Formation, Stability, and Mechanical Properties of Bovine Serum ...

Article pubs.acs.org/Langmuir

Formation, Stability, and Mechanical Properties of Bovine Serum Albumin Stabilized Air Bubbles Produced Using Coaxial Electrohydrodynamic Atomization S. Mahalingam,*,†,‡ M. B. J. Meinders,† and M. Edirisinghe‡ †

TopInstitute Food and Nutrition (TIFN), P.O. Box 557, Wageningen, 6700 AN, The Netherlands Department of Mechanical Engineering, University College London, London WC1E 7JE, United Kingdom

‡

ABSTRACT: Bovine serum albumin (BSA) microbubbles were generated using coaxial electrohydrodynamic atomization (CEDHA) using various concentrations of BSA solutions. The bubble characteristics and the long-term stability of the microbubbles were studied through adjustment of processing parameters and the collection media. Bubbles in the range of 40−800 μm were obtained in a controlled fashion, and increasing the flow rate of the BSA solution reduced the polydispersity of the microbubbles. Use of distilled water−glutaraldehyde, glycerol, and glycerol−Tween 80 collection media allowed a remarkable improvement in bubble stability compared to BSA solution collection medium. Possible physical mechanisms were developed to explain the stability of the microbubbles. The collection distance showed a marked influence on stability of the microbubbles. Near-monodisperse particle-reinforced microbubbles were formed with various concentrations of 2,2′-azobis(isobutyramidine) dihydrochloride (AIBA)−polystyrene particle in BSA solution. The bubble size and the size distribution showed negligible change over a period of time irrespective of the concentration of particles at the bubble surface. The compression stiffness of the microbubbles was determined using nanoindentation at ambient temperature and showed that the stiffness of the microbubbles increased from 8 N/m to 20 N/m upon changing the concentration of BSA solution from 5 wt % to 15 wt %. micro- and nano-encapsulated structures.12 Microbubbles prepared using this robust and adaptable technique depend on fluid properties and processing parameters such as flow rate and applied voltage.10 While sonication generates a high yield of microbubbles with a wide size distribution, microfluidic techniques offer excellent size control of the microbubbles. Recent work13 has also helped immensely to enhance microfluidic microbubble yield. CEDHA seems to be in between sonication and microfluidic methods; it offers high microbubble yield, but bubble size is not as monodisperse as in the microfluidics methods. However, the technique can be easily adapted to generate a variety of multilayered structures which can be commercially very attractive, e.g., in biomedical engineering.14 The stability of microbubbles is vital for many applications; thus, any shrinkage due to gas diffusion from the bubbles will impair their effectiveness, and the deterioration of the shell material results in premature release of encapsulated material.15 It has been shown that stability of the CEDHA prepared microbubbles could be improved by saturating the collection fluid and by using solid particles in the bubbling suspension.16,17 The mechanical properties of the microbubbles are an important quantity that essentially gives information about bubble deformation and the elasticity of the shell material. Mechanical and viscoelastic properties of the microbubbles

1. INTRODUCTION Microbubbles are spherical gas-filled structures characterized by their core−shell composition. They have been used in diagnostic applications, in particular, as ultrasound contrasting agents for imaging.1,2 In therapeutic applications, they are very promising and potential vehicles for drug and gene delivery where targeting molecules such as antibodies, peptides, or genes could be delivered at specific sites in a controlled manner.3−5 In the food industry, surfactant-stabilized microbubbles are used for protein recovery.6 Indeed they have been shown to improve the desired properties in food systems, including texture, digestibility, and flavor intensity.7,8 It has been an enormous challenge to make microbubbles and potentially coat them to attain functionality in foods. The physical stability and amount of air incorporated in the microbubbles can affect the appearance and shelf life of food. The addition of such bubbles, in sufficiently small amounts, significantly alters the macroscopic rheological properties, while it can lead to a healthier food with a similar taste. Additionally, microbubbles coated with nutritional ingredients or drugs could help to improve the nutrition or act as a medicinal aid in food.9 Coaxial electrohydrodynamic atomization (CEHDA) is a well-known and well-established technique where two coflowing media are subjected to a high voltage under ambient conditions to generate coaxial jetting and subsequently jet break up to form different products. CEDHA was used to generate microbubbles for the first time in 2007. 10 Subsequently, it was adapted to produce lipid-coated microbubbles.11 CEDHA is also very useful for the preparation of © 2014 American Chemical Society

Received: March 28, 2014 Revised: May 15, 2014 Published: May 19, 2014 6694

dx.doi.org/10.1021/la5011715 | Langmuir 2014, 30, 6694−6703

Langmuir

Article

generated by other techniques have been reported and the stiffness and elastic modulus of the shell material have been evaluated.18−21 However, the long-term stability and mechanical properties of the CEDHA generated microbubbles are lacking in the literature. Bovine serum albumin (BSA) is a low cost protein and has the ability to bind to a variety of biological molecules and impart stability. The primary aim of this work is to study the long-term stability of food grade BSA microbubbles in various collection media at a low temperature (5 °C). We also aim to determine how the processing parameters (such as solution concentration, flow rate, and applied voltage) affect the stability of the microbubbles. In addition, we also report for the first time the mechanical properties of CEDHA generated microbubbles using nanoindentation.

2. MATERIALS AND METHODS 2.1. Materials. Food grade bovine serum albumin (BSA, >96% lyophilized powder, essentially fatty acid and globulin free), phosphatebuffered saline (PBS; one tablet in 200 mL of distilled water gives 137 mM NaCl, 2.7 mM KCl, and 10 mM phosphate buffer solution), cross-linking agent glutaraldehyde and glycerol were supplied by Sigma-Aldrich (Dorset, UK). 2 wt % Tween 80 surfactant was obtained from Lind-Chem Ltd. (Great Yarmouth, UK). The 2,2′azobis(isobutyramidine) dihydrochloride (AIBA)−polystyrene (PS) particles in water were provided by TIFN/Wageningen University and Research Centre, Wageningen, 6700 AN, The Netherlands. 2.2. Solution Preparation and Characterization. 5, 10, and 15 wt % BSA solutions were prepared using PBS. The solutions were heated to 60 °C and mechanically stirred for 2 h prior to CEDHA. Differential scanning calorimeter (DSC) studies of BSA solutions was performed using Universal TA Instruments (USA) equipment. ∼1.8 mg samples were loaded in the cell and tested in a temperature range of 0−100 °C. The heating rate for the studies was chosen as 5 °C/min. In addition, BSA solutions were characterized for their density, surface tension, viscosity, electrical conductivity, and pH immediately after preparation. The densities of the samples were calculated using a standard 10 mL density bottle (VWR, UK). The static surface tension was measured using a Kruss tensiometer K9 (Krus GmbH, Germany), the Wilhelmy plate method was adapted to record the surface tension values of the samples. Electrical conductivity and pH of the samples were assessed using a Jenway 3540 Instruments (UK) device with respective probes. All measurements were performed at ambient temperature (∼23 °C). The particle jammed microbubbles were prepared using AIBA-PS particles with a wt % of 0.15 and 1.5 added to 15 wt % BSA solution. Again, these solutions were heated to 60 °C and mechanically stirred for 2 h prior to CEDHA. 2.3. Processing. The CEDHA setup used is shown in Figure 1. The coaxial device consisted of an inner needle of 0.2 mm and an outer needle of 0.5 mm. Air was infused through the inner needle of the stainless steel coaxial device and the BSA solutions was passed through the outer needle using precision Harvard syringe pumps. The inner needle was placed 2 mm above the exit of the outer needle. As both fluids are allowed to flow simultaneously with optimized infusion rates, the air becomes encapsulated within the protein solution at the orifice, forming bubbles. These bubbles were further atomized by subjecting the flow to an electric field using a Glassman high voltage (kV) unit coupled to the needles. For this study, BSA solutions, distilled water−glutaraldehyde, glycerol, and glycerol−Tween 80 solutions were used as collection media. The microbubbles were collected using vials containing various media that were placed below the coaxial needle at appropriate distances. 2.4. Structural Characterization. Once the microbubbles were collected, they were sealed and stored at ∼5 °C. The microbubbles and their stability were observed using a Nikon Eclipse ME600 optical microscope over a period of time. For studies on stability of the rehydrated microbubbles, the microbubbles were collected on glass slides. They were dried (for 2 h) and stored for 15 days at ambient

Figure 1. Needle setup performing CEDHA to generate BSA microbubbles.

temperature. They were rehydrated in deionized water after a certain period of time (5 days, 10 days, and 15 days) at ambient temperature. More than 100 microbubbles were studied and averaged in each case. Fourier-transform infrared (FTIR) spectroscopy was utilized to assist in structural characterization. FTIR measurements were obtained with a PerkinElmer 2000 unit in a wavenumber range 400−4000 cm−1 with a resolution of 4 cm−1 after accumulation of 30 scans. The spectra were taken in transmission mode for all samples. Background correction for each measurement was done with nitrogen gas. UV/ vis spectroscopy was performed on the microbubble samples collected on glass slides. The transmittance spectra were obtained using a PerkinElmer Lambda 35 spectrometer in the wavelength range 200− 700 nm. 2.5. Mechanical Testing. The nanoindentation on the microbubbles was carried out using an Atomic Force Microscope (Bruker3100, Bruker Instruments, UK) at ambient temperature. Before carrying out the experiments the instrument was calibrated using the standard Veeco silicon tip and fused silica glass. The laser sensitivity was determined by pressing the cantilever on a hard surface. The spring constant was 25.09 N/m and obtained by thermal tuning. The indentation load and the displacement were continuously recorded during one complete cycle of loading and unloading. The tip velocity was kept constant at 0.5 μm/s during the measurements. The microbubbles were loaded and unloaded for each case and the resulting force−displacement was plotted. The instrument was operated in a closed loop. The thermal drift of the transducer was corrected during each measurement.22 Five microbubbles were measured for each condition and in the size ranging between 150 and 200 μm. 6695


Langmuir

Article

3. RESULTS AND DISCUSSION Figure 2 shows the DSC curve for the BSA solutions. The denaturation temperature of the BSA was found to be ∼63 °C

parametric plot constructed for microbubbling of BSA solution−air system as a function of the BSA flow rate and the air flow rate. A range of BSA solution flow rates from 30 μL/min to 800 μL/min and air flow rates from 500 μL/min to 6000 μL/min were selected to construct the map. There were four distinct regimes identified. Below a certain flow rate of air, the BSA solution was subjected to conventional electrohydrodynamic atomization on application of applied voltage. This region is marked as Zone 1 in Figure 3, the air has no influence on the atomization of BSA solution. For air flow rates above this critical minimum level, until a threshold value of BSA solution flow rate was reached, the microbubbling mode was not detected on application of applied voltage due to insufficient BSA solution flow rate. This region is noticeable in Figure 3 as Zone 2. Zone 3 represents the microbubbling region where all BSA solution is bubbled and the yield of the microbubbles was high (106 bubbles/min). In Zone 4 (that is, large BSA flow rate and air flow rate) near monodisperse microbubbling could not be obtained. At these flow rates the microbubbles lost their near-monodispersity. Figure 4 shows the microbubbles produced using various concentrations of BSA solution. For 5 wt % BSA solution for an

Figure 2. DSC curve indicating the denaturation temperature of BSA solution.

at a heating rate of 5 °C/min. A heating rate of >5 °C/min did not produce a meaningful result due to rapid melting and evaporation of BSA. The DSC curve shows a single peak and a two-state transition, native and denatured. There is no intermediate state observed. This also indicates that any significant population of other small globular additives are not present in the BSA used. The modes of atomization are an important characteristic of electrohydrodynamic processing: spraying23 or spinning.24 However, the jetting modes of electrohydrodynamic microbubbling differs from the above processes having distinctive features: bubble dripping, coning, and microbubbling, caused by presence of air instead of liquid flow through the inner needle.25 Generally, the bubble dripping mode is characterized where air and BSA solution are allowed to flow simultaneously without application of high voltage. The bubbles generated in this mode are very coarse and polydisperse. Once the voltage has been applied, the concentric air−BSA solution coflow leads to the coning mode. Further increasing the applied voltage leads to the microbubbling mode where near monodisperse and finer microbubbles were produced. Optimizing the processing parameters is crucial in the determination of the size and size distribution of the microbubbles. This could be done by either adjusting the solution physical properties26 or varying the infusion rate of the fluid media and the applied voltage.27 Figure 3 shows a

Figure 4. Optical micrographs of microbubbles at (a) air flow rate 1000 μL/min, BSA flow rate 100 μL/min; (b) air flow rate 2000 μL/ min, BSA flow rate 200 μL/min; (c) air flow rate 4000 μL/min, BSA flow rate 400 μL/min; (d) air flow rate 6000 μL/min, BSA flow rate 400 μL/min. The applied voltage is 14−18 kV in all these.

air flow rate 1000 μL/min and a BSA flow rate 100 μL/min the bubbles are coarser and the air in these quickly diffused out and/or the bubbles burst immediately on collection (Figure 4a). Figure 4b shows the microbubbles produced at an air flow rate of 2000 μL/min and a BSA flow rate of 200 μL/min. For this case the microbubble sizes were bimodal. At an air flow rate of 4000 μL/min and a BSA flow rate of 400 μL/min, the produced microbubbles showed near-monodispersity. The polydispersity index of the microbubbles is calculated to be ∼13%, which is significantly lower than the sonication method.28 The stability of the microbubbles could be increased by adding a surfactant such as polyoxyethylene or Tween 80 to the shell material.29 However, it has been shown that the stability of microbubbles could be increased by simply adjusting the flow ratio of BSA solution and air.29 In this way, the shell

Figure 3. Parametric plot for bubbling of the BSA solutions. 6696


Langmuir

Article

10 wt % BSA solution, respectively (Figure 6c). This might be due to unfolding of the BSA in the bubble shell and therefore the microbubbles tend to grow in the presence of excess BSA solution. The effect of collection distance on stability of the microbubbles is shown in Figure 6d. A gradual reduction in microbubble size is observed for all three collection distances investigated in this work. However, there is a greater stability observed for the collection distance of ∼10 cm. The collection distance plays a significant role in microbubble formation, like in other electrically driven processes such as electrospinning and electrospraying. The collection distance has an effect on the electric field and the subsequent instabilities during processing. This ultimately affects the shape and size of the formed structues.23 As can be seen in Figure 6d, a collection distance ∼10 cm gives the best results compared to shorter and longer collection distance. This may be associated with the number of the microbubbles/m2 and/or evaporation of the solvent during processing. Whereas the applied voltage can deliver a systematic reduction in bubble size, the process is governed by the electric field, surface tension, and gravitational force which give rise to various jetting stabilities and associated modes that could lead to unwanted regions such as dripping in the parametric processing map.34 The number of microbubbles in a milliliter of storage solution is gradually reduced from ∼150 to ∼40 for 5 wt % BSA, ∼150 to ∼60 for 10 wt % BSA, and ∼140 to ∼60 to 15 wt % BSA (Figure 6e). Moreover, the number of microbubbles reduced from ∼160 to ∼120 in distilled water−glutaraldehyde collection medium, ∼150 to 120 in glycerol−Tween 80 system, and ∼150 to ∼80 for glycerol system (Figure 6f). These are significantly higher than storing the microbubbles in BSA solution itself. This remarkable stability is attributed to interfacial elasticity and interfacial viscosity in various collection media and is discussed below. The stability of the microbubbles could be explained by the elasticity of the shell material. The elastic interface can be formed by proteins that adsorb at the interface and form a viscous elastic gel like shell. There will always be some viscous behavior due to intra- or intermolecular rearrangements. The interface becomes more elastic if the proteins form extra molecular interactions.31 In addition, the increase in shell thickness increases the elasticity of the microbubble interface and will stop the shrinkage due to disproportionation. Second, intrabubble interaction gives a greater stability to microbubbles over a period of time. In the food and beverages industry, glycerol serves as a humectant, solvent, and sweetener and help to preserve foods. The glycerol backbone is central to all lipids known as triglycerides. Tween 80 is a well-known surfactant and the stabilization of the microbubbles could be obtained for a longer period by suppressing the Laplace surface tension force. Moreover, interfacial viscosity will prevent bubble shrinkage and promote the stabilization. It has been shown that, when there is low interfacial viscosity there is no clear effect of interfacial viscosity on the dissolution. However, the bubbles’ shrinkage could be delayed. The higher interfacial viscosity significantly reduces the bubble dissolution and improves the stability.30 Figure 7a shows the stability of rehydrated microbubbles, which were dried and stored at ambient temperature, over a period of time. The large drop in the curve is related to the time the bubbles have been left before rehydration. It could also be related to the elasticity of the BSA shell. The dried microbubbles are rehydrated and an air bubble with a shell is

thickness can be increased providing a larger shell elasticity that prevents the rapid diffusion of air.30−32 Figure 4d shows the microbubbles produced at an air flow rate of 6000 μL/min and a BSA flow rate of 400 μL/min. The shell thickness of the BSA is found to be ∼500 nm and the BSA molecules per bubble is 0.456 × 10−9 g corresponding to 6.9 × 10−15 mol. Generally, heating a solution leads to reduction in viscosity. However, it has been shown that the viscosity increases with temperature for BSA solution due to change in the protein structure, and this has an effect on the stability of the microbubbles.29,33 Figure 5a shows the bubble size variation with the BSA solution flow rate for various concentrations of BSA solution. A

Figure 5. (a) Bubble size variation with the flow rate of BSA solution for various concentrations of BSA.

gradual increase in bubble size was observed with increase in flow rate (Table 1). Generally, an increase in flow rate results in an increase in the size of the microbubbles. Table 1. Mean Bubble Size for Various Concentrations of BSA Solution BSA solution

flow rate (μL/min)

mean bubble size (μm)

polydispersity index %

15 wt % 15 wt % 10 wt % 10 wt % 5 wt % 5 wt %

150 400 150 400 150 400

72 180 42 157 39 155

22 6 28 13 28 13

These results show that the flow rate of the BSA solutions has a significant effect on the mean bubble size and the polydispersity. The polydispersity of the microbubbles improved with increase in concentration and the flow rate of BSA solution. The long-term stability of the microbubbles was studied in various collection media. Figure 6a shows a gradual reduction bubble size is observed over a period of time in distilled water− glutaraldehyde collection medium. The bubble size is reduced from ∼150 μm to ∼100 and 50 μm for 10 and 5 wt % of BSA solution, respectively. The bubble size is reduced from ∼200 μm to ∼100 μm for 15 wt % of BSA solution. Similarly, the bubble size is reduced from ∼200 μm to ∼100 μm for 15 wt % of BSA solution in glycerol and glycerol−Tween 80 collection media (Figure 6b). However, the bubble size is increased for 5, 10, and 15 wt % BSA solutions in BSA solution collection medium. The bubble size increased from ∼200 μm to ∼800 μm for 15 wt % BSA solution and ∼150 μm to ∼500 μm for 5 and 6697


Langmuir

Article

Figure 6. Bubble size variation over a period of time in (a) distilled water−glutaraldehyde collection medium; (b) glycerol, glycerol−Tween 80 system and 27 wt % BSA solution; (c) various concentrations of BSA solutions. (d) Effect of working distance on the stability of the microbubbles. Number of microbubbles collected over a period of time in (e) various concentrations of BSA solutions; (f) glycerol, glycerol−Tween 80 system; and water−glutaraldehyde collection media.

reduced from ∼200 to 60 μm in this case. The corresponding optical micrograph of the rehydrated microbubbles is shown in Figure 7b. This indicates that the microbubbles could be rehydrated at any time and used for specific purposes without

formed. This bubble may disproportionate until the elasticity in the shell is large enough to overcome the Laplace pressure and stop disproportionation. The microbubble size remains constant after a specific time interval. The bubble size is 6698


Langmuir

Article

Figure 7. continued

6699


Langmuir

Article

Figure 7. (a) Stability of rehydrated microbubbles. (b) Optical micrograph of the rehydrated microbubbles. (c) Effect of flow rate on bubble size for AIBA-PS particle-reinforced microbubbles for two different concentrations of particles. (d) Stability of the microbubbles for two different concentrations of particles. (e) Optical micrograph of the particle-reinforced microbubbles. (f) FTIR spectrum of the particle-reinforced microbubbles. (g) UV/visible spectra for BSA and particle-reinforced microbubbles.

microbubble acoustic response by a ’jamming’ effect.35 It is very similar to the well-known phenomenon of Pickering stabilization in liquid−liquid emulsions36 and has also been observed in foams.37 Thus, surface active particles with high adsorption energy can generate a sufficiently rigid shell to prevent microbubble shrinkage due to disproportionation. Therefore, the formation of particle-reinforced microbubbles depends on a delicate balance between the tendency of the hydrophobic particles to adsorb at microbubble surfaces and their tendency to aggregate rather than disperse in water. Moreover, a microbubble needs to shrink appreciably to achieve stable disproportionation, and this suggests that during shrinkage there may be some rearrangement of the particles adsorbed on the microbubble surface and/or more further adsorption of particles to the microbubbles, resulting in a more close-packed particle layer necessary for long-term stability.37 There is no appreciable change in bubble size for 0.15 and 1.5 wt % AIBA−PS particles after a certain time interval. This indicates that there was no marked effect caused by the concentration of the particles in BSA solution. This may be attributed to the AIBA−PS particles reaching their packing density on the bubble surface, thereby minimizing the surface area available for gas diffusion and significantly reducing the effective interfacial tension due to a decrease in the surface to volume ratio.38 In addition, over a period of time rates of gas diffusion and Ostwald ripening become negligible and the AIBA−PS particles are assumed to reach their packing density and thus significantly inhibit further mass transfer.39 A typical load−displacement curve obtained during the nanoindentation of different BSA microbubbles (five bubbles were measured for each condition with sizes ranging from 150 to 200 μm) is shown in Figure 8. The compression stiffness of microbubbles was derived from the gradient of the straight line fitted to the upper portion of the curves.40 The stiffness was found to be 8 ± 0.5 N/m, 12 ± 1 N/m, and 20 ± 2 N/m for 5, 10, and 20 wt % BSA microbubbles, respectively, at ambient temperature. It shows a two orders of magnitude increase in stiffness compared with phospholipid microbubbles21 and an order of magnitude higher than the protein vesicle.41 However, the values are remarkably similar to the stiffness of polyelectrolyte microspheres.19 The difference in stiffness of

any significant change in their size and the size distribution. Moreover, the microbubbles could be dried at any temperature and/or even collection temperature could be varied for a specific usage. Figure 7c shows the bubble size variation with the flow rate of the 15 wt % BSA solution containing particles. It is clearly seen that bubble size increases with the flow rate of the solutions. The bubble size shows a gradual increase from ∼100 μm to ∼200 μm for a flow rate of