Atom traps

Light forces

It is well known that a light beam carries momentum and that the scattering of light by an object produces a force. This property of light was first demonstrated by Frish [139] through the observation of a very small transverse deflection (3 × 10–5 rad) in a sodium atomic beam exposed to light from a resonance lamp. With the invention of the laser, it became easier to observe effects of this kind because the strength of the force is greatly enhanced by the use of intense and highly directional light fields, as demonstrated by Ashkin [20, 21] with the manipulation of transparent dielectric spheres suspended in water. The results obtained by Frish and Ashkin raised the possibility of using light forces to control the motion of neutral atoms. Although the understanding of light forces acting on atoms was already established by the end of the 1970s, unambiguous demonstration of atom cooling and trapping was not accomplished before the mid 1980s. In this section we discuss some fundamental aspects of light forces and schemes employed to cool and trap neutral atoms.

The light force exerted on an atom can be of two types: a dissipative, spontaneous force and a conservative, dipole force. The spontaneous force arises from the impulse experienced by an atom when it absorbs or emits a quantum of photon momentum.

An exoergic collision converts internal atomic energy to kinetic energy of the colliding species. When there is only one species in the trap (the usual case) this kinetic energy is equally divided between the two partners. If the net gain in kinetic energy exceeds the trapping potential or the ability of the trap to recapture, the atoms escape; and the exoergic collision leads to trap loss.

Of the several trapping possibilities described in Chapter 3, by far the most popular choice for collision studies has been the magneto-optical trap (MOT). A MOT uses spatially dependent resonant scattering to cool and confine atoms. If these atoms also absorb the trapping light at the initial stage of a binary collision and approach each other on an excited molecular potential, then during the time of approach the colliding partners can undergo a fine-structure-changing collision (FCC) or relax to the ground state by spontaneously emitting a photon. In either case electronic energy of the quasimolecule converts to nuclear kinetic energy. If both atoms are in their electronic ground states from the beginning to the end of the collision, only elastic and hyperfine changing collisions (HCC) can take place. Elastic collisions (identical scattering entrance and exit states) are not exoergic but figure importantly in the production of Bose–Einstein condensates (BEC). At the very lowest energies only s-waves contribute to the elastic scattering, and in this regime the collisional interaction is characterized by the scattering length.

Most of this review has focused on collisions of cold, trapped atoms in a light field. Understanding such collisions is clearly a significant issue for atoms trapped by optical methods, and historically this subject has received much attention by the laser cooling community. However, there is also great interest in ground-state collisions of cold neutral atoms in the absence of light. Most of the early interest in this area was in the context of the cryogenic hydrogen maser or the attempt to achieve Bose–Einstein condensation (BEC) of trapped doubly spin-polarized hydrogen. More recently the interest has turned to new areas such as pressure shifts in atomic clocks or the achievement of BEC in alkali systems. The actual realization of BEC in 87Rb [15], 23Na [103], 7Li [56, 57], 4He* [310, 330] and H [138] has given a tremendous impetus to the study of collisions in the ultracold regime. Collisions are important to all aspects of condensates and condensate dynamics. The process of evaporative cooling which leads to condensate formation relies on elastic collisions to thermalize the atoms. The highly successful mean field theory of condensates depends on the sign and magnitude of the s-wave scattering length to parameterize the atom interaction energy that determines the mean field wavefunction. The success of evaporative cooling, and having a reasonably long lifetime of the condensate, depend on having sufficiently small inelastic collision rates that remove trapped atoms through destructive processes.

Introduction

If, while approaching on an unbound ground-state potential, two atoms absorb a photon and couple to an excited bound molecular state, they are said to undergo photoassociation. Figure 5.1 illustrates the process. At long range electrostatic dispersion forces give rise to the ground-state molecular potential varying as C6/R6. If the two atoms are homonuclear, then a resonant dipole–dipole interaction sets up ±C3/R3 excited-state repulsive and attractive potentials. Figure 5.2 shows the actual long-range excited potential curves for the sodium dimer, originating from the 2S½ + 2P3/2 and 2S½ + 2P½ separated atom states. For cold and ultracold photoassociation processes the long-range attractive potentials play the key role; the repulsive potentials figure importantly in optical shielding and suppression, the subject of Chapter 6. In the presence of a photon with frequency ωp the colliding pair with kinetic energy kBT couples from the ground-state to the attractive molecular state in a free–bound transition near the Condon point RC, the point at which the difference potential just matches ћωp.

Scanning the probe laser ωp excites population of vibration–rotation states in the excited bound potential and generates a free–bound spectrum. This general class of measurements is called photoassociative spectroscopy (PAS) and can be observed in several different ways. The observation may consist of bound-state decay by spontaneous emission, most probably as the nuclei move slowly around the outer turning point, to some distribution of continuum states on the ground potential controlled by bound–free nuclear Franck–Condon overlap factors.

In the 1980s the first successful experiments [312] and theory [98], demonstrating that light could be used to cool and confine atoms to submillikelvin temperatures, opened several exciting new chapters in atomic, molecular, and optical (AMO) physics. Atom interferometry [6, 8], matter–wave holography [294], optical lattices [192], and Bose–Einstein condensation in dilute gases [18, 95] all exemplified significant new physics where collisions between atoms cooled with light play a pivotal role. The nature of these collisions has become the subject of intensive study not only because of their importance to these new areas of AMO physics but also because their investigation has led to new insights into how cold collision spectroscopy can lead to precision measurements of atomic and molecular parameters and how radiation fields can manipulate the outcome of a collision itself. As a general orientation Fig. 1.1 shows how a typical atomic de Broglie wavelength varies with temperature and where various physical phenomena situate along the scale. With de Broglie wavelengths on the order of a few thousandths of a nanometer, conventional gas-phase chemistry can usually be interpreted as the interaction of classical nuclear point particles moving along potential surfaces defined by their associated electronic charge distribution. At one time liquid helium was thought to define a regime of cryogenic physics, but it is clear from Fig. 1.1 that optical and evaporative cooling have created “cryogenic” environments below liquid helium by many orders of magnitude.

Cold and ultracold collisions occupy a strategic position at the intersection of several powerful themes of current research in chemical physics, in atomic, molecular and optical physics, and even in condensed matter. The nature of these collisions has important consequences for optical manipulation of inelastic and reactive processes, precision measurement of molecular and atomic properties, matter–wave coherence and quantum-statistical condensates of dilute, weakly interacting atoms. This crucial position explains the wide interest and explosive growth of the field since its inception in 1987. Obviously due to continuing rapid developments the very latest new results cannot appear in book form, but the field is sufficiently mature that a fairly comprehensive account of the principal research themes can now be undertaken. The hope is that this account will prove useful to newcomers seeking a point of entry and as a reference for those already initiated.

After a general introduction and a brief review of the elements of scattering theory in Chapters 1 and 2, the next four chapters treat collisions taking place in the presence of one or more light fields. The reason for this is simply historical. After the development of the physics of optical cooling and trapping from the early to mid 1980s, the first generation of collisions experiments applied this light-force physics to cool and confine atoms in traps and beams.

Introduction

Photoassociation uses optical fields to produce bound molecules from free atoms. Optical fields can also prevent atoms from closely approaching, thereby shielding them from shortrange inelastic or reactive interactions and suppressing the rates of these processes. Recently several groups have demonstrated shielding and suppression by shining an optical field on a cold atom sample. Figure 6.1(a) shows how a simple semiclassical picture can be used to interpret the shielding effect as the rerouting of a ground-state entrance channel scattering flux to an excited repulsive curve at an internuclear distance localized around a Condon point. An optical field, blue detuned with respect to the asymptotic atomic transition, resonantly couples the ground and excited states. In the cold and ultracold regime particles approach on the ground state with very little kinetic energy. Excitation to the repulsive state effectively halts their approach in the immediate vicinity of the Condon point, and the scattering flux then exits either on the repulsive excited state or on the ground state. Figure 6.1(b) shows how this picture can be represented as a Landau–Zener (LZ) avoided crossing of field-dressed potentials. As the blue-detuned suppressor laser intensity increases, the avoided crossing gap around the Condon point widens, and the semiclassical particle moves through the optical coupling region adiabatically. The flux effectively enters and exits on the ground state, and the collision becomes elastic.