This action might not be possible to undo. Are you sure you want to continue?
CohenTannoudji
Vol. 6, No. 11/November 1989/J. Opt. Soc. Am. B
2023
Laser cooling below the Doppler limit by polarization gradients: simple theoretical models
J. Dalibard and C. CohenTannoudji Collage de France et Laboratoire de Spectroscopie Hertzienne de 1'EcoleNormale Sup6rieure [Laboratoire
associ6 au Centre National de la Recherche Scientifique (LA18) et i l'Universit6 Paris VI], 24, rue Lhomond,
F75231Paris Cedex 05, France
Received April 3, 1989; accepted June 29, 1989
We present two coolingmechanisms that lead to temperatures wellbelow the Doppler limit. These mechanisms are based on laser polarization gradients and work at low laser power when the opticalpumping time between different groundstate sublevels becomes long. There is then a large time lag between the internal atomic response and the
atomic motion, which leads to a large cooling force. In the simple case of onedimensional molasses, we identify two
types of polarization gradient that occur when the two counterpropagating waves have either orthogonal linear polarizations or orthogonal circular polarizations. In the first case, the light shifts of the groundstate Zeeman sublevels are spatially modulated, and optical pumping among them leads to dipole forces and to a Sisyphus effect analogous to the one that occurs in stimulated molasses. In the second case (+ar configuration), the cooling mechanism is radically different. Even at very low velocity, atomic motion produces a population difference among groundstate sublevels, which gives rise to unbalanced radiation pressures. From semiclassical optical Bloch
equations, we derive for the two cases quantitative expressions for friction coefficients and velocity capture ranges. The friction coefficients are shown in both cases to be independent of the laser power, which produces an
equilibrium temperature proportional to the laser power. The lowest achievable temperatures then approach the onephoton recoil energy. We briefly outline a full quantum treatment of such a limit.
1.
INTRODUCTION kBTR
=
h 2 k'
The physical mechanism that underlies the first proposals for laser cooling of free atoms' or trapped ions2 is the Doppler effect. Consider, for example, a free atom moving in a
2M'
(1.3)
weak standing wave, slightly detuned to the red. Because of the Doppler effect, the counterpropagating wave gets closer to resonance and exerts a stronger radiation pressure on the atom than the copropagating wave. It follows that the
atomic velocity is damped, as if the atom were moving in a
where M is the atomic mass, can be observed on lasercooled sodium atoms at lowlaser powers. Such an important result was confirmed soon after by other experiments on sodium8
and cesium. 9
A possible explanation for these low temperatures based
on new cooling mechanisms resulting from polarization gra
viscous medium (optical molasses). The velocity capture range Av of such a process is obviously determined by the natural width of the atomic excited state kA  r, (1.1)
dients was presented independently by two groups at the last International Conference on Atomic Physics in Paris. 910 We summarize below the broad outlines of the argument": (i) The friction force experienced by an atom moving in a laser wave is due to the fact that the atomic internal state does not follow adiabatically the variations of the laser field resulting from atomic motion.1 2" 3 Such effects are characterized by a nonadiabaticity parameter (,defined as the ratio between the distance VTcovered by the atom with a velocity v during its internal relaxation time r and the laser wavelength X= 1/k, which in a standing wave is the characteristic length for the spatial variations of the laser field
E=
VT

where k is the wave number of the laser wave. On the other
hand, by studying the competition between laser coolingand diffusion heating introduced by the random nature of spontaneous emission, one finds that for twolevel atoms the
lowest temperature od is given by3 ' 4
TD that can be achieved by such a meth
kBTD =
r
(1.2)
TD is called the Doppler limit. The first experimental dem
=
kvi.
(1.4)
onstrations of optical molasses seemed to agree with such a
limit. 5 6 (ii) For a twolevel atom, there is a single internal time,
1
TrR =
,
In 1988, it appeared that such a limit could be overcome. More precise measurements by the National Institute of Standards and Technology Washington group7 showed that temperatures much lower than TD, and even approaching
the recoil limit TR given by
07403224/89/11202323$02.00
which is the radiative lifetime of the excited state
r
(1.5)
so that
© 1989 Optical Society of America
2024
J. Opt. Soc. Am. B/Vol. 6, No. 11/November 1989 e = kvrR = kv
J. Dalibard and C. CohenTannoudji
(1.6)
But for atoms, such as alkali atoms, that have several Zeeman sublevels gr g', ... in the ground state g, there is another internal time, which is the opticalpumping time rp, characterizing the mean time that it takes for an atom to be transferred by a fluorescence cycle from one sublevel gmto
another gm'. We can write
__1
calculation of the new friction force and of the equilibrium temperature. Our treatment will be limited here to a Jg to Je = Jg + 1transition, neglecting all other possible hyperfine levels of the optical transition.
We first introduce, for a 1D molasses, two types of polar
(1.7)
where r is the mean scattering rate of incident photons and also can be considered the width of the ground state. It followsthat for multilevel atoms we must introduce a second nonadiabaticity parameter e' = kvrp = kv
(1.8)
At low laser power, i.e., when the Rabi frequency Qis small compared with r, we have Tp >> TR and consequently r << r: u << r  r, << r  E >> E.
(1.9)
friction forces can be experienced by very slow atoms.' 4
It followsthat nonadiabatic effects can appear at velocities (kv ') much smaller than those required by the usual Dopplercooling scheme (kv  r). This explains why large
(iii) The last point concerns the importance of polarization gradients. Long pumping times can give rise to large friction forces only if the internal atomic state in g strongly depends on the position of the atom in the laser wave, so that
when the atom is moving there are large changes in its inter
nal state and, consequently, large nonadiabatic effects. By internal atomic state in g, we mean actually the anisotropy in g (usually described in terms of orientation or alignment) that results from the existence of large population differences among the Zeeman sublevels of g or from coherences among these sublevels. Polarization gradients are essential if there are to be important spatial variations of the groundstate anisotropy. For example, if the polarization changes from a+ to or, the equilibrium internal state in g changes
from a situation in which the atom is pumped in g, with m =
ization gradient (see Section 2). In the first case, which occurs with two counterpropagating waves with opposite (a+ and r) circular polarizations, the polarization vector rotates when one moves along the standing wave, but it keeps the same ellipticity. In the second case, which occurs, for example, with two counterpropagating waves with orthogonal linear polarizations, the ellipticity of the laser polarization varies in space, but the principal axis of polarization remain fixed. The basic difference between these two situations is that the second configuration can give rise to dipole or gradient forces but the first one cannot. Section 3 is devoted to a physical discussion of the cooling mechanisms associated with these two types of polarization gradient; they are shown to be quite different. In the configuration with orthogonal linear polarizations, hereafter denoted as the lin I lin configuration, the light shifts of the various Zeeman sublevels of g oscillate in space, and optical pumping among these sublevels provides a cooling mechanism analogous to the Sisyphus effect occurring in highintensity stimulated molasses'6 7 : The atom is always climbing potential hills. In the +o configuration, the combined effect of the rotation of the polarization and of optical pumping and light shifts produces a highly sensitive motioninduced population difference among the Zeeman sublevels of g (defined with respect to the axis of the standing wave) and, consequently, a large imbalance between the radiation pressures of the two counterpropagating waves. In Sections 4 and 5 some quantitative results for 1D molasses and simple atomic transitions are presented. In
Section 4 the case of a transition Jg = 1/2 Je = 3/2 is
considered for an atom moving in the lin I lin configuration, whereas in Section 5 the case of a Jg = 1 Je = 2 transition is considered for an atom moving in the o+o configuration. In Sections 4 and 5, atomic motion is treated semiclassically: the spatial extent of the atomic wavepacket is neglected and the force at a given point in the laser wave is calculated.
Since the new cooling mechanisms work at low power, the
Jg to a situation in which it is pumped in gm'with m' = Jg; if the polarization e is linear and rotates, the atomic alignment in g is parallel to e and rotates with E. By contrast, in the
lowpower regime considered here [see expression (1.9)], a
gradient of light intensity without gradient of polarization would produce only a slight change of the total population in g (which remains close to 1) without any change of the
anisotropy in g.' 5
Finally, note that the laser field does not produce only optical pumping between the Zeeman sublevels of g; it also induces light shifts Am' that can vary from one sublevel to the other. Another consequence of polarization gradients is
that the various Zeeman sublevels in g have not only a popu
calculations are limited to the perturbative regime (Q << F), where it is possible to derive from optical Bloch equations a subset of equations that involve only the populations and Zeeman coherences in the atomic ground state g. In both configurations, analytical or numerical solutions of Bloch equations are derived that are then used to analyze the velocity dependence of the mean radiative force. Quantitative results are derived for the friction coefficient, the velocity capture range, and the equilibrium temperature, which is
shown to be proportional to the laser power Q2.
When the laser power is low enough, the equilibrium temperature approaches the recoil limit TR. It is then clear that
the semiclassical treatment breaks down, since the de Broglie wavelength of an atom with T = TR is equal to the laser wavelength. At the end of Section 5, a full quantum treatment is presented of the cooling process in the o+o configuration for a simplified atomiclevel scheme. Such a treatment is similar to the one used in the analysis of other cooling schemes allowing temperatures of the order of or below TR to be reached.' 8 " 9 We show that the velocity
lation but also a lightshifted energy and a wave function
that can vary in space.
The purpose of this paper is to analyze in detail the physical mechanisms of these new cooling schemes by polarization
gradients and to present a few simple theoretical models for onedimensional (1D) molasses, permitting a quantitative
that the two waves have equal amplitudes. No.  (2. C. We consider first the simple case in which e = e+ =I (ex + iy). From Eq. 1(b)]. 1(a) and 1(b) differ radically with regard to dipole forces. for a +a. (2. The total electric field E(z. E)/ parallel to the two bisectrices of e. since the ClebschGordan coefficients of the various transitions g . according to expression (3. in addition.em' are not the same.7) we see that the polarization is linear along e = (e.5b) The two laser configurations of Figs. The lin lin ConfigurationGradient of Ellipticity We suppose now that the two counterpropagating waves have orthogonal linear polarizations e= v (2. we get from Eqs. the case when both waves have the same polariza +(z) = where t ( 60)ex  ( 0 + O)Ey.. a. [Fig. and its extremity forms a helix with a pitch X [Fig. For example. we can always take Go and 6 0' real. By pure we mean that the polarization rotates but keeps the same ellipticity. and ey.1 2 sublevel has the largest shift for a a. we still have the superposition of two fields in quadrature.&o)/ (60' + 0) for all z. the laser polarization is always linear. we consider the simple case of a Jg = 1/2 Je = 3/2 transition for which there are two Zeeman sublevels. the total electric field is. circular (a) in z = /8. TWO TYPES OF POLARIZATION GRADIENT IN A ONEDIMENSIONAL MOLASSES In this section. 6.2). and the laser intensity is the same for all z. The two waves have opposite circular polarizations. their polarizations. is linear along el in z = X/2. for all configurations other than the o+aone. with amplitudes (' . 1(c) and 1(d) . Suppose that we have in z an atom with several Zeeman sublevels in the ground state g. A. 1(a) and 1(b). where the positivefrequency (2. y coincides with e When one moves along Oz. Let 60 and 6o' be the amplitudes of the two waves and e and e' be polarized along y. Opt.cos kz and 62 sin kz.1) that the two lightshifted energies oscillate with z for the lin I lin configuration [Fig.3a) 'e'= e_ (= (if). Dalibard and C. the periodicity along z is determined by the laser wavelength = 2r/k. The total electric field in z is the superposition of two fields in quadrature. g_1/2 and g+11 2 . On the other hand. circular (+) in z = 3X/8.1) If we suppose. 1(c)]. linear along 2 = ( y)l J.5a) (2. ey rotates.3b) If the two amplitudes 60 and ' are not equal. One can showthat the af.4) tion). When one moves along Oz.6) &+(z) = 6oA(cos kz  i sin kz X) (2. (2. however. B.. It is clear that the ellipticity changes now when one moves along Oz. and so on. Connection with Dipole Forces and Redistribution ex = excos kzey ey = ex sin kz + sin kz.polarization. we get i. t) in z at time t can be written as E(z. Soc.in z = X/4.J. B 2025 distribution curves exhibit a very narrow structure around v = 0. (o' . Inserting Eqs. linearly shifted energies are equal and do not vary with z [Fig.2) 6 (z) = 6 0 ee'z + 0 o efeikz. 1(d)]: the g2 sublevel has the largest shift for a a+ polarization the g. Previous analysis shows that. one can easily show (see Subsection 3. t) = +G(z)exp(iWLt) + c.A. with amplitudes 6J. y Cos kz. As expected.configu Oz. For the +o.. The a+a. there are gradients of ellipticity when one moves along Oz (excluding.3) into Eq. (2. The striking difference between the z dependences of the lightshifted energies represented in Figs.e. For z = 0. (2.6b) C'= ey.7) component 6+(z) is given by (2. We conclude that the polarization of the total electric field is elliptical and keeps the same ellipticity. of course. we have a pure rotation of polarization along Oz.have the same amplitude = o'. in agreement with the semiclassical predictions. 11/November 1989/J. whereas both sublevels are equally shifted for a linear polarization. (2. and e by a rotation of angle s = kz around It is easy to see that the z dependence of the light shifts of the two groundstate sublevels is quite different for the two laser configurations of Figs.6a) (2. In the simple case in which the two counterpropagating waves a+and a. the axes of ration.ConfigurationPure Polarization Rotation of The total electric field is the superposition of two fields in quadrature. It follows that the two light the ellipse just rotate around Oz by an angle ( = kz. with a width of a few recoil velocities. 2. and polarized along two fixed orthogonal vectors (e. (2. Such a result generally holds. in g and four Zeeman sublevels in e.configuration. .4) below. By a convenient choice of the origin on the Oz axis. 1(a)].configuration is the only one that gives such a result.c.for the wave propagating toward z < 0 and u+for the other wave. Am. (2.60 )/F and (60' + 60)/ j and polarized along two orthogonal directions ex and y deduced from e. now they are polarized along two fixed but nonorthogonal directions. we consider two laser plane waves with the same frequency 0L that propagate along opposite directions on the Oz axis and we study how the polarization vector of the total electric field varies when one moves along Oz..2) and (2. CohenTannoudji Vol. For 60 = 6 o' the nature of the polarization of the total field changes along Oz. and since the nature of the polarization changes with z. + ey)/ V'2 in z = 0.
and one can show that. they become large only at high intensity. 1. B/Vol. We begin (Subsection 3. for z = X/8 a.. 1(d).polarization. here we consider an atom that has several sublevels in the ground state. whereas such forces do not exist in the configuration of Fig. redistribution is not completely forbidden but is limited to a finite number of processes.configuration: two counterpropagating waves. CohenTannoudji (a) Cy+ h.a+ and a. by stimulated emission of a a. 11/November 1989 J. Soc. and we show how optical pumping between the two oscillating levels of Fig. such as for a Jg = 1 . Am.photon. we show in Subsection 3. 1(d) can give rise to a new a. dipole forces in terms of gradients of dressedstate energies. We thus have in this case a limited redistribution. the atom can no longer make a stimulated emission in the awave.2 that for a moving atom. 1(b). the lightshifted energies in the ground state do not vary with z (see Subsection 3. For more complex situations.. The has an ellipticity that varies in space: for z = 0 linear polarization along el = (eY+ e)/12. the combination of long pumping times and dipole forces can produce a highly efficient new cooling Fig. The lightshift splitting between the two oscillating levels of Fig. in general. let us note that. The two types of polarization gradient in a 1D molasses and the corresponding lightshifted groundstate sublevels for a Jg 1/2 Je = 3/2 atomic transition. once in e+2. since no a.transition can be excited by both linear cooling mechanism analogous to the Sisyphus effect occur16 7 ring in stimulated molasses. Furthermore. 1(b).16 Another equivalent interpretation can be given in terms of redistribution of photons between the two counterpropagating waves. (a) a+a..transition starts from e+2.. 1(c). each a+ or a.photon. We use here the interpretation of polarizations e.. from where it can be reexcited to e+2 by absorption of a o+photon. 1(a).A) by analyzing the lin I lin configuration of Fig. resulting polarization now for z = X/4 linear polarizaa+a configuration: The The lightshifted energies means that there are dipole or gradient forces in the configuration of Fig. 1(d) can be large compared with the transitions starting from these levels and that could be used for the stimulated emission of a a.em+l or em1. No.ey)/v2. and ey. we consider a multilevel atom moving in a laser configuration exhibiting a polarization gradient.X~ I_ 'C4Zi   Fy t ~ F2 Y __ 6 C 1 C OX 0 lI A18 y I I I > X14 3A/8 A/2 Z y (C)  _ ' e3/ 2 el/2 e1/2 e3 /2 Cd) I / / 9i 91/2 I _K 2 K 2 = Fig. "1 Such an effect cannot exist . PHYSICAL ANALYSIS OF TWO NEW COOLING MECHANISMS In this section. even with weak dipole forces. The situation is completely different for the configuration of Fig. create a linear polarization that rotates in space. 1(b).' 6 Actually.' 2 " 0 It is obvious that conservation of angu lar momentum prevents such a redistribution from occurring in the configuration of Fig.and an infinite number of redistribution processes between the two counterpropagating waves can take place via the same transition g . (b) + %. tion along 62 = ( . for z = 3X/8 a+ circular polarization . Then. that the atom is initially in g1. the atom is put into e+1/2 or e+3/2. However. for example. 6. Suppose. 3. when the atom absorbs a photon from one wave and transfers it via stimulated emission into the opposite wave. and there are no a This is why the lightshifted energies vary with z in Fig. Then.A. Opt. (b) lin lin configuration: The two counterpropagating waves have orthogonal linear polarizations. one would expect dipole forces to be inefficient in the weakintensity limit considered in this paper since. 5 below).2026 J. at first sight.photon. Finally. (d) Lightshifted groundstate sublevels for the lin oscillate in space with a period X/2.1). Dalibard and C.J. When it absorbs a mechanism. as in Fig.polarized. 1(a).2 ' After it absorbs ao+ photon. (c) Lightshifted groundstate sublevels for the lin configuration: lightshifted energies do not vary with z. = 2 transition (see width I' of these groundstate sublevels. it jumps to eo..B. it falls to g+. when the splitting among dressed states is large compared with the natural width r.
as usual in Dopplercooling experiments. are indicated in Fig. i. 3. respectively. steadystate population of this sublevel. that the effect discussed in this paper appears at much lower intensities than in Ref. Soc. and reach the top of this hill before being optically pumped to the other sublevel. X/4.1) between the laser frequency WL and the atomic frequency WA is negative so that both light shifts are negative. in Ref. From there. 2. The square of these coefficients give the transition probabilities of the corresponding transitions. Atomic level scheme and ClebshGordan Jg = 1/2 Je = 3/2 transition. A more quantitative analysis. 16 since it involves two groundstate sublevels with spatially modulated light shifts A' much smaller than r.. Sisyphus Effect for a Moving Atom We suppose now that the atom is moving along Oz. All these results are summarized in Fig.J.transition starting from g. 1(c) are flat. 6. The atom is optically pumped in g. on the contrary.112 so Fig. . CohenTannoudji for the configuration of Fig. 1(a) since the energy levels of Fig. respectively. where the polarization is o+ [Fig. Am.C. for an atom at rest in z. is also the most populated one. if z = 0. Finally. the atom will on average remain on the same sublevel. 3 that the energies of the groundstate sublevels oscillate in space with a spatial period /2 and that the lowestenergy sublevel is also the most populated one.[Fig. Equilibrium Internal State for an Atom at Rest We first show that. at z = X/8 (see Fig. to the bottom of the next potential valley at z = 3X/ 8. since the atransition starting from g. we take a Jg = 1/2 gm Lin transition. 16the modulation of the dressedstate energies is much larger than r. B 2027 Energy new cooling mechanism associated with the o+oo configuration. one can also understand how the atomic mo mentum changes during the climbing. It thus appears that. 4). 2. 1(b)]. because of the time lag T. The lowest sublevel. Here we have an atomic example of the Sisyphus myth that is quite analogous to the cooling effect that occurs in stimulated molasses and discussed in Ref.1 1 2 and g11 2 are equal to 1and 0. the atom is always climbing potential hills. 11/November 1989/J. 1(b)].1/2 is three times as intense as the a. the energies and the populations of the two groundstate sublevels depend on z. The populations of g1/2 and g2 are equal to 0 and 1. Dalibard and C. The emphasis in this section will be on physical ideas. 1. The sizes of the black circles represented on each sublevel are proportional to the e 3 e_ 1 1 2 e+1/2 try to understand how its velocity can be damped. Mechanism of Energy Dissipation and Order of Magnitude of the Friction Coefficient The previous physical picture clearly shows how the atomic kinetic energy is converted into potential energy: The atom climbs a potential hill. having the largest negative light shift. No. the detuning is a. The key point is that optical pumping between the two groundstate sublevels takes a finite time rp. but we now have A+' = 3'. Multilevel Atom Moving in a Gradient of Ellipticity 0 Lin X/8 C A/4 Lin 3X/8 C+ A/2  5A/8 Y z  The laser configuration is the lin transitions lin configuration of Fig. We show in Subsection 3. The ClebschGordan coefficients of the various em. but it has a completely different physical interpretation. There is a corre . A/2. 1(b)]. A. Je = 3/2 1(b). for example.e. If the velocity v is such that the atom travels over a distance of the order of /4 during p. as a function of z the lightshifted energies of the two ground state sublevels of an atom at rest in z. Opt. We assume here that. 2. symmetry considerations show that both sublevels are equally populated and undergo the same (negative) light shift equal to 2/3 times the maximum light shift occurring for a + or polarization. because the atom is now optically pumped into g 2. climb the potential hill. for example. 91/2 9+1/2 9_1/2 Suppose. that z = X/8so that the polarization that the steadystate populations of g. if the atom is in a place where the polarization is linear. based on optical Bloch equations. As in Subsection 2. dipole forces. 3. and that it moves to the right. 4). 16. Wermust also note that. 3. is presented in the followingsections. Note. the same sequence can be repeated (see the solid lines in Fig.. Lightshifted energiesand steadystate populations (represented by filled circles) for a Jg = 1/2 ground state in the lin lin configuration and for negative detuning. [Fig. the previous conclusions are reversed. 3X/8. transforming part of its kinetic energy into potential energy.B that there is a l Vol. In the same way as for the usual 91/2 9+1/2 coefficients for a Fig. and we a= W If the atom is at z  WA (3. Both light shifts are still negative. the light shift A' of g11 is three times larger (in modulus) than the lightshift A+' of g112 . Suppose that the atom starts from the bottom of a potential valley. however.12. which shows. It clearly appears in Fig. for example.
Note.7) that hIA'I is divided by r'. howev ton.8a) (3.3) els remain constant when the atom moves along Oz. is much smaller than the velocity capture range for Doppler cooling (given by kv . It then follows from expression (3. 1(c) several horizontal lines instead of a single one. 11/November 1989 J. The atom could slide down a potential hill and transform its potential energy into kinetic energy. Multilevel Atom Moving in a Rotating Laser Polarization The laser configuration is now the o+alaser configuration of Fig. that the velocity capture range of this new friction force. blueshifted by an amount corresponding to the lightshift splitting between the two groundstate sublevels.hk2 . which is of the order of hk 2 . there is a great probability that it will absorb a laser photon and emit a fluorescence pho 11>> r) than the optimal friction coefficient for the usual Doppler cooling. This clearly distinguishes this new friction force from the usual one occurring in Doppler cooling. We see from Fig. in which case the atom o Since all the previous considerations are restricted to the <<r). B/Vol. in addition. The cooling force is then sponding change of momentum of the laser field because of a coherent redistribution of photons between the two counterpropagating waves. Dalibard and C. IA'I are both proportional to the laser intensity. (3. even when v is given by expression (3. when of the friction force. i. we can keep the linear expression (3. Assuming. due to optical pumping. so that dW 2 dt (3. Atomic Sisyphus effect in the lin lin configuration. CohenTannoudji On the other hand. we finally obtain (3.3). Opt. and even larger (since reaches the top of the hill. the energy dissipated during Tp is of the order of hA' (since A' < 0). finally.4 One must not forget. If Jg were larger than 1/2.3) of v. It is clear in Fig. we can derive an order of magnitude of the friction coefficient a appearing in the lowvelocity expression In other words. As shown in Fig.2028 J. and Such a result is where r' = 1/Tp. Because of the time lag r.6) and using expression ahk 2 I' A. The velocity of the atom represented here is such that vrp . The gain of potential energy at the expense of kinetic energy is dissipated by spontaneous Raman antiStokes photons that carry away the excess of energy. since it corresponds to the light shift of the ground state at low laser power.8b) A. however. 1(c). 4 that when the atom hWL (3. .3 . From the previous discussion. 4 is so small. No. (3. 4. we get rF' so that g2r/52. the atom sees on the average more uphill parts than downhill ones. forming an helix with a pitch X. we would have in Fig. (3. 6.X.3). TP.2) .9) Note.7) or (3. (3. A' and r' lowintensity limit (we want to have rF. 1(a) for which the laser polarization remains linear and rotates around Oz. We see in expression (3.. nism quite analogous to the one occurring in stimulated 16 molasses. Optical pumping is the mechanism of energy dissipation essential for introducing irreversibility into the process and for producing cooling. the weakness of dipole forces is compensated for by the length of the opticalpumping times.9)] is large. the light shifts of the groundstate there is no possibility of a Sisyphus effect. We can still transform expression travels overs X in a relaxation time close to its maximal value.7) that the friction coefficient of this new cooling mechanism is independent of the laser power at low power. One can also understand why a is so large. which is also very small since the opticalpumping time is very long. Here also we find a mecha er. Tp (3. 4 that the maximum kv r.7) by using the expressions of r' and A' at lowpower (Q << >>r) in r).5) II I l l Since we evaluate only orders of magnitude. Photons are absorbed from one wave and transferred by stimulated emission to the other wave. despite the fact that the size hA'! of the potential hills of Fig. that the friction coefficient [expression (3. (3.TlI I I II I I dW dt Fv.6) Equating expressions (3. which is given by expression (3. B. sublev F= av value of the friction force occurs when vrp (3. _ Q2/6' a .X/4. Soc.4) have different light shifts and since there are several possible values for mlwhen Jg > 1/2.r). we can also calculate dW/dt from F: Energy . All these processes are conservative and could occur in both ways. which is linear in laser power. For this value of v. so the energy dissipated per unit time is easily extended to all values of Jg.2) of F. since sublevels gm with different values of Iml dW hA = hA'r'.e. a large detuning (131 order to have in the groundstate light shifts larger than the level widths. Am.4) and (3.7) I I I I 1 2! z X/8 A/4 3A/8 A/2 5A/8 Fig. that the energy dissipated here is much smaller.
the two counterpropagating waves are absorbed with different efficiencies. Larmor's theorem tells us that. B 2029 sublevels adjust themselves to values such that the mean number of Stokes processes from Igo). (3. 11/November 1989/J. (3. 6. which governs the light shift A. the transition Jg = 1 Je = 2 (see Fig.12) the eigenstates of Jy (J: angular momentum). both sublevels Ig~. one can show (see Appendix A) that the new Hamiltonian. The steadystate populations of these states (4/17.'. there is an for the light shifts (since the laser intensity does not change with z) and to the same steadystate populations. (2. It is thus clear that in steady state.Ig_)y proportional to (T73)2(11/@)2 = 1/9. Ig_1)y. in this moving rotating frame. To the lowest order in wL 2/62. spectrum. and g+. Atomic level scheme and ClebshGordan coefficients for a Jg = 1 Je = 2 transition. an inertial field will appear as a result of the rotation.10) As in the previous subsection. balances the mean number of antiStokes processes from Ig_)y to Igo)y. and a RamanantiStokes line at WL . which rotates around Oz in the plane xOy.11) In its rest frame. If the atom is in a different location but still at rest. nonadiabatic atomic orientation along Oz that appears in the ground state as a result of atomic motion. that the wave functions vary in space. since the lightshifted Zeeman sublevels are the eigenstates of the component of J along the rotating laser polarization ey..)yor IgI)y and ends in Igo)y. We also have a RamanStokes line at L + (A0'/4) (remember that A0' < 0).5b)] = kz = kvt. In steady state. CohenTannoudji Vol. corresponding to cycles where the atom starts from Ig 0 )y and ends in Ig+. If we take the quantization axis along the local polarization. tA!I=3tS 4 0 I 91> we find first a Rayleigh line at Y 90 >y / 19+1>y corresponding to fluorescence cycles where the atom starts and ends in the same groundstate sublevel.(A 0 '/ 4) will be equal. 5). The double arrows represent couplings be . the same calculations can be repeated. with a 7rpolarization along Oy. even at very low velocity. Moving AtomTransformation to a Moving Rotating Frame The atom is now moving with a velocity v along Oz: z = Vt. 1(a)]. 1.A. The quantization axis Oy is chosen along the resulting linear laser polarization. Equilibrium Internal State for an Atom at Rest We suppose first that the atom is at rest in z = 0.)y or Ig. We must also note that. 6. For subsequent discussions.)y undergo the same It is then convenient to introduce./2)2(1/)2 = 1/4 is greater than the rate Igo)y .)y are equal to 9/17. This inertial field looks like a (fictitious) magnetic field parallel to the rotation axis Oz and has an amplitude such that the corresponding Larmor frequency is equal to the rotation speed k. which is at z = 0 [see Fig.)y. will concen trate atoms in Igo)y. 5. Am. We consider here the simplest possible atomic transition for such a scheme.to Ig1). we see that optical pumping. in the atomic rest frame. Lightshifted groundstate tween Zeeman sublevels owing to the transformation to the moving rotating frame. Figure 6 represents the lightshifted groundstate sublevels in z = 0 with their steadystate populations. 1(a) and Eq. the mean number of photons emitted per unit time at WL + ( 0'/4) and WL . we take a red detuning so that A0' and A' are both negative. we have considered only an atom at rest in z = 0. 4/17) are represented by the filled circles. The steadystate populations of g0)y. however.configuration. We suppose that the laser power is very weak ( << r) and that the detuning is large (161 >>r).(A 0'/4). More precisely. giving rise to the same values In this subsection we describe a new cooling mechanism that works in the o+olaser configuration for atoms with Jg > 1 and is quite different from the one discussed in Subsection 3. (3. We show that. 2. which moves with the same velocity v. Ig)y. > Igo)y proportional to (1/. Because of this highly sensitive motioninduced atomic orientation. It follows that. and if we note that Ig_)y. the populations of the various groundstate Fig. since the r transition starting from Igo)y is 4/3 as more intense as the two r transitions starting from lg9+)y. 9/17. which gives rise to unbalanced radiation pressures and consequently to a net friction force. giving rise to a symmetrical fluorescence 91 9o 9+1 Fig. No. Of course. Dalibard and C. the atom sees a linear polarization ey. and 4/17. Opt.J. when the atom moves along Oz. couplings can appear among the various Zeeman sublevels undergoing different light shifts. sublevels of a Jg = 1 Je = 2 transition in the a+a. Igo)y. corresponding to the inverse processes where the atom starts from Ig+. making an angle with Oy [see Fig. So far. it will be useful to analyze briefly the spectrum of the fluorescence light emitted by an atom at rest in z = 0. 4/17. We must note.since the opticalpumping rate Ig). smaller (in modulus) than the lightshift A' of Igo)Y A0' = 4/. respectively. a rotating frame such that in this moving rotating frame the laser polarization keeps a fixed direction. Soc.
y basis and since Eqs. Opt. More important is the modification of the wavefunction. we get from Eq. A' is negative.yj basis to state that is parallel to Oz and sensitive to kv.19) ll~il.)y = Ig+1)Y + i.)y Ig1)yj order in kv/A'. From Eqs (3. one finds for the steadystate value of J. (3._ Ao') .15) and (3. the modifications introduced by atomic motion concern only the wave functions. Igo) (3. 6. for these populations and using (Jz)st = h(l+ 1 .16) give the new states in terms of the old ones. To demonstrate this result.17b) y. Soc. polarization.). so that the energy splitting between the groundstate energy levels is much larger than their widths: r << IAI.15) and (3.1.. 11/November 1989 J. 6). this inertial term introduces couplings proportional to kv between Igo)y and Igai)y (double arrows in Fig. Since we have chosen a red detuning (6 < 0).A' (3.B.. Noting 11+1and II.14b) Y(g'11Jzlg+1) = Y(g 1IJzg.e. We also suppose that kv <<IA'.17)] of the corresponding levels and summing over m. (kv/ If we compare the new Hamiltonian in the moving rotating frame with the Hamiltonian for an atom at rest in z = 0 considered in Subsection 3.20) we get from firstorder perturbation theory go). it is important to determine first what the new energy levels are in the ground state as well as the new steadystate the lIgm)}.16). as they are in Ig)y. Since J. in the basis {Igm).16a) population > 0. These couplings are sometimes called nonadiabatic since they vanish when v tends to 0. the energies and populations of the groundstate sublevels are the same as in Subsection 3.. Now. (3.13)] and also to understand the energy exchanges between the atom and the laser field.(g+Ipst g+. The state Igo)yis contaminated by Ig)y and Ig1)y to first order in kv/A'. that v > 0. Since we know the matrix elements of J. . consider the level Igo)yin Fig. the important point is that in the perturbed states 1Igm).which is proportional to this population difference.1: Y(golpsaIgo)y = 9/17. becoming the perturbed state go)y.17a) (3.1)y = 4/17.. provided that we change from the nonzero matrix elements among the eigenstates of Jy. we get WJ. To first order in kvIA' and to zeroth order in I'/A'.C.13)].. No.(g_1 Ptg. (3.C. in term with a fixedpolarization laser term resulting from the rotation and Vrot = kvJ.16b) 4. contains.B. i. (3. = Fgoy Ig~y+ 1g 0). that 9 V )~ =2hkv st AOz Al '\17 2 17 2 17) 40 hkv 17 AO' (3. are presented in Section 5. B/Vol.2 that the problem studied We now study the steadystate density matrix in the ener here is formally equivalent to the problem of an atom at rest. atomic orientation in the ground state. while their wave functions become fg+. It follows from Eq.21) We show in the next subsection how this motioninduced difference can give rise to a new much more efficient friction mechanism than the one used in Doppler cooling.)y = y. (3. We show in the next subsection how they can give rise to an atomic orientation in the ground IIgm). + Y g_') kv J_2(A'l . it is possible to show (see Appendix A) that. the steadystate density matrix Petis diagonal in the 1Igm)y basis and has the same diagonal elements as those calculated in Subsection 3.A'l) (3. (3..y of eigenstates of Jy (see Appendix A).14a) Weighting these values by the populations [Eqs. and Ig.+ At§(kv Al') ~2_Ao Ig+ 1). has A')(P'/A) or higher. (3. To sum up. when the atom moves along Oz in a rotating First. CohenTannoudji gy basis jgm)y}.> H+. (3. (3.18b) (3. Thus we have shown that. not restricted by conditions such as Eqs.19) which permits a perturbative treatment of the effect of Vrot. To simplify the calculations. we assume here that the light shift IA'Iis much larger than F'. The perturbation kvJZ that has no diagonal element in Igo)yshifts this level only to second order in kv/A'. 3.H. Dalibard and C. MotionInduced Orientation in the Atomic Ground State In order to get a clear insight into the modifications introduced by the inertial term [Eq. in the manifold tIg+l )y.1)z is more populated than V Since the matrix elements of J. ~__A'AO')0 ±2_ Ig = Igl)y + . an extra inertial equal to (3.13) in the lIgm). have different steadystate populations.15) Such a result is in quantitative agreement with the more detailed calculations in Section 5. Since we know the effect of optical pumping atomic evolution after the unitary moving rotating frame. are not changed to first are zero. 6 <0 .. goy=2hkv =AO' ..1)Y = A .10). 6. we calculate the average value of J in Igm)y. Am.y the populations of the two eigenstates Ig+ 1). of J. Suppose that the atom moves toward z > 0. (3.ll1). (3. basis.1 we see that all the new effects that are due to atomic motion in a rotating laser polarization must come from the inertial term [Eq. Remarks (i) We saw in Subsection 3. if we neglect terms of order (kV/A') 2 . the two eigenstates Igu)z of J.~ 17AO' = kv (3. are not equal. (JZ)st is a motioninduced laser Quantitative calculations.18a) density matrix in such an energy basis. transformation to the addition to a coupling field. . the energies of Ig).AO' y ~i__ hkv h (3. where we have used Eqs.14) and therefore valid for any velocity..20) that Ig.2030 J. (3.
however. Consequently. giving rise to a new nonzero component of the alignment (JJy + J. kV/IA'I is much larger than the corresponding parameter kV/r characterizing Doppler cooling.23) F is proportional to v and opposite v since the light shift A' is negative for 6 < 0. is very small. At low laser power A'1 << r. (3. Soc.. that gating a. The atom will scatter more counterpropa .22) where r is of the order of the mean scattering rate of photons by an atom in the ground state.e. which has the same symmetry axis.23). i. No.21) that the sublevel Ig1). Opt. with an amplitude such that the Larmor frequency in Bo is equal to kv. We must emphasize here that the fact that the two radiation pressures become unbalanced when the atom moves is due not to the Doppler effect. The reverse conclusions can be drawn for an atom in Igi)z. It is well known. Getting Unbalanced Radiation Pressures with a MotionInduced Atomic Orientation Looking at Figs. Dalibard and C. We can now give the order of magnitude of the new friction force. that the friction coefficient associated with expression (3. (ii) One can easily understand why the new cooling rI. we must work in the energy basis Igm). is more populated than g+ 1)z. the mean force F acting upon the atom. as in Doppler cooling. but to a difference of populations in the ground state that is induced by the inertial term [Eq.Jx). If the atom moves toward z > 0 and if the detuning 6 is negative.) along Oz. We will see. 1(a) and 5.photons than copropagating a+ ones. It follows that the radiation pressures exerted by the two oand o+waves will be unbalanced. cannot be connected to the same excited state by the two laser polarizations + and o. for example.. as in the Sisyphus effect of Subsection 3.24) is smaller than the one associated with expression (3.so that no coherence can build up between these two sublevels.B. and its veloc ity will be damped.. 24 and 25 and dealing with the orientation produced by the interaction of an atomic alignment with a real or fictitious electric field. such a formulation of the problem allows us to establish a connection between the effects studied in this paper and other wellknown effects that were previously observed in opticalpumping experiments. we saw above in expression (3. however. leading to usual Doppler cooling. Optical pumping with a linearly polarized light cannot therefore introduce any anisotropy in a Jg = 1/2 ground state.states. CohenTannoudji Vol. is of the order of Fhk2L A' V (3.3. As in mechanism studied in this section does not work for a ground state Jg = 1/2. B 2031 interacting with a laser field linearly polarized along Oy and submitted to a static magnetic field Bo along Oz. 6. The difference between the number of c+ and o Actually. to a coherent redistribution of photons between the two counterpropagating waves but to the fact that the atom scatters more photons from one wave than from the other. Examples of such structures are zerofield levelcrossingresonances and Hanle resonances in 22 2 3 atomic ground states. (3.13)].). We can now try to understand how the atomic kinetic energy is dissipated.ll. It appears clearly in Eq. Another way of interpreting this result is to note that the two sublevels Iga12). In the absence of Bo._J2) produced by optical pumping does not precess around Eo. Changing v is equivalent to changing IBoI. The two radiation pressures get unbalanced when the atom moves. therefore F is a friction force. This is due not.A. i. will absorb a photon propagating toward z < 0 than that it will absorb a + photon propagating toward z > 0. Note.20) and neglecting numerical factors. of the order of the Stark shifts that would be produced by a fictitious static electric field Eo parallel to Oy.kvA' r.2. the problem studied here is a little more complicated than a pure Hanle effect in the ground state since the laser beam not only introduces an atomic alignment along Oy ((3J. CY hk 2rA Al (3. 11/November 1989/J. Mechanism of Energy Dissipation The physical picture presented in the previous subsection clearly shows how the atomic momentum decreases in this new cooling scheme. one sees there is a six times is much smaller than expression (3. (3. We will see in Section the application of a static field Bo in a direction different from the symmetry axis of the laser polarization can give rise to very narrow structures in the variations with B 01of the light absorbed or emitted. Note that here we ignore the Zeeman coherence between Ig i)z and Ig+i)z.14a)].. which can be very small at low power.24) Zeeman detuning comparable with r that the two counterpropagating laser beams begin to be scattered with different efficiencies. B 01.J. In order to have a precise energy balance. at least when v.e. which have the same symmetry as photons scattered per unit time is.). It is the interaction of this alignment with Eo that gives rise to the orientation (J. and we have to wait long enough that we greater probability that an atom in Ig.7).y introduced in Subsection 3.7).. since it varies as r/i instead of A'/r' and since we suppose here that r << 1A'! [see expression (3. (3. there is an analogy between the motioninduced atomic orientation studied here and the effects described in Refs. 5 that the contribution of such a Zeeman coherence does not change the previous conclusion..2. in Sections 4 and 5 that the diffusion coefficient associated with expression (3. the new cooling mechanism works at velocities much lower than for Doppler cooling.A we find that F is independent of the laser power at low power since F' and A' are both proportional to the laser power. Am. ' These resonances have a narrow width ABO such that the Larmor frequency in AB(is equal to the width r' of the ground state.2J2) differs from zero) but also produces light shifts of the Igm). the alignment (3J. 5. When Bo is applied this alignment starts to precess around Bo. It is only when B0! is large enough to produce Subsection 3. Since each + () scattered photon transfers to the atom a mean momentum + hk (hk). In a J = 1/2 state no alignment can exist (WignerEckart theorem). so that both configurations lead to equilibrium temperatures of the same order. In a certain sense. according to Eq.20) that the dimensionless parameter characterizing this new cooling mechanism is kV/IA'I. we conclude that the mean momentum transferred to the atom per unit time. 6.
and A+ and 6. Am. Let us. due to the fact that in the laboratory frame the fluorescence photons have on average a blue Doppler shift on the Rayleigh line as well as on the two Raman lines. which in turn may lead to temperatures well interaction of the atom with two standing waves. and shifted with respect to each other by X/4. The average force acting on the atom can now be derived below hr.2) fin 1 fin CONFIGURATION This section is devoted to a quantitative study of the cooling mechanism presented in Subsection 3.25) (.c. 4.3) as describing the r). this ensures that pumping times rPmuch longer than the radiative lifetime TR can appear. First. adjust themselves to values such that the rate of Stokes processes balances the rate of antiStokes processes from go)y to g±L)_y from Ig+)y to Igo). This also introduces an important simplification. we study the internal atomic evolution.A.23)].1 /2.27 spatial modulation of the light shifts of the atomic Zeeman ground sublevels. in order to simplify the mathematical expression of V.are the raising and lowering parts of the atomic electricdipole operator. Consider.3 /2 ) (g.1. On average. A. (2. we shifted the origin on the z axis by an amount /8.) It is then clear that the steadystate populations of Igo). whereas it has a Doppler shift 2kv when it is reemit ted toward z > 0. the fluorescence pho V =[D+ *G6(r)exp(ioLt) + D (r)exp(iWLt)]. it has no Doppler shift in the laboratory frame.c. No.1 /2 ) + C. we evaluate the momentum diffusion coefficient and derive the equilibrium temperature of atoms cooled in the lin 1 lin configuration. giving the evolution of the groundstate populations.2) also can be interpreted approach to evaluate the friction force and the equilibrium temperature reached by the atom. We then calculate the average radiative force as a function of the atomic velocity.1 /2 ) (9l/2 1exp(HiWLt)+ h. we get V = a sin kzi e3 / 2 ) (g 1/ 21'+ X exp(iWLt) + 1lel/ 2 ) dW__r FT'(+lIIl)hkv dt  A'l 2V 2 hk (3.photon. Starting from the optical Bloch equations that describe the evolution 26 we obtain two equations of the atomic density operator. to much simpler calculations since it permits a perturbative treatment of the problem.4) of these quantities is negligible. We also restrict our calculation of the radiative force to from the spatial gradient of V: / dV\ dz hkd cos kZ[P g 1 2 . Dalibard and C. This lowpower hypothesis leads. and Igl). Indeed.2032 J. The laser field for the lin 1 lin configuration was given in Eq. Internal Atomic Evolution Actually. for example.= (+)* are the positive. As in Subsection 3. in this lowvelocity domain. 11/November 1989 J. The total Doppler energy dissipated per unit time is therefore equal to D+ and D.1/ 2lj which coincides with the power Fv dissipated by the friction force [expression (3. e. Our treatment is limited to the lowpower domain (Q << (4. Let us recall that this mechanism is based on a Sisyphus effect induced by a Note that. this potential energy then being dissipated by Raman antiStokes processes. Opt.1 2 1 + 1 le. For a a+ photon a similar argument gives hkv. as was done in the first explanations of this cooling mechanism. As in Subsection 3.photon is hkv. which gives the friction coefficient and the velocity capture range for this coolingmechanism. a Cos kz[le.. we find a fluo rescence spectrum that remains symmetric even when the atom moves. e 3 /2 ++ hkQ sin kz PD(i/2. to invoke di rect nonadiabatic transitions between Igo)yand Ig± 1)y converting kinetic energy into potential energy. a+ and apolarized.1) ton followingthe absorption of a . As explained in the Introduction.2). B/Vol. e.B. Soc. (Of course. Consequently. Finally. THEORY OF LASER COOLING IN THE (4. All the calculations of this paper are done in the electricdipole and rotatingwave approximations so that the atomlaser field coupling can be written as as in Doppler cooling. in addition. 2.7). since there are always as many Raman Stokes as Raman antiStokes photons emitted per unit time. . Atomic motion can therefore greatly affect the atomic groundstate dynamics and then induce a large velocitydependent force.A. Inserting its value into the atomfield coupling and using the ClebshGordan coefficients indicated in Fig. Qrepresents the Rabi frequency for each of the two running waves calculated for an atomic transition with a ClebshGordan coefficient equal to 1 and with a reduced dipole moment for the transition equal to d: Q = 2 dho * h Expression (4.3 / 2) + Ag. however. 6. the Doppler energy dissipated by a fluorescence cycle involving a . CohenTannoudji can reach a steady state and get an energy resolution better than hlA'I for the scattered photons. this time should not be too long so that we can neglect any velocity change resulting from the friction force. el/ 2 ) + CC the lowvelocity domain (Doppler shift kv << r). it does not seem appropriate.and negativefrequency components of the laser electric field. (4. (4. emphasize that we do not impose any condition on kvand the inverse r' of the pumping time r. If it is reemitted toward z < 0. we study here a Jg = 1/2 J = 3/2 atomic transition and use a semiclassical In Eq.. optical coherences (densitymatrix elements between ground and excited states) and excitedstate populations are almost unaffected by atomic motion: The effect of the Doppler shift kvduring the relaxation time TR (or 2 TR)  P( 1 /2. for this second cooling scheme the dissipation is. (4.
e3/2) = + )p(gj.8) Consequently. we can again note that they follow adiabatically the variations of the groundstate populations so that (e3 /2 1ple 3 /2 ) = SO111/2 sin2 kz.J.we first replace the optical coherences contributing to f by their expressions in terms of the groundstate populations. Opt. CohenTannoudji Vol. 2.12. h~so COS 2k z+!1sin2hz) (4. e3 /2)] (4.13) where the pumping time rpand the stationary populations 11ist(z) are given by 1/rp = r = 2MOs/. e). Force in the lin 1 lin Configuration In order to calculate the radiative force f [Eq. we obtain AE = hA+' = h6so(sin2 kz + 1 cos2 kz) +i a sin kz[p(e 3 /2. 12 .12) This equation is valid for any power and for any atomic velocity 2(t). 2 I11/2sinkz. Dalibard and C. we can neglect (eilplej) in Eq. For example. e3/2) = In112St(z) = sin2 kz. (4. = Eo AE12= hA. the relaxation time We insert the expressions of optical coherences and excitedstate populations in terms of Ih11/2 to get I~i T [Ili Ii (z)]. (4. we now need to calculate the groundstate populations II41/2 in order to evaluate the force [Eq. since the relaxation time of the excitedstate populations is r1. 11/November 1989/J.9) In this equation we replace the optical coherences by their expressions in terms of groundstate populations. for example. (4.1/ 2)cos kz + c. we now write the optical Bloch for excitedstate (e3.6) and calculate p(g. e 3 /2 ) Finally. (4.10) where =E + h6so 3 cos 2kz. B. We can indeed add these two terms independently in the lowintensity domain since the two states g1/2 and g1/ 2 are not connected to the same excited sublevel.7) For an atom at rest in z the populations IHi(z)reach their stationary value in a time Tp inversely proportional to the laser power.' hbs( 3 = . since it is not coupled to the groundstate populations.4)]. populations. Now. We also note that these stationary populations are strongly modulated in space (see Fig.2) of the laseratom coupling V. e3 /2 ) to first order in U.14a) (4.16) where sois the detuningdependent saturation parameter . (4. It can be simplified in the lowpowerand lowvelocity domains in the followingways: (i) The lowintensity hypothesis ( << r) implies that the populations and coherences of the excited state remain very small compared with the populations of the ground state. (4. we have for II1/2 I1/2= r (e3/21Ple3/2)+ (el/21piel/2) +  (e1/2LPe1/2) p(gj 12. ej) = (gjIplej)exp(iWLt). l1 1/2 st(z) = cos2 kz. as mentioned in the Introduction.6) in p(g.i + a sin kz[(g 12 jpjg. e3 2) .2Iple3/2): (e 312 1ple312 ) = r(e 3 /2 1p1e 3/ 2 ) expression (4. We now need to calculate the steadystate value of the optical coherences p(gi. This means that the optical coherence p(g 1 /2 . B u2/2 62+ 2033 with p(gi.21pIg112 ) does not contribute to the calculation and actually has a zero steadystate value for this laser configuration.4)]. Am. Using the ClebshGordan coefficients given in Fig. For example. e. (4.12. using optical Bloch equations.5) SO= L 4  (4. 1 2)  cos kz(e_/21p1e 3 /2) + [g p(g. 1 2 . . 3). for hk5sO (I11/2 112 )sin 2kz.11) where p is the steadystate density operator. with II=1/2 (gai/21PIgi/ 2) (4.cos 2kz. we have for p(g. (4. (ii) Since the velocity is low (kv <<r). we write the optical Bloch equations for the groundstate populations. (4. (4.15) This expression has a clear physical meaning in terms of light shifts: The two levelsgl/2 and g1/2are lightshifted by a quantity that corresponds to the sum of the two light shifts created by the two 0+ and o standing waves appearing in In order to calculate II±1/2.c . This leads to f = 3 Note that the coherence between the two ground states (gi. Soc. e3 / 2 )sin kz + (e3 1ple 2 3/ 2 )]. This would no longer be true for a more complicated atomic transition for which nonzero steadystate coherences can appear among groundstate equations sublevels. No. e/2) follows adiabatically the groundstate population so that we can write p(g. 1 2. g 1 / 2) (g 12 . Consequently. 6.14b) 2r1 of optical coherences is much shorter than the typical evolution time of sin(kz) (of the order of /kv) or of the groundstate populations (rp).r 2 (4.
13): 1:11/(Z with a long residence time p. where the friction coefficient a is equal to a = As sketched in Fig.15)] can be written as f = fl/2]Il/2 + f.26) so that the average force [Eq.22) We insert these expressions into the result [Eq.20) khV = 1/2TP.14a).18) r (6 0). 11/November 1989 J.3. Note that expression (4. v = As predicted in Subsection 3.17) E = 32h6so. we come to the calculation of the force on the whole range kv << r with no restriction on the relative values of kv and 1/Pr.14b)]. ) = flSt(Z) . (4. it is possible to get an exact solution for the forced regime of Eq. On the other hand.7) or (4. we did not calculate the full velocity dependence of the force.19) for the force and averaging over a wavelength. We first evaluate the momentum diffusion coefficient DP and then calculate the equilibrium temperature resulting from the competition between the cooling described above and the heating from diffusion fi* i=±1/2 0) + 4 hk2 6sOvTp sin 2 (2kz). (4. we get f12 or. To sum up.29) 6 h6so sin 2 2kz. 7 we have plotted this force versus v (solid curve). and we get f(Z.28)] that gives both the friction coefficient a (slope at origin) and the velocity capture range vC.23) We now average this result over a wavelength. We now calculate the populations 11±1/2 in order to evaluate the force from Eq. The average (4. that is. and the dominant process is Doppler cooling. (4.hkhso sin 2z cos 2kz 3 2 =1 F cos 2kz + 2kvrP sin 2kz\ 11412(Z V)2 1F 1 + 4V2Tp ) (4. By contrast. so that we obtain f(z. the polarization gradient coolingbecomes inefficient. These two populations are then equal to their steadystate value [Eq.19). (4.the force is maximal in this configuration when v = vu. (4. (4. This spatial 2 hk 2 5sOTp (6<0) (4. Equilibrium Temperature in the lin lin Configuration We now turn to the problem of evaluating the equilibrium temperature in this new cooling scheme.28)of the force is valid only for kv << r. (4. v = 0) .24)] is proportional to laser power.28) =_ . For such very slow atoms. B/Vol. The differences between the slopes at v = 0 (friction coefficients) and between the capture ranges appear clearly. in the twolevel problem the two dressed states were strongly contaminated by the excited state. In Fig. (4. but they can no longer be solved by a perturbative expansion such as Eq. the effect of atomic motion on the populations 11/2 can be treated pertubatively by an expansion in terms of the small parameter kvTp: 11i(Z. (4. dz where the potential U is given by U(z) =  dU where the critical velocity V is (4.A. Opt.19) We therefore get a friction coefficient independent of the laser power [see relation (3. the expansion (4. the two lightshifted groundstate sublevels oscillate in space with a period X/2. an important difference arises here since each of the two levels involved in Eq.2034 J. taking the gradient of AE±1/2 .24) .22). provided that the levels g+1/ 2 are replaced by the two dressed levels.25) dependence then causes a statedependent force. v = )VTpr = f(z.9)].VT dz + . (4.10). so that their lifetime was r1. we obtained an analytical expression for the velocitydependent force [Eq. Such an expression is similar to the one obtained for a twolevel atom in intense laser light. of f(z. Indeed. a = 3hk2 = =d dz AE+1/2 2 . We now consider a very slow atom for which the Doppler shift kv is smaller than 1/TP.. (4. In Fig. (4. where 1/TP is proportional to laser power. Dalibard and C. Soc. V) = f(Z. The evolution equations (4.. (4. (4.1/2 . v = 0) derives from a potential is again similar to the corresponding result for the dipole force acting on a twolevel atom in a standing wave.22) requires kv << 1/ Tp. (4. CohenTannoudji (4.21) The fact that f(z. v = 0) is zero.13)are still valid. 3. the range of validity of the result [Eq. we obtain aV 2 A)=1 + (V /V 2)I (4. 6. which would allow one to study the transition between the two cooling regimes. 7 we plotted just (dotted curve) the force that one would get by independently summing the two radiation pressure forces exerted by each Dopplershifted wave.19) is essentially a groundstate sublevel Finally.11/2 . Am. C. when the distance covered during a pumping time is of the order of the spatial period of the modulated light shifts.hkhso sin 2kz. Note that this condition is much more stringent than the condition kv << r required for writing Eq. Fortunately.27) Inserting this result into expression (4. No. using Eq. (4. This force f is then just the average of the two statedependent forces f±1/2 weighted by the populations of these states. First.. so that we get f(v) = av.' 6 However. (4. For this lin I lin configuration. we take an atom at rest. Outside this range.19)] for the force. 3 (4.
11/November 1989/J.3r.34) D '' 10 7h 2k2rso./ 2(z) and f. it gives Dp"_ (4. 2 (z)]2 1/ 2 St(z) II_/ 2 St(z)Tp 1 transition.33) (ii) There are fluctuations in the difference among the number of photons absorbed in each of the two laser waves. The values of the parameters are Q2 = 0./2(z) and f/ 2 (z) at a rate 1/Tp. the two first ones are already present for a Jg = 0 Je = 1 transition. t. a = r. we discuss the validity of the semiclassical approximation used throughout this calculation. 7. Variations with velocity v of the force due to polarization gradients in the lin lin configuration for a Jg = 1/2 Je = 3/2 transition (solid curve). and it leads to Dp Je = 4[f. To evaluate the third contribution Finally.35) . t + T).J. CohenTannoudji Vol. so that we prefer to use a heuristic calculation here. such a calculation would be rather tedious.31) We assume that Eq. 1 2 (z). where P(i. Subsection 4B). (4. Once this is averaged over a wavelength. 2 ' the two first contributions give for a dipole radiation pattern For a Jg = 0 = 2h2 k2 . No.j. t. (coefficient Dp"). j. (4. 622r DrTh (4. Soc.32) which must be calculated for an atom at rest in z (the label z was omitted for simplification). (4. one could 3 compute the correlation function of the force operator. The dotted curve shows sum of the two radiation pressure forces exerted in dependently by the two Dopplershifted counterpropagating waves.04 VELOCITY unit I w /k 2 __ _ 2 0. Am.'2 "1 For a multilevel atom. B 2035 0. i=+1/2 j=1/2 (4. and its correlation function can be written as f(t)f(t + T)= > Y fjP(i. There are three main contributions to Dp. The force f(t) oscillates between f. (iii) There are fluctuations of the instantaneous dipole force oscillating back and forth between f. 16. 2 Fig.30) nitude for these two contributions in the case of a Jg = 1/2 Je = 3/2 transition. t + r) represents the probability of being in state i at time t and in state j at time t + T. 2 T FORCE (Ur i t k r/2) 0. Opt. Dalibard and C. The force due to polarization gradients leads to a much higher friction coefficient (slope at v = 0) but acts on a much narrower velocity range. kBT=  D a (4. 2 ' and the third one is specific of an atom with several groundstate sublevels: (i) There are fluctuations of the momentum carried away by fluorescence photons.so sin 4(2kz).31)still gives the good order of mag 3 h 2 So. 6. The calculation is then similar to the one done to evaluate the fluctuations of the dipole force for a twolevel atom (Ref. we start from = D dr[f(t)f(t+ T)  2]. In order to calculate the exact value of Dp.
5). and these coherences cannot build up. [Eq. Finally. we can distinguish three velocity domains: (i) For very low velocities (kv <<I). and the Doppler shift is now comparable to the natural width. one should then correct expression (4. Dalibard and C. we get 16 » r==. for instance. we expect the force ty classescontributing to the stationary state. In this lowintensity regime. . is probably the best way to derive precise results concerning the stationary state in this configuration. le. leading to rp >> Tror equivalently to r << r). B/Vol. Expression (4. qualitatively studied in Subsection 3. originates from the enhancement of radiation pressure imbalance that is due to a sensitive motioninduced atomic orientation.37) is well below the critical velocity v. much lower temperature. (iii) For higher velocities (kv > r).as long as the other hyperfine levels of the transition can be ignored: (i) (ii) At a given power ( fixed). this is not true. Note that such a treatment in the present configuration is more complicated than for the o+ar one. let us look for the lowest temperatures achievable in this configuration. At a given detuning. Opt. CohenTannoudji This second coefficient exceeds the first one as soon as 161 is larger than r. We therefore get Vrms<< Vc >> hr M 3 (4. In this domain. This simple result must be compared with the one obtained for a twolevel system [Eq.2036 J. le+. This also allows for a perturbative treatment of the problem: One needs to consider only the density matrix in the ground state (3 X 3 elements) instead of the total atomic density matrix (8 X 8 elements). 5.B. The first is the usual Doppler cooling that results from differential absorption of the C+and cr waves when the Doppler shift kv is a nonnegligible part of when the detuning increases. Consequently. THEORY OF LASER COOLING IN THE o. owing to B. the polarization gradient is then quite different from the one studied in Section 4: The polarization is linear in any place and rotates along the propagation axis Oz on the length scale X. (1. cooling in the lin I lin configuration.37) suggests that an arbitrarily low temperature could be reached. 161 >>r . the lowest achievable rms velocity in this model remains larger than the recoil velocity. since the coupling between populations and coherences is responsible for the motioninduced orientation.). which indicates that in the stationary state atoms are bunched around the points z = nX/4 rather than uniformly distributed.39) Consequently. we start with the atomlaser coupling [Eq. (4. by decreasing the laser power.13)] is large compared with their damping time rp. (4. Indeed one must check that the rms velocity deduced from expression (4. and it is consistent with two ex7 9 perimental observations.D for the a+of configuration. which was obtained by assuming a constant velocity. 11/November 1989 J.1)]. (4.B. the Doppler shift kv is still small compared with F.11) for so. we restrict our treatment to the lowintensity domain (Q << r.38) to be linear with velocity. 6.kBT 8161* (4. which we . We are currently working on a full quantum treatment of sublevels [owing to the inertial term in Eq. so that the usual Dopplercooling force remains small. analogous to the one presented at the end of Subsection 5. Let us recall that the recoil velocity was.). the precession frequency kv of the coherences among groundstate which puts a lower bound on the laser power required for our treatment to be valid. neglecting Dp'. we need to take into account all coherences among the various groundstate sublevels.37) Now we come to this final section of this paper. The second mechanism. the force is then practically equal to the usual Doppler cooling force. Various Velocity Domains As in Section 4. It appears in these expressions that the residual kinetic energy is of the order of the light shift hA' of the ground state. We study here the case of a Jg = 1 Je = 2 transition (Fig. This in turn gives a lower bound on the achievable rms velocity: hh 161 Vrms>> r (4. the temperature decreases Cooling in this configuration originates from two quite distinct processes.24) for the force.2)]. a limit for the validity of our semiclassical treatment. Actually. In order to calculate this force.21). which can be formally obtained from the real 12 transition by removing the V system formed by Igo). A Monte Carlo simulation. as explained in Subsection 3. Calculation of the Cooling Force In order to calculate the steadystate radiative force. We actually find that these two quantities are of the same order. 28. Therefore. On the other hand. Soc. reducing the power decreases the temperature. It may lead to a As shown in Section 2. which is devoted to the quantitative study of laser coolingin a a+ofconfiguration for a Jg = 1 Je = 2 atomic transition. the coherences among the groundstate sublevels are completely negligible. No.CONFIGURATION + or. For simplification. Another important remark concerns the comparison between the residual kinetic energy kBT/2 and the potential U(z) derived in Eq. (3. the new cooling force decreases. A. (ii) In the intermediate regime (r << kv << F). is the simplest atomic transition exhibiting this sensitive velocityinduced orientation. so that the cooling force is indeed linear for all veloci the natural width r. In this regime.36) the possibility that photons are coherently redistributed between the two waves.29)]. we also use at the end of this section a fictitious W atom. using definition (4. in any event. hBT Dp a h1I 4 so 416 (4. similar to the one in Ref. which. Am.
5) In this equation. Note that. g3). B 2037 write in the atomic reference frame. Assuming +C(h+ 1 C 6 k) 6 + k(5  + (s  + + where S= Q2/2 (6 F kv)2 +4 << 1 (5.8) X exp[i(wt 2 [ I 1O)(elI+1g) (e1 * Ig1) h e2 + aW2  X exp[i(w+t where we have put hz)] + h. (5. (5. and g1. The resulting radiation pressure force. in this way.C). which involve the laserinduced coherence between g. The first (third) line of Eq. g1) +  I p(el.c. . Cr. these coefficients are nothing but the matrix elements of p in the moving rotating frame defined in Subsection 3. Take... The second and third terms of Eq. which give the evolution of the atomic density operator p. No. .7) C= 56 '34 We can. an atom in the state gl.11) where we have put. (5. (4. weighted by the population II. as in Section 4 [cf.3)]. J Ci = Im[(glplgl)exp(2ikvt)J. (4. the last two terms. 11/November 1989/J. The detailed calculation is rather long and tedious and is presented in Appendix B. V = 2 FIgl) (e2l + [  if Igo)(ell + A Igl) (eol kz)] + h.17 sor 5r2 + 462 (5. 6. go): represents the saturation parameter of each of the two waves A for the strongest corresponding transition (g±l e+2) Each term has a clear meaning in expression (5. We put in the following: Cr = Re[(glplgl)exp(2ikvt)]. write down the equation for j(el. to first order in kvTP = kv/r" kv/sor: Ra1 1 (13 34 240 kV ~sor 5P 2 + ar 462 /2 no. In the very lowvelocity domain (vp <<1) these equations lead to the following solutions. (5.kv + i 2 2 (5.(el. Soc. we get f = ihkf{p(e2 .1) Once all optical coherences have been calculated. As in Section 4. calculate all optical coherences in terms of the groundstate populations Hiand coherences among these states. 1 Ipigi)exp(iwt).C. of gl.c. We now need to evaluate the steadystate value of the optical coherences p(ej. Finally.10) that the atom is at the point z = 0 in its reference frame.4) Actually. Using expressions (2. because of the structure of the laser C_60 k r2 . describe the force induced by the limited photon redistribution that takes place when the atom jumps between g. through elimination of the excitedstate populations and coherences. Am.2 . CohenTannoudji Vol. Here. 5). and g1. Opt.c.(golplgo)].9) for the force. for instance. This is done again by using optical Bloch equations. (5.1) describes the coupling with the a+(cr) laser wave propagating toward z > 0 (z < 0). (5. the only nonzero Zeeman coherence in steady state is the coherence between g. 2 6 Let us. go)+ 1(eo gl) + CC(5.3)].J. (5. This is done using optical Bloch equations.Ci. (ellplel) and (eilplel) can be neglected compared with (golplgo) because of the lowpower hypothesis. (5. This elimination is done in Appendix B of this paper. ihkQ {(e. II1l.kV) P(e1 . (5.6) II4 17 where Hi = (gilpjgi). the semiclassical force is obtained from the average value of the gradient of the coupling V. and g1 by absorptionstimulated emission cycles (see Subsection 2. we get excitation. it is possible to get a new expression for the force [Eq. for example. go) + I (e gl)] + c. (5. is therefore equal to the first term of Eq. In a time interval dt.2. 6 .2) and (2.11)].3) for the laser electric field. it scatters on the average Fs+dt/2 photons from the a+ wave and rsdt/12 photons from the cr wave (see the ClebshGordan coefficients of Fig.9). we give only the main results. First we get for the force f = hk 2 [identical to Eq.2) = 2d~o/h W = L + kV. We then get in steady state p(el go) = Now we must calculate the five groundstate variables Ho. Dalibard and C. gl) + A (e.9) describe in the same way the radiation pressure force when the atom is in the states goand g1.go) = [i(6 .3) where the coefficients p are defined by P(eisl gi) = (ei. Eq. go) + i[(ellplel) + (e1IpIel)exp(2ikvt) . One is then left with a closed system of five equations involving only these five required groundstate variables.
9) for the force instead of the simplified expression (5. by neglecting all terms in kv/r and keeping only terms in kv/r'. we get in this lowvelocity domain a force We can note that these results are in agreement with the ones obtained in Section 3. 8). The result of such a calculation (which is detailed in Appendix B) is represented in Fig. 8.20) in the limit 161 >> r. (3. Soc. In the lowvelocity domain..6) and by using Eq.8hk2 v. (5.damping coefficient is four times smaller than the lin lin one. (5. Dalibard and C CohenTannoudji (5. showing that there is polarization gradient cooling.11)] for the atomic density matrix into the force. the a+a. one indeed gets for the light shift AO'of go)y the value A0 ' = f22 /36.14) any atomic velocity in the lowpower approximation.2038 J. No. Outside this range.B in the Igm)ybasis. . so that groundstate populations are obtained only from rate equations: Fig.2 Ci] (5. Am.A'. 8. Variations with velocity of the steadystate radiative force for a Jg = 1 Je = 2 transition in the a+a. a = 120  6P 2 6 17 5p + 46 hh2 . the order of 0.configuration (1 = 0. This Doppler force is calculated by neglecting all coherences among groundstate sublevels. calculated from Eqs. with a damping coefficient (for negative detunings) independent of power. which is proportional to where the force is of 6/r [Eq. even for the optimum detuning.26)]. in particular. the force appears to be close to the usual Doppler force. (5. for an atom at rest. we again find the friction force just calculated above (see. the population difference 11 II. the force is nearly equal to the Doppler force (shown by the dotted curve) calculated by neglecting all coherences between groundstate sublevels Ig)z. (5. The slope of the force near v = 0 is very high (see also inset)./3. it is easy to show that the nondiagonal density matrix obtained here in the Igm)zbasis indeed leads to the diagonal density matrix obtained in Subsection 3. Opt. As expected. is much smaller than the friction coefficient for the lin lin configuration.13). First. First we note that Eq. This new cooling force acts in the velocity range kv . we obtain f = av. B/Vol.5 r).the friction coefficient in the 0+cr configuration. the inset of Fig. 2 (5. This is done by keeping all the velocitydependent terms in expressions like (5. Second. 6. we get Furthermore. (4. Now by inserting the result [Eqs. (In this limit.13) has already been estimated in Subsection 3.) linear with velocity. which varies as r/a.) . we see here that.13) The first term of Eq. We are now able to calculate the force in this lowvelocity domain. (5.11) is equal to the one predicted in Eq.12) S = 62+ 2/ The contribution of the population term 5/6(H 1) represents 4/5 of the total result for the friction coefficient a. the remaining 1/5 part being due to 26C.B: For 161 >>r. Outside this domain. We now come to the complete calculation of the force for f= hkr so (1  .9) can be greatly simplified.25 r. 11/November 1989 g2 /2 J.B. whereas the second one describes the effect of limited photon redistribution. a = 0. This damping coefficient is maximal for a detuning 6 = r5N. We find here a result that was already mentioned in Subsection 3. represented by dotted curves.
leading to a small value of D2 and thus to low temperatures. the steps of the random walk in momentum space (due to absorption) can be several hk instead of hk. for sufficiently low power.. as we show now.)z for a time Tp = 1/rP. and D2 are of the same order. the atom is more likelyto be in the Igj)Zstate and is therefore more likely to absorb another a+ withr D.) One clearly sees that during the evolution a narrower peak appears around the zero velocity with a width of 2hk/M.(s. 18 concept of closed families of states introduced and 19: In the 0+cr configuration. D = 18 (5.C. as for a Jg = 0 .15) = hBT  h92 F 29 + 254 75 r 2 /4 62 + 1 161 300 (r 2 /4)J (5. 11/November 1989/J.18) We therefore confirm quantitatively that the new cooling force mainly acts in the kv < r' domain. Using the generalized optical Bloch equations including 8 9 recoil. Dalibard and C. (5.)zcan be considered nearly stationary on a time scale Ip. For large detunings (161 >> r).J. The first term of Eq.39)] when the detuning 6 is large compared Such an expression can be understood by simple momentumconservation arguments. The exact calculation of D is For large 161. In particular. the largestep random walk.16) (D1) corresponds to the fluctuations of the momentum carried away by fluorescence photons. occur at the frequency IA'I.18)] is obtained as in Section 4. disappear. 9(b). Opt.16)] can now be used to get the equilibrium temperature k (5.' " we studied the evolution of the atomic momentum distribution for a W atom with a linewidth hr equal to 400 times the recoil energy h2k2 /2M (analogous to sodium). If this is not the case. matrix.) 2 + 52) + 14s+s15(s+ Io = 8s+s15(s+ + s_ ) + 14s+s2 2 The result [Eq. It uses a method sug29 gested to us by Castin and Molmer and leads to D = D. the cooling limit becomes of the order. i. Soc. if 161 >>r. We use the in Refs. N is the size [(2Jg + 1) + (2Je + 1)]2 of the internal atomic density matrix. the Rabi oscillations in g are fast enough to redistribute completely the populations in g in a time p. we considered a W transition.Je = 1 transition. of the onephoton recoil energy. An argument sketched at the end of Appendix B.p). if IA'I >>r. (5. + D 2 . CohenTannoudji Vol. such a fast redistribu square tion of populations in g occurs at large detuning. We present here the principle of a treatment in which atomic motion is treated in a full quantum way. Am. C. we first need to evaluate the momentum diffusion coefficient D. p  2hk). For simplicity. and transfers between families occur only via photon rather than a cr one.e. If A'l << r. 1000points in momentum space. it is proportional to laser power and decreases as 1/6 when 161 increases.17) (5. and this can increase D 2 for a given saturation parameter by a factor as large as 10. The second term (D2) corresponds to the fluctuations of the difference between the number of photons absorbed in each wave.19) 2 [17 1 + (4 2/5 2) + This limit is smaller than the one found in the lin I lin case [expression (4. such an argument holds only if the atom remains in Ig. As a consequence. for small . At low velocities (v <<IA'). B 2039 3s. + 5s. The initial distribution is Gaussian with a rms width of 16hk/M. 9(a). consequently.. 2 By contrast. p . the momentum diffusion is due only to the first two mechanisms mentioned at the beginning of Subsection 4. the set of states 5r(p) = le 2. This peak is superimposed upon a much broader background.hk). Then it is possible to show that the total atomic density operator in steady state has nonzero matrix elements only among states belonging to the same family.. The reason that such a background remains present is because the con . sketched in Fig. (5. if 161 <<r. As 29 pointed out to us by Molmer and Castin. (5. Equilibrium Temperature in the 0+acConfiguration In order to evaluate the equilibrium temperature in this configuration. Rabi oscillations between Ig+i). separated by a splitting of the order of hA'. No. Immediately after a cycle involving the absorption of a o+ photon. this enhancement of D2 is a consequence of correlations introduced by optical pumping between the directions of two successively absorbed photons. Of course. A typical result is presented in Fig. i. D. this result is quite similar to the one obtained in the lin I lin configuration [expression (4. Since the true stationary states of the systems are the 11gm)y} states. Note that for intermediate detunings (r/2 < 161 s< 3r). Since there are no dipole forces in the 0+cr configuration. so that kBTvaries approximately as 1/63. This explains why the enhancement of D2 disappears at large 6 [see Eqs. D 2 becomes much larger than D.and Ig1).16) analogous to the one given at the end of Section 4 then leads to (for 161 > r) 170 h h P h2k2rso. Ig. a semiclassical treatment then is no longer possible. (5. which is the mean time between two successive fluorescence cycles. This greatly simplifies the study of the evolution of this density matrix: If one wants to take.. Such a condition is now kv <<IA' (see Section 3). one has to consider only a Np X 1000 vector instead of a Np X [1000]2 the populations are redistributed among the three Zeeman sublevels in g in a time shorter than r. Ig. Ig+.e. Ie 2 . By contrast.p + hk). for instance.e. the previous memory effect and. Principle of a Full Quantum Treatment It is clear from the result obtained above that. if spontaneousemission processes. instead of the real J 1 J = 2 transition. D2 remains larger than DI. i.37)]. (The standard Doppler limit kBT = hF/3 for an isotropic radiation pattern leads to a rms velocity of 12hk/M. As emphasized in the Introduction. Ieo. 6.17)]. k << =Q >>CM>>kM h22so. The validity of the result [Eq.p + 2hk)I is closed with respect to absorption and stimulatedemission processes. The only change concerns the condition that the velocities must satisfy in order to get a linear force.
so that the steadystate temperature is proportional to the laser power for polarization gradient cooling. the coefficient of proportionality between the friction force and the velocity v near v = 0. Soc. Ref. the pare expression (4. which works for Jg > 1/2. These two schemes are much more efficient than usual Doppler cooling. In the second case.. (5.37) with Eq. Q= 0. whereas. (5.26) and (4. CohenTannoudji (a) (b) :t2000r1 t= 1000 r9 1 9+1 t= 300r' t=O 50tk 0 50Mk Fig.35)with Eqs. (b) Time 2 evolution of the atomic velocity distribution for a Wtransition (6 =r.2040 J. In the first case. the steadystate temperatures are actually found to be the same (in 2/6 for large 161). (4.18)].e. Am. 6. There are therefore still warmer atoms remaining in this equilibrium state. 30). We are currently working to improve this quantum treatment to take into account real atomic transitions (Jg Je = Jg + 1). 6. Dalibard and C. although the two new mechanisms lead to a different variation with detuning for both the friction and the diffusion coefficients [compare Eqs. We showed that the cooling mechanisms are quite different in these two cases. 31). the velocitydistribution becomes non dition kv << A' for a linear force is not fulfilled for all velocity classes contributing to the equilibrium state.e. and we are studying the variations of the characteristics of the stationary state with laser power and detuning (see. These two new cooling mechanisms have a common char In conclusion. Opt. whereas it is independent of this power for Doppler cooling. we have studied in this paper two new laser acteristic: The friction coefficient a. tions including recoil. This imbalance results from an ultrasensitive motioninduced population difference appearing among the groundstate sublevels. We clearly identified two types of polarization gradient. is independent of the laser power at low Rabi frequency. which works for Jg 1. On the contrary. the first one corresponds to a gradient of ellipticity with fixed polarization axis (lin 1 Doppler cooling{ friction proportional to power [capture range independent of power Polarization gradient J friction independent of power cooling Icapture range proportional to power On the other hand. hr = 200h2k /Mas for a sodium atom). the momentum diffusion coefficient is in both cases proportional to the laser power. with a very narrow peak (HWHM _ 2 recoil velocities) superimposed upon a much broader background. where polarization gradients are certainly always present. it is independent of this power for Doppler cooling. This can be summarized a follows: cooling mechanisms that are based on laser polarization gradients and work at low laser power ( <<r). No. 11/November 1989 J. and optical pumping between these states gives rise to a Sisyphus effect analogous to the one occurring in stimulated molasses but that works here at low intensity. For simplicity. 9.. we limited our treatment to 1D molasses and to transitions Jg J = Jg + 1. The best way to check the theoretical predictions of this . the capture range of the cooling. Note here that. i.within a numerical factor [com lin configuration) and the second to a pure rotation of the polarization axis with a fixed ellipticity (+cr configuration). (a) Fictitious W atom: This atom is the simplest atomiclevel scheme leading to extra coolingin the a+ar configuration.14) and (5. B/Vol. the light shifts of the groundstate sublevels are spatially modulated. The evolution is calculated via the generalized optical Bloch equaGaussian.16)]. is now proportional to the laser power. This must be contrasted with the result for usual Doppler cooling for which a is proportional to the laser power.2 r. The initial velocity distribution is chosen to be Gaussian with a rms width of 16 recoil velocities. for example. the range of velocities over which the force is approximately linear. i. Note that such a nonGaussian momentum distribution could be at the origin of the recent doublecomponents velocity profiles measured in sodium molasses8 (see also Ref.which permit a full quantum treatment of atomic motion. and they could be responsible for the anomalously low temperatures recently observed in 3D molasses. CONCLUSION cooling is due to an imbalance between the radiation pressures exerted by the two counterpropagating laser waves. During the evolution.
CONFIGURATION In this appendix.J. with a polarization gradient) is more efficient than with parallel linear polarizations (no polarization gradient). We then determine the new groundstate energy sublevels in such a From Eqs. (A5)]. No.Al') Mg)~hAO' (A8b) . (A3) polarizations. in the moving nisms since both types of polarization gradient are usually simultaneously present in a real 3D experiment. (A7a) (A7b) hkv/1.laser configuration with Go = Go' [see Eq.gK) limits.. CohenTannoudji Vol. (3. laser waves with controllable  tor.2.)y. let us emphasize that the steadystate rms velocities that can be achieved in these new cooling schemes are very low. the atomic dynamics is due to coupling with a laser field with fixed polarization eyand to the inertial term [Eq.D) because of the vicinity of the Doppler (TD = 23. Finally. (A5) of both coolings: atoms with velocities of up to P/k are first Dopplercooled and then reach lower temperatures by polarization gradient cooling.1 )y = moving rotating frame (for a Jg = 1 ground state) and the expression of the steadystate density matrix in this energy basis. 31).B. Am. for example. B 2041 paper experimentally is to investigate the velocity dependence of the friction force in 1D molasses. These states can be easily expanded on the basis flgm)z1of eigenstates of J.. Ref. ihtdT(t)/dt]T+(t). 1 Igo). we neglect the inertial term [Eq. by The only term of the initial Hamiltonian H that changes in such a transformation is V.5b)] of the dipole moment operator D. for example. 6) is not changed to order 1 in Vrot. (2.39) and (5. One finds that the only nonzero matrix elements of Vrot= kvJz are y(g+lIVrotlgo)y = y(golVrotlg+i)y = hk/V. in addition to T(t)H(t)T+(t) in the Hamiltonian H' governing the time evolution in the new representation. its components satisfy the wellknown commutation relations with J.4)]. (A4) full quantum treatment is certainly required in order to analyze them (see Subsection 5.5b) V Dey =D. Hamiltonian in the Moving Rotating Frame In the laboratory frame and in the a+o. Soc. we establish the expression of the new Hamiltonian that governs the atomic evolution in the moving rotating frame introduced in Subsection 3. One gets quantum approach of both internal and external atomic degrees of freedom is required. [J. (2.the energy diagram in the ground state (see Fig. the wave functions are changed to order 1 and become I Ig. preliminary results have already 23P2 transition of helium. Figs. To extend our results to real 3D molasses.1). 11/November 1989/J.B. Dalibard and C. To sum up. i I (A6a) IgO)y = 1t Ig+1) + 1 Ig l). According to Eq. (A5)]. On the other hand. The fact that T(t) is time dependent introduces a new term.uK)and recoil (TR = 4. (A2). Actually. one can then calculate the matrix elements of the inertial term [Eq. For a+o. a full rotating frame. [Jr. (Al) manifold Ig±. but these preliminary results already appear to be quite promising (see. The basis of such an ap proach and preliminary results are presented at the end of Section 5..32 They clearly show that coolingwith orthogonal linear polarizations (i.14b)]. (A5) [see expression (3. we just replace z with vt. since D is a vector opera studying the transverse collimation of an atomic beam crossing two counterpropagating been obtained on the 23S. In the atomicrest frame. (A6). Opt. y(g1IVrotgO)y = y(b'oIVrotIg. it does not seem easy to single out one of the two new cooling mecha which is nothing but Eqs.13). GroundState Energy Sublevels in the Moving Rotating Frame If.. and we get from Eq.the energy sublevels in the ground state are the eigenstates Igm)y of Jy (see Subsection 3. the laseratom coupling V is Since Vot has no diagonal elements in Igo)y and in the proportional to the component along Ey[givenby Eq. Dy] = ihD. but the velocity profiles are far from Gaussian. For such low velocities. A from which it is easy to show that T(t) [DXsin kt + Dy cos kt] T+(t) = Dy. one should first note that Doppler coolingis still present in these results (see. we have ih[dT(t) JT+(t) = hoJ_. sin kot +DY cos hot.which proves that T(t) is the unitary operator associated with the transformation to the moving rotating frame. 7 and 8). Therefore one takes advantage Thus it is clear that the transform of V by T(t) describes the coupling of D with a laser field keeping a fixed polarization ey. Ig. We are currently extending this approach to more realistic situations. since they can approach the recoil velocity [see expressions (4. D] = ihDy. (A5)] between the Igm)yI states that are necessary for a perturbative treatment of the effect of the term given by Eq. the residual kinetic energy remains of the order of the Doppler limit.1)Y =+ 1 2 1g~l) + .19)]. Transverse temperatures below the Doppler limit TDwere measured in the lin lin configuration. 6. (2.e. for example. (A2) F = Igo)y + Ig+ ) y(g+lrotAlo)y ig l~y g i rotlgo)y + . (A6b) APPENDIX A: INTERNAL ATOMIC STATE IN THE MOVING ROTATING FRAME FOR THE 9+T.)~ + I 0 ) y(goIVrotIl'±i)y (Aa) The transformation to the moving rotating frame is then achieved by applying a unitary transformation T(t) = exp(ikvtJl/h). Now. On the other hand. in a first step.
1.1 + Cr + ie). e2) =  2 11. we introduce the followingnotation: p(e2 . we used y(g+1IVrotlg_)y = 0. (3. go) [Eq. which is the population of the state e2.. Dalibard and C. Ho H p(e. g. Am.kv)s+]/6r.16). the steadystate density matrix PStis diagonal in the JIgm)y) basis and has the same diagonal elements as those calculated in Subsection 3. Take. Calculation of ExcitedState Populations and Coherences We now calculate the expressions of excitedstate populations and coherences in terms of groundstate variables.2 ) = 2 "1 (B6) r = 6 kv +i .6)] and p(e_. (5. (kv/ A')(r'/A') or higher. 11/November 1989 J.e2) by their expression in terms of groundstate variables [cf. we calculate the force in the atomic reference frame. and second a closed system of five equations dealing only with groundstate variables. + Cr ici) (H. and the ones that can be deduced from Eqs.3). the evolution of p(e2. (3. 2 (B4) 2 transition. Putting the results [Eqs. (5. P~~~~~e2. As explained at the beginning of Subsection 5. This system is finally solved numerically to get the value of the steadystate force. 2 p(eo. one sees that there will be source terms of order 1 in kv/A' in the 12 (II. Inserting Eqs.. (B3) by complex conjugation. (e2. go) = no. We first define (B2) A(el.g0 ) = 22 The same procedure allows one to calculate excitedstate coherences.15) and (3. + iii) P(e2 . (3.16). + T IH1 + vlCr + lCi. p(eo. The only nonzero matrix elements of p. (5. CohenTannoudji In Eq. H V = (s+ + s)/12. e. so that excitedstate populations remain small compared with groundstate populations. e 2) = 0]. e2) + i . el) = S.CONFIGURATION The purpose of this appendix is to outline the calculations of the radiative force and of the momentum diffusion coefficient in the  +ar configuration for the case of a Jg = 1 Je b(e2. Eqs.1) = 26. if we neglect all terms of order (kV/A') . one immediately gets Eq.(gl. and we replace the optical coherences p(e2. 2 r _ The coherences p(ei. The corresponding optical Bloch equation is APPENDIX B: Ig = 1 1 = 2 TRANSITION IN THE +a. (B7) written as p(el.[P(e 2 . This gives first the expression of the force only in terms of groundstate variables (densitymatrix elements in g).) = g1 263 (Cr iCe) ~~(e2 (B3) the other hand. ~(e. Note that several optical coherences. This is no longer the case in the {Igm)ybasis.. (A8b).B. g. P(e2. Opt. e.configuration. (5. to order 1 in kv/A'. for example. are thus y(golplgai)y On ~(e 2.lpg_l)y. 2 Consequently.) = p(eo. Calculation of Optical Coherences For simplicity. eo) = T II.2042 J.10).1) and the radiative force by Eq. which proves Eqs.B. This gives The first part of the calculation consists of calculating the optical coherences and the excitedstate populations and coherences in terms of the groundstate populations and coherences.in the moving rotating frame). (Cr + iCd). We proceed in the same way for the other excitedstate populations p(el.15) and (3. one calculates the other optical coherences (coherence between el and e. such as p(eOgo). e2) = rp(e 2 .have a zero steadystate value in this a+a.3)]. in which the atomlaser coupling is given by Eq..gl) = 26 III. g1) .1 UV (6 .3) for the force.)y will be reduced by an extra nonsecular term of the order of r/A. In a similar way. B/Vol.. gl) =  II1. with a damping rate of the order of r.) = (elIple_)exp(2ikvt) (B8) H0 n. No. and since the opticalpumping source term is static (at frequency 0). g.9). (B2) and (B3)] into expression (5..Al'). (B5) where s+ is defined in Eq. since the evolution frequencies of these coherences are (A' . (B1) where we have put Al = [(6 + k)s  63+ d 3kv + i . Cr. II+.17). SteadyState Density Matrix in the Energy Basis l~g)yJ The terms describing optical pumping in the master equation are diagonal in the lIgm)yJ basis. in which the force is expressed only in terms of Ho. g1)and P(g. Soc. (A7) into (A8) gives Eqs.2 . which has the followingequation of motion: . From Eqs. (5.) =  fl p(e. We take the steadystate value [(e2. 6. (B. go) can th n . The calculation is done in the lowpower limit ( <<r). the steadystate value of y(go plgt.g1) = 2 269 equation of motion ofy(go g)y and of order 2 in kv/A' in the equations of motion of the populations y(gmIPgm)y and of the coherence y(g+. e 2).and C. e. e 2)]. = 26 (C.1.
el) +   6 p(eo. and C for any value of the atomic velocity.J. Fig. e1) + (B12) rt.el. e..g1) = 2iko(gl.)exp(2ikt) ))exp(2ikt). (B15)(B19)] allows one to calculate numerically the five quantities 1+1. p )H (B13c) + (v2 s+ a 4r ) where $5 = ) & =Vs5  + V5Cr+ A 5C. Taking the real part and the imaginary part of the equation p (gl. we obtain from 11I = 0 + . eo) + A(eo.g. the five previous equations can be simplified by neglecting all terms in kv/r and keeping only terms in kvPSo. B 2043 (el. = Cr + iCi. (B2). Cr. (B9) 2Tt = sE1 +53 snI 5 s+IIl + (21S+)Cr + 2(. one then gets the value of the radiative force for any atomic velocity in the lowpower regime (see.2 ) = ((e2Ipleo) + (e 0 Iplewhich is found to be p(e2 . (B10) For the population IIO. = 0 and replace excited populations and optical coherences by their value in terms of groundstate variables. 2 V1 (B19c) ft. (B18a) X(gl. eo) + p(eo.e)].(6 .2 1kv #2 2 r' + 4k2VI 3r Air + 2vlkv 22 P+4h1o We also need for the following the quantity A(e2 . e2)] g.93. Inserting The first line describes how the state gl is fed by spontaneous emission from the three excited states e2.. eo) . We now write down the evolution equation of the groundstate coherence in the moving rotating frame: P(gl. e. g )] 1 + i [p(g1. In steady state. Am. eo)  + i . (B15) SO. We again replace the excitedstate coherences and the optical coherences by their steadystate values. e2) .[(el.[P(gl.g0' with 2 'fI3 2 =4 [r(vl + V) r+ 4a22 2 2 + 4h o  2kv(zll + $3)]. we have for III .2 ) = 2 since excitedstate populations are negligibleat this order in (Bll) 1. For example. e. e 2) 2 (e2 . 8).(e2. We obtain in this way the result into Eq.c1)= (r + 2ikv)(el.[p(e2.S)Cr + 2($ 6 put s+= S_ = +ho S)Ci. (B20) so that the previous set becomes . We then ° s+H 6 sII + 32 0 + 6 s+IIl + (2v.we just use 1 = Hr + II0 + II. e2) +  2 p(el. we can put II. CohenTannoudji Vol. (B17) 3r v 1 r . 11/November 1989/J. (B14) 3 + 3 A + 93 + IV4 2 'J2 P5 (B19d) The set formed by the five equations [Eqs. g) .. [A(gl.)] eo). = rp(e2 . 6. 6 +ho 3 )II + 2][10~ o0= V2+ _a \2. and the second line describes absorption and stimulatedemission processes. we get 2 2 + 2V2 3) + 2 [2kv(vl + V 2 + 4hko r(Al + $)] (B13b) 0= (2  S)II + 2 $210+ (2  ][1+ $5Cr VAi (B19a) and $3 = [(6 + 3kV)S3 . Dalibard and C. p(el.(go.3kv)s3+]/6r.1 ) + (4 + jV 4 )(Cr + i.9). g1) = 0. V3= (S3+ + S30)/12. (B13a) + iCe)I + i 2 [(gl. eo) + p(eo. gl)]. e) In a similar way.1 + H. (B19b) Calculation of GroundState Populations and Coherences We now write the equations of motion of the groundstate populations. e.g1) = (glIplg.l + kv S+)Ci (B16) The steadystate value for p(el. using Eqs. H1 0 . No. (5. e) = ( where we have put 2 + i 2)H0. Opt.A(eo. g1) + 2 p(el. e1) is then.p(eo. (B18b) 2 (2 + iV 2 )(H.. In the very lowvelocity domain (ho <<1/Tp = rso). Soc.
(B16)(B17)]. p) = dp' 8 (1 + h2k2) 1 = H. we can get a closed set of equations for these variables by Calculation of the Atomic Momentum Diffusion Coefficient In order to calculate the momentum diffusion coefficient.9) for the force. plpleo.plplei. p +p'). (5.15).p +p'). 1 Io sIl + 9 s_ . i2 elimination of the optical coherences and the excitedstate populations and coherences.which gives the simplified set 5sHl + 9s+IIo+ s+II. 3 C 6 C+2C r 62 ci. One can then neglect Cr and C in the three equations [Eqs.29 This method is well adapted to the present o+a. First. n=0. Opt. average over the momentum carried away by the fluorescence photon. and we integrate over p. we use a method introduced by Castin and Milmer. P~)+ I (eo. p) means an 11 36k 17 1 + (46 2 /5r 2 )=(H) I (pI1). 2.5 rs 4 Ci.IH1(p + hk) 6 H(P) +1 0(p) + 36 .2044 J. 8. It consists of writing down the generalized optical Bloch equations including recoil.p p) 3 +2 (e_1.plple1. we multiply the generalized optical Bloch equations (with v = p/M = 0) by p 2 . Then it is straightforward to check that the quantities given in Section 5 [Eqs.: (el. we get (pCi) = dppCi(p) = 2r (PI1).C(p+hk) 4 1 6 Cr(P) 3 3r Ci(P) Cp)= 2s{ [II(p). No. 6. Here we just give the result of this elimination: 2  I(p) + [o(phk) + (p + hk)] Ho(p) + 21 II(P) + g Cr(P)} 36 JI. the atom in the ground state g with momentum p: (go.plple. In order to get d(p 2)/dt. Inserting the corresponding values into expression (5. owing to the conservation of angular momentum. i. p . + o + . +1. Since we are interested here in the momentum diffusion coefficient. (B23) Now..hk)]exp(iWLt) + c. B/Vol.p . (Cr+ iCi)(p)= (g.p + hk) If one integrates these equations overp. The solution of this set is given in Eqs.5s+Il. which has been plotted in dotted curves in Fig. so that we just give the main results here. and g1 becomes negligible. The single or double bar over (ei. via spontaneousemission processes. This amounts to considering an atom with zero velocity but still exchanging momentum with the laser field.c.p + nhk)j. ) pc. 1=Irl 3 6 I. Am. we take the limit of infinite atomic mass. l/November 1989 J.11) for zero velocity. pplel.. 6 II 1n nJ_5 C 6hoCi 7r: (eo. either o± polarized or r polarized: 6 2 0=IH. CohenTannoudji ° = II + n + 6 1 II2 0 6 ~H. one just recoversthe stationary values deduced from Eqs.P) = 2 (el. 1.. so that (p 2 ) increases linearly with time as 2Dt. (5. Dalibard and C.= 1. for velocities such that kv >>1P. (B25) lowing notation: IIm(p) = (gm. p + hklpIgl.= 0. plplel. (B22) m=0.= 0. 18 and 19 for a Jg = 0 Je = 1 transition or Jg = 1 Je = 0 transition. multiplying these equations by p and integrating over p. Ig.laser configuration with a low laser power. and the last two lines describe departure due to absorption. II. We start with the equation of motion for the population of Igo. (B21) X (eo. p + p'lplel.p +plpleo. the coherence between the two ground states g.II ()J + 6 [Ci(phk)+ Ci(p +hk)]7 Ci(P)} 66 (B26) + 2 [(g0. we recover the usual Doppler force.11)] are indeed solutions of this simplified set. + (go. the followingfamily: lemp + mhk).6ho r0Cr .hk). On the other hand. p).plplgo. (5. f dP 4hk h2k2 O= 46II1 46 4r 41' II1 . + H0 + H1. Soc. (27 (B27) .plpleop) X (el. pIPlei.e. p + mhk).J1 remains globally invariant in absorption and stimulatedemission processes. This is done in detail in Refs..(p) 111(p hk) + . Transfers among families occur only Taking now the folm = 0.. The first two lines describe feeding by spontaneous emission.+H r. (B24) We can in the same way rewrite all the other optical Bloch equations.p + mhklplgn.IIl( + hk)+18 .
Phys. Metcalf. Nagourney. (World Scientific. R. Javanainen and S. S. Lett. 14. (SpringerVerlag. (4. Weiss. K. Soc. Gordon and A. CohenTannoudji. J.1661 (1985). 12. and C. Then. D. Vansteenkiste for many helpful discussions and remarks about this paper. (World Scientific. J. in Proceedings of the Ninth International Conference on Laser Spectroscopy. 22. Salomon. K. 28. Haroche. Arimondo. Shevy. Sesko. Fan. S. 5. 2072 (1989).1). and C. (NorthHolland. (B30). Haroche. (B28) Vol. J. Hollberg. Singapore. 2112 (1989). W. A 20. 1118 (1989). CohenTannoudji. 21. and A. F75231 Paris Cedex 05. J. J. (5. Stenholm. eds. and C. Ungar. A. 13. (2. Bull. A. Am. 21. Chu. Dalibard and C. Bambini. S. 3. Am. Opt. G. Phys. and C.68 (1975). 24 Rue Lhomond. A. Ungar. Gay. 1606 (1980). Opt. Hansch and A. Soc. J. 1989). 31. in Proceedings of the Confer ence on Spin Polarized Quantum Systems. They are also grateful to W. J. Y. No. Phys. Rev. Phys. B 6. Phys. Am. H. Opt. Trager. J. J. Kaiser. CohenTannoudji. C. S. F. P. eds. they acknowledge stimulating discussions with S. 7. Owing to symmetry around p = 0. R.16) can easily be deduced from this calculation. D. Janik. E.)). Opt. Feld. D. diffusion. and T. C.J. 27. (B23) the kernel describing the spontaneousemission pattern by just a 6(p') function. Schawlow. DupontRoc. CohenTannoudji and J. Other consequences of long atomic pumping times are described in W. A. A. Phys. ed. H. Salomon. Commun. and H. Salomon. Weiss. REFERENCES AND NOTES 1. 28A. and J. CohenTannoudji. Am. Heidmann. (SpringerVerlag. Soc. p. by difference between D and D2 ACKNOWLEDGMENTS The authors warmly thank all their colleagues from the Ecole Normale Superieure laboratory and A. and C. Grynberg. J. C. and J. and M. E. C. and G. Arimondo. Soc. B 2. Y. G. mechanisms overcoming the Doppler limit were also proposed. and C. Kaiser. Phys. 699 (1986). in Proceedings of the We now multiply Eqs. eds. Rev. Lombardi. N. Bull. S.Aspect. V. Balian. 11/November 1989/J. 191 (1967). N. M. Chu. in Frontiers in Laser Spectroscopy. Opt. Vansteenkiste. 6.) = (p 2 1ll). Watts. S. and N. P. 24. S. Dalibard. Wallis. Rev. R. J. we cancel out this cause of 13. B (to be published). Kaiser. Y. 25. Phys. 73. CohenTannoudji.. Gawlik. R. J. (Paris) 10. Westbrook. E. 61. B 5. J. Lett. one has (p fl. C. E. A. E. Weiss. Stenholm. 18. J. 32. S. Vansteenkiste. CohenTannoudji. Shevy. Ashkin. S. Harsche. 1989). For example. Y. 4. 631 (1969). J. 6. Soc. 57. Rev. and S. 26. Y.1707 (1985). Phillips. C. Salomon. B 3. E. D. R. Am. Aspect. 1977). Sci. 637 (1975). Berlin. W. A. we get an expression similar to Eq. Thomas. Liberman. Castin. CohenTannoudji. 761 (1985). 612 (1985). J. C. The two contributions D. Phys. 4463 (1964). R. C. A 21. Salomon. L. D. Dalibard and C. Lett. Phys. Rev. Dalibard. Rev. Ungar. Itano. J. 1225 (1988). Heidmann. 169 (1988). S. 9. Shevy. Universit6 Paris VI. Acad. . 19. D. 1989). Arimondo. J. We restrict ourselves here to neutral atoms. Cable. Y. Am. This represents the contribution of the fluctuations of the difference among the numbers of photons absorbed in each wave (D2 term). Phys. Castin. G. Am. 1688 (1986). Soc. A 24. 58. Kowalski. (World Scientific. A. and S. Finally. and N. 184 (1969). C. in Proceedings of dt 2 (B29) the 11th Conference on Atomic Physics. 638 (1969). Sci.997 (1987). 30. and C. P. Dalibard. Phys. represents the fluctuations of the momentum carried away by the fluorescence photons. Grynberg. Harsche. d (p 2 ) = h2 krso 72 1 + 8 (B30) which represents twice the total momentum diffusion coefficient D. Hansch last summer. in Fundamentals of Quantum Optics II. we obtain from Eqs. J. R. Am. Phys. Vollmer. Phys. and S. 15. Phillips and his group for their comments on this work and communicating their experimental results previous to publication. 55. Chu. Castin. Phys. where 58/85 is replaced by 8/17. Mooradian. J. Acad. and G. Dalibard. Ashkin. repeating the same calculation again. Metcalf. Rev. (B26) by p2 and integrate over p in order to get the rate of variation of (p2 ): d ( dt 2 2 2 H. France (personal communication). Lett. and C. 1989). Gay.1521 (1979). A. 35 (1980). Ecole Normale Superieure. 30. ) = d ((p 2 H 1 ) + (p Ho) + (p 11th Conference on Atomic Physics. S. Castin. Rev. Aspect. After some calculations. Wineland and W. 196. 20. Gould. J. K. 48 (1985). P. B 6. Laboratoire de Spectroscopic Hertzienne.898 (1981). Phys. and W. CohenTannoudji. Note that for trapped ions. D. J. E. J. D. B 17. and P. Y. Reynaud. Stringari. Wineland and H. Javanainen. CohenTannoudji. Ehlotzky. Am. Dalibard. Weiss. Metcalf. Opt. Riis. Commun. Bjorkholm. Opt. Am H. T.4577 (1984). M. 23. Van steenkiste. S. Phys. 265. Soc. Dehmelt. If we replace in Eq. DupontRoc. P. 2.7) is inserted into Eq. Appl. We then get D. ed. Chu. sterdam. Molmer. Opt. We also reincluded in the definition of the excited states the phase factors exp(+i7r/4) that appear when Eq. B 20. S. CohenTannoudji. C. F. B 6. Ann. J. Kaiser. 826 (1988). Dalibard. Lindberg and J. M. 30. Phys. Phys. and S. A. 1008 (1986). J. Lehmann and C. Toschek. R. Molmer. E. Minogin. 1. 17. Aspect. 10. Rev. C. C.C. Aspect. 29. B 2045 8. C. 61. Lett. They involve Raman twophoton processes: H. J. CohenTannoudji. Shevy. Dalibard and C. Lett. Aspect. and S. Singapore. J. Rev. B 18. 20. CohenTannoudji whereas for symmetry reasons one has (pH1 0 ) = (PC) = 0. Soc. 2046 (1989). W. and D2 appearing in Eq. 16. S. Wieman. Opt. J. 62. Phys.1 (1981). Arimondo. Letokhov and V. Arimondo. Berlin 1987). Lett. P. 258. Mod. Singapore. D. Molmer and Y. Dehmelt. (B26) and (B27) 11. eds. Soc. Chu. J. E. Stenholm. Chu.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.