Shear induced self-thickening of chitosan/β-cyclodextrin compound solution

Abstract

The molecular association and rheological behavior of the compound solutions of chitosan (CS) and β-cyclodextrin (CD) were investigated. The results show that an appropriate amount of CD could significantly affect the viscoelastic behavior of CS aqueous solution. The CS/CD compound solution presents an obvious multiple shear-thinning and a subsequent shear induced thickening (SIT) behavior which is similar to that of chitosan-grafted polyacrylamide aqueous solution, depending on the mass ratio of CD and CS. Too high or too low CD concentration is unfavorable for shear induced self-thickening of the CS/CD compound solution when CS concentration is fixed. TEM observations of the CS/CD compound solution show that larger CS–CD aggregates induced by hydrogen bonding appear with CD concentration increasing, and result in the increases of viscosity and moduli, which are in agreement with the results of steady and oscillation flows. The increased viscosity of the CS/CD compound solution after strong shearing is attributed mainly to the larger and irregular hydrogen-bonded aggregates inclined to connect into large pieces instead of isolated spheres. This superior and unique rheological property of the compound solution can be promisingly applied for various industrial fields in which accurate control of the solution rheology is required.

---

Introduction

Shear-thickening and rheopexy have been found in many kinds of associative polyelectrolytes possessing metal–ligand bonds,^1,2 host–guest complexation,³ hydrogen bonds⁴ and hydrophobic association.⁵ Usually, these shear thickening behaviors are transient and reversible, and could be ascribed to two possible mechanisms including shear-induced nonlinear high tension along stretched polymer chains beyond the Gaussian range and shear-induced increase in the number of elastically active chains.² Wang et al.⁶ investigated the shear-thickening behavior of partially hydrolyzed polyacrylamide (HPAM) solutions and found a critical shear rate, over which stress increased with the increase of shear time and gradually reached a constant value. However, the viscosity of the sheared solutions recovered once the shear revoked. Chassenieux et al.^5,7 found that amphiphilic copolymer of N,N-dimethylacrylamide and acrylic acid (DMA–AA) showed a shear-induced gelation phenomenon due to hydrophobic association, but the gelation vanished once stopping the high-rate shear. Xu and Craig et al.^8,9 introduced Pd(Pt)-graft-pyridine as the cross-linking agent into dimethylsulfoxide (DMSO) solution of poly-4-vinyl-pyrrolidone (PVP), and consequently observed a transient shear thickening and subsequent shear thinning when shear rate increasing, which was ascribed to macromolecular cross-linking resulting from metal complexation. It is noted that these traditional shear-thickening phenomena of polymer mentioned above strongly depend on shear condition and most of them disappear once ceasing shearing.

In previous reports, some special rheological behaviors including multiple shear thinning and subsequent static self-thickening were observed in chitosan-graft-polyacrylamide (GPAM) aqueous solution.^10,11 Most of multiple shear thinning are found in some associative polymer solutions. Usually, it indicates that the multistep corresponds to architecture changes including intermolecular/intramolecular interaction and molecular conformation with the increase of shear rate.^12,13 Although the GPAM aqueous solution also presents similar multiple shear thinning as most of associative polymers previously, its viscosity can recover promptly and finally exceed the original one when stopping the high-rate shear. Therefore, the whole rheological behavior of GPAM solution consisting of multiple shear thinning and subsequent static self-thickening is called as the shear induced self-thickening (SIT), differing from traditional shear thickening or/and negative thixotropy for polymer systems whose thickening processes strongly depend on shearing.^14–16 These superior and unique rheological properties may be promisingly and effectively applied in rheology modifiers,^17,18 flocculants for wastewater treatments,¹⁹ plugging oil-displace agent and so on. However, due to GPAM being a single-component system, its rheological property is only influenced by concentration and shear condition. Recently, it was found a binary system consisting of chitosan and β-cyclodextrin (β-CD) in aqueous solution can also display the similar SIT behavior.²⁰ Compared with GPAM, the binary system of CS/CD is easier to prepare, more likely to regulate special rheological behavior.

Similar to surfactants, cyclodextrin (CD) is also an important small molecule substance which is often used to compound with polyelectrolyte or associated polymer due to its abundant associate groups.²¹ In the past decades, some polymer complexes consisting of polyelectrolyte and surfactant have been reported. For example, some polyelectrolyte/surfactant systems could associate together to form complex micelles due to hydrophobic groups.^22–25 Some surfactants with charged groups could associate with oppositely charged polyelectrolytes by electrostatic interaction.^26–29 In recent years, polymer complexes containing CD have attracted much attention, especially the hydrophobic associated polymer/CD systems. CD is cyclic compound with 6, 7, or 8 dextrose unites in the ring, corresponding to α-, β-, and γ-CD, respectively. The central cavity is hydrophobic, thus CD can bind hydrophobic species to form ‘inclusion compounds’ through macromolecular interchain hydrophobic association.^30,31 In addition, the massive hydroxyl groups distributed on the outside of ring could form intermolecular hydrogen bonding. Since the outer ring complex of CD is as important as the inner ring inclusion complex, an exhaustive investigation about the former complex should also be conducted systematically.

In this work, a series of CS/CD compound solutions were prepared to investigate its special rheological behavior and consequently the compound effect of CS/CD was investigated. As a result, the correlation between special rheology characteristics and association structure of CS/CD were explored and discussed. In addition, to further confirm the role of hydrogen bonding in this SIT of CS/CD compound, a hydrogen bond breaker, ammonium acetate, is used to investigate rheological behavior.

Experiment section

Material

CS with weight-average molecular weight (M_w) of 1 × 10⁶ Da measured by gel permeation chromatography (GPC) in 1.0 wt% acetic acid was obtained from Aladdin Co. Ltd., China. The deacetylation degree of CS is 0.615, determined by titration and conductimetry. β-CD was obtained from Aladdin Co. Ltd., China. All materials above were used without further purification. All experiment were carried out with deionized water (conductivity is less than 5 μS cm⁻¹).

Solutions preparation

Compound solutions were prepared by dissolving the definite amount of CS and CD powder in 1.0 wt% acetic acid under mildly magnetic stirring at 30 rpm for 24 h at room temperature, respectively. Then all solutions stood for another 24 h before being tested.

Measurements

GPC measurements were carried out on WATERS2690 (WATER, USA) with acetic acid aqueous solution as the eluent. Chitosan standards were used for calibration.

Morphology observations were done by using a JEM-1200EX (JEOL, JPN) transmission electron microscope (TEM). The aqueous solutions were cast onto copper grids coated with carbon films followed by vacuum freeze-drying.

Rheological experiments were conducted by using a stress-controlled rotational rheometer AR-G2 (TA Instruments, USA). A 40 mm cone-plate geometry with a 2° cone angle and a 50 mm gap size was chosen for the steady shear test and a 40 mm parallel-plate geometry with 500 mm gap size for the oscillatory test, respectively. The sample solution was directly and very slowly poured onto the Peltier stage, in order to avoid the shear thinning effect and small air bubbles caused by using pipettes. A simple solvent trap near the margin of the Peltier stage was used to minimize evaporation of the solvent. Steady flow tests were performed in the shear rate range from 10⁻⁴ or 10⁻³ to 10³ s⁻¹ with 10 min as the maximum point time to ensure attain equilibrium at every test point. A peak hold step with defined shear rate and sampling delay time was chosen to monitor the instantaneous viscosity after high-rate shear. Oscillatory frequency sweeps from low to high frequency were performed at 0.1% strain amplitude which was determined as linear viscoelastic region (LVR). All tests were taken at 20 °C and all solution samples performed equilibration for at least 5 min before testing.

Result and discussion

Special rheological behaviors

Fig. 1A gives the steady flow curves for 1 wt% CS solution with different CD concentrations. It can be seen that the dependence of apparent viscosity, η_a, on shear rate for pure CS shows two-region characteristic: Newton region at low shear rate and shear thinning region at high shear rate.³² However, the steady flow curves of CS/CD compound solutions present different shear rate dependences of viscosity, depending on CD concentration. For the compound solution with 0.2 wt% CD, it presents an almost same initial apparent viscosity and steady flow curve as the pure CS sample. With the increase of CD concentration, the flow curves of CS/CD compound solutions present obvious shear thinning in low shear rate region no more than 1.0 s⁻¹. Furthermore, the initial apparent viscosities of them sharply increase when CD concentration is higher than 0.5 wt%. It indicates there is a complex effect between CS and CD molecules which can determine the rheological behavior, and this complex effect greatly depends on the mass ratio of CS/CD. Usually, the obvious increase of initial apparent viscosity for polymer complex is related to the associative structure induced by intermolecular interaction,^{11–13,33,34} this intermolecular interaction is so weak that can be broken at relative low shear rate and recover slowly, therefore it presents extra shear thinning at low shear rate. It is noted that the flow curves of all samples get close to each other or even overlap when the shear rate is higher than 1.0 s⁻¹, indicating that the possible associative structures in these CS/CD compound solutions are destroyed and the molecular architectures in them are similar as that in pure CS solution at high shear rate region. It is noticed that the flow curves at high shear rate (∼1000 s⁻¹) are smooth, indicating that no flow instabilities (like edge fracture and wall slip effects) happens. Fig. 1B presents the dynamic strain sweep curves for 1 wt% CS solution with different CD concentrations. It can be found that the complex viscosity, |η*|, of the CS/CD compound solution with 0.2 wt% CD seems to overlap with that of pure CS and they have little change with increase of strain while the CS/CD compound solutions with 0.5 and 1.0 wt% CD present the larger initial |η*| than pure CS and fast decrease.

Fig. 1 (A) Steady flow curves for 1% CS with different CD concentrations in 1% (w/w) CH₃COOH aqueous solution. (B) Strain sweep curves for 1% CS with different CD concentration in 1% (w/w) CH₃COOH aqueous solution. The frequency of 6.28 rad s⁻¹ of is adopted in dynamic strain sweep.

Fig. 2A shows the steady flow curves of the original and thickened samples undergoing the following process consisting of shearing at 1000 s⁻¹ for 2 min and subsequent recovering at 0.1 s⁻¹ for 30 min. Here, CS concentration is fixed at 1.0%. The steady flow curves of pure CS solution with and without pre-shearing and recovering overlaps each other, indicating that the viscosity drop induced by strong shearing can completely recover and couldn't influence the final state. For the CS/CD compound solutions with CD concentration of 0.2 and 1.0 wt%, their viscosities after thickening are slightly smaller than the original ones. Differing from these samples above, the initial viscosity of CS/CD compound solution containing 0.5 wt% CD after stronger shearing appears an obvious increase and consequently a remarkable shear thinning is observed, indicating a clear shear induced self-thickening behavior like previous reports.^11,35 Furthermore, considering that the results indicated in Fig. 1 and in S1 (ESI†), an interesting deduction can be made: although the initial viscosity of the compound solutions depends on CD concentration and the initial viscosity of sample containing 1.0 wt% CD is much higher than that of 0.5 wt% CD, only the compound solution containing 0.5 wt% CD can demonstrate shear induced self-thickening behavior. It indicates that too high or too low CD concentration is unfavorable for shear induced self-thickening of CS/CD compound solution. Furthermore, since the CS concentration is fixed, the varied CD concentration indeed represents the change of mass ratio in compound solution. In other words, the shear induced self-thickening behavior in CS/CD compound solution only appears in certain mass ratio of them. Fig. 2B presents the time evolution of apparent viscosity of thickened CS/CD compound solutions with different CD concentrations after being sheared at 1000 s⁻¹ for 2 min. It is found that the apparent viscosity of pure CS sample seems to keep constant, while other CS/CD samples present various degree of thickening with time increasing. Differing from the slight viscosity increase of the CS/CD solution sample with 0.2 wt% CD, the viscosities of both samples with 0.5 and 1.0 wt% CD remarkably increase. It is interesting that the CS/CD compound solution with 1.0 wt% CD presents a faster viscosity increase than that with 0.5 wt% CD and its |η*| reaches an constant in 200 seconds, while the compound solution with 0.5 wt% CD appears a relatively slow increase. It needs to be pointed out, as mentioned in Experimental section, steady flow tests were performed in the shear rate range from 10⁻⁴ or 10⁻³ to 10³ s⁻¹ with 10 min as the maximum point time to ensure attain equilibrium at every test point, so the instantaneous viscosity at low shear rates in Fig. 2A is constant. The separate time sweep experiments without pre-shearing like Fig. 2B also reveal that the viscosities of compound solutions are basically unchanged, as shown in Fig. 2C.

Fig. 2 (A) Comparison of steady shear viscosity curves for CS/CD in 1.0% (v/v) CH₃COOH with different CD concentrations, before (open) and after (solid) shear-induced thickening. The shear-induced thickening procedure consists of a high rate shearing at 1000 s⁻¹ for 2 min and subsequent a recovering at 0.1 s⁻¹ for 30 min. (B) Thickening process at 0.1 s⁻¹ for CS/CD in 1.0% (v/v) CH₃COOH with different CD concentration after being sheared at 1000 s⁻¹ for 2 min. (C) Time sweep curves of CS/CD in 1.0% (v/v) CH₃COOH without pre-shearing.

As a useful method to investigate molecular association behavior, dynamic rheological behavior can distinguish different aggregate structures according to distinct viscoelastic response.^36–38 Fig. 3A gives strain sweep results of different CS/CD solutions. With the increase of CD concentration, G′ begins to increase and the linear region at small strain range for CS/CD compound solutions obviously shortens. Both pure CS solution and CS/CD sample containing 0.2 wt% CD have a wide linear region. The CS/CD samples with 0.5 and 1.0 wt% CD concentration deviate from the linear region over the relatively small strain of 0.2% at the same frequency condition, indicating the new architectures in compound solutions are damaged obviously and can't be recovered or reconstructed transiently.³⁹ This extremely short strain linear region should be related to CS/CD associated or network structure induced by hydrogen bonding, and it also results in increasing G′ compared with pure CS sample. According to the above discussions, the addition of CD obviously changes CS aggregate structure from single relaxation into multiple relaxations resulting from intermolecular hydrogen bonding interaction. This transformation can't happen at low CD concentration, so the CS/CD sample of 0.2 wt% CD shows a constant G′ in wide strain range like the pure CS sample. Furthermore, G′ decreases in a multi-stage way with increasing strain, indicating a multiple-step damage of intermolecular structure. In addition, to further confirm recovery of associative structures in compound solutions after ceasing large amplitude oscillation, strain sweep curves after large amplitude oscillation are also given in Fig. 3A. It is noted that the second strain sweep curve get close to the first strain sweep curve or even overlap when the CD concentration is 0 wt%, 0.2 wt% and 1.0 wt%, indicating that associated structure in solution recovers promptly after ceasing large amplitude oscillation. However, the sample of 0.5 wt% CD presents increased G′ and shorter linear region after ceasing large amplitude oscillation. It shows the associated structure of this CS/CD compound induced by hydrogen bonding becomes more complete and stronger compared with the sample before large oscillation, which is consistent with the result of Fig. 2A.

Fig. 3 (A) Strain sweep curves of CS/CD in 1.0% (v/v) CH₃COOH solution with different CD concentration before (open) and after (solid) large amplitude oscillation. The frequency of 6.28 rad s⁻¹ of is adopted in dynamic strain sweep. (B) Frequency sweep curves at 0.1% strain CS/CD in 1.0% (v/v) CH₃COOH solution with different CD concentrations before large amplitude oscillation. (C) Frequency sweep curves of CS/CD in 1.0% (v/v) CH₃COOH solution before (original, open) and after (sheared, solid) large amplitude oscillation at 0.1% strain.

Fig. 3B gives frequency sweep results for CS/CD solutions with different CD concentrations. It can be seen that the frequency sweep curves of CS/CD solution with 0.2 wt% CD are nearly the same as those of pure CS solution and G′ < G′′, which clearly indicates that the molecular relaxation process is in a single relaxation mode.⁴⁰ For pure CS sample, according to the previous results,^41,42 the associated chain segments which are composed of glucose units without deacetylation collapse into aggregate nucleation and those de-acetyl amino glucose units form shell outside the core. The CS molecular movement dissociating from one aggregate to another aggregate in solution is indeed a dynamic equilibrium process, and there is no multiple relaxation induced by extra entanglement between molecular chains. Usually, this relaxation can be described by Single-Factor Maxwell Model.⁴³ Considering the similarity between pure CS and CS/CD solution with 0.2 wt% CD in Fig. 3A, it is suggested that too low CD concentration (0.2 wt%) could hardly change the previous aggregates structure and relaxation behavior of pure CS. However, when CD concentration increases to 0.5 wt%, there are some obvious changes: (1) the viscoelasticity transformation of G′ > G′′ occurs at a low frequency; (2) G′ and G′′ become less dependent on the oscillation frequency. These rheological features usually indicate the formation of an associate network or weak gel induced by hydrogen bonding in compound system.⁴⁰ Furthermore, when CD concentration reaches 1.0 wt%, the CS/CD solution shows higher moduli and less frequency-dependence of moduli, indicating the formation of stronger and/or more complete associate network or weak gel. Enough weak network structures are mainly induced by hydrogen bonding so that no more CS–CD hydrogen bonding forms after strong shearing, resulting in that only a viscosity recovery appears but never exceeds the original one. The above-mentioned results indicate that the aggregate structure of CS/CD compound system changes remarkably when CD concentration is higher than 0.5 wt% and its relaxation varies from single mode to multiple mode of reputation relaxation.³⁷ Fig. 3C gives G′ and G′′ of original and sheared CS/CD solutions. It can be found that after large amplitude oscillation both G′ and G′′ remarkably increase. For the sheared sample, G′ is larger than G′′ and they presents weak frequency dependence to some extent, like the sample of 1.0 wt% CS/1.0 wt% CD in Fig. 3B as mentioned above.

The reason for transformation occurring is probably that enough small CD molecules get into interchain of CS molecules and lead to the aggregates of CS transform into association structure induced by hydrogen bonding. With CD concentration increasing, enough interchain interactions of CS–CD–CS have been developed and lead to an obvious increase of moduli. Based on above rheological behavior of the CS/CD compound systems, a conclusion can be drawn: CD's addition could cause the initial hydrophobic association of CS into aggregates even a weak network structure mainly induced by hydrogen-bonded association and in turn the CS/CD compound systems demonstrate distinct rheological properties.

Molecular aggregates of in compound solutions

Fig. 4 gives TEM images of CS/CD compound samples with 1 wt% CS and different CD concentration. TEM image of pure CD solution can be seen in Fig. S2 (ESI†). As shown in Fig. 4a, some dark spherical domains with diameter of about 100 nm corresponding to CS hydrophobic aggregates exist in original pure CS solution with 1 wt%, which has been reported in chitosan and its derivative solution.⁴⁴ With the addition of 0.2 wt% CD, it can be seen that the morphology of CS/CD compound sample slightly changes into roughly elliptical dark domains, but the size remains unchanged. In Fig. 4c, some dark domains about 100 nm distribute as the dark spherical domains in Fig. 4a, but other larger dark domains with diameter of about 200 nm are observed for the addition of 0.5 wt% CD, indicating that at least two types of association exist in this sample. The observed domains are consistent with the above rheological analysis.

Fig. 4 TEM micrographs of (a) 1 wt% CS/0 wt% CD (b) 1 wt% CS/0.2 wt% CD (c) 1 wt% CS/0.5 wt% CD in 1.0% (v/v) CH₃COOH solution.

Aggregation behavior in pure chitosan solution can occur at very low concentration. Philippova et al. found that chitosan with M_w of 190 kDa can aggregate in 0.02 wt% solution and proposed the critical concentration should be lower for chitosan solution with higher molecular weight.⁴⁵ According to the structure of chitosan, the compound effect after the addition of CD is discussed as following: in CS/CD compound system, CD firstly gives priority to act with hydrophilic chain section of CS aggregate surface because of the hydroxyls outside the CD ring. When the CD concentration is 0.2 wt%, the amount of CD is too few to crosslink CS aggregates and CD molecules only adhere to the surface of CS aggregates, resulting in the unchanged size of aggregate. In addition, there are some hydrophilic segment involving hydrogen bonding sites, such as –OH\–NH₂\–C=O groups at the aggregates surface,^44,46,47 and the CDs attached to the surface may slightly change the aggregates from sphere into ellipsoid. However, when CD concentration is 0.5 wt%, there may be two possible locations for the CDs: some CDs are still attached to the surface of CS aggregates; some CDs serve as crosslink agent to bridge more CS aggregates or individual molecules chains together to form a CS–CD compound association induced by hydrogen bonding. Compared with the former, the latter make the dense aggregate become looser and larger, causing the distance between hydrophobic groups to expand further impairing the hydrophobic association effect. Finally the CS–CD–CS hydrogen bonding interaction becomes the dominate role instead of hydrophobic interaction.

Fig. 5 gives the TEM micrographs of 1% CS/0.5% CD samples before and after shear-induced thickening. As shown in Fig. 5a, some dark spherical domains whose size is about 200 nm evenly distribute in the field of vision. However, domains with larger and irregular shape are observed in the thickened solution experienced recovery course of 30 min, and these domains are inclined to connect into large pieces instead of the isolated spheres. These TEM observations are in agreement with the rheological result analysis above and the dark domain should be attributed to the formation of aggregates, which indicates the increased viscosity during recovery step of 1 wt% CS/0.5 wt% CD should mainly ascribed to the formation of lager scale aggregates after strong shear.¹⁰

Fig. 5 TEM micrographs of 1 wt% CS/0.5 wt% CD in 1.0% (v/v) CH₃COOH solution before (a) and after (b) shear-induced thickening for 30 min. Thickening processes for 1 wt% CS/0.5 wt% CD after being sheared at 1000 s⁻¹.

To further confirm the role of hydrogen bonding in this SIT of CS/CD compound, a hydrogen bond breaker is used to investigate rheological behavior.^44,46,48,49 Urea is one of the most common hydrogen bond breakers, but it can simultaneously destroy hydrogen bonding and hydrophobic interaction which both exist in CS/CD system.⁴³ Thus ammonium acetate (AcONH₄) is chosen as a hydrogen bond breaker, because it only destroy hydrogen bond but has little impact on hydrophobic effect.^48,49 As shown in Fig. 6, at the low shear rate level (≤1 s⁻¹), the thickening effect which means the change of viscosity between original and sheared samples at the same shear rate keeps decreasing obviously with the concentration of AcONH₄ increases. Since the fact that new proton donor NH₄⁺ ions can replace the original proton donors (–NH₂ and –OH in CS, –OH in CD) and destroy the formed intermolecular hydrogen bonds,⁵⁰ so that the formation of lager scale hydrogen bonding aggregates after strong shear becomes frustrated. However, at the high shear rate (>1 s⁻¹), no thickening effect exists in all samples with different AcONH₄ concentrations, which contributes to the complete destruction of the weak hydrogen bonding structure between CS and CD at such strong shear, as indicated by the results of Fig. 1 and 3. Therefore, the effect of SIT obviously decreases for the samples added AcONH₄ as a result of weakening of CS–CD hydrogen bonding intermolecular interactions. In other words, CS–CD intermolecular hydrogen bonds play a main role in the shear-induced thickening progress.

Fig. 6 Influence of added hydrogen bond breaker AcONH₄ on steady state shear for 1 wt% CS/0.5 wt% CD solution, before (open) and after (solid) shear-induced thickening. The shear-induced thickening procedure consists of a high rate shearing at 1000 s⁻¹ for 2 min and subsequent a recovering at 0.1 s⁻¹ for 30 min.

The shear induced thickening of CS/CD compound solution can be rationalized in following way. Firstly, with the addition of CD, some CS aggregates turn into CS/CD compound associations induced by hydrogen bonding, but some CS hydrophobic aggregates still remain unchanged, as shown in Fig. 4c. Strong shear deformation destroys the whole associations especially the CS hydrophobic aggregates in the original solution, and induces the alignment and stretching of semirigid chains along the shear flow.³⁵ Both the alignment and stretching of macromolecules result in more sites available for intermolecular junctions through hydrogen bonds. As a result, more hydrogen bonds between CS and CD are formed in solution to form larger and stronger interconnected three-dimensional CS/CD associations or networks, which causes an increased initial viscosity. It had been suggested that the network means the formation of the staggered ladder-like structure.⁵¹ The larger-scale intermolecular hydrogen bonding associations among CS and CD molecules give rise to the thickening.

Conclusions

The compound solution of CS/β-CD presents an obvious multiple shear thinning and a subsequent shear induced thickening (SIT) behavior, depending on the mass ratio of CD and CS. Too high or too low CD concentration is unfavorable for shear induced self-thickening of CS/CD compound solution when the CS concentration is fixed. According to results of steady and oscillatory shear flows, the introduction of CD could cause the initial hydrophobic association of CS into aggregates even a weak network structure mainly induced by hydrogen bonding and in turn the CS/CD compound systems demonstrate distinct rheological properties. TEM observations of CS/CD compound solution show that adding a small quantity of CD slightly changes morphology of CS aggregates, but has little effect on the size of aggregates. With CD concentration increasing, larger CS–CD aggregates induced by hydrogen bonding formed, resulting in the increase of viscosity and modulus. Compared with the original one, the increased viscosity for compound solution of 1 wt% CS/0.5 wt% CD after strong shear should mainly owe to that the formation of lager and irregular hydrogen-bonded aggregates incline to connect into large pieces instead of the isolated spheres. The addition of hydrogen bond breaker, AcONH₄, can obviously decreases the SIT behavior of CS/CD compound solution due to weakening of CS–CD intermolecular hydrogen bonding, reflecting that CS–CD intermolecular hydrogen bonds play a main role in the shear-induced thickening progress.

Acknowledgements

This work was supported by National Nature Science Foundation of China (No. 51473145) and Zhejiang Provincial Natural Science Foundation of China (No. R16E030003).

References

1. W. C. Yount, D. M. Loveless and S. L. Craig, J. Am. Chem. Soc., 2005, 127, 14488–14496.
2. D. Xu, J. L. Hawk, D. M. Loveless, S. L. Jeon and S. L. Craig, Macromolecules, 2010, 43, 3556–3565.
3. L. Zhu, Y. Shangguan, Y. Sun, J. Ji and Q. Zheng, Soft Matter, 2010, 6, 5541–5546.
4. I. Tho, A. L. Kjøniksen, B. Nyström and J. Roots, Biomacromolecules, 2003, 4, 1623–1629.
5. A. Cadix, C. Chassenieux, F. Lafuma and F. Lequeux, Macromolecules, 2005, 38, 527–536.
6. Y. Hu, S. Wang and A. Jamieson, Macromolecules, 1995, 28, 1847–1853.
7. J. Wang, L. Benyahia, C. Chassenieux, J. F. Tassin and T. Nicolai, Polymer, 2010, 51, 1964–1971.
8. D. Xu and S. L. Craig, J. Phys. Chem. Lett., 2010, 1, 1683–1686.
9. D. Xu, C. Y. Liu and S. L. Craig, Macromolecules, 2011, 44, 2343–2353.
10. L. Jin, Y. Shangguan, T. Ye, H. Yang, Q. An and Q. Zheng, Soft Matter, 2013, 9, 1835–1843.
11. L. Jin, Y. Tan, Y. Shangguan, Y. Lin, B. Xu, Q. Wu and Q. Zheng, J. Phys. Chem. B, 2013, 117, 15111–15121.
12. K. Tam, R. Jenkins, M. Winnik and D. Bassett, Macromolecules, 1998, 31, 4149–4159.
13. V. Tirtaatmadja, K. Tam and R. Jenkins, Macromolecules, 1997, 30, 1426–1433.
14. A. Lele, A. Shedge, M. Badiger, P. Wadgaonkar and C. Chassenieux, Macromolecules, 2010, 43, 10055–10063.
15. N. Bo, A. L. Kjøniksen, N. Beheshti, A. Maleki, K. Zhu, K. D. Knudsen, R. Pamies, J. G. H. Cifre and J. G. D. L. Torre, Adv. Colloid Interface Sci., 2009, 158, 108–118.
16. Z. Liu, A. Maleki, K. Zhu, A. L. Kjøniksen and N. Bo, J. Phys. Chem. B, 2008, 112, 1082–1089.
17. D. C. Boris and R. H. Colby, Macromolecules, 1998, 31, 5746–5755.
18. J. Hui, T. Prasad, J. R. Reynolds and K. S. Schanze, Angew. Chem., Int. Ed., 2009, 48, 4300–4316.
19. G. Petzold, H. M. Buchhammer and K. Lunkwitz, Colloids Surf., A, 1996, 119, 87–92.
20. G. H. Luo, Y. G. Shangguan and Q. Zheng, J. Mater. Sci. Eng., 2016 DOI:10.15234/j.cnki.issn 1673-2812.2016.06.031.
21. N. Bo, A. L. Kjøniksen, N. Beheshti, K. Zhu and K. D. Knudsen, Soft Matter, 2009, 5, 1328–1339.
22. K. Hayakawa and J. C. T. Kwak, J. Phys. Chem. B, 2002, 86, 3866–3870.
23. A. Bhattacharyya, F. Monroy, D. Langevin and J. F. Argillier, Langmuir, 2000, 16, 8727–8732.
24. J. Nirmesh, T. Siwar, G. Samuel, M. L. Daragh, L. Dominique, L. Pierre and T. Mireille, Langmuir, 2004, 20, 8496–8503.
25. B. A. Noskov, G. Loglio and R. Miller, Adv. Colloid Interface Sci., 2011, 168, 179–197.
26. Y. V. Shulevich, G. Petzold, A. V. Navrotskii and I. A. Novakov, Colloids Surf., A, 2012, 415, 148–152.
27. R. Petkova, S. Tcholakova and N. D. Denkov, Langmuir, 2012, 28, 4996–5009.
28. R. Ahmed, M. S. Hsiao, Y. Matsuura, N. Houbenov, C. F. J. Faul and I. Manners, Soft Matter, 2011, 7, 10462–10471.
29. K. Kogej, Adv. Colloid Interface Sci., 2010, 158, 68–83.
30. M. Weickenmeier, G. Wenz and J. Huff, Astron. Astrophys., 1997, 18, 1117–1123.
31. X. Guo, A. A. Abdala, B. L. May, S. F. Lincoln, S. A. Khan and R. K. Prud'Homme, Macromolecules, 2005, 38, 3037–3040.
32. A. Martínez, E. Chornet and D. Rodrigue, J. Texture Stud., 2004, 35, 53–74.
33. F. Müller-Plathe and W. F. V. Gunsteren, Polymer, 1997, 38, 2259–2268.
34. R. Buscall, I. J. Mcgowan and C. A. Mumme-Young, Faraday Discuss. Chem. Soc., 1990, 90, 115–127.
35. A. L. Kjøniksen, M. Hiorth, J. Roots and N. Bo, J. Phys. Chem. B, 2003, 107, 6324–6328.
36. M. Rubinstein, R. H. Colby and A. V. Dobrynin, Phys. Rev. Lett., 1994, 73, 2776–2779.
37. L. Leibler, M. Rubinstein and R. H. Colby, Macromolecules, 1991, 24, 4701–4707.
38. A. Martínez-Ruvalcaba, E. Chornet and D. Rodrigue, Appl. Rheol., 2004, 14, 140–147.
39. W. M. Kulicke and R. S. Porter, Rheol. Acta, 1980, 19, 601–605.
40. R. G. Larson, The Structure and Rheology of Complex Fluids, Oxford University Press, New York, 1999, p. 223.
41. E. V. Korchagina and O. E. Philippova, Langmuir, 2012, 28, 7880–7888.
42. L. Liu, J. P. Yang, X. J. Ju, R. Xie, Y. M. Liu, W. Wang, J. J. Zhang, C. H. Niu and L. Y. Chu, Soft Matter, 2011, 7, 4821–4827.
43. H. Rehage and H. Hoffmann, J. Phys. Chem., 2010, 92, 4712–4719.
44. O. E. Philippova, E. V. Korchagina, E. V. Volkov, V. A. Smirnov, A. R. Khokhlov and M. Rinaudo, Carbohydr. Polym., 2012, 87, 687–694.
45. O. E. Philippova, E. V. Volkov, N. L. Sitnikova, A. R. Khokhlov, J. Desbrieres and M. Rinaudo, Biomacromolecules, 2001, 2, 483–490.
46. T. Ouchi, H. Nishizawa and Y. Ohya, Polymer, 1998, 39, 5171–5175.
47. Y. He, B. Zhu and Y. Inoue, Prog. Polym. Sci., 2004, 29, 1021–1051.
48. C. Schatz, C. Viton, T. Delair, C. Pichot and A. Domard, Biomacromolecules, 2003, 4, 641–648.
49. G. Lamarque, A. Christophe Viton and A. Domard, Biomacromolecules, 2004, 5, 1899–1907.
50. H. C. Lee and D. A. Brant, Macromolecules, 2002, 35, 2223–2234.
51. A. K. Lele and R. A. Mashelkar, J. Non-Newtonian Fluid Mech., 1998, 75, 99–115.

---

Footnote

† Electronic supplementary information (ESI) available: Complementary TEM observations of pure CD samples and rheological results of CS/CD compound samples. See DOI: 10.1039/c6ra24608g

---