- Email: [email protected]
Exfoliation of montmorillonite in protein solutions
Journal of Colloid and Interface Science 374 (2012) 135–140
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Exfoliation of montmorillonite in protein solutions Krzysztof Kolman a,b, Werner Steffen a, Gabriela Bugla-Płoskon´ska c, Aleksandra Skwara c, Jacek Pigłowski b, Hans-Jürgen Butt a, Adam Kiersnowski a,b,⇑ a
Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany Polymer Technology and Engineering Division, Wroclaw University of Technology, Wybrzeze Wyspianskiego 27, 50-370 Wroclaw, Poland c Institute of Genetics and Microbiology, University of Wroclaw, Przybyszewskiego 63/77, 51-148 Wroclaw, Poland b
a r t i c l e
i n f o
Article history: Received 2 December 2011 Accepted 6 February 2012 Available online 18 February 2012 Keywords: Montmorillonite Exfoliation Nanoparticles Proteins Adsorption TEM DLS Bionanocomposites
a b s t r a c t In the study we demonstrate a method to obtain stable, exfoliated montmorillonite–protein complexes by adsorption of the proteins extracted from hen-egg albumen. Analysis of the process by means of sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) revealed that the complexes are formed by sequential adsorption of ovotransferrin, ovalbumins, ovomucoid and lysozyme on the surface of the silicate. Structural studies performed by X-ray diffraction (XRD) and transmission electron microscopy (TEM) indicated that the adsorption of ovotransferrin and albumins is accompanied by disintegration of clay stacks into discrete platelets. Further analysis by dynamic light scattering (DLS) revealed that at protein to silicate weight ratios exceeding 20, the synergistic adsorption of albumen components leads to reaggregation of silicate platelets into disordered, microgel-like particles. By means of DLS it was found that exfoliation predominantly leads to formation of particles with average hydrodynamic radii (Rh) of 0.19 lm while their aggregation causes formation of particles having Rh in of approx. 0.5 lm and larger. Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction Exfoliation of montmorillonite is crucial for the successful formation of clay-based nanocomposites. Dispersion of clay platelets in a polymer matrix at the nanoscale allows obtaining materials with improved mechanical performance, decreased permeability or better thermal stability [1,2]. The usual way to obtain exfoliated nanocomposites involves chemical modification of the clay and either chemical synthesis [3] or relatively harsh processing of the material, which are often inappropriate for e.g. biopolymers or drug delivery systems. This is because the chemical modification and synthesis often involves toxic alkylammonium compounds [4,5], whereas biopolymers are usually susceptible to denaturation and decomposition at elevated temperatures and/or pressures. Thus simple alternatives to exfoliate montmorillonite in mild conditions may contribute to development of materials such as scaffolds for tissue engineering [6], hybrid materials for drug delivery [7,8] or, more generally, functional bionanocomposites [9–11]. The influence of adsorption of ionic compounds on dispersion and aggregation of clay particles in aqueous media was extensively reported by Lagaly and coworkers [12–14]. It is also known that
⇑ Corresponding author at: Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany. E-mail addresses: [email protected], adam.kiersnowski@pwr. wroc.pl (A. Kiersnowski). 0021-9797/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2012.02.009
water-soluble, polar polymers, like polyvinylpyrrolidone or ionic liquids can stabilize dispersions of exfoliated particles of clay minerals [15,16]. Stabilization of clay particles in dispersions by proteins or polysaccharides seems however the most relevant for the application in bionanocomposites and biofunctional materials. [6–11,17] This approach not only allows obtaining stable dispersions of nanoparticles but also confers them with e.g. bacteriostatic properties [18]. At certain conditions accumulation of protein molecules on the particle clusters may trigger separation of their aggregates into smaller moieties. Some proteins are known to separate and stabilize dispersions of individual carbon nanotubes in aqueous solutions [19,20]. In the case of layered compounds adsorption of single proteins usually leads to formation of stable intercalated complexes [7–9,21,22]. In contrast, however, to nanotube bundles, layered silicates reveal also an electrostatically stabilized secondary structure of piled, approx. 10-nm thick, stacks of platelets [23–26]. Although separation or aggregation of bigger clusters is important for the formation of nanocomposites, the influence of protein molecules on general aggregation behavior of layered clays was not reported so far. Results of X-ray diffraction measurements revealed that the disintegration of the primary layered structure of the silicate (the exfoliation) occurs only upon the adsorption from the complex solutions as e.g. soy protein isolate or serum [17,18]. The exfoliation under the influence of soy-proteins may result from electrostatic interactions as well as hydrogen bonding between protein macromolecules and the silicate [17].
136
K. Kolman et al. / Journal of Colloid and Interface Science 374 (2012) 135–140
In the case of serum/montmorillonite complexes it was hypothesized that disintegration of the layered structure of the clay is due to synergistic, sequential adsorption of several proteins [18]. The studies on adsorption of lysozyme and albumin on the silica surface suggest that interactions between the proteins and solid surfaces are purely electrostatic [27]. The analysis of competitive adsorption of albumin and immuno-c-globulins (IgG) on a charged polystyrene surface leads to similar conclusions [28]. In the latter case, the electrostatic interactions were reported to enhance the adsorption of IgG on the surface pre-coated with albumin. In the same report however the initial stage of adsorption of IgG on the bare surface was found independent of pH, which indicated that the adsorption cannot be explained exclusively on the basis of electrostatic interactions [28]. Additionally, the co-adsorption of proteins from mixtures of e.g. b-lactoglobulin and lactoferrin boosts the accumulation of both the proteins at the surface. This synergistic adsorption was found irreversible, which suggests different complexation mechanisms between the surface and the molecules at different conditions [29]. It is thus clear that although electrostatic interactions play a major role in adsorption and formation of intermolecular complexes, the mechanisms depend also on specific and non-specific interactions in the system as well as the composition of the protein solution. In comparison to the macroscopic solid surfaces, the adsorption of proteins on layered silicates has not been studied in much detail although these processes reveal similarities. In both the cases electrostatic interactions and synergism in the adsorption from complex systems were pointed out as main driving forces [17,18]. Lack of the clear connection between the charge of protein molecules and their ability to intercalate [7,18,21] may indicate that the adsorption on layered silicates might also be site-specific with regard to protein molecules [29]. This could explain why some proteins reveal certain preferred orientation at the surface and the adsorption triggers structural changes in their molecules [30–32]. In this work we report the sequence of protein adsorption as well as the influence of the adsorption on exfoliation and aggregation of natural sodium montmorillonite in bulk and in the solution. In comparison to our previous study [18] we use here a commonly available protein mixture extracted from the hen egg albumen instead of normal human serum. The mixture contained mainly ovalbumin, ovotransferrin, ovomucin, ovomucoid and lysozyme [33]. It seems that due to bactericidal properties of lysozyme [34,35], the adsorption of this protein on montmorillonite may be a simple way to obtain hybrid antimicrobial complexes. Moreover, the positive charge of lysozyme over a broad range of pH, should not only help binding other hen-egg proteins but also is expected to have a stabilizing effect on the adsorbed proteins [29,33].
(14.4 kg/mol, pI = 9.36), ovomucoid (25 kg/mol, pI = 4.75), ovalbumins (40–45 kg/mol, pI = 5.19) and ovotransferrin (78 kg/ mol, pI = 6.85). This protein mixture is abbreviated as EPx (Egg Protein extracted). 2.2. The adsorption The buffer solution of the EPx was pumped at a constant rate (5 mL/min) through a 70 mg MMT bed using the apparatus schematically shown in Fig. 1. The flow rate was controlled by a peristaltic pump. The time necessary to stabilize the flow in the whole solution channel was reached after approx. 15 s. The protein solution samples (approx. 0.5 mL) were collected through a samples port at 20, 40, 60, 300, 600, 900 s and then analyzed by electrophoresis. Given that MMT was found already exfoliated after 20 s, the exfoliation process was studied in a different manner. In this case the dilute protein solutions containing amount of proteins corresponding to the 1, 2, 3, 5, 10, 20 s of the flow in previous experiments were pumped through the MMT bed. The montmorillonite–protein complexes (MMT-EPx) were then dried in vacuum at 30 °C. 2.3. Sodium dodecyl sulfate – polyacrylamide gel electrophoresis (SDS–PAGE) Protein quantification in the analyzed solutions was performed using BCA Protein Assay Kit (PierceÒ) containing bicinchoninic acid (BCA) and copper(II) sulfate as indicators for the colorimetric estimation of protein concentration and internal standard (bovine serum albumine). The detection range was 20–20000 lg/mL [36]. In our study, the average concentration of proteins in analyzed solutions after diluting by 100 was found to be 1000 ± 90 lg/mL. Sodium dodecyl sulfate discontinuous gel electrophoresis (SDS– PAGE) was performed according to the method established by Laemmli [37] using 10% separating gel and 4% stacking gel. All samples were heated at 100 °C for 4 min and 10 lg of proteins was applied onto each slab. For the calibration of molecular weight the Bio-Rad (161-0317) broad range molecular weight standard (6.5–200 kg/ mol) was used. The electrophoresis was carried out at 25 mA for 165 min, using glycine electrophoresis buffer. The visualization of protein bands was made using the Coomassie Brilliant Blue staining method. The electrophoregrams were analyzed with the Quantity OneÒ 1-D Analysis Software and ImageJ 1.44p. 2.4. Dynamic light scattering (DLS) Due to the high sensitivity of DLS to solid impurities, prior to the measurements both the clay and EPx solution were additionally
2. Experimental section 2.1. Materials The sodium montmorillonite (NanofilÒ116, MMT) with a cation exchange capacity of 1.16 meq/g was a gift of of Southern Clay Products/German Division in Moosburg. The protein substrate was obtained from 48 hen (Gallus gallus domesticus, untreated with antibiotics) egg whites. The substrate, after drying at 30 °C and pulverizing, was stored in antiseptic conditions at 20 °C. Directly prior to the experiments, 13.1 g of the protein substrate was dissolved in 100 mL of Sørensen’s phosphate buffer solution (pH = 7.7), vigorously stirred at ambient temperature for 1 h and filtered through a dense sintered-glass filter to remove any solid residues. The electrophoretic analysis of the solution, revealed presence of four main proteins: lysozyme
Fig. 1. Schematic illustration of the setup used for adsorption experiments. Arrows (?) indicate the flow direction.
K. Kolman et al. / Journal of Colloid and Interface Science 374 (2012) 135–140
137
purified. MMT was dispersed in 18.2 MO Milli-Q water (10.0 g/L) and sedimented for 24 h at room temperature. After the sedimentation, the supernatant containing 3.12 g/L of MMT (estimated by centrifugation) was separated from the residue and diluted to obtain a stock dispersion with a concentration of 0.1 g/L. The EPx solution was centrifuged for 600 s at 5000 rpm and filtered through cellulose acetate membranes with a pore size of 0.45 lm (Millipore). Mixtures of 0.5 mL MMT stock dispersion and 0.5 mL of EPx diluted to 0.1, 0.2, 0.5, 1.0, 2.0, 5.0 and 10 g/L) were placed in dust free cuvettes and thermally equilibrated in a toluene bath held at a constant temperature of 20 °C. DLS measurements were performed on a custom setup consisting of ALV-6010/160 digital full correlator (ALV), Nd-YAG laser (coherent, k = 532 nm) and two avalanche photo detectors (Perkin-Elmer) working in pseudo-cross-correlation mode. Data were collected in the range from 30° to 130° in 10° steps. The relaxation function g1(q, t) was calculated from the experimental, normalized autocorrelation function g2(q, t) by applying the Siegert relation. To analyze the g1(q, t) function the CONTIN inverse Laplace fitting routine as well as a Kohlrausch–Williams–Watts (KWW) fitting function were applied [38,39]. It is worthwhile to stress that in our case both the analyses resulted in apparent hydrodynamic radii (Rh) since both theories are based on the diffusion of a sphere in a homogeneous medium [40,41]. Nevertheless the change in Rh and its distribution function delivers a measure for aggregation. 2.5. Powder X-ray diffraction (XRD) and transmission electron microscopy (TEM) XRD measurements were performed on Ultima IV Bragg–Brentano X-ray diffractometer fitted with 2 kW X-ray tube (40 kV/30 mA CuKa, k = 0.1542 nm). Diffractograms were recorded for powdered MMT and MMT-EP complexes within the 2h range from 1.1° to 10.0° in 0.05° steps. For TEM measurements powdered MMT and MMT-EP samples were applied on semi-cured, viscous epoxy-resin slabs (EpoFix Kit, Struers), sealed with epoxy, finally cured at room temperature for 24 h and microtomed into 60–100 nm slices. The slices were placed on standard copper grids. The micrographs were recorded using Zeiss EM900 microscope operated at 80 kV.
Fig. 2. (A) The electrophoretic gels of EPx permeate after adsorption of proteins on MMT. White rectangle on the ‘300 s’ gel marks the integration area for the densitometric profiles shown in B. OVT, OVA, OVM and OVL stand for ovotransferrin, ovalbumin, ovomucoid and lysozyme respectively.
3. Results and discussion In order to trace the adsorption of a particular protein on the MMT surface we analyzed the variation of the EPx solution composition as a function of time by means of SDS–PAGE (Fig. 2A). More specifically, we analyzed changes in intensity of the bands corresponding to a particular protein to trace the variation in composition of the EPx upon the adsorption on MMT. The electrophoretic gel lanes (the electrophoregrams) were integrated along their axes and after correction of the baseline they were plotted as densitometric profiles (Fig. 2B). Numeric integration of the profiles allowed estimating the intensity of bands corresponding to a particular protein. Since the integral intensities obtained in this way do not reflect the protein concentrations directly, their values were divided by the intensities of respective bands in the reference gel (Fig. 2A). In this way we obtained relative fractions of proteins in the permeate ðX p Þ. The repetition of the SDS–PAGE for the reference sample indicated that the estimation error in this method ranges from 5% to 10%, depending on the contrast between the band and the background as well as geometrical distortion of bands [42]. The variation of X p vs. time (Fig. 3) allowed estimating which proteins are adsorbed at a given time as well as giving a clue about the adsorption rate. Values of X p around the unity indicate no adsorption (the same content in the EPx solution and in the permeate) while the lower X p values indicate the higher adsorption
Fig. 3. The adsorption of EPx components on MMT vs. time. X p (relative fractions of proteins in the permeate) denote the ratio of the particular proteins in the EPx and the permeate (lines serve as a visual guide).
rates. The values above one would indicate the elution of the protein from the surface. Thus, indirectly, the analysis of X p plots provides us with some qualitative information about the composition of the adsorbate. The analysis of graphs in Fig. 3 indicates that within the first minute the adsorption rate of transferrin and albumins is the highest. Adsorption of lysozyme is much slower while nearly no accumulation of ovomucoid on MMT is observed. After 40 s of the process, one can see an indistinct increase in the adsorption rate
138
K. Kolman et al. / Journal of Colloid and Interface Science 374 (2012) 135–140
of lysozyme, the rise in adsorption of ovomucoid and a decrease in the uptake of albumins. After 60 s, the adsorption of all except transferrin is nearly ceased while later after approx. 300 s of continuous flow of the EPx solution over MMT bed one can see that there is a dramatic increase in the adsorption rate of ovomucoid. Together with this protein, adsorption of lysozyme seems increasing as well. Finally after 900 s of adsorption, the process is ceased and no more protein uptake occurs (Fig. 3). Only lysozyme has a positive charge under the pH = 7.7 assumed in the adsorption experiment. Thus by analogy to ion exchange with alkylammonium compounds one could expect, that this protein should preferably adsorb on the MMT surface by exchanging sodium cations. The results of electrophoretic analysis indicate however that, at the beginning (0–20 s), the adsorbate contains mainly the ovalbumin (pI = 5.19) and the ovotransferrin (pI = 6.85) despite their negative charge. This indicates that the formation of the first layer of proteins on the silicate surface does not depend on simple electrostatic interactions between the proteins and the silicate. At this stage we may suppose that these proteins are bound to the surface either by specific interactions, like e.g. hydrogen bonding [28], or by interactions between ‘‘positive patches’’ on the macromolecules and the negatively charged silicate surface [17]. Disambiguation between these two concepts however would demand either spectroscopic studies or analysis of orientation of protein macromolecules in the deposited layers, which is beyond the scope of this work. It is also important that the fractions of the albumin and the transferrin removed from the EPx are similar (Fig. 3), which additionally indicates that, like in [29], the synergistic action may also be considered as the driving force behind the deposition of ovalbumin and ovotransferrin on the MMT surface. Most likely, further complexation of proteins on MMT surface relies on the electrostatic interactions between the adsorbed layer and the moieties in solution. This would explain the significant increase in ovomucoid (pI = 4.75) uptake after formation of the first layer and simultaneous increase in lysozyme adsorption. Further discussion of the adsorption solely on the basis of variation in adsorbate composition seems impossible. It is known, that in the case of mixtures, the proteins accumulated at the surface at early stages of adsorption may be replaced even without substantially affecting the adsorbed amount. This is due to competitive adsorption of proteins at solid surfaces. In our experiments for none of the proteins we recorded X p > 1, which would suggest that here the adsorption dominated over the removal of the EPx components from the surface. Nevertheless, since we are unable to distinguish between the not adsorbed and the eluted macromolecules, this is difficult to quantitatively assess the composition of MMT-EPx systems at the end of adsorption experiment or discuss the structure of the adsorbate. Generally, it seems that ovalbumin and ovotransferrin are present in the adsorbate after 900s of adsorption. Our results indicate that they form the very first protein layer of at the surface of the MMT. A minimum in the X p plot observed at 600 s of adsorption of ovomucoid indicates that this protein is also present in the adsorbate in the top layer. Similarly, the X p plot obtained for lysozyme (Fig. 3) suggests that after 60 s, and most likely also at the end of adsorption, this protein is also present in the MMT-EPx complex. In order to analyze the influence of protein adsorption on layered structure of MMT we recorded a series of diffraction patterns of MMT treated with amount of EPx corresponding to the defined periods of time (Fig. 4). Analysis of XRD results (Fig. 4) indicates that trace amounts of EPx does not affect the layered architecture of the silicate. Up to the 0.03 g/g the interlayer distance, calculated from the maximum at 7°, approximates 1.30 nm which is the same as the basal spacing in the pristine MMT. At the ratio of 0.06 g/g on the left side of the diffraction maximum a weak shoulder appears at 6.3° that corresponds to spacing of 1.40 nm. When the
Fig. 4. XRD patterns recorded for MMT after adsorption of EPx. Numbers indicate protein to silicate ratios. The values in round brackets indicate the time at which the given ratio would be reached at the constant flow of EPx in the experiment shown in Fig. 1 (cf. text for details).
ratio between components reaches 0.14 g/g, the diffraction profile dramatically changes. At this point the curve consists of at least three overlapping maxima: the two mentioned above and dominating one, located at 4.70° (1.9 nm). Further increase in the protein to silicate ratio causes a gradual shift of this maximum to smaller diffraction angles, which, according to Braggs’ law, means gradual increase in the interlayer distance from 3.1 nm to 3.8 nm observed at 0.6 g/g protein to silicate ratio. In the latter case however, the peak (or rather a shoulder) is broad and its integral intensity is very low, that implies that the ordering is poor and the mass fraction of the ordered material is small. Increasing the protein to silicate ratio during further adsorption does not affect the layered structure of the silicate in the system. Exfoliation occurs up to 0.6 g/g of EPx:MMT ratio. A comparison of this observation with results of SDS–PAGE analysis (Fig. 3) leads to the conclusion that the disintegration of MMT layered structure is mainly a result of adsorption of ovotransferrin and albumins. As mentioned previously both the proteins have a weak total negative charge, which suggests non-electrostatic nature of adsorption process [17,28]. Despite positive overall charge suggesting favorable electrostatic interactions with negatively charged silicate particles [43], lysozyme seems not contributing to the exfoliation. From Fig. 3 one can conclude that lysozyme fraction in the adsorbate is low. The comparison between results presented in Figs. 3 and 4 suggests that there is little or no correlation between the adsorption rate of lysozyme and the exfoliation. To additionally clarify the XRD results we analyzed the same samples using a TEM. Fig. 5A shows an edge view of a montmorillonite primary particle before adsorption of proteins. In Fig. 5D one can see the densitometric cross-section of the particle view along the A-A profile showing regular stratification with a characteristic average length of 1.3 nm – exactly the same value as determined by XRD. Fig. 5B and C show typical images of silicate particles after adsorption of 0.6 g/g of proteins. Their appearance indicates that although periodicity of montmorillonite platelets within particles is lost after the adsorption, a large number of primary aggregates retain their integral granular structure. The particles form a kind of microgels consisting of discrete silicate platelets ‘‘glued’’ with protein molecules. In some of these aggregates one can see groups of several stacked platelets separated by gaps bigger than the basal spacing in the bare silicate. Densitometric analysis (the linear cross-section, Fig. 5E) of 5C micrograph reveals presence of aggregates consisting of 2–4 platelets spaced by 1.2–1.4 nm separated by a gap of approx. 3.5–4.5 nm. This observation is also consistent with XRD results and additionally clarifies the occurrence of a
K. Kolman et al. / Journal of Colloid and Interface Science 374 (2012) 135–140
139
Fig. 5. Transmission electron micrographs of the untreated MMT (A) and MMT-EPx 0.6 g/g complex (B and C) plots D and E shows densitometric profiles of particle images measured along A–A and B–B lines respectively.
small diffraction maximum recorded by XRD at 2.3° for MMT-EPx 0.60 g/g complex. In order to study aggregation of the particles in dispersions, we performed DLS experiments (Fig. 6). Analysis of the autocorrelation functions using CONTIN algorithm (Fig. 6B) revealed that the dispersion of MMT displays a broad distribution of relaxation times and hence apparent hydrodynamic radii (Rh). The values between 0.05 and 1.75 lm suggest that the dispersion contains the discrete platelets, their stacks and also larger aggregates of the stacks [14,26]. The profiles of distribution functions of Rh shown in Fig. 6B indicate occurrence of two groups: the first, relatively monodisperse, having Rh of 0.19 lm and the second with significantly broader, asymmetric size distribution around approx. 0.5 lm. Taking into account dimensions of clay particles (cf. Fig. 5) and the fact that proteins cause them to exfoliate, one can assume that in both the cases we deal with relatively loose aggregates of platelets. At this stage we can only speculate that their sizes depend on whether the particles are aggregated through direct interactions of clay particles (the aggregates with Rh of 0.19 lm) or if they are formed by binding the particles with protein molecules (the aggregates with Rh of 0.5 lm). While dimensions of particles with Rh of 0.19 lm remain constant upon increasing EPx concentration in the solution, changing tail profiles of CONTIN fits (Fig. 6B) indicate variation in sizes of larger aggregates. Up to the EPx:MMT ratio of 20 g/g one can see a clear decrease in number of large aggregates having apparent hydrodynamic radii above 1.25 lm. This drop is accompanied by an increase in the number of particles with Rh of 0.19 lm. Unfortunately we cannot assign discrete sizes to peaks appearing in the distribution function for the aggregates since the CONTIN algorithm might decompose a very broad distribution into a series of more or less discrete peaks. At higher protein to silicate ratios aggregates with larger sizes reoccurred. To describe the disintegration and aggregation phenomena in a different manner we applied
KWW functions to fit autocorrelation curves (Fig. 6A). As a result of this analysis a function of mean sizes of aggregates vs. EPx:MMT ratio was obtained (Fig. 6C). The plot in Fig. 6C shows that at the initial stages of adsorption mean sizes of larger aggregates linearly decrease with increasing EPx:MMT ratio. The drop continues until Rh reaches its minimum at 0.445 lm at EPx:MMT ratio of 20 g/g. From this point up to approx. 50 g/g the particle sizes insignificantly grow, and beyond this point they remain constant suggesting that bigger aggregates no longer interact with molecules from EPx (Fig. 6B). This behavior suggests that increase in EPx:MMT ratio within 0–20 g/g induces a gain in the small aggregates concomitant to a loss in big clusters. Going beyond 20 g/g triggers most likely the reaccumulation of small particles to larger entities or alternatively swelling the latter by incorporation of protein molecules within existing aggregates (the pseudo-microgels observed by TEM).
4. Conclusion We describe a method to obtain stable exfoliated montmorillonite–protein hybrid complexes. The exfoliation of montmorillonite occurs at low protein to silicate ratios during the early stages of the adsorption process. Results of electrophoretic analysis clearly suggest that disintegration of MMT tactoids occurs due to adsorption of ovotransferrin and ovalbumins. The ultimate separation of clay particles is mainly a result of interaction between MMT platelets and all main components of albumen, namely ovotransferrin, ovalbumin, ovomucoid and lysozyme. In these structures, resembling pseudo-microgels, one can distinguish discrete clay platelets as well as their stacks consisting of 2–3 platelets. In aqueous dispersions, a secondary aggregation of clay particles into larger aggregates was evidenced by DLS. While the exfoliation of MMT occurs already when protein to MMT weight ratio is equal to
140
K. Kolman et al. / Journal of Colloid and Interface Science 374 (2012) 135–140
method can be applied also to stabilize dispersions of clay nanoparticles in aqueous solution. Acknowledgments K. Kolman thanks Yi Sun from Max Planck Institute for Polymer Research for help with obtaining TEM images. The work was supported by the International Max Planck Research School for Polymer Materials Science (IMPRS-PMS) at Max Planck Institute for Polymer Research in Mainz. A. Kiersnowski acknowledges the support from Marie Curie Intra European Fellowship (PIEF-GA-2009253521) granted within 7th EU Framework program. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31]
Fig. 6. Autocorrelation function recorded for EPx:MMT 20 g/g system (A). Distribution functions of MMT particles radii from the CONTIN fit (B) and variation in MMT particle sizes as a function of mass ratio between proteins and clay from the KWW fit (C). For the sake of clarity in Fig. B the profiles are shifted; zero intensity is marked by dashed lines.
0.6 g/g, formation of large aggregates demands high excess of protein over the silicate. DLS study revealed also that the structural effects of adsorption are relatively complex and rely on simultaneous disintegration of large clay particles and reaggregation of exfoliated platelets into larger entities. Interestingly, despite submicron dimensions, dispersions of these microgel-like particles were found to be very stable over time, which proves that the described
[32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43]
S. Ray, M. Okamoto, Prog. Polym. Sci. 40 (2003) 1539. L. Utracki, M. Sepher, E. Boccaleri, Polym. Adv. Technol. 18 (2007) 1. J. Di, D. Sogah, Macromolecules 39 (2006) 5052. T. Thorsteinsson, T. Loftsson, M. Masson, Curr. Med. Chem. 10 (2003) 1129. A. Abott, G. Capper, D. Davies, R. Rashed, V. Tambyrajah, Trans. Inst. Met. Finish 79 (2001) 204. J. Zheng, C. Wang, X. Wang, H. Wang, H. Zhuang, K. Yao, React. Funct. Polym. 67 (2007) 780. J.-J. Lin, J.-C. Wei, T.-Y. Juang, W.-C. Tsai, Langmuir 23 (2007) 1995. G. Joshi, B. Kevadiya, H. Mody, H. Bajaj, J. Polym. Sci. Pol. Chem. 50 (2012) 423. E. Ruiz-Hitzky, M. Darder, P. Aranda, J. Mater. Chem. 15 (2005) 3650. S. Pandey, S. Mishra, J. Colloid Interface Sci. 361 (2011) 509. T. Duncan, J. Colloid Interface Sci. 363 (2011) 1. G. Lagaly, Angew. Chem. Int. Ed. 15 (1976) 575. G. Lagaly, Solid State Ionics 22 (1986) 43. G. Lagaly, S. Ziesmer, Adv. Colloid Interface 100–102 (2003) 105. M. Zabska, K. Jaskiewicz, A. Kiersnowski, K. Szustakiewicz, S. Rathgeber, J. Piglowski, Radiat. Phys. Chem. 80 (2011) 1125. A. Letaief, J. Leclerq, Y. Liu, C. Detellier, Langmuir 27 (2011) 15248. P. Chen, L. Zhang, Biomacromolecules 7 (2006) 1700. A. Kiersnowski, M. Serwadczak, E. Kułaga, B. Futoma-Kołoch, R. Kwiatkowski, W. Doroszkiewicz, J. Pigłowski, Appl. Clay Sci. 44 (2009) 225. D. Nepal, K. Geckeler, Small 3 (2007) 1259. Y. Lee, K. Geckeler, Adv. Mater. 22 (2010) 4076. A. McLaren, G. Peterson, Nature 192 (1961) 960. A. De Cristofaro, A. Violante, Appl. Clay Sci. 44 (2001) 225. J. McAtee, F. Cheng, Am. Mineral. 52 (1967) 1386. J. McAtee, F. Cheng, Am. Mineral. 52 (1967) 1725. H. Van Olphen, Clay Colloid Chemistry, second ed., J. Wiley & Sons, New York, 1977. C. Johnston, Clay Miner. 45 (2010) 245. M. Wahlgren, M. Paulsson, T. Arnebrant, Colloids Surf., A 70 (1993) 139. A. Elgersma, L. Zsom, J. Lyklema, W. Norde, Colloids Surf. 65 (1992) 17. M. Wahlgren, T. Arnebrant, M. Paulsson, J. Colloid Interface 158 (1993) 46. M. Revault, H. Quiquampoix, M. Baron, S. Noinville, Biochim. Biophys. Acta 1724 (2005) 367. N. Hellasa, M. Revault, H. Quiquampoix, P. Dejardin, S. Staunton, S. Noinville, J. Colloid Interface Sci. 356 (2011) 718. K. Kubiak-Ossowska, P.A. Mulheran, Langmuir 26 (2010) 15954. L. Stevens, Comp. Biochem. Phys. B 100 (1991) 1. V.L. Hughey, P.A. Wilger, E.A. Johnson, Appl. Environ. Microbiol. 55 (1989) 631. G. Bugla-Płoskon´ska, A. Kiersnowski, B. Futoma-Kołoch, W. Doroszkiewicz, Microbiol. Ecol. 58 (2009) 276. P. Smith, R. Krohn, G. Hermanson, A. Mallia, F. Gartner, M. Provenzano, E. Fujimoto, N. Goeke, B. Olson, D. Klenk, Anal. Biochem. 150 (1985) 76. U. Laemmli, Nature 15 (1970) 680. S. Provencher, Comput. Phys. Commun. 27 (1982) 213. S. Provencher, Comput. Phys. Commun. 27 (1982) 229. W. Schärtl, Light scattering from polymer solutions and nanoparticle dispersions, Springer, Berlin, 2007. B. Berne, R. Pecora, Dynamic light scattering: with applications to chemistry, biology, and physics, Dover Publications Inc., New York, 2000. I. Bajla, I. Hollander, S. Fluch, K. Burg, M. Kollar, Comput. Meth. Prog. Bio. 77 (2005) 209. K. Ralla, U. Sohling, D. Riechers, K. Kasper, F. Ruf, T. Scheper, Bioproc. Biosyst. Eng. 33 (2010) 847.
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Exfoliation of montmorillonite in protein solutions Krzysztof Kolman a,b, Werner Steffen a, Gabriela Bugla-Płoskon´ska c, Aleksandra Skwara c, Jacek Pigłowski b, Hans-Jürgen Butt a, Adam Kiersnowski a,b,⇑ a
Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany Polymer Technology and Engineering Division, Wroclaw University of Technology, Wybrzeze Wyspianskiego 27, 50-370 Wroclaw, Poland c Institute of Genetics and Microbiology, University of Wroclaw, Przybyszewskiego 63/77, 51-148 Wroclaw, Poland b
a r t i c l e
i n f o
Article history: Received 2 December 2011 Accepted 6 February 2012 Available online 18 February 2012 Keywords: Montmorillonite Exfoliation Nanoparticles Proteins Adsorption TEM DLS Bionanocomposites
a b s t r a c t In the study we demonstrate a method to obtain stable, exfoliated montmorillonite–protein complexes by adsorption of the proteins extracted from hen-egg albumen. Analysis of the process by means of sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) revealed that the complexes are formed by sequential adsorption of ovotransferrin, ovalbumins, ovomucoid and lysozyme on the surface of the silicate. Structural studies performed by X-ray diffraction (XRD) and transmission electron microscopy (TEM) indicated that the adsorption of ovotransferrin and albumins is accompanied by disintegration of clay stacks into discrete platelets. Further analysis by dynamic light scattering (DLS) revealed that at protein to silicate weight ratios exceeding 20, the synergistic adsorption of albumen components leads to reaggregation of silicate platelets into disordered, microgel-like particles. By means of DLS it was found that exfoliation predominantly leads to formation of particles with average hydrodynamic radii (Rh) of 0.19 lm while their aggregation causes formation of particles having Rh in of approx. 0.5 lm and larger. Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction Exfoliation of montmorillonite is crucial for the successful formation of clay-based nanocomposites. Dispersion of clay platelets in a polymer matrix at the nanoscale allows obtaining materials with improved mechanical performance, decreased permeability or better thermal stability [1,2]. The usual way to obtain exfoliated nanocomposites involves chemical modification of the clay and either chemical synthesis [3] or relatively harsh processing of the material, which are often inappropriate for e.g. biopolymers or drug delivery systems. This is because the chemical modification and synthesis often involves toxic alkylammonium compounds [4,5], whereas biopolymers are usually susceptible to denaturation and decomposition at elevated temperatures and/or pressures. Thus simple alternatives to exfoliate montmorillonite in mild conditions may contribute to development of materials such as scaffolds for tissue engineering [6], hybrid materials for drug delivery [7,8] or, more generally, functional bionanocomposites [9–11]. The influence of adsorption of ionic compounds on dispersion and aggregation of clay particles in aqueous media was extensively reported by Lagaly and coworkers [12–14]. It is also known that
⇑ Corresponding author at: Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany. E-mail addresses: [email protected], adam.kiersnowski@pwr. wroc.pl (A. Kiersnowski). 0021-9797/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2012.02.009
water-soluble, polar polymers, like polyvinylpyrrolidone or ionic liquids can stabilize dispersions of exfoliated particles of clay minerals [15,16]. Stabilization of clay particles in dispersions by proteins or polysaccharides seems however the most relevant for the application in bionanocomposites and biofunctional materials. [6–11,17] This approach not only allows obtaining stable dispersions of nanoparticles but also confers them with e.g. bacteriostatic properties [18]. At certain conditions accumulation of protein molecules on the particle clusters may trigger separation of their aggregates into smaller moieties. Some proteins are known to separate and stabilize dispersions of individual carbon nanotubes in aqueous solutions [19,20]. In the case of layered compounds adsorption of single proteins usually leads to formation of stable intercalated complexes [7–9,21,22]. In contrast, however, to nanotube bundles, layered silicates reveal also an electrostatically stabilized secondary structure of piled, approx. 10-nm thick, stacks of platelets [23–26]. Although separation or aggregation of bigger clusters is important for the formation of nanocomposites, the influence of protein molecules on general aggregation behavior of layered clays was not reported so far. Results of X-ray diffraction measurements revealed that the disintegration of the primary layered structure of the silicate (the exfoliation) occurs only upon the adsorption from the complex solutions as e.g. soy protein isolate or serum [17,18]. The exfoliation under the influence of soy-proteins may result from electrostatic interactions as well as hydrogen bonding between protein macromolecules and the silicate [17].
136
K. Kolman et al. / Journal of Colloid and Interface Science 374 (2012) 135–140
In the case of serum/montmorillonite complexes it was hypothesized that disintegration of the layered structure of the clay is due to synergistic, sequential adsorption of several proteins [18]. The studies on adsorption of lysozyme and albumin on the silica surface suggest that interactions between the proteins and solid surfaces are purely electrostatic [27]. The analysis of competitive adsorption of albumin and immuno-c-globulins (IgG) on a charged polystyrene surface leads to similar conclusions [28]. In the latter case, the electrostatic interactions were reported to enhance the adsorption of IgG on the surface pre-coated with albumin. In the same report however the initial stage of adsorption of IgG on the bare surface was found independent of pH, which indicated that the adsorption cannot be explained exclusively on the basis of electrostatic interactions [28]. Additionally, the co-adsorption of proteins from mixtures of e.g. b-lactoglobulin and lactoferrin boosts the accumulation of both the proteins at the surface. This synergistic adsorption was found irreversible, which suggests different complexation mechanisms between the surface and the molecules at different conditions [29]. It is thus clear that although electrostatic interactions play a major role in adsorption and formation of intermolecular complexes, the mechanisms depend also on specific and non-specific interactions in the system as well as the composition of the protein solution. In comparison to the macroscopic solid surfaces, the adsorption of proteins on layered silicates has not been studied in much detail although these processes reveal similarities. In both the cases electrostatic interactions and synergism in the adsorption from complex systems were pointed out as main driving forces [17,18]. Lack of the clear connection between the charge of protein molecules and their ability to intercalate [7,18,21] may indicate that the adsorption on layered silicates might also be site-specific with regard to protein molecules [29]. This could explain why some proteins reveal certain preferred orientation at the surface and the adsorption triggers structural changes in their molecules [30–32]. In this work we report the sequence of protein adsorption as well as the influence of the adsorption on exfoliation and aggregation of natural sodium montmorillonite in bulk and in the solution. In comparison to our previous study [18] we use here a commonly available protein mixture extracted from the hen egg albumen instead of normal human serum. The mixture contained mainly ovalbumin, ovotransferrin, ovomucin, ovomucoid and lysozyme [33]. It seems that due to bactericidal properties of lysozyme [34,35], the adsorption of this protein on montmorillonite may be a simple way to obtain hybrid antimicrobial complexes. Moreover, the positive charge of lysozyme over a broad range of pH, should not only help binding other hen-egg proteins but also is expected to have a stabilizing effect on the adsorbed proteins [29,33].
(14.4 kg/mol, pI = 9.36), ovomucoid (25 kg/mol, pI = 4.75), ovalbumins (40–45 kg/mol, pI = 5.19) and ovotransferrin (78 kg/ mol, pI = 6.85). This protein mixture is abbreviated as EPx (Egg Protein extracted). 2.2. The adsorption The buffer solution of the EPx was pumped at a constant rate (5 mL/min) through a 70 mg MMT bed using the apparatus schematically shown in Fig. 1. The flow rate was controlled by a peristaltic pump. The time necessary to stabilize the flow in the whole solution channel was reached after approx. 15 s. The protein solution samples (approx. 0.5 mL) were collected through a samples port at 20, 40, 60, 300, 600, 900 s and then analyzed by electrophoresis. Given that MMT was found already exfoliated after 20 s, the exfoliation process was studied in a different manner. In this case the dilute protein solutions containing amount of proteins corresponding to the 1, 2, 3, 5, 10, 20 s of the flow in previous experiments were pumped through the MMT bed. The montmorillonite–protein complexes (MMT-EPx) were then dried in vacuum at 30 °C. 2.3. Sodium dodecyl sulfate – polyacrylamide gel electrophoresis (SDS–PAGE) Protein quantification in the analyzed solutions was performed using BCA Protein Assay Kit (PierceÒ) containing bicinchoninic acid (BCA) and copper(II) sulfate as indicators for the colorimetric estimation of protein concentration and internal standard (bovine serum albumine). The detection range was 20–20000 lg/mL [36]. In our study, the average concentration of proteins in analyzed solutions after diluting by 100 was found to be 1000 ± 90 lg/mL. Sodium dodecyl sulfate discontinuous gel electrophoresis (SDS– PAGE) was performed according to the method established by Laemmli [37] using 10% separating gel and 4% stacking gel. All samples were heated at 100 °C for 4 min and 10 lg of proteins was applied onto each slab. For the calibration of molecular weight the Bio-Rad (161-0317) broad range molecular weight standard (6.5–200 kg/ mol) was used. The electrophoresis was carried out at 25 mA for 165 min, using glycine electrophoresis buffer. The visualization of protein bands was made using the Coomassie Brilliant Blue staining method. The electrophoregrams were analyzed with the Quantity OneÒ 1-D Analysis Software and ImageJ 1.44p. 2.4. Dynamic light scattering (DLS) Due to the high sensitivity of DLS to solid impurities, prior to the measurements both the clay and EPx solution were additionally
2. Experimental section 2.1. Materials The sodium montmorillonite (NanofilÒ116, MMT) with a cation exchange capacity of 1.16 meq/g was a gift of of Southern Clay Products/German Division in Moosburg. The protein substrate was obtained from 48 hen (Gallus gallus domesticus, untreated with antibiotics) egg whites. The substrate, after drying at 30 °C and pulverizing, was stored in antiseptic conditions at 20 °C. Directly prior to the experiments, 13.1 g of the protein substrate was dissolved in 100 mL of Sørensen’s phosphate buffer solution (pH = 7.7), vigorously stirred at ambient temperature for 1 h and filtered through a dense sintered-glass filter to remove any solid residues. The electrophoretic analysis of the solution, revealed presence of four main proteins: lysozyme
Fig. 1. Schematic illustration of the setup used for adsorption experiments. Arrows (?) indicate the flow direction.
K. Kolman et al. / Journal of Colloid and Interface Science 374 (2012) 135–140
137
purified. MMT was dispersed in 18.2 MO Milli-Q water (10.0 g/L) and sedimented for 24 h at room temperature. After the sedimentation, the supernatant containing 3.12 g/L of MMT (estimated by centrifugation) was separated from the residue and diluted to obtain a stock dispersion with a concentration of 0.1 g/L. The EPx solution was centrifuged for 600 s at 5000 rpm and filtered through cellulose acetate membranes with a pore size of 0.45 lm (Millipore). Mixtures of 0.5 mL MMT stock dispersion and 0.5 mL of EPx diluted to 0.1, 0.2, 0.5, 1.0, 2.0, 5.0 and 10 g/L) were placed in dust free cuvettes and thermally equilibrated in a toluene bath held at a constant temperature of 20 °C. DLS measurements were performed on a custom setup consisting of ALV-6010/160 digital full correlator (ALV), Nd-YAG laser (coherent, k = 532 nm) and two avalanche photo detectors (Perkin-Elmer) working in pseudo-cross-correlation mode. Data were collected in the range from 30° to 130° in 10° steps. The relaxation function g1(q, t) was calculated from the experimental, normalized autocorrelation function g2(q, t) by applying the Siegert relation. To analyze the g1(q, t) function the CONTIN inverse Laplace fitting routine as well as a Kohlrausch–Williams–Watts (KWW) fitting function were applied [38,39]. It is worthwhile to stress that in our case both the analyses resulted in apparent hydrodynamic radii (Rh) since both theories are based on the diffusion of a sphere in a homogeneous medium [40,41]. Nevertheless the change in Rh and its distribution function delivers a measure for aggregation. 2.5. Powder X-ray diffraction (XRD) and transmission electron microscopy (TEM) XRD measurements were performed on Ultima IV Bragg–Brentano X-ray diffractometer fitted with 2 kW X-ray tube (40 kV/30 mA CuKa, k = 0.1542 nm). Diffractograms were recorded for powdered MMT and MMT-EP complexes within the 2h range from 1.1° to 10.0° in 0.05° steps. For TEM measurements powdered MMT and MMT-EP samples were applied on semi-cured, viscous epoxy-resin slabs (EpoFix Kit, Struers), sealed with epoxy, finally cured at room temperature for 24 h and microtomed into 60–100 nm slices. The slices were placed on standard copper grids. The micrographs were recorded using Zeiss EM900 microscope operated at 80 kV.
Fig. 2. (A) The electrophoretic gels of EPx permeate after adsorption of proteins on MMT. White rectangle on the ‘300 s’ gel marks the integration area for the densitometric profiles shown in B. OVT, OVA, OVM and OVL stand for ovotransferrin, ovalbumin, ovomucoid and lysozyme respectively.
3. Results and discussion In order to trace the adsorption of a particular protein on the MMT surface we analyzed the variation of the EPx solution composition as a function of time by means of SDS–PAGE (Fig. 2A). More specifically, we analyzed changes in intensity of the bands corresponding to a particular protein to trace the variation in composition of the EPx upon the adsorption on MMT. The electrophoretic gel lanes (the electrophoregrams) were integrated along their axes and after correction of the baseline they were plotted as densitometric profiles (Fig. 2B). Numeric integration of the profiles allowed estimating the intensity of bands corresponding to a particular protein. Since the integral intensities obtained in this way do not reflect the protein concentrations directly, their values were divided by the intensities of respective bands in the reference gel (Fig. 2A). In this way we obtained relative fractions of proteins in the permeate ðX p Þ. The repetition of the SDS–PAGE for the reference sample indicated that the estimation error in this method ranges from 5% to 10%, depending on the contrast between the band and the background as well as geometrical distortion of bands [42]. The variation of X p vs. time (Fig. 3) allowed estimating which proteins are adsorbed at a given time as well as giving a clue about the adsorption rate. Values of X p around the unity indicate no adsorption (the same content in the EPx solution and in the permeate) while the lower X p values indicate the higher adsorption
Fig. 3. The adsorption of EPx components on MMT vs. time. X p (relative fractions of proteins in the permeate) denote the ratio of the particular proteins in the EPx and the permeate (lines serve as a visual guide).
rates. The values above one would indicate the elution of the protein from the surface. Thus, indirectly, the analysis of X p plots provides us with some qualitative information about the composition of the adsorbate. The analysis of graphs in Fig. 3 indicates that within the first minute the adsorption rate of transferrin and albumins is the highest. Adsorption of lysozyme is much slower while nearly no accumulation of ovomucoid on MMT is observed. After 40 s of the process, one can see an indistinct increase in the adsorption rate
138
K. Kolman et al. / Journal of Colloid and Interface Science 374 (2012) 135–140
of lysozyme, the rise in adsorption of ovomucoid and a decrease in the uptake of albumins. After 60 s, the adsorption of all except transferrin is nearly ceased while later after approx. 300 s of continuous flow of the EPx solution over MMT bed one can see that there is a dramatic increase in the adsorption rate of ovomucoid. Together with this protein, adsorption of lysozyme seems increasing as well. Finally after 900 s of adsorption, the process is ceased and no more protein uptake occurs (Fig. 3). Only lysozyme has a positive charge under the pH = 7.7 assumed in the adsorption experiment. Thus by analogy to ion exchange with alkylammonium compounds one could expect, that this protein should preferably adsorb on the MMT surface by exchanging sodium cations. The results of electrophoretic analysis indicate however that, at the beginning (0–20 s), the adsorbate contains mainly the ovalbumin (pI = 5.19) and the ovotransferrin (pI = 6.85) despite their negative charge. This indicates that the formation of the first layer of proteins on the silicate surface does not depend on simple electrostatic interactions between the proteins and the silicate. At this stage we may suppose that these proteins are bound to the surface either by specific interactions, like e.g. hydrogen bonding [28], or by interactions between ‘‘positive patches’’ on the macromolecules and the negatively charged silicate surface [17]. Disambiguation between these two concepts however would demand either spectroscopic studies or analysis of orientation of protein macromolecules in the deposited layers, which is beyond the scope of this work. It is also important that the fractions of the albumin and the transferrin removed from the EPx are similar (Fig. 3), which additionally indicates that, like in [29], the synergistic action may also be considered as the driving force behind the deposition of ovalbumin and ovotransferrin on the MMT surface. Most likely, further complexation of proteins on MMT surface relies on the electrostatic interactions between the adsorbed layer and the moieties in solution. This would explain the significant increase in ovomucoid (pI = 4.75) uptake after formation of the first layer and simultaneous increase in lysozyme adsorption. Further discussion of the adsorption solely on the basis of variation in adsorbate composition seems impossible. It is known, that in the case of mixtures, the proteins accumulated at the surface at early stages of adsorption may be replaced even without substantially affecting the adsorbed amount. This is due to competitive adsorption of proteins at solid surfaces. In our experiments for none of the proteins we recorded X p > 1, which would suggest that here the adsorption dominated over the removal of the EPx components from the surface. Nevertheless, since we are unable to distinguish between the not adsorbed and the eluted macromolecules, this is difficult to quantitatively assess the composition of MMT-EPx systems at the end of adsorption experiment or discuss the structure of the adsorbate. Generally, it seems that ovalbumin and ovotransferrin are present in the adsorbate after 900s of adsorption. Our results indicate that they form the very first protein layer of at the surface of the MMT. A minimum in the X p plot observed at 600 s of adsorption of ovomucoid indicates that this protein is also present in the adsorbate in the top layer. Similarly, the X p plot obtained for lysozyme (Fig. 3) suggests that after 60 s, and most likely also at the end of adsorption, this protein is also present in the MMT-EPx complex. In order to analyze the influence of protein adsorption on layered structure of MMT we recorded a series of diffraction patterns of MMT treated with amount of EPx corresponding to the defined periods of time (Fig. 4). Analysis of XRD results (Fig. 4) indicates that trace amounts of EPx does not affect the layered architecture of the silicate. Up to the 0.03 g/g the interlayer distance, calculated from the maximum at 7°, approximates 1.30 nm which is the same as the basal spacing in the pristine MMT. At the ratio of 0.06 g/g on the left side of the diffraction maximum a weak shoulder appears at 6.3° that corresponds to spacing of 1.40 nm. When the
Fig. 4. XRD patterns recorded for MMT after adsorption of EPx. Numbers indicate protein to silicate ratios. The values in round brackets indicate the time at which the given ratio would be reached at the constant flow of EPx in the experiment shown in Fig. 1 (cf. text for details).
ratio between components reaches 0.14 g/g, the diffraction profile dramatically changes. At this point the curve consists of at least three overlapping maxima: the two mentioned above and dominating one, located at 4.70° (1.9 nm). Further increase in the protein to silicate ratio causes a gradual shift of this maximum to smaller diffraction angles, which, according to Braggs’ law, means gradual increase in the interlayer distance from 3.1 nm to 3.8 nm observed at 0.6 g/g protein to silicate ratio. In the latter case however, the peak (or rather a shoulder) is broad and its integral intensity is very low, that implies that the ordering is poor and the mass fraction of the ordered material is small. Increasing the protein to silicate ratio during further adsorption does not affect the layered structure of the silicate in the system. Exfoliation occurs up to 0.6 g/g of EPx:MMT ratio. A comparison of this observation with results of SDS–PAGE analysis (Fig. 3) leads to the conclusion that the disintegration of MMT layered structure is mainly a result of adsorption of ovotransferrin and albumins. As mentioned previously both the proteins have a weak total negative charge, which suggests non-electrostatic nature of adsorption process [17,28]. Despite positive overall charge suggesting favorable electrostatic interactions with negatively charged silicate particles [43], lysozyme seems not contributing to the exfoliation. From Fig. 3 one can conclude that lysozyme fraction in the adsorbate is low. The comparison between results presented in Figs. 3 and 4 suggests that there is little or no correlation between the adsorption rate of lysozyme and the exfoliation. To additionally clarify the XRD results we analyzed the same samples using a TEM. Fig. 5A shows an edge view of a montmorillonite primary particle before adsorption of proteins. In Fig. 5D one can see the densitometric cross-section of the particle view along the A-A profile showing regular stratification with a characteristic average length of 1.3 nm – exactly the same value as determined by XRD. Fig. 5B and C show typical images of silicate particles after adsorption of 0.6 g/g of proteins. Their appearance indicates that although periodicity of montmorillonite platelets within particles is lost after the adsorption, a large number of primary aggregates retain their integral granular structure. The particles form a kind of microgels consisting of discrete silicate platelets ‘‘glued’’ with protein molecules. In some of these aggregates one can see groups of several stacked platelets separated by gaps bigger than the basal spacing in the bare silicate. Densitometric analysis (the linear cross-section, Fig. 5E) of 5C micrograph reveals presence of aggregates consisting of 2–4 platelets spaced by 1.2–1.4 nm separated by a gap of approx. 3.5–4.5 nm. This observation is also consistent with XRD results and additionally clarifies the occurrence of a
K. Kolman et al. / Journal of Colloid and Interface Science 374 (2012) 135–140
139
Fig. 5. Transmission electron micrographs of the untreated MMT (A) and MMT-EPx 0.6 g/g complex (B and C) plots D and E shows densitometric profiles of particle images measured along A–A and B–B lines respectively.
small diffraction maximum recorded by XRD at 2.3° for MMT-EPx 0.60 g/g complex. In order to study aggregation of the particles in dispersions, we performed DLS experiments (Fig. 6). Analysis of the autocorrelation functions using CONTIN algorithm (Fig. 6B) revealed that the dispersion of MMT displays a broad distribution of relaxation times and hence apparent hydrodynamic radii (Rh). The values between 0.05 and 1.75 lm suggest that the dispersion contains the discrete platelets, their stacks and also larger aggregates of the stacks [14,26]. The profiles of distribution functions of Rh shown in Fig. 6B indicate occurrence of two groups: the first, relatively monodisperse, having Rh of 0.19 lm and the second with significantly broader, asymmetric size distribution around approx. 0.5 lm. Taking into account dimensions of clay particles (cf. Fig. 5) and the fact that proteins cause them to exfoliate, one can assume that in both the cases we deal with relatively loose aggregates of platelets. At this stage we can only speculate that their sizes depend on whether the particles are aggregated through direct interactions of clay particles (the aggregates with Rh of 0.19 lm) or if they are formed by binding the particles with protein molecules (the aggregates with Rh of 0.5 lm). While dimensions of particles with Rh of 0.19 lm remain constant upon increasing EPx concentration in the solution, changing tail profiles of CONTIN fits (Fig. 6B) indicate variation in sizes of larger aggregates. Up to the EPx:MMT ratio of 20 g/g one can see a clear decrease in number of large aggregates having apparent hydrodynamic radii above 1.25 lm. This drop is accompanied by an increase in the number of particles with Rh of 0.19 lm. Unfortunately we cannot assign discrete sizes to peaks appearing in the distribution function for the aggregates since the CONTIN algorithm might decompose a very broad distribution into a series of more or less discrete peaks. At higher protein to silicate ratios aggregates with larger sizes reoccurred. To describe the disintegration and aggregation phenomena in a different manner we applied
KWW functions to fit autocorrelation curves (Fig. 6A). As a result of this analysis a function of mean sizes of aggregates vs. EPx:MMT ratio was obtained (Fig. 6C). The plot in Fig. 6C shows that at the initial stages of adsorption mean sizes of larger aggregates linearly decrease with increasing EPx:MMT ratio. The drop continues until Rh reaches its minimum at 0.445 lm at EPx:MMT ratio of 20 g/g. From this point up to approx. 50 g/g the particle sizes insignificantly grow, and beyond this point they remain constant suggesting that bigger aggregates no longer interact with molecules from EPx (Fig. 6B). This behavior suggests that increase in EPx:MMT ratio within 0–20 g/g induces a gain in the small aggregates concomitant to a loss in big clusters. Going beyond 20 g/g triggers most likely the reaccumulation of small particles to larger entities or alternatively swelling the latter by incorporation of protein molecules within existing aggregates (the pseudo-microgels observed by TEM).
4. Conclusion We describe a method to obtain stable exfoliated montmorillonite–protein hybrid complexes. The exfoliation of montmorillonite occurs at low protein to silicate ratios during the early stages of the adsorption process. Results of electrophoretic analysis clearly suggest that disintegration of MMT tactoids occurs due to adsorption of ovotransferrin and ovalbumins. The ultimate separation of clay particles is mainly a result of interaction between MMT platelets and all main components of albumen, namely ovotransferrin, ovalbumin, ovomucoid and lysozyme. In these structures, resembling pseudo-microgels, one can distinguish discrete clay platelets as well as their stacks consisting of 2–3 platelets. In aqueous dispersions, a secondary aggregation of clay particles into larger aggregates was evidenced by DLS. While the exfoliation of MMT occurs already when protein to MMT weight ratio is equal to
140
K. Kolman et al. / Journal of Colloid and Interface Science 374 (2012) 135–140
method can be applied also to stabilize dispersions of clay nanoparticles in aqueous solution. Acknowledgments K. Kolman thanks Yi Sun from Max Planck Institute for Polymer Research for help with obtaining TEM images. The work was supported by the International Max Planck Research School for Polymer Materials Science (IMPRS-PMS) at Max Planck Institute for Polymer Research in Mainz. A. Kiersnowski acknowledges the support from Marie Curie Intra European Fellowship (PIEF-GA-2009253521) granted within 7th EU Framework program. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31]
Fig. 6. Autocorrelation function recorded for EPx:MMT 20 g/g system (A). Distribution functions of MMT particles radii from the CONTIN fit (B) and variation in MMT particle sizes as a function of mass ratio between proteins and clay from the KWW fit (C). For the sake of clarity in Fig. B the profiles are shifted; zero intensity is marked by dashed lines.
0.6 g/g, formation of large aggregates demands high excess of protein over the silicate. DLS study revealed also that the structural effects of adsorption are relatively complex and rely on simultaneous disintegration of large clay particles and reaggregation of exfoliated platelets into larger entities. Interestingly, despite submicron dimensions, dispersions of these microgel-like particles were found to be very stable over time, which proves that the described
[32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43]
S. Ray, M. Okamoto, Prog. Polym. Sci. 40 (2003) 1539. L. Utracki, M. Sepher, E. Boccaleri, Polym. Adv. Technol. 18 (2007) 1. J. Di, D. Sogah, Macromolecules 39 (2006) 5052. T. Thorsteinsson, T. Loftsson, M. Masson, Curr. Med. Chem. 10 (2003) 1129. A. Abott, G. Capper, D. Davies, R. Rashed, V. Tambyrajah, Trans. Inst. Met. Finish 79 (2001) 204. J. Zheng, C. Wang, X. Wang, H. Wang, H. Zhuang, K. Yao, React. Funct. Polym. 67 (2007) 780. J.-J. Lin, J.-C. Wei, T.-Y. Juang, W.-C. Tsai, Langmuir 23 (2007) 1995. G. Joshi, B. Kevadiya, H. Mody, H. Bajaj, J. Polym. Sci. Pol. Chem. 50 (2012) 423. E. Ruiz-Hitzky, M. Darder, P. Aranda, J. Mater. Chem. 15 (2005) 3650. S. Pandey, S. Mishra, J. Colloid Interface Sci. 361 (2011) 509. T. Duncan, J. Colloid Interface Sci. 363 (2011) 1. G. Lagaly, Angew. Chem. Int. Ed. 15 (1976) 575. G. Lagaly, Solid State Ionics 22 (1986) 43. G. Lagaly, S. Ziesmer, Adv. Colloid Interface 100–102 (2003) 105. M. Zabska, K. Jaskiewicz, A. Kiersnowski, K. Szustakiewicz, S. Rathgeber, J. Piglowski, Radiat. Phys. Chem. 80 (2011) 1125. A. Letaief, J. Leclerq, Y. Liu, C. Detellier, Langmuir 27 (2011) 15248. P. Chen, L. Zhang, Biomacromolecules 7 (2006) 1700. A. Kiersnowski, M. Serwadczak, E. Kułaga, B. Futoma-Kołoch, R. Kwiatkowski, W. Doroszkiewicz, J. Pigłowski, Appl. Clay Sci. 44 (2009) 225. D. Nepal, K. Geckeler, Small 3 (2007) 1259. Y. Lee, K. Geckeler, Adv. Mater. 22 (2010) 4076. A. McLaren, G. Peterson, Nature 192 (1961) 960. A. De Cristofaro, A. Violante, Appl. Clay Sci. 44 (2001) 225. J. McAtee, F. Cheng, Am. Mineral. 52 (1967) 1386. J. McAtee, F. Cheng, Am. Mineral. 52 (1967) 1725. H. Van Olphen, Clay Colloid Chemistry, second ed., J. Wiley & Sons, New York, 1977. C. Johnston, Clay Miner. 45 (2010) 245. M. Wahlgren, M. Paulsson, T. Arnebrant, Colloids Surf., A 70 (1993) 139. A. Elgersma, L. Zsom, J. Lyklema, W. Norde, Colloids Surf. 65 (1992) 17. M. Wahlgren, T. Arnebrant, M. Paulsson, J. Colloid Interface 158 (1993) 46. M. Revault, H. Quiquampoix, M. Baron, S. Noinville, Biochim. Biophys. Acta 1724 (2005) 367. N. Hellasa, M. Revault, H. Quiquampoix, P. Dejardin, S. Staunton, S. Noinville, J. Colloid Interface Sci. 356 (2011) 718. K. Kubiak-Ossowska, P.A. Mulheran, Langmuir 26 (2010) 15954. L. Stevens, Comp. Biochem. Phys. B 100 (1991) 1. V.L. Hughey, P.A. Wilger, E.A. Johnson, Appl. Environ. Microbiol. 55 (1989) 631. G. Bugla-Płoskon´ska, A. Kiersnowski, B. Futoma-Kołoch, W. Doroszkiewicz, Microbiol. Ecol. 58 (2009) 276. P. Smith, R. Krohn, G. Hermanson, A. Mallia, F. Gartner, M. Provenzano, E. Fujimoto, N. Goeke, B. Olson, D. Klenk, Anal. Biochem. 150 (1985) 76. U. Laemmli, Nature 15 (1970) 680. S. Provencher, Comput. Phys. Commun. 27 (1982) 213. S. Provencher, Comput. Phys. Commun. 27 (1982) 229. W. Schärtl, Light scattering from polymer solutions and nanoparticle dispersions, Springer, Berlin, 2007. B. Berne, R. Pecora, Dynamic light scattering: with applications to chemistry, biology, and physics, Dover Publications Inc., New York, 2000. I. Bajla, I. Hollander, S. Fluch, K. Burg, M. Kollar, Comput. Meth. Prog. Bio. 77 (2005) 209. K. Ralla, U. Sohling, D. Riechers, K. Kasper, F. Ruf, T. Scheper, Bioproc. Biosyst. Eng. 33 (2010) 847.