Reply to "Comment on 'Semimetal-to-semiconductor transition in bismuth thin films'"

The suppression of intrinsic carrier generation does not in itself constitute proof that a semimetal-to-semiconductor transition has occurred. However, we present a realistic statistical analysis that demonstrates that the Bi film data do in fact imply the opening of a 56-meV energy gap in the thinnest sample (200) studied. The standard vanishing-wave-function model accounts for the experimental observations, whereas the alternative vanishing-gradient model yields results that are inherently inconsistent with the data.