DFT calculations for cadmium telluride (CdTe)

Abstract

Cadmium Telluride (CdTe) thin film photovoltaics (PV) has demonstrated low Levelized Cost of Energy (LCOE). CdTe technology also counted for half the thin film market in 2013 [3]. CdTe PV has the smallest carbon footprint a nd the energy payback time (less than one year) is the shortest of any current photovoltaic technology. The modules made of CdTe can also be recycled at the end of their lifetime. The attractiveness of these materials comes from their bandgap value (1.5 eV), which falls within the solar spectrum, thereby enabling the efficient creation of electron-hole pairs (or excitons) by solar photons. This has led to the research that dates back to 1950's and is currently ongoing in many parts of the world. A simple h eterojunction cell design was evolved in which p-type CdTe was matched with n-type Cadmium Sulfide (CdS) and by adding the top and bottom contacts. Today, multiple crystalline layers, of thicknesses ranging from a few nanometers (nm) to tens of micrometers (μm), are added to improve the efficiencies of the CdTe PV cells. The highest cell efficiency recorded to date is over 22%. Different computational tools and methods are used to study these effects, with Quantum ESPRESSO and VASP being used for many years now. QuantumATK, built in 2008 by the company Atomistix and acquired by Synopsys in 2017, is a simulation tool that uses Density Functional Theory (DFT) for atomistic-scale modelling of nanostructures. In this work, QuantumATK was used to predict the structural properties of bulk CdTe. Different exchange-correlation (XC) functionals were used to perform the calculations. Firstly, the crystal structure of bulk CdTe was predicted using the tool. Later the properties like lattice parameter, were calculated. In addition to structural properties, the electrical properties were also predicted using different XC functionals. Also, the XC functionals that correct the bandgap obtained from the standard functionals were used to predict the bandgaps and the results were also compared to the experimental values again to see how accurately does QuantumATK predicts the electrical properties of bulk CdTe. The LDA and GGA XC functionals, predicted the band gap for bulk CdTe with error percentages of 57% and 40% respectively, when compared to the experimental value. The more accurate MGGA predicted the band gap with a 24% error while HSE06 (hybrid functional) predicted within 4% of experimental value. The LDA-1/2 and GGA-1/2 predicted the band gap most accurately within 2% of the compared experimental value. All the different XC functionals predicted the crystal structure correctly and the lattice parameter was within 2.2% of the experimental value.