E-Skin Pressure Sensors Made by Laser Engraved PDMS Molds ^†

Abstract

This work describes the production of electronic-skin (e-skin) piezoresistive sensors, which micro-structuration is performed using laser engraved molds. With this fabrication approach, low-cost sensors are easily produced with a tailored performance. Sensors with micro-cones and a high sensitivity of −1 kPa⁻¹ under 600 Pa are more adequate for the blood pressure wave detection, while sensors micro-structured with semi-spheres and a maximum sensitivity of −6 × 10⁻³ kPa⁻¹ in a large pressure range (1.6 kPa to 100 kPa) are more suitable for robotics and functional prosthesis.

1. Introduction

Human skin is an inspiring organ due to its outstanding performance, namely in the perception of several mechanical stimuli [1]. Electronic skin (e-skin) is a flexible and stretchable electronic surrogate of skin with sensing properties [1] that shows an enormous potential for human health monitoring, humanoid robots, and functional prosthetic devices [2,3,4,5]. E-skin pressure sensors use some distinct effects for pressure transduction [6,7,8], but the attractive simplicity of piezoresistivity has fuelled a great research in the field [9,10,11,12]. When playing on contact resistance change, the micro-structuration of sensors’ films increases the contact area between them, hence increasing the R_OFF/R_ON ratio [12] and thus improving the sensitivity. Sensors typically have a higher sensitivity for low pressures and a significantly lower value for higher pressures [2,6,10,11,12,13]. For the blood pressure wave detection at the wrist, a sensor should have a high sensitivity below 400 Pa [9]. For robotics and prosthesis, a constant sensitivity in a range from less than 10 kPa (gentle touch) to 100 kPa (objects manipulation) is preferable [6]. Also, the majority of micro-structured sensors either use costly and time consuming photolithography techniques [2,6,8,9,11,12] or low customizable techniques based on using objects as molds [10,13] for the micro-structuration. Hence, there is a need for a cheap and easy-to-produce pressure sensor with a high and constant sensitivity in a wide pressure range. Herein, an easy and fast production of highly customizable molds through laser engraving is explored, and the versatility of this technique is demonstrated through the production of piezoresistive e-skin sensors with different performances which makes them more suitable for health monitoring applications or robotics and functional prosthesis.

2. Materials and Methods

The sensors were produced following a methodology described in a previous work [14]. The engraved patterns consisted in crosses or circles with a width/diameter of 100 µm or 200 µm and a pitch of 150 µm, 200 µm, or 300 µm, over 4 cm². The molds were either hard-PDMS (h-PDMS, 1:5 w/w ratio of curing agent to elastomer) for the engraving of circles or acrylic plates for the engraving of crosses. The laser speed was set to 0.1524 m s⁻¹ (for acrylic) or 0.254 m s⁻¹ (for h-PDMS), with a laser power from 2.5 W to 25 W. Micro-structured standard-PDMS (s-PDMS, 1:10 w/w ratio) films were observed with a tabletop scanning electron microscope (SEM) (Hitachi TM3030Plus). Several weights were placed on the sensors to apply pressures from 15 Pa to 100 kPa, under 1 V to 10 V. The relaxation time was estimated with the unloading of a small magnet (184 Pa). Semi-spheres sensors were mounted in a robotic arm, controlled by a servomotor, to monitor the pressure exerted during the grasping of an object, which was performed 2500 times. Sensors with cones were tested at the wrist for the monitor of the blood pressure pulse wave.

3. Results and Discussion

3.1. h-PDMS Patterning

The effect of laser power on h-PDMS molds was studied by varying the power from 25 W to 12.5 W, 7.5 W, and 2.5 W, while fixing the laser speed at 0.254 m/s for a design based on circles with a pitch of 200 µm. Figure 1 shows that the height of the s-PDMS micro-structures peeled off from such molds increases with the laser power and also with the diameter, because while the laser is engraving a larger area, it melts more h-PDMS in depth. To achieve a semi-sphere like microstructure, a laser power of 7.5 W and a designed diameter of 200 µm should be used because the resultant micro-structures are rounded and their ratio of height/diameter is the closest to 0.5.

3.2. Characterization of Micro-Structured Films Peeled off from h-PDMS Molds

Figure 2a–d show a general view of s-PDMS films with semi-spheres, where the homogeneity of the micro-structures and their correct alignment is evident. Figure 2e–h prove that PMMA and carbon coating seem to spread homogeneously over the semi-spheres surface. Additionally, the PMMA layer is thin enough to the point of having a null impact in the features of the semi-spheres (height, diameter, and pitch), as shown in the bottom of Figure 2.

3.3. Electrical Characterization of Sensors

A schematic illustration and a photograph of a semi-spheres sensor are shown in Figure 3a–b, respectively. These sensors may be subjected to bending and compressive forces without an impairment of their performance. Figure 3d illustrates the relative resistance change for cones sensors (produced with acrylic molds engraved with crosses) and semi-spheres sensors (produced with h-PDMS molds engraved with circles with the optimized parameters from Section 3.1) with different pitches, and the respective sensitivities (calculated as done elsewhere [12,13]) are presented in Figure 3e. Cones sensors generally have higher sensitivities for lower pressures, with a maximum of −1 kPa⁻¹ under 600 Pa for a pitch of 300 µm. However, semi-spheres sensors were tested until higher pressures and revealed a maximum sensitivity of −6 × 10⁻³ kPa⁻¹ from 1.6 kPa to 100 kPa when having a pitch of 200 µm.

The relaxation time for semi-spheres sensors was estimated as shown in Figure 3f, and the impact of the PMMA layer on this parameter was also evaluated. The PMMA layer seems not only to reduce the relaxation time of the sensors but also to smooth the sensors’ behavior during relaxation. Contrarily, sensors without PMMA have a peak before stabilizing their resistance.

To study the reproducibility and effectiveness of the sensors, semi-spheres sensors were subjected to a cyclic compression (2500 cycles) done with a robotic arm upon grabbing an object (the setup of the test is shown in Figure 3c and the results are in Figure 3g) while cones sensors were placed at the wrist to monitor the blood pressure wave (Figure 3h). The semi-spheres sensor presents a reproducible response, with a similar relative resistance change for the first and last cycles of (− 33 ± 1)% and (− 32 ± 1)%, respectively. The relative current change of the cones sensor placed at the wrist is coincident with the shape of the blood pressure wave.