Exploiting Ionic Radii for Rational Design of Halide Perovskites

Abstract

Emerging perovskite solar cell technology has inspired extensive study of perovskite materials, with a recent focus on the search of stable and environmentally benign lead-free alternatives. The huge compositional space of the perovskites family necessitates rational strategies, for which predictive descriptors that relate target properties to a few key variables are crucial. Among others, ionic radius plays key roles on many important materials properties. In this work, to unlock the full potential of ionic radius for rational materials design (1) the widely used Shannon radii data set were consistently expanded to more than twice as large by extending to unusual oxidation states and arbitrary coordination, (2) new criterion for improved prediction of perovskites formability was proposed, (3) a general and predictive band gap physical descriptor with radius and normalized volume as inputs was discovered for a wide range of three-dimensional halide perovskites. These results offer new opportunities for perovskites design by exploiting radii.

Thermodynamic Preference for Atom Adsorption on versus Intercalation into Multilayer Graphene

“…To test this expectation, in Figure 2 we plot the revised effective ionic radius (often called the Shannon radius) 19 as a function of ∆ . Also note that the Shannon radius for one specified element can vary, depending on charge 0 , coordination number , spin state, and experimental method or theoretical model 19,20 . In Figure 2, we use the values (Table S7.1) corresponding to the most common valences 0 .…”

“…For coordination numbers, we approximately take = 6, by considering the bonding structures between the atom and its surrounding C atoms in the gallery. For spin states, the values do not have significant difference between high and low spin states, and we take the values suggested by Ouyang 20 . From Figure 2, one can immediately see that Ag and Au with much more positive ∆ values, therefore being much less favorable for intercalation, have larger values, while the other 11 elements with negative ∆ values (Pd, Pt, Cu, Rh, Dy, Ir, Ru, Ni, Fe, Co, and Os from small to large), therefore being more favorable for intercalation, have smaller values.…”

Beyond Tolerance Factor: Using Deep Learning for Prediction Formability of ABX3 Perovskite Structures

“…Values of were taken the same as in the origin of the dataset. [ 12,13,29 ] We have to note that using different radii values [ 30,31 ] may lead to different results. The best of these models with input features reached accuracies: for up to 91.07% on the test set (86.92% on the training set; , ), for up to 98.31% on the test set (97.81% on the training set; , ), and for the merged case, the model reached accuracy up to 95.65% on the test set (95.51% on the training set; , ).…”

“…Values of r i were taken the same as in the origin of the dataset. [12,13,29] We have to note that using different radii values [30,31] may lead to different results. The best of these models with input features The growth of the model's accuracies correlates with our hypothesis noted above.…”

Influence of A site cation on nonlinear band gap dependence of 2D Ruddlesden–Popper A2Pb1−xSnxI4 perovskites

“…The ionic radius of Sn 2+ for a coordination number of 6, as found in perovskites is 1.04 Å, compared to 1.19 Å in the case of Pb 2+ so is a favourable substitution. 15 Tin based perovskites are generally more unstable in experimental conditions, mostly due to a lower oxidation threshold in Sn 2+ to Sn 4+ , leading to decomposition of the perovskite. 16 Mixed Pb-Sn perovskites have been shown to perform well, with a good combination of power conversion efficiencies and much higher stability than pure Sn systems.…”