Biomacromolecules 2010, 11, 187–191
187
Viscoelasticity and Phase Separation of Aqueous Na-Type Gellan Solution Yoko Nitta,*,† Rheo Takahashi,‡ and Katsuyoshi Nishinari§ School of Human Science and Environment, University of Hyogo, Hyogo 670-0092, Japan, Graduate School of Engineering, Gunma University, Kiryu, Gunma 376-8515, Japan, and Graduate School of Human Life Science, Osaka City University, Sugimoto, Sumiyoshiku, Osaka 558-8585, Japan Received September 18, 2009
Viscoelastic and phase behaviors of Na-type gellan were investigated. The complex shear modulus of aqueous gellan solutions was monitored as a function of concentration, temperature, and molar mass. At relatively low concentrations, the loss shear modulus, G′′, increased steeply at a certain temperature upon cooling, which was attributed to helix formation. Above a certain concentration, a steep increase and then gradual decrease of G′′ was observed. The decrease of G′′ was significant and is considered to be due to anisotropic phase formation, judging from the macroscopic phase separation into upper isotropic and lower anisotropic phases of a gellan solution kept quiescently in a vial. Polarized light microscopy produced an image showing a nematic liquid crystalline phase. Although gel formation was not observed when gellan alone was used in the present study, the gellan formed a gel in the presence of an appropriate amount of salt.
Introduction Gellan gum is a microbial polysaccharide produced by Sphingomonas elodea. It is composed of 1,3-β-D-glucose, 1,4β-D-glucuronic acid, 1,4-β-D-glucose, and 1,4-R-L-rhamnose.1,2 Native gellan has two acyl substituents present on the 3-linked glucose and is deacylated by heat treatment at high pH. Deacylated gellan, often just called gellan, is widely used as a gelling agent and texture modifier in the food industry. The term “gellan” in this article is also used to refer to the deacylated gellan. Upon cooling, the aqueous gellan solution forms a transparent and heat-resistant gel in the presence of cations. The effectiveness of cations to enhance gelation decreases in the order Ca2+ > Mg2+ . K+ > Na+.3-5 The gelation of gellan is now believed to occur under conditions favorable for the aggregation of double helices. Two disordered chains of gellan at higher temperatures were observed to form a double helix at lower temperatures in dilute solution by optical rotation, light scattering, and intrinsic viscosity measurements.6,7 Monovalent cations are thought to promote double helix formation and aggregation of the helices with an effectiveness of K+ > Na+. Divalent cations seem to bind directly with gellan molecules to form aggregates of gellan helices with the effectiveness of Ca2+ > Mg2+.8,9 If gel-promoting cations like Ca2+, Mg2+, and K+ exist as counterions of gellan, aggregation occurs even in dilute solution, which makes molar mass determination difficult. To determine the molar mass of gellan, it is necessary to prevent undesirable aggregation and microgel formation by removing gel-promoting cations. When most of the counterions were replaced by Na+, aggregation and microgel formation were not observed and the molar mass could be determined. Studies of the solution properties of a sodium form (Na-type) of gellan with different molar masses10-12 indicated that the persistence length of gellan was 9.4 nm in a disordered state (40 °C) and 98 nm in a helical state (25 * To whom correspondence should be addressed. E-mail:
[email protected]. † University of Hyogo. ‡ Gunma University. § Osaka City University.
°C) in 25 mM NaCl. When purified Na-type gellan is used, the viscoelastic behavior in a semidilute solution and critical behavior near the sol-gel transition can be studied, allowing for direct correlation with properties in a dilute solution. However, the effects of the molar mass of gellan on viscoelastic and critical behaviors remain to be clarified. In the present study, viscoelastic properties of Na-type gellan with different molar masses were investigated by dynamic shear oscillatory measurements. The phase behavior was also investigated.
Experimental Section Materials. A sodium-type gellan gum supplied by San-Ei Gen F. F. I., Osaka, Japan, was dissolved in 10-45 mM NaOH aqueous solution for 12 h at 50 °C. Each of the chemically degraded gellans was centrifuged at 1 × 104 × g for 1 h. The supernatant was poured into a large quantity of isopropyl alcohol to precipitate the gellan. Aqueous solutions of each fraction were treated using a mixed-bed-type ion exchanger and neutralized with 0.1 M NaOH to convert them into the Na-type. Dry Na-type gellan samples were obtained by freeze-drying the solutions for 1 week. Three samples were chosen for the present study and were designated G-10, G-15, and G-30 to represent the order of increasing molar mass. Dynamic Viscoelastic Measurements. The storage shear modulus G′ and the loss shear modulus G′′ were measured using the strain-controlled rheometer, RFSII (TA Instruments, New Castle, Delaware) with a transducer, the sensitivity limit of which was 0.002 g m. The aqueous gellan solution was prepared by heating the solution to 90 °C and then pouring it onto a parallel plate geometry (plate diameter: 50 mm) at 50 °C. The periphery of the sample was immediately covered with silicone oil to prevent the evaporation of water. Strain dependences of G′ and G′′ were examined to determine a linear viscoelastic regime. Within this regime, G′ and G′′ were measured at various temperatures and frequencies. The requisite strain became higher for less viscous samples because of the sensitivity limit of the instrument. For example, a 20% strain was necessary for a 5 wt % G-10 solution (Figure 2) where linear viscoelasticity was observed at least below a 40% strain. The temperature was controlled using the refrigerated circulator FS18-MV (Julabo, Seelbach, Germany). Polarized Light Microscopy. The samples were observed using the polarized light microscope Nikon Coolpix 4500 (Nikon, Tokyo, Japan) equipped with a camera at room temperature (∼20 °C).
10.1021/bm901063k CCC: $40.75 2010 American Chemical Society Published on Web 10/09/2009
188
Biomacromolecules, Vol. 11, No. 1, 2010
Nitta et al.
Figure 1. The storage shear modulus G′ (O) and the loss shear modulus G′′ (b) of G-10 aqueous solution upon cooling. Concentrations were (a) 1, (b) 2, (c) 2.5, (d) 3, (e) 3.5, and (f) 5 wt %. Frequency was 1 rad s-1. The sample was kept at least 20 min at each temperature.
Figure 2. Frequency dependence of G′ (O) and G′′ (b) of 5 wt % G-10 at 5 °C and 20% strain.
Figure 3. G′ of the G-10 aqueous solution as a function of concentration at 10 °C and at 10 rad/s (2), 46 rad/s (0), and 100 rad/s (b).
Results and Discussion The storage shear modulus G′ and the loss shear modulus G′′ of the G-10 solution upon cooling are shown in Figure 1. At a concentration from 1 to 2.5 wt % (Figure 1a-c), G′′ increased steeply around 30 °C. Upon cooling, a step-like increase of G′′ was observed at a certain temperature at which an exothermic peak appeared in the DSC and the molar ellipticity at 202 nm changed. This was attributed to helix formation.4,13 At concentrations from 3 to 5 wt % (Figure 1d-f), G′′ increased and then decreased with decreasing temperature. If a true gel was formed and G′ and G′′ decreased upon cooling, a reel-chain model14 or slippage15 could be the explanation. However, the frequency dependence of G′ and G′′ showed behavior typical of an associative polymer network (Figure 2),16 which is not the behavior of a true gel. Some associative polymer networks show a behavior fitting the Maxwell model, essentially a liquid, which flows after a long time even if it does not flow immediately just after the tilting.17 Therefore, the decrease of G′′ with a lowering of temperature in the present study must result from another cause. G′ as a function of concentration at different frequencies at 10 °C is shown in Figure 3. G′ showed a maximum at a certain concentration, which has also been observed for polymers that form liquid crystals such as xanthan,18 κ-carrageenan,19 and methyl cellulose.20 When an 8 wt % G-10 solution was kept quiescently in a vial and stored at low temperatures, phase separation was identified as shown in Figure 4. The separation into two phases was clearly observed through a polarizing plate, where the upper phase was transparent and the lower phase was
Figure 4. Appearance of 8 wt % G-10 between crossed polarizers.
Figure 5. Image produced by polarized light microscopy of 8 wt % G-10 at room temperature.
turbid. An image of the lower phase was obtained by polarized light microscopy and is shown in Figure 5. The lower phase was a nematic liquid crystalline phase. Therefore, it is thought that the decrease of G′′ was induced by biphasic phase formation, which consists of isotropic and anisotropic phases, and not by gel formation. In fact, the frequency dependence of
Aqueous Na-Type Gellan Solution
Biomacromolecules, Vol. 11, No. 1, 2010
189
Figure 6. G′ (O) and G′′ (b) upon cooling for (a) 1, (b) 2, and (c) 3 wt % of G-15 and (d) 1, (e) 2, and (f) 3 wt % of G-30. Frequency was 1 rad s-1. The sample was kept at least 20 min at each temperature.
Figure 7. Plots of Cb as a function of molar mass at room temperature.
G′ and G′′ in Figure 2 was quite similar to that of an actin filament solution, which showed a liquid crystalline behavior at higher concentrations.21 The temperature dependence of G′ and G′′ for gellan samples with the higher molar masses of G-15 and G-30 is shown in Figure 6. The decrease of G′′ upon cooling was observed for 3 wt % G-15 and 2-3 wt % G-30 (Figure 6c,e,f). Birefringence of G-10, G-15, and G-30 was monitored by polarized light microscopy and the concentration at which birefringence appeared, Cb, was plotted against molar mass (Figure 7). Birefringence appeared at a lower concentration with increasing gellan molar mass, which is consistent with previously reported results for xanthan18,22 and κ-carrageenan-NaI systems.23 The frequency dependence of G′ and G′′ of G-15 at each temperature (data not shown) indicated that G-15 did not form a true gel below 3 wt % (Figure 6a-c). The frequency dependence of G′ and G′′ for 3 wt % G-30 is shown in Figure 8. At 30 and 0 °C, G′ was larger than G′′ with increasing frequency at all tested frequencies. This behavior has previously been called a “weak gel” type behavior. Regrettably, this terminology is not the most accurate description because a “weak gel” is essentially a liquid and not a gel [Ross-Murphy SB (2008) in: Tanaka, F., Ed. Lecture Notes at Kyoto University]. Therefore, we concluded that G-30 also does not form a gel below 3 wt %. Gel formation of gellan is promoted by the addition of salt, which has been studied in detail.4 The gelation of the present gellan sample was examined in the presence of NaCl. G′ and G′′ of 2 wt % G-30 with 30 mM NaCl upon cooling are shown in Figure 9a. G′ and G′′ increased steeply around 40 °C below
Figure 8. Frequency dependence of G′ (open) and G′′ (closed) of 3 wt % G-30 at 30 °C (triangle) and 0 °C (diamond) at 10% strain.
Figure 9. (a) G′ (O) and G′′ (b) of 2 wt % G-30 with 30 mM NaCl upon cooling at 1 rad/s, and (b) frequency dependence of G′ and G′′ of the same sample at 5 °C and 1% strain.
which G′ became larger than G′′. The frequency dependence test showed G′ ∼ G′′ ∼ ωn (n ) 0.16), which represented a critical gel-like behavior (Figure 9b)24,25 and suggested that a gel can be obtained by addition of more than 30 mM NaCl. In the presence of 100 mM NaCl, a true gel was obtained for all tested gellan solutions. G′ and G′′ of 0.3 wt % G-10, G-15, and G-30 with 100 mM NaCl on cooling and subsequent
190
Biomacromolecules, Vol. 11, No. 1, 2010
Nitta et al.
Figure 10. Upper figures: Temperature dependence of G′ (O) and G′′ (b) of 0.3 wt % gellan with 100 mM NaCl on cooling and subsequent heating at 0.5 °C/min and 1 rad/s for (a) G-10, (b) G-15, and (c) G-30. Lower figures (d), (e), and (f): Frequency dependence of G′ and G′′ at 35 °C for each sample.
heating are shown in Figure 10a-c. Upon cooling, G′ and G′′ increased steeply around 40 °C, below which G′ became more than 10 times larger than G′′. The frequency dependence test showed that G′ and G′′ were independent of frequencies (Figure 10d-f), which is typical gel-like behavior. Thermal hysteresis was observed where G′ and G′′ decreased steeply around 70 °C on heating (Figure 10a-c), and such thermal hysteresis is often observed for a true gel.26 It was previously thought that Na-type gellan forms a gel without addition of salt, although its gelling ability is weak in comparison to a potassium-type gellan. The present study contradicts this, showing that Na-type gellan alone did not form a gel even at higher concentrations, but phase separated into isotropic and anisotropic phases. This result may originate from differences in (i) purity of Na-type gellan, (ii) molar mass, and (iii) polydispersity of the present sample compared to those of previous samples. First, insufficiently purified Na-type gellan contains a non-negligible amount of divalent cations. Because a slight amount of divalent cations strongly promotes gel formation, the Na-type gellan at high concentrations in previous studies might have formed a gel due to the presence of divalent cations. Second, G-30 was the sample with the highest molar mass among the gellan samples examined in the present study and showed a “weak gel”-like behavior that was previously and inaccurately classified as a gel by some authors and should be regarded as a liquid. In a previous study,4 both “weak gel” and “true gel” were classified as gels in a state diagram (see Figure 3 in ref 4). Third, the molar mass distribution of the gellan used in the present study is far narrower than that recorded in previous works. This quality might be necessary to make phase separation clear, because a broad molar mass distribution would result in solutions dominated by the viscoelasticity of gellan with higher molar masses, leading to highly viscous solutions. As observed in the κ-carrageenan-NaI system, a highly viscous sample made macroscopic phase separation difficult.23 A low viscous sample like G-10 (Mw ) 98000, Mw/Mn ) 1.31) showed phase separation, where the dynamic viscosity |η*| at 3 wt %, 1 rad/s, and 25 °C was 10 times smaller than that of 3 wt % gellan,
NaGG-3,27,4 as used by collaborative groups in 1999 (Mw ) 95000, Mw/Mn ) 1.63).28,29 The ionic content of both samples was almost the same: the contents of Na+, K+, Ca2+, and Mg2+ of NaGG-3 were 2.59, 0.009, 0.02, and 0.001%, respectively. Phase separation itself might occur for stiff chains. Macroscopic phase separation was probably hindered in previous works due to the high viscosity at high concentration, which was caused by a higher molar mass fraction. It appears that the gelation process driven by the addition of NaCl in the present study occurred through the same mechanism as proposed previously.4,31 Aggregation of helices is promoted by NaCl, which is thought to be due to an electrostatic shielding effect. Longitudinal and lateral aggregations lead to the formation of a bundle consisting of gellan helices, which was observed by atomic force microscopy.30 The elasticity of the gel is thought to originate from elastically active chains released from junction zones that are formed by the gellan bundle. Gelation then occurs when NaCl promotes the aggregation of gellan helices and the formation of a three-dimensional network consisting of helical bundles and elastically active chains. Recently, another model concerning gel structure was proposed in which the bundle itself is the origin of the elasticity of the gel.31 Either type of gel structure should form irrespective of polydispersity, and thus, the gelation of the present sample can be explained by existing models. However, such an interpretation is only applicable to gellan possessing a molar mass high enough for helix formation, because gellan with a low molar mass (
