Session X48: Scalable Hardware for Superconducting Qubits

High-Q 3D coaxial resonators for cavity QED

Three-dimensional microwave resonators provide an alternative approach to transmission-line resonators used in most current circuit QED experiments [1]. Their large mode volume greatly reduces the surface dielectric losses that limits the coherence of superconducting circuits, and the well-isolated and controlled cavity modes further suppress coupling to the environment. In this work, we focus on unibody 3D coaxial cavities which are only evanescently coupled and free from losses due to metal-metal interfaces, allowing us to reach extremely high quality-factors. We achieve quality-factor of up to 170 million using 4N6 Aluminum at superconducting temperatures, corresponding to an energy ringdown time of \textasciitilde 4ms. We extend our methods to other materials including Niobium, NbTi, and copper coated with Tin-Lead solder. These cavities can be further explored to study their properties under magnetic field or upon coupling to superconducting Josephson junction qubits, e.g. 3D transmon qubits. Such 3D cavity QED system can be used for quantum information applications, or quantum simulation in coupled cavity arrays. References: [1] Matthew Reagor et al.,~A quantum memory with near-millisecond coherence in circuit QED. arXiv:~1508.05882 (2015)   

Three-Dimensional Architecture at Chip Level for Large-Scale-Integration of Superconducting Quantum Electronic Devices

We propose a novel way to realize three-dimensional circuit QED systems at chip level. System components such as qubits, transmission lines, capacitors, inductors or cross-overs can be implemented as suspended, electromagnetically shielded and optionally, as hermetically sealed structures. Compared to known state-of-the-art devices, volumes of dielectrics penetrated by electromagnetic fields can be drastically reduced. Our intention is to harness process technologies for very-large-scale-integration, reliably applied and improved over decades in micro-sensor- and semiconductor industry, for the realization of highly integrated circuit QED systems. Process capabilities are demonstrated by fabricating first exploratory devices using the back-end-of-line part of a commercial 180\,nm CMOS foundry process in conjunction with HF vapor phase release etching.   

3D Integration for Superconducting Qubits

As the field of superconducting quantum computing advances from the few-qubit stage to large-scale fault-tolerant devices, scalability requirements will necessitate the use of standard 3D packaging and integration processes. While the field of 3D integration is well-developed, relatively little work has been performed to determine the compatibility of the associated processes with superconducting qubits. Qubit coherence time could potentially be affected by required process steps or by the proximity of an interposer that could introduce extra sources of charge or flux noise. As a first step towards a large-scale quantum information processor, we have used a flip-chip process to bond a chip with flux qubits to an interposer containing structures for qubit readout and control. We will present data on the effect of the presence of the interposer on qubit coherence time for various qubit-chip-interposer spacings and discuss the implications for integrated multi-qubit devices. This research was funded by the ODNI and IARPA under Air Force Contract No. FA8721-05-C-0002. The views and conclusions contained herein are those of the authors and should not be interpreted as representing the official policies or endorsements, either expressed or implied, of ODNI, IARPA, or the US Government.   

Extensible circuit QED processor architecture with vertical I/O

Achieving quantum fault tolerance in an extensible architecture is an outstanding challenge across experimental quantum computing platforms today. Traditionally, circuit QED processors have millimeter dimensions and lateral coupling for all input/output (I/O) signals, precluding the increase in qubit numbers beyond \textasciitilde 10. We present a scalable footprint for circuit QED processors with vertically coupled I/O. Our demonstration using centimeter scale chips can accommodate the \textasciitilde 50 qubits needed in next-generation processors targeting the experimental demonstration of quantum fault tolerance.   

The Quantum Socket: Wiring for Superconducting Qubits - Part 1

Quantum systems with ten superconducting quantum bits (qubits) have been realized, making it possible to show basic quantum error correction (QEC) algorithms. However, a truly scalable architecture has not been developed yet. QEC requires a two-dimensional array of qubits, restricting any interconnection to external classical systems to the third axis. In this talk, we introduce an interconnect solution for solid-state qubits: The quantum socket. The quantum socket employs three-dimensional wires and makes it possible to connect classical electronics with quantum circuits more densely and accurately than methods based on wire bonding. The three-dimensional wires are based on spring-loaded pins engineered to insure compatibility with quantum computing applications. Extensive design work and machining was required, with focus on material quality to prevent magnetic impurities. Microwave simulations were undertaken to optimize the design, focusing on the interface between the micro-connector and an on-chip coplanar waveguide pad. Simulations revealed good performance from DC to 10 GHz and were later confirmed against experimental measurements.   

The Quantum Socket: Wiring for Superconducting Qubits - Part 2

Quantum computing research has reached a level of maturity where quantum error correction (QEC) codes can be executed on linear arrays of superconducting quantum bits (qubits). A truly scalable quantum computing architecture, however, based on practical QEC algorithms, requires nearest neighbor interaction between qubits on a two-dimensional array. Such an arrangement is not possible with techniques that rely on wire bonding. To address this issue, we have developed the quantum socket, a device based on three-dimensional wires that enables the control of superconducting qubits on a two-dimensional grid. In this talk, we present experimental results characterizing this type of wiring. We will show that the quantum socket performs exceptionally well for the transmission and reflection of microwave signals up to 10 GHz, while minimizing crosstalk between adjacent wires. Under realistic conditions, we measured an $S_{21}$ of -5 dB at 6 GHz and an average crosstalk of -60 dB. We also describe time domain reflectometry results and arbitrary pulse transmission tests, showing that the quantum socket can be used to control superconducting qubits.   

The Quantum Socket: Wiring for Superconducting Qubits - Part 3

The implementation of a quantum computer requires quantum error correction codes, which allow to correct errors occurring on physical quantum bits (qubits). Ensemble of physical qubits will be grouped to form a logical qubit with a lower error rate. Reaching low error rates will necessitate a large number of physical qubits. Thus, a scalable qubit architecture must be developed. Superconducting qubits have been used to realize error correction. However, a truly scalable qubit architecture has yet to be demonstrated. A critical step towards scalability is the realization of a wiring method that allows to address qubits densely and accurately. A quantum socket that serves this purpose has been designed and tested at microwave frequencies. In this talk, we show results where the socket is used at millikelvin temperatures to measure an on-chip superconducting resonator. The control electronics is another fundamental element for scalability. We will present a proposal based on the quantum socket to interconnect a classical control hardware to a superconducting qubit hardware, where both are operated at millikelvin temperatures.   

Development of superconducting bonding for multilayer microwave integrated quantum circuits

Future quantum computers are likely to take the shape of multilayer microwave integrated quantum circuits.[1] The proposed physical architecture retains the superb coherence of 3D structures while achieving superior scalability and compatibility with planar circuitry and integrated readout electronics. This hardware platform utilizes known techniques of bulk etching in silicon wafers and requires metallic bonding of superconducting materials. Superconducting wafer bonding is a crucial tool in need of development. Whether micromachined in wafers or traditionally machined in bulk metal, 3D cavities typically posses a seam where two parts meet. Ideally, this seam consists of a perfect superconducting bond. Pursuing this goal, we have developed a new understanding of seams as a loss mechanism that is applicable to 3D cavities in general.[2] We present quality factor measurements of both 3D cavities and 2D stripline resonators to study the losses of superconducting bonds. [1]Brecht, T. \textit{et al.}, arXiv:1509.01127 (2015) [2]Brecht, T. \textit{et al.}, arXiv:1509.01119 (2015)   

High speed on-chip current measurement using a low-Q tunable LC resonator

Superconducting quantum computing technology requires precise high frequency analog waveforms to perform single and multi-qubit gates. Due to signal path irregularities, gates are tuned-up by perturbing the drive signal until qubit state populations indicate the desired gate function. A more direct approach is to measure the effect of circuit imperfections by sampling control waveforms directly, as seen by the qubits. We proceed by measuring the resonant frequency shift of a capacitively shunted SQUID and converting the control waveform to DC flux applied to the SQUID. By measuring the reflected phase of a CW tone applied to this resonant circuit while applying the resonance-shifting flux pulse, we are able to reconstruct the current waveform of the input pulse at the SQUID loop. This device's geometry is the same as the z-control lines used in qubit experiments to control the qubit frequency. I will present this method of on-chip waveform sampling for superconducting circuits in addition to proof of concept data. This technique opens the door for improved gate bring up and a deeper understanding of qubit control as well as the circuit parasitics that deform these waveforms.   

Maintaining Qubit Coherence in the face of Increased Superconducting Circuit Complexity

Maintaining qubit coherence in the face of increased superconducting circuit complexity is a challenge when designing an extensible quantum computing architecture. We consider this challenge in the context of inductively coupled, long-lived, capacitively-shunted flux qubits. Specifically, we discuss our efforts to mitigate the effects of radiation loss, parasitic chip-modes, cross-coupling, and Purcell decay. Our approach employs numerical modeling of the ideal Hamiltonian and electromagnetic analysis of the circuit, both of which are independently shown to be consistent with experimental results. This research was funded by the Office of the Director of National Intelligence (ODNI), Intelligence Advanced Research Projects Activity (IARPA) and by the Assistant Secretary of Defense for Research & Engineering under Air Force Contract No. FA8721-05-C-0002. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of ODNI, IARPA, or the US Government.   

Time Reversal Symmetry Breaking Microwave Resonators

In this talk we present our work towards realizing high Q, superconducting circulators to be employed in topological circuit QED lattices. These circulators generate gauge fields that produce protected edge states. We couple magnon excitations in spheres of the ferrite Yttrium Iron Garnet (YIG) to microwave cavity fields in order to break the degeneracy between modes that precess with different handedness. The YIG sphere only couples strongly (~1GHz) to cavity modes that precess with the same handedness. We tune the YIG sphere into resonance with degenerate cavity modes to shift only the frequency of the modes with the same handedness, leaving the uncoupled mode at its original frequency. Since this mode is dark to the YIG excitation, its quality factor is dependent only on the characteristics of the cavity. We make the cavities out of the Type II superconductor Niobium Titanium so that we achieve high quality factors while also tolerating the large magnetic fields acting on the YIG spheres within the cavities. These cavities can be evanescently coupled to create topologically nontrivial lattices. Photon-photon interactions can then be added via couplings to qubits to create fractional quantum hall states for microwave photons.   

On Chip Josephson Junction Microwave Switch

We report on the design and measurement of a reflective single-pole single-throw microwave switch based on a superconducting circuit containing a single Josephson junction. The device has no internal power dissipation, minimal insertion loss, and is controlled by $\Phi_{\mathrm{0}}$-level base-band signals. The data demonstrates the device operation with 2 GHz instantaneous bandwidth centered at 10 GHz and better than 20 dB on/off ratio for input powers up to -100 dBm.   

Realization of an on-chip superconducting microwave switch

As state-of-the-art superconducting quantum devices get increasingly complex, they require a growing number of control and detection channels. On-chip routing and multiplexing of signals presents a way to realize these without requiring an unrealistically large number of microwave lines. The ability to route signals on a chip will also be a useful tool for fast in-situ characterization of superconducting devices. Here, we describe and experimentally demonstrate a superconducting on-chip microwave switch which can be integrated with current superconducting quantum circuits. The device is based on interference effects and is in principle lossless, making it well-suited for operation in dilution cryostats and for routing of signals at the single quantum level with near-unity efficiency. The first proof-of-principle device has a bandwidth of $150\,\mathrm{MHz}$, a $1\,\mathrm{dB}$ compression point of $-80\,\mathrm{dBm}$ and turn-on/off times on the order of $5\,\mathrm{ns}$. On/off power ratios reach values of approximately $30\,\mathrm{dB}$. We expect that our device will find use in (de)multiplexing of control and readout in superconducting circuits and routing of microwave fields in quantum optical experiments and quantum communication applications.   

Coherent control of a linear microwave cavity via single flux quantum pulses

Classical Josephson digital logic based on single flux quantum (SFQ) pulses offers a path to robust, low-latency control of a large-scale quantum processor. Here we describe the coherent control of a linear superconducting cavity by direct excitation via SFQ pulses. Resonant trains of SFQ pulses are capacitively coupled to a thin-film coplanar waveguide cavity. We examine the resulting cavity states as a function of subharmonic drive and temperature. In addition, we describe first steps toward the coherent control of a superconducting qubit with SFQ pulses.   

Development of Integrated Single Flux Quantum - Superconducting Qubit Circuits

Significant theoretical and experimental progress has been made in recent years towards a scalable superconducting quantum circuit architecture. Here we present a first attempt to integrate classical control elements from the single flux quantum (SFQ) digital logic family with a superconducting transom qubit on a single chip. The SFQ driving circuit is fabricated in a six-layer high-$J_c$ Nb/Al-AlOx/Nb junction process while the transmon qubit is subsequently formed using submicron Al-AlOx-Al junctions grown by double-angle evaporation. We investigate sources of decoherence associated with the more complex fabrication process and describe first attempts to perform coherent qubit manipulations using resonant trains of SFQ pulses.