The primary mechanism of operation of almost all transistors today relies on electric-field effect in a semiconducting channel to tune its conductivity from the conducting 'on'-state to a nonconducting 'off'-state. As transistors continue to scale down to increase computational performance, physical limitations from nanoscale field-effect operation begin to cause undesirable current leakage that is detrimental to the continued advancement of computing 1,2 . Using a fundamentally different mechanism of operation, we show that through nanoscale strain engineering with thin films and ferroelectrics (FEs) the transition metal dichalcogenide (TMDC) MoTe2 can be reversibly switched with electric-field induced strain between the 1T'-MoTe2 (semimetallic) phase to a semiconducting MoTe2 phase in a field effect transistor geometry. This alternative mechanism for transistor switching sidesteps all the static and dynamic power consumption problems in conventional field-effect transistors (FETs) 3,4 . Using strain, we achieve large non-volatile changes in channel conductivity (Gon/Goff~10 7 vs. Gon/Goff~0.04 in the control device) at room temperature. Ferroelectric devices offer the potential to reach sub-ns nonvolatile strain switching at the attojoule/bit level 5-7 , having immediate applications in ultra-fast low-power non-volatile logic and memory 8 while also transforming the landscape of computational architectures since conventional power, speed, and volatility considerations for microelectronics may no longer exist.We design our device using single crystal oxide substrates of relaxor ferroelectric Pb(Mg1/3Nb2/3)0.71Ti0.29O3 (PMN-PT) as the gate dielectric (0.25-0.3 mm thickness). On top of this ferroelectric substrate we exfoliate 1T'-MoTe2 (13-70 nm) from a single crystal source, and pattern devices using Ni contact pads (Figure 1a,b). Exfoliation was performed in a humidity-controlled environment for increased adhesion. Depending on the contact material, stress from the deposited thin film strains the MoTe2 channel at the contact pads, analogous to the uniaxial strain techniques from strained silicon technology, widely adopted in industrial CMOS processes 9 . We find that contact metal stress and TMDC to substrate adhesion are both critically important for obtaining a functional device. During fabrication we are careful not to increase the temperature of the ferroelectric above 80° C, well below the Curie temperature of the ferroelectric at 135° C. Upon reaching the Curie temperature, the Device fabrication performed by W.H., A.S., T.P., and A.A.; Device characterization performed by W.H., A.A., and S.M.W.; Conductive atomic force microscopy performed by W.H., A.S., and S.M.W.; Strain gauge calibration performed by W.H. and S.M.W.; Topographic atomic force microscopy and optical contrast calibration performed by T.P.; Thin film stress measurements performed by C.W., A.A., W.H., and S.M.W.; Piezoresponse force microscopy performed by C.W.; Raman spectroscopy performed by A.A. and S.M.W.; Finite element analysis simul...

The interaction of intruding objects with deformable materials is a common phenomenon, arising in impact and penetration problems, animal and vehicle locomotion, and various geo-space applications. The dynamics of arbitrary intruders can be simplified using Resistive Force Theory (RFT), an empirical framework originally used for fluids but works surprisingly well, better in fact, in granular materials. That such a simple model describes behavior in dry grains, a complex nonlinear material, has invigorated a search to determine the underlying mechanism of RFT. We have discovered that a straightforward friction-based continuum model generates RFT, establishing a link between RFT and local material behavior. Our theory reproduces experimental RFT data without any parameter fitting and generates RFT's key simplifying assumption: a geometry-independent local force formula. Analysis of the system explains why RFT works better in grains than in viscous fluids, and leads to an analytical criterion to predict RFT's in other materials. IntroductionThe interaction of solid objects with a surrounding, plastically-deforming media is a common aspect of many natural and man-made processes. In the animal world, when organisms undulate, pulse, crawl, burrow, walk, or run on loose terrain they implicitly deform their environment to produce propulsive reaction forces giving rise to their motion (1). The physics of such interactions have been the subject of many studies, from aquatic organisms (2,3) to small insects and lizards (4,5) to humans and other legged-mammals (6,7). Similar principals are used for robotic applications to design machines that run (8), fly (9), swim (10), or walk in fluids or sand (11,12).Such complex interactions are also key to terramechanics of vehicular locomotion on granular substrates, models of excavation in sand and soil (13,14), and the study of similar problems in extraplanetary conditions (15,16). These topics and others, including cratering dynamics and penetration in plastic solids (17,18), all depend crucially on the way local material properties produce global resistive forces on arbitrary intruders.Motivated by past observations in fluids (19), a simple yet very effective empirical tool known as Resistive Force Theory (RFT) for granular materials has been proposed to approximate the forces on intruding objects moving through granular media. Despite the fact that a fundamental derivation is missing, when coupled with the force balance equations, the theory provides a simple and predictive tool for simulating the locomotion of arbitrarily shaped moving bodies in loose terrain (5,20,21). The simplicity of the theory and its predictive effectiveness are surprising in light of the complex, nonlinear, history-dependent, and oftentimes visibly nonlocal constitutive properties of granular media (22)(23)(24)(25)(26). RFT was initially developed to approximate 2 the speed of swimming micro-organisms at low Reynolds numbers (27) by studying the thrust and drag of individually moving elements of its bod...

No abstract

In this work, we induce on-chip static strain into the transition metal dichalcogenide (TMDC) MoS2 with e-beam evaporated stressed thin film multilayers. These thin film stressors are analogous to SiNx based stressors utilized in CMOS technology. We choose optically transparent thin film stressors to allow us to probe the strain transferred into the MoS2 with Raman spectroscopy. We combine thickness dependent analyses from Raman peak shifts in MoS2 and atomistic simulations to understand the strain transferred throughout each layer. This collaboration between experimental and theoretical efforts allows us to conclude that strain is transferred from the stressor into the top few layers of MoS2 and the bottom layer is always partially fixed to the substrate. This proof of concept suggests that commonly used industrial strain engineering techniques may be easily implemented with 2D materials, as long as the c-axis strain transfer is considered.

No abstract