**3. Results and Discussion**

### *3.1. Synthesis and Characterization of p(NiPAm-co-HEMA) Microgels*

The p(NiPAm-co-HEMA) microgels were synthesized by radical precipitation polymerization as illustrated in Figure 1a [23]. We simply mixed two monomers having unsaturated hydrocarbon groups with a crosslinker. Thus, it is assumed that the polymeric structure of the copolymer is random. For Fourier-transform infrared spectroscopy (FTIR) measurement, p(NiPAm-co-HEMA) microgel was freeze-dried to obtain a powdery sample. The FTIR data in Figure 1b confirms the successful synthesis of p(NiPAm-co-HEMA) in

accordance with the specified mole ratio as described in the experimental section. The absorption peak at 3275 cm<sup>−</sup><sup>1</sup> and 1640 cm<sup>−</sup><sup>1</sup> are assigned to stretching vibration of the amino group (N–H) and the amide I groups (C=O) in NiPAM, respectively [24]. The peak at 1734 cm<sup>−</sup><sup>1</sup> is assigned to the stretching vibration of the carbonyl group (C=O) in HEMA [17]. As the molar ratio of HEMA increases so does the peak intensity at 1734 cm<sup>−</sup>1, while it dwindles at 1640 cm<sup>−</sup>1. Normalized FTIR data for comparison of NiPAm and HEMA peak ratio of p(NiPAm-co-HEMA) microgels showed an increase of carbonyl group signal as the content of HEMA increased. For the 5:5 sample, however, the intensity of the carbonyl group signal decreased because the microgel was not successfully synthesized. Zeta potential of microgel dispersion with varying NiPAm:HEMA molar ratio was measured at room temperature. For a control sample, p(NiPAm) microgel dispersion, the value of zeta potential was at the isoelectric point with a slightly negative value. It showed a decreasing trend in values as the content of HEMA increased.

**Figure 1.** (**a**) Schematic illustration for the synthesis of p(NiPAm-co-HEMA) microgels. (**b**) Fourier-transform infrared spectroscopy (FTIR) data for p(HEMA) polymer, p(NiPAm), and p(NiPAm-co-HEMA) microgels with varying molar ratio between N-isopropylacrylamide (NiPAm) and 2-hydroxyethyl methacrylate (HEMA). For comparison of NiPAm and HEMA signal ratio among samples, normalized FTIR data are plotted for p(NiPAm-co-HEMA) microgels (right). (**c**) Zeta potential of p(NiPAm) and p(NiPAm-co-HEMA) microgels with varying NiPAm:HEMA molar ratio. (**d**–**f**) Scanning electron microscope (SEM) images of p(NiPAm-co-HEMA) microgels with molar ratios between NiPAm and HEMA of (**d**) 9:1, (**e**) 7:3, and (**f**) 5:5, respectively. Scale bar in (**d**–**f**) represents 100 nm. (**g**) Effect of molar composition of HEMA to particle diameter.

The average diameter of p(NiPAm-co-HEMA) microgels shows a degree of swelling. Scanning electron microscope (SEM) images of the freeze-dried p(NiPAm-co-HEMA) microgels in Figure 1d–f indicate uniformly sized, spherical microgels. It was confirmed that the diameter of microgels increased as the composition of HEMA increased, Figure 1g. The increment of particle size is attributed to the good water uptake capability of HEMA. As a result, sparse polymeric networks were formed during the radical polymerization as the amount of HEMA increased, resulting in the formation of larger-, and mechanically weaker microgels. Indeed, when we increased the molar ratio of HEMA to more than 0.5, the structural integrity of p(NIPAM-co-HEMA) microgels could not be maintained. As a result, bulk p(NIPAM-co-HEMA) polymer cakes were observed as gels precipitated. Based on the above results, further studies regarding investigations of colloidal stability and their thermogelling behaviors were mainly focused on the characteristics of p(NiPAm-HEMA) microgel dispersions with NiPAm to HEMA molar ratios of 9:1, 7:3, and 5:5. In addition, further experiments were conducted in neutral pH conditions because colloidal stability under acidic and alkalic conditions was unstable, as shown in Figure S1.

### *3.2. Effect of Salt on the Stability of p(NiPAm-co-HEMA) Microgel Dispersion*

A notable difference between p(NiPAm-co-HEMA) and p(NiPAm) microgel dispersion is that the former requires the addition of salt and appropriate temperature for successful gelation, Figure 2a. This is because the electrostatic repulsion between the microgels is increased by the hydroxyl group of HEMA. In the presence of HEMA, salt ions are required to screen the electrostatic double layers on the microgels to prompt the formation of a particle network that leads to gelation upon thermal heating. Because adding salt in conventional colloidal dispersion causes destabilization of the system, we first investigated the effect of salt on the stability of p(NiPAm-co-HEMA) microgel dispersions.

**Figure 2.** (**a**) Photograph images of thermogelling p(NiPAm-co-HEMA) microgel dispersion upon different temperature conditions. In the absence of salt, the dispersion remained at the sol state for both 25 ◦C and 33 ◦C (top panel). In the presence of 0.33 M of sodium chloride (NaCl), microgels dispersion showed volume phase transition from sol to gel state at 33 ◦C (bottom panel). (**b**–**d**) The transmittance of p(NiPAm-co-HEMA) microgel dispersion with molar ratios between NiPAm and HEMA of (**b**) 9:1, (**c**) 7:3, and (**d**) 5:5, respectively, with respect to NaCl concentration. The bulk aggregation of microgels by the strong ionic strength of solution was expressed in yellow background. (**e**) Summary of dispersion quality of p(NiPAm-co-HEMA) microgels. The microgel showed a well-dispersed (white background) phase, micro aggregation (blue background) phase, and bulk aggregation (yellow background) phase in accordance with NiPAm: HEMA molar ratios and NaCl concentrations. (**f**) Schematic illustration of deswelling of microgel with the increase of the salt concentration.

The stability of p(NiPAm-co-HEMA) microgel dispersion was investigated by observing the turbidity of the dispersion via a UV–visible spectrometer (SHIMAZU, Kyoto, Japan). 0.04 wt% of p(NiPAm-co-HEMA) microgel dispersion with varying NiPAm to HEMA molar ratios were prepared. As shown in Figure 2b–d, the transmittance of the p(NiPAm-co-HEMA) microgels composed of NiPAm to HEMA molar ratios of 9:1, 7:3, and 5:5 at different NaCl concentrations were measured. The transmittance for each sample without NaCl was measured as 59.7%, 34.7%, and 23.7% for 9:1, 7:3, and 5:5 samples, respectively, which showed a decreasing trend as the amount of HEMA increased. This is because the particle size-dependent Mie scattering was dominant for particles with a size similar to the wavelength of the incident light. In all cases, transmittance was slightly increased at low salt concentration and then started to decrease until it reached the critical salt concentration. Then, a dramatic increase in transmittance was observed. These results can be explained by the change of Mie scattering and Debye length of the system along with the change of the salt concentration. When the salt is added to the system, osmotic pressure applied to the microgel causes the deswelling of the microgels [16,25]. It

results in the decrease of the microgel size and thereby decrease of Mie scattering. Therefore, the transmittance of dispersion was increased. As we stated above, an increase of the salt concentration in colloidal dispersion also causes the screening of electric double layer, which results in the decrease of Debye length of microgels. When the Debye length reaches below the critical value for maintaining a stable dispersion, Van der Waals force causes aggregation of microgels followed by sedimentation. Here, a decrease and dramatic increase of transmittance can be explained by microaggregation and bulk aggregation (i.e., sedimentation), respectively. When the dispersion is at the state of microaggregation, the effective size of microgels can be considered to be larger and it results in the decrease of transmittance. Then, a sudden increase of transmittance at bulk aggregation is observed due to the sedimentation. Therefore, a critical salt concentration can be regarded as a point at which bulk aggregation of microgels occurs. The phase behaviors of p(NiPAmco-HEMA) microgel dispersion with varying NiPAm:HEMA compositions in Figure 2e revealed that the higher the amount of HEMA is, the more the salt concentration-initiated microaggregation occurs. By reflecting on the trends of transmittance for each sample, it is expected that the content of HEMA contributed to the increase of Debye length of microgels, which also agreed with the previous results. The comprehensive behavior of microgel dispersion is schematically described in Figure 2f.

### *3.3. Thermogelling Behaviors of p(NiPAm-co-HEMA) Microgel Dispersion*

We characterized the thermogelling behaviors of p(NiPAm-co-HEMA) microgel dispersion in accordance with the concentration of NaCl, concentration of microgels and composition of NiPAm and HEMA by the SAOS test results measured with a rotational rheometer (refer to the experimental section for the detailed conditions). Following the conventional method of measuring thermogelling behaviors, the gelation temperature at which the colloidal dispersion changes from sol to gel was determined by observing the cross-over behavior of the storage (G) and loss (G") moduli, as shown in Figure 3 [26].

**Figure 3.** Evolution of storage (G) and loss (G") moduli of the p(NiPAm-co-HEMA) microgel dispersion from 20 ◦C to 40 ◦C. The procedure was conducted under fixed stress of 0.05 Pa and a frequency of 0.1 Hz ( *ω* = 0.63 rad/s). (**a**) Change in dynamic modulus of the polymer at different salt concentrations. The concentration of polymer and molar ratio between NiPAm and HEMA were kept constant at 2.7 wt% and 7:3, respectively. (**b**) Change in dynamic modulus at different copolymer concentration (*cm*). The concentration of NaCl and molar ratio between NiPAm and HEMA were kept constant at 0.17 M and 7:3, respectively. Log plot between G and microgel concentration showed a power-law relationship with a slope of 2.06 (inset). (**c**) Change in dynamic modulus at different molar ratios between NiPAm and HEMA. The concentrations of polymer and NaCl were kept constant at 2.7 wt% and 0.33 M, respectively.

### 3.3.1. Effect of Salt Concentration

From the prior investigation of the dispersion stability, it can be concluded that the microgels under the speculation are stable up to 0.5 M NaCl. Thus, the effect of salt concentration for thermogelling behaviors was investigated for 2.7 wt% of 7:3 molar ratio of p(NiPAm-co-HEMA) microgel dispersion with NaCl concentrations of 0.17 M, 0.33 M, and 0.5 M. As shown in Figure 3a, larger G" than G was observed initially for all the samples with no significant fluctuation, which proves a liquid-like sol state. As temperature kept increasing, a crossover between G and G" occurred at 31.8 ◦C, 29.3 ◦C, and 27.3 ◦C for samples with salt concentrations of 0.17 M, 0.33 M, and 0.5 M, respectively. The result shows an aggregation of p(NiPAm-co-HEMA) microgels and thereby gelation of the system. Based on the findings, it was concluded that the gelation temperature decreases as the salt concentration increases. Decreasing trends of gelation temperature can be attributed to the screening of electrostatic repulsion by the salt addition, which results in the decrease of Debye length of p(NiPAm-co-HEMA) microgels. In addition, it is notable that the plateau of the G and G" after gelation was formed at a similar magnitude for all the samples. It implies that the mechanical strength of gels in different salt concentrations was not affected.

## 3.3.2. Effect of Microgel Concentration

Effect of microgel concentration was conducted by comparing 1.3 wt%, 2.7 wt%, and 4.0 wt% of 7:3 molar ratio of p(NiPAm-co-HEMA) microgel dispersion at a fixed NaCl concentration of 0.33 M. As shown in Figure 3b, gelation temperature where the crossover between G and G" occurs did not change by the concentration of p(NiPAm-co-HEMA) microgels. On the other hand, the mechanical strength of gels was affected by the concentration. When the temperature exceeds the gelation temperature, the magnitude of both G and G" at plateau were proportional to the microgel concentration. It is noted that the magnitude of G and G" at sol state (i.e., below gelation temperature) only showed little increases with increasing the concentration of p(NiPAm-co-HEMA) microgels. The current results sugges<sup>t</sup> that the viscous properties ( ∼=G"/ *ω*) at the sol states of the samples are all close to that of pure water, i.e., G"~O(10−3) [27], which are expected to be proportional to the volume fractions of the microgel. However, the change in the viscous properties with the increasing microgel concentration is too small to be captured within the sensitivity limit of the rotational rheometer. On the other hand, it is clear that their gel strengths are significantly affected by the formation of denser physical networks between microgels during the gelation process as the microgel concentration increases [26]. The relationship between *cm* and gel strength (G) clearly shows a powerlaw relationship (inset of Figure 3b), which is consistent with the previous studies on the gelation of particulate systems [28]. The power-law exponent is 1 when the particle volume fraction is close to the gelation particle volume fraction ( *φg*), which increases to 3–5 as the volume fraction significantly deviates from *φg* [28]. Therefore, it can be concluded that the current particle volume fraction ( ≈*kcm*) range is not far from *φg* for p(NiPAm-co-HEMA) microgels, in which *k* is a proportionality constant between the particle concentration and volume fraction.
