves were obtained by an up-down-up steps program with shear rate ranging from 0 to 300 s−1 and the data were fitted to the power law rheological model to obtain the consistency index (k) and the behavior index (n). The viscoelastic properties were evaluated by oscillatory measurements, using a frequency sweep between 0.1 and 10 Hz within the linear viscoelasticity domain. These measurements were done at 25 ± 0.1 ◦ C. The thermal behavior of the samples was also evaluated by oscillatory shear measurements using a temperature sweep between 25 and 90 ◦ C and fixed frequency (1 Hz) and strain within the linear viscoelastic domain. The contribution of the elastic and viscous characteristics was evaluated from storage (G ) and loss (G ) moduli. Evaluation of viscoelastic properties can help understanding the contribution of viscous and elastic components on emulsions stability since these measurements are performed at low deformation. All measurements were done in triplicate. 2.5. Emulsion breakup The cream phase of the emulsions was broken employing physical methods and the combination of physical and chemical methods.

Fig. 2. Particle size distribution by volume and particle size ranges (a) and micrographs (b) of the yeast suspension and cream phase of the emulsions prepared at varied yeast concentration (%, w/w).

G.F. Furtado et al. / Colloids and Surfaces B: Biointerfaces 128 (2015) 568–576

571

Fig. 3. Flow curves (a) and rheological parameters (b) of the cream phase of the emulsions prepared at varied yeast concentration (%, w/w).

2.5.1. Physical methods One part of the cream phase was transferred to graduated cylinders (50 ml) and stored at 25 ± 1 ◦ C during enough time to observe a continuous oil top phase (gravitational effects). Another part of the cream phase was centrifuged in an Allegra 25R centrifuge (Beckman Coulter, Brea, USA) at 3000 and 10,000 rpm (1378 and 15,317 × g, respectively). The cream phase was also centrifuged at 40,000 rpm (150,000 × g) using an ultracentrifuge L8-M (Beckman Coulter, Brea, USA). The phase volume was recorded for 2 h in all centrifugation conditions. A part of the cream phase was transferred to graduated cylinders (10 ml) at 40, 50, and 60 ± 1 ◦ C which were immersed in a water bath with controlled heating. The volume of the separated phases was recorded for 2 h.

was evaluated by the percentage of oil recovery, the droplet size distribution, rheology and zeta potential of the remaining cream phase. 2.6. Statistical analysis The results were evaluated by an analysis of variance (ANOVA), and significant differences (p < 0.05) between the treatments were evaluated by the Tukey procedure. The statistical analyses were carried out using the software STATISTICA 6.0 (Statsoft Inc., Tulsa, USA). 3. Results and discussion

2.5.2. Physical and chemical methods The demulsifier agents used were alcohols and magnetic nanoparticles covered with cationic surfactant. Ethanol and isobutanol were added at concentrations of 10, 30, 50 and 100% (v/v). Magnetic nanoparticles (nominal diameter 10 nm) coated with cationic surfactant were prepared at concentrations of 0.1, 1, 10 and 100% (w/v). Demulsifier agent solutions (1 ml) were added in a graduated cylinder containing 9 ml of the cream phase and the mixture was vortexed for 20 s [25]. Subsequently, the samples were stored at 25 ± 1 ◦ C and the volume of the phases was recorded at certain time intervals. In parallel to gravity separation, some systems with alcohol addition were also centrifuged at 10,000 rpm (15,317 × g) for 10 min. Destabilization of the cream phase containing magnetic nanoparticles was induced by applying a magnetic field using a neodymium magnet (N35). Demulsification efficiency

3.1. Yeast characterization The zeta potential value of the yeast aqueous suspensions in its natural pH (5.11) was −38.72 ± 0.65 mV and the hydrophobicity was 23.16 ± 1.18%. These results suggest that emulsion stability could be associated to electrostatic repulsion between droplets, with minor adhesion of cells to the hexadecane–aqueous interface. Interfacial tension decreased with increasing yeast concentration (Fig. 1a). The presence of a component with surface activity reduces the interfacial tension by its diffusion and adsorption onto the interface of the droplets, decreasing thermodynamically unfavorable contacts between the oil and water [10,26]. The higher yeast concentration also resulted in shorter time required to achieve interfacial tension equilibrium. The interfacial tension

Fig. 4. Frequency sweep (a) and temperature sweep (b) of the cream phase of the emulsions prepared at varied yeast concentration (%, w/w). The diagram shows the storage (G ; full symbols) and loss (G ; empty symbols) moduli.

572

G.F. Furtado et al. / Colloids and Surfaces B: Biointerfaces 128 (2015) 568–576

values become constant in a shorter time for the highest yeast concentration, thereby achieving an interfacial equilibrium. 3.2. Emulsion characterization The mixtures separated in three phases: cream (oil rich), intermediate (water rich) and bottom phase (yeast rich). Fig. 1b shows a higher turbidity of cream and intermediate phase at higher yeast concentration and a tendency to reduce the cream phase volume at higher yeast concentration. However, the creaming index values (CI) showed no significant differences independent of yeast concentration indicating that almost all the hexadecane was in the top phase, which was confirmed by the oil content (OC) values. Increasing density values reflected the increase in yeast concentration and a more packed cream phase with oil droplets completely covered by yeast cells (Fig. 1b). Fig. 1b shows a higher turbidity of cream and intermediate phase at higher yeast concentration and a tendency to reduce the cream phase volume at higher yeast concentration but with no significant differences. Almost all the hexadecane was in the top phase, which was confirmed by the oil content (OC) values and the increasing density values reflected the increase in yeast concentration and a more packed cream phase with oil droplets completely covered by yeast cells (Fig. 1b).

Fig. 2a shows a bimodal size distribution of particles in the cream phase of the emulsions. The first peak corresponds to the yeast cells and the second peak represents the hexadecane droplets. The yeast cells showed mode near 5 ␮m and droplets around 60 ␮m (Fig. 2a and b). The size distribution plays an important role on the stability of oil-in-water emulsions since larger droplets enhance creaming favoring destabilization of emulsions [27]. However the cream phase stabilized by yeast was stable during 7 days despite of the large droplets observed. The cream phase showed a non-Newtonian fluid behavior in spite of water and hexadecane (viscosity of 0.0032 Pa s) being Newtonian fluids. The behavior index (n) obtained from power law equation indicates the degree of deviation from the Newtonian linear rheological behavior. Newtonian fluids show n equals to the unit whilst shear-thinning samples present n < 1. The cream phase showed shear-thinning behavior (Fig. 3a) that is typically observed in concentrated suspensions of solid particles and liquid droplets interacting with each other [10]. Decreasing cream phase viscosity with an increasing shear rate can be associated to the formation of non-spherical droplets, cells deformation or breakage which tend to align under the action of a shear field [28,29]. In the present study the behavior index remained constant with the increase in yeast concentration, suggesting that the rheological properties of

Fig. 5. Oil recovery (%) of the cream phase centrifuged at 3000, 10,000 and 40,000 rpm (1378, 15,317 and 150,000 × g) (a), micrograph, size distribution by volume of the residual cream phase resulting from the centrifugation at 10,000 rpm (15,317 × g), oil content (OC), density () and viscosity () of the residual cream phase (b).

G.F. Furtado et al. / Colloids and Surfaces B: Biointerfaces 128 (2015) 568–576

the cream phase showed a similar dependence of shear rate independent on the initial yeast concentration. However, the values of viscosity and consistency index (k) (Fig. 3b) showed a statistical significant increase at higher yeast concentrations, confirming the increase of emulsion structuration due to a probable higher interaction between the droplets (Fig. 2b). However, the values of viscosity and consistency index (k) (Fig. 3b) showed a statistical significant increase at higher yeast concentrations, confirming the increase of emulsion structuration since smaller droplets size (Fig. 2a) favor higher interaction between them and a more packed system [10]. Oscillatory rheological measurements from the cream phase (Fig. 4a) indicate that all systems presented gel-like behavior (G higher than G throughout the frequency). The cream phase formed with the lowest yeast concentration showed a frequencydependence of the viscoelastic moduli which is associated to a weaker gel or a less stable emulsion (G /G lower than the other yeast concentration). However at higher yeast concentration (14.84 and 22.11% w/w) the elastic moduli (G ) was close and typical of strong gels [30]. Such increase of the elastic moduli could be associated to the formation of a droplets network due to the emulsions packing (Fig. 2b). Emulsions stabilized by colloidal particles can show adhesion between the dispersed phase droplets through the sharing of the colloidal particles providing an important stabilizing mechanism that hinders film drainage and retards drop/drop coalescence [30,31]. Therefore, viscoelastic parameters (Fig. 3a) indicate the presence of a weak elastic network structure [30]. The thermal rheological behavior of the cream phase presented in Fig. 4b shows that the systems were predominantly elastic at the beginning of heating (G > G ). All systems presented similar G values between 15 and 40 ◦ C but the cream phase containing the lowest concentration of yeast showed network weakening from 40 ◦ C while at higher yeast concentrations such network weakening occurred at temperatures higher than 60 ◦ C. The increase in the

573

storage modulus between 40 and 60 ◦ C could be attributed to the aggregation of droplets favored by hydrophobic or covalent interactions caused by the initial denaturation process of yeast which occurs between 50 and 55 ◦ C [32,33]. At 60 ◦ C the proteins from yeast cells could be mostly denatured leading to a protein network formation onto the interface with decreasing emulsifying properties [34,35]. These results indicate that a further increase of temperature could lead to the rupture of the emulsion due to the interface destabilization caused by the favored yeast aggregation. As consequence occurred a dramatic decrease in both viscoelastic functions. The droplets interface in the cream phase was completely covered by yeast cells (Fig. 2b) indicating that the yeast cells could be acting as stabilizing agents. Emulsions stabilized by inert particles are called Pickering emulsions [36] and the microorganisms allocated on the interface can be enabling a Pickering-type stabilization [37]. The adhesion of microorganisms in different types of interface has been explained by the classical theory of colloidal stability (DLVO theory). In such theory adhesion is governed by long-range interactions between the particles and liquid involving hydrophobic, acid–base and electrostatic interactions [38]. Some authors reported that Pickering stabilization does not affect the interfacial tension between the phases [2,39]. However some studies have identified a considerable reduction of the interfacial tension by cell adhesion, even without the production of bioemulsifiers [40,41]. Hence, the interfacial tension reduction observed at increased yeast concentration (Fig. 1a) could be associated to the cell adhesion onto the oil–water interface. In order to confirm if only cell adhesion was responsible for the emulsion stabilization a new emulsion was prepared from the separated phases without yeast. The recovered hexadecane mixed with water did not result in the formation of an emulsion, confirming that compounds responsible for the stability of the systems were

Fig. 6. Oil recovery (%) (a), volume size distribution of the residual cream phase after destabilization with different temperatures and rheological parameters of the cream phase subjected to different temperatures of destabilization (b).

574

G.F. Furtado et al. / Colloids and Surfaces B: Biointerfaces 128 (2015) 568–576

not lipophilic. The recovered aqueous fraction mixed with hexadecane resulted in an emulsion with two separated phases, showing a similar creaming index to the initial emulsion containing yeast. This result suggests that hydrophilic components are being produced or released by yeast, and that they could be also partly responsible for the stability of the cream phase formed after phase separation of the emulsion. Mannoproteins as minor components released by yeast could be associated to emulsion stability [42,43]. Mannoproteins are regarded as a filling material linked to the structural network of glucans and can be extracted from the cell wall without causing major changes in cell shape, because they are present in the outer surfaces of the cells determining most of their surface properties [44,45]. Therefore, the stabilization of emulsions by yeast at the evaluated cell concentration range could be associated to a combination of different mechanisms: yeast adhesion (Pickering-type), electrostatic repulsion between droplets and viscoelastic properties of the emulsion. 3.3. Demulsification procedures The demulsification aimed at destabilizing the cream phase since it is a necessary step for oil recovery. Based on the droplet size distribution (Fig. 2a), micrographs (Fig. 2b) and rheological properties of the cream phase, the model system containing 7.47% (w/w)

of yeast was chosen to be used in the demulsification tests because this concentration was enough to present droplets with reduced size, complete coverage of yeast cells on the interface and a similar rheological behavior between concentrations. Under gravity separation, the volume of the phases did not change and only after 7 days of observation a small oil continuous phase (less than 1 ml) was formed. The packing of droplets in the cream phase was not enough to induce a narrow interfacial film formation with a subsequent rupture of this film which joined the droplets to form a continuous phase [10]. A partial separation of hexadecane was observed using centrifugation (Fig. 5a), but a residual cream phase remained even at the highest centrifugation speed used. This residual cream phase presented similar droplet size to the initial emulsion but a smaller volume of yeast cells was observed (Fig. 5b). The oil content (OC), density and viscosity values showed a slight increase compared to the initial values (Fig. 5b) indicating the formation of a more concentrated emulsion thus making it more difficult to destabilize. It was possible to reach approximately 70% separation of oil after 2 h using 15,317 × g of centrifugation force. Centrifugation has been reported as an accelerated stability test [45], where the application of elevated centrifugal forces causes the separation of the emulsion phases since oil droplets can be distorted to polyhedral shapes with permanent loss of spherical shape [46]. The lower g forces

Fig. 7. Oil recovery (%, w/w) from cream phase using varying concentrations of ethanol and iso-butanol (a), followed by centrifugation at 10,000 rpm (15,317 × g) for 10 min and volume size distribution of the residual cream phase after destabilization with ethanol followed by centrifugation (b).

G.F. Furtado et al. / Colloids and Surfaces B: Biointerfaces 128 (2015) 568–576

575

Fig. 8. Oil recovery (%, w/w) from the cream phase (a) and zeta potential of the residual cream phase (b) using varied concentration of magnetic nanoparticles.

used were probably not sufficient to cause a pronounced change in the shape of droplets which may be related to the high viscosity of the continuous medium [47]. However the use of a force 10 times higher (150,000 × g) during 2 h was enough to obtain a complete separation of the components, breaking the elastic network formed between the droplets that kept the stability of the cream phase. The cream phase subjected to temperature variations did not show good performance for oil separation within the time–temperature range used, since at the highest temperature (60 ◦ C) it was possible to obtain an oil separation of only 20% (Fig. 6a) after 2 h. The temperature increase caused the decrease of viscosity of the cream phase (Fig. 6b) and consequently its continuous and dispersed phase [48] leading to increase of the droplet size (Fig. 6b). Therefore, despite of the elastic network weakening observed in the thermal rheological behavior of the cream phase (Fig. 4b), it was not sufficient for complete oil separation. Fig. 7a shows the oil recovery from ethanol and iso-butanol addition at varied concentration. It was possible to obtain 96% of separated oil with iso-butanol addition at a concentration of 10% (v/v). However, ethanol addition at the same concentration resulted in approximately 10% oil recovery. Butanol has the ability to alter the kinetic equilibrium of emulsions leading to an immediate phase separation [17]. The addition of alcohols increased the droplets coalescence probability due to the presence of polar and nonpolar moieties in the structure of these components. Thus, the alcohol molecules are able to destabilize the emulsions by displacing surfactants from the interface or connecting to the chain thereof [10]. In addition, the longer carbon chain of iso-butanol relative to that of ethanol might have led to a higher affinity for nonpolar phase and consequently, a higher efficiency to destabilize the cream phase. Oil recovery was approximately 84% using ethanol at a concentration of 10% (v/v) with a subsequent centrifugation step (Fig. 7b). The oil recovery was the same using butanol at concentration of 10% (v/v) followed by centrifugation. However, it is necessary to use a last step of liquid–liquid extraction or distillation for obtaining pure oil in both cases. The residual cream phase showed a similar particle size distribution (Fig. 7b) after the centrifugation process independent of ethanol concentration. This result indicates that a higher amount of ethanol would produce a minor effect on oil recovery. As shown in Fig. 8a, addition of a suspension of magnetic nanoparticles (initial charge density +39.65 ± 0.52 mV) to the cream phase led to an oil separation of approximately 96% using a concentration of 1% (w/v). Magnetic nanoparticles coated with surfactants form complexes at the interface that quickly migrate

towards an applied magnetic field, facilitating the coalescence of the droplets and the consequent formation of a continuous oil phase. A higher concentration (10% w/v), on the other hand, resulted in 88% oil recovery. This might be explained by analyzing the zeta potential values of the residual cream phase (Fig. 8b). Zeta potential was around +20 mV at the concentration of 10% (w/v) providing electrostatic stability to the droplets. However, the zeta potential was near to zero at 1% (w/v), favoring the interface instability by reducing the electrostatic repulsion between the droplets. Electrostatic attractive interaction was very low at the lowest concentrations but enough to cause some oil separation. 4. Conclusions The kinetic stability of emulsions composed by hexadecane, water and yeast was ensured by stabilizing mechanisms such as electrostatic repulsion between the droplets, a possible Pickering-type stabilization and the viscoelastic properties of the concentrated emulsion. To achieve the complete recovery of the oil using physical methods it is necessary to apply long times and/or high centrifugal forces. A temperature increase to 60 ◦ C allowed beginning the destabilization process of the cream phase but it was not enough for an efficient oil separation. The use of butanol followed by centrifugation allowed almost complete separation of the emulsions, but a subsequent step to remove the solvent might be required in order to obtain pure oil. Addition of magnetic nanoparticles has been shown to be effective in interacting electrostatically with the yeast cells and the hydrophilic compounds responsible for the stability of the emulsions. This method led to 96% of oil recovery and it is potentially advantageous, since no additional steps are required for further purifying the recovered oil. The difficulties observed in oil recovery show the importance of detailed knowledge of the emulsion stabilization mechanisms, especially in complex systems such as those from biotechnological processes. Acknowledgements This work was supported by the Fundac¸ão de Amparo à Pesquisa e Desenvolvimento de São Paulo–Brazil (2011/51707-1 and 2012/14003-9) and by the Conselho Nacional de Desenvolvimento Científico e Tecnológico–Brazil (305477/2012-9 and 130752/2012-6). Furthermore, this work was partly carried out within the BE-Basic R&D Program, which was granted a FES subsidy from the Dutch Ministry of Economic affairs, agriculture and innovation (EL&I).

576

G.F. Furtado et al. / Colloids and Surfaces B: Biointerfaces 128 (2015) 568–576

References [1] L. Lin, Z. Cunshan, S. Vittayapadung, S. Xiangqian, D. Mingdong, Opportunities and challenges for biodiesel fuel, Appl. Energy 88 (2011) 1020–1031. [2] L.S. Dorobantu, A.K. Yeung, J.M. Foght, M.R. Gray, Stabilization of oil–water emulsions by hydrophobic bacteria, Appl. Environ. Microbiol. 70 (2004) 6333–6336. [3] N.S.O.C.A. Renninger, J.D.B.C.A., Newman, K.K.O.C.A. Reiling, Fuel Components, Fuel Compositions and Methods of Making and Using Same, US, 2010. [4] K. Koczo, B. Falk, A. Palumbo, Demulsifying composition, useful e.g. for separating emulsions, and in petroleum industry to break water-in-oil and oil-in-water emulsions, comprises a blend of an emulsion, and silicon containing copolymer comprising a silicon compound, Momentive Performance Materials Inc. (MOME-Non-standard), pp. 18. [5] S. Lim, L.K. Teong, Recent trends, opportunities and challenges of biodiesel in Malaysia: an overview, Renewable Sustainable Energy Rev. 14 (2010) 938–954. [6] N.N.A.N. Yusuf, S.K. Kamarudin, Z. Yaakub, Overview on the current trends in biodiesel production, Energy Convers. Manage. 52 (2011) 2741–2751. [7] M. Balat, H. Balat, Progress in biodiesel processing, Appl. Energy 87 (2010) 1815–1835. [8] M. Balat, Potential alternatives to edible oils for biodiesel production—a review of current work, Energy Convers. Manage. 52 (2011) 1479–1492. [9] A.S. Heeres, C.S.F. Picone, L.A.M. van der Wielen, R.L. Cunha, M.C. Cuellar, Microbial advanced biofuels production: overcoming emulsification challenges for large-scale operation, Trends Biotechnol. 32 (2014) 221–229. [10] D.J. MCClements, Food Emulsions: Principles, Practice, and Techniques, CRC Press, Washington, 2005. [11] R.J. Hunter, Foundations of Colloid Science, Oxford University Press, 2001. [12] L. Schlieper, M. Chatterjee, M. Henschke, A. Pfennig, Liquid–liquid phase separation in gravity settler with inclined plates, AIChE J. 50 (2004) 802–811. [13] R. Grace, Commercial emulsion breaking, Adv. Chem. Ser. (1992) 313–339. [14] C. Brunschwig, W. Moussavou, J. Blin, Use of bioethanol for biodiesel production, Prog. Energy Combust. Sci. 38 (2012) 283–301. [15] S.S. Lee, M. Nakamura, Y. Takeuchi, Demulsification of o/w emulsion and subsequent water-treatment using powder adsorbent, Korean J. Chem. Eng. 6 (1989) 81–87. [16] S.S. Chandran, J.T. Kealey, C.D. Reeves, Microbial production of isoprenoids, Process Biochem. 46 (2011) 1703–1710. [17] K. Larson, B. Raghuraman, J. Wiencek, Electrical and chemical demulsification techniques for microemulsion liquid membranes, J. Membr. Sci. 91 (1994) 231–248. [18] W.F. Souza, M.C. Pereira, L.C.A. Oliveira, Amphiphilic catalysts based on onionlike carbon over magnetic iron oxide for petrochemical industry use, Fuel 96 (2012) 604–607. [19] A.P.C. Teixeira, A.D. Purceno, A.S. Barros, B.R.S. Lemos, J.D. Ardisson, W.A.A. Macedo, E.C.O. Nassor, C.C. Amorim, F.C.C. Moura, M.G. Hernandez-Terrones, F.M. Portela, R.M. Lago, Amphiphilic magnetic composites based on layered vermiculite and fibrous chrysotile with carbon nanostructures: application in catalysis, Catal. Today 190 (2012) 133–143. [20] J. Peng, Q. Liu, Z. Xu, J. Masliyah, Novel magnetic demulsifier for water removal from diluted bitumen emulsion, Energy Fuels 26 (2011) 2705–2710. [21] L. Caspeta, J. Nielsen, Economic and environmental impacts of microbial biodiesel, Nat. Biotechnol. 31 (2013) 789–793. [22] M. Rosenberg, D. Gutnick, E. Rosenberg, Adherence of bacteria to hydrocarbons: a simple method for measuring cell-surface hydrophobicity, FEMS Microbiol. Lett. 9 (1980) 29–33. [23] S.R. DeSousa, C. Laluce, M. Jafelicci Jr., Effects of organic and inorganic additives on flotation recovery of washed cells of Saccharomyces cerevisiae resuspended in water, Colloids Surf. B 48 (2006) 77–83. [24] E.G. Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification, Can. J. Biochem. Physiol. 37 (1959) 911–917. [25] N. Nadarajah, A. Singh, O.P. Ward, De-emulsification of petroleum oil emulsion by a mixed bacterial culture, Process Biochem. 37 (2002) 1135–1141.

[26] I˙ . Gülseren, M. Corredig, Interactions of chitin nanocrystals with ␤lactoglobulin at the oil–water interface, studied by drop shape tensiometry, Colloids Surf. B 111 (2013) 672–679. [27] G. Gutiérrez, A. Lobo, D. Allende, A. Cambiella, C. Pazos, J. Coca, J.M. Benito, Influence of coagulant salt addition on the treatment of oil-in-water emulsions by centrifugation, ultrafiltration, and vacuum evaporation, Sep. Sci. Technol. 43 (2008) 1884–1895. [28] C.W. Macosko, Rheology: Principles, Measurements and Applications, VCH, Publishers, New York, 1994. [29] D. Quemada, C. Berli, Energy of interaction in colloids and its implications in rheological modeling, Adv. Colloid Interface Sci. 98 (2002) 51–85. [30] L.G. Torres, R. Iturbe, M.J. Snowden, B.Z. Chowdhry, S.A. Leharne, Preparation of o/w emulsions stabilized by solid particles and their characterization by oscillatory rheology, Colloids Surf. A 302 (2007) 439–448. [31] E.J. Stancik, G.G. Fuller, Connect the drops: using solids as adhesives for liquids, Langmuir 20 (2004) 4805–4808. [32] J.E. Moros, F. Cordobés, J.M. Franco, C. Gallegos, Enhancement of gel strength by application of thermal treatments in highly flocculated emulsions, Food Hydrocolloids 17 (2003) 199–206. [33] P. Sceni, G.G. Palazolo, M.d.C. Vasallo, M.C. Puppo, M.A. Otero, J.R. Wagner, Thermal and surface behavior of yeast protein fractions from Saccharomyces cerevisiae, LWT - Food Sci. Technol. 42 (2009) 1098–1106. ˜ [34] M.A. Otero, J.R. Wagner, M.C. Vasallo, M.C. Anón, L. García, J.C. Jiménez, J.C. López, Thermal denaturation kinetics of yeast proteins in whole cells of Saccharomyces cerevisiae and Kluyveromyces fragilis, Food Sci. Technol. Int. 8 (2002) 163–167. [35] V. Raikos, Effect of heat treatment on milk protein functionality at emulsion interfaces. A review, Food Hydrocolloids 24 (2010) 259–265. [36] S.U. Pickering, Emulsions, J. Chem. Soc. Trans. 91 (1907) 2001–2021. [37] E. Dickinson, Use of nanoparticles and microparticles in the formation and stabilization of food emulsions, Trends Food Sci. Technol. 24 (2012) 4–12. [38] M. Hermansson, The DLVO theory in microbial adhesion, Colloids Surf. B 14 (1999) 105–119. [39] G. Mohebali, A. Ball, A. Kaytash, B. Rasekh, Stabilization of water/gas oil emulsions by desulfurizing cells of Gordonia alkanivorans RIPI90A, Microbiology 153 (2007) 1573–1581. [40] X. Huang, K. Peng, Y. Feng, J. Liu, L. Lu, Separation and characterization of effective demulsifying substances from surface of Alcaligenes sp. S-XJ-1 and its application in water-in-kerosene emulsion, Bioresour. Technol. 139 (2013) 257–264. [41] N.G. Vilkova, A.V. Nushtaeva, Influence of hydrophobized solid particles on the reduction of interfacial tension, Mendeleev Commun. 23 (2013) 155–156. [42] D.R. Cameron, D.G. Cooper, R.J. Neufeld, The mannoprotein of Saccharomyces cerevisiae is an effective bioemulsifier, Appl. Environ. Microbiol. 54 (1988) 1420–1425. [43] J.A.T. Barriga, D.G. Cooper, E.S. Idziak, D.R. Cameron, Components of the bioemulsifier from S. cerevisiae, Enzyme Microb. Technol. 25 (1999) 96–102. [44] H. Zlotnik, M.P. Fernandez, B. Bowers, E. Cabib, Saccharomyces cerevisiae mannoproteins form an external cell wall layer that determines wall porosity, J. Bacteriol. 159 (1984) 1018–1026. [45] M.P. Schreuder, A.T.A. Mooren, H.Y. Toschka, C. Theo Verrips, F.M. Klis, Immobilizing proteins on the surface of yeast cells, Trends Biotechnol. 14 (1996) 115–120. [46] S.A. Basha, K.R. Gopal, S. Jebaraj, A review on biodiesel production, combustion, emissions and performance, Renewable Sustainable Energy Rev. 13 (2009) 1628–1634. [47] A. Demirbas, Progress and recent trends in biodiesel fuels, Energy Convers. Manage. 50 (2009) 14–34. [48] G. Chen, D. Tao, An experimental study of stability of oil–water emulsion, Fuel Process. Technol. 86 (2005) 499–508.

Breaking oil-in-water emulsions stabilized by yeast.

Several biotechnological processes can show an undesirable formation of emulsions making difficult phase separation and product recovery. The breakup ...
2MB Sizes 3 Downloads 11 Views