Laser-cooled trapped ions platform is one of the best candidates for the development of future quantum computing. This has generated a major worldwide research effort aimed at scaling and integrating trapping devices. As part of this effort, we are developing miniature atomic ion traps in the laboratory: Paul linear surface traps manufactured in collaboration with Nanyang Technology University and the Microelectronics Institute of Singapore. An original feature of the manufacturing process is that all the electrical contacts of the trap electrodes are made through the Silicon substrate (TSV: through silicon vias). In this way, the wire-bondings usually soldered directly to the electrodes can be offset or completely eliminated [1]. This opens the way to complex architectures, in particular cylindrically symmetrical ring traps, which cannot be made with surface connections that would break the desired symmetry. The trapping and laser cooling of ions in cylindrically symmetrical linear Paul traps (called "ion storage rings" in their macroscopic version, diameters of around 100 mm) was demonstrated in pioneering work at Garching [2] and subsequently extended to microfabricated traps (multilayer technology) at Sandia National Labs (diameter 2.5 mm) [3] and Berkeley (diameter 95 μm) [4]. These devices are of interest because, in the absence of defects, they enable rotational symmetry for the trapped ions and thus periodic boundary conditions for the confined cold ion system (Coulomb crystal). They are also candidates for observing and manipulating the rotational quantum state of trapped ion ensembles [4]. We will present our latest results concerning the performance of TSV surface ring traps (diameters between 150 and 210 μm) manufactured in Singapore. In particular, we loaded these traps with Dopplercooled 88Sr+ ions (from a single ion to several hundreds). Laser cooling and image acquisition enable us to estimate the defects of the trapping potential with respect to perfect rotational symmetry. The application of DC voltages to a set of segmented electrodes then makes it possible to compensate for most of these defects, which nevertheless remain the major problem to be solved in order to achieve sub milikelvin free motion.
Laser-cooled trapped ions platform is one of the best candidates for the development of future quantum computing. This has generated a major worldwide research effort aimed at scaling and integrating trapping devices. As part of this effort, we are developing miniature atomic ion traps in the laboratory: Paul linear surface traps manufactured in collaboration with Nanyang Technology University and the Microelectronics Institute of Singapore. An original feature of the manufacturing process is that all the electrical contacts of the trap electrodes are made through the Silicon substrate (TSV: through silicon vias). In this way, the wire-bondings usually soldered directly to the electrodes can be offset or completely eliminated [1]. This opens the way to complex architectures, in particular cylindrically symmetrical ring traps, which cannot be made with surface connections that would break the desired symmetry. The trapping and laser cooling of ions in cylindrically symmetrical linear Paul traps (called "ion storage rings" in their macroscopic version, diameters of around 100 mm) was demonstrated in pioneering work at Garching [2] and subsequently extended to microfabricated traps (multilayer technology) at Sandia National Labs (diameter 2.5 mm) [3] and Berkeley (diameter 95 μm) [4]. These devices are of interest because, in the absence of defects, they enable rotational symmetry for the trapped ions and thus periodic boundary conditions for the confined cold ion system (Coulomb crystal). They are also candidates for observing and manipulating the rotational quantum state of trapped ion ensembles [4]. We will present our latest results concerning the performance of TSV surface ring traps (diameters between 150 and 210 μm) manufactured in Singapore. In particular, we loaded these traps with Dopplercooled 88Sr+ ions (from a single ion to several hundreds). Laser cooling and image acquisition enable us to estimate the defects of the trapping potential with respect to perfect rotational symmetry. The application of DC voltages to a set of segmented electrodes then makes it possible to compensate for most of these defects, which nevertheless remain the major problem to be solved in order to achieve sub milikelvin free motion.
Laser-cooled trapped ions platform is one of the best candidates for the development of future quantum computing. This has generated a worldwide research effort for scaling and integrating trapping devices. As part of this effort, we are developping cylindrically symmetrical linear Paul traps manufactured in collaboration with Nanyang Technology University and the Microelectronics Institute of Singapore. The electrical contacts of these traps electrodes are made through the Silicon substrate (TSV: through silicon vias), allowing not to break the desired symmetry with surface connections. The trapping and laser cooling of ions in ring traps was demonstrated at Garching [2] and subsequently extended to microfabricated traps at Sandia National Labs [3] and Berkeley [4]. In the absence of defects, these devices enable rotational symmetry for the trapped ions. They are also candidates for manipulating the rotational quantum state of trapped ion ensembles [4].We will present our results concerning the performance of TSV surface ring traps manufactured in Singapore. In particular, we loaded these traps with Doppler-cooled 88Sr+ ions and we estimate the defects of the trapping potential with respect to perfect rotational symmetry. The application of DC voltages to a set of segmented electrodes makes it possible to compensate for most of these defects for observing sub milikelvin free motion.[1] P. Zhao et al., Appl. Phys. Lett. 118, 124003 (2021).[2] I. Waki et al.,Phys. Rev. Lett. 68, 2007 (1992).[3] B. Tabakov et al., Phys. Rev. Applied 4, 031001 (2015). [4] E. Urban et al., Phys. Rev. Lett. 123, 133202 (2019)
We report the laser cooling and trapping of barium ions in a novel planar Paul trap. Unlike other rf quadrupole traps, the electrodes of a planar Paul trap are confined to one or more parallel planes. They may therefore be constructed on a micrometer scale by the same photolithographic techniques used for semiconductor devices. Such devices are under construction in our laboratory. Micro-traps permit entirely new studies, for example, superradiance of a few ions, which occurs as ion-ion distances approach one wavelength of light. Photolithography can also generate an array of traps on a single substrate for an atomic clock. Our traps consist of one or more parallel planes, one above the other, each plane having either a conducting ring or a hole in a conducting sheet. Thus, a ring or a hole results in a potential well for trapping. A three-ring design produces a quadrupole potential that is harmonic up to eighth-order for a particular choice of ring diameters and spacings. Three-hole traps have been constructed with inner hole radii ranging from 50 to 400 μm, and both single barium ions and clouds have been observed.
Ion traps are a foundational technology in quantum information processing, precision spectroscopy and foundational physics tests. At the heart of their performance is the inhibition of micromotion, an oscillation of ions through imperfect trapping potentials, that restricts the laser cooling efficiency, coherence times, and the accuracy of measurements. The paper provides a systematic computational analysis of four different electrode geometries (Linear Paul, Segmented, Surface and Ring) in 24 different RF / DC configurations (C01-C24) and 13 performance parameters that comprise trap depth (0.62 1.48 eV), radial secular frequency (1,100 1 Multi-objective optimization with weighted scoring of these metrics indicates that Ring trap geometry scores the highest total (80.2/100), followed by Linear Paul (72.0), Segmented (67.0) and Surface (55.1) configurations. ANOVA (one-way) statistically verified the presence of statistically significant inter-geometry differences in micro-motions amplitude (F = 18.43, p < .001) and cooling efficiency (F = 22.17, p < .001), with Tukey HSD post-hoc revealed that the most significant difference was between the Ring and the Surface. The Pearson correlation analysis shows that RF drive voltage and micromotion amplitude (r = -0.94 to -0.99) and cooling efficiency and micromotion (r = -1.00) have near-perfect inverse correlations, which prove that the quality of confinement is regulated by electrode geometry, RF amplitude, and spacing. These results give quantitative design principles of next-generation ion traps in quantum computing and atomic clock. Keywords: ion trap, micromotion minimization, RF/DC electrode optimization, laser cooling, Mathieu stability, secular frequency, quantum computing
Trapped ions within Paul traps have shown to be a promising architecture in the realisation of a quantum information processor together with the ability of providing quantum simulations. Linear Paul traps have demonstrated long coherence times with ions being well isolated from the environment, single and multi-qubit gates and the high fidelity detection of states. The scalability to large number of qubits, incorporating all the previous achievements requires an array of linear ion traps. Microfabrication techniques allow for fabrication and micron level accuracy of the trap electrode dimensions through photolithography techniques. The first part of this thesis presents the experiential setup and trapping of Yb+ ions needed to test large ion trap arrays. This include vacuum systems that can host advanced symmetric and asymmetric ion traps with up to 90 static voltage control electrodes. Demonstration of a single trapped Yb+ ion within a two-layer macroscopic ion trap is presented. with an ion-electrode distance of 310(10) µm. The anomalous heating rate and spectral noise density of the trap was measured, a main form of decoherence within ion traps. The second half of this thesis presents the design and fabrication of multi-layer asymmetric ion traps. This allows for isolated electrodes that cannot be accessed via surface pathways, allowing for higher density of electrodes as well as creating novel trap designs that allow for the potential of quantum simulations to be demonstrated. These include two-dimensional lattices and ring trap designs in which the isolated electrodes provide more control in the ion position. For the microfabrication of these traps I present a novel high-aspect ratio electroplated electrode design that provides shielding of the dielectric layer. This provides a means to mitigate stray electric field due to charge build up on the dielectric surfaces. Electrical testing of the trap structures was performed to test bulk breakdown and surface flashover of the ion trap architectures. Results showed sufficient isolation between electrodes for both radio frequency and static breakdown. Surface flashover voltage measurements over the dielectric layer showed an improvement of more than double over previous results using a new fabrication technique. This will allow for more powerful ion trap chips needed for the next generation of microfabricated ion trap arrays for scalable quantum technologies.
Trapped ions in Paul traps (radio frequency ion traps) are a promising candidate for the realization of a quantum computer. To date, experimental demonstrations of quantum computing and information processing with trapped ions have been limited to small numbers of quantum bits (qubits). The most prominent challenge to realise ion-trap quantum computation is to scale the system and control a large number of trapped ions coherently. Surface-electrode ion trap designs which are amenable to various microfabrication techniques offer a path to achieve this. In this work, design, fabrication and key operations of two types of surface-electrode ion traps, i.e. a Y-junction trap and a two-dimensional (2D) hexagonal lattice trap developed towards scaling ion-trap architectures, are presented. The steps involved in the fabrication of the ion-trap devices included photolithography, wet etch (hydrofluoric acid etch) and dry etch (deep reactive ion and inductively coupled plasma etching), and metal evaporation. The devices were fabricated on SOI substrates with different buried oxide thicknesses (i.e. 5 µm and 10 µm) and different metals for the trap electrodes (i.e. gold and aluminium). An extremely high-breakdown voltage up to 1 kV was measured in vacuum for the SOI-built test devices with a 10 µm-thick buried oxide formed by oxide film-to-oxide film bonding process. This is attributed to the deep V-shaped undercut profile of the buried oxide layer which resulted in a large breakdown path length, and so mitigating surface breakdown effects. Owing to the V-shaped undercut profile, the surface breakdown voltage was improved by 15% for the devices fabricated on SOI substrates with a 10 µm-thick buried oxide. In the first experiment, the demonstration of ytterbium (174Yb+) ion trapping in the 2D hexagonal lattice trap is achieved with a relatively long lifetime of a laser-cooled ion up to 90 minutes and ≤ 5 minutes without cooling in ultra-high vacuum system. The 2D lattice trap design allows for reliable trapping of 2D ion lattices, rudimentary shuttling between lattice sites and deterministic introducing of defects into the ion lattice. Based on all these abilities, such a 2D lattice trap with the predicted reduction of lattice spacing down to 32 µm can be configured as a versatile architecture for 2D quantum simulations with trapped ions. In the second experiment, microparticle trapping in the 2D lattice trap, performed in air, is achieved with the longest trapping lifetime of 60 minutes. Both gold and aluminium-coated electrodes also lead to comparable trapping performances. Furthermore, the in situ levitation of microparticles is successfully demonstrated at the maximum levitation height of 30 µm. This ability provides additional control on the trapping height which increases the overall functionality of the 2D lattice trap.
With all the key elements of quantum computing in ion traps demonstrated by the research community, the focus is now placed on building more sophisticated traps with larger numbers of ions to allow practical scale information processing. One promising avenue is to store ions in and shuttle them between many independent traps which serve as potential interaction sites. The core of the work described in this thesis is the experimental evaluation of a microfabricated segmented ion trap, built by Sandia National Laboratories ("Sandia trap"). These experiments required construction of a wholly new optical setup including laser and detection systems, a vacuum system and control electronics. Among our experimental achievements were: successful loading of single and pairs of ions in the microscale trap, measurement of ion storage lifetime, measurement of the motional heating rate with a time-resolved Doppler-cooling method — which showed above than average heating, and implemented a single-ion shuttling method — which reliably transferred the ion through a distance of 360 μm (two DC electrode widths away) and back. These results have been used to improve the next version of the Sandia trap design. We also used computer modelling to study several aspects of ion traps: a mesoscopic ion trap designed for fast ion separation, simulated ion loading to quantify requirements for successful trapping in small and shallow traps, and analyzed a precise shuttling method — where the time dependence of the trapping potential is engineered such that there is minimal motional heating. The results show that ion trap arrays at the 100 μm distance scale are feasible and suggests that such multiple trap designs merit further study.