All MoS2-Based Large Area, Skin-Attachable Active-Matrix Tactile

Feb 15, 2019 - Herein, we utilized the benefit of the semiconducting and mechanical excellence of MoS2 and placed it between high-k Al2O3 dielectric ...
0 downloads 0 Views 786KB Size
Subscriber access provided by WEBSTER UNIV

Article 2

All MoS Based Large Area, Skin-Attachable Active-Matrix Tactile Sensor Yong Ju Park, Bhupendra K. Sharma, Sachin M. Shinde, MinSeok Kim, Bongkyun Jang, Jae-Hyun Kim, and Jong-Hyun Ahn ACS Nano, Just Accepted Manuscript • Publication Date (Web): 15 Feb 2019 Downloaded from http://pubs.acs.org on February 15, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

All MoS2 Based Large Area, Skin-Attachable Active-Matrix Tactile Sensor Yong Ju Park†, Bhupendra K. Sharma†, Sachin M. Shinde†, Min-Seok Kim‡, Bongkyun Jang§, Jae-Hyun Kim§, and Jong-Hyun Ahn†,* †

School of Electrical and Electronic Engineering, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, Republic of Korea. ‡

Center for Mass and Related Quantities, Korea Research Institute of Standards and Science,

267 Gajeong-ro, Yuseong-gu, Daejeon 34113, Republic of Korea. §

Department of Applied Nano-mechanics, Nano-Convergence Mechanical Systems Research

Division, Korea Institute of Machinery & Materials, 156 Gajeongbuk-ro, Yuseong-gu, Daejeon 34103, Republic of Korea.

* Corresponding author: [email protected]

1 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT Large-area, ultra-thin flexible tactile sensors with conformal adherence are becoming crucial for advances in wearable electronics, electronic skins and bio-robotics. However, normal passive tactile sensors suffer from high crosstalk, resulting in inaccurate sensing, which consequently, limits their use in such advanced applications. Active-matrix driven tactile sensors could potentially overcome such hurdles but it demands the high performance and reliable operations of the thin-film-transistor array that could efficiently control integrated pressure gauges. Herein, we utilized the benefit of the semiconducting and mechanical excellence of MoS2 and placed it between high-k Al2O3 dielectric sandwich layers to achieve the high and reliable performance of MoS2 based back-plane circuitry and strain sensor. This strategical combination reduces the fabrication complexity and enables the demonstration of all MoS2 based large area (8x8 array) active-matrix tactile sensor offering a wide sensing range (1 to 120 kPa), sensitivity value (ΔR/R0: 0.011 kPa−1) and a response time (180 ms) with excellent linearity. In addition, it showed potential in sensing multi-touch accurately, tracking a stylus trajectory and detecting the shape of an external object by grasping it using the palm of the human hand. Keywords : MoS2, active-matrix, tactile sensor, wearable electronics, electronic skin.

2 ACS Paragon Plus Environment

Page 2 of 19

Page 3 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

To achieve improved functions in bio-medical, robotic, wearable and electronic skin applications, the development of large-area flexible tactile sensors with high-density arrays that enable accurate sensing of pressure from external environment/objects is in great demand.1-11 In this regard, extensive work has been conducted to demonstrate various types of tactile sensors based on interlocking nanofibres, hybrid composites and micro-structured pyramid or dome-shaped arrays for enhancing the sensitivity, response time and accuracy of the sensors.12-17 However, a simple passive-matrix (PM) tactile sensor arrays result in inaccurate sensing of pressure distribution owing to the substantial crosstalk among pressure gauges which can lead to a wrong estimation of the object’s shape.18 Therefore, to obtain a more precise readout of the pressure distribution based on the various shapes and sizes of the object, developing a tactile sensor with an active-matrix (AM) circuitry is necessary.19 Furthermore, integrating the developed tactile sensors in a conformal format without losing their tactility for real-time applications, including sensing an object, detecting the size, shape and orientation of an external object, picking up various objects using a robot hand and health monitoring such as measuring pulses and blood pressure, is necessary. Few attempts have been made to demonstrate conformal tactile sensors20,21; however, these sensors lacked a combination of excellent conformability, high sensitivity, the ability to maintain a sensing range of the human skin and low crosstalk. Two-dimensional (2D) transition metal dichalcogenides (TMDCs) have been proven to be potential semiconductors for flexible electronics, but they have not been explored much in the fabrication of large-area, flexible AM tactile sensors. Molybdenum disulfide (MoS2) grown using chemical vapour deposition (CVD) with excellent optical, mechanical and semiconducting properties has been recognized as a potential candidate for future flexible applications.22-24 Moreover, MoS2 exhibited a gauge factor (single crystal ~200, polycrystalline ~70),25,26 which facilitates its potential applications in tactile sensing.25 Recently, Park et al. integrated CVD-grown MoS2 into the fabrication of a simple PM tactile sensor array with excellent mechanical endurance.26 However, this integrated sensor array presented limited sensing capability due to the passive structure. A significant improvement in the sensing capability of MoS2-based tactile sensors could be made by controlling the sensing units through an AM backplane of the TFT array. Herein, we fabricated large-area, flexible MoS2 tactile sensors driven by an AM backplane circuitry. The representative tactile sensor array efficiently increased the crosstalk isolation value up to 24.8 dB as compared to its PM counterpart of 6.6 dB.27 This sensor array was also able to sense minimum and maximum pressure values up to 1 and 120 kPa, respectively, 3 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

which covered a reasonable sensing range of the human skin. Moreover, owing to low crosstalk, the tactile sensor array exhibited high multi-touch sensibility, thus being able to detect the shape of an object grasped by the human palm (e.g. grasping an apple). Given the achieved pressure range that covered the human skin’s sensing range and the multi-touch sensibility together with excellent features such as response time of 180 ms, sensitivity of 0.011 kPa−1, excellent linearity, quick information input and pressure mapping analysis, the prototype AM-enabled tactile sensor array established compatible applications for electronic skin (E-skin). RESULTS AND DISCUSSTION Figure 1 presents the detailed structure of large-area, flexible AM-driven tactile sensors. First, interdigitated electrodes based on pressure gauge units along with source-drain lines for the MoS2 TFTs were deposited on Al2O3 (50 nm)/SU-8 (500 nm)/PMMA (sacrificial layer), top to bottom in sequence, coated with a SiO2/Si wafer (the growth and characterizations of MoS2 are given in Supporting Information, Figure S1).28,29 To increase the effective pressure sensing area by more than 50%, the width-to-length (W/L) ratio for the pressure gauge was designed to be much larger (W/L: 14800/25) than that of the TFT channel (W/L: 300/4). The schematic of the AM-driven tactile sensor (AM-TS) along with its crosssectional view, operational mechanism of a single sensor unit controlled by TFT, and optical image are shown in Figure. 1(a), (b), (c) and (d) respectively. The tactile sensor arrays were assembled in an 8 × 8 matrix forming an active area of 2.8 × 2.8 cm2, wherein each sensor unit was controlled by a particular TFT. The complete assembly of the AM-TS could be formed in a freestanding flexible form by merely dissolving the PMMA sacrificial layer in acetone (Figure 1(e)). The AM-TS was made e-skin compatible by attaching it to 2-mm-thin polydimethylsiloxane (PDMS) prepared with a specific ratio of the curing agent (20:1), which exhibited a Young’s modulus value (200 kPa), comparable to the human skin (150–350 kPa).30,31 The integration of the AM backplane TFT array to the tactile sensors over the normal PM type resulted in the reduction of crosstalk among the pressure gauge units. Before evaluating the full potential of AM integration in the operation of a tactile sensor array, the pressure gauge unit, MoS2 TFT, and the unit pressure gauge driven by a TFT were evaluated individually. The current–voltage (I–V) characteristics of the pressure gauge exhibited symmetrically linear behaviours (Figure 2(a)) at a particular applied pressure. The current values in characteristics decrease with increasing the pressure which indicates the 4 ACS Paragon Plus Environment

Page 4 of 19

Page 5 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

piezoresistive property of MoS2. The piezoresistivity in MoS2 occurs due to the change in bandgap in response to generated strain upon exerting the external pressure.25 Furthermore, compressive strain (up to  -2%) increases the bandgap, while tensile strain ( 8%) decreases it which increases and decreases the resistance values, respectively.32 In the present case, the neutral mechanical plane (NMP) is positioned below the active layer MoS2, which bends towards tip under externally applied pressure. Therefore, active layer MoS2 feels compressive strain which results in the increase in resistance (Figure 2(a), inset). Moreover, the change in relative resistance at a particular voltage followed a linear behaviour with respect to the externally applied pressures (Figure 2(a), inset) which indicates good linearity from MoS2 based pressure gauge unit. The linear response of the output (here, change in resistance) against the input (applied external pressure) is a primary requirement for an accurate and reliable sensor which was very well fulfilled by the MoS2 pressure gauge unit. The observed excellent linear feature enabled the AM tactile sensor to detect the exact pressure without requiring any extra compensating circuit. The individual TFT exhibited excellent switching and output characteristics (Figure 2(b) and Supporting Information, Figure S2) with the average values of extracted parameters: current on/off ratio (Ion/off ratio) of 106, threshold voltage (Vth) of 3.1 ± 0.4 and mobility (µ) 16.2 ± 1.3 cm2/V·s. The integrated TFT enabled or disabled the individual pressure gauge unit depending on its ON- and OFF- states, thus avoiding the interference among neighbouring cells, thereby leading to low crosstalk. The sensing output (i.e. overall measured resistance) of the TFT-controlled pressure gauge unit against externally applied pressure included the sum of the resistance change of the TFT channel and the pressure gauge unit. In other words, the sensing ability of the pressure gauge unit controlled by the TFT relied on the resistance change of both, the semiconductor MoS2 belonging to the individual pressure gauge unit as well as the channel resistance of the integrated TFT. Therefore, for an accurate sensing operation, the TFT channel resistance must be minimal, and the resistance value of the whole assembly must be primarily contributed by the pressure gauge unit. To ensure this, the TFT parameters must be optimized within a certain value range, resulting in the minimal value of TFT resistance in the ON-state. In the present scenario, the Vth and the Ion played a crucial role in deciding the TFT resistance at a particular set of Vgs and Vds values. To achieve the required identical value of resistance estimated by Ion at a particular set of Vgs and Vds, the values of Vth for all 64 TFTs embedded in the AM must be matched, otherwise variations would be observed in the resistance of the TFTs (Supporting Information, Figure S3). Usually, MoS2 TFT presents a large Vgs switching range, a negative Vth, a low Ion and a variable performance on the SiO2/Si back-gate substrate 5 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

owing to its low dielectric, charge trapping and exposed back-channel surface to the ambient environment (Supporting Information, Figure S4). However, to achieve a reliable and highperforming MoS2 TFT, the recently developed high-κ Al2O3 sandwiched channel-based approach was adopted,33 which resulted in a quite low sweeping Vgs range, a nearly-zero Vth, high Ion and high mobility (Figure 2(b), (c) and Supporting Information, Figure S2). The resultant histogram of Vth and Ion for all 64 TFTs are presented in Figure 2(c), indicating their reliable values. The TFT characteristics at zero pressure showed a resistance value of 1 kΩ in the ON-state at Vgs 10 V (Figure S2), whereas the total resistance became 140 kΩ when measured in a combination series of pressure gauge, thereby revealing its dominant behaviour (Figure 2(d)). This result confirmed that the TFT in ON-state at Vgs 10 V became conducting enough to ignore its resistance as compared to the pressure gauge, thus fulfilling the criteria of accurate measurement of a tactile sensor. Here, the impact of TFT on the pressure gauge in terms of resistance was decided by two regions of the I–V curves labelled as ‘1’ and ‘2’ (Figure 2(d)), i.e. OFF- and ON-states. Initially, the channel resistance dominated and disabled the pressure gauge in the OFF-state; however, after crossing Vth, in the ON-state, the channel started to enable the pressure gauge sensor; once the pressure gauge resistance dominated over the channel resistance, the current output started to get limited, resulting in the sensing feature. The TFT enabled the pressure gauge to exhibit a different characteristic (in region ‘2’) against applied external pressures, showing its excellent sensing ability (Figure 2(d)). Furthermore, the change in resistance increased at high external pressure (Figure 2(e)) and followed a linear behaviour with increasing applied pressures (Figure 2(f)). The pressure gauge sensor exhibited a sensitivity of ∆R/R0: 0.011 kPa−1, a response time of 180 ms and a wide detection range of 1~120 kPa. Furthermore, the response time is reasonably low when considering the planar device structure, which can further be reduced if the delay of deformation by underneath viscoelastic PDMS substrate can be avoided. This was verified by measuring the device on rigid SiO2/Si substrate which showed lower response time (100 ms) than that on PDMS substrate (180 ms) (Figure S5). The TFT enabled pressure gauge also showed excellent repeatability in detecting a particular applied external pressure with a maximum variation of 5.3% for 10,000 cycles (Figure 2(g)). The large-area AM-TS, comprising TFT-driven pressure sensor units, was also found to exhibit excellent mechanical endurance when wrapped around a glass pipet with a radius of 1-mm (Figure 3(a)). The active layers MoS2 and Al2O3 in the tactile sensor played a crucial role in its sensing ability. Thus, the strain applied to these layers under bending states must be 6 ACS Paragon Plus Environment

Page 6 of 19

Page 7 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

investigated to observe the effect on reliable sensing. The finite element method (FEM) was employed to quantify the strain distribution on the top surface and in various layers along the cross-sectional direction under a bending state. We designed the structure of the AM-TS to be positioned a NMP near the MoS2 and Al2O3 layers. Hence, the device layers experienced compressive strain below this plane, whereas tensile strain appeared above it. However, strains applied to these layers are sufficiently low to endure the outer bending with the radius of 1-mm as shown in Figure 3(b). The FEM analysis showed their corresponding result for the strain distribution on the top surface (SU-8): maximum tensile strain (0.08 % on SU-8) was applied along the bending region, whereas the regions far from the bending were found to be strain free (Figure 3(b), left). For strain distribution in a cross-sectional direction, a partial region with the identical structure including all the layers (SU-8/Au/Al2O3/MoS2/Al2O3/SU-8) of the complete device was considered. The maximum strains of MoS2 and Al2O3 in a crosssectional manner were estimated to be approximately 1.5 × 10−3 % and 3.0 × 10−3 %, respectively (Figure 3(b), right). These values were calculated for different bending radii (Supporting Information, Figure S6), which were further used to estimate the lowest possible bending of MoS2 (10 µm) and Al2O3 (26 µm) before they fractured (Supporting Information, Figure S7). The overall FEM analysis revealed that the device’s active layers, MoS2 and Al2O3, experienced strain far less than their failure level when the fabricated AM-TS arrays were bent up to 1 mm of the bending radius. This result was corroborated by the robust performance of the MoS2 TFT and pressure gauge when tested under different bending and cyclic states. The transfer and I-V characteristics of TFT and pressure gauge were noted to be reliable with slight variations under cyclic test (~1000) (Figure S8a and b). The extracted mobility and resistance values also showed slight variations (Figure S8c). The pressure mapping indicates good cross-talk isolation even after 1000 cycles (Figure S8d). Furthermore, the gauge units showed similar values of resistance change under different bending radii (Figure S9). The AM-TS also showed an efficient reduction of cross-talk under bending states (Figure S9c). The reliable performances of AM-TS under cyclic and bending states indicate its excellent mechanical robustness. The transfer characteristics of the MoS2 TFT showed excellent repeatability in maintaining the Ioff, Ion and Ion/off values when its flat state bent to the corresponding 15, 6 and 1 mm bending radii followed by recovery (Figure 3(c)). The slight variation in the Ion, Vth, µ and R (Figure 3(e)) values were noted as 1.8 mA, 0.14 V, 0.12 (∆µ/µ0) and 0.07 (∆R/R0), respectively, which did not affect the dominance of the pressure gauge in terms of resistance 7 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

in the AM-TS; thus, the sensing capability functioned well even under these bending states. Furthermore, the I–V characteristics of the pressure gauge were well maintained with only minor variations in the resistance under similar bending states (Figure 3(d)). Next, the AM-TS performances were evaluated for different overlapping regions between tip and gauge unit (Figure S10), different thickness of the underneath substrate (SU-8/PDMS) (Figure S11), smaller sizes of interdigitated electrodes (Figure S12), and hysteresis for applied pressure (Figure S13). The AM-TS showed the minimum detection limit of ~1 kPa with almost no hysteresis (Figure S13b). Several previous studies have reported a superior detection limit (3~100 Pa), but employed engineered device structures rather than normal planar structure.34,35 Furthermore, the large-area, AM-TS was employed to demonstrate the multitouch ability, real-time pressure mapping and shape detection by grasping an external object with a human palm. Such applications demand excellent mechanical properties and conformal adherence to mount the device array on the rough surface of human skin and textile. The conformal contact is a consequence of low bending stiffness which relies on the material thickness. In the AM-TS, the thickness of SU-8 was much higher than that of the active layers of the device. Thus, the conformal ability was exclusively decided by the SU-8 thickness value. To achieve optimum thickness of SU-8, theoretical modelling was performed, and the interface energy as a function of thickness was estimated (Supporting Information, Figure S14). In the present case, the AM-TS was supported by a 1.5-µm-thick SU-8, well below the critical thickness required for conformal contact, thereby adhering to the PDMS-molded palm very well (Figure 4(a)). Moreover, multi-touch ability, real-time pressure mapping and accurate shape detection require low crosstalk among the assembled pressure gauge units, which was efficiently obtained using the integrated MoS2-based AM backplane (Supporting Information, Figure S15). The individual TFT in the AM backplane isolated its attached pressure gauge unit from others by enabling or disabling the unit with ON- or OFF- states. Therefore, during measurement, only those pressure gauge units responded to the ON-state at a time, which were affected by external applied pressure, whereas the others were disabled by the OFF-states; as a result, crosstalk among pressure gauge units reduced efficiently. The increased crosstalk isolation (Figure 4(b), Supporting Information, Figure S15) in the AM type (24.8 dB) over the normal PM type (6.6 dB), made the representative AM-based tactile sensor capable of detecting the multi-touch with excellent sensitivity. This fact was corroborated by placing a small miniature table on the surface of AM-TS within a 1-mm radius, having 9.75 g of weight with a leg spacing of 2.5 mm (Figure 4(c), Supporting Information, Movie S1). In addition, the low response time with low crosstalk further 8 ACS Paragon Plus Environment

Page 8 of 19

Page 9 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

enhanced the tactile sensor’s capability and allowed it to the map real-time trajectory of the stylus in an ‘M’ shape (Figure 4(d), (e) and Supporting Information, Movie S2). The most exciting feature showed by the AM-TS was its efficient shape detection making it compatible with E-skin in robotic applications. As proof, an external object, an apple, was grasped by the AM-TS adhered to a human palm. The individual pressure gauge unit which came in touch with the object surface detected different values of pressure exerted by the object; thus, the changes in relative resistance (from 0.1 to 0.8 corresponding to 10–80 kPa) were accurately sensed. The resistance change depended on the grip strength (Figure 4(f) and Supporting Information, Movie S3). Herein, the faded regions of the object (e.g. apple) may be due attributed to the fewer number of pressure gauge arrays (64, in fabricated AM-TS). However, a further increment in array numbers could improve shape detection more accurately. CONCLUSION The present work utilized the inherent benefit of the MoS2 characteristics in fabricating the AM as well as pressure gauge units, thus making the fabrication process simpler, and which in combination, allowed to demonstrate highly efficient, large-area, AMenabled tactile sensor arrays. The integrated MoS2-based AM potentially eliminated the crosstalk issue, and the ultra-low thickness of the active layers made it compatible with the conformal attachment to the human-like skin. In addition, the tactile sensor responded excellently to the grasping of an object using the palm of the hand, providing a sense of the object shape. The demonstrated tactile sensor prototype was loaded with such features and capability advances using MoS2 2D material in E-skin and robotics applications. METHODS Synthesis of MoS2. Bilayer MoS2 was grown by employing metal organic chemical vapour deposition (MOCVD) technique. A 4-inch Si wafer with thermally grown 300-nm-thick SiO2 was placed was cleaned with water, acetone and finally isopropanol (IPA) sequentially and placed in a 4.3-inch-diameter quartz tube. Molybdenum hexacarbonyl (MHC, Sigma Aldrich 577766) and dimethyl sulfide (DMS, Sigma Aldrich 471577) were chosen as the Mo and S precursors, respectively, and introduced into the quartz tube with the carrier gases of H2 and Ar. The pressure, temperature and growth time were kept of 7.5 Torr, 550°C and 20 h, respectively. To maintain the pressure during growth, the flow of gases was constant; MHC flow of 1.0 sccm, DMS flow of 0.3 sccm, Ar flow of 300 sccm and H2 flow of 10 sccm.

9 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Fabrication and characterization of the AM MoS2 tactile sensor. A 1.5-μm-thick SU-8 layer was coated on a PMMA/SiO2/Si substrate. Afterwards, a 50-nm-thick Al2O3 layer was deposited on the top of SU-8 layer using an atomic layer deposition system. Source–drain electrodes (Cr/Au: 3/30 nm) were patterned on the Al2O3/SU-8/PMMA/SiO2/Si substrate using conventional photolithography. Herein, PMMA is used as sacrificial layer which is dissolved for transferring the device. The bilayer MoS2 film was transferred onto the sourcedrain metal deposited substrate and patterned as a channel via reactive ion etching using CHF3/O2 plasma. Subsequently, a 50-nm-thick top Al2O3 dielectric layer was deposited on MoS2. Further, an SU-8 insulator was formed partially prior to depositing the top-gate electrode line to avoid the leakage with S-D line (not active region). The top-gate electrode (Cr/Au: 3/30 nm) was formed using photolithography and a lift-off process. A 500-nm-thick SU-8 layer was coated for passivation. Finally, the tactile sensor layer was delaminated by removing the PMMA sacrificial layer. The AM MoS2 tactile sensor was electrically characterised using a source meter unit (Keithley 4200 SCS parameter analyser, Keithley Instruments Inc.). Strain calculation of the AM MoS2 tactile sensor. The strain applied at each layer was calculated via the FEM using a commercial code of ABAQUS. The AM tactile sensor was modelled with 3-dimensional (3D) solid elements based on the actual shape and dimension of the layers, which were tightly bonded together at the interfaces without delamination and slippage. The material properties used in the simulation are listed in Supporting Information, Table S1. For a single unit cell of the tactile sensor, the bending analyses were performed by applying boundary conditions as explained in Supporting Information, Figure S16. After buckling simulation of the devices, the bending radii were obtained from the deformed shape, and the principal strain distributions were plotted in the entire area of the device. In addition, 2D local models illustrating the cross-section of TFTs were analysed using the identical boundary condition used in the global models.

AUTHOR INFORMATION Corresponding author *E-mail: [email protected]

CONFLICT OF INTEREST The authors declare no competing financial interest. 10 ACS Paragon Plus Environment

Page 10 of 19

Page 11 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

ACKNWLEDGEMENTS This work was supported by the National Research Foundation of Korea (NRF) funded by the Korean government (MSIF) (2015R1A3A2066337, CASE2014M3A6A5060933).

SUPPORTING INFORMATION AVAILABLE: Growth and characterization of MOCVD grown wafer-scale large area bilayer MoS2 film on SiO2 wafer, Electrical performance of topgated MoS2 TFT, ON-current variation of three types of top-gated MoS2 TFTs, Electrical performance of bottom-gated MoS2 TFT, Strain distribution analysis, Estimation of minimum bending radii for failure level of MoS2 and Al2O3, Critical SU-8 thickness for conformal contact, Reduced crosstalk of active matrix MoS2 tactile sensor, Boundary conditions used in finite element analysis, Material properties used in finite element models expressing the active-matrix tactile sensor, Movie S1, Movie S2, and Movie S3. This material is available free of charge via the Internet at http://pubs.acs.org.

REFERENCES 1. Kim, J.; Lee, M.; Shim, H. J.; Ghaffari, R.; Cho, H. R.; Son, D.; Jung, Y. H.; Soh, M.; Choi, C.; Jung, S.; Chu, K.; Jeon, D.; Lee, S. T.; Kim, J. H.; Choi, S. H.; Hyeon, T.; Kim, D. H., Stretchable Silicon Nanoribbon Electronics for Skin Prosthesis. Nat. Commun. 2014, 5, 5747. 2. Han, S.; Kim, J.; Won, S. M.; Ma, Y. J.; Kang, D.; Xie, Z. Q.; Lee, K. T.; Chung, H. U.; Banks, A.; Min, S.; Heo, S. Y.; Davies, C. R.; Lee, J. W.; Lee, C. H.; Kim, B. H.; Li, K.; Zhou, Y. D.; Wei, C.; Feng, X.; Huang, Y. G.; Rogers, J. A., Battery-Free, Wireless Sensors for Full-Body Pressure and Temperature Mapping. Sci. Transl. Med. 2018, 10, eaan4950. 3. Wang, C.; Hwang, D.; Yu, Z. B.; Takei, K.; Park, J.; Chen, T.; Ma, B. W.; Javey, A., User-Interactive Electronic Skin for Instantaneous Pressure Visualization. Nat. Mater. 2013, 12, 899-904. 4. Schwartz, G.; Tee, B. C. K.; Mei, J. G.; Appleton, A. L.; Kim, D. H.; Wang, H. L.; Bao, Z. N., Flexible Polymer Transistors with High Pressure Sensitivity for Application in Electronic Skin and Health Monitoring. Nat. Commun. 2013, 4, 1859. 5. Ying, M.; Bonifas, A. P.; Lu, N. S.; Su, Y. W.; Li, R.; Cheng, H. Y.; Ameen, A.; Huang, Y. G.; Rogers, J. A., Silicon Nanomembranes for Fingertip Electronics. Nanotechnology 2012, 23, 344004. 11 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

6. Yeom, C.; Chen, K.; Kiriya, D.; Yu, Z. B.; Cho, G.; Javey, A., Large-Area Compliant Tactile Sensors Using Printed Carbon Nanotube Active-Matrix Backplanes. Adv. Mater. 2015, 27, 1561-1566. 7. You, I.; Choi, S. E.; Hwang, H.; Han, S. W.; Kim, J. W.; Jeong, U., E-Skin Tactile Sensor Matrix Pixelated by Position-Registered Conductive Microparticles Creating Pressure-Sensitive Selectors. Adv. Funct. Mater. 2018, 28, 1801858. 8. Takei, K.; Takahashi, T.; Ho, J. C.; Ko, H.; Gillies, A. G.; Leu, P. W.; Fearing, R. S.; Javey, A., Nanowire Active-Matrix Circuitry for Low-Voltage Macroscale Artificial Skin. Nat. Mater. 2010, 9, 821-826. 9. Someya, T.; Sekitani, T.; Iba, S.; Kato, Y.; Kawaguchi, H.; Sakurai, T., A Large-area, Flexible Pressure Sensor Matrix with Organic Field-Effect Transistors for Artificial Skin Applications. P. Natl. Acad. Sci. USA. 2004, 101, 9966-9970. 10. Jason, N. N.; Ho, M. D.; Cheng, W. L., Resistive Electronic Skin. J. Mater. Chem. C 2017, 5, 5845-5866. 11. Ameri, S. K.; Ho, R.; Jang, H. W.; Tao, L.; Wang, Y. H.; Wang, L.; Schnyer, D. M.; Akinwande, D.; Lu, N. S., Graphene Electronic Tattoo Sensors. ACS Nano 2017, 11, 7634-7641. 12. Pang, C.; Lee, G. Y.; Kim, T. I.; Kim, S. M.; Kim, H. N.; Ahn, S. H.; Suh, K. Y., A Flexible and Highly Sensitive Strain-Gauge Sensor Using Reversible Interlocking of Nanofibres. Nat. Mater. 2012, 11, 795-801. 13. Littlejohn, S.; Nogaret, A.; Prentice, G. M.; Pantos, G. D., Pressure Sensing and Electronic Amplification with Functionalized Graphite-Silicone Composite. Adv. Funct. Mater. 2013, 23, 5398-5402. 14. Zhu, B. W.; Niu, Z. Q.; Wang, H.; Leow, W. R.; Wang, H.; Li, Y. G.; Zheng, L. Y.; Wei, J.; Huo, F. W.; Chen, X. D., Microstructured Graphene Arrays for Highly Sensitive Flexible Tactile Sensors. Small 2014, 10, 3625-3631. 15. Mannsfeld, S. C. B.; Tee, B. C. K.; Stoltenberg, R. M.; Chen, C. V. H. H.; Barman, S.; Muir, B. V. O.; Sokolov, A. N.; Reese, C.; Bao, Z. N., Highly Sensitive Flexible Pressure Sensors with Microstructured Rubber Dielectric Layers. Nat. Mater. 2010, 9, 859-864. 16. Park, J.; Kim, M.; Lee, Y.; Lee, H. S.; Ko, H., Fingertip Skin-Inspired Microstructured Ferroelectric Skins Discriminate Static/Dynamic Pressure and Temperature Stimuli. Sci. Adv. 2015, 1, 1500661 17. Cai, G. F.; Wang, J. X.; Lin, M. F.; Chen, J. W.; Cui, M. Q.; Qian, K.; Li, S. H.; Cui, P.; Lee, P. S., A Semitransparent Snake-Like Tactile and Olfactory Bionic Sensor with Reversibly Stretchable Properties. NPG Asia Mater. 2017, 9, e437. 18. Wang, X. D.; Zhang, H. L.; Dong, L.; Han, X.; Du, W. M.; Zhai, J. Y.; Pan, C. F.; 12 ACS Paragon Plus Environment

Page 12 of 19

Page 13 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Wang, Z. L., Self-Powered High-Resolution and Pressure-Sensitive Triboelectric Sensor Matrix for Real-Time Tactile Mapping. Adv. Mater. 2016, 28, 2896-2903. 19. Shu, L.; Tao, X. M.; Feng, D. D., A New Approach for Readout of Resistive Sensor Arrays for Wearable Electronic Applications. IEEE Sens. J. 2015, 15, 442-452. 20. Nawrocki, R. A.; Matsuhisa, N.; Yokota, T.; Someya, T., 300-nm Imperceptible, Ultraflexible, and Biocompatible E-Skin Fit with Tactile Sensors and Organic Transistors. Adv. Electron. Mater. 2016, 2, 1500452. 21. Lee, S.; Reuveny, A.; Reeder, J.; Lee, S.; Jin, H.; Liu, Q. H.; Yokota, T.; Sekitani, T.; Isoyama, T.; Abe, Y.; Suo, Z. G.; Someya, T., A Transparent Bending-Insensitive Pressure Sensor. Nat. Nanotechnol. 2016, 11, 472-478. 22. Ganatra, R.; Zhang, Q., Few-Layer MoS2: A Promising Layered Semiconductor. ACS Nano 2014, 8, 4074-4099. 23. Shih, F. Y.; Wu, Y. C.; Shih, Y. S.; Shih, M. C.; Wu, T. S.; Ho, P. H.; Chen, C. W.; Chen, Y. F.; Chiu, Y. P.; Wang, W. H., Environment-Insensitive and Gate-Controllable Photocurrent Enabled by Bandgap Engineering of MoS2 Junctions. Sci. Rep. 2017, 7, 44768. 24. Zhu, Y. B.; Li, Y. J.; Arefe, G.; Burke, R. A.; Tan, C.; Hao, Y. F.; Liu, X. C.; Liu, X.; Yoo, W. J.; Dubey, M.; Lin, Q.; Hone, J. C., Monolayer Molybdenum Disulfide Transistors with Single-Atom-Thick Gates. Nano Lett. 2018, 18, 3807-3813. 25. Manzeli, S.; Allain, A.; Ghadimi, A.; Kis, A., Piezoresistivity and Strain-Induced Band Gap Tuning in Atomically Thin MoS2. Nano Lett. 2015, 15, 5330-5335. 26. Park, M.; Park, Y. J.; Chen, X.; Park, Y. K.; Kim, M. S.; Ahn, J. H., MoS2-Based Tactile Sensor for Electronic Skin Applications. Adv. Mater. 2016, 28, 2556-2562. 27. Park, M.; Kim, M. S.; Park, Y. K.; Ahn, J. H., Si Membrane Based Tactile Sensor with Active Matrix Circuitry for Artificial Skin Applications. Appl. Phys. Lett. 2015, 106, 043502. 28. Kang, K.; Xie, S. E.; Huang, L. J.; Han, Y. M.; Huang, P. Y.; Mak, K. F.; Kim, C. J.; Muller, D.; Park, J., High-Mobility Three-Atom-Thick Semiconducting Films with Wafer-Scale Homogeneity. Nature 2015, 520, 656-660. 29. Jeon, J.; Jang, S. K.; Jeon, S. M.; Yoo, G.; Jang, Y. H.; Park, J. H.; Lee, S., LayerControlled CVD Growth of Large-Area Two-Dimensional MoS2 Films. Nanoscale 2015, 7, 1688-1695. 30. Michael, H.; Pinar, A.; Gaurav, G.;, Imaging in Dermatology 1st edition, Academic Press, 2016. 31. Xu, Q. W.; Li, C.; Kang, Y. J.; Zhang, Y. L., Long Term Effects of Substrate Stiffness on the Development of hMSC Mechanical Properties. RSC Adv. 2015, 5, 10565113 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

105660. 32. Lu, P.; Wu, X. J.; Guo, W. L.; Zeng, X. C., Strain-Dependent Electronic and Magnetic Properties of MoS2 Monolayer, Bilayer, Nanoribbons and Nanotubes. Phys. Chem. Chem. Phys. 2012, 14, 13035-13040. 33. Choi, M.; Park, Y. J.; Sharma, B. K.; Bae, S. R.; Kim, S. Y.; Ahn, J. H., Flexible Active-Matrix Organic Light-Emitting Diode Display Enabled by MoS2 Thin-Film Transistor. Sci. Adv. 2018, 4, eaas8721. 34. Mannsfeld, S. C. B.; Tee, B. C. K.; Stoltenberg, R. M.; Chen, C. V. H. H.; Barman, S.; Muir, B. V. O.; Sokolov, A. N.; Reese, C.; Bao, Z. N., Highly Sensitive Flexible Pressure Sensors with Microstructured Rubber Dielectric Layers. Nat. Mater. 2010, 9, 859-864. 35. Park, J.; Lee, Y.; Hong, J.; Lee, Y.; Ha, M.; Jung, Y.; Lim, H.; Kim, S. Y.; Ko, H., Tactile-Direction-Sensitive and Stretchable Electronic Skins Based on Human-SkinInspired Interlocked Microstructures. ACS Nano 2014, 8, 12020-12029.

14 ACS Paragon Plus Environment

Page 14 of 19

Page 15 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

FIGURES

Figure 1. Detailed structure of large-area active-matrix (AM) MoS2 tactile sensor. (a) Schematic illustration of the fully integrated AM MoS2 tactile sensor array. (b) The corresponding layouts in planar (top) and cross-sectional (bottom) of a single sensor unit. (c,d) Circuit diagram and optical micrograph of TFT enabled single pressure gauge unit. (e) Photograph of the large area flexible AM MoS2 tactile sensor array.

15 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2. Electrical characterization of the AM MoS2 tactile sensor. (a) I–V characteristics of the MoS2 pressure gauge (without TFT). (b) Transfer characteristics of the MoS2 TFTs (64 devices). (c) Histograms of ON-current and threshold voltage of the MoS2 TFTs showing their statistical distribution of. (d) Transfer characteristics of TFT integrated pressure gauge unit as a function of applied pressure (from 0 to 120 kPa) at a drain voltage of 1 V. (e) Relative resistance change of the TFT enabled pressure gauge unit as a function of time at various pressure levels (1 kPa to 120 kPa). Inset shows its response feature at a low-pressure level (1 kPa), resulting the response time (tr) of 0.18 s. (f) Variation of relative resistance of TFT enabled pressure gauge unit as a function of the applied pressure from 1 to 120 kPa. (g) Repeatable measurement of relative resistance changes at different applied pressures for 10,000 cycles.

16 ACS Paragon Plus Environment

Page 16 of 19

Page 17 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Figure 3. Electrical and mechanical testing of AM MoS2 tactile sensor under bending. (a) Photograph of the AM tactile sensor array wrapped around the glass pipet (diameter: 1 mm). (b) Strain distribution of the AM MoS2 sensor array obtained by FEM analysis. Inset shows the cross-sectional strain distribution for the TFT. (c) Transfer characteristics of the MoS2 transistor and (d) I–V characteristics of the MoS2 pressure gauge under different bending radii. (e) Variation of normalized mobility, Vth and ON-current for MoS2 TFT whereas the relative resistance change for MoS2 pressure gauge under different bending radii.

17 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4. E-skin application of the large-area AM MoS2 tactile sensor on palm. (a) SEM image of the AM MoS2 tactile sensor array attached on a PDMS based palm mold. (b) The value of crosstalk isolation for the AM and PM-type tactile sensors. (c) Sensing of pressure exerted by a standing miniature chair on the AM MoS2 tactile sensor attached to the human palm. (d) Photograph of the AM MoS2 tactile sensor on human palm. (e) Real-time mapping of trajectory of the stylus movement on AM MoS2 tactile sensor. The pressure exerted by the stylus motion was ~20 kPa. (f) Mapping of shape sensing of an external object (apple) while gripping it by the AM MoS2 tactile sensor attached to the human palm.

18 ACS Paragon Plus Environment

Page 18 of 19

Page 19 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Table of Contents

19 ACS Paragon Plus Environment