Effect of metal cation replacement on the electronic structure of metalorganic halide perovskites: Replacement of lead with alkaline-earth metals

Organic and inorganic lead halogen perovskites, and in particular, CH3NH3PbI3, have during the last years emerged as a class of highly efficient solar cell materials. Herein we introduce metalorganic halogen perovskite materials for energy-relevant applications based on alkaline-earth metals. Based on the classical notion of Goldschmidt's rules and quantum mechanical considerations, the three alkaline-earth metals, Ca, Sr, and Ba, are shown to be able to exchange lead in the perovskite structure. The three alkaline-earth perovskites, CH3NH3CaI3, CH3NH3SrI3, and CH3NH3BaI3, as well as the reference compound, CH3NH3PbI3, are in this paper investigated with density functional theory (DFT) calculations, which predict these compounds to exist as stable perovskite materials, and their electronic properties are explored. A detailed analysis of the projected molecular orbital density of states and electronic band structure from DFT calculations were used for interpretation of the band-gap variations in these materials and for estimation of the effective masses of the electrons and holes. Neglecting spin-orbit effects, the band gap of MACaI(3), MASrI(3), and MABaI(3) were estimated to be 2.95, 3.6, and 3.3 eV, respectively, showing the relative change expected for metal cation exchange. The shifts in the conduction band (CB) edges for the alkaline-earth perovskites were quantified using scalar relativistic DFT calculations and tight-binding analysis, and were compared to the situation in the more extensively studied lead halide perovskite, CH3NH3PbI3, where the change in the work function of the metal is the single most important factor in tuning the CB edge and band gap. The results show that alkaline-earth-based organometallic perovskites will not work as an efficient light absorber in photovoltaic applications but instead could be applicable as charge-selective contact materials. The rather high CB edge and the wide band gap together with the large difference of the electron and hole effective masses make them good candidates for n-type selective layers in hot carrier injection solar cell devices together with some light absorber candidates. The fact that they have similar lattice constants as the lead perovskite and suitable positions of the valence band edges open up the possibility to use them also as thin epitaxial p-type hole selective contacts in combination with the lead halogen perovskite materials. This can lead to both charge selectivity as well as to superior crystal growth of lead perovskite with less contact stress, which is interesting for further investigations.

Goldschmidt's Rules and Strontium Replacement in Lead Halogen Perovskite Solar Cells: Theory and Preliminary Experiments on CH3NH3SrI3

Ulf Anders Hagfeldt, Tor Jesper Jacobsson

During the past few years, organic lead halogen perovskites have emerged as a class of highly promising solar cell materials, with certified solar cell efficiencies now surpassing 20%. Concerns have, however, been raised about the possible environmental and legalization problems associated with a new solar cell technology based on a water-soluble lead compound. Replacing lead in the perovskite structure: with a less toxic element, without degrading the favorable photo physical properties, would therefore be of interest. In this paper, the possibility of replacing lead with other metal ions is explored by following the replacement rules of Goldschmidt together with additional quantum mechanical considerations. This analysis provides a conceptual toolbox toward replacing lead, as well as additional insights into the photo physics of the metal halogen perovskites. This approach is exemplified by focusing on strontium in particular, which is nontoxic and relatively inexpensive. The ionic radius of Sr2+ and Pb2+ are almost identical, suggesting an exchange could be made without affecting the crystal structure. Couple cluster calculations on the metal ions and their halogen salts give the bonding patterns to be sufficiently similar and density functional theory (DFT) revealed the strontium perovskite, CH3NH3SrI3, to be a stable phase, despite the difference in electronegativity between lead and strontium. This is further supported by the existence of binary PM, and SrI2 compounds and the beneficial formation energy of the strontium perovskite. The electronic properties of both CH3NH3SrI3 and CH3NH3PbI3 were simulated and compared, revealing a higher degree of ionic interaction in the metal halogen bound in the strontium perovskite. This is a consequence of the lower electronegativity of strontium, which, together with the lack of d-orbitals in the Valence of Sr2+, results, in a higher band gap. The band gap for the strontium perovskite was estimated to 3.6 eV, which unfortunately is too high for an efficient photo absorber. Initial investigations on experimental synthesis of the strontium perovskite, using wet chemical methods, revealed it to be harder to produce than the lead perovskite This is explained as a:consequence of different bonding patterns in the metal iodine salts, which obstruct the methylammonium intercalation pathway utilized for forming the perovskite. Vapor phase methods are instead suggested as more promising synthesis routes.

American Chemical Society2015

Nonempirical hybrid functionals for advanced electronic-structure calculations

Thomas Bischoff

Electronic-structure calculations based on hybrid functionals have emerged as a standard technique used in physics, chemistry, and material science. Despite this success, hybrid functionals have the drawback of containing undetermined parameters. To overcome this deficiency, two different nonempirical determination schemes are being investigated at present, namely dielectric-dependent hybrid (DDH) functionals and hybrid functionals that satisfy Koopmans' condition. This thesis is dedicated to the examination and further development of these approaches. In particular, we show that a precise description of band gaps in condensed-matter systems can be achieved with these functionals. Moreover, we demonstrate that their accuracy is comparable to state-of-the-art

GW

methods while requiring substantially lower computational cost. First, we focus on the development of hybrid functionals satisfying Koopmans' condition. We show that the construction of those functionals can be optimized through suitably defined potential probes. By monitoring the delocalized screening charge, we achieve a measure of the degree of hybridization with the band states, which can be used to improve the band-gap estimate. We show that the application of this methodology to common semiconducting and insulating materials yields band gaps differing by less than 0.2 eV from experiment. These conceptual developments are an important step towards establishing hybrid functionals satisfying Koopmans' condition as a robust approach for band-gap predictions. Second, we examine DDH functionals and hybrid functionals satisfying Koopmans' condition for band gaps of more sophisticated materials. In particular, we focus on inorganic metal-halide perovskites which have recently drawn great scientific attention. For this class of materials, we show that both nonempirical hybrid-functional schemes yield band gaps of comparable accuracy (

\sim

0.2 eV) with respect to

GW

reference calculations. Furthermore, we discuss the suitability of nonempirical hybrid-functional schemes for the application to the screening of large sets of perovskite materials. Third, we investigate the fundamental band gap of liquid water and hexagonal ice. These materials are particularly challenging since experimental studies have not reached a consensus on the band gap yet. Therefore, we first deduce robust benchmarks on the basis of a critical review of various experimental studies in the literature. Then, we compute the band gap through state-of-the-art

GW

methods as well as nonempirical hybrid functionals. We show that theoretical calculations and experimental references are in good agreement with each other and we discuss critical aspects which are essential to ensure a consistent description of the band gap. Finally, we investigate band-edge levels as obtained with hybrid-functional calculations. The CaF

_2

/Si(111) interface serves thereby as an ideal test case to examine the accuracy of different theoretical schemes. The comparison with experiment reveals that global hybrid functionals and self-consistent

GW

methods provide the most accurate description of the interfacial band alignment.

EPFL2021

Electronic Transport in 2D Materials with Strong Spin-orbit Coupling

Artem Pulkin

The thesis describes the computational study of structural, electonic and transport properties of monolayer transition metal dichalcogenides (TMDs) in the stable 2H and the metastable 1T' phases. Several aspects have been covered by the study including the electronic properties of the topological quantum spin Hall (QSH) state in the 1T' monolayer phase as well as the effects of strain, periodic line defects, interfaces and edges of monolayer TMDs. The electronic properties of the bulk monolayer phases were described by the ab-initio density functional theory framework while the electronic and transport properties of 1D defects were calculated using the non-equilibrium Green's function formalism. A specific focus was made on the transport of spin-polarized charge carriers across line defects in the monolayer 2H phase. Subject to energy, pseudomomentum and spin conservation, the size of the transport gap is governed by both bulk properties of a material and symmetries of a line defect. Outside the transport gap energy region, the charge carriers are discriminated with respect to their spin resulting in the spin polarization of the transmitted current. Next, the properties of the metastable monolayer 1T' phase were studied. The presence of a sufficiently large band gap is crucial to observe the QSH phase in the family of materials by probing the topological boundary states. The meV-order band gaps of the 1T' phase of monolayer TMDs were found to be sensitive to materials' lattice constants suggesting the control of the band gap size by strain. In particular, the electronic band structure and the size of the band gap in monolayer 1T'-WSe2 were found to be in agreement with spectroscopy studies. The topologically protected states at the edges of the monolayer 1T' phase as well as at the boundaries between the topological 1T' phase and the trivial 2H phase of monolayer TMDs were studied. Specific atomic structure configurations were suggested to observe experimentally the topological protection of the charge carrier transport against back-scattering. Finally, in the context of lateral semiconducting device engineering, the electronic and transverse transport properties of 2H-1T' phase boundaries as well as the dimerization defects in the 1T' phase were investigated. Both kinds of defects considered exhibit a relatively large transmission probability for the charge carriers crossing the defects. However, the differences between the shapes of bulk bands of the two phases open a sizeable transport gap for charge carriers crossing periodic domain boundaries between the monolayer 2H and 1T' phases. The calculated formation energies of dimerization defects were found to be relatively low suggesting their high concentration in real samples of monolayer 1T'-TMDs. Additionally, the thesis includes studies of magnetic dopants on the surface of Bi2Te3 and atomic vacancies in monolayer 2H-MoSe2 where the electronic properties of point defects were calculated and compared to experimental results. The two possible adsorption sites of Fe on the surface of Bi2Te3 both show a large out-of-plane magnetic anisotropy in agreement with experiments. The calculated local electronic properties of Se vacancies in monolayer 2H-MoSe2 show the presence of in-gap states which are not observed in experiment. Nevertheless, the combination of theoretical and experimental scanning tunneling microscopy images allowed the unambiguous identification of the vacancy defect.

EPFL2017