Influence of Ambient Humidity on the Voltage Response of Ionic

Subscriber access provided by RMIT University Library

Article

Influence of Ambient Humidity on the Voltage Response of Ionic Polymer-Metal composite Sensor Zicai Zhu, Tetsuya Horiuchi, Karl Kruusamäe, Longfei Chang, and Kinji Asaka J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.5b12634 • Publication Date (Web): 15 Mar 2016 Downloaded from http://pubs.acs.org on March 19, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry B is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Influence of Ambient Humidity on the Voltage Response of Ionic Polymer-Metal Composite Sensor Zicai Zhu†,‡, Tetsuya Horiuchi†, Karl Kruusamäe†, Longfei Chang§ and Kinji Asaka†* †

. Inorganic Functional Material Research Institute, National Institute of Advanced Industrial

Science and Technology, 1-8-31, Midorigaoka, Ikeda, Osaka, 563-8577, Japan ‡

. IMS Lab, University of Tartu, Nooruse 1, 50411 Tartu, Estonia

§.

Institute of Industry and Equipment Technology, Hefei University of Technology, Hefei,

Anhui, 230009, People's Republic of China *Corresponding author: [email protected]

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 49

Abstract Electrical potential based on ion-migration exists not only in natural systems but also in ionic polymer materials. In order to investigate the influence of ambient humidity on voltage response, classical Au-Nafion IPMC was chosen as the reference sample. Voltage response under a bending deformation was measured in two ways. Firstly, continuous measurement of voltage response in the process of absorption and desorption of water to study the tendency of voltage variation at all water states, secondly measurements at multiple fixed ambient humidity levels to characterize the process of voltage response quantitatively. Ambient humidity influences the voltage response mainly by varying water content in ionic polymer. Under a step bending, the amplitude of initial voltage peak first increases and then decreases as the ambient humidity and the inherent water content decreases. This tendency is explained semi quantitatively by mass storage capacity related to the stretchable state of the Nafion polymer network. Following the initial peak, the voltage shows a slow decay to a steady-state, which is first characterized in this paper. The relative voltage decay during the steady state always decreases as the ambient humidity is lowered. It is ascribed to progressive increase of the ratio between the water molecules in the cation hydration shell to the free water. Under sinusoidal mechanical bending excitation in the range of 0.1-10 Hz, the voltage magnitude increases with frequency at high ambient humidity but decreases with frequency at low ambient humidity. The relationship is mainly controlled by the voltage decay effect and the response speed.


Page 3 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Electric potential induced by ion migration widely exists in natural systems such as nerve fibers, tactile corpuscles and auditory hair cells. In a cell, membrane potential can be generated by ion transport through the phospholipid bilayer as shown in Figure 1(a). For instance, the cell membrane potential under a pressure, vibration or acoustic stimulus is the basis for our tactile and auditory sensing1. Ion-migration-based electrical response can also be realized by ionic polymer materials including ionic gels2, ion exchange membranes3, solid electrolyte composites containing ionic liquid4 and conducting polymers5. These ionic polymers can be used as mechanical sensors or energy conversion materials6,7 with advantages of light mass, flexibility and easy processing. Their electromechanical properties can be fine-tuned during the fabrication to meet specific requirements. As flexible sensors, they are considered attractive for wide range of potential applications, amongst which artificial skin8-11 - a compact flexible sensor array with pressure, tactile and/or vibration sensation - is one exceptional example.

(a)

(b)

Figure 1. Illustration of electrical potential caused by ion migration. (a) Cell membrane potential. (b) IPMC sensing under a bending deformation. Ionic polymer-metal composite (IPMC) is usually composed of an ion exchange membrane and two metallic electrodes which form a sandwich structure. In comparison to other ionic polymer sensors, water-containing IPMC is more similar to systems occurring in Nature. It has a simpler composition with only one kind of mobile ion and it often shows a higher voltage amplitude12-14 under the same bending deformation. Thus it was chosen to be a typical ionic sensor material for fundamental research on ionic sensing properties. As shown in Figure 1(b), the polymer substrate contains hydrated cations and water molecules. Generally, when applying a bending deformation, an elastic stress gradient is generated along 3 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 49

the thickness, thus the mobile cations and water molecules migrate towards the side with lower hydraulic pressure, i.e. the outside electrode15,16. Ionic charges redistributing along the thickness cause an electrical response between the two electrodes, and the outside usually shows a higher electric potential. Newbury and Leo investigated the sensing properties of ionic polymer transducers and compared to that of piezoelectric materials. It was reported that an IPMC sensor is one order of magnitude more sensitive than traditional piezoelectric materials17. Brunetto et al. showed experimentally that the magnitude of the electrical potential under a continuous sine vibration is primarily dependent on the humidity level18. Chen et al. observed that as a hydrated sample approaches equilibrium with the environment, sensitivity initially improves but is then followed by decay19. While the previous observations convey general trends on how the humidity influences sensorial properties, in these reports the initial or final water states of IPMC have not been clearly defined, thus making it difficult to draw explicit conclusions. On the other hand, Farinholt and Leo20 simulated the current responses to some extent by cation equation based on Nernst-Planck transport theory. Following the model, Chen19 proposed a dynamic model to predict the frequency response of IPMC current by considering electrode resistance. Gao and Weiland21 predicted the current response by depicting cations motion based on streaming potential hypothesis. All these studies have presented the basic physical process of cation distribution under a bending load. However, we are still far from comprehensive understanding of sensing properties and the underlying physics of ion-migration-based sensing. Due to the complex structure of IPMCs, there are several factors that influence the sensing performance such as the ambient humidity (water content), type of cations, dimensions of the polymer membrane (especially the thickness), physical properties of the electrode. This paper concentrates on the influence of ambient humidity on the electrical response of IPMCs. Ambient humidity influences the water content in IPMC and presumably the ratio of so-called states of water22-24. Water molecules also influence the properties of cation transport and therefore the electrical response. In contrast with the electrical current, the responded voltage directly reflects the state of cation distribution and more helpful to reveal the sensing physics. Therefore, in this paper we mainly focus on the voltage response of IPMC sensor influenced by ambient humidity. Based on our previous studies regarding the relationship between water content and actuation mechanism25,26, two series of experiments were designed and performed to investigate how ambient humidity influences the sensing and the underlying mechanisms. In the first series of experiments we investigate the voltage response of IPMC 4 ACS Paragon Plus Environment

Page 5 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


sensor at all states of water content. It was realized by a continuous measurement of voltage response during a spontaneous alteration of water content. The change in water content was triggered by the initial difference of humidity levels of IPMC sample and its surroundings as fully hydrated or completely dry IPMC sample is placed in an environment with predetermined humidity. This kind of scan test can show the voltage response evolvement at diverse water contents. In the second series of experiments we investigate the voltage responses, including both static and dynamic responses, at multiple fixed ambient humidity levels. We describe a quantitative relationship between the voltage response and the ambient humidity. Finally, we discuss the physical mechanisms to explain the peak voltage, steadystate voltage and frequency response.

2. Experimental Methods 2.1 Preparation of Au-Nafion 117 (Na+) IPMC. A Nafion® (DuPont) plated with gold electrodes (Au-Nafion 117 IPMC) was chosen as the test material for an IPMC sensor because of its good stability of performance27. After being roughened by sandblasting and boiled in 5% H2O2 and 0.5 mol/L H2SO4 solutions in sequence, a Nafion 117 membrane was immersed in an aqueous solution of gold complex ([Au[Phen]Cl2]Cl) for cation exchange overnight at room temperature. Then the membrane was washed with water to remove the excess metal species and immersed in reducing solution of 3-4 mmol/L sodium sulfite (Na2SO3) to precipitate metallic gold for 5 hours. The processes of cation exchange and reduction were repeated for four times to acquire an ideal electrode with good conductivity. Next it was cut into a strip with a size of 22 mm × 5 mm and immersed in 0.1 mol/L solution of NaOH so that the counter cation was exchanged to Na+. Two samples were used in this paper, one for the water desorption/absorption test and the other for ionic sensing properties.

2.2 Mechanoelectrical sensing test platform. A dedicated test bench consisting of PCcontrolled wavemaker and signal acquisition system was designed and set up to study the sensing behavior of different IPMCs. As illustrated in Figure 2(a), a control signal was generated by National Instruments LabVIEW and sent to an Asahi Seisakusyo APD-050FCA controller. A small displacement stimulus, generated by the wavemaker SL-0505, was applied at the tip of the IPMC sample to cause bending deformation. An IPMC sample was clamped with a gold foil covered clip and formed a cantilever structure. The free length between the fixing clamp and the point of application for displacement was set to 10 mm. A Keyence IL30 laser displacement meter was used to measure the actual deflection imposed by the wavemaker. The IPMC and the immediate supporting structure were placed in an ESPEC SH5 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 49

222 temperature and humidity chamber as shown in Figure 2(b). The voltage generated by IPMC sensor was amplified by custom-made circuit. All data were recorded using a National Instruments PCIe-6351 data acquisition board. In order to preclude any S-shape deformation, the head of the wavemaker was not fixed to the IPMC sample, leaving a small gap (400µm, double thickness of the sample) as depicted in Figure 2(c) and shown in Figure 2(d). The displacement of mechanical excitation was set to 2 mm in all the reported experiments.

(a)

(b)

(c)

(d)

Figure 2. Test platform for IPMC sensor. (a) Schematic of the custom-made setup; (b) Main test bench placed in the humidity chamber; (c) and (d) are configuration and view of the IPMC sample and the shaker head. 2.3 Processes of water absorption and desorption. Nafion-based IPMC is a water absorbing material and therefore ambient humidity influences the dynamic and steady-state water content of IPMC. It is thus necessary to study how long it takes for an IPMC sample to reach a steady-state of water content. The processes of absorbing and desorbing water were characterized by dynamic water content w(t). For measuring the water desorption, IPMC sample was taken from deionized (DI) water and after removing the excess surface water, electronic balance was used to measure its mass m(t) after every 2 minutes from the state of full saturation to a steady level at ambient conditions (25.1°C, 54%RH). For measuring water 6 ACS Paragon Plus Environment

Page 7 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


absorption, IPMC was first dried in vacuum oven at 60°C overnight, which completely removes the free water22,25. After cooling down, the mass at dry state mdry was measured immediately, and the change of mass m(t) in time was recorded similarly to the case of water desorption. The dynamic water content is obtained by:

=

× 100%

(1)

2.4 Voltage response when varying water content. To measure the voltage response of IPMC at all water content states (0 – w(t) at full saturation), the voltage response is continuously measured in the process of varying water content. To cover all the water states as comprehensive as possible, it was performed as follows. First the voltage response was measured from the full saturation (the state after taking the IPMC from DI water) to the room environment (25.1°C, 54%RH). In another experiment continuous cyclic response was recorded from the dry state to the room environment. Because the two equilibrium water contents of the previous were not coincident, a third measurement was conducted from full saturation to a predetermined environment (25°C, 20%RH, the minimum humidity that can be achieved by the humidity chamber). In these measurements, IPMC sample was subjected to step bending displacement with ±2mm amplitude and 50 s period. The response under a step bending reveals the transient sensing process. It takes about 1 hour for IPMC to get the equilibrium state in air, so each series of measurements lasted for one and a half hour. 2.5 Voltage response at multiple fixed ambient humidity levels. The voltage response of IPMC was also investigated at multiple fixed levels of ambient humidity, i.e. water contents here. After being taken from DI water and removing the surface water, the voltage responses of IPMC at full saturation, 90%, 70%, 50%, 30%, 20% relative humidity were investigated in sequence at a constant temperature of 25°C. According to experimental results in Sec. 2.4, the voltage response at these humidity levels were sufficient to reflect the whole evolvement with water content. For each case, it takes approximately 10 minutes for the chamber to adjust humidity and about 1 hour for the water equilibrium between IPMC sample and the atmosphere in the chamber. Therefore the interval of two consecutive measurements was at least 1 hour. Both static and dynamic voltage responses were acquired. After establishing initial level for 2 seconds, IPMC sample was subjected to 20 seconds of 2 mm step displacement and then forced to recover back to the initial position for 10 seconds. The duration of 20 s is a compromise of clear display between the initial transient and the steadystate responses. Response under a sinewave bending was also performed to investigate the



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 49

dynamic response. The bending amplitude was also set to 2 mm and the frequency ranged from 0.1 to 10 Hz. Identical measurements were carried out on an IPMC in its dry state.

3. Results and Discussion 3.1 Water absorption and desorption. During the processes of water absorption and desorption in air, the water content change of IPMC with time is shown in Figure 3. The water content at full saturation is about 24.44% and the equilibrium water contents at room environment (25.1°C and 51%RH) are 2.86% and 2.04% for desorption and absorption, respectively. It means that the equilibrium water content is decided not only by ambient humidity, but also by the initial water content state. The equilibrium water content in desorption is usually a little larger than that in absorption at the same ambient humidity28. Due to water molecules diffusion, the water content varies with time in exponential form, which decreases or increases acutely at the beginning. Using an exponential decay function to fit the experimental data, it was observed that the time constant of desorption (11.295 minutes) is smaller than that of absorption (20.753 minutes), which means the desorption process is much faster. Water molecules exchange at the polymer-air boundary. They go out of the polymer to an open space in desorption and go into the polymer in absorption. Water diffusion in the air is faster than that in the polymer. Faster diffusion further leads to low pressure of water at the boundary and accelerates the water exchange. It will lead to a faster water exchange between the polymer and the air in the desorption process, which shows a smaller time constant. We also see that for the measurements during the varying water content, one and a half hour is sufficient for the IPMC sample to reach equilibrium state. For the measurements at multiple fixed ambient humidity levels, the interval of two consecutive measurements should be at least 1 hour.

Figure 3 Water content of Au-Nafion (Na+) IPMC varies with time due to water desorption and absorption in air 8 ACS Paragon Plus Environment

Page 9 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3.2 Voltage response at all water content states. The voltage responses of the three continuous measurements when varying water content were performed for three times in different days. The experimental results are repeatable and one of them is shown in Figure 4. Subplots (a), (b), and (c) are the results from full saturation to 54%RH, from full saturation to 20%RH and from the dry state to 54%RH, respectively, which have been treated by a low pass filter with a cutoff frequency of 8Hz.

(a)

(b)

(c) Figure 4 Continuous electrical responses of IPMC under a cyclic square wave bending deformation. Three times measurements (a) from full water saturation to the room humidity 54%RH, (b) from full water saturation to the setting humidity 20%RH; (c) from dry state to the room humidity 54%RH. In desorption process, there was a clear large voltage drift during the first 1800 s in (a) and (b). The voltage drifted toward negative potential and then recovered back to positive potential. Compared to the process of water desorption in Figure 3, the voltage drift is strongly correlated to the considerable water loss in the first 30 minutes29. Usually it is difficult to fabricate an IPMC with exactly the same two electrode layers. Water loss at 9 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 49

different rates on the two electrode surfaces can cause a gradient of water concentration in the polymer. The gradient causes redistribution of water molecules and cations, and then induces a voltage drift because of non-uniform ionic charges in the polymer. This phenomenon is repeatable to some extent, but it is difficult to precisely control the absolute value of voltage. After the first 30 minutes, the voltage drift is attenuated as the water changes very little. For the absorption process, it is hard to observe any voltage drift, because the variation of water content is quite small, i.e. no more than 2.04%. Additionally, the bias voltage was not zero in most experiments. It reflects distribution of the ionic charges at natural state without any mechanical load. Usually, the distribution of movable cations, like that of residual strain, is non-homogeneous and causes a non-zero bias voltage.

(a)

(b)

Figure 5 Typical electrical response of IPMC under a cyclic square wave bending deformation. (a) Definition of peak-to-peak voltage VPP, steady-to-steady voltage VSS and transient voltage amplitude VA; (b) Voltage amplitude at a nearly dry state. In order to describe the voltage response evolvement with water content, the characteristic voltages are defined in Figure 5(a). The voltage difference between the maximum and the minimum is peak-to-peak voltage VPP, and the voltage difference between the two steady-state is steady-to-steady voltage VSS. Then the transient voltage under a 4 mm step bending stimulus can be obtained by, =

(2)

In the absorption process, the voltage response showed a random noise during the first 2000 s (Figure 4(c)). When approaching to the dry state (no free water), IPMC sensor, especially the Nafion membrane is more like a solid resistor with high resistance. The high noise level can be ascribed to thermal noise like that in a resistor. As water content increasing by absorbing, the noise level decreased with time from 3000s to 8000s continuously, then the voltage 10 ACS Paragon Plus Environment

Page 11 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


showed a similar tendency with the square waveform bending as shown in Figure 5(b). Due to high noise level, it is difficult to identify any voltage decay. So the peak-to-peak voltage, steady-to-steady voltage and the voltage amplitude are treated as one. The behavior of VPP,

VSS, and VA over the course of time is shown in Figure 6. In Figure 6(a), the peak-to-peak voltage VPP under a step bending first increased with time but then showed a drop at around 4000 s, whereas the steady-to-steady voltage VSS monotonously increased with time. Compared to the voltage VPP, the voltage amplitude VA increased with time from 0.83 mV to 1.51 mV at the first 3000 s, and then changed very little besides the slight fluctuation. Thus the voltage VA is much better to characterize the magnitude of initial transient voltage under a step bending. In Figure 6(b), the voltage VA first increased from 0.70mV to 1.40mV but then showed a monotonous decline with time to 1.18mV. Comparing the voltage variation with time in Figure 6(a) and (b), it can be deduced that the whole process in Figure 6(a) was compressed into the first 30 minutes in Figure 6(b). The maximum voltage (1.51mV) in Figure 6(a) is slightly larger than that (1.40mV) in Figure (b). It is likely that the water loss occurred so quickly that it was difficult to capture the optimum voltage during the first 1500 s in Figure 6(b). In Figure 6(c), the voltage amplitude

VA was about 1.13mV after the first 2000 s of noise. In general we can obtain that the initial transient voltage first increases and then decreases from the beginning of full saturation to dry state (the same tendency is also shown by a similar measurement under a sine waveform bending at 1Hz in our lab), whereas the steady-state voltage shows a continuous increase, except when IPMC is close to the dry state.

Figure 6 Peak-to-peak voltage VPP, steady-to-steady voltage VSS and transient voltage VA vary with time (a) from full water saturation to the room humidity 54%RH, (b) from full water saturation to the setting humidity 20%RH; (c) from dry state to the room humidity 54%RH.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 49

In addition, the voltage responses during the final 30 minutes in Figure 4 can be regarded as the responses at equilibrium states. Although in Figure 4(a) and (c) the ambient humidity levels were the same 54%RH, and the difference of the actual equilibrium water contents was very small, the voltage responses at the steady state were quite different (subfigures (3) and (9)). The voltage response is not decided solely by the ambient humidity. Considering that the equilibrium water contents were different, water content is most likely the direct acting factor of the electrical response. Even by comparing the voltage responses in subfigure (6) and (9), the former (20%RH) still shows larger magnitude and much less noise than the latter. 3.3 Voltage response at multiple fixed ambient humidity levels. The voltage responses at multiple fixed ambient humidity levels are shown in Figure 7. The static voltage response includes two processes: initial fast voltage rising and slow voltage decay (Similar tendency has been confirmed based on Au-Nafion117 IPMC with various cations, H+, Li+, Na+ and K+ cations and can be found in supplementary information). It is noteworthy that the steady-state voltage does not return to the initial zero. When IPMC contained enough water from “Water” (full saturation) to “RH70” (70% relative humidity), the initial voltage peak was very sharp and the voltage decay was considerable. Whereas for the rest, the peak got obtuse and the voltage decay grew smaller as the ambient humidity decreased. It means that the response speed i.e. the migration of cation and water became slower. At the dry state it also showed only random noise.


Page 13 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure.7 Voltage response of IPMC under a step bending deformation at multiple fixed levels of ambient humidity from full saturation (Water) to dry state (Dry). a. Initial peak voltage. The initial peak voltage varying with ambient humidity is shown in Figure 8. As ambient humidity (water content) decreases, the peak first increases, remains at the highest voltage level from 90% to 30%RH and then decreases. The tendency is in accordance to that of the voltage VA obtained previously. And a similar tendency can be observed in the deformation of IPMC actuator with ambient humidity25,26. The reason why the initial peak voltage cannot reach the optimum at full saturation but at a moderate water content state can be explained by material parameters (such as elastic modulus, diffusion coefficients etc.) variation with water content. However, the intrinsic mechanism is most likely related to mass storage capacity. And it can be explained by the stretchable state of the Nafion polymer network based on the widely accepted ionic cluster model30 as follows.

Figure. 8 The initial peak voltage of IPMC varies with ambient humidity. As shown in Figure 9(a), when IPMC is applied a bending stimulus, there are three different states for ionic clusters: expanded, neutralized and contracted. The cluster containing more water molecules and cations exhibits higher pressure in the liquid phase, i.e. larger elastic stress in the polymer, and vice versa. In comparison to the neutralized cluster located at the neutral plane, the polymer network is compressed (stretched) in the expanded cluster and relaxed (contracted) in the contracted cluster in radial (tangential) direction. The extent that the polymer network can be compressed decides the mass storage capacity and the initial peak voltage, because the cations and water molecules migrate toward the same direction at the beginning.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(a)

Page 14 of 49

(b)

Figure 9 Illustration of polymer network compressed states in IPMC. (a) Compressed states at three different swelling ionic clusters when IPMC bending, the yellow is solid polymer and the white is liquid phase; (b) Normalized pressure in a cluster controlled by water volume fraction ratio. Based on the micromechanical model and the representation volume element (RVE) in Ref. 31, on the macro scale, stress state is generated in the polymer under an external force i.e. bending displacement. Whereas on the micro scale in an ionic cluster, due to compression or stretch of the polymer chains, the radial elastic stress difference between the inner ( ) and outer ( ) surface of an ion cluster in Figure 9(b) is, =

'

"#$ ( ! & "#

%$

'

( # ! $& ) #

%$(3)

Where Gdry is bulk modulus of the polymer, is the water volume fraction (ratio of local

volume of water to local volume of polymer network) and * is the initial value. Usually for a Nafion based IPMC, * is about 0.5 at full saturation31. The elastic stress at the inner

surface is applied on the liquid phase directly, thus the hydrostatic pressure Ph equals to . On the micro scale, mass transport of water and cation is induced by the hydrostatic

pressure gradient. Equation (3) is a bridge to connect the force balance and the mass transport on different scales. From Ref. 32, the cluster pressure Pc mainly refers to the hydrostatic pressure Ph here. Other eigen stresses such as electrostatic pressure and osmotic pressure, which are related to cation concentration gradient along the thickness, can be ignored approximately here, because the ionic current of IPMC sensor is quite low (micro ampere). And since the elastic stress

is decided by the external force, it can be deemed as a constant under the same mechanical load. Then the cluster pressure Pc can be evaluated by setting = 0,


Page 15 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


+, =

'

"#$ ( ! & "#

%$

'

# ( ! $& ) #

%$(4)

Then normalized cluster pressure can be defined as a function of water volume fraction ratio (λ = /* ),

'

"#$ ( Pc*= =! & 1 "#$% /0

!

'

#$ (

#

%$&

'

"2#$% ( =! & "#

%$'

λ(

(5)

For example, when * = 0.3 (at a moderate ambient humidity), the cluster pressure varying with the ratio is shown in Figure 9(b). It shows higher cluster pressure in expanded cluster and lower pressure in contracted cluster. Let us set the reference state (Pc = 0) of the initial cluster pressure at a moderate water

content state. If we set Pc = 0 at * = 0.3, we can obtain the initial normalized cluster pressure varying with water volume fraction by Equation (4), i.e. the dashed line in Figure 10. At different initial water content states (* = 0.1, 0.2, 0.3, 0.4, 0.5), the cluster pressures Pc at

reference points ( = * ) have different initial values, because the polymer shows different stretchable states. And under a bending stimulus, the cluster pressures Pc varying with water volume fraction at different initial water content states (* = 0.1, 0.2, 0.3, 0.4, 0.5) are shown in Figure 10(a), which have a similar tendency with that in Figure 9(b). However, at full saturation the ionic cluster is extremely expanded as in Figure 9(a) at the initial state. As a rough estimation, setting the initial radius of an ionic cluster unit (a solid sphere, the radius of the inner pore is zero approximately) at the dry state is 1, the inner and outer radii of the ionic cluster (a spherical shell) are about 0.8 and 1.15 at full saturation when the water volume fraction is 0.5. Then the average compression ratio in the radius direction is 0.35, which is a relative high compression ratio. The elastic modulus of polymer increases greatly under a large stretch or compression, which is named as strain-stiffening effect33,34. We think the elastic modulus of Nafion membrane Gdry increases greatly at saturation state because of the large compression by swelling. The initial stretchable state of the polymer network reaches the maximum state, i.e. the polymer is too stiff to be compressed in the radial

direction and is difficult to absorb more water. So the purple curve at * = 0.5 predicted by Equation (4) is not effective any more. The bulk modulus of the polymer Gdry at full saturation increases to tens of times of that at moderate compression i.e. moderate water content state. The cluster pressure will change greatly with a little variation of water volume as illustrated by the green curve in Figure 10(b) (In Equation (5), Gdry in Pc/Gdry is a constant defined as the bulk modulus of the polymer at moderate water content state.).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 49

Under the same mechanical bending, it generates approximately the same elastic stress ( ) distribution and requires the same cluster pressure (Pc) gradient to balance. Usually the cluster pressure is higher at the outside electrode and lower at the inside electrode. From Figure 10, to generate the same level of pressure Pc gradient, less water volume fraction difference is required between two electrodes at full saturation. It means less water and cation can be driven to the outside electrode. Due to low mass storage capacity, the initial peak voltage is not the optimum at full saturation.

(a)

(b)

Figure 10 (a) Normalized cluster pressure varies with water volume fraction at diverse initial water content states, the dashed line is the reference pressure; (b) Illustration of the actual normalized cluster pressure (green) at full saturation (* = 0.5). On the other hand, when IPMC is close to the dry state, there are less water molecules in the polymer. It increases the drag force for both water molecules and cations. Therefore less and slower water and cations migrate to the outside electrode, leading to not only smaller peak voltage, but also to a slower response speed. b. Steady-state voltage. The steady-state voltage varying with ambient humidity is shown in Figure 11(a). It increases as the humidity decreases until 30%RH. The sudden drop of the steady-state voltage after 30%RH is possible due to the decline of the initial peak voltage. In Figure 11(b), we see that relative voltage decay decreases monotonously together with the humidity. The relative voltage decay is defined as: Decay =

/89 :8 /89

× 100%.

(6)


Page 17 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 11 (a) Steady-state voltage and (b) relative voltage decay varies with ambient humidity. As described in Refs. 22 and 25, two states of water, the free water and the water molecules in the cation hydration shell, are mainly responsible for the electromechanical properties of IPMC. Under a bending deformation, both the free water and the water molecules in the cation hydration shell migrate toward the outside electrode to balance the elastic stress gradient. Cations also move in the same direction mainly based on hydration, and then a part of these cations will diffuse back under the concentration gradient and cause the voltage decay. The steady-state voltage is determined by the amount of cations at the outside electrode. These cations are bonded with the water molecules in the cation hydration shell, so the steady-state voltage is also related to the amount of the water molecules in the cation hydration shell there. At different initial water content states, the amount of total migrated water is approximately the same at steady state under the same bending deformation except at full saturation or nearly dry state. At full saturation, less water is required at the outside electrode to balance the applied bending due to high elastic modulus as illustrated in Figure 10(b). As the water content decreases, the ratio of water molecules in the cation hydration shell to the free water increases monotonously25. For the migrated water at the outside electrode, the ratio is also increased. It means the amount of the water molecules in the cation hydration shell i.e. cations increases under the same bending. More cations lead to the increase of the steady-state voltage. At nearly dry state (20%RH), both the amount of the initial migrated water and the amount of the migrated water at steady state decrease greatly. It leads to a lower steady-state voltage but guarantees the monotonous decrease of the relative voltage decay in Figure 11(b). c. Frequency response. Under a sine waveform bending, a typical dynamic voltage response is shown in Figure 12(a), and the peak-to-peak value and the phase difference of the voltage response varying with frequency are shown in Figure 12(b) and (c). 17 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 49

At high ambient humidity levels (90% and 70% RH), the amplitude increased with frequency from 0.1 to 10Hz. And the slope decreases when reducing the humidity. In Ref. 35, it was reported that the voltage amplitude is linearly proportional to the bending velocity. To some extent, it is consistent with the results here, because higher frequency means larger velocity at the same deformation amplitude. However, at 50%RH the curve showed a decline after 1Hz. Furthermore, the curve showed a monotonous decline with frequency at 30%RH, which is contrary with that at high humidity.

(a)

(b)

(c)

Figure 12 Dynamic response of IPMC sensor under a sine waveform bending at multiple fixed levels of ambient humidity, (a) a typical voltage response with the applied bending displacement; (b) peak-to-peak value and (c) phase varies with frequency. At 90% RH, the peak-to-peak value at 10Hz is about 0.84mV, which is still a little less than double of the peak voltage (0.45mV) in Figure 7. At high ambient humidity, IPMC sensor shows a large and fast voltage decay caused by the back diffusion of cation. As frequency decreases, the slower vibration gives more time to the voltage decay of IPMC sensor. If the vibration is slow enough, the voltage decay or the steady-state voltage will


Page 19 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


dominate the amplitude in frequency response. So at 0.1Hz, the voltage amplitude shows RH90< RH70< RH50 < RH30. Increasing frequency can reduce effect of the voltage decay, thus lead to a larger voltage amplitude, which is coincident with the cases of 90% and 70% RH. At low ambient humidity, the response speed is much slower. In Figure 7, it takes about 0.2s for IPMC to reach the peak voltage at 20% RH. With frequency increasing, cations movement cannot catch up with the reverse speed of mechanical vibration. It makes the amplitude decline with frequency. At high frequency, the response speed shows more impact on the voltage amplitude. So at 10Hz, the voltage amplitude is opposite RH90> RH70> RH50 > RH30. Decreasing frequency can reduce effect of the response speed, thus lead to the voltage increase in the case of 30% RH. At a moderate ambient humidity, IPMC sensor stays at an intermediate sate. The influences of voltage decay and response speed on the voltage amplitude are comparable, so it first increases and then decreases with frequency. At a moderate frequency (0.5-1Hz), the tendency of the voltage amplitude with humidity is similar with that of the peak voltage in static response. In general, the frequency response of IPMC sensor is not only controlled by the peak voltage, but also greatly influenced by the voltage decay and the response speed identified from static response. For the phase diagram, the phase is very small (most 70%RH), but decreases with frequency at low ambient humidity (

Influence of Ambient Humidity on the Voltage Response of Ionic

Recommend Documents