Characterisation of xanthan gum solutions using dynamic light scattering and rheology

Abstract

The ‘weak gel’ formation mechanism of xanthan gum solutions has been examined using rheological and light scattering methods. Significant differences are observed between the critical overlap concentration (c^∗) measured using both techniques. The effect of shear on c^∗ for semi-flexible molecules such as xanthan is demonstrated. The approximate theories of Doi and Edwards (1984). The Theory of Polymer Dynamics. Oxford: Clarendon Press) for rigid rods and De Gennes (1979). Scaling Concepts in Polymer Solutions. London: Cornell University Press) for random coils are used to explain the differences observed between rheological and light scattering measurements of c^∗. A second concentration regime parameter due to aggregation, c^∗∗, is observed in dynamic light scattering results (DLS). The parameter c^∗∗ has been previously observed for xanthan systems. Light scattering measurements in various electrolytes suggest that the use of DLS is able to distinguish between the transition from truly dilute behaviour to semi-dilute behaviour (c^∗) and the critical concentration for aggregation (c^∗∗).

Introduction

Xanthan gum is a high molecular weight extracellular polysaccharide produced by bacteria of the genus Xanthomonas. Commercial xanthan gum is extracted from Xanthomonas campestris. However, for the purpose of most scientific studies, it is extracted from Xanthomonas phaseoli and Xanthomonas juglandis. Xanthan gum may be chemically considered as an anionic polyelectrolyte, with a backbone chain consisting of (1→4)β-d-glucan cellulose. The polymer backbone is substituted at C-3 on alternate glucose residues with a trisaccharide sidechain. The sidechain consists of β-d-mannopyranosyl-(1→4)-(α-$\text{d}$-glucuropyranosyl)-(1→2)-β-$\text{d}$-mannopyranoside 6-acetate. A pyruvic acid residue is linked to the 4 and 6 positions between 31–56% of the terminal d-mannose residues (Morris, 1982, Rinaudo, 1980, Ross-Murphy, 1983). The xanthan molecule undergoes a conformation transition from an ordered double helix to a random coil when heated to above, between 40 and 80°C depending on the ionic strength of solution (Paradossi, 1982, Sato, 1984). The ordered conformation of the xanthan molecule is semi-flexible with a hydrodynamic length varying between 600 and 2000 nm (Holzwarth, 1978, Ross-Murphy, 1983) and a hydrodynamic diameter of the order of 2 nm (Stokke, Elgsaeter & Smidsrod, 1986). A number of workers have investigated the degree of flexibility of the xanthan molecule, generally characterised in terms of persistence length (L_p). Static light scattering and sedimentation coefficients were used with application of the Benoit and Doty (1953) expansion of the Krakty Porod model for a wormlike chain to predict the flexibility (Richards, 1980). Experimental measurements of the persistence length have yielded values which vary between 130±20 nm (Berth, 1996, Sato, 1984) or 45–70 nm (Muller, 1984, Tinland, 1990, Tinland, 1989). The two regions of L_p determined suggest that thermal history and xanthan source are important parameters in considering the flexibility of xanthan molecules and its behaviour in solution.

Solutions of xanthan gum are one of the most intensively studied polysaccharide systems both in terms of rheological properties and physical chemistry. The behaviour of xanthan in both dilute and semi-dilute solutions has been extensively studied (Carriere, 1993, Coviello, 1986, Coviello, 1987, Southwick, 1981, Tinland, 1990, Whitcomb, 1978). The determination of concentration regimes is an important aspect of previous work and illustrates the complicated nature in which xanthan behaves in solution. The concentration at which individual polymer molecules begin to physically interact is defined as c^∗. The concept of c^∗ is based on the theory that the polymer coils in solution in their quiescent state occupy a hydrodynamic volume where above a critical concentration for close packing the molecules interact. DLS may be used to trace the diffusivity of molecules in solutions through a range of concentrations and gain insight into the solutions behaviour in a truly dilute environment. c^∗ has been evaluated by application of light scattering techniques through analysis of trends in apparent diffusion coefficients (Coviello, 1987, Tinland, 1990).

Rheologically, c^∗ of polymer solutions is determined by extrapolation of shear viscosity data to zero shear viscosities over a range of concentrations. On a double logarithmic plot of zero shear viscosity versus concentration, a slope of approximately one in the dilute regime and between three and four in the semi-dilute regime is observed, depending on the rigidity of the molecule (Lapasin, Pricl, Paoletti & Zanetti, 1990). The concentration at which the slope of zero shear viscosity versus concentration changes, is defined as c^∗. A limitation of the rheological measure of c^∗ is the measure of zero shear viscosities which involves the application of shear to the system. In the case of solutions of xanthan gum, shear forces will dominate Brownian motion at low shear rates, causing molecular alignment, resulting in the solution not being in a state of true random motion and Rotary Peclet Numbers >1 (Lapasin & Pricl, 1993). Doi and Edwards (1984) show theoretically that under the application of shear forces, rod-like molecules will align in solution and move longitudinally in a constrained tube before molecular interaction occurs. Although xanthan molecules may not be considered as rodlike (Coviello et al., 1987), their extended form in solution suggests the alignment of molecules in solution may be applied qualitatively.

Because of the semi-flexible nature of xanthan and the tendency for aggregation or self-association at low concentrations, it is difficult to obtain a measure of c^∗ when shear is applied (Kojima, 1988, Tinland, 1990). Southwick et al. (1981) applied quasi elastic light scattering to solutions of xanthan gum and determined a c^∗ of 0.02 wt%. Their calculations assumed sphericity and the authors speculated a much lower value to be more indicative of the initial concentration for molecular overlap. Southwick et al. (1981) observed a second transition at 0.07 wt% where a general decrease in relaxation times and the onset of bimodal behaviour occurs, which was attributed to the onset of anisotropy. Bimodal behaviour is defined as having two distinct diffusion coefficients. Jamieson, Southwick and Blackwell (1982) applied static light scattering through a range of concentrations and observed a discontinuity of translational diffusion at 0.02 wt%, interpreted as the onset of the semi-dilute regime, however, no diffusion plateau consistent with Stokes–Einstein measure of diluteness was observed. The workers also observed a linear relationship between intensity and concentration for flow birefringence of xanthan solutions with a concentration above 0.1 wt%. This observation agrees qualitatively, but not quantitatively with the predictions of Doi and Edwards (1984) for an anisotropic entangled solution of rigid-rod molecules.

Coviello et al. (1986) examined dynamic and static light scattering from native and modified xanthan and commented on the ambiguity in determining c^∗ for semi-flexible objects, suggesting that the value obtained is often dependent on experimental technique and interpretation of the data. Tinland et al. (1990) comment on the reptation of semidilute solutions of worm-like polymers and report a c^∗ of 0.04 wt% for xanthan gum of a similar molecular weight to that reported here from deviation of the self diffusion coefficient as measured through DLS.

Milas, Rinaudo, Knipper and Schuppsier (1990) applied steady shear and dynamic rheological measurements to an unpasteurized sample of xanthan gum, and through sensitive rheological measurements observed two concentration transitions which were assigned c^∗ and c^∗∗. These workers observed a departure from a slope of one of the specific viscosity versus c·[η] (concentration times intrinsic viscosity) at 0.126 wt%. A similar value was also obtained by applying large deformation rheology where a departure from Newtonian behaviour to pseudoplastic behaviour was observed. A second concentration transition was assigned c^∗∗ and observed at 0.60–0.78 wt%. c^∗∗ was assigned to the concentration where specific viscosity followed (c·[η])^3.7 which is similar to that observed for polymeric melts. Milas et al. (1990) speculate that c^∗∗ represents the onset of a concentrated domain in which a uniform distribution of polymer segments is present. Other workers have expanded on the concept of c^∗ and c^∗∗, by applying complementary rheooptical techniques of birefringence and dichroism to commercial xanthan gum samples. A steady state orientation angle after shearing was observed for samples with a concentration of 0.2 wt% or greater using dichroism. Although the presence of aggregates was observed down to concentrations of 50 ppm, no steady state orientation was observed (Meyer, Fuller, Clark & Kulicke, 1993). The presence of aggregates at such low concentrations is consistent with the observations of Jamieson et al. (1982).

For the most part, previous determinations of c^∗ and c^∗∗ have involved combination of two or more experimental techniques and/or the application of shear to solutions of xanthan gum. DLS is a non purtabative technique commonly used to measure the diffusion of polymer molecules in solution and has been used in the present study for the determination of both c^∗ and c^∗∗. DLS measures a diffusion coefficient which has dimensions of (length)²/time and is a space average measure of an individual molecules diffusional path which is interpreted by analysis of time correlation functions (TCF).

Light scattering of xanthan gum solutions is complicated by the semi-flexible nature of the xanthan molecules proposed to be a right hand double helix (Bezemer, 1993, Milas, 1995) and the presence of micro-aggregates at low concentrations caused by xanthan gum self-association (Meyer et al., 1993). Increased rigidity of the xanthan ordered conformation at higher salt conditions by the reduction of the electrical double layer, suggests that the hydrodynamic radius of an individual molecule would increase as electrolyte is added. Self-association processes in xanthan solutions complicates this hypothesis through the introduction of large aggregates at low concentrations enabling molecules to interact at lower concentrations than would be theoretically expected for a solution of semi-flexible non-interacting molecules (Ferry, 1980).

The work presented here aims to examine xanthan solutions using both rheological techniques and DLS in order to characterise the boundary between dilute and semi-dilute solutions. Experiments have been conducted to quantify c^∗ and c^∗∗ as defined by previous work (Meyer, 1993, Milas, 1990, Southwick, 1981) in the absence of shear. The results have been compared with shear rheology data and theoretical calculations. Varying electrolyte concentrations have been used to determine the effect of electrostatic screening on the light scattering and rheological data. The effect of heat treatment on xanthan solutions has been investigated to establish the effect of removing molecular aggregates by converting the xanthan molecule into its re-natured form ensuring a true molecular dispersion (Capron, Brigand & Muller, 1997). The high aspect ratio of xanthan molecules in solution requires characterisation of the effect of different lengths of the scattering vector (K) on the xanthan molecule diffusion (Berth, 1996, Gamini, 1994). The observed diffusional trends have been compared with theoretical predictions of rigid rods in solution (Doi & Edwards, 1984) and scaling concepts for random coils (De Gennes, 1979).