Theory of optical absorption in expanded fluid mercury

A theory is developed of optical absorption in mercury fluid which is correct in the atomic limit and describes densities up to 4-5 gcm-3. The model takes density fluctuations into account explicitly, and shows that the steep, near-exponential absorption edge observed in mercury can be explained quantitatively in terms of absorption by excitonic states of large randomly distributed clusters. This removes the discrepancy between optical-absorption measurements which indicated the band gap closes around 5 gcm-3 and transport and Knight-shift measurements which showed a metal-insulator transition around 8.5 gcm-3. The model predicts, in qualitative agreement with results of recent reflectivity measurements, that the excitonic absorption is separate at densities even up to 1/4 5 gcm-3.