Erratum: First-principles study of pyroelectricity in GaN and ZnO (Physical Review B (2016) 93 (081205R) DOI: 10.1103/PhysRevB.93.081205)

There was an error in Fig. 3 of this paper. The experimental temperature-dependent lattice parameters [1] a ( T ) and c ( T ) for ZnO were mistranscribed. This caused a significant error in the curve of Δ u ext . The error was not made when the derivative was plotted in Fig. 2. The results in Fig. 2 are correct, and the corrected version of Fig. 3 is given here. We also take this opportunity to add the predicted Δ u ( T ) for GaN, which was omitted from the original paper (the derivatives are shown in Fig. 1 of the original paper). We also wish to clarify some notations which were only implicit via the text but not included explicitly in the equations. The clarified version of Eq. (5) is (Formula Presented) and (Figure Presented). and the wurtzite version, two lines below in the text, should read pϵ ( T ) = ( 2 e / ω ) Z ∗ c ( Δ u / Δ T ) e . Equation (7) is written for Δ u / u , but, as the text explains, it is really ( Δ u ) ϵ / u . The Grüneisen parameter Δ γq j ( 0 ) used in that equation is identified correctly in the text as the "internal" Grüneisen parameter, defined in Eq. (A14). The second term in Eq. (8) is the internal shift [it should have been written as Δ u int ( T ) - Δ u int ( 0 ) ] , and the last two terms are the external shift [caused by external strains e ( T ) ] . Finally, recent work has clarified and extended some of this paper. A more thorough discussion on the relation between external and internal thermal expansions can be found in Ref. [5]. New measurements of p ( T ) are now available for GaN [6]. The electron-phonon contribution, omitted in this paper, has now been computed [7] and is not negligible for ZnO or GaN. (Figure Presented).