Development of a Large-Area 8 × 8 Tactile Sensing Array with High Sensitivity

In this work, a large-area tactile sensing array with high sensitivity is reported. The sensing elements of the array consist of multiwall carbon nanotubes dispersed in polydimethylsiloxane polymer. A novel fabrication process using a universal applicator is proposed for realizing largearea conducting polymer films. In addition, a nylon membrane filter is employed as a mold for creating microdome patterns, which are essential for increasing the sensitivity. The major advantages of the proposed device include ultrahigh sensitivity, flexibility, and a simple fabrication process. The sensing mechanism involves utilizing the contact resistance between the electrode and the conductive polymer film. A demonstration of the application of the tactile sensing array is presented.


Introduction
The development of tactile sensors has received attention because of their potential applications in many areas such as robotics, (1,2) biomedical devices, (3) and artificial skin. (4)(5)(6) Tactile sensors are one of the essential sensing mechanisms for providing information from the interaction between robots and both their environment and human beings. During the past decade, many novel designs of tactile sensors have been proposed. Zhong et al. reported a paper-based active tactile sensor array. (7) The performance of the proposed device remained satisfactory when parts of the array were removed, which indicates its potential application for customized electronic skins. Kolesar et al. realized a robotic tactile sensor by assembling a polyvinylidene fluoride (PVDF) polymer film with a silicon integrated circuit. (8) Yang et al. presented a system based on graphene woven fabrics (GWF) for highly sensitive strain sensing. (9) The sensors can be employed for applications such as human motion detection, sound signal acquisition, and external stress distribution monitoring. Saccomandi et al. presented a fiber-optic tactile sensor based on fiber Bragg grating (FBG) technology. (10) The static metrological properties and relatively high spatial resolution make the proposed device attractive for robotic hands.
Furthermore, tactile sensing elements embedded with microstructure arrays, which enable high sensitivity and fast response, have received attention in recent years. Su et al. proposed a bioinspired flexible pressure sensor with high sensitivity. The microstructures on the sensor substrates were created by the direct molding of natural mimosa leaves. (11) Zhang et al. developed flexible temperature-pressure sensors based on microstructure-frame-supported organic thermoelectric materials. (12) Obviously, the embedded microstructures on the sensing elements play an important role in providing excellent performance characteristics such as high sensitivity and a fast response. However, special and expensive molds have usually been needed to realize these microstructured arrays.
In our previous work, we proposed a novel lithography process for realizing a tactile sensor array using a mold made of SU-8 and for transferring microstructures from nylon membranes onto sensing elements. (13) In this work, we present a new process for fabricating a large-area tactile sensing array with microdome structures, and a new electrode configuration which simplifies the device assembly. The proposed device is designed for a sensing system deployed on the handlebar of a robot that assists humans in walking. A fabrication process is proposed to form conducting polymer films with large areas by using a universal applicator for transferring microdome structures from nylon membrane filters onto polymer films. The proposed sensing array is assembled by combining the conductive polymer film with a flexible printed circuit board (FPCB) with arrays of interdigital electrode pairs. The characteristics and uniformity of the large-area conductive polymer sheet are also investigated. Figure 1(a) shows a schematic of the proposed sensing array. The device consists of three layers: a PE film, a conductive polymer with microdome structures, and an FPCB with an array of interdigital electrodes. Figure 1(b) shows the cross-sectional view of a sensing element. Each sensing element includes a pair of interdigital electrodes and the conductive polymer above the electrode pair.

305
The operating principle of the proposed tactile sensing array is shown in Fig. 1(c). As an external pressure is applied, the contact area between the microdome structures and the interdigital electrodes increases sharply, which in turn rapidly reduces the resistance between the polymer and the electrode. (14) The change in resistance can be easily measured via the interdigital pairs. This type of tunneling piezoresistive device is much more sensitive (at least two orders of magnitude higher sensitivity) than the pressure-sensing devices fabricated with typical conductive polymer materials without microstructures. (14) In addition, the sharp change in the contact area due to pressure-induced deformation of the microdome structures gives a much more rapid response (about 10 times faster) in the form of a change in resistance than a device using a typical conductive polymer material. Hexane serves as the dispersant during blending. To cure the prepolymer, the mixture is mixed with a curing agent (Sylgard 184 B, Dow Corning) in a 1:10 ratio and is stirred for 50 min. Then, the prepolymer is degassed in a vacuum chamber for 3 h to evaporate the volatile hexane, (15) after which it is ready for patterning microstructures using an applicator (Zehntner Testing Instruments, ZUA 2000.220).

Polymer film fabrication
The fabrication process of the conductive polymer film is shown in Fig. 3. A nylon membrane filter (Finetech Research and Innovation Co., Nylon Membrane, M-Nylon 5.0I) is placed on a glass plate [ Fig. 3(a)]. The prepared prepolymer is dispensed on the membrane [ Fig. 3(b)] using a computer-controlled dispenser. By squeezing the dispensed prepolymer using the applicator, a conductive polymer film is formed [ Fig. 3(c)]. The film is held at room temperature for 12 h for   Fig. 3(d)]. Then, the film is cured at 50 °C for 3 h using a hot plate [ Fig. 3(e)]. The conductive polymer film is then peeled from the membrane filter [ Fig. 3(f)].
On the polymer surface, numerous microdome structures transferred from the filter membrane can be observed. A scanning electron microscopy (SEM) image of the surface of the nylon membrane filter is shown in Fig. 4(a). The figure clearly shows numerous filtering pores on the surface. Figure 4(b) is a SEM image of the conductive polymer film. The microdome structures, whose average base width is about 7 μm, were transferred from the membrane filter.

Results and Discussion
The experimented setup for characterizing the conductive polymer film is shown in Fig. 5. A force gauge (HF-1, ALGOL Engineering Co.) was fixed on a vertical (z-axis) translational stage that had a displacement resolution of 1 μm. The maximum resolution of the force gauge was 1 mN. A polymethylmethacrylate (PMMA) cylindrical rod was attached to the sensing probe of the force gauge. The PMMA rod was moved downward by the vertical translational stage. Normal forces on the conductive polymer film were induced as the PMMA rod came in contact with the conductive polymer film. The change in the resistance of the polymer film was measured using a source meter (Model 2400, Keithley Instruments). The dimensions of the conductive film were 30 × 22 cm 2 . The film was partitioned into 16 regions (i.e., a 4 × 4 array), and the resistance of the center of each region was measured. During the measurement of each region, an interdigital electrode pair (on a 5.5 × 3.5 cm 2 FPCB) was placed under the region of the polymer film. Figure 6(a) shows the measured relationships of electrical resistance versus applied pressure in different regions. Figure 6(b) shows the normalized resistance for these regions. Note that in the figure Rf is the final resistance in each measurement for each region. These results indicate that the variation between the regions near the edge of the film and the regions around the center is relatively large. Figure 7 shows a histogram of variations at the 64 sensing locations under two uniform pressures of 1.2 and 1.8 kPa. In this case, the polymer film was assembled with an FPCB with an 8 × 8 array of interdigital pairs. There were 64 sensing locations in total. A resistance value can be easily retrieved using a scanning circuit. (16) The abscissa in Fig. 7 is the resistance variation (ΔR) of a sensing location with respect to the average measured resistance of the 64 sensing locations (R A ) under a specific uniform pressure force. Figure 8(a) is a photograph of an assembled tactile sensing array. The sensing array can be fixed on a robot surface by attaching the array using a typical thin stick tape. Figures 8(b) and 8(c) show the images of the force induced by pressing the 8 × 8 sensing array with one hand. The image of the measured force shows that the shape of the hand is clearly resolved.

Conclusions
In this report, we presented the development of a large-area, highly sensitive tactile sensing array using a conductive polymer film embedded with microdome structures. A novel fabrication process was proposed for realizing large-area conducting polymer films using a nylon membrane filter and an applicator. The uniformity of the fabricated conductive polymer sheets was also investigated. The characteristics of the proposed device were measured and studied. A demonstration of the sensing capability of the array was also presented.