First-principles reinvestigation of bulk WO3

Abstract

6Using first-principles calculations, we analyze the structural properties of tungsten trioxide WO 3 . Our calculations rely on density functional theory and the use of the B1-WC hybrid functional, which provides very good agreement with experimental data. We show that the hypothetical high-symmetry cubic reference structure combines several ferroelectric and antiferrodistortive (antipolar cation motions, rotations, and tilts of oxygen octahedra) structural instabilities. Although the ferroelectric instability is the largest, the instability related to antipolar W motions combines with those associated to oxygen rotations and tilts to produce the biggest energy reduction, yielding a P 2 1 /c ground state. This nonpolar P 2 1 /c phase is only different from the experimentally reported P c ground state by the absence of a very tiny additional ferroelectric distortion. The calculations performed on a stoichiometric compound so suggest that the low-temperature phase of WO 3 is not intrinsically ferroelectric and that the experimentally observed ferroelectric character might arise from extrinsic defects such as oxygen vacancies. Independently, we also identify never observed R3m and R3c ferroelectric metastable phases with large polarizations and low energies close to the P 2 1 /c ground state, which makes WO 3 a potential antiferroelectric material. The relative stability of various phases is discussed in terms of the anharmonic couplings between different structural distortions, highlighting a very complex interplay.

Bi2W2O9: A potentially antiferroelectric Aurivillius phase

“…By contrast, distortions involving octahedral tilts produce significantly smaller energy lowerings with, in decreasing order, X − 3 (Amam phase, non standard setting of Cmcm), X − 4 (Bbcb phase, non standard setting of Cmma), X − 1 (Bbab phase, non standard setting of Ccca), X + 2 (Acam phase, non standard setting of Cmca) and X + 3 (Abam non standard setting of Cmca). This hierarchy corresponds to what is observed in WO 3 , in which the polar and antipolar distortions lower the total energy much more than the octa-hedral tilts [36]. This is in contrast with other Aurivillius phases like Bi 2 WO 6 , in which octahedral tilts and polar distortions give comparable energy gains (although the tilts are much less unstable than the polar mode).…”

“…Bi 2 W 2 O 9 shows many similarities with WO 3 , which was also recently proposed to be potentially antiferroelectric [54] and it is worth noting that the energy landscape of both compounds share very similar features: the strongest instability is polar and produces, together with the antipolar instability, a more substantial gain in energy than octahedral rotations. However, the combination of antipolar motions and octahedral rotations yield a slightly larger gain in energy than polar motions, yielding a non polar ground state.…”

Charge doping and large lattice expansion in oxygen-deficient heteroepitaxial WO3

“…The tungsten oxide WO 3 holds a special place in the family of complex oxides, since its perovskite ABO 3 crystal structure has an empty A-site. This characteristic determines an open crystalline structure, which is prone to host interstitial species which act as dopants for the otherwise insulating material [1][2][3][4][5]. For these reasons WO 3 finds wide use in electrochromic, optoelectronic and gas sensing applications [6][7][8].…”

Symmetry driven control of optical properties in WO3 films

“…This large strain response can be understood as the result of the strong coupling between electronic and structural instabilities. 13 The graphs in Fig. 3 depict the calculated changes of the octahedral rotations and tilts between +4% and 4% strain.…”

“…11 Upon cooling from high temperature, WO 3 undergoes a series of symmetry lowering phase transitions as cation displacement and oxygen octahedral rotation patterns shift. 12,13 These phases show drastic differences in optical, 14 ferroelectric, 13 and electronic transport 15 properties. For example, the optical bandgap (E g ) of the room temperature monoclinic γ-WO 3 phase (E g ∼ 3.9 eV ) is much larger than the high temperature tetragonal α-WO3 phase (E g ∼ 2.5 eV ).…”