A Polyimide Based Force Sensor Fabricated Using Additive Screen-Printing Process for Flexible Electronics

A flexible capacitive force sensor was developed for force sensing applications. The force sensor consists of two electrodes and a dielectric layer. The electrodes were fabricated by depositing conductive silver (Ag) ink on a flexible polyimide platform using screen-printing process. The dielectric layer was prepared by mixing the contents of polydimethylsiloxane (PDMS) (pre-polymer and curing agent) in a 16:1 ratio. Then the PDMS dielectric layer was sandwiched between the screen-printed Ag electrodes. The capability of the fabricated force sensor was investigated by recording its capacitive response for varying applied forces of 100 N. It was observed that the capacitance increased from 5.83 pF to 6.36 pF as the force was increased from 0 N (no load) to 100 N. A sensitivity and correlation coefficient of 0.081%N−1 and 0.998 were calculated for the force sensor. In addition, a response time and recovery time of 3.7 seconds and 5.7 seconds was measured for the fabricated force sensor. The relative humidity (% RH) tests performed from 20% RH to 80% RH, in steps of 20% RH, revealed that there was a minimal effect of RH on the base capacitance of the force sensor at room temperature.


I. INTRODUCTION
In recent years, the need for the development of flexible force sensors has been increased due to its wide range of application in infrastructure monitoring, robotics, electronic skins, health care and diagnostic purposes, and smart displays [1]- [5]. Piezoelectricity, piezo resistivity, and capacitance are the three commonly employed force sensing mechanisms. Among them, capacitance-based sensing mechanisms are widely preferred due to its multiple advantages including relatively less power consumption, repeatability of the responses, less prone to humidity and temperature The associate editor coordinating the review of this manuscript and approving it for publication was Gautam Srivastava . changes, and less susceptibility to noise [6]- [9]. The force sensors are typically fabricated using traditional/conventional methods that often employs rigid platforms with either hanging or cavity structured designs [10]- [15]. Thus, these sensors do not offer the required flexibility and conformability for the emerging smart wearable industry, specifically e-skins [10]- [15]. In addition, these force sensors are not suitable for asset/structural health monitoring of aircrafts or any civil structures due to the limitations associated with conformability as well as less resistance towards harsh environments including variations in humidity levels. These drawbacks can be alleviated by exploring some novel fabrication techniques such as additive print manufacturing processes that can enable the development VOLUME 8, 2020 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ of relatively cost-effective flexible and conformal force sensors.
Additive print manufacturing processes such as screen, inkjet, flexography and gravure provides numerous advantages for the fabrication of various electronic devices including device flexibility, conformability, low wastage of materials as well as processing temperatures, relatively fewer and simpler fabrication steps [16]- [31]. In addition, these processes have roll-to-roll (R2R) capability that enables the large-scale manufacturing of functional devices in a relatively short time and this reduces the production cost [32]- [34]. Among the additive manufacturing process, screen printing process is the most reliable and durable deposition method with high yields over longer print runs providing relatively thicker ink layers [32].
The working principle/mechanism of a capacitive force sensor is same as a parallel plate capacitor and can be mathematically expressed using Eq. (1), where C -capacitance of the force sensor, ε r -dielectric constant of dielectric layer, ε 0 -dielectric constant of vacuum, A -overlapping electrodes area, d -distance between the electrodes (thickness of the dielectric layer). When an external force is perpendicularly applied on the force sensor, the dielectric layer gets compressed (the thickness of the dielectric layer decreases) and this results in a proportional capacitance change.
There are many studies performed on the development of flexible force sensors using dielectric layers made of polydimethylsiloxane (PDMS) and liquid metals, but the sensitivity of these sensors are relatively low [35]- [37]. Typically, the PDMS dielectric layers employed in these sensors use a standard fabrication method in which the pre-polymer and curing agent of the silicone elastomer kit are mixed in a 10:1 (w/w) ratio. However, if the hardness of the PDMS can be decreased, this can increase the compressibility of the PDMS and results in higher relative capacitance change and increases the sensitivity of the force sensor. The hardness of the PDMS can be decreased by increasing the ratio of pre-polymer content to curing agent during the fabrication process. A PDMS dielectric layer prepared by mixing the contents of silicone elastomer kit, pre-polymer and curing agent in a 16:1 (w/w), will provide high stress to strain ratio and this will result in obtaining higher relative capacitance change [38].
Recently, force/pressure sensors have been fabricated using plastic materials (polyethylene terephthalate (PET)), conductive materials (graphene, copper, carbon nanotubes) and dielectric materials (PDMS) with pores or micro-structures using 3D templates and laser machining [39]- [44]. Even though the sensitivity of some of these sensors is typically high, the structures in the dielectric layers are prone to collapse or break easily and are not suitable for high force measurements. In addition, the fabrication of these sensors is complex and difficult to obtain the uniformity in pores or microstructures resulting in low fabrication yields and inconsistent output performances from the sensor to sensor. The asset/structural health monitoring of aircrafts or any civil structures requires force sensors that are not only conformable and resistant to environmental changes (variations in humidity levels), but also be able to detect high forces while being robust, stable and compatible for mass production with high fabrication repeatability.
In this work, a flexible capacitive force sensor was developed by using a PDMS based dielectric layer prepared by mixing pre-polymer and curing agent in a 16:1 ratio. The top and bottom electrodes were fabricated by depositing silver (Ag) ink on a polyimide substrate using screen printing process. The flexible polyimide substrate was used as substrate for the electrodes since it provides relatively higher thermal stability, tensile strength, chemical resistance as well as excellent mechanical properties when compared to the mostly used other plastic based materials such as PET (force sensor fabricated on PET was demonstrated in our previous work) [45], [46]. The capability of the fabricated highly robust force sensor to detect varying applied forces was demonstrated by investigating its capacitive response. In addition, response and recovery time as well as the effect of varying relative humidity (% RH) and temperature levels on the sensor was investigated and is presented.

A. CHEMICALS AND MATERIALS
Ag conductive ink (Ag 800) from Kayaku Advanced Materials, Inc. was used for fabricating the electrodes. A 127 µm thick polyimide film, Kapton R 500HN from Dupont TM was used as the substrate. Computer numerical control (CNC) machined mold made of aluminium, and Sylgard R 184 PDMS silicone elastomer kit from Dow Corning was used in fabricating the dielectric layer. A 100 µm thick double-sided polyimide tape consisting of 25 µm thick polyimide film and 38 µm of silicone adhesive coating on each side of polyimide film from BERTECH R was used to attach the top and bottom electrodes to the PDMS dielectric layer. A flat flexible connector (FFC) with crimp-based contact termination, from TE Connectivity AMP Connectors (1-88997-2) and male to female jumper wires (PRT-12794) from SparkFun Electronics were used to connect the sensor to an Agilent E4980A precision LCR meter.  1320 thermal oven. The PDMS was prepared by adding the Sylgard R 184 pre-polymer to a curing agent (in a 16:1 (w/w)) and stirring in Thinky AR-100 planetary centrifugal mixer for 3 minutes at 1500 rpm. This resulted in a homogenous mixed PDMS without any bubbles. Following this, the bubble-free PDMS was poured into the 15 mm diameter aluminium mold and annealed for 10 minutes at 110 • C in the VWR 1320 thermal oven. Then, the electrodes (top and bottom) ( Fig. 1(c)) and the PDMS dielectric layer (thickness 0.6 mm) ( Fig. 1(d)) of the force sensor were then patterned by laser cutting (Universal Laser Systems -PLS6MW) the screen printed samples and PDMS. Finally, the PDMS based dielectric layer with a dielectric constant of 2.65 was sandwiched between the Ag based top and bottom electrodes, using a double-sided polyimide tape to form the robust force sensor as shown in Fig. 1(e) [38].
A thickness of 4.46 µm and roughness of 0.8 µm was measured for the screen-printed Ag layer on polyimide substrate using a Bruker Contour GT-K white light interferometer (5X lens). The 3D output of the white light vertical scanning interferometry with thickness and roughness measurements are shown in Fig. 2(a) and Fig. 2(b), respectively. The print quality of the Ag electrode was examined using a JEOL JSM-IT500HR InTouchScope TM scanning electron microscope (SEM). The SEM micrograph of the Ag electrode was captured in a low vacuum mode (30 Pa) using an accelerating voltage of 20 keV, emission current of 90 µA and backscattered electron detector at a working distance of 6.8 mm. As shown in Fig. 2(c), it is evident that the Ag flakes are homogenously distributed on the polyimide film and connected to each other without any cracks or voids that influences the electrode conductivity.

C. EXPERIMENT SETUP
The experiment setup for characterizing the sensor is shown in Fig. 3(a). The fabricated capacitive force sensor was placed between the clamp holders of a Mark-10 model M5-200 force gauge and Mark-10 ESM 301 motorized test stand. Varying forces up to 100 N were applied perpendicularly to the force sensor using automated Mark-10 force gauge. Then, the sensor with contact crimp socket connectors was connected to an Agilent E4980A precision LCR meter using alligator clips for measuring the capacitive responses of the sensor. An operating voltage and a frequency of 1 V and 1 kHz was applied to the force sensor for capacitive measurements. The LCR meter was connected to a PC with a custom-built LabVIEW TM program via a USB cable to automatically record the real-time capacitive responses of the force sensor. Similarly, for measuring the effect of humidity on the force sensors, a Thermotron R SE 3000 environmental chamber was used as shown in Fig. 3(b). The force sensors were placed in the environmental chamber and the % RH levels inside the chamber was programmed to vary from 20% RH to 80% RH, in steps of 20% RH at 23.5 • C. In addition, the temperature inside the chamber was increased from 55 • F to 73 • F to 85 • F for investigating the effect of temperature on the sensor. The sensor responses were sequentially recorded by an Instek 6100 LCR meter using Keithley 2700 mainframe with 7700 multiplexer. The multiplexer and LCR meter were connected to a desktop with MATLAB program via general-purpose interface bus (GPIB) and serial RS232 cables, respectively. Figure 4 shows the dynamic capacitive response of the sensor from 0 N to 10 N, in steps of 2 N and 20 N to 100 N, in steps of 20 N. Initially, no force was applied on the force sensor and its base capacitance value was recorded. Then, the force 207816 VOLUME 8, 2020 sensor was subjected to the varying applied forces up to 100 N. It was observed that the capacitance increased from ∼5.83 pF to ∼5.85 pF to ∼5.87 pF to ∼5.89 pF to ∼5.91 pF to ∼5.94 pF to ∼6.03 pF to ∼6.11 pF to ∼6.19 pF to ∼6.27 pF to ∼6.36 pF as the applied forces were increased from 0 N to 100 N, respectively ( Fig. 4(a)). These results correspond to a relative capacitance change of ∼0.3 %, ∼0.7%, ∼1.0 %, ∼1.4 %, ∼1.8 %, ∼3.0 %, ∼4.7%, ∼6.4 %, ∼7.7 %, ∼9.5 % for 0 N to 100 N, when compared to the base capacitance value (Fig. 4(b)). A linear response with a sensitivity of 0.185%N −1 and 0.081%N −1 as well as a correlation coefficient of 0.998 were obtained for the force sensors subjected to applied forces 0 N to 10 N and 20 N to 100 N, respectively. It is worth noting that the fabricated force sensor attains its base value after removing the applied forces (six sensors; each three times tested). The response time of the force sensor was measured by applying a force of 100 N using the Mark-10 force gauge ( Fig. 5(a)). The capacitance of the sensor increased from 5.82 pF to 6.34 pF in 3.7 seconds. Similarly, the recovery time of the force sensor was measured when the applied force of 100 N was removed. This resulted in a recovery time of 5.7 seconds and the capacitance decreased from 6.34 pF to 5.81 pF (Fig. 5(b)). Then the hysteresis of the force sensors was investigated by loading and unloading the applied forces from 0 N to 100 N using Mark-10 force gauge (Fig. 6). The  hysteresis of the sensor was mathematically calculated using Eq. (2) [47].

Max Hysteresis
where fsd -full-scale deflection, O and I is the output (y-axis) and input (x-axis) values of the sensor, respectively. A maximum hysteresis of 3.2% was calculated at the force of 40 N. This indicates that the force sensor has relatively better recovery and elasticity characteristics. Further, the stability of the sensor was investigated by storing the sensor samples (over 2 months) in a Plas-Labs 860-CGA acrylic vacuum cabinet with desiccants. Fig. 7 shows the stability results of the sensor for varying applied pressures (0 to 100 N) before and after the two-month time duration. A maximum change of 0.5% was calculated at 20 N, before and after the two-month time period, indicating that the fabricated force sensor is very stable for over two months. In addition to hysteresis and stability tests, the effect of humidity on the sensor performance was investigated by varying % RH levels from 20% RH to 80% RH, in steps of 20% RH at 23.5 • C in a programmable Thermotron R SE-3000 environmental chamber (Fig. 8). It was observed that the base capacitance of the force sensor increased from 5.83 pF to 5.86 pF due to accumulation of parasitic or stray capacitances in the connecting cables. The varying humidity levels had minimal effect on the sensor response. Apart from humidity, the effect of temperature on the sensor was also investigated by varying the temperature of the environmental chamber from 55 • F (12.8 • C) to 73 • F (23 • C) to 85 • F (29.8 • C). It was observed that the capacitance increased from 5.81 pF to 5.84 pF to 5.86 pF when the temperature of the environmental chamber was increased from 55 • F to 73 • F to 85 • F, respectively and affects the capacitance measurements of forces below 10 N (Fig. 9). This effect can be attributed to the thermal expansion coefficient of the PDMS dielectric layer that leads to micro-thermal deformations in the sensor [48], [49]. The temperature effect on the force sensor response can be mitigated using well known reference gauge compensation techniques [49], [50]. In this technique, a reference force sensor (same as the printed force sensor) placed under the same ambient conditions of force sensor serves as a reference capacitor. The reference force sensor is not subjected to external applied forces. During the capacitance measurements, the absolute capacitance value of the reference force sensor is deducted from the printed force sensor at any given time (under both loading and unloading of forces) to alleviate/mitigate the effect of temperature.
A comparison table of some recently reported force sensors along with the current work is provided in Table 1. It is worth noting that the printed force sensor has better sensitivity and stability when compared to some recently reported force sensors. The results thus demonstrated the feasibility of employing a polyimide based flexible capacitive force sensor that was fabricated using additive screen-printing process and 16:1 PDMS as an efficient and cost-effective way to monitor forces in wide variety of applications.

IV. CONCLUSION
In this work, a capacitive force sensor was successfully fabricated on a flexible polyimide platform. The top and bottom electrodes of the sensor were fabricated by screen printing Ag ink on polyimide substrate. A PDMS based dielectric layer was prepared by mixing pre-polymer and curing agent in 16:1 (w/w) ratio. The force sensor was assembled by attaching the top and bottom electrodes to the PDMS dielectric layer using a double-sided polyimide tape. An increase in capacitance from 5.83 pF to 6.36 pF was observed when the applied forces were increased from 0 N to 100 N resulting in an overall relative capacitance change of 9.5%. A sensitivity of 0.081%/N −1 , a response time of 3.7s and a recovery time of 5.7s was obtained for the force sensor. Future work includes implementation of reference gauge compensation technique to mitigate the effect of temperature on sensor as well as investigate the mechanical stresses such as bending on the performance of the force sensor.

ACKNOWLEDGMENT
The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation thereon. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of Air