Faraday Rotation of the CaO<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"><mml:mi>F</mml:mi></mml:math>Band and 3557 Å Zero-Phonon Peak

Abstract

The absorption band centered at 3.65 eV (3350 \AA{}) seen in both neutron-irradiated and Ca-excess CaO crystals has been assigned to the $F$ center. A sharp peak is seen adjacent to this band at 3557 \AA{}, and has long been conjectured to be the zero-phonon line of the $F$ center. Evidence is given here that supports this identification. Faraday-rotation experiments have been carried out on both the $F$ band and the 3557 \AA{} line, including studies using the method of ESR-sensitive Faraday rotation, which directly identifies rotation peaks due to paramagnetic centers. Such studies were earlier reported only on the 3.65-eV band, with low spectral resolution; in the new data, the spectral rotation of the sharp peak is revealed, and a more accurate pattern for the band is given. The magneto-optic behavior of the sharp peak is found to be similar to that of the band; both are paramagnetic and react in the same way to saturation of the $F$-center ESR line. A model featuring large Jahn-Teller distortion $\ensuremath{\epsilon}$ of the $F$-center $p$ state predicts a dispersionlike rotation for the zero-phonon peak, with amplitude of order $(\frac{\ensuremath{\lambda}}{\ensuremath{\epsilon}})tanh(\frac{\ensuremath{\mu}H}{\mathrm{kT}})$ times the amplitude of the dispersion associated with the absorption peak. The measured pattern has this shape, and agrees in sign and amplitude with the model if we take $\ensuremath{\lambda}=\ensuremath{-}7\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}3}$ eV, from the second moment of the circular dichroism of the $F$ band; and $\ensuremath{\epsilon}\ensuremath{\simeq}0.12$ eV \ensuremath{\simeq}1000 ${\mathrm{cm}}^{\ensuremath{-}1}$, comparable with the distance between the zero-phonon and band peaks (\ensuremath{\simeq}0.16 eV). Though our magneto-optical pattern is thus explainable in terms of a quenched-orbital-momentum model, certain other considerations suggest that quenching is small or marginal, the Jahn-Teller well depth $\frac{1}{4}\ensuremath{\epsilon}\ensuremath{\sim}250$ ${\mathrm{cm}}^{\ensuremath{-}1}$ being of the order of the mean vibrational quantum. Presumably the Faraday rotation of the zero-phonon peak is a mixture of dispersionlike and dispersion-derivativelike terms of comparable magnitudes, but the second is suppressed under our experimental conditions.