Current research

My current interests include the following. Interested students are encouraged to contact me at redwansajjad@gmail.com for potential projects.

- - Fabrication and characterization of thin-film photovoltaics.
  - Various coatings on glass for energy-efficient glass, smart windows.
  - Energy harvesting through plasmonic nanoparticles.
  - Modeling and characterization of thermal properties of coated glass.
  - Physics-based device modeling and DFT simulation of materials.

Current students

- - Mayrazul Hoque (funded through EPRC) - Thin film deposition using physical vapor and chemical methods for solar cell application.
  - Shamsun Nahar - Nanoparticle synthesis with the sol-gel method.
  - Muntasir Hassan (co-supervision with EEE, BUET) - Energy saving potential of photovoltaics windows.
  - Mamunur Rahman - Fabrication of Electrical Porcelain Insulator
  - Tareq Morshed - Ceramic thermal barrier coating

Past students

- - Eymana Maria (co-supervision with EEE, BUET) - Carrier transport in ultra thin-film solar cells.
  - Khalid Hossain (GCE, BUET), MSc thesis – Performance estimate of Semi-Transparent Photovoltaics Through Modeling and

Characterization - December, 2020

- - Genya Crossman, Undergraduate student intern, MIT - Novel two-dimensional heterostructure Tunnel FET devices - Summer 2016.
  - Mark Cheung, Undergraduate thesis, UVA – Performance limits of Graphene Nano-Ribbon (GNR) based Tunnel FET - Summer 2013.

Carrier transport in semi-transparent photovoltaic (STPV) devices

STPV is an emerging variant of thin-film photovoltaic technology. It partially absorbs incident solar radiation while transmitting the rest thus working as a transparent power generating window. The ultra-thin absorber leads to novel transport phenomena -1) depleted absorber resulting in negligible effect of bulk and back-surface recombination, 2) small grain size, very short carrier life-time, 3) unusually high series resistance resulting from grain boundary resistance, 4) small shunt resistance due to ultra-thin absorber. Transport phenomena are explained by FDTD simulation, drift-diffusion transport, and compact models of series and shunt resistances with benchmarking against recently published experiments. Performance estimate is projected for improved material parameters (JAP, 2021).

Energy-saving potential of photovoltaic windows

A major portion of electricity generated worldwide is used in buildings - responsible for nearly 75% of electricity consumption and a major player in global climate change. A large share of this electricity is used for HVAC and lighting loads. Although windows are the pathways for natural light, which is vital for human health, they are poor thermal insulators thus allowing heat leakage in and out of the building. This fundamental trade-off between enjoying natural light and lower electricity consumption can only be avoided with power or photovoltaic windows, where semi-transparent photovoltaics is deposited on clear glass using thin-film technology. Although thin-film PV technology has matured over the last few decades, STPV device fabrication for optimized building energy has been initiated only recently. In our work, we have developed a device-to-system level model to estimate the energy-saving potential of power windows ( Applied Energy, 2021, Solar Energy, 2022). Device performance based on thermodynamics limits and practical limits in presence of non-idealities has been used in performing system-level (building) energy simulation using Energyplus. Our model includes a solar insolation model (using PVLIB, Sandia National Laboratory), cloud model (NASA), thermal models for U-value and solar heat gain coefficient and shadow model for urban environment. Our model projects a 40-60% saving in building energy consumption (depending on location).

Deposition of solar materials

To get the optimum efficiency for a given visible transparency, each layer in the STPV film needs to be optimized. In this work, we optimize the cadmium sulfide (CdS) buffer layer for the CIGS based STPV using the chemical bath deposition technique. Several samples are deposited on clear glass and the optimum combination of deposition time, solution temperature, stirring rate and cadmium-sulfur reactant ratio is found. The thickness measured from SEM micrograph is found to be 66 nm, which is the range of optimum thickness for this type of application (ICTP, 2021).

Previous research

My graduate studies and the subsequent postdoctoral research involved the computational study of quantum-mechanical effects in carrier transport – namely the quantum transport - in exploratory nanoelectronic devices. Due to the miniaturization of transistors, the building blocks of modern microprocessors, our understanding of how current flows at the nanoscale has become quite important. The complete understanding and the solutions, which may incorporate the use of new materials, often require a multidisciplinary approach to model, characterize and design devices of such scale. I studied several novel materials, studied the new physics and their material aspects and how they affect the device performance.

Tunnel FETs (MIT): TFETs have the potential for very low power switching due to their low standby power and steep subthreshold swing. Despite the promise, experimental evidence of the tunnel driven switching has been weak. My work at MIT focused on understanding the role of leakage paths from interface traps and intrinsic band steepness on the TFET performance. I developed a multi-phonon based trap assisted tunneling (TAT) model that for the first time explained how TAT can obscure the steep transition of current in TFET (IEEE TED 2016). My model has already been tested with experimental data from leading groups (IEDM 2016, IEEE TED 2017) and has made significant impact towards understanding TFET performance limits. Based on the full model, a compact model has been developed to simulate circuits made of TFETs (SSE 2018).

(a) Tunnel FET has the promise of low power switching due to its tunneling-triggered steep subthreshold swing. (b) Because of non-idealities such as the lack of pristine material interfaces, practical TFETs produce higher leakage currents than ideal. (c) Numerical as well as compact models are developed to explain the experimental results, which 1) allows to predict performance in the absence of interface traps shown in (d), 2) provides guidance for future material growth and device fabrication.

2D materials (MIT): 2D materials comprise a promising set of materials that have short natural scale length λ due to its thin body structure. MoS2 is among the most promising for ultra short channel transistors due to its suitable effective mass and electronic bandgap. To test its potential, Prof. Palacios group at MIT built serially connected multiple 7.5 nm MoS2 transistors with promising switching behavior. With NEGF simulation combined with the MIT Virtual Source (MVS) model, I modeled the details of the transistor behavior and extracted the intrinsic device performance of each individual device after eliminating parasitic effects (Nano Letters 2016). This work has established that in the limit of low contact resistance and scaled oxide, sub-10 nm MoS2 transistor is a promising choice for low power logic applications.

A rich set of phenomena and device structures is possible with 2D materials based on the choice of material (left). Our recent work (right) on 2D-MoS2 based transistors show promising results for transistor applications in the deeply scaled sub-10 nm regime, where it shows certain advantages over conventional semiconductors.

Graphene (UVa): A rich set of phenomena emerge once we look at heterojunctions made of gateable gapless 2D materials, e.g. graphene, Topological Insulators (TI). Since the bandstructure of graphene resembles photons, there are a lot of similarities that can be exploited to build new kinds of devices. The devices are based upon the pseudospin (or spin) - momentum locking in graphene (or TI), electrical analogues of photonic principles (such as Snell’s law, negative refractive index) and the ability to control the heterojunctions with gate voltage. Since the lack of bandgap limits graphene’s application, I proposed a device concept to turn OFF graphene by using a unique angular transmission property of GPNJ (ACS Nano 2013). The OFF state resistance is increased by placing the gates at different angles relative to each other to filter out all propagating modes. This creates a transmission gap, which disappears in the uniformly doped limit (no junction). Experimental efforts demonstrated the angle dependent transmission and signatures of increased ON-OFF ratio (PRB 2012).

While regular materials suffer from Boltzmann tail (the high energy part of Fermi-Dirac distribution) and thermally limited subhthreshold swing (kBTlog10/decade), a technology with collapsible transport/bandgap will avoid such limit. This is demonstrated in graphene pn junction utilizing its voltage dependent transport gap.

Topological insulators (UVa): Topological Insulator (TI) is a new class of materials that has been described as a new state of material, which is insulating in the bulk while exhibiting conducting spin polarized states on the surface (or on the edge of a 2D TI). It has an intriguing similarity with graphene; its surface states (for 3D TI) can be described with a Dirac Hamiltonian. Using the lessons learned with graphene, we reported an extremely large spin-charge current gain, β=Js/Jc using TI pn junctions (PRL 2015). Our reported β is larger (20-30 vs. <1) than any other spin systems (such as tantalum).

(Left) Top view of the device showing the directions of incident, the reflected and transmitted electrons, and their spins. (right) Charge and spin current vs gate voltage.

Quantum transport modeling (UVa): During my graduate studies at the University of Virginia, I developed expertise in quantum transport modeling using Non-Equilibrium Green’s Function (NEGF)-Landauer formalism. NEGF is a powerful, bottom-up approach to understand nanoscale device physics. It can capture transport in both ballistic and diffusive regimes and model contact and different boundary conditions. The method is sufficiently general to model different kinds of interactions, such as, electron-phonon, electron-photon etc.

I used NEGF and analytical formalism to study graphene, graphene pn junction (GPNJ) and topological insulators. My studies on graphene involved detailed bandstructure calculations from both Extended Hückel theory and atomistic tight binding methods (JCE 2013). In order to match experimental results of large scale graphene devices (up to micron), we developed an efficient numerical platform (incorporating parallel cluster computing and matrix manipulations) to capture diffusive transport for the first time using NEGF (PRB 2015, APL 2016).

Tracking electron trajectories in Graphene pn junction - a testbed for optics like electron behavior in materials. Carrier density mapping from NEGF matches closely with the classical ray tracing analysis. Efficient matrix inversion algorithm combined with parallel programming enabled simulation for a significantly large graphene sheet (100nm x 100nm).