A variation of the usual method of atomic orientation of alkali metal vapors by optical pumping is proposed. This variation consists of employing a single circularly polarized
D1 line as pumping radiation, instead of a mixture of the two
D lines. Under these conditions, one magnetic substate of the ground state is transparent to the pumping radiation. It is shown that for a sufficiently intense light source and relatively weak relaxation, the entire population of the vapor will tend to be pumped into this single state, resulting in nearly one-hundred percent atomic orientation. The pumping process is analyzed in detail for sodium vapor illuminated with $3{P}_{\frac{1}{2}}\ensuremath{-}3{S}_{\frac{1}{2}}$ radiation, and the time dependence of the populations of the eight magnetic sublevels of the ground state is computed. It is shown that in the steady state, the light absorption by the vapor is linear rather than exponential with distance. A simple method of measuring relaxation times based on this observation is suggested. The conditions under which the imprisonment of resonance radiation limits the size of the oriented sample are discussed, and it is suggested that the proposed technique will allow the alignment of larger samples than is possible by the usual method of optical pumping.
A new optical method for studying the spin relaxation of optically aligned rubidium vapor is described. In this method, the pumping radiation, consisting of circularly polarized
D1 resonance radiation, is suddenly shut off and then turned on again after a known time interval. The relaxation which takes place during the interval of darkness causes the vapor to become more opaque to the pumping radiation at a rate which is determined by the relaxation time. A large degree of alignment, as well as a relaxation time of about 80 milliseconds, were observed in a closed-off evacuated cylindrical glass cell completely lined, except for
2001 of the wall area, with a thin film of tetracontane (
C40H82). The variation of relaxation time with buffer gas pressure was studied in this cell and in an unlined glass cell. From the observations with the unlined cell, diffusion coefficients for rubidium in neon and argon of 0.31
cm2/sec and 0.24
cm2/sec, respectively, can be deduced. Observed cross sections for disorientation collisions between aligned ground-state rubidium atoms and neon, argon, krypton, and xenon atoms are 5.2\ifmmode\times\else\texttimes\fi{}${10}^{\ensuremath{-}23}$
cm2, 3.7\ifmmode\times\else\texttimes\fi{}${10}^{\ensuremath{-}22}$
cm2, 5.9\ifmmode\times\else\texttimes\fi{}${10}^{\ensuremath{-}21}$
cm2, and 1.3\ifmmode\times\else\texttimes\fi{}${10}^{\ensuremath{-}20}$
cm2, respectively. In the evacuated tetracontane-lined cell, the relaxation time decreased by 30% for a tenfold increase in rubidium vapor pressure. An explanation for this relatively weak dependence is suggested. The longest observed relaxation time was approximately 0.4 seconds in a tetracontanelined cell filled with neon to a pressure of 3 cm Hg.
We have developed a general method to perform optical pumping by a pulsed laser with the aid of an optical cavity (cavity-assisted optical pumping). Optical pumping is achieved by repetitive interaction of a single laser pulse with a target material in the cavity. This method is demonstrated for manganese ions,
Mn+, stored in a linear radio-frequency ion trap; about
108 ions are spin polarized by a 5-ns laser pulse via ${^{7}P}_{J}\ensuremath{\leftarrow}{^{7}S}_{3}$ (
J=3 or 4) transition in the ultraviolet region. The linewidth of the pulsed light source is broad enough to transfer the populations of lower hyperfine levels to the highest one
(F=11/2) in the
7S3 ground state; the nuclear spin is polarized as well.
III-8.Comparison of optical pumping signals for two spherical cells with different buffer gas pressures as a function of density times diameter of cell for equal D2/D1.60 IV-1.Energy level diagram showing hyperfine structure of Rb Di line.Units are 10*3 cm-1.67 IV-2.Equivalent circuit for rate equations involving ex change polarization.73 IV-3.Diagram of optical system for optical pumping of Rb. 83 IV-4.Block diagram of Rb optical pumping signal detection system.85 IV-5.Block diagram of Rb light detection system.90 IV-6.Typical photograph from which relaxation time measure ments are made..Time increases from left to right.Scale is 1 millisecond per division.92 IV-7.Lines determined from measurements on time dependence of Rb87 signals at 50°C for the two different light in tensities.Scatter of points is too small to be notice able on graph.93 A-1.Spectral profile of D1 radiation from a Varian spectral lamp containing natural Rb.Ill A-2.Spectral profile of Di radiation from a Varian spectral lamp containing Rb87.112 A-3-Spectral profile of Di radiation from a Varian spectral lamp containing Rb87 after filtering by an Rb85 absorption cell.113 vi CHAPTER I2 atomic ground states and the parameters of various relaxation processes of gaseous systems.These investigations led to the development of mag netometers of unique sensitivity.More recently, very precise frequency standards have been developed using the methods of optical pumping and the knowledge gained through their application.The present work was undertaken in the context of extending our knowledge of the dynamics of optical pumping methods and of exploiting these methods for the deter mination of an interesting atomic exchange cross section.Optical pumping may be considered as a specialized technique used in conjunction with magnetic resonance phenomena.All magnetic reso nance techniques depend upon obtaining an oriented (polarized) macro scopic sample of matter, that is, matter with a net magnetic moment.In the atomic beam magnetic resonance method, this is accomplished by sort ing out, spatially, atoms whose magnetic moments have a particular ori entation with respect to a magnetic field.In nuclear or electron para magnetic resonance work in solids, liquids, or gases, the small thermal equilibrium polarization that matter possesses in a magnetic field is exploited.Optical pumping enhances this equilibrium polarization through selective optical absorption, by a gaseous atomic system, of polarized resonance radiation.Optical pumping offers the great advan tage of experimental simplicity, as compared to other magnetic resonance techniques, but suffers from severe limitations in application.Since the original suggestion by Brossel and Nastier^ and subse quent development by many others, optical pumping, in conjunction with the transmission monitoring technique of signal detection, has became a standard technique in the study of alkali metal ground states.By far the greatest amount of work in optical pumping has been concerned with alkali metal vapors.Bell, Bloom/ and others^ have considered the process of the optical pumping of the alkalis with attention to the spec tral composition of the pumping light, that is, the resonance radiation incident on the alkali vapor.Colegrove and Franken? have recently dis cussed the complication in the analyses of optical pumping signals for Q helium when the helium absorption cell is optically thick.Hobart has considered this same complication for sodium.Prediction of signals must take into account the fact that the pumping light is absorbed ap proximately exponentially in the optical pumping absorption cell.This means that the absorption coefficient, which is proportional to the al kali density, times the length of the cell is a significant parameter in determining the strength of the signal.Similar calculations of the dependence of optical pumping signals on alkali density and spectral com position of the pumping light will be presented in this work.In order to test the signal predictions> a high resolution Fabry-Perot spectrometer was constructed.This instrument can be used for two purposes : the investigation of the intensity distribution of the pumping light, and, more importantly, determinations of the alkali vapor density in the absorption cell.The desired data are curves of optical pumping signals, for a typical sodium optical pumping experiment, as a function of density.It is often necessary to know these densities in order to is limited to the accuracy to which the density is known.Since the Fabry-Perot spectrometer used to measure sodium vapor densities in conjunction with the signal strength versus density meas urements can be applied to different systems, it was decided to make a determination of the rubidium 85-rubidium 87 spin exchange cross sec tion.The reasons for this choice are discussed in detail in Chapt.IV.In brief, work on rubidium exchange offered some experimental sim plifications and rubidium has previously been studied more thoroughly than other systems.
We report on the realization of a magneto-optical trap (MOT) for europium atoms using a narrow-line cooling transition with a natural linewidth of 97 kHz. Europium is a strongly dipolar species with interesting properties, such as the presence of hyperfine structure and the absence of electronic orbital angular momentum. Our starting point of the MOT is continuous capturing and cooling of optically pumped metastable europium atoms. We have employed simultaneous MOT for the metastable and ground-state atoms. The trapped metastable atoms are successively pumped back to the ground state and then continuously loaded to the narrow-line MOT, where up to $4.7\ifmmode\times\else\texttimes\fi{}{10}^{7}$ atoms are captured. A spin-polarized sample at a temperature of $6\phantom{\rule{0.16em}{0ex}}\ensuremath{\mu}\mathrm{K}$ and with a peak number density of $2.2\ifmmode\times\else\texttimes\fi{}{10}^{11}\phantom{\rule{0.16em}{0ex}}{\mathrm{cm}}^{\ensuremath{-}3}$ is obtained through the compression process, resulting in a phase space density of $3\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}5}$.