Abstract We present a series of recent experimental and numerical results which allow us to propose and explain the mechanisms for collisional cooling of a trapped ion by sequential binary collisions with atoms. Our experiments in hybrid atom-ion traps study the cooling of ions when the reservoir of the coolant atoms is spatially localized at the centre of the ion trap. Under these conditions the widely and long held expectation that for the ion to collisionally cool, the atom must be of lighter mass (i.e. mI > mA ) is overturned. Instead we show ion cooling for mI ⩽ mA in addition to mI > mA , and explain why earlier work establishing mass ratios does not apply to the experiments in hybrid traps. Further, for cooling of an ion A+ by the parent atom A, the mechanism of resonant charge exchange (RCE) allows for extremely efficient cooling of the ion A+. This mechanism is demonstrated by comparing the cooling rate of an ion with a localized ensemble of its parent atoms and with a lighter atomic species, where in the latter case the RCE mechanism does not exist. The difference in measured cooling rates for a given number density of the atoms and the theoretically calculated difference between elastic and RCE cross sections are used to show that the cooling efficiency per collision with RCE is much greater than the cooling by elastic collision. We conclude with some perspectives and prospects of future experiments.
Abstract We present a series of recent experimental and numerical results which allow us to propose and explain the mechanisms for collisional cooling of a trapped ion by sequential binary collisions with atoms. Our experiments in hybrid atom-ion traps study the cooling of ions when the reservoir of the coolant atoms is spatially localized at the centre of the ion trap. Under these conditions the widely and long held expectation that for the ion to collisionally cool, the atom must be of lighter mass (i.e. mI > mA ) is overturned. Instead we show ion cooling for mI ⩽ mA in addition to mI > mA , and explain why earlier work establishing mass ratios does not apply to the experiments in hybrid traps. Further, for cooling of an ion A+ by the parent atom A, the mechanism of resonant charge exchange (RCE) allows for extremely efficient cooling of the ion A+. This mechanism is demonstrated by comparing the cooling rate of an ion with a localized ensemble of its parent atoms and with a lighter atomic species, where in the latter case the RCE mechanism does not exist. The difference in measured cooling rates for a given number density of the atoms and the theoretically calculated difference between elastic and RCE cross sections are used to show that the cooling efficiency per collision with RCE is much greater than the cooling by elastic collision. We conclude with some perspectives and prospects of future experiments.
The two most widely used ion cooling methods are laser cooling and sympathetic cooling by elastic collisions (ECs). Here, we demonstrate another method of cooling ions that is based on resonant charge exchange (RCE) between the trapped ion and the ultracold parent atom. Specifically, trapped Cs+ ions are cooled by collisions with cotrapped, ultracold Cs atoms and, separately, by collisions with cotrapped, ultracold Rb atoms. We observe that the cooling of Cs+ ions by Cs atoms is more efficient than the cooling of Cs+ ions by Rb atoms. This signals the presence of a cooling mechanism apart from the elastic ion-atom collision channel for the Cs-Cs+ case, which is cooling by RCE. The efficiency of cooling by RCE is experimentally determined and the per-collision cooling is found to be two orders of magnitude higher than cooling by EC. The result provides the experimental basis for future studies on charge transport by electron hopping in atom-ion hybrid systems.
We experimentally demonstrate cooling of trapped ions by collisions with cotrapped, higher-mass neutral atoms. It is shown that the lighter ^{39}K^{+} ions, created by ionizing ^{39}K atoms in a magneto-optical trap (MOT), when trapped in an ion trap and subsequently allowed to cool by collisions with ultracold, heavier ^{85}Rb atoms in a MOT, exhibit a longer trap lifetime than without the localized ^{85}Rb MOT atoms. A similar cooling of trapped ^{85}Rb^{+} ions by ultracold ^{133}Cs atoms in a MOT is also demonstrated in a different experimental configuration to validate this mechanism of ion cooling by localized and centered ultracold neutral atoms. Our results suggest that the cooling of ions by localized cold atoms holds for any mass ratio, thereby enabling studies on a wider class of atom-ion systems irrespective of their masses.
This work presents various novel results in the field of experimental and theoretical trapped ion studies. All investigations involve laser-cooled 138Ba+ ions confined in a linear quadrupole ion trap, which serve as coolant for complex molecular ions, as target for collisions with neutral atoms, or as model system in theoretical analyses. The special feature of the apparatus used in this work, is a molecular ion source based on electrospray ionization that allows for the production of gas-phase molecular ions of almost arbitrary species with a maximum mass-to-charge ratio of 2,000 Da. Trapped together with laser-cooled 138Ba+ ions, the molecular ions can be cooled down, kept cold and investigated for many minutes and in principle for up to hours in a nearly collisionless environment.
In this work, the sympathetic cooling of complex molecular ions to translational temperatures below 1 K is demonstrated for different types of molecules covering a mass range from 182 to 12,400 Da. Using the example of the protein cytochrome c, with a mass of about 12,400 Da the most massive molecular species sympathetically cooled in an ion trap so far, the methods for detection of successful trapping and cooling, and temperature determination are presented in detail. In one case, multiply protonated molecules of the protein cytochrome c have been cooled to less than 0.75 K. Furthermore, two methods for measuring photodestruction rates of cold, trapped complex molecular ions have been developed that can, for example, be applied in destructive spectroscopy schemes. Both techniques are demonstrated using the example of singly protonated glycyrrhetinic acid molecules (GAH+) dissociated by an ultraviolet (UV) laser. Measurements of the photodestruction rate were performed at different intensities of the UV laser, and rates as low as 0.05 s-1 and less have been determined, which is only possible due to the long ion storage times that can be achieved with this apparatus. For the UV laser wavelength of 266 nm, a destruction cross section of (1.1±0.1)∙10E-17 cm2 for GAH+ has been determined.
In this work, feasibility studies towards internal cooling of trapped complex molecular ions have been performed. Such an additional internal cooling could be realized via collisions with laser-cooled neutral atoms. To test their suitability, neutral atoms of Li, Na, K, Rb, Cs, and Yb have been collided with trapped laser-cooled and not cooled Ba+ ions as well as with complex molecular ions. For all species charge exchange reactions with Ba+ ions have been observed excluding these species for collisional cooling. While K, Rb, and Cs showed charge exchange reactions with ground state and laser-excited Ba+ ions, the charge exchange rates of Yb and Ba+ were found to be state dependent. Collisions of neutral Yb atoms were found to be reactive and lead to the formation of adducts and fragments, in contrast to the alkali metals for which no such reactions have been observed. In addition, the motion of cooled, trapped ions has been studied theoretically via simulations. When the translational energy of a trapped ion ensemble is sufficiently reduced, a first-order phase transition from a liquid to a solid state occurs with the ion ensemble changing from a disordered ion cloud to an ordered ion crystal. As ions in a linear quadrupole trap arrange in shell structures, a free motion in radial direction is disabled, and ions can only move to neighboring shells when their kinetic energy is high enough to overcome a certain potential barrier. The simulations have shown that intershell diffusion rates in the considered ion crystals increase exponentially when the ion temperatures exceed values that well agree with the predictions of other established models. Ion ensembles of different ion numbers (300, 500, and 1000 138Ba+ ions) and symmetries (prolate and spherical) have been investigated showing a clear tendency for the phase transition temperatures: The higher the ion number, the lower is the transition temperature.
We report multi-hour stable trapping and steady state of non-laser-coolable Cs+ ions in a linear Paul trap, achieved via resonant charge-exchange (RCE) cooling with a precisely centered ultracold Cs cloud. Without cold atoms, all ions are lost within 3 minutes. With centered cold atoms, RCE cooling establishes a stable population of ions for 6 hours with no measurable decay -- more than two orders of magnitude longer hold times compared to prior hybrid atom-ion systems. We observe a threshold behavior in the steady-state number of ions (Ns): initial ion loadings above Ns ions converge downward to Ns ions and remain constant thereafter; initial loadings below Ns ions show no measurable decay. The dynamics is governed by a competition between ion-ion rf heating and ion-atom collisional cooling. The prolonged, simultaneous ion-atom trapping suggests an upper bound on the three-body recombination-induced Cs2+ formation rate constant: k3<<1.4×10−25cm6s−1. This remarkable realization of long-lived steady state overcomes a critical limitation of prior hybrid atom-ion systems and enables extended studies of ultracold ion-neutral chemistry, rare inelastic collisions, and sympathetic cooling routes for complex molecular ions.
Collisions between ultracold neutral atoms have for some time been the subject of investigation, initially with hydrogen and more recently with laser cooled alkali atoms. Advances in laser cooling and trapping of neutral atoms in a Magneto-Optic Trap (MOT) have made cold atoms available as the starting point for many laser cooled atomic physics investigations. The most spectacularly successful of these, the observation of Bose-Einstein Condensation (BEC) in a dilute ultra-cold spin polarized atomic vapor, has accelerated the study of cold collisions. Experimental and theoretical studies of BEC and the long range interaction between cold alkali atoms is at the boundary of atomic and low temperature physics. Such studies have been difficult and would not have been possible without the development and advancement of laser cooling and trapping of neutral atoms. By contrast, ion-atom interactions at low temperature, also very difficult to study prior to modern day laser cooling, have remained largely unexplored. But now, many laboratories worldwide have almost routine access to cold neutral atoms. The combined technologies of ion trapping, together with laser cooling of neutrals has made these studies experimentally feasible and several very important, novel applications might come out of such investigations . This paper is an investigation of ion-atom interactions in the cold and ultra-cold temperature regime. Some of the collisional ion-atom interactions present at room temperature are very much reduced in the low temperature regime. Reaction rates for charge transfer between unlike atoms, A + B(+) approaches A(+) + B, are expected to fall rapidly with temperature, approximately as T(sup 5/2). Thus, cold mixtures of atoms and ions are expected to coexist for very long times, unlike room temperature mixtures of the same ion-atom combination. Thus, it seems feasible to cool ions via collisions with laser cooled atoms. Many of the conventional collisional interactions, exploited as a useful tool at room temperature and higher, are greatly enhanced at low energy. For example, collisional spin transfer from one species of polarized atoms to another has long been a useful method for polarizing a sample of atoms where no other means was available. Because optical pumping cannot be used to polarize the nuclear spin of Xe-129 or He-3 (for use in nmr imaging of the lungs), the nuclear spins are polarized via collisions with an optically pumped Rb vapor in a cell containing both gases. In another case, a spin polarized thermal Cs beam was used to polarize the hyperfine states of trapped He(+)-3 ions in order to measure their hyperfine clock transition frequency. The absence of an x-ray light source to optically pump the ground state of the He(+)-3 ion necessitated this alternative state preparation. Similarly, Cd(+) and Sr(+) ions were spin-oriented via collisions in a cell with optically pumped Rb vapor. Resonant RF spin changing transitions in the ground state of the ions were detected by changes in the Rb resonance light absorption. Because cold collision spin exchange rates scale with temperature as T(sup -1/2) this technique is expected to be a far more powerful tool than the room temperature counterpart. This factor of 100 or more enhancement in spin exchange reaction rates at low temperatures is the basis for a novel trapped ion clock where laser cooled neutrals will cool, state select and monitor the ion clock transition. The advantage over conventional direct laser cooling of trapped ions is that the very expensive and cumbersome UV laser light sources, required to excite the ionic cooling transition, are effectively replaced by simple diode lasers.