Flexible capacitive pressure sensor based on interdigital electrodes with porous microneedle arrays for physiological signal monitoring

Because they can provide real-time feedback by converting pressure signals into detectable electrical signals, flexible pressure sensors are favored in various applications, including electronic skin, intelligent medical detection, smartphones, human–computer interaction systems, and intelligent robots.^1–8 The sensing mechanisms of sensors are classified into several categories, and these include capacitive,^9,10 piezoresistive,¹¹ and piezoelectric¹² approaches. Capacitive pressure sensors have attracted wide attention due to their advantages of simple processing, low cost, and low power consumption.¹³ They can be used to collect information about human respiration, pulse, and joint movements due to their ability to detect subtle pressure changes, their ease of integration into textiles, and their strong biological compatibility.⁸ However, compared with piezoresistive and piezoelectric sensors, the sensitivity of capacitive pressure sensor is lower, and this is a problem that requires an urgent solution.

In general, capacitive pressure sensors consist of two electrode layers with a dielectric layer sandwiched between them.^9,10 When pressure is applied to these electrode layers, this causes the sensor to deform,^14–16 changing the distance between the two and affecting the capacitance value. The composition of the dielectric will also greatly affect the dielectric coefficient and will thus directly affect the capacitance. Therefore, improving the sensitivity of capacitive pressure sensors can be considered from two perspectives. In the first, the structure of the dielectric layer can be modified to improve the deformation capacity of the sensor. In the second, the composition of the dielectric can be modified to increase its dielectric coefficient.^17,18 Transformation of the dielectric layer structure can be accomplished by preparing a microneedle array (MNA)^19–22 or constructing a microporous air-gap structure.

Luo et al.¹⁴ used photolithography to prepare a sensor based on the bending deformation of an inclined microcolumn array instead of the compression deformation of a traditional vertical structure. Shi et al.²³ prepared a composite dielectric by solvent-free planetary mixing to improve the dielectric constant, and they solidified a mold to obtain a microneedle structure, resulting in a sensitivity of 3.97 × 10⁻³ kPa⁻¹ within the range 0–1.3 MPa. Zhao et al.²⁴ prepared microtower arrays inspired by the stomatal structure on the surface of a lotus leaf; the sensing performance of this device remained unchanged after 4000 repeated tests. Zhang et al.²⁵ used the surface micropores of rose petals to prepare a bilayer biomimetic microstructure to form a dielectric layer. Li et al.²⁶ prepared a pressure sensor using an electrostatically spun thermoplastic polyurethane nanofiber film in the middle as the dielectric and conductive fabric on the upper and lower sides as the electrodes, thus forming a gap to improve the sensitivity (0.28 kPa⁻¹). Yu et al.²⁷ used multi-walled carbon nanotubes (MWCNTs) as a conductive material and a composite film of barium titanate (BaTiO₃) with polydimethylsiloxane (PDMS) as an intermediate dielectric layer. Atalay et al.⁹ used sugar particles and salt crystals as sacrificial fillers to mix with silicone, followed by dissolving and curing to create voids.

As illustrated in Table I, most of these studies improved the sensitivity of the devices by optimizing the dielectric layer structure and increasing the numbers of voids, but the optimization of the conductive electrode has so far been ignored. This means that the electrodes of the sensor are subject to being affected by the environment, thus reducing the stability and the transmission efficiency of the sensor signal. In this work, two porous PDMS substrates with MNA structures were first prepared using a mold. Then, parylene–Cr–Au–parylene layers were deposited successively to serve as conductive electrodes, in which the two layers of parylene are used to enhance adhesion and insulate the conductive layer. Two porous MNA electrodes with the same structure were spin-coated and bonded with the porous PDMS to form an interdigital electrode structure. This porous MNA interdigital electrode structure can enhance the deformation of both the electrodes and the dielectric and completely insulate the conductive layers inside the substrates, improving the sensitivity of the sensor and reducing interference from the external environment.²⁸ Compared to previous approaches, this work presents a novel method for improving the sensitivity and stability of flexible capacitive pressure sensors by modifying the electrode layer rather than the dielectric layer. Although the sensors presented in Refs. 14, 26, and 27 are slightly more sensitive, their pressure-measurement ranges are narrower. Therefore, if the factors of measuring range, sensitivity, and stability are considered at the same time, the present approach has notable advantages.

PDMS (Sylgard 184 Silicone elastomer base) was obtained from Dow Corning Corporation (USA). The PDMS mold was obtained from Microchip Medicine Co., Ltd. (China). The Parylene powder was obtained from the Suzhou Kerui Nano Company (China). Solid NaHCO₃ powder was obtained from Sinopmedicine Group Chemical Reagent Co., Ltd. (China). The conductive silver paste was obtained from Xinwei Xincai Co., Ltd. (China). The ultrasonic mixer (HZ-TB28-250) was obtained from Hongzhen Co., Ltd. (China). The magnetron sputtering machine (PVD-150A) was acquired from Beijing Sinotron Electronics Co., Ltd. (China). The parylene CVD coating machine (KR350HSU) was acquired from Suzhou Kerry Nano Co., Ltd. (China). The glue applicator (TL-6A) was obtained from Beijing ZhongCotyrone Electronics Co., Ltd. (China). The metallographic microscope (DM2700M) was obtained from Leica (Germany). The electronic universal testing machine (P300) was from Dongri Instrument Co., Ltd. (Taiwan). Impedance analyzer (Keysight E4990A) was purchased from Zide Technology Company (USA).

The fabrication process for the porous MNA interdigital-electrode-based pressure sensor is shown in Fig. 1(a). First, a 2.2-μm-thick layer of parylene is deposited on the surface of the PDMS mold for complete release of the MNA. Second, the PDMS precursor and curing agent are mixed at a mass ratio of 9:1. This is then mixed with a saturated aqueous NaHCO₃ solution in a 2:1 ratio so that the NaHCO₃ will decompose and produce CO₂ to form bubbles when the mixture is heated. Third, the prepared mixture is filled into the PDMS mold and placed under low pressure using a vacuum pump for half an hour until the bubbles are completely eliminated. After being evenly baked for 4 h on a heating plate at 90 °C, the porous PDMS MNA has solidified and can be released from the mold. Fourth, another 1.8-μm-thick parylene layer is deposited on the cured MNA surface as the metal attachment layer. Fifth, 20 nm Cr and 60 nm Au are deposited on the surface of the MNA as a metal conductive electrode; the signal lines are connected to these electrode pads with conductive silver paste. Then, a 1.8-μm-thick layer of parylene is deposited on top of the MNA electrode as an insulating layer. Finally, two identical porous MNA electrodes are bonded face to face with a PDMS and NaHCO₃ mixture and cured at 90 °C for 1 h to form the PM-type pressure sensor.

MNA electrodes with an imporous PDMS substrate are fabricated in the same way as the PM-type electrode. However, NaHCO₃ is not added to the PDMS to avoid the formation of air bubbles. Then, two identical MNA electrodes are bonded face to face with PDMS solution and cured at 90 °C for 1 h to form the M-type pressure sensor.

To create a flat electrode without an MNA or porous structures, PDMS is spun on a glass wafer at a speed of 600 rpm for 30 s and solidified to form a substrate with 2 μm height, which is consistent with the height of the MNA substrate. A metal conductive layer and an insulating layer are then fabricated in the same way as for the PM-type electrode. Then, two identical flat electrodes are bonded face to face using PDMS solution and cured at 90 °C for 1 h to form the F-type pressure sensor.

Figures 1(b)–1(f) show schematic diagrams and photographs of the three different types of sensor. It can be seen that there are obvious differences in their physical appearance. The cross sections of the PM- and M-type sensor in Figs. 1(g) and 1(h) show that the microneedle structures of the upper and lower MNAs are interleaved to form an interdigital electrode structure. However, for two main reasons, there are some inconsistencies between the structures shown in Figs. 1(c) and 1(d) and in Figs. 1(g) and 1(h). First, some of the microneedles may squeeze and deform during bonding, resulting in the two MNA arrays not forming a perfect interdigital structure. Second, the interdigital structure formed by the bonded MNA array can be seen only when observations are made in specific directions. The fillings between the PM- and M-type interdigital electrodes are porous and imporous PDMS, respectively. In comparison, the substrate of the M-type sensor is relatively flat and smooth. However, the substrate of the PM-type sensor contains a large number of bubbles and will be more prone to deformation under external forces.

A pressure-sensor array based on as-fabricated PM-type sensors was constructed with a 3 × 3 structure. First, three identical PM-type electrodes were fabricated on the same porous PDMS substrate to form a 3 × 1 electrode array. Then, three of these 3 × 1 electrode arrays were aligned in a row with a spacing of 20 mm to form the top 3 × 3 electrode array; the other three 3 × 1 electrode arrays were aligned in a column with a spacing of 20 mm to form the bottom 3 × 3 electrode array. The fabrication processes of the conductive and insulating layers were similar to that for the single PM-type electrode. After that, the row and column sensor array was orthogonally bonded using a PDMS–NaHCO₃ mixture to form a 3 × 3 PM-type sensor array.

PDMS is widely used in the fabrication of flexible pressure sensors due to its easy deformability.²⁹ To further explore the effect of the interdigital MNA electrode structure on the deformability of the sensor, finite-element analysis of different MNA structures was conducted under different pressures. As shown in Fig. 2, finite-element modeling and analysis were carried out on a cylindrical array [Fig. 2(a)], inverted MNA [Fig. 2(b)], forward MNA [Fig. 2(c)], and interdigital MNA [Fig. 2(d)]. In these structures, the height, top/bottom diameter, and center pitch of the microneedles/cylinders remained the same. The density of the material was set to 970 kg/m³, the relative dielectric constant was 2.75, the Young’s modulus was 750 kPa, and the Poisson ratio was 0.49 uniformly. The lower substrate of each structure was set to a fixed constraint. Then, the upper surfaces of the structures were subjected to the application of 0.5, 2.0, 5.0, and 10 kPa of pressure.

According to the simulation results, the internal stress and strain of the structures increase with the gradual application of pressure on the upper electrode. For close-packed cylindrical structures, the stress is mainly concentrated on the upper surface of the electrode. Because the distribution area of the stress is large, the strain generated is small. The stress of the structure in Fig. 2(b) is concentrated on the downward tips and is therefore capable of generating a relatively large strain. The stress of the structure in Fig. 2(c) is mainly concentrated on the surface of the forward tips of the upper electrode and is therefore able to generate a relatively large strain on both the tips and the electrode. As shown in Fig. 2(d), the combination of the above two structures is able to generate large stresses on the surfaces of both the tips and the upper electrode at the same time, resulting in large strains on both.

Three kind of pressure sensor with structures as described in Sec. II A were subjected to 0–80 kPa pressure to test their sensitivity and stability. In the low-pressure range, the sensitivity of the PM-type sensor was found to be much higher than that of either of the other two. The PM-type sensor remained stable and had a faster response time when the pressure was applied cyclically. In the process of testing multiple biological signals, its capacitance value was found to change appropriately. The PM-type MNA structure could also precisely sense local pressure changes in a 3 × 3 sensor array.

The MNA electrode as constructed was 25 mm long, 15 mm wide, and 1 mm thick. The electrode had a MNA structure (10 × 10) with a height of 600 μm. The top and bottom diameters of the microneedles were 10 and 300 μm, respectively, and the center pitch of each microneedle was 600 μm. The morphologies of the flexible pressure sensors based on porous MNA interdigital electrodes (PM-type) and imporous MNA interdigital electrodes (M-type) were observed using a metallography microscope. As shown in Figs. 3(a)–3(b), compared with the P-type sensor, a large number of air gaps (white bubbles) appeared on the surface of the PM-type electrode due to the presence of NaHCO₃ decomposition products. Furthermore, as depicted in Figs. 3(c)–3(d), the surface of the PM-type electrode was rough and wrinkly, and this wrinkling increases the surface area of the metal electrode. However, the tips of the microneedles on the PM-type sensor are more likely to experience irreversible bending and deformation under high pressure (80 kPa), while the tips of the M-type sensor remained the same as before. This strongly demonstrates that the microneedle structure of the PM-type sensor is more likely to deform under a given pressure.

As the PM-type pressure sensor is used to detect physiological signals, which are in the low-pressure range, it has the advantage of high sensitivity while avoiding its disadvantage of low stability. As shown in Figs. 3(e) and 3(f), a large number of small bubbles appeared in the unstructured region of the PM-type sensor, but these were not present in the M-type sensor. These bubbles make the PM-type sensor to have a larger deformation space for a given external force. The sensitivity of the flexible capacitive pressure sensors was measured from 0 to 80 kPa using a pressure puller and an impedance analyzer.

To improve the stability of the PM-type sensor, two layers of parylene are used to enhance the bonding force and insulation properties of the metal electrode layer. To test the stability of the sensor, 2 N of external force was applied to it in 1500 loading and unloading cycles, as depicted in Fig. 5(a). The results show that the amplitude of the capacitive signal remained in the range 7–10 pF, while the amplitude of the noise signal was greater than 10 pF. To eliminate the influence of noise signals on the test results, the signals in Figs. 5(b)–5(d) were extracted from the beginning, middle, and end periods of the signal in Fig. 5(a), in which the vibration noise of the machine was deliberately avoided. No data adjustments were made to these extracted signals. As can be seen in Figs. 5(b)–5(d) in detail, the stability of the sensor did not decay over time in the tests, and the performance remained basically the same.

To test the dynamic response of the PM-type sensor, external forces of 4.00, 1.00, and 0.25 N were applied to it, as shown in Fig. 5(e). The central gray area represents the stable loading period; the light-red areas on either side of the gray area represent the loading and unloading periods. The dotted lines represent the starting points of the loading times and the ending points of the unloading times of the different forces. As the force is increased, the response time also gradually increases. This is mainly because the material inside the sensor will take longer to restore deformation when the sensor is under a greater pressure. When 0.25 N of force is applied, the loading response delay of the sensor is 100 ms, and the unloading response delay is 110 ms. During the dynamic tests, the capacitance output undergoes slight fluctuations, which are mainly caused by the vibration of the pressure-testing machine.

For robotics, haptic and motion feedback from the surfaces of manipulators is critical for accurate gripping.^31–33 For this reason, a PM-type pressure sensor was attached to the surface of a human thumb to detect the force feedback from the fingers as they gripped beakers of different weights. As shown in Fig. 6(a), a beaker without water, a beaker loaded with 40 g of water, and a beaker loaded with 80 g of water were lifted in turn by the hand for these pressure tests. The results showed that the capacitive output of the sensor increased with the weight of the beaker, which indicates that the thumb needed to apply more pressure hold it.

The movement of a wrist joint of the hand was also tested by attaching a sensor, and the results are shown in in Fig. 6(b). The bending of the hand squeezed the sensor, thus causing the capacitance to rise sharply. When the wrist was relaxed, the sensor quickly returned to its initial shape, and the capacitance decreased. Figure 6(c) shows the response of a sensor pasted onto the muscle group of a volunteer’s arm to test the forearm movement. In this situation, muscle contraction applies pressure to the sensor, and the capacitance rises rapidly with this motion. The movement of facial muscles was also tested [Fig. 6(d)] by attaching a sensor to the surface of a volunteer’s face in the expressionless state. When the volunteer smiled, the facial muscles exerted pressure on the sensor, causing the capacitance to rise. Figure 6(e) shows the results of pasting a sensor on a volunteer’s abdomen to detect breathing. When the volunteer inhaled, the abdominal cavity expanded and the capacitance of the sensor rose. When the volunteer exhaled, the abdominal cavity contracted and the capacitance of the sensor decreased. Figure 6(f) shows the results of pasting a sensor onto a volunteer’s temple to detect blinking. The blinking motion is the result of muscle movements around the eyes, and the contraction of these muscles exerted a pressure on the sensor through the deformation of the skin, causing its capacitance to increase. Figure 6(h) shows the results of pasting a sensor onto a volunteer’s Adam’s apple to test its movement. The Adam’s apple moved up and down when the volunteer swallowed, thus affecting the capacitance value. In addition, when the volunteer spoke different words with a certain pace, the changes of capacitance value detected by the sensor had clear waveform specificity, which could be used for semantic recognition, as shown in Fig. 6(i).

The ability to detect a pressure distribution in space may be the key to intelligent robotics and artificial electronic skin applications for prosthetic limbs.^23,34 Compared with a single sensor, a pressure-sensor array can systematically detect the active muscle groups of the human body over a wider area to enable more precise control. For this purpose, a 3 × 3 pressure-sensor array based on the as-fabricated PM-type sensor was constructed to test the pressure distribution. Objects with different masses were placed in different positions on the sensor array, and the capacitance variations were visualized using three-dimensional histograms. The results show that the sensor array can accurately distinguish objects with different masses placed at different positions. As shown in Fig. 7(a), weights with masses of 20 and 50 g were placed at coordinates (1,3) and (3,1). By visualizing the capacitance data measured by the sensor array, the masses and positions of the objects can be accurately identified in Fig. 7(b). As shown in Fig. 7(c), a 50 g weight was placed at coordinate (2,2), and the mass and position of this object can also be accurately identified from Fig. 7(d). As shown in Fig. 7(e), a beaker with mass of 100 g was placed on the sensor array. It can be seen from Fig. 7(f) that the gravity center of the beaker was located between coordinates (2,2) and (2,3), where there was a notch in the beaker.

This work developed a fabrication method for flexible pressure sensors based on an interdigital electrode structure composed of porous MNAs. Compared to previous research, this work presents a novel method for improving the sensitivity and stability of the device by modifying the electrode layer rather than the dielectric layer. The porous MNA substrate was prepared by molding from NaHCO₃ mixed with PDMS. Two layers of parylene were deposited on the MNA substrate to enhance the stability of the conductive layer. By comparing the PM-, M-, and F-type sensors, it was shown that the interdigital electrode structure with porous MNA plays a key role in improving the sensitivity. The PM-type sensor was found to have a maximum sensitivity of 0.07 kPa⁻¹ in the range 0–80 kPa, which is seven times larger than that found for the M-type sensor. In addition, the stability of the PM-type sensor remained stable during 1500 loading cycles of 2 N. The response times during loading and unloading of 0.25 N were 100 and 110 ms, respectively. The results of bio-signal monitoring tests showed that the pressure sensor can accurately monitor the motion of the finger, wrist joint, arm, face, abdomen, eye, and Adam’s apple. By monitoring the movement of the Adam’s apple, a preliminary semantic-recognition function was achieved. Finally, multiple pressure sensors were integrated into a 3 × 3 array to the detect pressure distribution in space, which is an important ability for intelligent robotics and artificial electronic skin applications in prosthetic limbs.

It should be noted that the initial capacitances of different sensors under a given pressure and the initial capacitance a given sensor under different pressures show slight variations. This indicates that the consistency of the sensor-preparation process and the working stability under high pressures still need to be improved. This is a problem that needs to be addressed in subsequent work.

This work was supported in part by the National Natural Science Foundation of China (Grant No. 62104056), the Zhejiang Provincial Natural Science Foundation of China (Grant No. LQ21F010010), the National Natural Science Foundation of China (Grant Nos. 62141409 and 62204204), the National Key R&D Program of China (Grant No. 2022ZD0208602), the Zhejiang Provincial Key Research & Development Fund (Grant Nos. 2019C04003 and 2021C01041), the Shanghai Sailing Program (Grant No. 21YF1451000), and the Key Research and Development Program of Shaanxi (Grant No. 2022GY-001).

The authors have no conflicts to disclose.

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Jiahui Xu received a B.S. degree in electronic science and technology from Hangzhou Dianzi University, Hangzhou, China, in 2021. He is currently pursuing an M.S. degree in a new generation of electronic information technology at Hangzhou Dianzi University, Hangzhou, China. His current research interests include the design of flexible wearable biosensors, implantable devices, and human–machine interaction.

Minghao Wang received a B.S. degree in applied physics from Henan University of Science and Technology, Luoyang, China, in 2011, an M.S. degree in microelectronics and solid-state electronics from Xiangtan University, Xiangtan, China, in 2014, and a Ph.D. in electronics science and technology from Shanghai Jiao Tong University, Shanghai, China, in 2019. He is currently an associate professor at Hangzhou Dianzi University. His current research interests include the design of biotic microelectrodes, MEMS sensors, implantable devices, and neural interface devices.

Minyi Jin received a B.S. degree in electronic and information engineering from China Jiliang University, Hangzhou, China. He is currently pursuing an M.S. degree in new generation electronic information technology at Hangzhou Dianzi University, Hangzhou, China. His current research interests include design, fabrication, and multifunctional integration of implantable devices and neural interface devices.

Siyan Shang received a B.S. degree in electronic science and technology from Hangzhou Dianzi University, Hangzhou, China, in 2021. She is currently pursuing an M.S. degree in electronic science and technology at Hangzhou Dianzi University, Hangzhou, China. Her current research interests include the design of implantable devices and neural interface devices.

Chuner Ni received a B.S. degree in electronic information engineering from Hangzhou Dianzi University, Hangzhou, China, in 2022. He is currently pursuing an M.S. degree in integrated circuit engineering at Hangzhou Dianzi University, Hangzhou, China. His current research interests include EOG electrodes and high-precision pressure sensors.

Yili Hu was born in Hangzhou, China, 1989. He received a B.S. degree in mechanical design, manufacturing, and automation and an M.S. degree in physical electronics from Zhejiang Normal University, Jinhua, China, in 2008 and 2012, respectively. He then received a Ph.D. from Shanghai Jiao Tong University, Shanghai, China, in 2020. He is currently working as a lecturer in Zhejiang Normal University, Jinhua, China. His current research interests include piezoelectric actuators, sensors, and energy harvesters.

Xun Sun received a Ph.D. in electronic science and technology from Shanghai Jiao Tong University in 2019. From 2019 to 2020, he was an R&D engineer at Intel Semiconductor (Dalian) Co., Ltd. Since 2021, he has been an academic leader in the Institute of Guizhou Aerospace Measuring and Testing Technology, which is part of China Aerospace Science and Industry Corporation. He is the author of nine articles and is responsible for five inventions. His research interests include advanced pressure sensors and new testing technologies based on MEMS technology.

Jun Xu received a B.S. degree in automation technology from Harbin Engineering University, Harbin, China, in 2008. He has been a Senior Engineer at the Institute of Guizhou Aerospace Measuring and Testing Technology, which is part of the China Aerospace Science and Industry Corporation. He is the author of three articles and is responsible for three inventions. His research interests include advanced biosensors and testing systems.

Bowen Ji received a Ph.D. in electronics science and technology at Shanghai Jiao Tong University, Shanghai, China, in 2019. He was also a joint Ph.D. in solid mechanics at Northwestern University, Evanston, USA, from 2016 to 2018. Now, he is an associate professor in Northwestern Polytechnical University, Xi’an, China. His research interests include brain–computer interfaces, implantable MEMS devices, and flexible sensors.

Dr. Le Li is currently a professor with the Institute of Medical Research, Northwestern Polytechnical University, China. He received his Ph.D. from the Hong Kong Polytechnic University (2007) and completed his postdoctoral training in the same university (2010). He worked as a clinical researcher at First Affiliated Hospital, Sun Yat-sen University from 2010–2021. Previously, Dr. Li worked as a visiting scientist at TIRR Memorial Hermann Research Center and Department of Physical Medicine and Rehabilitation, University of Texas Health Science Center (UTHealth) in Houston, USA (2014–2016). Dr. Li’s research interests include bio-signal processing (EMG), neuromusculoskeletal modeling of normal and spastic subjects, and musculoskeletal ultrasound applications.

Yuhua Cheng received a B.S. degree in electrical and electronic engineering and a Ph.D. in circuits and system in 2005 and 2011, respectively, both from Zhejiang University, China. He was a visiting scholar at GT-Bionics Lab, Georgia Institute of Technology, USA, from 2015 to 2016. He is a professor at Wenzhou Institute of Hangzhou Dianzi University, China. His current research interests include wireless power transfer for implantable biomedical devices, energy harvesting, and power management circuits and systems.

Gaofeng Wang received a Ph.D. in electrical engineering from the University of Wisconsin, Milwaukee, in 1993 and a Ph.D. in scientific computing from Stanford University, Stanford, California, in 2001. From 1993 to1996, he was a scientist at Tanner Research Inc., Pasadena, CA. From 1996 to 2001, he was a Principal Research and Development Engineer at Synopsys Inc., Mountain View, CA. From 2001 to 2003, he was Chief Technology Officer (CTO) at Intpax, Inc., San Jose, CA. From 2004 to 2010, he was CTO at Siargo Inc., Santa Clara, CA. From 2010 to 2013, he was Chief Scientist at Lorentz Solution, Inc., Santa Clara, CA. From 2004 to 2013, he was also Adjunct Professor and Founding Director at the Institute of Microelectronics & Information Technology, Wuhan University, Hubei, China. He is currently a Distinguished Professor at the School of Electronics and Information, Hangzhou Dianzi University, Zhejiang, China. He has published over 500 peer-reviewed journal articles and conference papers and is the holder of over 190 patents. His research and development interests include integrated circuits, MEMS and sensor technology, electronic design automation, and computational electromagnetics. He is a Fellow of the Institution of Engineering and Technology (IET).