REEXAMINATION OF INFRARED SPECTRA OF Bi NANORODS: L–T TRANSITION OR EXTRINSIC PHASES

Abstract

Bismuth is a fascinating material system owing to its unusual Fermi surface topology, which depends on size and temperature. Theoretical calculations predict that Bi should undergo a semimetal-to-semiconductor transition as at least one of its dimensions becomes < 50 nm. This prediction was experimentally confirmed by infrared (IR) absorption spectra, which is largely underlain by transitions between the L (electron) and T (hole) pockets of the Fermi surface. In this work, however, we report that in our nanosize samples, the observed IR peak positions are practically independent of temperature, which is hard to reconcile with the predicted behavior of the L–T transition. To help elucidate the origin of these IR peaks, we performed a careful analysis of the IR spectra of Bi nanorods, as well as those of bulk Bi, Bi samples prepared under different conditions and Bi2(CO3)O2 using Fourier transform infrared and photoacoustic spectroscopy measurements. We propose that the observed IR peaks in Bi nanorods arise from the oxygen–carbon containing secondary phases formed on the surface of Bi rather than from the Bi itself. We believe that secondary phases must be taken into account on a general basis in modeling the IR spectra of Bi and that the scenario that ascribes these IR peaks solely to the L–T transitions may not be correct. The results reported herein may also impact the research of Bi-based thermoelectric nanostructures and bulk materials.