2D Materials for Energy Harvesting and Sensing

The NASA NSTRF proposal entitled 2D Materials for Energy Harvesting and Sensing aimed to develop electronic devices using 2D materials for a self-powered sensing system that can conform to an arbitrary surface and monitor its environment. The first half of the project focused on developing ultrathin film Tungsten Diselenide (WSe2) solar cells, while the second half of the project focused on developing 2D material and other thin film infrared detectors. The Pt/WSe2 vertical schottky junction solar cells were fabricated and characterized to better understand the performance limitations and possibility of EQE, short circuit current and open circuit voltage improvement via surface passivation techniques. Specific accomplishments within this subproject include: (1) identifying an optimal WSe2 absorber thickness and (2) showing that the devices’ photovoltaic performance can be improved via Al2O3 passivation, which increases the EQE from 17.2% up to 29.5 % at 410 nm wavelength incident light. The overall resulting short circuit current improves through antireflection coating, surface doping, and surface trap passivation effects. Thanks to the Al2O3 coating, this work demonstrated a device with open circuit voltage (VOC) of 380 mV and short circuit current density (JSC) of 10.7 mA/cm2. In addition, the impact of Schottky barrier height inhomogeneity at the Pt/WSe2 contact was investigated as a source of open circuit voltage lowering in these devices. The second half of this project focused on the design and fabrication of two different types of infrared detectors: (1) a molecular junction bolometer (MJB) and (2) a pyroelectric (PE) gated MoS2 transistor. In addition, we developed a data processing scheme based on compressive sensing techniques to use an electric-field tunable bilayer graphene detector to reconstruct sparse signals in the 10 μm – 20 μm band with high accuracy. The proof-of-concept molecular junction bolometer is a metal – self-assembled monolayer (SAM) – metal suspended tunnel junction. Incident infrared light is absorbed in the metal arms, causing the material to expand and close the SAM nanogap. This then leads to a large increase in the tunneling current through the device. Thus far, we have demonstrated devices with temperature coefficient of resistance (TCR) as high as 28 %/K. In addition, we have characterized the low frequency 1/f noise in our devices and have found that both SAM growth across the nanogap junction and increasing the thickness of the metal cantilever arms can reduce high frequency noise. At 10 kHz, the 1/f noise is at 10x the shot noise level. Our calculations show that these devices can have thermal time constant limited bandwidths of up to 1 MHz. Our experimental results show that these devices have the potential to be several times faster than the than state of the art (with time constants on the order of 1 ms or larger) and also more sensitive due to the large TCR. In addition to the molecular junction bolometer, we have been developing pyroelectric (PE) gated MoS2 transistors for mid-infrared detection. Our material stack consists of a TiN gate, 10 nm thick hafnium zirconium oxide (Hf0.5Zr0.5O2), 10 nm non-ferroelectric oxide, and the semiconductor layer. These devices offer new functionality in that they can be operated as photoconductor detectors at shorts wavelengths (with energy greater than the semiconductor bandgap), and therefore are not limited by the background fluctuation noise. Meanwhile, the PE MoS2 transistor’s large TCR can enable Johnson Noise limited detectivities that can compete with the state of the art, while maintaining sub-1ms speeds. Early stage measurements show that we can achieve TCR values as high as 4.5 %/K at room temperature when biased in the subthreshold regime, which is greater than the state of the art ~2.5 %/K of vanadium oxide bolometers. Furthermore, we have demonstrated ferroelectric switching behavior in our devices that indicates the quality of our fabrication process. Ongoing work is being done to further characterize the TCR, detectivity, and time constants of these devices. Finally, we have explored signal processing techniques that can enable nm scale resolution sparse signal hyperspectral imaging using a gate tunable bilayer graphene device fabricated by Prof. Long Ju’s group at MIT. Exciton peaks in the 10 μm to 20 μm band can be blue-shifted by applying a large electric field (0.5 V/nm – 1.5 V/nm) across the active material. Simulations show that, by using a compressive sensing technique, we can reconstruct a signal of five superimposed gaussians with 10% broadband background noise. It is therefore possible to use a small array of programmable bilayer graphene photodetectors for either high spatial resolution or high temporal resolution hyperspectral imaging depending on the application of interest. No other infrared sensing technology that has this kind of functionality is currently available. We are continuing to explore the algorithm development and device characterization. Benefits from Further Development and Application: WSe2 and other Transition Metal Dichalcogenide (TMD) solar cells have the potenial to offer high efficiency with a thin form factor due to the large absorption coefficient in the visible wavelengths, enabling microscale sensing systems and self-powered flexible electronics. The detailed passivation study conducted in this work will be beneficial for improving the performance of single material, heterostructure, and tandem devices. The thermal infrared detector devices that we have fabricated have the potential to either compete or outperform the state of the art in terms of both speed and sensitivity. Furthermore, 2D-material based infrared detectors can offer new functionality as demonstrated by the gate tunable bilayer graphene detectors.