Quantum-limited traveling-wave parametric amplifier based on DUV lithography-defined planar structures

Abstract

The relentless scaling of classical microelectronics has been enabled by the precision and reproducibility of deep-ultraviolet (DUV) optical lithography. Implementing large-scale superconducting quantum processors will require cryogenic microwave components that follow a similarly scalable fabrication path. This need is particularly acute for high circuit-density devices such as traveling-wave parametric amplifiers (TWPAs), where recent implementations have demonstrated high gain, broad bandwidth, high saturation power, and near-quantum-limited noise, but trade-offs between footprint, insertion loss, and scalable integration remain. Here, we demonstrate a four-wave-mixing TWPA fabricated via a hybrid scheme that combines DUV-defined planar circuit elements with electron-beam-patterned Josephson junctions, constituting a first step toward fully scalable manufacturing. The device combines a compact footprint with broadband gain from 3 to 11 GHz and an average 1 dB compression point of -102 dBm. By using planar capacitors to reduce loss, it operates near the quantum limit, with added noise near 0 and 1.5 photons above the standard quantum limit and an average of 0.4 photons in the 4 to 8 GHz band. The phase-matching stopband remains narrow, with a bandwidth of 43 MHz, consistent with resonator-frequency variation below 1% and indicative of the uniformity enabled by DUV lithography. These results show that DUV-defined planar elements can enable compact, low-loss, near-quantum-limited TWPAs and provide a promising route toward high-density cryogenic microwave hardware for large-scale quantum systems.

Figures (4)

• Figure 1: Wafer-scale manufacturing of near quantum-limited Josephson traveling-wave parametric amplifier.a, the TWPA is realized on a 100 mm wafer, hosting 145 individual chips. The colors arise from the interference of scattered light by circuit features comparable to the optical wavelength. b, the core fabrication process relies on two primary lithography steps. Deep-ultraviolet (DUV) lithography (utilizing a 4x magnification mask) defines the waveguide capacitors, resonator capacitors, and meandered inductors. Electron-beam lithography (EBL) is subsequently used to define the Josephson junctions (JJs). The process is completed with junction patches and air-bridges to connect the ground planes (not shown). c, optical micrograph of a single TWPA chip post-dicing. d, room-temperature resistance mapping of single JJs, measured on test structures comprising 10- or 11-junction arrays. Junction dimensions are swept across nine discrete values. e, equivalent circuit schematic of a single TWPA unit cell, consisting of eight JJs and a phase-matching resonator. This unit cell is cascaded 256 times along the transmission line. f, amplification relies on a four-wave mixing process, wherein two pump photons at frequency $\omega_{\rm p}$ are converted into a signal photon at $\omega_{\rm s}$ and an idler photon at $\omega_{\rm i}$. g, false-color scanning electron microscope (SEM) micrograph of a unit cell. Highlighted components include the phase-matching resonator capacitor (green) and inductor (orange), detailed further in h. i, Close-up of the resonator coupling capacitor (purple) and the capacitance to ground (yellow). j, SEM micrograph detailing the Josephson junctions (blue).
• Figure 2: Performance characterization of the traveling-wave parametric amplifier.a, Simplified schematic of the measurement setup. The TWPA is mounted on the mixing chamber stage of a dilution refrigerator, with the signal and the pump tone injected through a directional coupler. A bypass coaxial cable and a qubit directly coupled to the waveguide allow for direct transmission measurements without the amplifier and power calibration, respectively. b, Transmittance $|S_{21}|^2$ and reflectance $|S_{11}|^2$ of the sample packaging at room temperature for a gold straight CPW. c, Packaged TWPA without the box's lid. d, Insertion loss of the TWPA without the pump, measured by comparing the transmission through the TWPA with the bypass cable. e, A detailed scan of the phase-matching stopband, highlighting the narrow 43 MHz rejection region. The continuous line represents a theoretical fit to the propagation in the TWPA line. f, Gain profile measured by comparing the TWPA output with the transmission through the bypass coaxial cable. Around the pump frequency at 6.688 GHz, two regions of no gain are visible due to the phase-matching stopband. The continuous orange line represents the simulated gain using extracted parameters. Crucially, the ripples observed in the measured gain directly correlate with the impedance mismatch and package resonances characterized in panel b.
• Figure 3: Noise performance of the near quantum-limited Josephson junctions based TWPA.a, Optical micrograph of the frequency-tunable waveguide transmon used for signal and pump power calibration. The inset shows a magnified view of the SQUID. b, Transmittance ($|S_{21}|^2$) of the qubit feedline at varying driving powers. At low power, the resonance indicates the qubit is strongly overcoupled to the feedline. Solid lines represent a global complex fit. c, Power dependence of the feedline transmittance on resonance with the qubit. The solid line indicates the fit. d, Signal line attenuation extracted across the 4 to 8 GHz qubit tuning range, utilizing the calibrated power at the qubit. The attenuation exhibits a linear frequency dependence in dB (solid line indicates linear fit, shaded blue area indicates $\pm$1 dB deviation from the linear fit). e, Calibrated power spectral density (PSD) of a 5 GHz signal, measured with the parametric pump turned on and off. f, Calibrated added noise ($n_{add}$) in units of photons (solid orange points), with the shaded region representing the measurement uncertainty. The standard quantum limit (SQL, solid black line) accounts for the frequency-dependent parametric gain. The greyed-out band indicates the frequency range where the phase-matching stopband on either the signal or idler inhibits amplification. g, Signal-to-noise ratio improvement (SNRi), determined by subtracting the baseline transmission through a bypass cable from the amplified transmission.
• Figure 4: Power handling of the amplifiera, Small signal gain and signal-to-noise ratio improvement versus pump power at the input of the amplifier. b, Measured powers of different intermodulation products of two tones centered at 5GHz and separated by 5MHz versus their power at the input of the amplifier. c, Amplifier 1dB compression point for various signal frequencies within the operational range of frequencies d, 5$^o$ phase distortion point, i.e. power at which the phase of the amplified signal is shifted 5$^o$ with respect to the phase of a small signal at the same frequency.