During carrier multiplication (CM), also known as multiexciton generation (MEG), absorption of a single photon produces multiple electron-hole pairs, or excitons. This process can appreciably increase the efficiency of photoconversion, which is especially beneficial in photocatalysis and photovoltaics.
This Account reviews recent progress in understanding the CM process in semiconductor nanocrystals (NCs), motivated by the challenge researchers face to quickly identify candidate nanomaterials with enhanced CM. We present a possible solution to this problem by showing that, using measured biexciton Auger lifetimes and intraband relaxation rates as surrogates for, respectively, CM time constants and non-CM energy-loss rates, we can predict relative changes in CM yields as a function of composition. Indeed, by studying PbS, PbSe, and PbTe NCs of a variety of sizes we determine that the significant difference in CM yields for these compounds comes from the dissimilarities in their non-CM relaxation channels, i.e., the processes that compete with CM. This finding is likely general, as previous observations of a material-independent, “universal” volume-scaling of Auger lifetimes suggest that the timescale of the CM process itself is only weakly affected by NC composition.
We further explore the role of nanostructure shape in the CM process. We observe that a moderate elongation (aspect ratio of 6–7) of PbSe NCs can cause up to an approximately two-fold increase in the multiexciton yield compared to spherical nanoparticles. The increased Auger lifetimes and improved charge transport properties generally associated with elongated nanostructures suggest that lead chalcogenide nanorods are a promising system for testing CM concepts in practical photovoltaics.
Historically, experimental considerations have been an important factor influencing CM studies. To this end, we discuss the role of NC photocharging in CM measurements. Photocharging can distort multiexciton dynamics, leading to erroneous estimations of the CM yield. Here, we show that in addition to distorting time-resolved CM signals, photocharging also creates spectral signatures that mimic CM. This re-emphasizes the importance of a careful analysis of the potential effect of charged species in both optical and photocurrent-based measurements of this process.