Holographic detection of spectral holes is demonstrated in a crystalline host material with signal-to-noise ratios of up to 104. Hole burning occurs in two Pr3+ sites in the Y2SiO5 lattice, in both cases due to population redistribution between the ground-state quadrupole levels. The signal contains contributions due to a resonant hole and several side holes and antiholes, a phenomenon not previously observed using the holographic technique. The diffracted spectrum is modeled in two ways. In the first case the transmission spectrum is used to determine the population gratings and thus the diffraction efficiency. In the second case the transition probabilities between ground- and excited-state Kramer's doublets are used to model the population gratings. The technique is applied to pseudo-Stark-effect measurements from which the crystallographic sites as determined by x-ray analysis are matched to the spectroscopic data presented here. The time decay of the diffracted signal is used to study nuclear spin-lattice relaxation. It is shown that at 1.6 K temperature-dependent phonon-induced processes make no contribution to this decay. The nonexponential time decay of the population upon radio-frequency irradiation resonant with a ground-state quadrupole splitting is attributed to Pr-Pr cross relaxation

The constructive interference between two Stark-effect-broadened holograms produced by spectral hole burning is discussed. The holograms are burned at the same frequency but at different external electric-field values. The phase difference is selected to be zero so that constructive interference between the waves diffracted by each grating occurs. Experimentally it is found that a dip in the hologram efficiency that is not predicted by previous theory occurs for all reconstruction external electric-field values in the region between the original burn values. This dip is interpreted as being due to the time nonlinearity of the hologram burn process. The dip corresponds to those molecules, oriented in a specific direction with respect to the electric field, for which no Stark shift occurs and that are therefore resonant with the laser during the production of both holograms. The width of the anomalous feature is close to that of the hologram when the hologram is reconstructed at the original burn external electric-field strength. Other molecular orientations may be selected by burning pairs of holographic gratings at other combinations of the frequency and the electric field.