Efficient carrier multiplication in CsPbI3 perovskite nanocrystals

Abstract

The all-inorganic perovskite nanocrystals are currently in the research spotlight owing to their physical stability and superior optical properties-these features make them interesting for optoelectronic and photovoltaic applications. Here, we report on the observation of highly efficient carrier multiplication in colloidal CsPbI₃ nanocrystals prepared by a hot-injection method. The carrier multiplication process counteracts thermalization of hot carriers and as such provides the potential to increase the conversion efficiency of solar cells. We demonstrate that carrier multiplication commences at the threshold excitation energy near the energy conservation limit of twice the band gap, and has step-like characteristics with an extremely high quantum yield of up to 98%. Using ultrahigh temporal resolution, we show that carrier multiplication induces a longer build-up of the free carrier concentration, thus providing important insights into the physical mechanism responsible for this phenomenon. The evidence is obtained using three independent experimental approaches, and is conclusive.

Fig. 1

Microscopic characterization of the CsPbI₃ nanocrystals. a Annular dark field image of the freshly drop casted sample. b Atomic resolution scanning transmission microscopy image of a nanocrystal showing clearly their mostly cubic structure. c, d Energy dispersive X-ray spectrum of the same sample (c) and the core-loss electron energy loss spectrum (d) where mainly Cs, Pb and I are detected. The Cs:Pb:I ratio is 1:1:3 ± 10%, as determined from the quantification of the energy dispersive X-ray signal for the L-lines of the corresponding elements, which suggests (mostly) the perovskite composition. e Low-loss electron energy loss spectrum of a large, 12 nm nanocrystal, with a band gap energy of 1.77 eV. The dotted line indicates the first derivative of the spectrum around the absorption onset, where its maximum is taken as the NC band gap energy

Fig. 2

Optical characterization. a Absorbance (dotted) and PL (solid) spectra. The colored arrows indicate the photon energies used in the transient absorption experiment. b Time-resolved photoluminescence (PL) measurement for λ_det = 695 nm which was fitted using a bi-exponential function and yielding decay times τ₁ = 3.3 ns and τ₂ = 45 ns. The inset shows the obtained lifetimes for all detection wavelengths. The PL lifetimes with corresponding amplitudes are shown in the inset. The uncertainty is determined by the statistical error from the fitting method. c 2D contour plot of the PL excitation, showing the effect of excitation energy on the PL emission and intensity. A PL maximum is observed around <250 nm (>4.96 eV), which could be the first sign of carrier multiplication. The excitation and emission intensities are corrected for the wavelength dependent components of the setup and spectral sensitivity

Fig. 3

Transient absorption dynamics. a, b Dynamics below (a) and above (b) the carrier multiplication (CM) threshold, i.e., at pump wavelengths of 500 nm (2.48 eV) and 295 nm (4.2 eV), respectively. The dashed lines represent the exponential fit to the data. The appearance of the additional fast component when pumping at 4.2 eV is the fingerprint of CM. The insets shows the initial transient amplitude A and its ratio to the single exciton decay tail A/B, as a function of the absorbed photon fluence, demonstrating the single photon absorption (linear) regime. All dynamics are measured at probe wavelengths around the photo-induced bleach maximum (680 nm) by integrating the signal from 675 to 685 nm. The latter determines the error bars in the y-direction. The error in the x-direction arises from small fluctuations in the pump power. c Linear vs. nonlinear regime, showing the decay through Auger recombination with pumping outside the linear regime (i.e., by multi-photon absorption) and through CM, yields the same dynamics

Fig. 4

Carrier multiplication efficiency and Auger recombination. a ΔA as a function of the absorbed photon fluence. The solid lines represent a linear fit through the data points. b CM efficiency plotted as a function of excitation energy normalized to the band gap energy of the CsPbI₃ nanocrystals. The blue and pink data points correspond to the yield calculated from the A/B ratios deduced from the photo-induced-bleach and -absorption respectively. c Initial rise of the photo-induced bleach signal (normalized) which becomes slower when the nanocrystals are excited with higher photon energies, and carrier multiplication sets in. The error bars are determined by small fluctuations in the pump power (x-direction) and from integrating the dynamics between 4 and 8 ps yielding an upper and lower limit for ΔA (y-direction)