Tunable Viscoelastic Properties of Sodium Polyacrylate Solution via CO2-Responsive Switchable Water

Upon stimulus by CO2, CO2-switchable viscoelastic fluids experience a deliberate transition between non-viscous and highly viscous solution states. Despite attracting considerable recent attention, most such fluids have not been applied at a large- scale due to their high costs and/or complex synthesis processes. Here, we report the development of CO2-switchable viscoelastic fluids using commercially available sodium polyacrylate (NaPAA) and N,N-dimethyl ethanol amine (DMEA)-based switchable water. Upon bubbling CO2, into the solutions under study, DMEA molecules are protonated to generate quaternary ammonium salts, resulting in pronounced decreases in solutions viscosity and elasticity due to the influence of increased ionic strength on NaPAA molecular conformations. Upon removal of CO2 via introduction of N2, quaternary salts are deprotonated to tertiary amines, allowing recovery of fluid viscosity and elasticity to near the initial state. This work provides a simple approach to fabricating CO2-switchable viscoelastic fluids, widening the potential use of CO2 in stimuli-responsive applications.

So far, CO 2 -responsive viscoelastic systems based on surfactants and/or polymers have been developed. Surfactant wormlike micelles (WLMs) are long flexible aggregates that can entangle to form three-dimensional networks that impart remarkable viscoelastic properties to the bulk solution. Thus, the reversible generation and destruction of wormlike micelles via external stimulations can realize switchable viscoelastic fluids [23]. Feng and

Rheological Properties of NaPAA Aqueous Solutions and Their CO 2 -Responsive Behavior
Owing to its strong thickening power and easy availability, NaPAA is widely used in various industries as rheology modifier. The NaPAA used in this work was a commercial product with M W of 4 × 10 6 -5 × 10 6 g·mol −1 , and its thickening ability was investigated. A series of aqueous solutions with different NaPAA concentrations were prepared, and the viscosity-shear rate curves ( Figure S1) of each were measured to obtain the zero-shear viscosity (η 0 ), which is the value when the shear rate tends to zero. Figure 1a shows a plot of η 0 at varying concentrations of NaPAA in pure water. One can find that the curve has been divided into three concentration regimes with two clear breakpoints. At low concentrations, the NaPAA aqueous solutions exhibited Newtonian fluids properties, and the η 0 was found to be very close to that of the solvent, classifying these fluids as dilute [31]. The critical overlap concentrations (C*), at which the solution transitions from dilute to semidilute behavior, were determined from the concentrations at which the measured solution viscosity began to abruptly increase. Herein, C* (≈ 0.0001 wt%) falls in the second area shown in Figure 1a, in which the solution viscosity was found to increase with the increment of NaPAA concentration according to a power law with an exponent of 0.52, approaching a theoretically predicted value (0.5) for semidilute, unentangled polyelectrolyte solutions in previous studies [32][33][34]. In the third region of Figure 1a, η 0 of the fluids increased more dramatically with concentration than it did in semidilute, unentangled regime, demonstrating a power law with an exponent of 1.77, near the expected power index (1.5) of semidilute, entangled polyelectrolyte fluids [33,34]. Based on the semidilute unentangled and entangled regime scaling, the second turning point was defined as the critical entanglement concentration, i.e., C e (≈ 0.0007 wt%), as labeled in Figure 1a. A series of aqueous solutions with different NaPAA concentrations were prepared, and the viscosity-shear rate curves ( Figure S1) of each were measured to obtain the zero-shear viscosity (η0), which is the value when the shear rate tends to zero. Figure 1a shows a plot of η0 at varying concentrations of NaPAA in pure water. One can find that the curve has been divided into three concentration regimes with two clear breakpoints. At low concentrations, the NaPAA aqueous solutions exhibited Newtonian fluids properties, and the η0 was found to be very close to that of the solvent, classifying these fluids as dilute [31]. The critical overlap concentrations (C*), at which the solution transitions from dilute to semidilute behavior, were determined from the concentrations at which the measured solution viscosity began to abruptly increase. Herein, C* (≈ 0.0001 wt%) falls in the second area shown in Figure 1a, in which the solution viscosity was found to increase with the increment of NaPAA concentration according to a power law with an exponent of 0.52, approaching a theoretically predicted value (0.5) for semidilute, unentangled polyelectrolyte solutions in previous studies [32][33][34]. In the third region of Figure 1a, η0 of the fluids increased more dramatically with concentration than it did in semidilute, unentangled regime, demonstrating a power law with an exponent of 1.77, near the expected power index (1.5) of semidilute, entangled polyelectrolyte fluids [33,34]. Based on the semidilute unentangled and entangled regime scaling, the second turning point was defined as the critical entanglement concentration, i.e., Ce (≈ 0.0007 wt%), as labeled in Figure 1a. To investigate the influence of CO2 on the NaPAA fluid, two different concentrations of NaPAA, 0.65 and 1.30 wt%, representing solutions with high viscosity and elasticity in the semidilute entangled regime, were selected. The NaPAA aqueous solutions with concentrations of 0.65 wt% and 1.30 wt% were prepared, and their rheological behaviors were examined at 25 °C under steady and dynamic shear conditions. Figure S2 shows the viscosity-shear rate curves of the samples containing 0.65 wt% and 1.30 wt% NaPAA in the absence and presence of CO2. It can be found that the fluids showed Newtonian behavior with high, near-constant viscosities at low shear rates, then exhibited shear thinning properties in which the viscosities sharply decreased with the increment in shear rate, which was consistent with typical rheological properties of viscoelastic fluids [3,9]. After bubbling CO2 for 10 min until the conductivity of the solutions stabilized, the viscosity of each fluid underwent an apparent decrease over the entire range of tested shear rates, and the lower the shear rate, the larger the decrement in viscosity. To investigate the influence of CO 2 on the NaPAA fluid, two different concentrations of NaPAA, 0.65 and 1.30 wt%, representing solutions with high viscosity and elasticity in the semidilute entangled regime, were selected. The NaPAA aqueous solutions with concentrations of 0.65 wt% and 1.30 wt% were prepared, and their rheological behaviors were examined at 25 • C under steady and dynamic shear conditions. Figure S2 shows the viscosity-shear rate curves of the samples containing 0.65 wt% and 1.30 wt% NaPAA in the absence and presence of CO 2 . It can be found that the fluids showed Newtonian behavior with high, near-constant viscosities at low shear rates, then exhibited shear thinning properties in which the viscosities sharply decreased with the increment in shear rate, which was consistent with typical rheological properties of viscoelastic fluids [3,9]. After bubbling CO 2 for 10 min until the conductivity of the solutions stabilized, the viscosity of each fluid underwent an apparent decrease over the entire range of tested shear rates, and the lower the shear rate, the larger the decrement in viscosity. N 2 was subsequently introduced into the solutions at 60 • C to remove CO 2 , and the viscosities of both the 0.65 wt% and 1.30 wt% NaPAA solution were partially recovered. Figure 1b presents the variation in viscosity that arose for each fluid at a shear rate of 10 s −1 upon alternately bubbling CO 2 and N 2 ; both solutions exhibited a clear CO 2responsive change in viscosity. For the 0.65 wt% NaPAA solution, its viscosity was initially 1821 mPa·s, and dropped sharply to 1056 mPa·s upon CO 2 treatment, then increased to 1553 mPa·s when N 2 was bubbled in. In contrast, the initial viscosity of the 1.30 wt% NaPAA in pure water decreased from 4314 mPa·s to 2956 mPa·s after treating with CO 2 , then recovered to 3962 mPa·s when bubbling N 2 . For the 0.65 wt% and 1.30 wt% NaPAA aqueous solutions at a shear rate of 10 s −1 , the degrees of viscosity recovery after alternating CO 2 /N 2 treatments were 85% and 92%, respectively. These results clearly indicate that alternating addition and removal of CO 2 facilitated a reversible change in the viscosities of the NaPAA aqueous solutions.
The effect of CO 2 on the viscoelasticity of the NaPAA aqueous solutions was then investigated via oscillatory-shear measurements. Before measurements, the linear viscoelastic region of the NaPAA aqueous solutions was confirmed through a strain sweep test. As shown in Figure S3, at strains lower than 40 % at 25 • C, the fluids exhibited linear viscoelastic behavior, defining the region in which the solutions could retain a stable structure without being destroyed [35]. Here, storage modulus (G ) characterized elastic behavior of the NaPAA solutions, while loss modulus (G") represented its viscous properties. Under low strain, G > G", for both samples, indicating that they exhibited solid gel-like behaviors. Contrarily, in the high strain regime, G < G", demonstrating that the both samples showed fluid-like properties. Therefore, frequency sweep curves for the 0.65 wt% and 1.30 wt% NaPAA aqueous solutions were obtained at a strain of 10 %, as shown in Figure 2.
N2 was subsequently introduced into the solutions at 60 °C to remove CO2, and the viscosities of both the 0.65 wt% and 1.30 wt% NaPAA solution were partially recovered. Figure 1b presents the variation in viscosity that arose for each fluid at a shear rate of 10 s −1 upon alternately bubbling CO2 and N2; both solutions exhibited a clear CO2-responsive change in viscosity. For the 0.65 wt% NaPAA solution, its viscosity was initially 1821 mPa·s, and dropped sharply to 1056 mPa·s upon CO2 treatment, then increased to 1553 mPa·s when N2 was bubbled in. In contrast, the initial viscosity of the 1.30 wt% NaPAA in pure water decreased from 4314 mPa·s to 2956 mPa·s after treating with CO2, then recovered to 3962 mPa·s when bubbling N2. For the 0.65 wt% and 1.30 wt% NaPAA aqueous solutions at a shear rate of 10 s −1 , the degrees of viscosity recovery after alternating CO2/N2 treatments were 85% and 92%, respectively. These results clearly indicate that alternating addition and removal of CO2 facilitated a reversible change in the viscosities of the Na-PAA aqueous solutions.
The effect of CO2 on the viscoelasticity of the NaPAA aqueous solutions was then investigated via oscillatory-shear measurements. Before measurements, the linear viscoelastic region of the NaPAA aqueous solutions was confirmed through a strain sweep test. As shown in Figure S3, at strains lower than 40 % at 25 °C, the fluids exhibited linear viscoelastic behavior, defining the region in which the solutions could retain a stable structure without being destroyed [35]. Here, storage modulus (G′) characterized elastic behavior of the NaPAA solutions, while loss modulus (G″) represented its viscous properties. Under low strain, G′ > G″, for both samples, indicating that they exhibited solid gellike behaviors. Contrarily, in the high strain regime, G′ < G″, demonstrating that the both samples showed fluid-like properties. Therefore, frequency sweep curves for the 0.65 wt% and 1.30 wt% NaPAA aqueous solutions were obtained at a strain of 10 %, as shown in Figure 2. Regardless of whether CO 2 was present, the solutions displayed classic viscoelastic character, i.e., they showed a crossover of G and G" within the tested range of strain frequencies. The maximum relaxation time (τ R ) for a viscoelastic fluid is generally defined as the inverse of this intersectant frequency (ω c ) [36]; thus, the fluid's response can be divided into two regimes based on τ R . At low frequency (ω << ω c ), G < G", indicating that samples possessed a viscous behavior; while at high frequency (ω >> ω c ), G > G", the fluids behaved an elastic property. Additionally, for the 0.65 wt% NaPAA aqueous solution, the point of intersection shifted to a higher frequency after CO 2 treatment, corresponding to a decrease in τ R from~25 s to~1.45 s, then recovered to a lower frequency upon bubbling N 2 , i.e., τ R increased to~5.26 s. In comparison, τ R of the 1.30 wt% NaPAA aqueous solution decreased from~50 s to~5.56 s after bubbling CO 2 , and then increased to~8.33 s with the introduction of N 2 . These results clearly suggest that the viscoelastic character of NaPAA aqueous solution can be reversibly tuned through alternating introduction of CO 2 and N 2 . Moreover, both G and G" were dependent on frequency and much lower in the presence of CO 2 , confirming that the introduction of CO 2 weakened the strength of the fluids.

Effect of DMEA CO 2 -Switchable Water on Rheological Properties of NaPAA Solutions
DMEA contains a tertiary amino group that can be well dissolved in water and protonated by CO 2 to produce ammonium bicarbonate salt, yielding a large increase in ionic strength of the aqueous solution. The protonated product can be converted back into to the original amine by bubbling N 2 , showing good reversibility [22,37]. For these reasons, DMEA was chosen to prepare CO 2 -switchable water, and the effect of both DEMA concentration and the introduction of CO 2 on the rheological behavior of NaPAA aqueous solution was investigated.
The influence of different DMEA concentrations on the steady rheology properties of 0.65 wt% and 1.30 wt% NaPAA solutions in the presence and absence of CO 2 is shown in Figures S4 and S5, respectively. One can find that the initial solution showed the largest viscosities in low shear rate region, while the CO 2 -treated solutions exhibited the lowest viscosities at the same shear rate. Values of η 0 were obtained from Figures S4 and S5, and further plotted against DMEA concentrations. As displayed in Figure 3a, η 0 of the 0.65% NaPAA aqueous solution without the addition of DMEA was initially 303. the inverse of this intersectant frequency (ωc) [36]; thus, the fluid's response can be divided into two regimes based on τR. At low frequency (ω << ωc), G′ < G″, indicating that samples possessed a viscous behavior; while at high frequency (ω >> ωc), G′ > G″, the fluids behaved an elastic property. Additionally, for the 0.65 wt% NaPAA aqueous solution, the point of intersection shifted to a higher frequency after CO2 treatment, corresponding to a decrease in τR from ~25 s to ~1.45 s, then recovered to a lower frequency upon bubbling N2, i.e., τR increased to ~5.26 s. In comparison, τR of the 1.30 wt% NaPAA aqueous solution decreased from ~50 s to ~5.56 s after bubbling CO2, and then increased to ~8.33 s with the introduction of N2. These results clearly suggest that the viscoelastic character of NaPAA aqueous solution can be reversibly tuned through alternating introduction of CO2 and N2.
Moreover, both G′ and G″ were dependent on frequency and much lower in the presence of CO2, confirming that the introduction of CO2 weakened the strength of the fluids.

Effect of DMEA CO2-Switchable Water on Rheological Properties of NaPAA Solutions
DMEA contains a tertiary amino group that can be well dissolved in water and protonated by CO2 to produce ammonium bicarbonate salt, yielding a large increase in ionic strength of the aqueous solution. The protonated product can be converted back into to the original amine by bubbling N2, showing good reversibility [22,37]. For these reasons, DMEA was chosen to prepare CO2-switchable water, and the effect of both DEMA concentration and the introduction of CO2 on the rheological behavior of NaPAA aqueous solution was investigated.
The influence of different DMEA concentrations on the steady rheology properties of 0.65 wt% and 1.30 wt% NaPAA solutions in the presence and absence of CO2 is shown in Figures S4 and S5, respectively. One can find that the initial solution showed the largest viscosities in low shear rate region, while the CO2-treated solutions exhibited the lowest viscosities at the same shear rate. Values of η0 were obtained from Figures S4 and S5, and further plotted against DMEA concentrations. As displayed in Figure 3a, η0 of the 0.65% NaPAA aqueous solution without the addition of DMEA was initially 303.1 Pa·s −1 , then decreased to 55.3 Pa·s −1 in the presence of CO2, and recovered to 273.1 Pa·s −1 after introducing N2. Upon adding DMEA to the 0.65% NaPAA aqueous solution, the initial η 0 of the 0.65wt% NaPAA solution decreased with the increment of DMEA, which can be interpreted as the introduction of DMEA increased the ionic strength of the mixed solution, weakening electrostatic repulsions between NaPAA chains. While bubbling CO 2 , tertiary amino groups in DMEA reacted with CO 2 to produce ammonium bicarbonate salts, which further remarkably increased the ionic strength of the solution, resulting in a sharp decrease in η 0 of the 0.65 wt% NaPAA solutions. With respect to the CO 2 -treated solutions, varying DMEA concentration from 0.6 wt% to 4.8 wt% led to decreases in η 0 from 12.6 Pa·s −1 to 2.6 Pa·s −1 . Bubbling N 2 to remove CO 2 led to partial recovery of the initial solution viscosity across all concentrations of DMEA, as shown in Figure 3a.
To investigate the effect of CO 2 on η 0 in the presence of DMEA, a loss ratio index, i.e., the ratio of lost η 0 in CO 2 -treated solution to its initial value, was introduced, as displayed in Figure 3b. The data indicates that the loss ratio of η 0 upon introduction of CO 2 increased from 94.6 % to 97.9 % with the increment of DMEA, implying that the protonated DMEA could greatly weaken the viscosity of NaPAA solution, demonstrating behavior similar to that of Na + and Ca 2+ in prior studies [26,27]. However, unlike Na + and Ca 2+ , the loss of NaPAA solution viscosity induced by protonated DMEA could be reversed to some extent, as shown in Figure 3a. Therefore, to assess the degree of η 0 that occurred after introducing N 2 , a recovery ratio, i.e., the ratio of η 0 of the N 2 -treated solution to its initial value, is also presented in Figure 3b. It can be found that the recovery ratio varied between 8.3 % and 18.2 %, indicating that the DMEA was not completely deprotonated by bubbling N 2 at 60 • C.
The influence of different DMEA concentrations on η 0 of the 1.30 wt% NaPAA solution, in the presence or absence of CO 2 , is displayed in Figure S6. The variation of η 0 with concentrations of DMEA exhibited similar tendencies to that of the 0.65 wt% NaPAA solution. However, the loss ratio in η 0 of 1.30 wt% NaPAA solution was lower than that of 0.65 wt% solution, and the recovery ratio in η 0 of 1.30 wt% NaPAA solution was higher than that of 0.65 wt% solution, indicating that the inter-and intramolecular entanglement of 1.30 wt% NaPAA solution is stronger than that of 0.65 wt% solution. These results prove that the viscosity of NaPAA solutions could be reversibly tuned via the protonation and deprotonation of DMEA.
Then, the influence of DMEA concentration on the dynamic rheological behavior of NaPAA fluids upon CO 2 /N 2 treatment was also investigated. Figure 4 presents frequency sweep curves of 0.65 wt% NaPAA solutions with different DMEA contents. As mentioned previously, the G /G"-crossover point on the frequency sweep curves provides the frequency of the viscous-to-elastic transition corresponding to the maximum relaxation time, τ R . Measured values of τ R are plotted against variations in DMEA concentration upon alternate gas treatments in Figure 5.
Upon adding DMEA to the 0.65% NaPAA aqueous solution, the initial η0 of the 0.65wt% NaPAA solution decreased with the increment of DMEA, which can be interpreted as the introduction of DMEA increased the ionic strength of the mixed solution, weakening electrostatic repulsions between NaPAA chains. While bubbling CO2, tertiary amino groups in DMEA reacted with CO2 to produce ammonium bicarbonate salts, which further remarkably increased the ionic strength of the solution, resulting in a sharp decrease in η0 of the 0.65 wt% NaPAA solutions. With respect to the CO2-treated solutions, varying DMEA concentration from 0.6 wt% to 4.8 wt% led to decreases in η0 from 12.6 Pa·s −1 to 2.6 Pa·s −1 . Bubbling N2 to remove CO2 led to partial recovery of the initial solution viscosity across all concentrations of DMEA, as shown in Figure 3a.
To investigate the effect of CO2 on η0 in the presence of DMEA, a loss ratio index, i.e., the ratio of lost η0 in CO2-treated solution to its initial value, was introduced, as displayed in Figure 3b. The data indicates that the loss ratio of η0 upon introduction of CO2 increased from 94.6 % to 97.9 % with the increment of DMEA, implying that the protonated DMEA could greatly weaken the viscosity of NaPAA solution, demonstrating behavior similar to that of Na + and Ca 2+ in prior studies [26,27]. However, unlike Na + and Ca 2+ , the loss of NaPAA solution viscosity induced by protonated DMEA could be reversed to some extent, as shown in Figure 3a. Therefore, to assess the degree of η0 that occurred after introducing N2, a recovery ratio, i.e., the ratio of η0 of the N2-treated solution to its initial value, is also presented in Figure 3b. It can be found that the recovery ratio varied between 8.3 % and 18.2 %, indicating that the DMEA was not completely deprotonated by bubbling N2 at 60 °C.
The influence of different DMEA concentrations on η0 of the 1.30 wt% NaPAA solution, in the presence or absence of CO2, is displayed in Figure S6. The variation of η0 with concentrations of DMEA exhibited similar tendencies to that of the 0.65 wt% NaPAA solution. However, the loss ratio in η0 of 1.30 wt% NaPAA solution was lower than that of 0.65 wt% solution, and the recovery ratio in η0 of 1.30 wt% NaPAA solution was higher than that of 0.65 wt% solution, indicating that the inter-and intramolecular entanglement of 1.30 wt% NaPAA solution is stronger than that of 0.65 wt% solution. These results prove that the viscosity of NaPAA solutions could be reversibly tuned via the protonation and deprotonation of DMEA.
Then, the influence of DMEA concentration on the dynamic rheological behavior of NaPAA fluids upon CO2/N2 treatment was also investigated. Figure 4 presents frequency sweep curves of 0.65 wt% NaPAA solutions with different DMEA contents. As mentioned previously, the G′/G″-crossover point on the frequency sweep curves provides the frequency of the viscous-to-elastic transition corresponding to the maximum relaxation time, τR. Measured values of τR are plotted against variations in DMEA concentration upon alternate gas treatments in Figure 5.   deprotonated by N2 treatment.
The variation of τR with different DMEA concentrations obtained for the 1.30 wt% NAPAA solution exhibited a similar variation tendency upon alternately bubbling CO2 and N2, as shown in Figure S7 and Figure 5b. These results illustrate that the viscoelasticity of NaPAA solutions could be reversibly adjusted through the incorporation of DMEAbased CO2-switchable water.

Mechanism of CO2-Switchable Viscoelasticity
Rheological characterizations showed that CO2 treatment could decrease the viscosity and weaken the elasticity of NaPAA fluids, which might be ascribe to the reaction between incorporated DMEA and CO2. To elucidate the corresponding mechanism, the pH and conductivity of NaPAA solutions with different DMEA concentrations were continuously monitored during cyclical CO2/N2 treatment at 25 °C.
In general, the basicity of functional groups can be judged based on the dissociation constant (pKaH) of its conjugate acid; the larger the pKaH, the stronger the basicity [20]. The pH titration curves of aqueous solutions of DMEA and NaPAA against HCl solution were obtained to determine the pKaH of DMEA and NaPAA. In a typical titration curve, the pH corresponding to the half equivalence points is taken as the average pKaH [20]. The reaction between DMEA and HCl is exhibited in Scheme 1, and the titration curve of the DMEA solution is shown in Figure 6a, in which the pKaH of DMEA was found to be 9.56. Similarly, the pKaH of NaPAA used in this study was found to be 5.93, as shown in Figure 6b Figure 5a demonstrates that, for the 0.65 wt% NaPAA solutions, the initial τ R gradually decreased from 25.00 s to 9.09 s with the DMEA content increased from 0 to 4.80 wt%. Thus, the strength of the NaPAA fluids was weakened by the addition of DMEA. Upon treatment of CO 2 until conductivity stabilized, DMEA was protonated by CO 2 , resulting in the presence of quaternary ammonium salts in the solution [22,37]. Consequently, τ R of the solutions further declined from 1.45 s to 0.32 s with increasing DMEA concentrations, suggesting that protonated DMEA enhanced CO 2 -induced weaking of the fluid. To remove CO 2 , N 2 was introduced into the mixture, and τ R was partially recovered to its initial value, again proving that the quaternary ammonium salts of DMEA were only partially deprotonated by N 2 treatment.
The variation of τ R with different DMEA concentrations obtained for the 1.30 wt% NAPAA solution exhibited a similar variation tendency upon alternately bubbling CO 2 and N 2 , as shown in Figure S7 and Figure 5b. These results illustrate that the viscoelasticity of NaPAA solutions could be reversibly adjusted through the incorporation of DMEA-based CO 2 -switchable water.

Mechanism of CO 2 -Switchable Viscoelasticity
Rheological characterizations showed that CO 2 treatment could decrease the viscosity and weaken the elasticity of NaPAA fluids, which might be ascribe to the reaction between incorporated DMEA and CO 2 . To elucidate the corresponding mechanism, the pH and conductivity of NaPAA solutions with different DMEA concentrations were continuously monitored during cyclical CO 2 /N 2 treatment at 25 • C.
In general, the basicity of functional groups can be judged based on the dissociation constant (pK aH ) of its conjugate acid; the larger the pK aH , the stronger the basicity [20]. The pH titration curves of aqueous solutions of DMEA and NaPAA against HCl solution were obtained to determine the pK aH of DMEA and NaPAA. In a typical titration curve, the pH corresponding to the half equivalence points is taken as the average pK aH [20]. The reaction between DMEA and HCl is exhibited in Scheme 1, and the titration curve of the DMEA solution is shown in Figure 6a, in which the pK aH of DMEA was found to be 9.56. Similarly, the pK aH of NaPAA used in this study was found to be 5.93, as shown in Figure 6b.  In terms of the pKaH values determined above, the degree of protonation (δ), i.e., the proportion of protonated organ-base, can be calculated using Equations (1) and (2) [20,38].
Upon bubbling CO2, the pH of the 0.65 wt% NaPAA aqueous solution decreased from 9.17 to 5.82, as displayed in Figure 7a. According to Equation (2), the initial δ of NaPAA was ~0.06 mol%. The change in pH that occurred after sparging with CO2 indicates that δ increased to ~56.30 mol%, and thus ~56.24 mol% carboxylate groups were more protonated in the presence of CO2. These results coincide with results of the rheology experiments: viscosities η and relaxation times τR of the 0.65 wt% aqueous solutions dropped sharply after bubbling CO2, revealing that protonation of carboxylate side groups reduced the charge density and corresponding electrostatic repulsions between NaPAA chains to allow a smaller average molecular size than in the initial state. After bubbling N2, the pH of 0.65 wt% NaPAA aqueous solution recovered to 7.99, signifying deprotonation of the carboxylate groups. The pH was not totally restored to its initial value, indicating that 0.86 mol% carboxylate groups were still protonated. This phenomenon also coincides with the prior rheological results, in which the viscosities η and relaxation times τR of NaPAA aqueous solutions were not fully restored to their initial values.  In terms of the pKaH values determined above, the degree of protonation (δ), i.e., the proportion of protonated organ-base, can be calculated using Equations (1) and (2) [20,38].
Upon bubbling CO2, the pH of the 0.65 wt% NaPAA aqueous solution decreased from 9.17 to 5.82, as displayed in Figure 7a. According to Equation (2), the initial δ of NaPAA was ~0.06 mol%. The change in pH that occurred after sparging with CO2 indicates that δ increased to ~56.30 mol%, and thus ~56.24 mol% carboxylate groups were more protonated in the presence of CO2. These results coincide with results of the rheology experiments: viscosities η and relaxation times τR of the 0.65 wt% aqueous solutions dropped sharply after bubbling CO2, revealing that protonation of carboxylate side groups reduced the charge density and corresponding electrostatic repulsions between NaPAA chains to allow a smaller average molecular size than in the initial state. After bubbling N2, the pH of 0.65 wt% NaPAA aqueous solution recovered to 7.99, signifying deprotonation of the carboxylate groups. The pH was not totally restored to its initial value, indicating that 0.86 mol% carboxylate groups were still protonated. This phenomenon also coincides with the prior rheological results, in which the viscosities η and relaxation times τR of NaPAA aqueous solutions were not fully restored to their initial values. In terms of the pK aH values determined above, the degree of protonation (δ), i.e., the proportion of protonated organ-base, can be calculated using Equations (1) and (2) [20,38].
Upon bubbling CO 2 , the pH of the 0.65 wt% NaPAA aqueous solution decreased from 9.17 to 5.82, as displayed in Figure 7a. According to Equation (2), the initial δ of NaPAA was~0.06 mol%. The change in pH that occurred after sparging with CO 2 indicates that δ increased to~56.30 mol%, and thus~56.24 mol% carboxylate groups were more protonated in the presence of CO 2 . These results coincide with results of the rheology experiments: viscosities η and relaxation times τ R of the 0.65 wt% aqueous solutions dropped sharply after bubbling CO 2 , revealing that protonation of carboxylate side groups reduced the charge density and corresponding electrostatic repulsions between NaPAA chains to allow a smaller average molecular size than in the initial state. After bubbling N 2 , the pH of 0.65 wt% NaPAA aqueous solution recovered to 7.99, signifying deprotonation of the carboxylate groups. The pH was not totally restored to its initial value, indicating that 0.86 mol% carboxylate groups were still protonated. This phenomenon also coincides with the prior rheological results, in which the viscosities η and relaxation times τ R of NaPAA aqueous solutions were not fully restored to their initial values. The pH of 0.65 wt% NaPAA solutions containing different DMEA concentrations was also examined (Figure 7a). It can be found that the pH of 0.65 wt% NaPAA solutions increased apparently with the increment of CDMEA, regardless of the presence of CO2 (Table  S1). As an example, the variation in pH for the 0.65 wt% NaPAA solution with 0.6 wt% DMEA upon the introduction of CO2 and N2 is discussed. As a typical tertiary amine, DMEA can be protonated in water in the presence of CO2 to generate ammonium bicarbonate salts (Scheme 1) [39][40][41][42]. The initial pH of 11.14 was reduced to 6.75 upon bubbling CO2, and then recovered to 8.94 after introducing N2. Via Equation (2), the initial δ of DMEA can be calculated at ~2.56 mol%, followed by an increase to ~99.85 mol% after the introduction of CO2, indicating that the tertiary amino groups on DMEA molecules were almost completely protonated. After bubbling N2, the pH of the solution did not recover to its initial value; rather, 80.65 mol% of the tertiary groups were still protonated, which was caused by the larger the basicity, the poorer the "switch" capability [20]. Besides, the initial δ for NaPAA of 6.17 × 10 −4 mol% was increased to 13.15 mol% after bubbling CO2, then again restored to 9.76 × 10 −2 mol%. Thus, one can find that even if CO2 is not available, there was at least some small amount of protonated DMEA in the solution. These results properly clarify why viscoelasticity of the NaPAA solutions containing DMEA evidently decreased before bubbling CO2. By contrast, in the presence of CO2, a large number of protonated DMEA increased the charge density of the fluids, together with protonated b a The pH of 0.65 wt% NaPAA solutions containing different DMEA concentrations was also examined (Figure 7a). It can be found that the pH of 0.65 wt% NaPAA solutions increased apparently with the increment of C DMEA , regardless of the presence of CO 2 (Table S1). As an example, the variation in pH for the 0.65 wt% NaPAA solution with 0.6 wt% DMEA upon the introduction of CO 2 and N 2 is discussed. As a typical tertiary amine, DMEA can be protonated in water in the presence of CO 2 to generate ammonium bicarbonate salts (Scheme 1) [39][40][41][42]. The initial pH of 11.14 was reduced to 6.75 upon bubbling CO 2 , and then recovered to 8.94 after introducing N 2 . Via Equation (2), the initial δ of DMEA can be calculated at~2.56 mol%, followed by an increase to~99.85 mol% after the introduction of CO 2 , indicating that the tertiary amino groups on DMEA molecules were almost completely protonated. After bubbling N 2 , the pH of the solution did not recover to its initial value; rather, 80.65 mol% of the tertiary groups were still protonated, which was caused by the larger the basicity, the poorer the "switch" capability [20]. Besides, the initial δ for NaPAA of 6.17 × 10 −4 mol% was increased to 13.15 mol% after bubbling CO 2 , then again restored to 9.76 × 10 −2 mol%. Thus, one can find that even if CO 2 is not available, there was at least some small amount of protonated DMEA in the solution. These results properly clarify why viscoelasticity of the NaPAA solutions containing DMEA evidently decreased before bubbling CO 2 . By contrast, in the presence of CO 2 , a large number of protonated DMEA increased the charge density of the fluids, together with protonated carboxylate groups reduced the charge density and electrostatic repulsions between molecular chains, then the synergy of the two effects diminished the size of NaPAA molecules, resulting in the decrease of η and τ R with the addition of DMEA. The change in pH of 1.30 wt% NaPAA solution with different DMEA concentrations demonstrated similar tendencies, as shown in Figure 7b and Table S2.
As a typical polyelectrolyte, the conformation of NaPAA molecule chains is usually affected by salts in solutions, which influences the rheological properties of the fluids. To verify the changes in charge density in the NaPAA/DMEA solution at different states, Figure 8a provides conductivity of 0.65 wt% solution with different DMEA concentrations in the presence or absence of CO 2 at 25 • C. The conductivity of the 0.65 wt% NaPAA aqueous solution was initially 2.06 mS·cm −1 , then increased to 3.53 mS·cm −1 upon CO 2 treatment, suggesting that various inorganic ions (H + , HCO 3 − , CO 3 2− ) were produced. After bubbling N 2 to remove CO 2 , the conductivity of the 0.65 wt% aqueous solution returned to 2.54 mS·cm −1 , which was very close to its initial value. Hence, the enhancement of ionic strength weakened electrostatic repulsions between inter-and intramolecular chains, resulting in the observed drop in η and τ R .  Figure 7b and Table S2. As a typical polyelectrolyte, the conformation of NaPAA molecule chains is usually affected by salts in solutions, which influences the rheological properties of the fluids. To verify the changes in charge density in the NaPAA/DMEA solution at different states, Figure 8a provides conductivity of 0.65 wt% solution with different DMEA concentrations in the presence or absence of CO2 at 25 °C. The conductivity of the 0.65 wt% NaPAA aqueous solution was initially 2.06 mS·cm −1 , then increased to 3.53 mS·cm −1 upon CO2 treatment, suggesting that various inorganic ions (H + , HCO3 − , CO3 2− ) were produced. After bubbling N2 to remove CO2, the conductivity of the 0.65 wt% aqueous solution returned to 2.54 mS·cm −1 , which was very close to its initial value. Hence, the enhancement of ionic strength weakened electrostatic repulsions between inter-and intramolecular chains, resulting in the observed drop in η and τR. Besides, the protonated DMEA, i.e., DMEAH + , further increased the ionic strength of the fluids. Data shown in Figure 8a demonstrates that, when bubbling CO 2 , the maximum conductivity of 0.65 wt% NaPAA solution increased with the addition of DMEA, implying that DMEAH + was generated in those fluids. The charged DMEAH + molecules shielded static repulsions among carboxylate groups (-COO − ) of the NaPAA molecules, allowing the polymer chains to become coiled and thus reducing electrostatic repulsions between them. Thus, in the presence of CO 2 , the η and τ R of 0.65 wt% NaPAA solution decreased with the increment of DMEA. After introducing N 2 , the conductivity of the fluids partially returned to their initial values, as exhibited in Figure 8a and Table S3. These results properly explain why η and τ R could not completely go back to their initial values when introducing N 2 to remove CO 2 . Three repeated cycles indicated that DMEA possesses good CO 2 -induced switching behavior. In addition, the data in Figure 8b and Table S4 also show that variation in conductivity of the 1.30 wt% NaPAA soution with different DMEA concentrations also obeyed the same rule.
Consequently, the mechanism for CO 2 -switchable rheological behavior of NaPAA solutions, with or without the addition of DMEA, can be summarized in Scheme 2. On the one hand, for NaPAA aqueous solution, bubbling CO 2 leads to -COO − in side chains will be protonated to carboxylic acid (-COOH), resulting in a decrease in charge density on the molecular chains relative to that of the initial solution state, as shown in Scheme 2a. Therefore, electrostatic repulsions between intra-and inter-molecules due to -COO − groups are weakened, and the viscoelasticity of the NaPAA aqueous solution decreases accordingly. Treatment with N 2 allows viscoelasticity of the solution to partially recover, due to the partial deprotonation of -COOH groups. On the other hand, for NaPAA solutions that contain DMEA, in addition to the influence of protonation of -COO − , the protonated DMEA plays a crucial role in determining the viscoelasticity of the solution upon the addition of CO 2 , as displayed in Scheme 2b. When CO 2 is bubbled into the solutions, the DMEA molecules are protonated to generate quaternary ammonium salts, which shield charges remaining on the NaPAA chains, allowing them to coil more tightly, resulting in a sharp decrease in viscoelasticity of the fluids. Upon removal of CO 2 , the quaternary salts are deprotonated, causing partial recovery of the viscosity and elasticity of the solutions.
Molecules 2021, 26, x FOR PEER REVIEW 11 of that DMEAH + was generated in those fluids. The charged DMEAH + molecules shield static repulsions among carboxylate groups (-COO − ) of the NaPAA molecules, allowi the polymer chains to become coiled and thus reducing electrostatic repulsions betwe them. Thus, in the presence of CO2, the η and τR of 0.65 wt% NaPAA solution decreas with the increment of DMEA. After introducing N2, the conductivity of the fluids partia returned to their initial values, as exhibited in Figure 8a and Table S3. These resu properly explain why η and τR could not completely go back to their initial values wh introducing N2 to remove CO2. Three repeated cycles indicated that DMEA possess good CO2-induced switching behavior. In addition, the data in Figure 8b and Table S4 a show that variation in conductivity of the 1.30 wt% NaPAA soution with different DME concentrations also obeyed the same rule. Consequently, the mechanism for CO2-switchable rheological behavior of NaPA solutions, with or without the addition of DMEA, can be summarized in Scheme 2. On t one hand, for NaPAA aqueous solution, bubbling CO2 leads to -COO − in side chains w be protonated to carboxylic acid (-COOH), resulting in a decrease in charge density the molecular chains relative to that of the initial solution state, as shown in Scheme Therefore, electrostatic repulsions between intra-and inter-molecules due to -CO groups are weakened, and the viscoelasticity of the NaPAA aqueous solution decreas accordingly. Treatment with N2 allows viscoelasticity of the solution to partially recov due to the partial deprotonation of -COOH groups. On the other hand, for NaPAA so tions that contain DMEA, in addition to the influence of protonation of -COO − , the pro b a Scheme 2. Schematic illustration of the mechanism for CO 2 -switchable viscoelasticity of NaPAA in (a) pure water, and (b) DMEA solution.
Ultrapure water with resistivity of 18.25 MΩ cm −1 was produced by an ultrapure water purification system (Chengdu Ultrapure Technology Co., Ltd., Chengdu, China) and used throughout this study.

Sample Preparation
A general method of preparing NaPAA-DMEA aqueous solutions is described as follows. First, a designated amount of NaPAA was added to a beaker with ultrapure water, and the mixture was stirred until NaPAA was completely dissolved. Second, a designated amount of DMEA was completely dissolved in ultrapure water. Then, the two solutions were mixed and stirred for 24 h at room temperature to produce the final solution.
3.2.1. CO 2 Treatment CO 2 gas was bubbled into the fluid via a syringe needle at 25 • C, until the measured conductivity of the solution stabilized.

N 2 Treatment
To remove CO 2 , N 2 gas was bubbled into the previously CO 2 -treated fluid via a syringe needle at 60 • C, until the measured conductivity of the solution remains stable.

Rheological Tests
Both steady and dynamic rheological measurements were performed via a Physica MCR 302 rotational rheometer (Anton Paar, Rannachstrasse, Austria) equipped with concentric cylinder geometry (CC27). All measurements were carried out in stress-controlled mode at 25 • C, and Cannon standard oil was used to calibrate the instrument before experimentation. All samples were centrifuged to eliminate the interior bubbles prior to measurements, then equilibrated at 25 • C for at least 10 min prior to experimentation. Dynamic frequency spectra were conducted in the linear viscoelastic region, as determined from dynamic stress sweep measurements.

pK aH Determination
The pK aH (pK a of the protonated species) values of DMEA and NaPAA were determined by titrating 20 mL of 0.1 M aqueous solutions with 0.1 M hydrochloric acid. The pH was continuously monitored at 25 • C with a S2-T Kit pH-meter (Mettler Toledo, ±0.01, Zurich, Switzerland) calibrated with standard buffer solution. The pK aH values were obtained by taking the pH readings at the mid-point between two pH jumps.

Conductivity Measurements
The conductivity of NaPAA-DMEA solution was recorded with a S230-K conductometer (Mettler Toledo, Zurich, Switzerland) at 25 • C while bubbling CO 2 or N 2 alternatively.

Conclusions
In summary, the rheological properties of NaPAA aqueous solutions with different DMEA concentrations were investigated in the presence and absence of CO 2 . The effects of NaPAA and DMEA concentrations on measured solution properties were examined to elucidate the mechanism of CO 2 -responsive tunable rheological properties. It was found that NaPAA showed strong thickening power in pure water. The introduction/removal of CO 2 imparted tunable viscoelasticity to the bulk solution, attributed to protonation/deprotonation of carboxylate groups and corresponding reduction/increase in charge density among molecular chains, allowing for reduction/increase in the size of solvated molecular chains. Moreover, the addition of DMEA gave rise to further reductions in viscoelasticity upon bubbling CO 2 , but initial values could be restored to some extent via N 2 treatment. Besides, higher concentrations of DMEA corresponded to greater losses in viscoelasticity. The CO 2 -responsive switching behavior of DMEA-containing solutions was attributed to protonation of DMEA, which enabled NaPAA molecules to coil up more tightly, thus lowering the viscoelasticity, and to deprotonation of DMEAH + via bubbling N 2 , which enabled partial recuperation of the viscoelasticity. This work not only widens the utilization of CO 2 in preparing smart viscoelastic fluids but also demonstrates fabrication of such systems using a low-cost, industrial polymer and stimuli-responsive additives.
Supplementary Materials: The following are available online, Figure S1: viscosity as a function of shear rate for NaPAA aqueous solutions at 25 • C; Figure S2: viscosity-shear rate curves of 0.65 wt% and 1.30 wt% NaPAA solution while alternatively bubbling CO 2 and N 2 at 25 • C; Figure S3: strain sweep curves of 0.65 wt% and 1.30 wt% NaPAA aqueous solution at 25 • C; Figure S4: viscosity-shear rate curves of 0.65 wt% NaPAA solution with different DMEA concentrations while alternatively bubbling CO 2 and N 2 at 25 • C; Figure S5: viscosity-shear rate curves of 1.30 wt% NaPAA solution with different DMEA concentrations while alternatively bubbling CO 2 and N 2 at 25 • C; Figure S6: η 0 of 1.30 wt% NaPAA solution with different DMEA concentrations before and after bubbling CO 2 or N 2 , and the loss and recovery ratio for 1.30 wt% NaPAA solution with different DMEA concentrations at 25 • C after alternative treatment of CO 2 and N 2 ; Figure S7: frequency sweep curves of 1.30 wt% NaPAA solution with different DMEA concentrations while alternatively bubbling CO 2 and N 2 at 25 • C; Table S1: pH of 0.65 wt% NaPAA solutions with different DMEA concentration at different conditions;

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.