Evaluation of Local Gelation Behavior of Aqueous Methylcellulose Solution Using Quartz Crystal Microbalance

Abstract

The physical gelation of an aqueous methylcellulose (MC) solution in response to temperature change was evaluated using a quartz crystal microbalance (QCM), which is an extremely sensitive mass balance that measures changes in mass per unit area from nanogram to microgram level. Then, the potential use of QCM for interfacial selective viscoelasticity measurements was investigated. The viscosity changes accompanying gelation were observed as resonance frequency shifts. The gelation temperature determined from the temperature dependence of the resonance frequency shifts showed good agreement with the gelation temperatures obtained by visual inclination observation and rheology measurements. Furthermore, MC molecules were adsorbed, and the local concentration increased at the interface with hydrophobic quartz units due to the surface properties. We believe that QCM enables the evaluation of interfacial viscoelasticity.

 

This Paper was Originally Published in Japanese in J. Japan Inst. Met. Mater. 85 (2021) 23–29. Captions of all Figures and Tables are modified.

1. Introductions

Electronic materials, adhesives, lubricants, selectively permeable membranes, and biomaterials exhibit their functionality when in contact with different materials. Therefore, for these polymeric materials to achieve high performance, the structure and properties at the interface between the polymer and a dissimilar material must be accurately understood for material design. The polymer interface is at a significantly different energy state compared with the bulk material,^1,2) and its structure and properties are notably different. Presently, spectroscopy using X-rays, neutrons, and sum frequency generation provides a nondestructive and accurate method to analyze the structures of material interfaces,^3–7) thereby enabling the incorporation of interfacial structures in material designs. However, when analyzing viscoelasticity, it is extremely difficult to selectively apply and detect micro-strain and force without destroying the material structures near the interface; thus, evaluation methods are limited.

The piezoelectric oscillation of a quartz resonator has been used as an ultra sensitive mass sensor, utilizing the Sauerbrey relationship between the resonance frequency and the mass per unit area deposited on the crystal.⁸⁾ This relationship has enabled the quartz crystal microbalance (QCM) to be a mainstay of vacuum science. Kanazawa and co-workers demonstrated that QCM operation in liquids was possible,⁹⁾ opening opportunities for QCM to contribute to many electrochemical and biological investigations.¹⁰⁾ However, the frequency changes depending not only on mass but also on viscoelasticity in liquids. Consequently, we focused on the depth of ultra-small strains and high-frequency vibrations from the probe of a QCM quartz crystal resonator propagated to a liquid at a distance from the interface. We conceived that by applying vibrations from the quartz crystal resonator to create the strain necessary to evaluate viscoelasticity, it would be possible to selectively evaluate localized regions near the interface.

Methylcellulose (MC) is a chemically modified cellulose where some or all of the hydrophilic hydroxyl groups (OH groups) at C2, C3, and C6 of the anhydro-β-glucose ring repeating unit are replaced with the hydrophobic methoxy group (CH₃O). The chemical structure of MC was showed in Fig. 1. It is produced from cellulose molecules that are isolated and purified from trees; therefore, it is a natural resource with a low environmental burden. MC with moderate methoxy group substitution per glucose ring (degree of substitution (DS) of 1.5–2.0) has a nonuniform DS in a chain; thus, it behaves as a water-soluble polymer at low temperatures, reversibly transitioning to a cloudy hydrogel as the temperature increases.^11,12) Heyman believed that the solution-to-gel (sol-gel) transition of MC is caused by dehydration of the molecular chain during heating.¹³⁾ Kato et al. proposed hydrogen bond and dipole–dipole interaction operating between molecular chains, as well as hydrophobic interaction between chain segments with a high DS, as candidates for reversible physical crosslinking resulting in reversible gelation.¹⁴⁾ Kobayashi et al. showed that MC first undergoes liquid-liquid phase separation forming a polymer dense phase and a dilute phase, followed by the formation of physical crosslinking in the polymer dense phase. Thus, gelation occurs in two steps.¹⁵⁾ However, much of the initial path of phase separation is unknown, and many models have been proposed. Takeshita et al. and Fairclough et al. proposed that the phase separation of MC is spinodal decomposition.^16,17) On the other hand, Lodge et al. concluded that the process involves nucleation and growth mechanism,¹⁸⁾ while Tanaka et al. explained that it was viscoelastic phase separation.¹⁹⁾ Therefore, the phase separation of aqueous MC solutions and the detailed gelation mechanism are still unclear.

Fig. 1

Chemical structure of methylcellulose used in this study.

In this study, QCM was used to examine the physical gelation behavior of an aqueous methylcellulose (MC) solution that changed in a thermoreversible manner to gain new insights into MC gelation. Changes in the resonance frequency of the quartz crystal resonator in the aqueous MC solution and the dissipation rate were evaluated as a function of temperature, which enabled the gelation behavior of the aqueous MC solution to be measured. Results were compared with the bulk gelation behavior obtained via traditional transition evaluation methods, namely visual inclination observation, light transmittance measurement, and the measurement of rheological properties. The latter method is frequently used. In addition, by regulating the electrode surface properties of the quartz crystal resonator, the interfacial interaction between the quartz crystal resonator electrodes and the aqueous MC solution was changed, the impact of the interface on the gelation of the aqueous MC solution was evaluated, and the interface selectivity of the viscoelasticity measurement method using the quartz crystal resonator was examined.

2. Quartz Crystal Microbalance

The QCM method is an extremely sensitive weighing method that detects changes in mass at the molecular level on the quartz crystal resonator electrodes through changes in resonance frequency.⁸⁾ The AT-cut quartz crystal resonator is a typical quartz crystal resonator comprising an extremely thin quartz crystal cut along AT plane with thin metal film electrodes attached to both sides (Fig. 2(a)). Due to the inverse piezoelectric effect of the crystal, when an alternating-current (AC) voltage is applied to electrodes, thickness-shear vibration occurs in the direction parallel to the crystal surface at a certain resonance frequency. The resonance frequency of the quartz crystal resonator depends on the thickness of the crystal and is typically high (in the order of 10⁶ Hz). In addition, mechanical strain induced by the quartz crystal resonator has been reported to be extremely small, at a sub-nanometer scale.

Fig. 2

(a) Optical image of a quartz oscillator with gold electrodes. (b) Diagram of the equivalent circuit of a quartz oscillator. C₀ is the capacitance of electrode. L₁, R₁ and C₁ are the inductance, resistance, and capacitance of the AT-cut quartz, respectively. (c) Spectrum of electrical conductance obtained via QCM with the corresponding resonance frequency (f) and dissipation (Γ). (d) Schematic representation of a quartz oscillator in a Newtonian liquid. The solid red line represents the propagation of vibrations damped depending on distance from the interface (z). u is the displacement field of a shear wave. δ is the penetration depth represented by the analysis depth of the quartz oscillator in the liquid.²⁰⁾

When the quartz crystal resonator is vibrating at the resonance frequency, it can be represented by the equivalent circuit in Fig. 2(b). The electrical characteristics of the quartz crystal resonator change in response to the environment and the application of mechanical power.²⁰⁾ The QCM can evaluate changes in the mass on the electrode substrate and changes in the viscoelasticity of a substance adhering to the electrode substrate from the electrical characteristics of the quartz crystal resonator. Figure 2(c) shows the conductance spectrum of the quartz crystal resonator measured via QCM. The peak frequency is referred to as the resonance frequency (f), while the half width at half maximum (Γ) of the peak is the dissipation rate due to the viscoelasticity of substance adhering to the electrodes. Changes in f and Γ (Δf and ΔΓ) are used to evaluate changes in mass and viscoelasticity.

Complex resonance frequency ($\varDelta f^{*}$) is expressed as a function of Δf and ΔΓ in the following equation:²⁰⁾   

\begin{equation} \varDelta f^{*} = \varDelta f + i \times \varDelta \varGamma \end{equation} (1)

When a minute amount of a rigid substance comes in contact with electrodes of the quartz crystal resonator, the complex resonance frequency changes in proportion to the change in the mass on the electrodes, which is at the nanogram scale. However, since changes in the dissipation rate are extremely small compared with changes in the resonance frequency (|ΔΓ| ≪ |Δf|), changes in mass on the electrodes and changes in the resonance frequency are expressed by eq. (2):⁸⁾   

\begin{equation} \varDelta f^{*} \approx \varDelta f = -2 \times n \times f_{0}{}^{2} \times \varDelta m/\text{Z}_{\text{q}} \end{equation} (2)

where n represents harmonics, f₀ is the basic resonance frequency of the quartz crystal resonator, Δm is the change in mass per unit volume on the quartz crystal resonator electrodes, and Z_q is the acoustic impedance of AT-cut quartz crystal (8.8 × 10⁶ kg m⁻² s⁻¹).

However, when the quartz crystal resonator is in contact with a homogeneous Newtonian fluid in a semi-infinite region wider than the limit of vibration propagation, the complex resonance frequency is proportional to the product of the viscosity and the density of liquid and is expressed by the following equation:^9,20–25)   

\begin{align} \varDelta f^{*}/f_{0} &= (-1+i) \times (2nf_{0})^{1/2}\\ &\quad \times (\eta_{\text{liq}}\times \rho_{\text{liq}})^{1/2}/(\pi^{1/2}\times \text{Z}_{\text{q}}) \end{align} (3)

η_liq and ρ_liq are the viscosity and density of the liquid, respectively. When $\varDelta f^{*}$ is replaced by Δf and ΔΓ according to eq. (1), eq. (3) can be re-organized and expressed as:   

\begin{equation} |\varDelta f| = |\varDelta\varGamma| = n^{1/2}\times f_{0}{}^{3/2}\times(\eta_{\text{liq}}\times\rho_{\text{liq}})^{1/2}/(\pi^{1/2}\times \text{Z}_{\text{q}}) \end{equation} (4)

As shown in eq. (4), the absolute value of change of the resonance frequency and dissipation rate are equal, moreover, Δf and ΔΓ are reciprocals of each other. However, this relationship does not apply to non-Newtonian fluids.

Furthermore, the vibration amplitude of the quartz crystal resonator attenuates exponentially from the interface. Thus, the distance at which the amplitude is 1/e of vibration amplitude at the interface is called the viscous invasiveness (δ), which is the analytical depth of the quartz crystal resonator in a liquid (Fig. 2(d)). δ is expressed by the following equation:^26,27)   

\begin{equation} \delta = [\eta_{\text{liq}}/(\pi\times f_{0}\times\rho_{\text{liq}})]^{1/2} \end{equation} (5)

When the quartz crystal resonator has a basic resonance frequency of 9 MHz, the viscous invasiveness in water is approximately 190 nm. Therefore, the extremely small amplitude of the quartz crystal resonator can be directly applied to the interface as a stimulant and the viscoelasticity of a microregion near the interface can be measured.

3. Experimental

3.1 Sample and solution preparation

We used Metolose^® SM-25 provided by Shin-Etsu Chemical Co., Ltd. as MC with a weight-average molecular weight (M_w) of 5.1 × 10⁴ g/mol, a polydispersity (M_w/M_n) of 1.52, and a DS of 1.8. Vacuum-dried MC powder was weighed using an electronic balance. An aqueous solution with a concentration of 10 times that of the critical entanglement concentration ($\text{C}^{*}$) was prepared. Here, $\text{C}^{*}$ is the concentration where adjacent polymer chains in the solution come in contact resulting in entanglement. Moreover, it is the concentration where the dilute solution transitions to a semi-dilute solution. Since the viscosity of the polymer solution increases significantly above $\text{C}^{*}$, it is a key concentration that characterizes the viscosity of a polymer solution. $\text{C}^{*}$ is expressed as the inverse of limiting viscosity [η], which represents the coefficient of viscosity per molecule:²⁸⁾   

\begin{equation} \text{C}^{*} \approx 1/[\eta] \end{equation} (6)

The $\text{C}^{*}$ of the MC used in the present experiment was 0.58 mass% in water at 25°C. Viscosity was measured using an Ubbelohde-type viscometer. When water was directly added to the MC powder, only the powder surface became wet and partially dissolved aggregates formed; thus, we prepared the solution via the hydrothermal method where water heated to 70°C or higher was added. The prepared aqueous MC solution was stored overnight at 4°C in a refrigerator before use.

3.2 Visual inclination observation

We visually observed the gelation behavior of the bulk aqueous MC solution. The solution was heated from 10°C at a rate of 1°C/min. At pre-determined temperatures, the screw-cap vial containing the solution was tilted 90° to visually observe if there was a change in state and fluidity. Subsequently, the solution was cooled to 10°C at the same rate and the change from gel to solution was visually observed. The temperature of the solution was recorded using a thermocouple thermometer. When tilting the screw-cap vial, a solution that flowed under its own weight was defined as “sol” and a solution that did not flow was defined as “gel”. The temperature at which fluidity was lost was defined as the gelation temperature (T_gel). Each experiment was performed five times and the average value was used.

3.3 Light transmittance measurements

A spectrophotometer (V-650, JASCO Corporation) was used to evaluate the temperature dependence of transmittance to assess the phase separation behavior that induces the gelation of the aqueous MC solution. An aqueous MC solution with concentration of 10 $\text{C}^{*}$ was placed in a quartz cell with an optical path length of 1 cm. The cell was sealed with a rubber stopper to avoid the evaporation of water during heating. An aluminum heating block was used to increase the temperature of the solution from 20 to 70°C at a rate of 1°C/min. The transmittance of light with a wavelength of 380–780 nm was measured every 5°C, as well as every 2°C between 40 and 60°C in the vicinity of the gelation temperature. Subsequently, the MC gel that was heated to 70°C was cooled at a rate of 1°C/min, and the transmittance of light with a wavelength of 380–780 nm was measured every 5°C. This measurement was performed every 2°C between 40 and 20°C in the vicinity of the temperature where the gel returned to sol.

3.4 Rheology measurements

A rheometer (MCR302, Anton Paar GmbH) was used to evaluate changes in viscoelasticity associated with the gelation of the bulk aqueous MC solution. We poured approximately 20 mL of the solution into the cup of a coaxial cylindrical jig, and after moving the rotor (inner cylinder) to the measurement position, a sample from the upper part of the rotor was removed using a pipette (trimming) to improve the reproducibility of the experimental data. The upper part of the sample was sealed with silicon oil with viscosity of 10 cS (Shin-Etsu Chemical Co., Ltd.). The provided lid for the prevention of solvent evaporation was applied from the top of the jig to minimize changes in concentration through solvent evaporation during measurement. The resonance frequency was set at 1 Hz and strain was fixed at 1%, which is the linear range. The storage modulus (G′) and the loss modulus (G′′) were measured in 1°C increments. Temperature was regulated via a Peltier temperature control system (C-PTD200, Anton Paar GmbH) and was increased from 10 to 70°C at a rate of 1°C/min. Subsequently, the aqueous MC solution was cooled down to 10°C at the same rate and the temperature dependence of the moduli during gel-to-sol transition was evaluated.

3.5 QCM measurements

Figure 3 illustrates a schematic of the experimental device. The quartz crystal resonator had a basic resonance frequency of 9 MHz and gold (Au) electrodes. The surface of the electrodes was ultrasonically cleaned for 15 min in ethanol. The quartz crystal resonator with a Teflon dip-type cell, which allows for measurement in liquids, was immersed in the aqueous MC solution. The temperature of the solution was regulated using an aluminum heating block and was heated from 10 to 70°C at a heating rate of 1°C/min. The temperature of the solution near the quartz crystal resonator was recorded using a thermocouple thermometer. Δf and ΔΓ were measured via a quartz crystal microbalance measurement system, QCM922A (SEIKO EG&G Co., Ltd.). The MC gel heated to 70°C was cooled to 10°C at a rate of 1°C/min, and the temperature dependence of Δf and ΔΓ during transition from gel to sol was evaluated.

Fig. 3

Schematic illustration of QCM measurement equipment.

In addition to the Au electrode quartz crystal resonator, we used a silica (SiO₂) electrode as the hydrophilic surface. The natural oxide layer (Si–OH group) at the outermost surface of the silicon (Si) electrode quartz crystal resonator was hydrophobized (Si–H groups) using a 1% of hydrofluoric acid aqueous solution. We examined the impact on gelation of the interaction at the interface between these three electrodes and the aqueous MC solution with a concentration of 10$\text{C}^{*}$.

4. Results and Discussion

4.1 Visual inclination observation of the gelation behavior of the aqueous MC solution

Figure 4(a) shows photographs of the state change associated with increased temperature of the aqueous MC solution with a concentration of 10$\text{C}^{*}$. At lower temperatures, MC dissolved in water forming a clear and colorless aqueous solution. However, as the temperature increased, the solution became cloudy due to the change in the solubility of the MC molecules in water. Methoxy groups within the MC molecules dehydrated as the temperature increased.²⁹⁾ As chain segments with numerous hydrophobic methoxy groups aggregated through hydrophobic interaction, phase separation into a polymer dense phase and a dilute phase occurred,¹⁷⁾ resulting in the clouding of the aqueous solution. As clouding progressed, the viscosity of the solution increased. As the temperature continued to increase, at a specific temperature, the solution completely lost its fluidity and changed to a gel. Within the polymer dense phase, physical crosslinking occurred leading to aggregation. The hydrophobic parts of the MC acted as crosslinking points, leading to the reversible formation of a network structure. The temperature at which the fluidity of the solution was completely lost was 50.9 ± 0.9°C, which was set as the visual T_gel. During cooling (Fig. 4(b)), the solution cleared with decreasing temperature and fluidity re-appeared at approximately 30°C, which was lower than that in case of the T_gel obtained during heating. Thus, hysteresis was observed in the gelation behavior of the aqueous MC solution.

Fig. 4

Optical images of the aqueous methylcellulose solution at various temperatures during (a) heating and (b) cooling.

4.2 Coarsening of the aggregate structure associated with gelation

Figure 5(a) shows the temperature dependence of transmittance during heating measured in the wavelength band of 380–780 nm. When the aqueous methylcellulose solution was clear and colorless, transmittance was almost 100%. However, transmittance at 380 nm was lower at approximately 80% because the MC molecules absorb light near 210 nm within the ultraviolet region. When heated, transmittance rapidly decreased at approximately 35–40°C. The temperature at which transmittance began to decrease shifted toward higher temperatures as the wavelength of the light increased. We believe this was due to the size of aggregates consisting of MC molecules. When the temperature of the aqueous MC solution was low (20–30°C), the molecules dissolved in water and minimal aggregation of the molecules occurred.³⁰⁾ Therefore, most light passed through the aqueous MC solution without scattering. However, since light with short wavelengths was scattered by the MC molecules, short-wavelength transmittance was reduced even at the low temperatures. As the temperature of the solution increased, MC molecules aggregated. As the size of the aggregates increased, initially only short-wavelength light was scattered, reducing transmittance. As the size of aggregates further increased, even longer wavelength light was scattered At temperatures of 60°C or higher, transmittance of all wavelengths reduced to 0% and visual observation confirmed complete clouding of the MC gel.

Fig. 5

Temperature dependence of the transmittance of the aqueous methylcellulose solution during (a) heating and (b) cooling.

The temperature dependence of transmittance during cooling (Fig. 5(b)) displayed a different behavior from that during heating. At all wavelengths, transmittance rapidly increased at temperatures, above which transmittance rapidly decreased during heating, i.e., 20°C. This confirmed hysteresis and the thermal reversibility of the gelation of aqueous MC solutions with respect to the temperature dependence of transmittance. In addition, since the transmittance of long wavelengths gradually increased with cooling, it is assumed that the size of the aggregates in the MC molecular chain gradually decreased during cooling.

4.3 Moduli changes associated with the gelation of the aqueous MC solution

Figure 6 shows the temperature dependence of the storage modulus (G′) and loss modulus (G′′) of the aqueous MC solution with a concentration of 10$\text{C}^{*}$. At lower temperatures, G′′ (viscosity component) was larger than G′ (elasticity component), indicating that the aqueous MC solution was in the sol state. The gradual decrease in the moduli between 10 and 40°C was caused by the increased thermal activity of molecules with increasing temperature that led to decreasing intermolecular interaction, which in turn lowered the solution viscosity.³¹⁾ Above approximately 40°C, all moduli rapidly increased. At higher temperatures, G′ was larger than G′′ and the aqueous MC solution transitioned to the gel state. Thus, we defined the temperature at which G′ and G′′ reversed as the “rheometer T_gel”. The rheometer T_gel of the aqueous MC solution with a concentration of 10$\text{C}^{*}$ was 50.4°C. On the other hand, during the cooling of the MC gel, G′ and G′′ both displayed constant values down to 40°C, followed by a rapid decrease from approximately 35°C. The relative values of G′ and G′′ reversed at 25°C. The moduli of the aqueous MC solution followed different paths during heating and cooling, thus displaying hysteresis, which was attributed to the gelation of MC being an entropy-driven reaction.³²⁾ To hydrate the dehydrated MC molecules, entropy must be lowered to change water molecules from a random state to a relatively ordered state. To produce the required energy state, the aqueous solution must be cooled. Therefore, the network structure of the MC molecular chain was maintained at a lower temperature, leading to observation of hysteresis. After cooling to below 15°C, the values of the moduli were similar to those before heating. This indicates that the gelation of the aqueous MC solution is thermally reversible.

Fig. 6

Temperature dependence of the storage modulus (G′) and loss modulus (G′′) of the aqueous methylcellulose solution during heating and cooling.

4.4 Investigation of gelation behavior of the aqueous MC solution via QCM

We used a quartz crystal resonator with Au electrodes to measure the temperature dependence of changes in resonance frequency (Δf) and dissipation rate (ΔΓ) associated with the gelation of the aqueous MC solution with a concentration of 10$\text{C}^{*}$. The results are shown in Fig. 7. Δf and ΔΓ were dependent on the solution viscosity. The gradual increase in Δf (decrease in ΔΓ) between 10 and 40°C was caused by a decrease in the solution viscosity associated with increasing temperature, similar to the gradual decrease in moduli observed during the measurement of rheological properties.³³⁾ Δf decreased rapidly (increase in ΔΓ) at temperatures above 45°C because the gelation of the aqueous MC solution rapidly increased the solution viscosity. Subsequently, at 60°C and higher, the gelation of the aqueous MC solution was complete; therefore, Δf and ΔΓ displayed constant values. As such, Δf and ΔΓ changed due to the gelation of the aqueous MC solution. Therefore, we defined the inflection point where Δf rapidly decreased as the “QCM T_gel”. The QCM T_gel of the aqueous MC solution with the concentration of 10$\text{C}^{*}$ was 50.4 ± 0.5°C. It is listed in Table 1 along with T_gel obtained from the measurement of rheological properties. The T_gel values obtained via the different measurement methods were consistent. When the MC gel was cooled from 70 to 10°C, Δf and ΔΓ did not change until near 40°C, displaying constant values. From approximately 35°C, Δf rapidly increased (ΔΓ decreased) to a value similar to the pre-heating value at 20°C and below. Hysteresis and thermal reversibility of the aqueous MC solution observed during rheological measurements were also observed as changes in Δf and ΔΓ during the QCM measurements, empirically demonstrating that QCM can be used to evaluate the gelation of the aqueous MC solution.

Fig. 7

Temperature dependence of (a) the resonance frequency shifts and (b) the dissipation shift of the aqueous methylcellulose solution during heating and cooling.

Table 1 Gelation temperature of an aqueous methylcellulose solution determined by visual observation, rheometer measurements and QCM evaluation.

4.5 Effect of the surface properties of the quartz crystal resonator

Figure 8 shows the temperature dependence of Δf and ΔΓ measured via quartz crystal resonators with three different electrodes, namely Au, hydrophilic SiO₂, and hydrophobic Si. There was no notable difference in the temperature dependence of Δf and ΔΓ for the aqueous MC solution when using the Au and SiO₂ electrodes. However, when using the hydrophobic Si electrode, the resonance frequency was approximately 1000 Hz lower than that measured with the Au and SiO₂ electrodes, while the dissipation rate was approximately 500 Hz higher, indicating that the viscosity of the solution was high near the interface. Since the change in the resonance frequency was greater than the change in the dissipation rate, it was assumed that MC molecular chains were adsorbed onto the electrode thereby increasing the viscosity. In addition, the T_gel values obtained from temperature dependence of Δf for each electrode are summarized in Table 2. The lowest value was observed in case of the Si electrode. It was implied that at the interface with the Si electrode substrate, the local MC concentration was higher than at the Au and SiO₂ electrode interfaces.

Fig. 8

Temperature dependence of (a) the resonance frequency shifts and (b) the dissipation shifts of the aqueous methylcellulose solution with Au (yellow circles), SiO₂ (gray squares) and Si (blue triangles) electrodes.

Table 2 Gelation temperature (T_gel) of an aqueous methylcellulose solution via a quartz oscillator with various electrodes. γ is the surface free energy and RMS is the root mean square of the surface roughness of the electrodes.

To determine the reason for the differences in T_gel for the different electrodes, we evaluated the surface free energy (γ) and root mean square (RMS) roughness of each electrode surface. γ was calculated from the contact angle of the electrode surface to water and diiodomethane, while RMS roughness was evaluated via atomic force microscopy of the electrode surface (Table 2).

The SiO₂ and the Si electrodes that had been hydrophobized with hydrofluoric acid displayed similar γ values that were larger than that of the Au electrode. γ of the Au electrode was close to the theoretical value;^34,35) however, the Si and SiO₂ electrodes deviated from the hydrophilic and hydrophobic behavior observed for a typical Si substrate surface. In case of the Si electrode, we believe that this was due to the extremely unstable nature of the Si–H group that covers the outermost surface of the dehydrated Si electrode, along with the impact of being oxidized even in ambient atmosphere. For the SiO₂ electrode, the deviation was due to the inadequate acidification of the outermost surface of the electrode because of the structure of the quartz crystal resonator. In addition, MC has both hydrophilic hydroxyl groups and hydrophobic methoxy groups, thus displaying amphiphilicity. Therefore, adhesion occurred on all surfaces independent of the hydrophilic or hydrophobic nature of the electrode resulting in no significant difference in γ. The above results are expected since surface free energy exhibits extremely short distance interaction compared with the depth of analysis of the quartz crystal resonator.

The RMS roughness was approximately 0.8 nm for the Au and SiO₂ electrodes, but notably lower than 1.8 nm for the hydrophobized Si electrode surface. Since the RMS roughness of the Si electrode surface prior to hydrophobization by hydrofluoric acid was 0.76 nm, the hydrofluoric acid treatment likely increased the surface roughness of the Si electrode. Consequently, the surface area of the Si electrode with hydrophobization treatment increased and more MC molecules adhered to the interface than in case of other electrodes. (This was reflected by the change in Δf.) Thus, it can be concluded that an increase in the surface area of electrodes resulted in a higher local concentration of the aqueous MC solution near the interface, leading to a decrease in T_gel.

5. Conclusions

We successfully observed changes in solution viscosity associated with the gelation of an aqueous MC solution with the concentration of 10$\text{C}^{*}$ as changes in resonance frequency by via QCM and thereby determined the gelation temperature. Similar to the rheological behavior, the hysteresis and thermal reversibility of the aqueous MC solution were successfully demonstrated using the temperature dependence of Δf and ΔΓ. In addition, the temperature dependence of Δf and ΔΓ associated with the gelation of the solution using three different electrodes was investigated. The measurements confirmed an increase in the adsorption of MC molecules onto the increased surface area of the quartz crystal resonator electrodes and an associated decrease in T_gel, indicating that QCM can measure viscoelasticity near the interface.

Acknowledgement

This work was supported by JSPS KAKENHI Grant Numbers JP19H05720 and JP16K05926. Part of this study utilized the Alumni Association research fund of the Faculty of Engineering at Mie University. In addition, the measurement of rheological properties was performed at National Institute for Materials Science (NIMS) supported by NIMS Joint Research Hub Program. We would like to extend our most sincere appreciation to the NIMS Data-driven Polymer Design Group Leader, Dr. Masanobu Naito, for providing an opportunity for measurement.

REFERENCES

• 1)  M. Kawaguchi and K. Hayakawa: Interfaces and Dispersion Colloids Basic Principles and Applications, (CMC Publishing CO., LTD., Tokyo, Japan, 2017).
• 2)  Y. Fujii, H. Morita, A. Takahara and K. Tanaka: in Glass Transition, Dynamics and Heterogeneity of Polymer Thin Films, (Springer Verlag Berlin, Heidelberg, 2013) pp. 1–27.
• 3)   K.  Mitamura,  N.L.  Yamada,  H.  Sagehashi,  N.  Torikai,  H.  Arita,  M.  Terada,  M.  Kobayashi,  S.  Sato,  H.  Seto,  S.  Goko,  M.  Furusaka,  T.  Oda,  M.  Hino,  H.  Jinnai and  A.  Takahara: Polym. J. 45 (2013) 100–108.
• 4)   H.  Tanoue,  N.L.  Yamada,  K.  Ito and  H.  Yokoyama: Langmuir 33 (2017) 5166–5172.
• 5)   D.A.  Barkley,  N.S.  Jiang,  M.  Sen,  M.K.  Endoh,  J.G.  Rudick,  T.  Koga,  Y.G.  Zhang,  O.  Gang,  G.C.  Yuan,  S.K.  Satija,  D.  Kawaguchi,  K.  Tanaka and  A.  Karim: Macromolecules 50 (2017) 7657–7665.
• 6)   X.  Li and  X.L.  Lu: Macromolecules 51 (2018) 6653–6660.
• 7)   H.  Ogawa,  M.  Takenaka,  T.  Miyazaki,  T.  Kabe and  T.  Kanaya: Macromolecules 51 (2018) 10040–10051.
• 8)   G.  Sauerbrey: Z. Phys. 155 (1959) 206–222.
• 9)   K.K.  Kanazawa and  J.G.  Gordon: Anal. Chem. 57 (1985) 1770–1771.
• 10)   K.  Sadman,  C.G.  Wiener,  R.A.  Weiss,  C.C.  White,  K.R.  Shull and  B.D.  Vogt: Anal. Chem. 90 (2018) 4079–4088.
• 11)   M.  Hirrien,  J.  Desbrieres and  M.  Rinaudo: Carbohydr. Polym. 31 (1996) 243–252.
• 12)   L.  Li,  P.M.  Thangamathesvaran,  C.Y.  Yue,  K.C.  Tam,  X.  Hu and  Y.C.  Lam: Langmuir 17 (2001) 8062–8068.
• 13)   E.  Heymann: Trans. Faraday Soc. 31 (1935) 846–864.
• 14)   T.  Kato,  M.  Yokoyama and  A.  Takahashi: Colloid Polym. Sci. 256 (1978) 15–21.
• 15)   K.  Kobayashi,  C.-i.  Huang and  T.P.  Lodge: Macromolecules 32 (1999) 7070–7077.
• 16)   H.  Takeshita,  K.  Saito,  M.  Miya,  K.  Takenaka and  T.  Shiomi: J. Polym. Sci. B Polym. Phys. 48 (2010) 168–174.
• 17)   J.P.  Fairclough,  H.  Yu,  O.  Kelly,  A.J.  Ryan,  R.L.  Sammler and  M.  Radler: Langmuir 28 (2012) 10551–10557.
• 18)   J.R.  Lott,  J.W.  McAllister,  M.  Wasbrough,  R.L.  Sammler,  F.S.  Bates and  T.P.  Lodge: Macromolecules 46 (2013) 9760–9771.
• 19)   H.  Tanaka: J. Phys. Condens. Matter 12 (2000) R207–R264.
• 20)  D. Johannsmann: The Quartz Crystal Microbalance in Soft Matter Research: Fundamentals and Modeling, (Springer International Publishing, Cham, Switzerland, 2015).
• 21)   T.  Nomura and  O.  Hattori: Anal. Chim. Acta 115 (1980) 323–326.
• 22)   T.  Nomura and  M.  Okuhara: Anal. Chim. Acta 142 (1982) 281–284.
• 23)   S.  Bruckenstein and  M.  Shay: Electrochim. Acta 30 (1985) 1295–1300.
• 24)   P.  Schön,  R.  Michalek and  L.  Walder: Anal. Chem. 71 (1999) 3305–3310.
• 25)   I.  Goubaidoulline,  J.  Reuber,  F.  Merz and  D.  Johannsmann: J. Appl. Phys. 98 (2005) 014305.
• 26)  L.D. Landau and E.M. Lifshitz: Fluid Mechanics, (Pergamon, Oxford, 1987) pp. 44–94.
• 27)  H. Muramatsu: in Handbook of Biosensor and Chemical Sensors, (TECHNO SYSTEM CO., LTD, Tokyo, Japan, 2007) pp. 374–383.
• 28)   S.A.  Arvidson,  J.R.  Lott,  J.W.  McAllister,  J.  Zhang,  F.S.  Bates,  T.P.  Lodge,  R.L.  Sammler,  Y.  Li and  M.  Brackhagen: Macromolecules 46 (2013) 300–309.
• 29)   J.  Desbrieres,  M.  Hirrien and  M.  Rinaudo: Carbohydr. Polym. 37 (1998) 145–152.
• 30)   R.  Bodvik,  A.  Dedinaite,  L.  Karlson,  M.  Bergstrom,  P.  Baverback,  J.S.  Pedersen,  K.  Edwards,  G.  Karlsson,  I.  Varga and  P.M.  Claesson: Colloids Surf. A 354 (2010) 162–171.
• 31)   S.M.  Silva,  F.V.  Pinto,  F.E.  Antunes,  M.G.  Miguel,  J.J.  Sousa and  A.A.  Pais: J. Colloid Interface Sci. 327 (2008) 333–340.
• 32)   L.  Li: Macromolecules 35 (2002) 5990–5998.
• 33)   M.I.S.  Veríssimo,  A.A.C.C.  Pais and  M.T.S.R.  Gomes: Carbohydr. Polym. 82 (2010) 363–369.
• 34)   R.A.  Erb: J. Phys. Chem. 72 (1968) 2412–2417.
• 35)   D.  Parobek and  H.  Liu: 2D Materials 2 (2015) 032001.