Generalized optical Bloch equations are derived and applied to free-induction decay. The resultant expression for the free decay rate is shown to give a reasonably good fit to recent experimental data of DeVoe and Brewer that cannot be explained by the usual Bloch equations with the conventional 1/T1 and 1/T2 relaxation parameters.
Recently, DeVoe and Brewer [Phys. Rev. Lett. 50, 1269 (1983)] observed a striking deviation from a prediction of the optical Bloch equations by monitoring optical free-induction decay in the impurity-ion crystal Pr3+:LaF3. At low optical fields, the Pr3+ optical dephasing time T2 arises from magnetic fluctuations of the local environment, but at elevated optical fields, T2 is no longer a constant, as assumed in the Bloch equations, because the magnetic line-broadening process is quenched. Several theories have been developed to explain the phenomenon. In this paper we present a simple theory of relaxation in solids which allows for comparison with earlier work. A ``strong-redistribution'' model is proposed where the optically excited impurity ions experience frequency shifts \ensuremath{\epsilon} induced by a thermal bath. Frequency jumps occur at an average rate \ensuremath{\Gamma} and with an rms value of ${\ensuremath{\epsilon}}_{0}$/ \ensuremath{\surd}2 , where ${\ensuremath{\epsilon}}_{0}$ is the thermal width associated with the frequency shifts. Modified Bloch equations (MBE) follow that are solved explicitly for two limiting cases, ${\ensuremath{\epsilon}}_{0}$\ensuremath{\ll}\ensuremath{\Gamma} and ${\ensuremath{\epsilon}}_{0}$\ensuremath{\gg}\ensuremath{\Gamma}, and qualitatively for the more general case where the ratio \ensuremath{\Gamma}/${\ensuremath{\epsilon}}_{0}$ is arbitrary. Since many of the earlier theories are equivalent to the strong-redistribution model in the limit ${\ensuremath{\epsilon}}_{0}$\ensuremath{\ll}\ensuremath{\Gamma}, we can assess the validity of the approximations made by these authors. Finally, we compare our MBE to the analogous transport equations that describe the effects of collisions in atomic vapors. We conclude that the problem addressed here is a general one, common to solids and gases alike, and is not restricted to impurity-ion crystals but will occur whenever frequency fluctuations are important.
In a recent experiment [ DevoeR. G.BrewerR. G., Phys. Rev. Lett.50, 1269 ( 1983)], it was found that the optical Bloch equations could not satisfactorily explain the signal that was observed for free-induction decay in the impurity ion cyrstal Pr3+:LaF3. Several theories have been proposed to explain this failure of the Bloch equations. In this paper, the general validity conditions for the optical Bloch equations are examined within the limits of a Markovian relaxation model. The specific problem to be considered is the interaction of an optical field with two-level atoms. The atoms undergo relaxation as a result of coupling to a perturber bath that, itself, is negligibly affected by the relaxation process. First, a simple decay-parameter model is assumed for the relaxation of atomic density-matrix elements. Such a model is found to lead to a set of generalized Bloch equations of which the conventional Bloch equations form a subset. Subsequently, more-realistic models for relaxation in both vapors and solids are considered within the limits of the impact approximation (i.e., the duration of a fluctuation can be viewed as instantaneous with respect to all relevant time scales in the problem). It is found that, even in the impact (Markovian) approximation, the generalized Bloch equations cannot be expected to provide an adequate description of relaxation, owing to effects in which a fluctuation-induced change in the atomic transition frequency persists between fluctuations. In vapors this frequency shift is produced by velocity-changing collisions that change the atomic resonance frequency (as seen in the laboratory frame), whereas in solids it is produced by local-field fluctuations. The limiting conditions under which one can expect both the generalized and the conventional Bloch equations to retain their validity are explored.
We derive the effects of random telegraph noise on the equations of optical resonance. The random-telegraph-signal model allows some results to be found exactly, independent of the strength of the noise or the strength of the field that induces the resonant transitions. Our optical resonance equations depart from the well-known empirical optical Bloch equations in several ways. We obtain excellent agreement at high, low, and intermediate Rabi frequencies with the data of DeVoe and Brewer [Phys. Rev. Lett. 50, 1269 (1983)]. This is the first such agreement that has been reported with a value of the noise coherence time smaller than about 20 \ensuremath{\mu}sec.
In a recent experiment of DEVOE and BREWER [1], it was concluded that the optical Bloch equations are incapable of describing the saturation phenomena observed. Optical free-induction decay (FID) measurements of the impurity ion crystal Pr3+:LaF3 were conducted,where the Pr3+ ions are cohertntly prepared by a laser field under steady-state conditions,and then freely precess when the driving field is removed. At low optical fields, the observed Pr3+ optical line width is dominated by magnetic fluctuations.arising from pairs of fluorine nuclear flip-flops. At high optical fields, this nuclear broadening mechanism is quenched.and the Bloch equations are seriously violated. On physical grounds, this failure is due to a time-averaging of the magnetic interaction as the optical nutation frequency increases [2]. The phenomenological dipole dephasing time T2 of the Bloch equations is therefore not a true constant,but lengthens with increasing field strength.
In a previous paper [ J. Opt. Soc. Am. B3, 564 ( 1986)], the validity conditions for the optical Bloch equations were reviewed. It was shown that, even within the limits of an impact or Markovian approximation, the optical Bloch equations fail to account properly for fluctuation-induced changes in atomic transition frequencies. Such changes are properly incorporated in a quantum-mechanical transport equation (QMTE) in which the fluctuation-induced frequency shifts are totally characterized by kernels W(∊′ → ∊) that give the probability density per unit time for a fluctuation to change the frequency shift from ∊′ to ∊. The QMTE describes the interaction of atoms with both an external radiation field and the perturber bath producing the fluctuations. A general method for solving the QMTE as a perturbation series in the external field is presented. Specific calculations are carried out for strong-redistribution, difference [W(∊′ → ∊) is a function of (∊–∊′) only], and Brownian motion kernels. It is shown that, although the kernels possess fundamental differences, they can yield similar results in certain limits. As an example, a perturbation calculation is performed for the free-induction decay (FID) of atoms prepared by a cw laser field and then allowed to radiate when the field is suddenly removed. Radical departures from the predictions of the conventional Bloch equations are found in certain limits, including a first-order contribution to FID in vapors and a nonexponential FID decay for atoms in vapors or solids. The implications of these results to a consistent interpretation of a recent experiment [ Phys. Rev. Lett.50, 1269 ( 1983)] on FID in the impurity ion crystal Pr3+:LaF3 are explored.
Recent studies by DeVoe and Brewer [Phys. Rev. Lett. 50, 1269 (1983)] have shown that the conventional optical Bloch equations markedly fail to describe the optical saturation behavior of the D1 line in the solid Pr3+:LaF3. In this paper we extend these studies to another solid, ruby, using free-induction-decay observations obtained by pulse excitation of the R1 line at 693.4 nm with an ultranarrow-linewidth dye laser. Comparison of the results with Gauss-Markov and random-telegraph-dephasing theories shows approximate agreement for a fluctuation correlation time ${\ensuremath{\tau}}_{c}$=T2, the dephasing time. This result is remarkably similar to that obtained for Pr3+:LaF3. However, for theoretical and experimental reasons, we conclude that the theories do not consistently explain the current as well as other data. A qualitative discussion of another dephasing model is given.