Raman spectroscopy of submicronKNO3films

We have obtained Raman spectra of phase-I, -II, and -III ${\mathrm{KNO}}_3$ films 0.26 μm thick. The primary aim was to establish the physical mechanism for stabilization of the ferroelectric phase III in thin films. In bulk, phase III exists only as a reentrant phase stable from ∼113 °C to 120 °C. However, application of hydrostatic or uniaxial pressure greatly enlarges the temperature range over which phase III is stable. This has led to the conjecture that stress (e.g., from differential thermal expansion of film and substrate) stabilizes phase III in thin films. Our vibrational spectroscopy shows that this is not likely to be the dominant effect. The dominant effect appears to be surface electric fields. This is shown in the shift and splitting of low-energy odd-parity modes from an unresolved broad peak at 120 ${\mathrm{cm}}^{\mathrm{-}1}$ in bulk crystals to resolved TO-LO pairs at 92 and 97 ${\mathrm{cm}}^{\mathrm{-}1}$ and 107 and 113 ${\mathrm{cm}}^{\mathrm{-}1}$. The Raman data show that the ${\mathrm{KNO}}_3$ thin films are more highly ordered than in bulk. This is compatible with spontaneous polarization values, which are 50% or more greater in the films, and with much higher $T_C$ values in the thinner films. The values of $T_C$ as a function of thickness d are compatible with the theory of Tilley and Zeks [Solid State Commun. 49, 823 (1984)] as well as with similar experimental results by Hadni and Thomas [Thin Solid Films 81, 247 (1981); Ferroelectrics 59, 221 (1989)] on triglycine sulfate.