Investigations on the interactions between xanthan gum and poly(vinyl alcohol) in solid state and aqueous solutions

European Polymer Journal 84 (2016) 161–172

Contents lists available at ScienceDirect

European Polymer Journal journal homepage: www.elsevier.com/locate/europolj

Macromolecular Nanotechnology

Investigations on the interactions between xanthan gum and poly(vinyl alcohol) in solid state and aqueous solutions Cristina-Eliza Brunchi ⇑, Maria Bercea, Simona Morariu, Mihaela Avadanei ‘‘Petru Poni” Institute of Macromolecular Chemistry, 41-A Grigore Ghica Voda Alley, 700487 Iasi, Romania

a r t i c l e

i n f o

Article history: Received 25 March 2016 Received in revised form 19 August 2016 Accepted 6 September 2016 Available online 10 September 2016 Paper dedicated to the 150th anniversary of the Romanian Academy. Keywords: Xanthan gum Poly(vinyl alcohol) FTIR Viscosity Zeta potential Interpolymer interactions

a b s t r a c t In the present study, the polymer mixtures based on an anionic polysaccharide, namely xanthan gum (XG), and a neutral polymer, poly(vinyl alcohol) (PVA), were investigated by FTIR spectroscopy, viscometry and zeta potential measurements. The hydrogen bond intermolecular interactions between functional groups of xanthan gum and those of PVA were evidenced by FTIR spectroscopy. The experimental viscometric data of polymer mixtures in aqueous solutions were interpreted by using the Wolf approach. This model allows the determination of the intrinsic viscosity and other hydrodynamic parameters irrespective of structural characteristics of macromolecular chains, polymer mixture composition and polymer concentration. The chain conformation and polyelectrolyte feature of xanthan gum have a high contribution on the behavior of XG/PVA mixtures; high values of the hydrodynamic parameters were obtained for XG rich polymer mixtures. The viscometric results and FTIR analysis revealed the formation of intermolecular associations between XG and PVA due to the hydrogen bonds interactions. The absolute value of the zeta potential for XG/PVA mixtures decreases as the poly(vinyl alcohol) content in the polymer mixtures increases. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction The design of new materials with improved or tailored properties without altering the individual structure of polymers or development new synthesis represents a goal of the chemists. On the other hand, the increasing concern for environment made that the researchers attention to be focused on environmentally friendly technologies and materials. The natural polymers are biocompatible, biodegradable, bioadhesive, abundant in nature and cheap, but exceedingly poor cohesive strength of them makes inappropriate their use in certain applications. At the opposite pole are the synthetic polymers which, due to excellent combination between physical and mechanical properties, are widely used in the formulations of various systems, increasing their strength. However, they may contain some unreacted monomers which are usually toxic. Therefore, physical mixing of natural and synthetic polymers represents an attractive way to obtain polymeric materials with desired features, improved mechanical characteristics and biological performances. These materials have found applications in wastewaters treatment, food industry, medicine, etc. [1]. Xanthan gum (XG) is an anionic polysaccharide composed of (1,4)-b-D-glucose residues as main chain at which a trisaccharide side chain is linked at C3 position to every other glucose residues. The side chain consists of (1,4)-b-D-glucuronate unit between (1,2)-a-D-mannose and b-D-mannose units [2] (Scheme 1a). ⇑ Corresponding author. E-mail address: [email protected] (C.-E. Brunchi). http://dx.doi.org/10.1016/j.eurpolymj.2016.09.006 0014-3057/Ó 2016 Elsevier Ltd. All rights reserved.

162

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

Scheme 1. Chemical structure for (a) xanthan gum and (b) PVA.

XG is used as thickening, suspending and emulsifying agent for various kinds of water-based systems from food, pharmaceutical, cosmetic, agricultural, textile, ceramic and petroleum industries and in conjunction with other biocompatible polymers it can form stable gel systems [3,4] or it is involved in matrix formulations for oral controlled-release drugs [5–8]. Poly(vinyl alcohol) (PVA) (Scheme 1b) is a water soluble polymer synthesized by hydrolysis of poly(vinyl acetate) and commercially available with different molecular weights and hydrolysis degrees; the content of the residual acetyl groups strongly influences the reactivity of hydroxyl groups of PVA and its biodegradability [8,9]. Chain flexibility, biocompatibility, air permeability, excellent film-forming and water absorption properties, ability to form physical aggregates in aqueous solution by hydrogen bonds, etc., are some features required in different applications (such as, in formulation of biomedical and pharmaceutical devices [10–13], paper coating, packaging manufacturing or other materials [14]). XG/PVA system was used as basis in preparation of controlled drug delivery formulations [6–8], pH and electrolyte sensors [3] and some plastic materials [14] for which different cross-linking agents (i.e., epichlorohydrin, glutaraldehyde, poly(acrylic acid) or low dose electron beam) and synthesis conditions (temperature and time of reaction, emulsion or basic medium, a given ratio of XG/ PVA) were used. As far as we know, there are no reported data regarding the behavior of XG/PVA physical mixtures in dilute aqueous solutions. In this context, the current study aims to investigate XG/PVA mixtures and to discuss the interactions between XG and PVA by using FTIR spectroscopy, viscometry and zeta potential measurements. 2. Experimental 2.1. Materials Commercial samples of XG and PVA were used as received without further purification. XG in a powder form was provided by Sigma-Aldrich. Previous papers reported the viscometric molecular weight of xanthan gum used in the present study as being 1.165
106 g/mol [15] and the type of metallic ion from xanthan sample is potassium (K+) [16]. PVA with the degree of hydrolysis of 80% and molecular weight of 6.5
104 g/mol was purchased from Loba Feinchemie AG (Austria Chemical Companies). 2.2. Samples preparation Pure polymer stock solutions were prepared in Millipore water (obtained from a Milli-Q PF apparatus). Xanthan gum was dissolved at room temperature under gentle stirring. In order to prevent the degradation of polysaccharide chains, the stock solution was kept at temperature of 5 °C. PVA dissolution was carried out at 80–90 °C under magnetic stirring. After preparation, PVA solution was allowed to equilibrate overnight at room temperature. XG/PVA mixture solutions were obtained by combining the stock solutions in different ratios. The composition of polymer mixtures was expressed as weight fractions of PVA (wPVA). All investigated solutions were prepared with one day before measurements. 2.3. Measurements 2.3.1. Fourier transform infrared spectroscopy (FTIR) The infrared measurements were carried out on a Bruker Vertex 70 spectrometer (Bruker Optics, Germany). The spectra in solid state were recorded in transmission mode in the range of 4000–400 cm1 at a resolution of 2 cm1 by using the

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

163

conventional KBr disk method. The aqueous solution spectra were measured in the ATR geometry, using a single bounce Golden GateTM accessory. 128 scans were collected at a 4 cm1 resolution, for all the solution samples and for water as a reference spectrum. Each sample was analyzed in triplicate at 37 °C. The data processing was performed by means of the OPUS 6.5 software (Bruker Optics, Germany) on small spectral regions cut from the whole spectrum. The spectral regions of interest were baseline-corrected by using the interactive concave rubber band method. The multiple peak curve fitting procedure used mixed Gaussian-Lorentzian functions. The exact sub-bands position and their estimated number were determined using the second derivative spectrum and the Savitzky-Golay algorithm with a nine-point smoothing factor. In the curvefitting procedure, the peak position was maintained fixed and the intensity, shape and width were considered as variable parameters. 2.3.2. Viscometry The viscometric investigations for dilute solutions of polymer mixtures with different PVA contents (wPVA) were carried out at 37 °C with an automated viscosity measuring system (LAUDA LMV 830 Instrument, Germany) equipped with an Ubbelohde viscometer type I having the capillary diameter of 0.63 mm. The viscometer has enabled to obtain solutions with different concentrations directly inside, through sequential dilution of initial polymer solution with pure water as solvent. Prior the measurements, the pure polymer solutions and their mixtures were kept about 15 min for their equilibration at the working temperature. 2.3.3. Zeta potential Zeta potential (f) for 0.01 g/dL XG/PVA aqueous solution at 37 °C was determined by using electrophoretic mobility (l) and Henry and Smoluchowski approximation for f(k  a) = 1.5 as follows:

f ¼ g  l=e

ð1Þ

where g is the viscosity, e represents the dielectric constant of the medium, f(k  a) is the Henry’s function, k represents Debye-Huckel parameter and a is the particle radius. The electrophoretic mobility, l, was determined on Malvern Zetasizer ZS (Malvern Instruments, UK); the temperature of 37 °C was maintained with a Peltier device. For each composition of XG/PVA mixture, the measurement was made three times and the average value was taken into consideration. 3. Results and discussion Generally, when two polymer solutions are mixed, the diffusion of one flexible polymer chain (such as PVA) into a structured polymer (with rigid chains in double helix conformation, in our case xanthan gum) is accompanied by the destruction of the existing weak bonds, the formation of other new and finally, of the structural rearrangement of the entire macromolecular system in solution. FTIR spectroscopy is a technique widely used to study macromolecular structures and to detect the interactions between polymer chains. The abundance of hydroxyl groups in the chemical structure of XG and PVA leads to formation of the intermolecular interactions in their physical mixtures based on hydrogen bonding between the hydroxyl and carbonyl/carboxylate groups of XG and the hydroxyl and acetate groups of PVA (Scheme 2). FTIR studies in the solid state were employed to identify the sites of hydrogen bonding in the absence of water as additional source of interaction. The spectral patterns of XG and PVA alone are well known [17,18], while the FTIR spectra of their mixtures show distinctive features which confirm the interactions of the two polymers. The interactions occuring between the two polymers of XG/PVA mixtures are visible by the marked spectral changes especially in the region of m(OH) and m(C–OH) vibrations. In solid state, the narrow profile of the absorption band assigned to hydroxyl groups, m(OH), and the distinctive appearance of the m(C–OH) absorption are common for all mixtures of XG and PVA. The maximum of the OH stretching band of the mixtures is located between that corresponding to PVA (3326 cm1) and XG (3435 cm1), but there is no regular trend of its evolution as the PVA content increases (Fig. 1a). For wPVA lower than 0.5, the value of m(OH) is around 3400 cm1. Higher PVA content (wPVA of 0.6 and 0.7) shifted m(OH) to 3426 cm1, which is approaching that of XG. The diversity of the chain segments at which the OH groups are linked in XG produces a very broad m(OH) band, with two visible shoulders around 3570 and 3220 cm1. The shoulder at higher wavenumbers can be assigned to mono hydrogen-bonded OH groups of XG and it no longer appears in the mixtures spectra. The shoulder that is observed around 3270 cm1 in the mixtures spectra can be rather connected with the contribution of PVA hydroxyl groups that are intermolecularly associated and it overlaps the sub-band of the double H-bonded OH groups from XG. The characteristic band of pyranose sugar which refers to the signals from C-O-C stretching, C–OH stretching and bending, is peaking around 1125 cm1 in pure XG. Even for the lowest PVA concentration, the band profile in mixture is unique and is centered on 1065 cm1. This position is characteristic to the D-mannopyranose ring [19,20], being given by a bending vibration of the C1–H bond, with contributions from the stretching C–O vibration of the C4–OH and C1–OH groups of the 1 D-glucopyrannose ring [20,21]. The m(C–OH) vibrations in the H-bonded C3–OH bonds are found around 1090 cm [22].

164

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

Scheme 2. Schematic representation of interactions occurring for XG/PVA mixture in (a) solid state and (b) aqueous solution. The molecules of water present as traces in solid state are also involved in hydrogen bonding.

For the highest PVA concentration, the band maximum is shifted to 1090 cm1 and the profile resembles very closely that of the m(C–O) of PVA. The order – sensitive band around 1142 cm1 is no longer seen in the spectra of all mixtures, and the second derivative spectra have confirmed its major decrease due to PVA interaction with XG. The second derivative spectra (Fig. S1) have revealed that the different profile of the m(C–OH) band in mixtures is given by a simultaneous increase of three adjacent bands, positioned at 1093, 1071 and 1052 cm1, the last two being very weak in

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

165

Fig. 1. FTIR spectra for XG, PVA and XG/PVA mixtures with different weight fractions of PVA in solid state (a) and in aqueous solution, as scale – expanded ATR – FTIR spectra in the stretching C=O region (b) and in the C–OH stretching region (c). The spectra of PVA and the mixtures are offset in absorbance.

the spectrum of pure XG. By addition of PVA, the sub-band at 1052 cm1 becomes substantially strong at the expense of that at 1071 cm1, while the band at 1093 cm1 keeps its intensity. The first two sub-bands seem to be specific to C2–OH and C3–OH hydroxyl groups engaged in hydrogen bonds [22], in particular those belonging to the mannose residue, with a contribution from the m(C–O) from PVA. The central peak at 1125 cm1 in pure XG is resolved in two sub-maxima, at 1118 cm1 and 1133 cm1. In the spectra of the mixtures, the first sub-band completely disappeared, even at the lowest PVA concentration. This happened simultaneously with the development of a new sub-band around 1128 cm1, although PVA does not have any significant absorption in this area. This spectral range is characteristic to m(C4–OH) in H-bonded C4–OH groups and also to m(C1–O–C40 ) of the glycosidic linkage in XG. Meanwhile, the component around 1021 cm1 is more intense than in pure XG, but does not change its intensity regardless of PVA content. This suggests that some intermolecular bonds in the mixtures appear between the OH group of the primary alcohol C6–OH of XG and the alcohol groups of PVA. Therefore, a complex network of hydrogen bonds connects the carboxylate and hydroxyl groups of XG with acetate and hydroxyl groups of PVA (Scheme 2a). The stretching vibration of the acetate carbonyl (maximum at 1743 cm1) from both PVA and XG decreased in intensity, while the combined band of the pyruvate mas(COO) (1670 cm1) and glucuronic mas(COO) (1614 cm1) increased. The second derivative spectra (Fig. S2) reveal a diversity of carboxylate groups from pyruvate and glucuronic moieties. The broad sub-band of the carboxylate absorption in glucuronic groups suggests their involvement in inter- or intramolecular interactions with different strengths. This band has been resolved into four components, located at 1599, 1605, 1620 and 1630 cm1. The sub-band at 1630 cm1 has been found to be correlated to the residual water in the sample and has been removed from the analysis. The lowest frequency (1599 cm1) is very close to the value observed in aqueous solution (1592 cm1) (Fig. 1b), and it could correspond to those C=O groups involved in two hydrogen bonds. The possible proton donors could be some residual water molecules and OH groups [22,23]. The highest frequency (1620 cm1) is given by the free glucuronate groups. The wavelength of 1605 cm1 could correspond to glucuronic carbonyls accepting one

166

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

hydrogen bond, most probably from the surrounding secondary OH groups. These assignments suggest that a great fraction of glucuronic COO groups in the dried XG are bonded by strong inter- or intramolecular hydrogen bonds. The corresponding m(COO) is observed in second derivative spectra at 1405 cm1 in solid state and 1402 cm1 in aqueous solution. The m(C=O) absorption of pyruvate COO groups has been tentatively resolved in four components, placed at 1640, 1657, 1672 and 1684 cm1 (Fig. S2). The sub-bands at 1657 and 1672 cm1 are the most intense and can be attributed to pyruvate C=O groups that are mono H-bonded and non H-bonded, respectively. The weaker sub-band at 1640 cm1 may belong to carboxylate groups acting as double – proton acceptor groups and probably would carry the signal from residual water bending vibration. The relatively high affinity of XG to water and the permanent presence of several water molecules around the hydrophilic groups, even in dried state [22,23], constitute a reasonable justification for those carboxylate bands identified as being double H-bonded. The relative proportions of the free pyruvate and glucuronic COO groups in solid state of the mixtures were extracted from curve fitting and were determined with Eq. (2) [24]:

f free ¼

ODfree ODfree þ 1:5
ODbond

ð2Þ

where ffree is fraction of free groups and OD represents the optical density of the absorption bands corresponding to free and bonded COO groups. For a low content of PVA in the polymer mixture (up to wPVA = 0.5), both pyruvate and glucuronate COO groups are still involved in intra- or intermolecular bonds within XG units (Fig. 2). When the XG concentration in mixture becomes low enough and its conformation is a partially unfolded helix, a number of previously bonded glucuronic carboxylates turns into free ones. These free groups will associate with the hydroxyls from PVA as the PVA concentration increases. The variation of the free COO groups from the pyruvate units is less significant. For a moderate PVA concentration there is a significant intermolecular association between the pyruvic COO groups and PVA, which is followed by a slight increase of the free pyruvic groups at higher PVA amounts. The band with peaks at 1275 cm1 (mas(C–O–O) from pyruvate residue) and 1248 cm1 (mas(C–O–O) from acetate group) is observed in mixtures as a single band centered on 1256 cm1. When the XG amount increased, the pyruvate mas(C–O–O) redshifted from 1273 cm1 in the mixture with wPVA = 0.9–1266 cm1 in the mixture with wPVA = 0.5, while the acetate mas(C– O–O) was stable. This fact confirms that an important part from the mixtures stabilization is based on the intermolecular associations via hydrogen bonding of the carboxylate groups of pyruvate. The FTIR spectroscopy of the XG/PVA mixtures in aqueous solution (1 g/dL) evidenced some supplementary associations that are obviously related with the partially disordered conformation of XG in water. The spectral variations of the m(C=O) bands of the mixtures in aqueous solution as compared to solid state indicate associations with water of both carboxylate groups from the XG component and of the residual acetate units of PVA. The mas(COO) peak of glucuronic carboxylate, with the maximum at 1592 cm1 in the XG aqueous solution, is gradually displacing towards higher frequencies by addition of PVA. The peak corresponding to glucuronic COO groups was found at 1597 cm1, 1616 cm1 and 1606 cm1 for the mixtures with wPVA of 0.1, 0.5 and 0.9, respectively. The spectra for XG aqueous solution and the mixture with low PVA content (wPVA = 0.1) show a relatively intense shoulder at 1633 cm1 that develops into the more intense peak around 1650 cm1 for higher PVA content. This peak corresponds to C=O groups from pyruvate carboxylates which interact with the OH groups of PVA and their intermolecular interactions cannot be broken by water molecules. From the above observations it appears that the carboxylate groups from the glucuronate

Fig. 2. Variation of free carboxylate groups as a function of wPVA in XG/PVA mixtures (solid state).

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

167

units were extensively bonded with water, while those from the pyruvate units reacted in a smaller extent (Scheme 2b). This would be valid if a greater number of glucuronate units are free in solid state as compared to those of pyruvate units, which makes them available for interactions with water molecules. One intriguing aspect is the remarkable similarity of the region of m(C–OH) vibration for XG aqueous solution and the mixtures with different content of PVA in solid state. By comparing the two types of spectra (Fig. 1a and c), the same band profile with maximum around 1065 cm1 is found in both cases, which suggests that the XG hydration and the XG – PVA intermolecular interactions would be based on the same OH group association. In the XG aqueous solution, this band can be assigned to solvated C2–OH groups [22,25], which would suggest that the glucuronate units and the terminal mannose rings form a strong H-bonded complex with water molecules. The whole m(C–OH) band in aqueous solution maintained its profile and maximum position independent of wPVA. The spectra of the mixtures can be discussed in terms of the available hydroxyl groups for association with water, due to the extended chain conformation of XG and the fraction of OH groups already linked to PVA. Secondly, several absorption bands are characteristic only to aqueous solutions and are given by the hydroxyl – water complexes. The OH groups that are more prone to associate with water would be C3–OH and C2–OH and they are the most abundant in the XG structure. The primary C6–OH groups from the backbone and the C4–OH units from the internal D-mannopyranose ring are also available for H-bonding with water. The peaks connected to m(C–OH) of each of these groups are found in the solutions spectra in the same position as in solid state, only the band profile is narrower. Bellow 1000 cm1 the libration modes of water are found and they are very informative regarding the water organization with and around hydrophilic units of the mixtures. The peak around 868 cm1 is evident in the spectrum of XG aqueous solution and is very weak and displaced to 872 cm1 in the XG/PVA mixtures spectra. This mode is specific to those water molecules in monohydrate complexes that are acting as bridges between two OH groups [25]. As the chains of XG and PVA are already linked by H-bonds, there is a sterical hindrance to the free OH groups that can be connected by water bridges. This fact would explain the very low intensity of the 872 cm1 band in the mixtures spectra. The bands at 712, 750 and 780 cm1 in the XG aqueous solution spectrum and at 706 and 755 cm1 in the XG/PVA mixtures spectra describe the solvation water molecules bonded to one OH group. Beside the hydrogen bond type interaction mentioned above, the electrostatic interactions between the COO groups of xanthan gum have to be taken into consideration. In the presence of a common solvent, all these interactions can influence the chains conformation, compatibility between the two polymers as well as association phenomena. Viscometry represents an easy way to assess the interactions that occur in dilute polymer solutions and allows the determination of intrinsic viscosity ([g]) and other polymers chains characteristics in given thermodynamic conditions. In the determination of intrinsic viscosity, the extrapolation methods of the reduced viscosity to zero polymer concentration can be applied only to dilute solutions of uncharged polymers in a proper solvent [26] or charged macromolecules in solvents containing sufficient amount of salt [15]. In the case of polyelectrolytes solution containing an insufficient quantity of salt for complete charges screen or in absence of salts, the Debye length of polymer increases as the polymer concentration decreases and so, the intramolecular repulsive interactions between charged groups along the main chain increase. Consequently, an expansion of the polymer chain takes place, the reduced viscosity continuously increases with dilution (Fig. 3a) and the determination of intrinsic viscosity by using the Huggins method is not possible. In the present paper, we applied the new theoretical approach developed by Wolf [27] to a binary mixture of polyelectrolyte/neutral polymer in aqueous solution in order to determine the intrinsic viscosity. Wolf approach is able to describe the viscometric behavior of charged polymers in the absence/presence of salts on the whole range of concentration and hence, it allows an accurate determination of the intrinsic viscosity. Previously, this model was successfully applied to uncharged synthetic (co)polymers [28–34] and natural polymers [15,35]. The new empirical Wolf model was also applied for polyelectrolyte mixtures [36], polyelectrolyte/neutral polymer mixtures [37,38] and its validity was also demonstrated for neutral polymer mixtures [32,39]. According to Wolf approach, the intrinsic viscosity can be determined from the initial slope of the dependence between natural logarithm of relative viscosity (ln gr) and polymer concentration (c) at constant temperature, pressure and shear rate according to the following equation:

ln gr ¼

c½g þ Bc2 ½g½g 1 þ Bc½g

ð3Þ

where [g] is the intrinsic viscosity, B represents the viscometric interaction parameter with the same physical meaning as the Huggins constant and [g]⁄ is the characteristic specific hydrodynamic volume. Fig. 3b shows the evaluation of the experimental data for all studied XG/PVA mixtures in terms of the relative viscosity dependence on polymer concentration according to Eq. (3) and the determined hydrodynamic parameters are presented in Table 1. The analysis of the experimental data revealed the following aspects: (i) the new approach proposed by Wolf was appropriate to describe the viscometric behavior for XG/PVA solutions irrespective of the polymer mixture composition; (ii) a high value of the intrinsic viscosity was obtained for XG solution as compared with that of PVA (the difference is of two orders of magnitude, whereas the molecular weight of XG is 20 times higher than of PVA); (iii) the viscosity of XG/PVA mixtures has

168

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

wPVA

60

0 0.1 0.29 0.7 0.84 0.9 0.95

50

ηsp/c (dL/g)

40 30 10

5 PVA 0 0.00

0.01

0.02

0.05

0.10

0.15

0.20

c (g/dL)

(a) wPVA 0 0.1 0.29 0.51 0.59 0.7 0.84 0.9 0.95

0.6

ln ηr

0.4

0.2

PVA 0.0 0.00

0.02

0.04

0.06

0.08

0.10

c (g/dL)

(b) Fig. 3. Dependence of (a) reduced viscosity and (b) natural logarithm of the relative viscosity as a function of the polymer concentration for XG/PVA mixtures in aqueous solutions for different wPVA values, at 37 °C; the dashed lines represent the dependences obtained for PVA solutions.

Table 1 Values of the hydrodynamic parameters determined according to Eq. (3) for XG/PVA mixtures in aqueous solutions. wPVA

[g] (dL/g)

B

½g (dL/g)

0 0.10 0.29 0.51 0.59 0.67 0.70 0.84 0.90 0.95 1

43.02 46.59 34.91 22.43 17.30 8.26 8.60 4.72 2.20 1.77 0.44

1.12 3.72 2.43 2.18 2.01 0.48 0.15 0.35 0.44 0.02 0.008

15.28 23.11 16.52 10.18 10.30 0 0 0 0 0 0

evolved in a nonpredictible manner on polymer composition, expressing the interpolymer interactions between XG and PVA in aqueous solution. [g]⁄ is an adjustable parameter only required for the polyelectrolytes in salt free aqueous solutions, incorporating the effect of electrostatic interactions on hydrodynamic volume of charged macromolecules at finite concentration [27]. For neutral polymers or by adding an enough amount of extra-salt, its value is zero and two parameters suffice for modelling the viscometric behavior [15,40–42]. The addition of a small PVA amount (wPVA = 0.1) induces a change of the XG conformation from helix into partial disordered form and the electrostatic interactions between XG molecules are favored (the number of free charged groups is higher, [g]⁄ increases). At the same time, the interactions between XG and PVA chains become stronger leading to the

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

169

increase of [g]. The electrostatic interactions between charged groups, reflected by the viscometric parameters ([g]⁄ and B as well as [g],) have shown a maximum for XG/PVA with wPVA = 0.1 (Table 1). For polymer mixtures with low amount of XG, the PVA contribution is dominant and [g]⁄ = 0 (Table 1). In this case, a simplified version of Eq. (3) is applied, as was shown previously in the case of pullulan and dextran mixtures [39] or oxidized pullulan [41] in aqueous solutions. On the other hand, the hydrodynamic interaction parameter B from Eq. (3) can quantify the interactions that occur between polymer segments belonging to different macromolecules as well as those established between polymer and solvent. For almost all uncharged polymers, B has positive values, indicating the existence for polymer–solvent interactions whereas, for charged polymers the value of B reflects a cumulative effect of Coulombic (BCoulomb) and non-Coulombic (Bnon-Coulomb) interactions [15,27,40,42,43]:

B ¼ BCoulomb þ Bnon-Coulomb

ð4Þ

For XG/PVA aqueous solutions, the positive values for B were obtained irrespective of polymer composition (Table 1). Taking into account that the term BCoulomb is positive and Bnon-Coulomb is negative [42,44], we suppose that significant contribution of Coulombic interactions were counterbalanced by non-Coulombic ones as the PVA content in polymer mixtures increases and for PVA rich solutions B becomes very close to zero. Concomitantly, the polymer–water interactions can affect the values of B parameter. Because the Krigbaum-Wall formalism [45] cannot be applied to XG/PVA aqueous solutions due to the impossibility to determine the viscometric interaction parameter between chains, the deviation from the ideal values ([g]id) of the intrinsic viscosity ([g]) has been proposed as a criterion for the evaluation of the compatibility or association phenomena occuring in a solution of polymer mixtures [46–48]. [g]id value represents the viscosity of the polymer mixture in the absence of specific interactions between the two polymers in solution and can be obtained by additive rule:

½gid ¼ wPVA ½gPVA þ wXG ½gXG

ð5Þ

There are some other criteria concerning the compatibility in mixtures of polyanions [49] or neutral and charged polymers in aqueous solutions [50]. The degree of compatibility can be discussed by comparing the predicted and experimental values of the interaction parameter between unlike chains. The intrinsic viscosity and viscometric interaction parameters were determined by different methodologies: either by considering the mixtures of the two polymers in water as solvent or by ‘‘the polymer solvent method”, the solvent being considered the solution of the second polymer. The variation of the [g] values as a function of weight fraction of PVA in the polymer mixture was plotted in Fig. 4. [g] presents positive deviations from the additivity for XG/PVA mixtures with weight fraction of PVA up to 0.5 and negative deviations for higher values of wPVA. These deviations indicate the formation of interpolymer associations with different structuring [46–48]. The stability of secondary structure of xanthan (consisting of double-helix strands - the predominant conformation at equilibrium) is influenced by temperature and also by the presence of PVA chains. Above 36 °C, XG chains exhibit conformational changes from a rigid state to a disordered state [16], favoring the intermolecular interactions between partially unfolded XG helix and PVA chains through hydrogen bonds (as it was also evidenced by FTIR spectroscopy of XG/PVA mixture in solution, Fig. 1b and c). The addition of a small amount of PVA (wPVA = 0.1) determines the increase of [g] due to the interactions established between unlike chains (PVA and XG) considering that there is no penetration of PVA and XG chains (the disordered state is not completely established for XG). For wPVA > 0.5, there are also interactions between XG and PVA chains but the number of heterocontacts (XG-PVA) is lower. The random PVA coils form the majority in solution and they can penetrate each other due to strong affinity between OH groups. This could explain the decrease in viscosity observed for high PVA content. The synergistic effects of the interactions between unlike and like macromolecules are influenced by the conformation adopted by XG at 37 °C. The xanthan conformation depends on many factors, such as the sample characteristics (molecular weight, pyruvate and acetate contents, the patterns of attachment of pyruvate or acetate to the structural units in xanthan), sample thermal history (conditions of preparation, storage and experimental investigations) and salt addition. Native xanthan presents a single helical chain conformation which is lost by denaturation (heating or salt addition) and by renaturation (fast cooling of the hot solution to avoid chain degradation) a locally double helical structure is reformed [51]. Milas et al. have shown that the most available samples on the market were previously submitted to a thermal treatment by the producers, so, in many cases, the xanthan samples present a renaturated conformation reformed at room temperature or bellow [52]. AFM investigations [53] have evidenced the double helix conformation of xanthan macromolecules at equilibrium due to either intramolecular (anti-parallel) or intermolecular (anti-parallel or parallel) association. Recently, temperature and pH dependences of zeta potential and viscosity for XG aqueous solutions were reported and conformational changes were observed above 36 °C [16]. By increasing the temperature, a conformational transition occurs from a rigid state (characterized by a high viscosity) to a more flexible conformation in a disordered state (which presents a lower viscosity). At 37 °C, disordered form coexists with helix conformation, order-disorder transition being an equilibrium process [54]. There are many discussions concerning the intermolecular interactions of xanthan with other polysaccharides showing that the conformation adopted by XG chains plays a very important role [54–56].

170

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

Fig. 4. Dependence of [g] as a function of wPVA for XG/PVA mixture solutions at 37 °C. The dashed line represents the additive rule.

Milas and coworkers [51,56] suggested a more pronounced interaction of the galactomannan chains with the disordered xanthan segments. Even at room temperature, low amounts of disordered conformations able to interact with other macromolecules exist [55]. On the other side, Morris and Foster [57] considered that not only disordered xanthan sequences give intermolecular interactions, for the xanthan based mixtures the thermodynamic stability of heterocontacts is in competition with the 5-fold helix conformation of xanthan. Ordered xanthan macromolecules can undergo a conformation change into a geometry which permits an efficient binding of xanthan with a co-synergist, as for example galactomannan chains. The disordered form of XG is favored by the pyruvate residue, situated at the periphery of the helix structure which together with carboxylate group (from glucuronic acid residue), is engaged in achieving of intermolecular interactions [58]. At it turn, the acetate residue, placed closer to the center of the helix structure of XG macromolecule, tends to stabilize the ordered form due to intramolecular interactions established between its methyl groups and adjacent hemiacetal oxigen atoms of the b-D-glucose residue; the alternate hydroxil group at C3 is also involved into intramolecular interactions by bind to adjacent hemiacetal oxigen atom of the b-D-glucose residue [59]. In the disordered form, the side chains of XG are free to rotate [6], and their mobility favors the formation of interpolymer associations between XG and PVA. Electrophoretic measurements allow the evaluation of the charged species mobility that move relative to the medium in an electric field through zeta potential (f). This parameter provides information about the changes in the surface charges as a function of PVA content in the XG/PVA solution (Fig. 5). The highest negative value of f was recorded for XG solution (57.2 mV) and the lowest for PVA (9.71 mV). As the PVA content in polymer mixture increased, the dynamic of xanthan side-chains was affected and so, the f values ranged from 56.4 mV (wPVA = 0.1) to 17.5 mV (wPVA = 0.9). The negative value obtained for f in the case of PVA solution was due to the residual acetate groups contained in its structure; it is well known that the hydrolysis reaction of poly(vinyl acetate) in the production process of PVA is never 100% [60,61]. We assume that the evolution of zeta potential of XG/PVA aqueous solutions, is not related only by decreasing of XG content in the polymer mixtures. A shift of the shear surface (where the solution and the particle move in opposite directions when an external field is applied) or screening of carboxylate group (situated at the periphery of the helix structure) can occur as a result of the intermolecular interactions between XG and PVA. Similar observations were reported on suspensions of poly(ethylene oxide), Pluronics and TiO2 [30,62] or PVA and silica [61].

XG

PVA

0 0.0

0.2

0.4

wPVA

0.6

0.8

1.0

ζ (mV)

-15

-30

-45

-60 Fig. 5. Variation of zeta potential as a function of PVA weight fraction for XG/PVA polymer mixtures solutions at 37 °C.

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

171

When the XG content in XG/PVA mixture prevails, the charged particles repel each other and no agglomeration takes place, and f presents negative values lower than < 30 mV. For wPVA > 0.6, the electrostatic repulsions between particles become insufficient to prevent the agglomeration and, as a consequence, f varies from 30 mV to 9.71 mV. Therefore, the interactions between PVA and the side-chains of xanthan gradually increase and the structure of intermolecular associations evolves from a slight expanded structure to more compact one. 4. Conclusion The aim of the present paper was to investigate the interactions that occur in polymer mixtures of water soluble polymers with different thermodynamic features: xanthan, a polyelectrolyte, and poly(vinyl alcohol), a neutral polymer. The conclusions of this study can be summarized as follows:  The analysis of FTIR spectra obtained in solid state, for all polymer mixtures, revealed that the carboxylate groups from xanthan interact preferentially with OH groups from poly(vinyl alcohol) through hydrogen bonding interactions.  The FTIR spectra of aqueous solutions evidenced the preservation of the associations between the carboxylate groups from the pyruvic moieties and the hydrophilic units from PVA even in the presence of water; at the level of the side chain, a greater number of intermolecular associations with water molecules has been made by the glucuronate units.  The interpolymer interactions occuring in solution of XG/PVA mixtures were evidenced by viscometry.  On the basis of the approach proposed by Wolf, we evaluated the viscometric parameters which are able to describe the conformation of polymer chains on the whole range of XG/PVA mixtures composition by the intrinsic viscosity, the hydrodynamic interaction parameter and the characteristic specific hydrodynamic volume.  The existence of XG/PVA interpolymer associations with less or more compact structures were evidenced by the deviation from additivity of the intrinsic viscosity and zeta potential values bellow and above 30 mV.

Acknowledgement This work was supported by a grant of the Romanian National Authority for Scientific Research, CNCS-UEFISCDI, project number PN–II–ID–PCE–2011–3–0199 (contract 300/2011). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.eurpolymj.2016.09.006. References [1] L.M. Robeson, Polymer Blends: A Comprehensive Review, Carl Hanser Verlag GmbH & Co. KG, 2007. [2] P. Fitzpatrick, J. Meadows, I. Ratcliffe, P.A. Williams, Control of the properties of xanthan/glucomannan mixed gels by varying xanthan fine structure, Carbohydr. Polym. 92 (2013) 1018–1025. [3] I.C. Alupei, M. Popa, M. Hamcerencu, M.J.M. Abadie, Superabsorbant hydrogels based on xanthan and poly(vinyl alcohol): 1. The study of the swelling properties, Eur. Polym. J. 38 (2002) 2313–2320. [4] D. Ciolacu, M. Cazacu, Synthesis of new hydrogels based on xanthan and cellulose allomorphs, Cell. Chem. Technol. 45 (2011) 163–169. [5] F. Garcia-Ochoa, V.E. Santos, J.A. Casas, E. Gomez, Xanthan gum: production, recovery, and properties, Biotechnol. Adv. 18 (2000) 549–579. [6] S.S. Bhattacharya, A. Mishra, D. Pal, A.K. Ghosh, A. Ghosh, S. Banerjee, et al, Synthesis and characterization of poly(acrylic acid)/poly(vinyl alcohol)xanthan gum interpenetrating network (IPN) superabsorbent polymeric composites, Polym. Plast. Technol. 5 (2012) 878–884. [7] S. Ray, S. Banerjee, S. Maiti, B. Laha, S. Barik, B. Sa, et al, Novel interpenetrating network microspheres of xanthan gum-poly(vinyl alcohol) for the delivery of diclofenac sodium to the intestine-in vitro and in vivo evaluation, Drug Delivery 17 (2010) 508–519. [8] T. Bhunia, A. Giri, T. Nasim, D. Chattopadhyay, A. Bandyopadhyay, Uniquely different PVA-xanthan gum irradiated membranes as transdermal diltiazem delivery device, Carbohydr. Polym. 95 (2013) 252–261. [9] K. Deshmukh, M.B. Ahamed, R.R. Deshmukh, S.K. Khadheer Pasha, K.K. Sadasivuni, D. Ponnamma, K. Chidambaram, Synergistic effect of vanadium pentoxide and graphene oxide in polyvinyl alcohol for energy storage application, Eur. Polym. J. 76 (2016) 14–27. [10] E. Chiellini, A. Corti, S. D’Antone, R. Solaro, Biodegradation of poly(vinyl alcohol) based materials, Prog. Polym. Sci. 28 (2003) 963–1014. [11] A.P.V. Pereira, L.V. Wander, R.L. Orefice, Novel multicomponent silicate-poly(vinyl alcohol) hybrids with controlled reactivity, J. Non-Cryst. Solids 273 (2000) 180–185. [12] T. Yamaoka, Y. Tabata, Y. Ikada, Comparison of body distribution of poly(vinyl alcohol) with other water-soluble polymers after intravenous administration, J. Pharmaceut. Pharmacol. 47 (1995) 479–486. [13] Y.-F. Tang, Y.-M. Du, X.-W. Hu, X.-W. Shi, J.F. Kennedy, Rheological characterization of a novel thermosensitive chitosan/poly(vinyl alcohol) blend hydrogel, Carbohydr. Polym. 67 (2007) 491–499. [14] M. Kopcˇilová, J. Hubácˇková, J. Ru˚zˇicˇka, M. Dvorˇácˇková, M. Julinová, M. Koutny´, M. Tomalová, P. Alexy, P. Bugaj, J. Filip, Biodegradability and mechanical properties of poly(vinyl alcohol)-based blend plastics prepared through extrusion method, J. Polym. Environ. 21 (2013) 88–94. [15] C.-E. Brunchi, S. Morariu, M. Bercea, Intrinsic viscosity and conformational parameters of xanthan in aqueous solutions: salt addition effect, Colloids Surf. B: Biointerfaces 122 (2014) 512–519. [16] C.-E. Brunchi, M. Bercea, S. Morariu, M. Dascalu, Some properties of xanthan gum in aqueous solutions: effect of temperature and pH, J. Polym. Res. 23 (7) (2016) 1–8. [17] S. Krimm, C.Y. Liang, G.B.B.M. Sutherland, Infrared spectra of high polymers. V. Polyvinyl Alcohol, J. Polym. Sci. 22 (1956) 227–247. [18] L. Su, W.K. Ji, W.Z. Lan, X.Q. Dong, Chemical modification of xanthan gum to increase dissolution rate, Carbohydr. Polym. 53 (2003) 497–499.

172

C.-E. Brunchi et al. / European Polymer Journal 84 (2016) 161–172

[19] M. Kanou, K. Nakanishi, A. Hashimoto, T. Kameoka, Influences of monosaccharides and its glycosidic linkage on infrared spectral characteristics of disaccharides in aqueous solutions, Appl. Spectrosc. 59 (2005) 885–892. [20] S. Barsberg, A.R. Sanadi, H. Jørgensen, A new density functional theory (DFT) based method for supporting the assignment of vibrational signatures of mannan and cellulose – analysis of palm kernel cake hydrolysis by ATR-FT-IR spectroscopy as a case study, Carbohydr. Polym. 85 (2011) 457–464. [21] M. Kacˇuráková, M. Mathlouthi, FTIR and laser-Raman spectra of oligosaccharides in water: characterization of the glycosidic bond, Carbohydr. Res. 284 (1996) 145–157. [22] Y. Maréchal, M. Milas, M. Rinaudo, Hydration of hyaluronan polysaccharide observed by IR spectrometry. III. Structure and mechanism of hydration, Biopolymers (Biospectroscopy) 72 (2003) 162–173. [23] K. Haxaire, Y. Maréchal, M. Milas, M. Rinaudo, Hydration of polysaccharide hyaluronan observed by IR spectrometry. I. Preliminary experiments and band assignments, Biopolymers (Biospectroscopy) 72 (2003) 10–20. [24] H. Bourara, S. Hadjout, Z. Benabdelghani, A. Etxeberria, Miscibility and hydrogen bonding in blends of poly(4-vinylphenol)/poly(vinyl methyl ketone), Polymers 6 (2014) 2752–2763. [25] A.N.J. Heyn, The infrared absorption spectrum of dextran and its bound water, Biopolymers 13 (1974) 475–506. [26] C.-E. Brunchi, M. Bercea, S. Morariu, The behavior of polyacrylonitrile solutions in dimethylformamide at very low concentrations, High. Perform. Polym. 20 (2008) 311–322. [27] B.A. Wolf, Polyelectrolytes revisited: reliable determination of intrinsic viscosities, Macromol. Rapid Commun. 28 (2007) 164–170. [28] C.-E. Brunchi, M. Bercea, S. Morariu, Viscometric and rheological study of polyacrylonitrile solutions, e-Polymers 065 (2009) 1–10. [29] I.E. Lamatic, M. Bercea, S. Morariu, Intrinsic viscosity of aqueous polyvinyl alcohol solutions, Rev. Roum. Chim. 54 (2009) 981–986. [30] L. Ghimici, C.-E. Brunchi, Titanium dioxide separation from water by PEG and Pluronic type polymers, Sep. Purif. Technol. 103 (2013) 306–312. [31] C.-E. Brunchi, L. Ghimici, PEG in aqueous salt solutions. Viscosity and separation ability in a TiO2 suspension, Rev. Roum. Chim. 58 (2013) 183–188. [32] C.-E. Brunchi, M. Bercea, S. Morariu, Hydrodynamic properties of polymer mixtures in solution, J. Chem. Eng. Data 55 (2010) 4399–4405. [33] C.-E. Brunchi, S. Morariu, M. Cazacu, M. Bercea, Influence of the solvent quality on the thermodynamic behavior of polymethylphenylsiloxane solutions, Ind. Eng. Chem. Res. 49 (2010) 12740–12746. [34] S. Morariu, C.-E. Brunchi, M. Cazacu, M. Bercea, The behavior of poly(dimethylsiloxane-co-diphenylsiloxane)s in good and theta solvents, J. Chem. Eng. Data 56 (2011) 1468–1475. [35] S. Morariu, C.-E. Brunchi, M. Bercea, The behavior of chitosan in solvents with different ionic strengths, Ind. Eng. Chem. Res. 51 (2012) 12959–12966. [36] M. Bercea, L.E. Nita, J. Eckelt, B.A. Wolf, Polyelectrolyte complexes: phase diagram and intrinsic viscosities of the system water/poly(2vinylpyridinium-Br)/poly(styrene sulfonate-Na), Macromol. Chem. Phys. 213 (2012) 2504–2513. [37] L.E. Nita, A. Chiriac, M. Bercea, B.A. Wolf, Synergistic behavior of poly(aspartic acid) and Pluronic F127 in aqueous solution as studied by viscometry and dynamic light scattering, Colloids Surf. B: Biointerfaces 103 (2013) 544–549. [38] S. Morariu, M. Bercea, C.-E. Brunchi, Effect of cryogenic treatment on the rheological properties of chitosan/poly(vinyl alcohol) hydrogels, Ind. Eng. Chem. Res. 54 (2015) 11475–11482. [39] A. Eckelt, J. Eckelt, B.A. Wolf, Interpolymer complexes and polymer compatibility, Macromol. Rapid Commun. 33 (2012) 1933–1937. [40] A. Eich, B.A. Wolf, Intrinsic viscosities of polyelectrolytes: determination and modeling of the effects of extra salt, Chem. Phys. Chem. 12 (2011) 2786– 2790. [41] A. Spatareanu, M. Bercea, T. Budtova, V. Harabagiu, L. Sacarescu, S. Coseri, Synthesis, characterization and solution behaviour of oxidized pullulan, Carbohydr. Polym. 111 (2014) 63–71. [42] P.R. Suresha, M.V. Badiger, B.A. Wolf, Polyelectrolytes in dilute solution: viscometric access to coil dimensions and salt effects, RSC Adv. 5 (2015) 27674–27681. [43] J. Eckelt, A. Knopf, B.A. Wolf, Polyelectrolytes: intrinsic viscosities in the absence and in the presence of salt, Macromolecules 41 (2008) 912–918. [44] F. Samadi, B.A. Wolf, Y. Guo, A. Zhang, A.D. Schluter, Branched versus linear polyelectrolytes: intrinsic viscosity of peripherically charged dendronized poly(methyl methacrylate)s and of their uncharged analogues, Macromolecules 41 (2008) 8173–8180. [45] W.R. Krigbaum, F.T. Wall, Viscosities of binary polymeric mixtures, J. Polym. Sci. V (4) (1950) 505–514. [46] G. Staikos, C. Tsitsilianis, Viscometric investigation of the poly(acrylic acid)-polyacrylamide interpolymer association, J. Appl. Polym. Sci. 42 (1991) 867–872. [47] G. Staikos, G. Bokias, C. Tsitsilianis, The viscometric methods in the investigation of the polyacid-polybase interpolymer complexes, J. Appl. Polym. Sci. 48 (1993) 215–217. [48] G.-G. Bumbu, C. Vasile, J. Eckelt, B.A. Wolf, Investigation of the interpolymer complex between hydroxypropyl cellulose and maleic acid-styrene copolymer, Macromol. Chem. Phys. 205 (2004) 1869–1876. [49] M.J. Parets, R. Garcia, V. Soria, A. Campos, Solution properties of polyelectrolytes-V. Viscometric study of mixed polyanions in pure water, Eur. Polym. J. 26 (1990) 767–773. [50] R. Garcia, C.M. Gomez, I. Porcar, J.E. Figueruelo, A. Campos, Eur. Polym. J. 33 (1997) 1723–1730. [51] M. Rinaudo, M. Milas, T.M.B. Bresolin, J. Ganter, Physical properties of xanthan, galactomannan and their mixtures in aqueous solutions, Macromol. Symp. 140 (1999) 115–124. [52] M. Milas, W.F. Reed, S. Printz, Conformations and flexibility of native and re-natured xanthan in aqueous solutions, Int. J. Biol. Macromol. 18 (1996) 211–221. [53] J. Moffat, V.J. Morris, S. Al-Assaf, A.P. Gunning, Visualisation of xanthan conformation by atomic force microscopy, Carbohydr. Polym. 148 (2016) 380– 389. [54] A. Abbaszadeh, W. MacNaughtan, G. Sworn, T.J. Foster, New insights into xanthan synergistic interactions with konjac glucomannan: a novel interaction mechanism proposal, Carbohydr. Polym. 144 (2016) 168–177. [55] S.M. Fitzsimons, J.T. Tobin, E.R. Morris, Synergistic binding of konjac glucomannan to xanthan on mixing at room temperature, Food Hydrocoll. 22 (2008) 36–46. [56] T. Bresolin, M. Milas, M. Rinaudo, J. Ganter, Xanthan–galactomannan interactions as related to xanthan conformations, Int. J. Biol. Macromol. 23 (1998) 263–275. [57] E.R. Morris, T.J. Foster, Role of conformation in synergistic interactions of xanthan, Carbohydr. Polym. 23 (1994) 133–135. [58] I.H. Smith, K.C. Symes, C.J. Lawson, E.R. Morris, Influence of the pyruvate content of xanthan on macromolecular association in solution, Int. J. Biol. Macromol. 3 (1981) 129–134. [59] M. Tako, S. Nakamura, Evidence for intramolecular associations in xanthan molecules in aqueous media, Agric. Biol. Chem. 53 (1989) 1941–1946. [60] F.S. Moolman, M. Meunier, P.W. Labuschagnea, P.-A. Truter, Compatibility of polyvinyl alcohol and poly(methyl vinyl ether-co-maleic acid) blends estimated by molecular dynamics, Polymer 46 (2005) 6192–6200. [61] M. Wis´niewska, The temperature effect on electrokinetic properties of the silica-polyvinyl alcohol (PVA) system, Colloid Polym. Sci. 289 (2011) 341– 344. [62] S. Mathur, B.M. Moudgil, Adsorption mechanism(s) on poly(ethylene oxide) on oxide surfaces, J. Colloid Interface Sci. 196 (1997) 92–98.