Arkadas Ozakin

Recent advances in quantum information processing with trapped ions have demonstrated the need for new ion trap architectures capable of holding and manipulating chains of many (>10) ions. Here we present the design and detailed characterization of a new linear trap, microfabricated with scalable complementary metal-oxide-semiconductor (CMOS) techniques, that is well-suited to this challenge. Forty-four individually controlled DC electrodes provide the many degrees of freedom required to construct anharmonic potential wells, shuttle ions, merge and split ion chains, precisely tune secular mode frequencies, and adjust the orientation of trap axes. Microfabricated capacitors on DC electrodes suppress radio-frequency pickup and excess micromotion, while a top-level ground layer simplifies modeling of electric fields and protects trap structures underneath. A localized aperture in the substrate provides access to the trapping region from an oven below, permitting deterministic loading of particular isotopic/elemental sequences via species-selective photoionization. The shapes of the aperture and radio-frequency electrodes are optimized to minimize perturbation of the trapping pseudopotential. Laboratory experiments verify simulated potentials and characterize trapping lifetimes, stray electric fields, and ion heating rates, while measurement and cancellation of spatially-varying stray electric fields permits the formation of nearly-equally spaced ion chains.

Motivated by recent developments in ion trap design and fabrication, we investigate the stability of ion motion in asymmetrical, planar versions of the classic Paul trap. The equations of motion of an ion in such a trap are generally coupled due to a nonzero relative angle $θ$ between the principal axes of RF and DC fields, invalidating the assumptions behind the standard stability analysis for symmetric Paul traps. We obtain stability diagrams for the coupled system for various values of $θ$, generalizing the standard $q$-$a$ stability diagrams. We use multi-scale perturbation theory to obtain approximate formulas for the boundaries of the primary stability region and obtain some of the stability boundaries independently by using the method of infinite determinants. We cross-check the consistency of the results of these methods. Our results show that while the primary stability region is quite robust to changes in $θ$, a secondary stability region is highly variable, joining the primary stability region at the special case of $θ=45^{\circ}$, which results in a significantly enlarged stability region for this particular angle. We conclude that while the stability diagrams for classical, symmetric Paul traps are not entirely accurate for asymmetric surface traps (or for other types of traps with a relative angle between the RF and DC axes), they are safe in the sense that operating conditions deemed stable according to standard stability plots are in fact stable for asymmetric traps, as well. By ignoring the coupling in the equations, one only underestimates the size of the primary stability region.

We describe a novel monolithic ion trap that combines the flexibility and scalability of silicon microfabrication technologies with the superior trapping characteristics of traditional four-rod Paul traps. The performace of the proposed microfabricated trap approaches that of the macroscopic structures. The fabrication process creates an angled through-chip slot which allows backside ion loading and through-laser access while avoiding surface light scattering and dielectric charging. The trap geometry and dimensions are optimized for confining long ion chains with equal ion spacing [G.-D. Lin, et al., Europhys. Lett. 86, 60004 (2009)]. Control potentials have been derived to produce linear, equally spaced ion chains of up to 50 ions spaced at 10 um. With the deep trapping depths achievable in this design, we expect that these chains will be sufficiently long-lived to be used in quantum simulations of magnetic systems [E.E. Edwards, et al., Phys. Rev. B 82, 060412(R) (2010)]. The trap is currently being fabricated at Georgia Tech using VLSI techniques.