**3.3 Results and discussion**

Figure 3.1a shows the time dependences of the strain of the actuators, as measured by the displacement of the weight as a function of time under a load stress of 0.3 MPa. The strain in Fig. 3.1a is defined as the length change of the PPy actuators divided by the length prior to deformation. No potential voltage was given for the first 30 s, and after that repeated voltage sweeps with a period of 400 s were applied to the PPy actuators, as shown in Fig. 3.1b. The characteristics of the actuators were measured in aqueous solutions of LiTFSI with different 2-propanol concentrations of 0, 20, 80, and 100%. During the initial 130 s of the first cycle, the PPy actuators immersed in the electrolyte solutions with 2-propanol concentrations of 0 and 20% showed slight reduction of the PPy length. Notable actuator elongation of approximately 3.5% due to swelling was observed in the electrolyte solutions with the 2-propanol concentrations of 80 and 100% even when the PPy potential was negative for the first 130 s. In this case, there should be no penetration of TFSI- anions into the PPy actuators. This may be due to the fact that neutral 2-propanol molecules penetrate into the PPy porous structures resulting in the elongation of the actuators (swelling) under stress.

Here, it should be noted that the strain of the measured PPy actuator consists of the electrochemical strain caused by anion motions and the swelling (creeping) strain. Therefore, the electrochemical strain is defined by the difference in the peak value minus the lowest value of the strain for each potential cycle. The electrochemical strain of the PPy actuator driven in the 0% 2-propanol electrolyte for the second cycle was nearly 3%. In contrast, the electrochemical strain of the PPy actuator driven in the aqueous solution of LiTFSI containing 20% 2-propanol was increased up to 12%. However, the electrochemical strain of the actuator driven in the LiTFSI solution of 80% 2-propanol became nearly 5%, and it continued to decrease as time elapsed. In this case, the maximum strain of the PPy actuator stayed at approximately 12.5%, but the minimum strain of the actuators continuously increased. The electrochemical strain of the actuator driven in the LiTFSI solution of 100% 2-propanol decreased to 2%. Here, the creeping effect seems to dominate, but the electrochemical strain is suppressed. This may be due to the fact that the ionic conductivity of the LiTFSI electrolyte solution is much smaller than that of the aqueous solution. The minimum strain of the actuator continues to increase as time elapses. This behavior appears to be similar to the creeping effect in metal deformation processes. Sendai et al. recently reported their detailed study on the creeping effect of PPy actuators, and concluded that this elongation could be recovered by releasing the stress during the deformation [15]. Hence, they called this phenomenon the memory effect.

Figure 3.2. shows the 2-propanol concentration dependence of the electrochemical strain. Note that the electrochemical strain of the PPy actuators shows the maximum at a 2 propanol concentration between 20 and 40%. Since this 2-propanol concentration range is fairly wide, slight change in the 2-propanol concentration may not affect the performance of the PPy actuators.

Polypyrrole Soft Actuators 171

Fig. 3.2. Relationship between the electrochemical strain and the 2-propanol concentration of

Figure 3.3. shows comparisons of (a) the potential voltage dependences of the strain of the second cycle for the PPy actuators deformed in the electrolyte solutions with 0 and 20% 2 propanol concentrations, and (b) the potential voltage dependence of the ionic current flowing into the PPy actuators for the second cycles. The curve (b) is called a "cyclic voltammogram (CV)". When the current increases, the PPy actuator expands. This

contrast, Li+ cations will penetrate into the PPy film when the actuator is negatively biased.

elongation during the negative bias should be negligibly small. It is clear that both CV characteristics corresponding to the 0 and 20% 2-propanol concentrations suggest that the deformation of the PPy actuator is due to the penetration of the TFSI- anions. The electrochemical strain is much larger for the PPy actuator driven in the 20% 2-propanol electrolyte solution than that in the 0% 2-propanol electrolyte solution. As shown in Fig. 3.3.b, the hysteresis of the PPy actuator in the 20% 2-propanol electrolyte is much larger than that in the 0% 2-propanol electrolyte. This means that the number of TFSI- ions penetrating into the PPy actuator and out diffusing from that in the 20% 2-propanol electrolyte solution is much larger than that in the 0% 2-propanol electrolyte solution, which explains the

increased deformation of the PPy actuator in the 20% 2-propanol electrolyte solution.

Figure 3.4. shows a comparison of the CV characteristics of the PPy actuators deformed in the LiTFSI aqueous solutions with the 2-propanol concentrations of 30, 40, 60, 80, and 100%. The large hysteresis was maintained for the CV curves for 20 [Fig. 3.3.b], 30, and 40%. These CV curves became smaller as the concentration of 2-propanol increases above 60%. Both cation and anion currents are suppressed, possibly due to the reduced ionization ratio of LiTFSI in the 60, 80 and 100% 2-propanol solutions, which explains the reduced deformation

However, the diameter of the cations is much smaller than that of TFSI-

) penetration into the positively biased PPy actuator. In

anions. Thus, the

LiTFSI electrolyte solutions.

corresponds to the anion (TFSI-

range above 60%.

Fig. 3.1. (a) Relationship between the strain of the actuator and time under the repeated operation of the actuator for 8 times. The actuators were electrochemically deformed in aqueous solutions of 1 M LiTFSI mixed with 2-propanol concentrations of 0, 20, 80, and 100%. The load stress during the deformation was 0.3 MPa. (b) Potential voltage shape shown as a function of time. The potential voltage during the first 30 s was maintained at 0 V, and no stress was applied. A stress of 0.3 MPa was applied after 30 s.

(a)

(b)

Fig. 3.1. (a) Relationship between the strain of the actuator and time under the repeated operation of the actuator for 8 times. The actuators were electrochemically deformed in aqueous solutions of 1 M LiTFSI mixed with 2-propanol concentrations of 0, 20, 80, and 100%. The load stress during the deformation was 0.3 MPa. (b) Potential voltage shape shown as a function of time. The potential voltage during the first 30 s was maintained at 0

V, and no stress was applied. A stress of 0.3 MPa was applied after 30 s.

Fig. 3.2. Relationship between the electrochemical strain and the 2-propanol concentration of LiTFSI electrolyte solutions.

Figure 3.3. shows comparisons of (a) the potential voltage dependences of the strain of the second cycle for the PPy actuators deformed in the electrolyte solutions with 0 and 20% 2 propanol concentrations, and (b) the potential voltage dependence of the ionic current flowing into the PPy actuators for the second cycles. The curve (b) is called a "cyclic voltammogram (CV)". When the current increases, the PPy actuator expands. This corresponds to the anion (TFSI- ) penetration into the positively biased PPy actuator. In contrast, Li+ cations will penetrate into the PPy film when the actuator is negatively biased. However, the diameter of the cations is much smaller than that of TFSI anions. Thus, the elongation during the negative bias should be negligibly small. It is clear that both CV characteristics corresponding to the 0 and 20% 2-propanol concentrations suggest that the deformation of the PPy actuator is due to the penetration of the TFSI- anions. The electrochemical strain is much larger for the PPy actuator driven in the 20% 2-propanol electrolyte solution than that in the 0% 2-propanol electrolyte solution. As shown in Fig. 3.3.b, the hysteresis of the PPy actuator in the 20% 2-propanol electrolyte is much larger than that in the 0% 2-propanol electrolyte. This means that the number of TFSI- ions penetrating into the PPy actuator and out diffusing from that in the 20% 2-propanol electrolyte solution is much larger than that in the 0% 2-propanol electrolyte solution, which explains the increased deformation of the PPy actuator in the 20% 2-propanol electrolyte solution. Figure 3.4. shows a comparison of the CV characteristics of the PPy actuators deformed in the LiTFSI aqueous solutions with the 2-propanol concentrations of 30, 40, 60, 80, and 100%. The large hysteresis was maintained for the CV curves for 20 [Fig. 3.3.b], 30, and 40%. These

CV curves became smaller as the concentration of 2-propanol increases above 60%. Both cation and anion currents are suppressed, possibly due to the reduced ionization ratio of LiTFSI in the 60, 80 and 100% 2-propanol solutions, which explains the reduced deformation range above 60%.

Polypyrrole Soft Actuators 173

Fig. 3.4. Cyclic voltammograms for deformations in electrolyte solutions of LiTFSI mixed

Hara et al. reported that their TFSI-doped PPy actuators exhibited the maximum deformation when their aqueous LiTFSI electrolyte solution contained 20-40% of propylene carbonate (Hara et al., 2005). They pointed out that these phenomena were due to the swelling of the PPy film with propylene carbonate. The PPy film was reported to have porous and sponge-like structures, and the swelling of the film increased the porous spacing. This will make the anions more easily penetrate into the PPy film, which results in

It should be pointed out that the PPy actuators investigated in this research showed smaller levels of the creep effects in the electrolyte solutions containing 20-40% 2-propanol while they showed the maximum deformation range. This could not be always explained only by the swelling of the PPy films. It is speculated that the surface tension and viscosity of water and 2-propanol affect the electrochemical deformation of the PPy actuators. Table 2 compares the surface tension and viscosity of pure water and 2-propanol at 20 oC. The data were taken from the web page of the National Institute of Standards and Technology (NIST). The surface tension of 2-propanol is 21.7 dyn.cm-1, which is nearly 30% that of water.

molecules might more easily penetrate into the porous structure of PPy. Thus, the increased deformation was observed for the 2-propanol concentration range between 20 and 40%. On the other hand, the contraction of the PPy actuator was disturbed in the electrolyte solutions with the 2-propanol concentrations larger that 60%. The TFSI- ions diffused into the PPy porous structure in the positive potential region could be disturbed to escape from the PPy

The mechanisms for these behaviors still need to be more carefully investigated. However, the introduction of optimised amounts of 2-propanol into the LiTFSI electrolyte solution

structures due to 2-propanol's high viscosity nature in the negative potential region.

anions along with 2-propanol

with 2-propanol concentrations of 30, 40, 60, 80, and 100%.

the increase in the deformation of the PPy actuators.

Therefore, when the PPy actuator is positively biased, TFSI-

significantly improves the deformation ranges of the PPy actuators.

Fig. 3.3. (a) Relationships between potential voltage applied on PPy actuators and the strain in the LiTFSI solutions with 2-propanol concentrations of 0 and 20 %, and (b) the corresponding cyclic voltammograms of the PPy actuators and the strain in the LiTFSI solutions with 2-propanol concentrations of 0 and 20 %.

(a)

**Potential voltage (V)** 

0 %

(b)

Fig. 3.3. (a) Relationships between potential voltage applied on PPy actuators and the strain

in the LiTFSI solutions with 2-propanol concentrations of 0 and 20 %, and (b) the corresponding cyclic voltammograms of the PPy actuators and the strain in the LiTFSI

solutions with 2-propanol concentrations of 0 and 20 %.

**Strain (**%)

Fig. 3.4. Cyclic voltammograms for deformations in electrolyte solutions of LiTFSI mixed with 2-propanol concentrations of 30, 40, 60, 80, and 100%.

Hara et al. reported that their TFSI-doped PPy actuators exhibited the maximum deformation when their aqueous LiTFSI electrolyte solution contained 20-40% of propylene carbonate (Hara et al., 2005). They pointed out that these phenomena were due to the swelling of the PPy film with propylene carbonate. The PPy film was reported to have porous and sponge-like structures, and the swelling of the film increased the porous spacing. This will make the anions more easily penetrate into the PPy film, which results in the increase in the deformation of the PPy actuators.

It should be pointed out that the PPy actuators investigated in this research showed smaller levels of the creep effects in the electrolyte solutions containing 20-40% 2-propanol while they showed the maximum deformation range. This could not be always explained only by the swelling of the PPy films. It is speculated that the surface tension and viscosity of water and 2-propanol affect the electrochemical deformation of the PPy actuators. Table 2 compares the surface tension and viscosity of pure water and 2-propanol at 20 oC. The data were taken from the web page of the National Institute of Standards and Technology (NIST). The surface tension of 2-propanol is 21.7 dyn.cm-1, which is nearly 30% that of water. Therefore, when the PPy actuator is positively biased, TFSI- anions along with 2-propanol molecules might more easily penetrate into the porous structure of PPy. Thus, the increased deformation was observed for the 2-propanol concentration range between 20 and 40%. On the other hand, the contraction of the PPy actuator was disturbed in the electrolyte solutions with the 2-propanol concentrations larger that 60%. The TFSI- ions diffused into the PPy porous structure in the positive potential region could be disturbed to escape from the PPy structures due to 2-propanol's high viscosity nature in the negative potential region.

The mechanisms for these behaviors still need to be more carefully investigated. However, the introduction of optimised amounts of 2-propanol into the LiTFSI electrolyte solution significantly improves the deformation ranges of the PPy actuators.

Polypyrrole Soft Actuators 175

Fig. 4.1. Relationships between the strain of the actuator and time under the repeated operation of the actuator for 10 times with the potential sweep rate of 10 mVs-1. The actuators were electrochemically deformed in aqueous solutions of 1 M LiTFSI mixed with 2-propanol concentrations of 0 (water) and 20%. The load stress during the deformation was

The characteristics of the actuators were measured in aqueous solutions of LiTFSI with the different 2-propanol concentrations of 0 (water) and 20%. During the initial 130 s of the first cycle, the PPy actuators immersed in the electrolyte solutions with the 2-propanol concentrations of 0 and 20% showed slight reduction of the PPy length. Here, it should be noted that the strain of the measured PPy actuator consists of the electrochemical strain caused by anion motions and the swelling (creep) strain. Therefore, the electrochemical strain was defined by the difference of the peak value minus the lowest values of the strain for each potential cycle. The electrochemical strain of the PPy actuator driven in the 0% 2 propanol electrolyte for the second cycle exhibited nearly 7%. In contrast, the electrochemical strain of the PPy actuator driven in the aqueous solution of LiTFSI containing 20% of 2-propanol was 12%. In this case, the electrochemical strain of the PPy actuator gradually decreased after the repeated potential cycles but it stayed around 10%, while the creep strain of the actuators continuously increased. This creep strain of the PPy actuator in the 20% 2-propanol electrolyte was larger than that in the 0% 2-propanol (water) electrolyte. This seems to suggest that the creep strain is larger in the former case due to the swelling of the PPy film in 2-propanol. These creep behaviors look similar to creep effects in metal deformation processes. Sendai et al. recently reported in their detailed study on creep effects of PPy actuators, and concluded that this elongation could be recovered by releasing the stress during the deformation (Chida et al., 2010). Hence, they called this phenomenon the memory effect. The dotted straight lines correspond to the tangents of the second peaks, and the tangential slopes correspond to the electrochemical strain rate of 0.25%s-1 for the actuator in the 20% 2-propanol electrolyte and 0.08%s-1 for the actuator in the 0% 2-propanol

0.3 MPa.

(water) electrolyte, respectively.


Table 2. Comparison of surface tension and viscocity for water and 2-propanol at 20 oC (data taken from the web page of the National Institute of Standards and Technology (http://webbook.nist.gov/chemistry/fluid/).
