Effects of Copper Substitution by Alkali Metals on the Properties of Chalcopyrites for Tandem Applications: Insights from Theory

Abstract

The effect of copper substitution by alkali metals on the properties of chalcopyrite-type materials for tandem applications in photovoltaics is investigated at the first-principles level, using an exchange-correlation hybrid functional optimized to yield a description of the structural, electronic, and dynamic properties of these materials in good agreement with experiment. Since the target values of the band gap for tandem applications should be between 1.5 and 1.8 eV, one part of the results concerned the variation of calculated band gap values under the effect of substitution. A systematic study of the effects of Li, Na, K, Rb, and Cs on the structural, electronic, and thermodynamic properties of CuGaS2, CuGaSe2, CuInS2, and CuInSe2 has been performed. The evolution of the crystallographic cell with the concentration of alkali metals turned out to be of two types: (i) the substitution of Cu with Li and Na in CuInS2, irrespective of concentration, leaves the underlying chalcopyrite structure unchanged, affecting only the lattice parameters; (ii) the substitution of Cu with Na (with the exception of CuInS2), K, Rb, and Cs at sufficiently high concentration brings about a phase transition. In all cases, the band gap increases with the alkali concentrations, whereby only the indium-based chalcopyrites reach the above-mentioned target values for applications in tandem photovoltaic devices. The static stabilities of the substituted materials have been further discussed in terms of substitution energies and energies of formation, with the latter being evaluated to secondary phases plausible in the process of synthesis. The comparison with the experimental situation and the impact of the novel predictions are discussed.

Cu1–xCsxInSe2: Candidate Absorber for Higher Conversion Efficiency and Broader Band Photodetector

“…Figure shows the band structures along the high-symmetry k -point paths of Cu 1– x Cs x InSe 2 compared with those of pristine CISe. For doping cases of low Cs content of x ≤ 0.25, partial substitution of Cs for Cu results in a slight decrease in band gap E g from 1.04 eV ( x = 0) to 1.019 eV ( x = 0.0625) and then to 1.009 eV ( x = 0.125), which is consistent with previous studies . In particular, continuing to increase the content of Cs will further reduce the band gap, such as the band gaps of 0.861 eV for x = 0.25 and 0.761 eV for x = 0.5, which perfectly meets the demand for broadband photodetectors.…”

“…For doping cases of low Cs content of x ≤ 0.25, partial substitution of Cs for Cu results in a slight decrease in band gap E g from 1.04 eV (x = 0) to 1.019 eV (x = 0.0625) and then to 1.009 eV (x = 0.125), which is consistent with previous studies. 33 In particular, continuing to increase the content of Cs will further reduce the band gap, such as the band gaps of 0.861 eV for x = 0.25 and 0.761 eV for x = 0.5, which perfectly meets the demand for broadband photodetectors. However, when the doping concentration is 100%, the direct band gap is estimated to be 2.534 eV, which is in agreement with the reported 2.66 eV.…”

Search of chalcopyrite materials based on hybrid density functional theory calculation

“…We, however, did not consider the alkali metals for the oxidation of +1 because they prefer other phases than chalcopyrite [36][37][38]. The alkali metals also likely distort the chalcopyrite structure as those have no d orbitals that form the valence bands in CIGS [39,40]. B, Al, Ga, In, and Tl were included in the set as a group III element.…”

Validation of Density Functional Theory Methods for Predicting the Optical Properties of Cu-Based Multinary Chalcogenide Semiconductors

“…To gain a deep understanding of the distinctive physical properties, it is critical to obtain an accurate description of the electronic structures of copper chalcogenide semiconductors, especially the information about band gap and band structure topology. Correspondingly, theoretical investigations on the electronic structures of Cu chalcogenides have attracted increasing interest in the past two decades. − However, the presence of strongly localized d electrons of Cu atoms leads to that the density functional theory (DFT) calculations within the local density approximation (LDA) and the generalized gradient approximation (GGA) are usually difficult to accurately describe the electronic structures of Cu-based multinary semiconductors, including the substantial underestimation of the band gap or the wrong prediction of energy band ordering. Even by using the DFT plus the Hubbard U correction (DFT + U ) approach that has been widely used in first-principles studies for strongly correlated systems, the calculated results still deviate obviously from the experimental measurements, although the prediction of the band gaps has been improved .…”