Comment on “Interplay of Structural and Optoelectronic Properties in Formamidinium Mixed Tin‐Lead Triiodide Perovskites”

Abstract

Studying optoelectronic properties inFAPb 1−x Sn x I 3 and in FA 0.83 Cs 0.17 Pb 1−x Sn x I 3 perovskites as a function of the lead:tin content, Parrott et al. (2018) and Savill et al. (2020) observed the broadest luminescence linewidth and the largest Stokes shift in mixed compositions with Sn <25% and with >85%. It is in contrast to the intuitive expectation of the largest effects of alloy disorder for the 50:50 composition. This comment addresses the alloy disorder caused by statistical local spatial fluctuations of the alloy composition and shows that the largest effects of alloy disorder for perfectly random fluctuations in FAPb 1−x Sn x I 3 and FA 0.83 Cs 0.17 Pb 1−x Sn x I 3 are, in fact, expected for x < 0.25 and for x > 0.85. It can be one of the reasons why Pb-rich and Sn-rich Sn-Pb perovskites typically show shorter photoluminescence (PL) lifetimes, broader emission, increased Stokes shifts, reduced PL quantum yield, and higher Urbach tails, compared with their lead-only counterparts.

Energy Scales of Compositional Disorder in Alloy Semiconductors

“…Using eq , we show below that the largest effects of alloy disorder are, in fact, expected for x a < 0.25 and for x b > 0.85. Moreover, the effect at x a should be much more pronounced than that at x b , in agreement with the experimental data. , Therefore, further crystal growth optimization might not be helpful to diminish the effects of alloy disorder in FAPb 1– x Sn x I 3 and FA 0.83 Cs 0.17 Pb 1– x Sn x I 3 perovskites . In other words, the samples have already achieved a high quality with respect to the effects of alloy disorder.…”

“…In Figure , the composition dependence ε 0 ( x ) in accord with these equations is plotted. In the calculations, a concentration of lattice sites N = 4 × 10 21 cm –3 was used, based on the estimate for the lattice constant a ≃ 6.3 × 10 –8 cm . There is uncertainty in the choice of the effective mass m in eq .…”

“…3,11 Therefore, further crystal growth optimization might not be helpful to diminish the effects of alloy disorder in FAPb 1−x Sn x I 3 and FA 0.83 Cs 0.17 Pb 1−x Sn x I 3 perovskites. 21 In other words, the samples have already achieved a high quality with respect to the effects of alloy disorder. If the parameters α, m, and N in eq 9 did not depend on x, ε 0 (x) would solely be proportional to x 2 (1 − x) 2 and would display a maximum at x = 0.5.…”

“… 3 , 11 Therefore, further crystal growth optimization might not be helpful to diminish the effects of alloy disorder in FAPb 1– x Sn x I 3 and FA 0.83 Cs 0.17 Pb 1– x Sn x I 3 perovskites. 21 In other words, the samples have already achieved a high quality with respect to the effects of alloy disorder.…”

“…In the calculations, a concentration of lattice sites N = 4 × 10 21 cm –3 was used, based on the estimate for the lattice constant a ≃ 6.3 × 10 –8 cm. 21 There is uncertainty in the choice of the effective mass m in eq 9 . If electrons and holes appear in the form of excitons, the sum m e + m h should be used.…”