Asymmetric electrode metal/insulator/metal (MIM) tunnel diodes can perform as ultra-fast rectifiers for applications in THz energy harvesting and IR detection, but require low turn on voltage (V_ON) and low zero bias resistance (ZBR) as well as current-voltage asymmetry (f_asym) and non-linearity (f_NL). Combining bilayer insulators as tunnel barriers (MIIM diodes) can improve performance over conventional MIM diodes via asymmetric resonant tunneling or "step" tunneling [1]. Utilizing insulating materials with intrinsic defects can further improve η_asym and f_NL as defect levels in the smaller band gap insulator can provide additional conduction pathways [2]. Finally, it has also been demonstrated that intentionally introduced extrinsic defect levels, precisely introduced into the insulator using atomic layer deposition (ALD) [3] can be used to engineer MIM diode performance. In this work, we investigate the use of ALD to intentionally introduce impurity defect levels into the large bandgap insulator of dual insulator MIIM diodes.

Three Al/HfO₂/Al₂O₃/Pt MIIM diodes were investigated: (i) Ti doped: in which a two ALD cycle Ti defect layer was positioned within the middle of the Al₂O₃, and (ii) Ni doped: in which a two ALD cycle Ni defect layer was also positioned within the middle of the Al₂O₃, and (iii) an undoped control. ALD of HfO₂, Al₂O₃, NiO, and Ti₂O₅ was performed using TEMAHf/H₂O,TMA/H₂O, Ni(tBu₂DAD)₂/O₃, and TTIP/H₂O. For all devices, ALD was performed onto a bottom Pt electrode. After ALD, Al was e-beam evaporated through a shadow mask with 250 µm diameter holes to form top electrodes.

The Ni doped diodes were found to have improved maximum f_asym over the undoped control, but increased V_ON, likely due to suppression of conduction by to negative charge trapped at Ni defect levels lying energetically near or below the equilibrium Fermi level (E_F,equil) [4]. The Ti doped diodes showed slightly reduced leakage current, likely due to positive trapped charge in Ti defect levels near or above the E_F,equil, and also increased maximum f_asym. However V_ON was not reduced. Compared to the undoped control, introducing either Ni or Ti defect levels resulted in an increase in f_asym at higher fields, but a slight decrease at low fields due to charge induced band bending. Ni doped devices also demonstrated a slight increase in breakdown field strength. Additional results will be presented at the meeting including capacitance-voltage measurements. This work shows that ALD can be an effective tool for engineering device behavior.

1. Alimardani et al. Appl. Phys. Lett.102, 143501 (2013)
2. Alimardani et al. Appl. Phys. Lett. 105, 082902 (2014).
3. Holden et al. J. Appl. Phys. 129,144502 (2021).
4. Ichimura, J. Electron. Mat. 48, 583 (2019).

Transparent conductors (TCs) are unusual materials that are optically transparent to visible light like insulating glass yet have electrical conductivity like opaque metals. The TCs are useful for a broad range of applications including flat panel displays, light-emitting diodes, solar cells, Particularly rare but useful for optoelectronic energy conversion devices are transparent materials that have p-type electrical conductivity with holes rather than electrons (n-type) as majority charge carriers. In contrast to n-type TCs that are usually oxides, some of the top performing p-type TCs are nitrides (e.g. Mg:GaN) or chalcogenides (i.e. sulfides, selenides, tellurides).

In this presentation, I will focus on wide band gap chalcogenide materials as p-type transparent conductors for photovoltaic and photoelectrochemical solar cells. First, I will give an overview desired physical properties of TCs besides transparency and conductivity, and present high-throughput research workshop that can be used to experimentally and theoretically screen candidate materials for TC applications [1]. Then I will give two examples of how these design principles and research methods can be used to synthesize and characterize Zn1-xCuxS [2] and ZnTe1-xSex [3] chalcogenide p-type transparent conductors, and integrate them in CdTe thin film photovoltaic devices [4].

[1] Chem. Rev. 2020, 120, 4007; [2] Matter 1 862 (2019); [3] J. Mater. Chem. C, 10, 15806 (2022); [4] ACS Applied Energy Materials 3 5427 (2020)

The ability to create and manipulate materials in two-dimensional form has repeatedly had transformative impact on science and technology. We have developed a general method to create freestanding complex oxide membranes and heterostructures using epitaxial water-soluble buffer layers, with millimeter-scale lateral dimensions and nanometer-scale thickness. This facilitates many new opportunities we are beginning to explore; here we will focus on the use of tensile strain and strain gradients to control the ferroelectric and flexoelectric response of oxide membranes.