Applied Mechanics and Materials Vols. 110-116 (2012) pp 1336-1344 © (2012) Trans Tech Publications, Switzerland doi:10.4028/www.scientific.net/AMM.110-116.1336
Comprehensive Characterization of Large Piezoresistive Variation of NiPDMS Composites Giancarlo Canavese1a, Mariangela Lombardi2b, Stefano Stassi3c, Candido F. Pirri4d 1,2
Center for Space Human Robotics, IIT-Italian Institute of Technology, C.so Trento 21. Torino, 10129, Italy.
3,4
Dept. of Materials Science and Chemical Eng., Politecnico di Torino, C.so Duca degli Abruzzi 24, Torino, 10129, Italy a,b
[email protected], c,d
[email protected]
Keywords: Quantum tunneling, piezoresistivity, polymer-metal composite, tactile sensor, robot sensing skin
Abstract. This work presents a comprehensive investigation of the piezoresistive response of a metal-polymer composite for robotic tactile sensor application. Composite samples, based on nickel nanostructured conductive filler in a polydimetihylsiloxane (PDMS) insulating elastomeric matrix, were prepared changing several process parameters like thickness, composition of the polymer and nickel filler content. A variation of electric resistance up to nine orders of magnitude under applied uniaxial load was measured in the fabricated samples. Cost efficient materials, simplicity of the process, large sensibility, and harsh environment compatibility make this quantum tunnelling composite adapted to be integrated as sensing coating in space robotic applications. Introduction The research fields concerning industrial automation, service robots, and in particular space robotics have generated the most interest in tactile sensor technology since contact interactions are a fundamental feature of any physical manipulation system. Despite tactile sensing is basically conceived like a spatially distributed function, much of the published works describes devices and applications based on localized sensing[1]. Many tactile devices have been developed in terms of tactile sensing points or pads glued to a surface with the aim to realize robust and reliable locally discrete devices e.g., tactile sensors mounted on the gripping surface[2]. Since one of the hardest problems in the fabrication of a distributed tactile sensor system is the wiring topology, only few studies consider a whole body sensing coating[3, 4]. For this last reason a more ambitious related aim is to develop a fully distributed “sensing skin” able to completely cover the robot surface in order to detect collision when operating in unstructured environments, to refine the human machine interaction, and to guarantee dexterous object manipulation [5]. Moreover distributed tactile sensors are essential for humanoid and space service robot to realize dynamic whole-body motion control [6]. Literature proposes sensors based on different working principles to transduce a stain, or an induced stress due to a tactile interaction: piezoelectric [7], optical [4], magnetic [8], piezoresistive [2] and capacitive principle [9] have been reported. In order to coupled tactile device system to robot surface two main sensor architectural strategies are reported in literature. The former deal with the integration of a network of arrays of separated devices [2], the latter, based on continuous sensible films, consist in solutions able to better promote the idea of cover the complex robot surface by a continuous smart sensing skin [10]. Among this last group, the piezoresistive principle has received most attention in tactile sensor design. This transduction method is based on the change in resistance of a conductive material under an applied load. This technique involves measuring the variable resistance either through or across the thickness of a conductive film. Besides standard materials used in MEMS technology, during the last year research efforts in tactile sensing have been oriented toward hybrid piezoresistive composites and carbon black, graphite flakes, fibers semiconducting metal-oxides (V2O3, TiO, etc.) and various metal powders All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of TTP, www.ttp.net. (ID: 130.192.47.152-03/10/11,11:56:19)
Applied Mechanics and Materials Vols. 110-116
1337
such Zn, Ni, Cu, Ag and Fe have been used as active fillers. Several polymers have been adopted as deformable isolating matrices: polyethylene, polyimides, polyesters, PTFE, polyurethane, PVA, epoxies, acrylics, etc.. Anyway the most diffuse conductive polymer composites are based on silicon rubber (PDMS) [10, 11]. Composite piezoresistive polymeric materials are the best candidate to constitute a smart sensing skin for robotic applications. In particular these composites could satisfy the main requirements for space robotics applications as: conformability (adaptable to arbitrary curved surfaces), wide range and high sensitivity, compliance (soft surface), large area coverage, lower power consumption, low payload, large working temperature range, insensitivity to electromagnetic field noise. Moreover, due to their simple construction, generally these composites are very robust on overpressure, shock and vibrations. The electrical resistance changes with the amount of force applied to the material, resulting from the deformation of the polymeric matrix altering the particle density within it. The polymer composites filled with a conductive phase should by divided in two main groups for what concern the conductive mechanisms. Regarding the first group of these materials well known as pressure conductive rubber, it has been demonstrated, that the randomly distributed electrical conductive particles are in physical contact with each other when the samples are mechanically loaded [12-15]. These composites have been largely investigated in different applications [10] and several electrode configurations have been reported [16-18]. The explanation for the conductivity variation of the sample with strain can be ascribed to the change in the conducting particles contacts, that increase when the samples are mechanically deformed [14]. The dependence of the conductivity on filler concentration slowly increases at low conductive filler loading, then over a narrow concentration range, rises fast and a high conductivity is reached with only a weak further increase in filler concentration. The change in resistivity as a function of filler concentration in conductive polymeric composites can be described by using the concept of percolation. It describes the conduction with the presence of electrically paths between two filler particles and different theoretical models have been presented [19, 20] but they generally fail below the percolation threshold. Luheng et al. [21] tested a carbon black silicon rubber composite that showed a critical pressure varying with the carbon black content. Based on the shell structure theory, the experimental phenomena were explained supposing that the external pressure induced the formation and destruction of effective conductive paths, leading to the changes in the resistance of the composite. In the second group of conductive polymers the conductive particles are dispersed very close each other, but not in physical contact being fully coated by the polymer. During the last decade different models have been consider to explain the conduction mechanism, such as electrical field induced emission[22], Richardson-Schottky transmission types and Pole-Frenkel conduction [19], but the well established and wide accepted model used to predict electrical conduction in piezoresistive polymer composites is the tunnelling effect[19, 23-26]. Without external deformation, the conductive particles are positioned apart from each other; consequently, the resistance value is infinitely large. However, when a compressive load is applied, the insulating layer between the particles decreases and the fillers form a chain of tunnelling paths: this brings about a large reduction in resistance value. Bloor et al. [23] tested different elastomeric matrices filled with nickel powders prepared as described by a patented process [27]. They showed for the first time a variation of twelve orders of magnitude of the resistance of a polymer composite piezoresistive samples[23]. This extreme large electrical conduction in response to a mechanical strain is assured by the presence of nanostructured extremely sharp tips on the nickel surface particles that enhance the local charge density. The mechanism of conduction results in a field assisted Fowler-Nordheim tunneling, because the charge injected in the composite will reside on the filler, generating very large electric local field on the surface of the tips. The field enhancement factor on the tip can be as high as 1000[28].
1338
Mechanical and Aerospace Engineering
The fundamental role of piezoresistive enhancement played by the tips is inferred from the results reported by Abyaneh et al. [29]where silicone filled with smooth zinc particle showed a considerable lower piezoresistive response at the same applied pressure. In the present work, for the first time, the effects of sample thickness, Nickel to PDMS ratio, PDMS Young’s modulus and temperature on the piezoresistive behaviour of PDMS composites based on nanostructured surface nickel particles are fine evaluated. Experimental The nickel powder used in this work was supplied by Vale Inco Ltd. (type 123) and the polydimethylsiloxane (PDMS) was purchased by Dow Corning Corporation (SYLGARD 184). The composites were prepared by varying three parameters: the nickel filler to PDMS ratio, the copolymer to curing agent ratio and the composite thickness. The combination obtained for all the samples are summarized in Tab.1. The mixture was prepared by dispersing the metallic powder in the PDMS by gently mixing, in order to avoid the destruction of the tips on the surface of the particles. In fact it has been demonstrated that a vigorous mixing destroy the nanometric tips on the surface of the particles, strongly decreasing the piezoresistive response of the final composite [23]. The curing agent was then added to the viscous mixture and the solution was gently mixed at room temperature. The resulting paste was degassed for 1 hour, poured in PMMA molds with different hollow cavity shapes and then cured in oven at 75ºC for three hours. All the square samples realized had a footprint of 10x10 mm2 and the thickness varying from 1 to 3 mm. TABLE I.
VARIATION OF NI-PDMS RATIO AS FUNCTION OF THE THICKNESS AND PDMS COMPOSITION
Thickness
PDMS copolymer - curing agent ratio 3.33:1
10:1
1 mm
From 3:1 to 5.5:1
From 2.5:1 to 5.5:1
2 mm
From 3:1 to 5.5:1
From 2.5:1 to 5.5:1
3 mm
From 3:1 to 5.5:1
From 2.5:1 to 5.5:1
X-ray diffraction (XRD) spectrum were taken using a JEOL JDX8P diffractometer in the 2θ=10°- 80° range and the X-ray tube with Cu catode (λ=1.541 Å) was operating at 40 kV and 40 mA. Field Emission Scanning Slectron Microscopy (FESEM) images were realized with a Zeiss SupraTM apparatus with a 5kV operating voltage. Except for the evaluation of the resistance dependence on the temperature the electrical and mechanical characterizations of the composites were performed at room temperature. The apparatus was composed of a Keithley 2635A sourcemeter connected to a home-made sample holder and coupled with a universal mechanical testing machine (MTS QTest/10) with a load cell of 500 N. The samples were placed between two Cu plates used as electrodes for applying a voltage in the direction parallel to the applied uniaxial pressure. The voltage was fixed and the currents were measured coupling them with the applied load. The maximum applied load was 200 N and the test speed was 0,3 mm/min. All the operations and measurements performed by the whole apparatus were synchronized and collected with the use of a computer. Discussion and results Structural and morphological analysis XRD analyses were performed to observed the crystallinity, the level of oxidation of the spiky nickel particles inside PDMS and the possible formation of strong bond between the metal powder and the polymeric matrix. Firstly, the pure nickel powder (Type123) and on the PDMS-Nickel composites were characterized. As shown in Fig.1, the XRD patterns of both the samples exhibit the characteristic peaks of crystalline Ni (JCPDS 4-850). No shifts or broadenings were observed in the
Applied Mechanics and Materials Vols. 110-116
1339
Ni signals of the metal-polymer composite, with respect to the pure powder, indicating the absence of nickel oxide and bonds between particles and matrix. There is just a decreasing in intensity due to the presence of the PDMS that cover the surround powder. FESEM analyses were used to investigate on the morphology of the Ni powder showing the presence of the very sharp nanometric spikes on the particle surface. As shown in Fig.2, the size of a single particle is in the range between 3.5 to 7 µm (close to the 3.5-4.5 µm range quoted by the manufacturer). Fig.2 shows a relevant presence of big aggregates which did not affect the final piezoresistive behavior being dispersed during the mixing FESEM image (Fig.2) of the PDMS-Ni composite shows that the nickel particles are well uniformly dispersed in the polymeric matrix with no presence of aggregates.
Figure 1. XRD patterns of pure Ni powder and PDMS filled by Ni powder
Figure2. Upper panel: FESEM image of pure Ni powder. Lower panel: FESEM images of Ni-PDMS composite
This examination shows that the morphology of the particles has not changed significantly during the preparation of the composite, because the original sharp nanometric protrusions are kept after the process. The nickel content inside the PDMS is highly above the percolation threshold, which has to result in the presence of a consistent number of percolation paths and a high value of conductivity. In our composite the particles are intimately coated with the insulating polymer, then they do not come in physical contact, as shown in the inset of Fig.2, so that the undeformed composite samples could presents an insulating electrical behavior, even with an high filler loading. Then the mechanism of conduction results along the percolation path as a tunneling process [30]. When the composite is compressed the particles come closer, without get in touch with each other but decreasing the insulating barrier. By increasing the sample deformation, the probability of tunneling increases, and consequently the resistance decreases. Electrical and mechanical behavior The piezoresistive behavior of the composite was evaluate as function of the content of nickel respect to the polymer, of the ratio copolymer-curing agent of the PDMS and of the thickness of the prepared samples.
1340
Mechanical and Aerospace Engineering
Since all the PDMS-Ni composites realized are very soft and easily deformable materials, even at the highest filler loading, we have to take in account the dimension and shape changes under external forces. Considering that for soft materials physical quantities such as resistivity and electric field are not constant throughout the sample, piezoresistive data are presented in terms of the electrical resistance as a function of compressive tension. The variation in resistance as a function of the applied uniaxial pressure for samples with different nickel particles loading is shown in Fig.3.
Figure 3. Electrical resistance of Ni-PDMS composites as a function of compressive stress and Ni amount
Figure 4. Electrical resistance of Ni-PDMS composites as a function of compressive stress, Ni amount and PDMS composition
Figure 5. Electrical resistance of Ni-PDMS composites as a function of compressive stress, Ni amount and sample thickness
The composites present in the graph were prepared with a PDMS copolymer-curing agent ratio of 3.33:1 by weight, a thickness of 3 mm and a content of nickel filler varying from 3:1 to 5.5:1 by weight (300-550 phr) respect to PDMS. The applied pressure was varied in a very narrow range between zero and 1,5 MPa. Even if the percentage of metal particles is very high and they are very well dispersed into the polymer, as observed in the FESEM image Fig.2, there are no conductive paths into the composite. All the fillers are completely covered by the PDMS and this implies a very high resistance (more than 109 Ω) for all the samples, without any applied pressure. As can be noticed there is no variation of resistance for composite with 3:1 filler to polymer ratio, while increasing the ratio up to 3.5:1 the resistance starts to vary around 0.6MPa and decreasing up to six order of magnitude at 1.5MPa. Increasing the content of nickel the change in resistance is enhanced and the variation from the insulating state starts at lower pressure. The bigger variation of resistance was registered for the higher nickel content respect to PDMS, 5.5:1 by weight, and is more than nine orders of magnitude over the whole range of applied uniaxial pressure. We try to realize samples with even a higher concentration of the solid metallic filler, but the composite becomes difficult to process, stiff and fragile. The same analysis was performed for the composites with a PDMS copolymer-curing agent ratio of 10:1 by weight, varying the filler to polymer ratio from 2.5:1 to 5.5:1 by weight (250-550 phr). The results were quite similar with even a better sensibility to pressure.
Applied Mechanics and Materials Vols. 110-116
1341
For these samples the composite 3:1 showed a variation in resistance of about three orders of magnitude at high pressure while the 2.5:1 one remained in the insulating state all over the pressure range. The better sensitivity of this kind of PDMS composition depends on the mechanical properties of the samples. A lower quantity of PDMS curing agent implies the production of a softer composite [31], allowing a higher deformation for a given load with a large relative displacement of the conductive filler and therefore a more enhanced decrease of the insulating barrier between the spiky particles. This concept can be seen in Fig.4 that shows the dependence of the variation of electrical resistance as function of the PDMS composition. For a better clarity of the graph, we have only reported the data related to the composites with nickel filler to polymer ratio of 3.5:1 and 5:1, but the results were similar also for the other samples. As already stated the change in resistance is more pronounced for the composites with a lower content of curing agent (that has a lower Young modulus [32]) and, moreover, this effect is more enhanced at lower metallic content due to the presence of larger amount of soft polymer between the particles. In order to establish the suitable thickness that could be used in tactile sensors, we have considered the piezoresistive response of all the samples with thickness of 1, 2 and 3 mm. Fig.5 shows this dependence for samples with a PDMS copolymer-curing agent ratio of 10:1 by weight and nickel filler to polymer ratio of 3.5:1 and 5:1 by weight. In the case of the thinner samples, a higher variation in resistance was
Figure 6. Electrical resistance of Ni-PDMS composites as a function of compressive stress time variable compressive load
recorded at lower pressure, as expected. It is important to notice that the variation of the piezoresistive response as function of the thickness is not linear, as predicted by the second Ohm’s law, but it is more pronounced and can be up to three orders of magnitude for some applied pressure As discussed above, these results underline the need to refer the piezoresistive behavior of a soft sample to the resistance instead of the resistivity. The magnitude of deformation and electrical resistance is in fact strongly dependent on the initial shape, dimension and composition. Since reusability and stability of the sensing material are very important parameters in sensor application, the piezoresistive response of samples after 10 cycles of compression and decompression were evaluated. Results from the 3mm thick sample with 5.5:1 Ni-PDMS ratio and 10:1 PDMS composition are reported in Fig.6. The compression and decompression cycle was repeated for 10 times varying the pressure between zero and 2 MPa for a total measurement time of 1760 s. The curve shows that after every cycle the initial value of resistance for the insulating state is restored. This evidences the ability of the composite
1342
Mechanical and Aerospace Engineering
Figure 7. Electrical resistance of Ni-PDMS composites as a function of temperature
to have an elastic behavior with a good repeatability in recovering the initial dimension and starting particle separation. On the other hand, concerning the maximum applied pressure the value of resistance is not constant during the cycles, but it increases of quite one order of magnitude at the tenth cycle: this is probably due to hysteresis phenomena. The inset in Fig.6 shows the presence of hysteresis behaviour during the compression and decompression cycle. Fig.7 shows the dependence of resistance of the composite from the temperature. The curve is referred to a sample 2 mm thick (4:1 Ni to PDMS ratio 4:1 and copolymer to curing agent ratio 3.33:1) compressed at room temperature to an initial value corresponding to an electrical resistance of 700 Ω. It is possible to notice that the Ni-PDMS has a very high PTC (positive temperature coefficient), because the resistance increases of more than six orders of magnitude heating the sample from 25° up to 100°C. Conclusion In this work a complete characterization of the piezoresistive response of the conductive NiPDMS composite has been reported. XRD analyses indicated that there is no chemical interaction between nickel and silicone matrix. FESEM analyses demonstrate that the composite fabrication process preserves the sharp nanostructures on the surface of the Ni particles and they prove that the polymeric matrix intimately coat the conductive particles. Also for very high filler loading the undeformed samples present an insulating electrical behavior. The resistance of the composite is found to be extremely sensitive to mechanical compressive load. The electrical resistance of the composite has been evaluated as a function of several process parameters like thickness, composition of the polymer and nickel filler content. Moreover the repeatability of the piezoresistive response as a function of a variable uniaxial load was reported as well as the dependence on the temperature of the electrical resistance of the composite. The variation of the piezoresistance up to nine orders of magnitude, the simplicity of the process, and harsh environment compatibility make this quantum tunnelling composite adapted to robotic tactile sensor application. Acknowledgment The authors would like to thank Prof. Cristina Bignardi and Eng. Raffaella Sesana for the help in mechanical tests, Dr. Angelica Chiodoni for the FESEM characterization, and Eng. Diego Manfredi for the XRD analysis.
Applied Mechanics and Materials Vols. 110-116
1343
References [1] M. H. Lee and H.R. Nicholls, Review Article Tactile sensing for mechatronics--a state of the art survey. Mechatronics, 1999. 9(1): p. 1-31. [2] K. Kim, e.a., A silicon-based flexible tactile sensor for ubiquitous robot companion applications Journal of Physics: Conference Series, 2006. 34: p. 399-403. [3] M. Inaba, et al., A full-body tactile sensor suit using electrically conductive fabric and strings. Proc. IEEE/RSJ Int. Conf. Intelligent Robots and Systems, 1996: p. 450-457. [4] Y. Ohmura, Y. Kuniyoshi, and A. Nagakubo, Conformable and scalable tactile sensor skin for curved surfaces. Proceeding of 2006 IEEE International Conference on Robotics and Automation. Orlando, Florida., 2006: p. 1348-1353. [5] V. J. Lumelsky, M.S. Shur, and Sigurd Wagner, Sensitive skin. IEEE SENSORS JOURNAL, 2001. 1(1): p. 41-50. [6] Y. Kuniyoshi, et al., Dynamic Roll-and-Rise Motion by an Adult-Size Humanoid Robot. International Journal of Humanoid Robotics, 2004. I(3): p. 497-516. [7] K. Takashima, e.a., Piezoelectric properties of vinylidene fluoride oligomer for use in medical tactile sensor applications. Sensors and Actuators A, 2008. 144: p. 90-96. [8] S. Takenawa, A magnetic type tactile sensor using a two-dimensional array of inductors. IEEE International Conference on Robotics and Automation Kobe, Japan, 2009: p. 3295-3300. [9] M. Y. Cheng, et al., A flexible tactile sensing array based on novel capacitance mechanism. IEEE Transducers 2009, Denver, CO, USA, 2009: p. 2182-2185. [10] G. Harsaanyi, Polymer films in sensor applications: a review of present uses and future possibilities. Sensor Review, 2000. 20(2): p. 98-105. [11] R.H. Norman, Conductive Rubber and Plastics 1970, London: Elsevier. [12] F. G. Souza, R. C. Michel, and B. G. Soares, A methodology for studying the dependence of electrical resistivity with pressure in conducting composites. Polymer Testing, 2005. 24: p. 9981004. [13] Y.J. Yang, et al., A 32×32 temperature and tactile sensing array using PI-copper films. Int J Adv Manuf Technol, 2010. 46: p. 945-956. [14] X. Niu, et al., Characterizing and patterning of pdms-based conducting composites. Advanced Materials, 2007. 19: p. 2682-2686. [15] Q. W. Yuan, et al., Simulations on the reinforcement of poly(dimethylsiloxane) elastomers by randomly distributed filler particles. Journal of Polymer Science Part B: Polymer Physics, 1996. 34(9): p. 1647-1657. [16] V. Duchaine, et al., A Flexible Robot Skin for Safe Physical Human Robot Interaction IEEE International Conference on Robotics and Automation Kobe, Japan., 2009: p. 3676-3681. [17] M. Shimojo, et al., A system for simultanious measuring grasping posture and pressure distribution. Proc. IEEE Int. Conf. Robotics and Automation, Nagoya, Japan, 1995: p. 831-836. [18] K. Weiss and H. Wörn, The Working Principle of Resistive Tactile Sensor Cells. Proceedings of the IEEE International Conference on Mechatronics & Automation Niagara Falls, Canada 2005: p. 471-476. [19] R. Strumpler and J. Glatz-Reichenbach, Conducting Polymer Composites. Journal of Electroceramics, 1999. 3(4): p. 329-346.
1344
Mechanical and Aerospace Engineering
[20] F. Carmona, Conducting filled polymers. Physica A: Statistical Mechanics and its Applications, 1989. 157(1): p. 461-469. [21] W. Luheng, D. Tianhuai, and W. Peng, Effects of conductive phase content on critical pressure of carbon black filled silicone rubber composite. Sensors and Actuators A: Physical, 2007. 135(2): p. 587-592. [22] L. K. H. Beek and B. I. C. F. van Pul, Internal field emission in carbon black-loaded natural rubber vulcanizates. Journal of Applied Polymer Science, 1962. 6(24): p. 651-655. [23] D. Bloor, et al., A metal–polymer composite with unusual properties. J. Phys. D: Appl. Phys., 2005. 38: p. 2851-2860. [24] F. G. Chang, et al., Enhanced piezoresistivity in Ni–silicone rubber composites. Chinese physics B, 2009. 18(2): p. 652-657. [25] X.W. Zhang, Z. P.Y, and X Q. Yi, Time dependence of piezoresistance for the conductor-filled polymer composites. J Polym Sci Part B: Polym Phys, 2000. 38(21): p. 2739-2749. [26] W. Luheng, D. Tianhuai, and W. Peng, Influence of carbon black concentration on piezoresistivity for carbon-black-filled silicone rubber composite. Carbon, 2009. 47(14): p. 31513157. [27] UK, P.L., Patent PCT/GB98/00206 (WO 98/33193). [28] C. J. Edgcombe and U. Valdrè, Microscopy and computational modelling to elucidate the enhancement factor for field electron emitters. Journal of Microscopy, 2001. 203: p. 188-194. [29] M.K. Abyaneh and S.K. Kulkarni, Giant piezoresistive response in zinc–polydimethylsiloxane composites under uniaxial pressure. J. Phys. D: Appl. Phys., 2008. 41: p. 135405. [30] D. Toker, et al., Tunneling and percolation in metal-insulator composite materials. Phisical Review B 2003. 68(041403). [31] F. M. Sasoglu, A. J Bohl, and B. E. Layton, Design and microfabrication of a high-aspect-ratio PDMS microbeam array for parallel nanonewton force measurement and protein printing. J. Micromech. Microeng. 17, 2007. 17: p. 623-632. [32] F. M. Sasoglu, A J Bohl, and B. E. Layton, Design and microfabrication of a high-aspect-ratio PDMS microbeam array for parallel nanonewton force measurement and protein printing. J. Micromech. Microeng., 2007. 17: p. 623-632.
