The purely thermal infrared emission spectra of a resonance medium (sodium vapor) are investigated experimentally. It is shown that the emission intensity in the 2–3 μm range at temperatures of 600–1200 K is several orders of magnitude higher than the intensity obtained from the standard theory of resonance radiation transfer. This phenomenon can be conventionally termed an “infrared catastrophe.” The form of the recorded spectra and the absolute intensity of the emission in both the infrared and visible regions of the spectrum are in agreement with the theory developed by Yu. L. Zemtsov and A. M. Starostin, Zh. Éksp. Teor. Fiz. 103 , 345 (1993) [JETP 76 , 186 (1993)], in which the Boltzmann spectral distribution of the population of the resonance level is proportional to exp(−ħω/ T ).
The problem of the laser-induced fluorescence of sodium vapor at a wavelength of 589.6 nm in the case of photoexcitation of a small volume of the medium by a laser beam has been solved numerically. The photon transfer from the zone of excitation to the periphery in the direction perpendicular to the direction of propagation of the laser radiation leads to a significant extension of the volume with a high density of excited atoms. The influence of this factor on the formation of the resonance fluorescence of sodium vapor is discussed.
Experimental line profiles and integrated intensities for backscattered and forward-scattered resonance radiation using white-light excitation are compared with theory for Na densities between
1019 and
1022 atoms/
m3. Absorption measurements are used to verify linewidth and line-wing asymmetry parameters and Na-vapor densities. The scattered spectral intensities are normalized to the incident flux using calibrated diffusers. Measurements of scattered intensities were also made as
N2 was added to the Na. The line profiles and integrated intensities were calculated using a model which assumes complete redistribution of the radiation scattered by an atom and takes into account both radiative and nonradiative transport of the resonance excitation. The theory includes the effects of quenching by
Na2 and by admixed
N2. With the white-light excitation used in these experiments the radiative transport is found to dominate the nonradiative transport except at wavelengths very near line center. The agreement between experimental and theoretical line profiles is good at high sodium densities (\ensuremath{\sim} 5 \ifmmode\times\else\texttimes\fi{}
1021 atoms/
m3) where instrumental resolution is not a serious limitation. The agreement between experimental and theoretical integrated intensities is within experimental error for Na densities of (1-10) \ifmmode\times\else\texttimes\fi{}
1019 and (2-5) \ifmmode\times\else\texttimes\fi{}
1021 atoms/
m3. At the intermediate densities the approximately 20% discrepancy may be caused by the incomplete redistribution of radiation in the scattering process.
Sodium vapor, in the density range
1013 to 5 \ifmmode\times\else\texttimes\fi{}
1014 ${\mathrm{cm}}^{\ensuremath{-}3}$, was excited by a cw dye laser, tuned 20-150 GHz from either the
D1 or
D2 resonance line. We observed a three-peak scattered spectrum, consisting of the Rayleigh component at the laser frequency, and the two fluorescence components (direct and sensitized) at the atomic resonance-line frequencies. Corrections to the Rayleigh signals for anisotropy and polarization effects, and to the fluorescence signals for radiation trapping, were made in order to obtain the ratio of the sum of the total intensities of the two fluorescence components to that of the Rayleigh component. This ratio combined with a measurement of the line-wing absorption coefficient yields the sodium density and the
D-line self-broadening rate coefficients [${k}_{\mathrm{br}}=4.67\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}7}$
cm3${\mathrm{s}}^{\ensuremath{-}1}$ (\ifmmode\pm\else\textpm\fi{}15%) for the
D2 line and ${k}_{\mathrm{br}}=3.07\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}7}$
cm3${\mathrm{s}}^{\ensuremath{-}1}$ (\ifmmode\pm\else\textpm\fi{}15%) for the
D1 line]. Asymmetry in the self-broadened line wings due to fine-structure recoupling was observed. The measured intensity ratio of the
D lines, combined with pulsed measurements of the effective radiative decay rates in the presence of radiation trapping, yields the fine-structure collisional-mixing cross section [$\ensuremath{\sigma}(3{P}_{\frac{3}{2}}\ensuremath{\rightarrow}3{P}_{\frac{1}{2}})=172$
A˚2(\ifmmode\pm\else\textpm\fi{}18%)] at $T\ensuremath{\cong}300\ifmmode^\circ\else\textdegree\fi{}$ C. Our results are compared to other experiments and to theory.
An integral equation is derived for the spectral density of excited atoms using optically dense bounded dispersion media as an example. It is found that the inclusion of the thermal motion of atoms and of the effects due to the existence of a boundary brings about a nonlocal correlation between the concentration of excited particles and the intensity of electromagnetic field in the medium. It is demonstrated that, when the spatial dispersion of permittivity and the boundary effects are disregarded, the obtained result transforms to the well-known Biberman-Holstein equation. The problem on the spectral intensity of radiation of a heated half-space is also investigated.
Abstract The problem of transfer of line radiation in sodium vapor under photoexcitation by laser radiation of a resonance line with wavelength λ = 589.6 nm is numerically solved. The influence of Doppler frequency shifts on photon emission coherent in the atomic reference system is taken into account in the formation of an emission line profile using the model of partial frequency redistribution. In the case of a small optical thickness of a medium τ0 ≤ 1, the Doppler frequency redistribution in the laboratory reference system yields higher intensity values in the emission line profile core in comparison with the model of complete frequency redistribution. For an optically dense medium, thermal motion of atoms leads to an increase of reabsorption in the emission line core, and intensity increases in the line wings as compared to the complete redistribution model. The dependence of the Holstein trapping factor on optical thickness is closer to the analytic dependence than the complete redistribution model. This is explained by the fact that the frequencies of emitted photons in the laboratory reference system fall into the spectral profile core, which increases the effects of resonance radiation trapping in a dense medium.