Session S41: Novel Superconducting Qubits I

Improving quantum control of the soft 0-pi qubit through device engineering

Ideally, 0-Pi qubit is fully protected against both depolarization and dephasing noise. However, realizing it is a difficult task because of the strict device-fabrication constraints. Therefore, soft 0-Pi qubit has been introduced as an alternative, which is achievable with current technology at the expense of weaker flux-noise protection. However, implementing fast qubit control while avoiding leakage is one of the big challenges that soft 0-Pi qubit faces. In this talk, we explore the parameter space of the soft 0-Pi qubit to maximize coherence and gate speed through low-lying Raman transitions. Moreover, we discuss fabrication constraints and technological considerations that could help push the qubit's design beyond the soft 0-Pi qubit.   

Bifluxon Qubit with Reduce Sensitivity to Charge Noise

The Bifluxon qubit is a symmetry-protected qubit whose states are encoded in the parity of fluxons in a superconducting loop [1]. Protection of fluxon-parity in the Bifluxon qubit is preserved through Aharonov-Casher destructive interference of single-fluxon tunneling events. Charge noise detrimentally affects parity protection by increasing the probability of single-fluxon tunneling. With the aim of protection improvement, we propose a modified design of the Bifluxon qubit which consists of two parallel split-Cooper-pair boxes shunted by a superinductor arranged in a superconducting loop. We show that the addition of a second Cooper-pair box suppresses the effect of charge noise and results in improved protection against energy relaxation. We will present preliminary results on the implementation of this Bifluxon qubit with the gate control of induced charges on two superconducting islands; the data demonstrate exceptional charge stability and suppressed quasi-particle poisoning due to the gap engineering in circuit elements. The spectroscopy data will be compared with numerical diagonalization of circuit Hamiltonian.

[1] K. Kalashnikov et. al. PRX Quantum 1, 010307 (2020).   

Experimental Verification of Time-Dependent Flux Allocation in Fluxonium

External flux in the fluxonium Hamiltonian is generally allocated entirely to the inductor or to the Josephson junction, both of which are equally valid in the case of static external flux. However, recent theory work has shown that given a time-dependent external flux, allocation of the flux to the junction leads to terms proportional to the time derivative of the flux appearing in the Hamiltonian [1]. When the flux is allocated to the inductor, the usual fluxonium Hamiltonian is recovered. Here, we experimentally verify the theory predictions by rapidly modulating the external flux and recording qubit state transfer. We find results consistent with grouping the flux with the inductor, and inconsistent with grouping the flux with the junction while ignoring the derivative term.

 

[1] X. You, et. al, Phys. Rev. B (2019).   

Blochnium: A flux-insensitive qubit with flux-tunable interactions

We introduce a flux-like qubit, nicknamed 'blochnium', whose 0-1 transition frequency is nearly insensitive to the external flux bias and is dual to a transmon. Blochnium is created by shunting a small Josephson junction with a hyperinductance, that is -- a maximal per-unit-length inductance that has minimal stray capacitance and is probably the highest impedance electromagnetic structure available today (L_tot > 1  μH, Z > 200 kΩ). To achieve such inductances, we release the entire circuit from the high dielectric substrate and suspend it almost entirely in vacuum. The qubit’s spectrum reveals a 0-1 flux dispersion on the order of a hundred MHz which becomes exponentially suppressed when decreasing the ratio EL/EC providing protection against 1/f flux noise. However, the higher transitions remain flux-sensitive allowing for multi-qubit operations. We report on recent coherence properties of blochnium with varying etching techniques.   

Charge-parity qubits based on concatenation of π-periodic Josephson elements: Part 2

For charge-parity qubits formed from concatenated π-periodic plaquette elements, the dispersion of the qubit energy bands with flux depends on the path through the multi-dimensional flux bias space. With only one plaquette biased near π-periodicity, the dispersion is linear near frustration, while biasing two concatenated plaquettes near π-periodicity results in a quadratic dispersion. Here, we describe our experimental characterization of devices with three concatenated π-periodic elements through spectroscopy and time-domain measurements at various flux and charge bias regimes.   

Charge-parity qubits based on concatenation of \pi-periodic Josephson elements: Part 1

Superconducting qubits with topological protection against local noise hold the promise of significantly enhanced coherence times and higher gate fidelities than are possible with conventional qubits. We are pursuing a design that uses compact, high inductances and conventional Josephson junctions with individual flux control for tuning each plaquette to reach a regime with a \pi-periodic Josephson energy. Concatenating multiple \pi-periodic elements further enhances the degree of protection that can be achieved. Here we describe the design, fabrication, and flux characterization of such a qubit.   

Implementation of cross resonance gate for a capacitively coupled two fluxonium device in 3D cavity

Fluxonium qubits stand out with their high anharmonicity and longer coherence times [1,2]. For this reason they are promising candidates for cross resonance gate applications where the weak transmon anharmonicity is one of the main limitations [3]. Our work focuses on the implementation of the cross resonance gate on a capacitively coupled two fluxonium system and suggests an 'all microwave' scheme in a simple 3D resonator geometry. We will present our improved toolbox for the realization of the CR gate in our system and report recent benchmarking results. In particular we will discuss observations of the effect of strong drives on fluxonium qubits in a 3D cavity.

[1] L. B. Nguyen, et. al., Phys. Rev. X 9, 041041 (2019)

[2] A. Somoroff, et. al., arXiv: 2103.08578 (2021)

[3] S. Sheldon, et al., Phys. Rev. A 93, 060302 (2016)   

Inductively shunted transmon: a qubit with phase flip protected plasmon and bit flip protected fluxon states

The inductively shunted transmon (IST) is a superconducting qubit with exponentially suppressed fluxon transitions and a plasmon spectrum approximating that of the transmon. It shares many characteristics with the transmon but offers charge offset insensitivity for all levels and precise flux tunability with quadratic flux noise suppression. In this work we propose and realize IST qubits deep in the transmon limit where the large geometric inductance acts as a mere perturbation. With a flux dispersion of only 5 MHz we reach the 'sweet spot everywhere' regime of a SQUID device with a measured constant coherence time over the full flux range. At the flux degeneracy point the device reveals tunneling physics between the two quasi-degenerate ground states with typical observed lifetimes on the order of hours. In the future, a smaller inductor could be used to avoid leakage to unconfined transmon states in high power read-out or driven-dissipative bosonic qubit realizations. Moreover, the combination of well controllable plasmon transitions together with stable fluxon states in a single device might offer a way forward towards improved qubit encoding schemes.   

Design of a granular aluminum fluxonium qubit in a coplanar waveguide architecture

Fluxonium qubits are often embedded in rectangular waveguides which dilute the electric field and favor high coherence [1,2]. However, this configuration complicates in-situ flux gates and multi-qubit experiments. Here, we present a fluxonium qubit placed in a coplanar waveguide architecture with an integrated fast-flux coil, surrounded by a normal metal ground plane. The superinductor is made out of granular aluminum (grAl) [3] and the use of a comparably large silver ground plane potentially decreases the number of quasiparticles in the system via phonon trapping [4].

[1] Pop et al., Nature 508, 369–372 (2014)

[2] Somoroff et al., arXiv:2103.08578 (2021)

[3] Grünhaupt and Spiecker et al., Nat. Mater. 18, 816–819 (2019)

[4] Henriques and Valenti et al., Appl. Phys. Lett. 115, 212601 (2019)   

A microwave-activated controlled-phase gate between a transmon and a fluxonium

In the quest for high qubit coherence, fluxonium qubits have emerged as promising candidates for

storing quantum information, reaching 1 ms coherence times. On the other hand, large

superconducting qubit devices based on transmon qubits have been built due to the relative

simplicity of the circuit and the ability to perform fast two-qubit gates compared to their coherence

times. We here analyze how two-qubit gates between a transmon and a fluxonium can be realized in

a possible hybrid architecture. The observation that the typical transmon frequencies are at least one

order of magnitude larger than those of the fluxonia leads immediately to the following question:

how can we couple qubits with such different frequencies? We show that this is possible by

microwave-activating the coupling by virtually exciting the fluxonium to its higher levels. In

particular, we consider a microwave-driven controlled-phase gate between capacitively-coupled

transmon and fluxonium qubits, similar to the approach proposed in [Ficheux et al, Phys. Rev. X 11,

021026, (2021)] for two fluxonia. We perform simulations of the gate, including relaxation and

dephasing noise, and show that the gate can be realized with high fidelity while allowing for low

leakage and low residual ZZ coupling.   

High-fidelity QND readout of Heavy-Fluxonium Qubit

Heavy fluxonium qubits with their rich anharmonic spectrum, high coherence,and high-fidelity two-qubit gates are emerging as a potential alternative to transmon qubits for building superconducting quantum processors. In addition,it is also necessary to have a high-fidelity, quantum non-demolition (QND) readout to realize a quantum processor. In principle, the fidelity of the circuit-QED based readout for the fluxonium qubit can be increased by using larger microwave power. But in some earlier work, it was observed that the qubit lifetime decreases significantly with increasing readout power [1] which resulted in lower fidelity.This effect also takes place in the transmon qubit readout but was observed to be stronger in fluxonium qubits. Recently,  high fidelity readout was demonstrated in fluxonium qubit with a super-inductance made using granular Aluminium [2]. Here, we describe a series of experiments to characterize and optimize measurement fidelity in heavy-fluxoium qubits which use the conventional Josephson-junction array for the super-inductance. We will discuss the relaxation induced by cavity photons, signal-to-noise ratio, and the QND aspect of the readout in the presence of a high readout photon number.

[1] V. Manucharyan, Superinductance, Ph.D. thesis, Yale University (2007).

[2] Daria Gusenkova, et.al, Phys. Rev. Applied 15, 064030 (2021).   

Understanding and mitigating decoherence in fluxonium qubits

The fluxonium qubit, with its long coherence time and large anharmonicity, is a promising candidate for the physical realization of quantum computation. State-of-the-art devices have achieved a relaxation time (T₁) and a coherence time (T₂) in the millisecond regime, and various decoherence mechanisms have been studied. We present experimental results from an tunable-E_J device to study dissipation in fluxonium qubits, using high-bandwidth flux controls to tune the device and measure the relaxation time in a large parameter space. We then analyze the data to characterize the noise spectrum and identify dissipation sources such as tunneling two-level systems and quasiparticles. Statistics and temporal fluctuations of coherence times in fixed-E_J devices will also be discussed. This work deepens our understanding of decoherence and enables its mitigation, thus paves the way for the implementation of high-fidelity fluxonium quantum processors.