betaFeedback

WorksPhysics and Astronomy

Photon-Counting Readout via Electro-Optic Microwave-to-Optical Transduction for Axion Haloscopes

Abstract

Axion dark matter searches require highly sensitive detection schemes with high scan rate to efficiently explore the parameter space. Building upon the conventional Sikivie haloscope, where axions convert into microwave photons inside a magnetic-field-immersed resonant cavity, we present a photon-counting scheme based on an electro-optic microwave-to-optical transducer. Through a comprehensive theoretical model under realistic experimental constraints, we find that the optimal operating regime employs an overcoupled microwave cavity, trading peak transduction efficiency for reduced transducer noise. Furthermore, combining parameters already demonstrated in existing devices, in a device with optimized extrinsic coupling rates, would yield a scan-rate enhancement of 5-10 relative to quantum-limited homodyne detection. Improvements in quality factors, coupling rates, and thermal management push this enhancement beyond 50. While our quantitative model is benchmarked at 9 GHz, the scheme adapts well to haloscopes across the 5-12 GHz range, where shrinking cavity volumes severely degrade the achievable scan rate. These results demonstrate the potential of electro-optical photon detection as a high-speed detection technique for axion dark matter searches.

Introduction

Dark matter is the non-luminous, non-absorbing matter that constitutes approximately 84.4% of the universe’s matter and 26.4% of its energy density [1]. Its gravitational effects are essential for understanding cosmic evolution and structure formation. The evidence for dark matter is overwhelming [2, 3, 4, 5, 6, 7], yet its fundamental nature remains unknown. Among the many proposed candidates, including weakly interacting massive particles (WIMPs), sterile neutrinos, axions, axion-like particles, and dark photons [1], axions stand out as particularly compelling. Beyond being a viable dark matter candidate, axions address the long-standing strong CP problem in quantum chromodynamics (QCD): why does the neutron have an electric dipole moment close to zero, despite nonzero values being allowed in the QCD Lagrangian? [8, 9, 10]

Axions could be detected through their coupling to electromagnetism. A broad class of axion search experiments is designed to convert axions into photons within a strong static magnetic field. If this conversion happens inside a resonant cavity, whose resonant frequency fcf_c matches the axion mass through hfc≈mac2h f_c \approx m_a c^2, the axion signal power can be resonantly enhanced known as a Sikivie haloscope [11]. The axion-photon coupling is parameterized by gaγγg_{a\gamma\gamma} and is constrained by the solution to the strong CP problem. Because the axion mass is unknown, haloscopes must be tunable to search a wide range of frequencies. In current experiments, the signal power is expected to be 10−2410^{-24}–10−2210^{-22} W. Detecting such a weak signal requires a strong magnetic field, a high-QQ cavity in millikelvin temperatures, and low-noise quantum sensors. The scan rate is a crucial figure of merit and is related to various experimental parameters by

R∝B04Veff2QLTn−2,(1)R \propto B_0^4 V_{\mathrm{eff}}^2 Q_L T_n^{-2}, \tag*{(1)}

where B0B_0 is the external magnetic field, VeffV_{\mathrm{eff}} is the effective detection volume, QLQ_L is the loaded quality factor, and TnT_n is the system noise temperature. The Axion Dark Matter Experiment (ADMX) uses a 136L cavity operating at the TM010_{010} mode at 1 GHz to search for QCD axions with masses of a few μeV\mu\mathrm{eV} at a scan rate of 543 MHz/year [12]. However, higher masses lead to smaller volume cavities (V∝f−3V \propto f^{-3}), more metallic losses (QL∝f−2/3Q_L \propto f^{-2/3}), and increased noise from the Standard Quantum Limit (SQL) (Tn∝fT_n \propto f), causing the scan rate to scale as R∝f−8.66R \propto f^{-8.66}. The same ADMX haloscope scaled to 10 GHz would have a scan rate of R=1.2R = 1.2 Hz/year, which is untenably slow. For this reason, the QCD axion parameter space is largely unexplored above 1 GHz. Yet many theoretical predictions place the axion mass above 2 GHz, strongly motivating the exploration of this parameter space. It is imperative to develop detector and quantum sensing technologies to make this parameter space accessible 1.

Conventional axion haloscopes rely predominantly on microwave amplifiers and homodyne-based readout schemes [19, 20, 21, 22]. For a phase-preserving linear amplifier, the added quantum noise imposes a fundamental noise floor, corresponding to at least one input-referred noise photon [23, 24, 25, 26, 27]. Although single-mode squeezing has been developed to enhance the scan rate beyond the quantum limit, its practical performance is constrained by finite squeezing and degradation of the squeezed state during amplification and transmission [28, 29, 30]. Photon-counting detection provides a different route that can circumvent the added-noise constraint associated with amplification, since it measures photon number rather than field quadratures [31, 32]. In the microwave domain, however, photon-counting schemes are challenged by limited detection efficiency, bandwidth, and relatively high dark count rates [33, 34, 35]. Microwave-to-optical quantum transduction offers an alternative approach that could potentially overcome these limitations by transferring the microwave photons to the optical regime [36, 37, 38, 39, 40]. Once converted, the signal is read out using superconducting nanowire single-photon detectors (SNSPDs), which provide exceptional detection efficiency with an ultra-low dark count rate [18, 41].

In this work, we propose and analyze a readout scheme for axion haloscopes based on microwave-to-optical transduction followed by optical photon counting. We evaluate the scheme using an electro-optic transducer designed at 9 GHz as a benchmark, although it can be adapted to other haloscope frequencies by redesigning the transducer’s microwave resonator. Our theoretical model incorporates key experimental parameters of the transducer, including the single-photon coupling rate, the quality factors (intrinsic loss rates), microwave and optical extrinsic coupling rates, pump power, and repetition rate. Furthermore, the model accounts for noise contributions from the sensing cavity, the transducer microwave mode, and detector dark counts. We investigate the trade-off between the transduction efficiency and total noise, and optimize the extrinsic coupling coefficients to maximize the scan-rate enhancement with respect to homodyne-detection baseline. Under benchmark experimental conditions, we find that enhancements of E≈5−10E \approx5 - 10 are possible with parameters demonstrated in existing devices, operating with optimized extrinsic coupling rates. Furthermore, enhancements in excess of 50 are accessible when pushing experimental conditions to their limits. These results highlight the potential of low-loss, low-noise, and high-repetition-rate electro-optic transducers for rapid axion searches and provide a promising pathway toward high-speed axion experiments in high frequencies.

Axion signal model

In a haloscope setup, axions from the galactic dark matter halo are converted into microwave photons inside a resonant cavity permeated by a strong static magnetic field B0B_0, via the inverse Primakoff effect. In natural units, the expected signal power PP is given by [26]:

P=gaγγ2ρamaB02VeffQLββ+1L(Δ),(2)P = g_{a\gamma\gamma}^{2}\frac{\rho_a}{m_a}B_0^{2}V_{\mathrm{eff}}Q_L\frac{\beta}{\beta+1}\mathcal{L}(\Delta), \tag*{(2)}

where gaγγg_{a\gamma\gamma} is the axion-photon coupling constant, ρa≈0.45 GeV/cm3\rho_a \approx0.45\ \mathrm{GeV}/\mathrm{cm}^{3} is the local dark matter density, mam_a is the axion mass, VeffV_{\mathrm{eff}} is the effective cavity volume including the geometric form factor, QLQ_L is the loaded quality factor of the cavity and β=κext/κi\beta= \kappa_{\mathrm{ext}}/\kappa_i is the coupling coefficient of sensing cavity. Here, κext\kappa_{\mathrm{ext}} is the extrinsic coupling rate, κi\kappa_i is the intrinsic loss rate and L(Δ)\mathcal{L}(\Delta) is the Lorentzian lineshape function given by

L(Δ)=11+4Δ2,(3)\mathcal{L}(\Delta) = \frac{1}{1+4\Delta^{2}}, \tag*{(3)}

where the dimensionless frequency detuning is defined as Δ=(fa−fc)/Δfc\Delta= (f_a-f_c)/\Delta f_c. Here, faf_a and fcf_c denote the axion frequency and the cavity’s resonance frequency, respectively, while Δfc\Delta f_c represents the cavity’s bandwidth. Due to the virialized velocity dispersion of dark matter in the galactic halo, the axion signal exhibits a characteristic Maxwell-Boltzmann-like lineshape centered at frequency fa=mac2/hf_a = m_a c^2/h, with a relative bandwidth of Δfa/fa∼10−6\Delta f_a/f_a \sim10^{-6} (corresponding to a quality factor of Qa≡fa/Δfa=106Q_a \equiv f_a/\Delta f_a = 10^6) [42]. To incorporate this weak signal into our quantum input-output framework, we describe the axion as a coherent drive in sensing cavity with annihilation operator b^(t)\hat{b}(t). The intracavity field operator a^(t)\hat{a}(t) evolves according to [23, 43]

da^dt=−iωca^−κtotal2a^−κexta^in−κia^i−iκab^,(4)\frac{d\hat{a}}{dt}=-i\omega_c\hat{a}-\frac{\kappa_{\mathrm{total}}}{2}\hat{a}-\sqrt{\kappa_{\mathrm{ext}}}\hat{a}_{\mathrm{in}}-\sqrt{\kappa_i}\hat{a}_i-i\kappa_a\hat{b}, \tag*{(4)}

where ωc=2πfc\omega_c = 2\pi f_c is the angular resonance frequency of the cavity, κtotal=κext+κi\kappa_{\mathrm{total}} = \kappa_{\mathrm{ext}}+\kappa_i represents the total cavity decay rate, and κa\kappa_a is the axion-photon coupling rate. The operators a^in(t)\hat{a}_{\mathrm{in}}(t) and a^i(t)\hat{a}_i(t) denote the input vacuum and noise field from the transmission line and the intrinsic bath noise, respectively. Solving for the steady state (see Appendix A), we obtain the output state from the sensing cavity as a Gaussian state with mean vector and covariance matrix given by

dout=2na(cos⁡ϕa, sin⁡ϕa),(5)\mathbf{d}_{\mathrm{out}}=\sqrt{2n_a}\left(\cos\phi_a,\ \sin\phi_a\right), \tag*{(5)}
Vout=(12+4β(1+β)2L(Δ)ns)I,(6)\mathbf{V}_{\mathrm{out}}=\left(\frac{1}{2}+\frac{4\beta}{(1+\beta)^2}\mathcal{L}(\Delta)n_s\right)\mathbf{I}, \tag*{(6)}

where na=P/(ℏωaΔfa)n_a = P/(\hbar\omega_a\Delta f_a) is the photon number of axion, ϕa\phi_a is the phase of the axion-induced microwave field and nsn_s is the thermal photon number of sensing cavity.

Transducer-based axion sensing

In this section, we present a comprehensive theoretical analysis of axion readout scheme based on the microwave-to-optical transduction and benchmark its performance with microwave homodyne detection. We trace the signal and noise propagation across the conversion and photon detection, and evaluate the resulting scan rate and the scan-rate enhancement relative to homodyne detection.

Figure 1. Schematic of the transducer-based axion sensing scheme. A strong magnetic field in the sensing cavity generates axion-induced microwave photons, which are routed to a transducer cavity and converted into optical photons through an optical pump and a nonlinear crystal for detection with a superconducting nanowire single-photon detector (SNSPD).

Signal and Noise Transfer

Assuming a steady-state conversion process, the optical output field operator a^out\hat{a}_{\mathrm{out}} can be obtained from the transducer’s dynamical equations, together with the corresponding input-output relations [44, 45]:

0=iΔoa^−igm^−γo2a^+γoea^in+γoia^i,(7)0 = i\Delta_{\mathrm{o}}\hat{a} - ig\hat{m} - \frac{\gamma_{\mathrm{o}}}{2}\hat{a} + \sqrt{\gamma_{\mathrm{oe}}}\hat{a}_{\mathrm{in}} + \sqrt{\gamma_{\mathrm{oi}}}\hat{a}_{\mathrm{i}}, \tag*{(7)}
0=iΔmm^−iga^−γm2m^+γmem^in+γmim^i,(8)0 = i\Delta_{\mathrm{m}}\hat{m} - ig\hat{a} - \frac{\gamma_{\mathrm{m}}}{2}\hat{m} + \sqrt{\gamma_{\mathrm{me}}}\hat{m}_{\mathrm{in}} + \sqrt{\gamma_{\mathrm{mi}}}\hat{m}_{\mathrm{i}}, \tag*{(8)}
a^out=−γoea^+a^in,(9)\hat{a}_{\mathrm{out}} = -\sqrt{\gamma_{\mathrm{oe}}}\hat{a} + \hat{a}_{\mathrm{in}}, \tag*{(9)}
m^out=−γmem^+m^in,(10)\hat{m}_{\mathrm{out}} = -\sqrt{\gamma_{\mathrm{me}}}\hat{m} + \hat{m}_{\mathrm{in}}, \tag*{(10)}

Here, we adopt a systematic subscript convention: the primary index (o or m) denotes the subsystem (optical or microwave), whereas the secondary index (c or i) denotes the extrinsic coupling and intrinsic loss channels, as summarized in Table II of Appendix B 2. Δo\Delta_{\mathrm{o}} and Δm\Delta_{\mathrm{m}} are the optical and microwave detunings with respect to the output signals. The total decay rate of each mode is the sum of these two contributions, γo=γoe+γoi\gamma_{\mathrm{o}} = \gamma_{\mathrm{oe}} + \gamma_{\mathrm{oi}} and γm=γme+γmi\gamma_{\mathrm{m}} = \gamma_{\mathrm{me}} + \gamma_{\mathrm{mi}}, and g=npgeog = \sqrt{n_{\mathrm{p}}}g_{\mathrm{eo}} is the pump-enhanced coupling rate, where npn_{\mathrm{p}} is the intracavity pump photon number and geog_{\mathrm{eo}} is the single-photon electro-optic coupling rate. From the derivations presented in Appendix B 2, the microwave-to-optical conversion efficiency ηT\eta_{\mathrm{T}} and the associated added noise naddn_{\mathrm{add}} are given by

SymbolDefinition
Optical mode
ωo,out\omega_{\mathrm{o,out}}Output optical frequency
γoe\gamma_{\mathrm{oe}}Extrinsic coupling rate
QoQ_{\mathrm{o}}Intrinsic quality factor
γoi\gamma_{\mathrm{oi}}Intrinsic loss rate, ωo/Qo\omega_{\mathrm{o}}/Q_{\mathrm{o}}
γo\gamma_{\mathrm{o}}Total damping rate, γoe+γoi\gamma_{\mathrm{oe}}+\gamma_{\mathrm{oi}}
Δo\Delta_{\mathrm{o}}Detuning, ωo,out−ωo\omega_{\mathrm{o,out}}-\omega_{\mathrm{o}}
Microwave mode
ωm\omega_{\mathrm{m}}Microwave mode frequency
ωm,out\omega_{\mathrm{m,out}}Output microwave mode frequency
γme\gamma_{\mathrm{me}}Extrinsic coupling rate
QmQ_{\mathrm{m}}Intrinsic quality factor
γmi\gamma_{\mathrm{mi}}Intrinsic loss rate, ωm/Qm\omega_{\mathrm{m}}/Q_{\mathrm{m}}
γm\gamma_{\mathrm{m}}Total damping rate, γme+γmi\gamma_{\mathrm{me}}+\gamma_{\mathrm{mi}}
Δm\Delta_{\mathrm{m}}Detuning, ωm,out−ωm\omega_{\mathrm{m,out}}-\omega_{\mathrm{m}}
Electro-optic coupling
geog_{\mathrm{eo}}Single-photon coupling rate
npn_pPump photon number
ggEffective coupling strength, npgeo\sqrt{n_p}g_{\mathrm{eo}}

Table II. Definitions of the parameters used in the transducer model.

ηT=ζoζm4C(1+C)2,(11)\eta_{\mathrm{T}} = \zeta_{\mathrm{o}}\zeta_{\mathrm{m}}\frac{4C}{(1+C)^2}, \tag*{(11)}
nadd=ζo(1−ζm)4C(1+C)2nm,(12)n_{\mathrm{add}} = \zeta_{\mathrm{o}}(1-\zeta_{\mathrm{m}})\frac{4C}{(1+C)^2}n_{\mathrm{m}}, \tag*{(12)}

where C=4g2/(γoγm)C = 4g^2/(\gamma_{\mathrm{o}}\gamma_{\mathrm{m}}) is the cooperativity, ζo=γoe/γo\zeta_{\mathrm{o}} = \gamma_{\mathrm{oe}}/\gamma_{\mathrm{o}} and ζm=γme/γm\zeta_{\mathrm{m}} = \gamma_{\mathrm{me}}/\gamma_{\mathrm{m}} are extraction efficiencies of the optical and microwave modes, respectively, and nmn_{\mathrm{m}} denotes the effective thermal occupancy of the transducer’s microwave mode. The quantum state by quantum transduction is given by

dT=2ηDηTna(cos⁡ϕa,sin⁡ϕa),(13)\mathbf{d}_{\mathrm{T}} = \sqrt{2\eta_{\mathrm{D}}\eta_{\mathrm{T}}n_{\mathrm{a}}}(\cos\phi_{\mathrm{a}},\sin\phi_{\mathrm{a}}), \tag*{(13)}
VT=(12+nT)I,(14)\mathbf{V}_{\mathrm{T}} = \left(\frac{1}{2}+n_{\mathrm{T}}\right)\mathbf{I}, \tag*{(14)}

where the transducer noise photon nTn_{\mathrm{T}} is given by

nT=ζo4C(1+C)2[(1−ζm)nm+ζm4β(1+β)2ns].(15)n_{\mathrm{T}} = \zeta_{\mathrm{o}}\frac{4C}{(1+C)^2}\left[(1-\zeta_{\mathrm{m}})n_{\mathrm{m}}+\zeta_{\mathrm{m}}\frac{4\beta}{(1+\beta)^2}n_{\mathrm{s}}\right]. \tag*{(15)}

Scan-Rate Enhancement Relative to Homodyne Detection

The ultimate performance of a dark matter haloscope experiment is governed by its capability to scan a broad frequency range for axion masses. Having established the signal and noise transfer of the transducer-based readout chain, we now formulate the signal-to-noise ratio (SNR), the scan rate, and its performance enhancement relative to microwave homodyne detection.

A detection event occurs when one or more photons are registered on the SNSPD. For a quantum transducer operating at a photon-counting measurement rate frf_{\mathrm{r}}, a total of N=tintfrN = t_{\mathrm{int}}f_{\mathrm{r}} individual photon detection cycles are performed over an integration time tintt_{\mathrm{int}}. For a pulsed pump, frf_{\mathrm{r}} is the repetition rate, whereas for a continuous pump, fr=Δfaf_{\mathrm{r}} = \Delta f_{\mathrm{a}}. As presented in Appendix § 10 2, the SNR is given by

SNR=ηDηTnaηDnT+nDtintfr,(16)\mathrm{SNR} = \frac{\eta_{\mathrm{D}}\eta_{\mathrm{T}}n_{\mathrm{a}}}{\sqrt{\eta_{\mathrm{D}}n_{\mathrm{T}}+n_{\mathrm{D}}}}\sqrt{t_{\mathrm{int}}f_{\mathrm{r}}}, \tag*{(16)}

where ηD\eta_{\mathrm{D}} is the efficiency of the single photon detector, nDn_{\mathrm{D}} is the dark count photon number per detection window. We next consider a frequency scan over a total range Δfscan\Delta f_{\mathrm{scan}}, discretized into N2N_2 frequency steps with a step size δ<Δfc\delta< \Delta f_{\mathrm{c}}. Following the derivation detailed in Appendix § 10 2, the resulting scan rate RTR_{\mathrm{T}} required to achieve a target total signal-to-noise ratio SNRtotal\mathrm{SNR}_{\mathrm{total}} is given by

RT=π4SNRtotal2ηD2ηT2na,02ηDnT+nDΔfcfr,(17)R_{\mathrm{T}} = \frac{\pi}{4\mathrm{SNR}_{\mathrm{total}}^2}\frac{\eta_{\mathrm{D}}^2\eta_{\mathrm{T}}^2n_{\mathrm{a},0}^2}{\eta_{\mathrm{D}}n_{\mathrm{T}}+n_{\mathrm{D}}}\Delta f_{\mathrm{c}}f_{\mathrm{r}}, \tag*{(17)}

where na,0n_{\mathrm{a},0} is the peak axion-induced photon number na(Δ=0)n_{\mathrm{a}}(\Delta= 0). For conventional microwave homodyne detection, the SNR and scan rate are derived in Appendix C 1 as:

SNR=na2nin,equivtintΔν,(18)\mathrm{SNR} = \frac{n_{\mathrm{a}}}{\sqrt{2}n_{\mathrm{in,equiv}}}\sqrt{t_{\mathrm{int}}\Delta\nu}, \tag*{(18)}
RH=π8SNRtotal2na,02npeak3/2nbase1/2ΔfcΔν,(19)R_{\mathrm{H}} = \frac{\pi}{8\mathrm{SNR}_{\mathrm{total}}^2}\frac{n_{\mathrm{a},0}^2}{n_{\mathrm{peak}}^{3/2}n_{\mathrm{base}}^{1/2}}\Delta f_{\mathrm{c}}\Delta\nu, \tag*{(19)}

where Δν\Delta\nu is the resolution bandwidth of the spectrum analyzer, and nin,equivn_{\mathrm{in,equiv}} is the input-equivalent noise photon number of system given by Friis formula (See Eq. (66) of Appendix B 1)

nin,equiv=1+nJPA+nHEMTGJPA+⋯⏟nbase+4β(1+β)2nsL(Δ)≡nbase+(npeak−nbase)L(Δ),(20)\begin{aligned} n_{\mathrm{in,equiv}} = \underbrace{1+n_{\mathrm{JPA}}+\frac{n_{\mathrm{HEMT}}}{G_{\mathrm{JPA}}}+\cdots}_{n_{\mathrm{base}}}+\frac{4\beta}{(1+\beta)^2}n_{\mathrm{s}}\mathcal{L}(\Delta) \\ &\equiv n_{\mathrm{base}}+(n_{\mathrm{peak}}-n_{\mathrm{base}})\mathcal{L}(\Delta), \tag*{(20)} \end{aligned}

where npeakn_{\mathrm{peak}} and nbasen_{\mathrm{base}} are the peak and base noise photon numbers, respectively. Here, nJPAn_{\mathrm{JPA}} and nHEMTn_{\mathrm{HEMT}} denote the noise contributions from the Josephson parametric amplifier (JPA) and the high electron mobility transistor (HEMT) amplifier, respectively, with GJPAG_{\mathrm{JPA}} being the JPA gain. Combining Eq. (17) and Eq. (19), the enhancement EE in scan rate with the same total SNR is given by:

E≡RTRH=2ηD2ηT2npeak3/2nbase1/2ηDnT+nDfrΔν.(21)E \equiv\frac{R_{\mathrm{T}}}{R_{\mathrm{H}}}=\frac{2\eta_{\mathrm{D}}^2\eta_{\mathrm{T}}^2n_{\mathrm{peak}}^{3/2}n_{\mathrm{base}}^{1/2}}{\eta_{\mathrm{D}}n_{\mathrm{T}}+n_{\mathrm{D}}}\frac{f_{\mathrm{r}}}{\Delta\nu}. \tag*{(21)}

Performance Analysis

Evaluating the performance enhancement requires knowledge of both the transducer efficiency and noise, both of which depend on the cooperativity CC. In this section, we first express the cooperativity in terms of the intrinsic cooperativity C0C_0 and the tunable coupling coefficients (βo,βm)(\beta_{\mathrm{o}},\beta_{\mathrm{m}}). Then we summarize the transducer’s device and noise parameters at 9 GHz across a realistic range of values in Table I. After setting up the benchmark parameters, we maximize the enhancement by optimizing the tunable coupling coefficients (βo,βm)(\beta_{\mathrm{o}},\beta_{\mathrm{m}}) and characterize the trade-off between the transducer efficiency and the noise governed by the microwave coupling coefficient βm\beta_{\mathrm{m}}. Furthermore, we explore the scan-rate enhancement of transducer-based photon counting over conventional homodyne detection across key operational parameters, including the single-photon coupling rate geog_{\mathrm{eo}}, the microwave quality factor QmQ_{\mathrm{m}}, the pump power PpumpP_{\mathrm{pump}}, and the thermal noise TmT_{\mathrm{m}} (nmn_{\mathrm{m}}). Our analysis considers two operating schemes of quantum transducer: continuous-wave pumping, which provides low added noise but is limited by the transduction efficiency, and pulsed pumping, which offers higher efficiency at the cost of increased noise and a lower repetition rate.

ParameterSymbolValue
Axion Sensing Cavity
Sensing resonance frequencyωc/(2π)\omega_{\mathrm{c}}/(2\pi)9.0 GHz
Loaded quality factorQLQ_{\mathrm{L}}10510^{5}
Axion quality factorQaQ_{\mathrm{a}}10610^{6}
extrinsic coupling ratioβ\beta2.0
Cavity temperatureTs (ns)T_{\mathrm{s}}\ (n_{\mathrm{s}})50 mK
Electro-Optic Transducer
Pump frequencyωp/(2π)\omega_{\mathrm{p}}/(2\pi)192.421 THz
Optical resonance frequencyωo/(2π)\omega_{\mathrm{o}}/(2\pi)192.43 THz
Microwave resonance frequencyωm/(2π)\omega_{\mathrm{m}}/(2\pi)9.0 GHz
Optical quality factorQoQ_{\mathrm{o}}10710^{7}
Microwave quality factorQmQ_{\mathrm{m}}10310^{3}–10510^{5}
Single-photon coupling rategeo/(2π)g_{\mathrm{eo}}/(2\pi)40–140 Hz
Pump detuningΔp/(2π)\Delta_{\mathrm{p}}/(2\pi)10 MHz
Continuous-wave operation
Pump powerPpumpP_{\mathrm{pump}}100–800 μ\muW
Transducer temperatureTm (nm)T_{\mathrm{m}}\ (n_{\mathrm{m}})100–300 mK
Photon counting measurement rateΔfa\Delta f_{\mathrm{a}}9.0 kHz
Pulsed operation
Pump powerPpumpP_{\mathrm{pump}}50–300 mW
Microwave noise photonnmn_{\mathrm{m}}0.01–0.1
Intrinsic cooperativityC0C_{0}1–10
Repetition ratefrf_{\mathrm{r}}100 Hz
Single Photon Detector
Single-photon detection efficiencyηD\eta_{\mathrm{D}}0.98
Dark photon counting raterDr_{\mathrm{D}}1 Hz
Homodyne Detection
Frequency bin resolutionΔν\Delta\nu3 kHz
Total temperatureTHom (nHom)T_{\mathrm{Hom}}\ (n_{\mathrm{Hom}})100 mK

Table I. Benchmark parameters for the transducer-based axion haloscope. Values are chosen based on experimental capabilities for searches near 9 GHz

Parametric Characterization of Cooperativity

The overall system performance is dictated by the conversion efficiency ηT\eta_{\mathrm{T}} and noise nTn_{\mathrm{T}}. A critical engineering trade-off arises from the dual role of the extrinsic coupling rates (γoe,γme)(\gamma_{\mathrm{oe}},\gamma_{\mathrm{me}}). High extrinsic coupling rates are essential extracting the converted signals. On the other hand, increasing the extrinsic coupling also increases the total damping rate (γo,γm)(\gamma_{\mathrm{o}},\gamma_{\mathrm{m}}), which in turn degrades the cooperativity since C∝1/(γoγm)C \propto1/(\gamma_{\mathrm{o}}\gamma_{\mathrm{m}}). We parameterize the cooperativity CC using the tunable coupling coefficients (βo,βm)(\beta_{\mathrm{o}},\beta_{\mathrm{m}}) and the intrinsic cooperativity C0C_0, which depends on the intra-cavity pump photon number npn_{\mathrm{p}}, single-photon coupling rate and intrinsic loss rates. Using standard input-output theory, npn_{\mathrm{p}} is directly related to the input pump power PpP_{\mathrm{p}}.

For electro-optic transducers, the intrinsic loss rates are set by the intrinsic quality factors of the two resonators, γoi=ωo/Qo\gamma_{\mathrm{oi}}=\omega_{\mathrm{o}}/Q_{\mathrm{o}} and γmi=ωm/Qm\gamma_{\mathrm{mi}}=\omega_{\mathrm{m}}/Q_{\mathrm{m}}, where ωo,m\omega_{\mathrm{o,m}} and Qo,mQ_{\mathrm{o,m}} are the operating angular frequencies and the intrinsic quality factor of the optical and microwave modes, respectively. These parameters are determined by material absorption, surface scattering, and fabrication quality, and are independent of the optical pump during normal operation, except in high-power regimes where pump-induced losses will arise. In contrast, the extrinsic coupling rates (γoe,γme)(\gamma_{\mathrm{oe}},\gamma_{\mathrm{me}}) can be engineered and optimized, for instance, by designing the waveguide-cavity spacing or antenna geometry. We parameterize the tunable extrinsic coupling rates using the optical and microwave coupling coefficient defined by βo≡γoe/γoi\beta_{\mathrm{o}}\equiv\gamma_{\mathrm{oe}}/\gamma_{\mathrm{oi}} and βm≡γme/γmi\beta_{\mathrm{m}}\equiv\gamma_{\mathrm{me}}/\gamma_{\mathrm{mi}}. Consequently, the extraction efficiency is expressed as ζo=βo/(1+βo)\zeta_{\mathrm{o}}=\beta_{\mathrm{o}}/(1+\beta_{\mathrm{o}}). We define the intrinsic cooperativity

C0=4geo2npγoiγmi.(22)C_0=\frac{4g_{\mathrm{eo}}^2n_{\mathrm{p}}}{\gamma_{\mathrm{oi}}\gamma_{\mathrm{mi}}}. \tag*{(22)}

The electro-optic cooperativity is expressed as:

C=4geo2npγoγm=C0(1+βm)(1+βo).(23)C=\frac{4g_{\mathrm{eo}}^2n_{\mathrm{p}}}{\gamma_{\mathrm{o}}\gamma_{\mathrm{m}}}=\frac{C_0}{(1+\beta_{\mathrm{m}})(1+\beta_{\mathrm{o}})}. \tag*{(23)}

Electro-optic transducers are commonly operated with pulsed optical pumps. The cooperativity CC is proportional to pump photon number npn_{\mathrm{p}}, whereas heating and thus noise nmn_{\mathrm{m}} follows the average power. A low duty cycle therefore buys high cooperativity within fixed cooling budget, a strategy applied in bulk [38, 39, 46] and thin-film [47, 48] lithium niobate as well as in aluminum nitride [49]. For example, in the high-cooperativity regime demonstrated in Ref. [38], the intrinsic cooperativity reaches C0≈4C_0\approx4 for C=1C=1, with ζo=0.58\zeta_{\mathrm{o}}=0.58 and ζm=0.408\zeta_{\mathrm{m}}=0.408. Consequently, we consider a realistic range of intrinsic cooperativity C0∈[1,10]C_0\in[1,10] for our performance analysis. Both in the steady-state regime (under continuous pumping) and the quasi-steady-state regime (where the pulse duration exceeds the cavity ring-up time), the intrinsic cooperativity C0C_0 can be further parameterized by the single-photon coupling rate geog_{\mathrm{eo}} and the pump power PpumpP_{\mathrm{pump}}. The single-photon coupling rate geog_{\mathrm{eo}} is determined by the field overlap between the microwave mode and optical mode. Based on the steady-state input-output relation, for a given pump power PpumpP_{\mathrm{pump}}, the intra-cavity pump photon number is expressed as [40]

np=γoeΔp2+(γo/2)2Ppumpℏωp,(24)n_{\mathrm{p}}=\frac{\gamma_{\mathrm{oe}}}{\Delta_{\mathrm{p}}^{2}+(\gamma_{\mathrm{o}}/2)^{2}}\frac{P_{\mathrm{pump}}}{\hbar\omega_{\mathrm{p}}}, \tag*{(24)}

where Δp\Delta_{\mathrm{p}} is the pump detuning and ωp\omega_{\mathrm{p}} is the angular frequency of the pump field. Therefore, given the intrinsic loss rates (γoi,γmi)(\gamma_{\mathrm{oi}},\gamma_{\mathrm{mi}}), single-photon coupling rate geog_{\mathrm{eo}}, pump power PpumpP_{\mathrm{pump}}, and their respective resonance frequencies, the resulting cooperativity CC depends on the tunable coupling coefficients (βo,βm)(\beta_{\mathrm{o}},\beta_{\mathrm{m}}) to be optimized.

Enhancement Regime for Electro-Optic Transducers

To evaluate the practical performance of the transducer-based axion haloscope, we consider an axion search targeting a frequency around 9 GHz. Table I summarizes the benchmark parameters in this subsection.

The axion signal is evaluated at a benchmark frequency of fa=9.0f_{\mathrm{a}}=9.0 GHz with a bandwidth of Δfa=fa/Qa=9.0\Delta f_{\mathrm{a}}=f_{\mathrm{a}}/Q_{\mathrm{a}}=9.0 kHz. The effective temperature of the sensing cavity has been reported to be below 60 mK in Ref. [30]. We adopt Ts=50T_{\mathrm{s}}=50 mK, within this range and consistent with the temperatures routinely achieved in dilution refrigerators hosting haloscope cavities. The thermal photon number in the sensing cavity is then determined by the Bose-Einstein distribution:

ns=1eℏωckBTs−1.(25)n_{\mathrm{s}}=\frac{1}{e^{\frac{\hbar\omega_{\mathrm{c}}}{k_{\mathrm{B}}T_{\mathrm{s}}}}-1}. \tag*{(25)}

For conventional homodyne detection, the effective noise photon number for standard microwave homodyne detection is defined by

nHom≡nJPA+nHEMTGJPA+⋯ .(26)n_{\mathrm{Hom}}\equiv n_{\mathrm{JPA}}+\frac{n_{\mathrm{HEMT}}}{G_{\mathrm{JPA}}}+\cdots. \tag*{(26)}

The dominant contribution comes from the JPA. We evaluate nHomn_{\mathrm{Hom}} (i.e., nbasen_{\mathrm{base}} in Eq. (20)) assuming a total effective noise temperature of 100 mK for the microwave components. The frequency resolution Δν\Delta\nu is typically chosen to be 1/3 to 1/5 of the axion bandwidth Δfa=9.0\Delta f_{\mathrm{a}}=9.0 kHz. We set Δν=3\Delta\nu=3 kHz for homodyne detection.

Continuous-wave pump operation

For a continuous-wave (CW) electro-optic transducer operating in steady-state conversion, the total number of modes is given by the time-bandwidth product Δfatint\Delta f_{\mathrm{a}}t_{\mathrm{int}} over an integration time tintt_{\mathrm{int}}. Thus the photon-counting

measurement rate fr=Δfaf_{\mathrm{r}}=\Delta f_{\mathrm{a}}. The average dark count per mode is given by nD=1/Δfan_{\mathrm{D}}=1/\Delta f_{\mathrm{a}} for a dark count rate of rD=1r_{\mathrm{D}}=1 Hz. A pump power in the order 100 μ\muW yields lower added noise (i.e., the effective temperature TmT_{\mathrm{m}}) but also a lower transduction efficiency ηT\eta_{\mathrm{T}} than pulsed operation [46, 47, 48]. Pump detuning (Δp)(\Delta_{\mathrm{p}}) is essential to mitigate photothermal instabilities and optical bistability caused by high-power pump absorption. Accordingly, we set the pump detuning to Δp=10\Delta_{\mathrm{p}}=10 MHz as a representative experimental benchmark [40]. The optical quality factor is set to 10710^{7}. The microwave quality factor is 10310^{3}–10410^{4} for on-chip resonators and projected 10410^{4}–10510^{5} for state-of-the-art 3D hybrid microwave cavities [40, 50, 51]. In steady-state conversion, the relation between nmn_{\mathrm{m}} and TmT_{\mathrm{m}} follows the Bose-Einstein distribution in Eq. (25). The steady-state input-output relation directly establishes the relation between the input pump power PpumpP_{\mathrm{pump}} and the intra-cavity photon number, uniquely determining C0C_{0} and CC. In contrast, pulsed operation outside the quasi-steady-state regime requires solving the time-dependent intra-cavity dynamics or utilizing an effective C0C_{0}.

With Tm=100T_{\mathrm{m}}=100 mK, geo/(2π)=50g_{\mathrm{eo}}/(2\pi)=50 Hz, Qm=5×104Q_{\mathrm{m}}=5\times10^{4} and Ppump=400P_{\mathrm{pump}}=400 μ\muW, optimal choices of (βm,βo)(\beta_{\mathrm{m}},\beta_{\mathrm{o}}) emerge from the trade-off between the overall transduction efficiency ηT\eta_{\mathrm{T}} and the noise nTn_{\mathrm{T}}. Through numerical simulations, we find a peak enhancement of E=5.13E = 5.13 at βo=1.23\beta_{\mathrm{o}} = 1.23 and βm=1.89\beta_{\mathrm{m}} = 1.89, with a transducer efficiency of ηT=1.34%\eta_{\mathrm{T}} = 1.34\%. With βo\beta_{\mathrm{o}} fixed at its optimum, we show the trade-off versus βm\beta_{\mathrm{m}} in Figure 2. Existing microwave-to-optical transducers were designed for quantum networking and qubit readout, where the goal is to maximize the conversion efficiency. They therefore operate near critical microwave coupling. The axion photon-counting scheme analyzed here instead favors a strongly overcoupled microwave port. We observe that a scan-rate enhancement is obtained at a relatively low transduction efficiency. This is because photon counting is limited by noise photon nTn_{\mathrm{T}} rather than by the standard quantum limit (nin,equiv≥1n_{\mathrm{in,equiv}} \ge1), i.e., a weak signal can still be detected with a favorable signal-to-noise ratio.

Trade-off between transduction efficiency and noise in the continuous-wave regime

Figure 2. Trade-off between transduction efficiency and noise in the continuous-wave regime. Transduction efficiency ηT\eta_{\mathrm{T}} (red solid), transducer noise nTn_{\mathrm{T}} (blue dashed), and scan-rate enhancement EE (black solid) versus the microwave coupling ratio βm\beta_{\mathrm{m}}, with the optical coupling fixed at its optimal value βo=1.23\beta_{\mathrm{o}} = 1.23. Each curve is normalized to its own maximum (ηTmax⁡=1.47%\eta_{\mathrm{T}}^{\max} = 1.47\%, nTmax⁡=3.5×10−4n_{\mathrm{T}}^{\max} = 3.5 \times10^{-4}, Emax⁡=5.13E^{\max} = 5.13). Vertical dashed lines mark the maxima of ηT\eta_{\mathrm{T}} (βm=1.03\beta_{\mathrm{m}} = 1.03) and of EE (βm=1.89\beta_{\mathrm{m}} = 1.89); the shaded band indicates the displacement between them. Here, Tm=100T_{\mathrm{m}} = 100 mK and Ppump=400 μWP_{\mathrm{pump}} = 400\,\mu\mathrm{W}, geo/(2π)=50g_{\mathrm{eo}}/(2\pi) = 50 Hz, Qm=5×104Q_{\mathrm{m}} = 5 \times10^{4}.

Beyond these benchmark values, we evaluate the enhancement EE, maximized over (βm,βo)(\beta_{\mathrm{m}}, \beta_{\mathrm{o}}) for the parameters listed in Table I. Figure 3(a) displays the enhancement over a range of single-photon coupling rate geog_{\mathrm{eo}} and microwave quality factor QmQ_{\mathrm{m}}, while Figure 3(b) shows the enhancement over a range of transducer temperature TmT_{\mathrm{m}} and pump power PpumpP_{\mathrm{pump}}, with EE optimized over (βm,βo)(\beta_{\mathrm{m}}, \beta_{\mathrm{o}}) at each point. The 3D microwave cavity transducer design in Ref. [40] features quality factors of 10410^{4}–10510^{5} with a relatively small single-photon coupling rate of geo/(2π)≈46g_{\mathrm{eo}}/(2\pi) \approx46 Hz, placing its performance in the bottom-right regime of Figure 3(a) with E=11E = 11. Conversely, on-chip thin-film lithium niobate transducers experiments in Ref. [48] offer a complementary route, exhibiting a much larger single-photon coupling rate (geo/(2π)≈945g_{\mathrm{eo}}/(2\pi) \approx945 Hz) but lower quality factors (10210^{2}–10310^{3}). Consequently, such devices operate in the upper-left regime of Figure 3(a), yielding E=12E = 12 at Qm=2.5×102Q_{\mathrm{m}} = 2.5 \times10^{2}. Although Figure 3(b) suggests that both a higher pump power and a lower transducer temperature increase the enhancement, these two parameters are not independent. A larger pump power raises the effective mode temperature through optical absorption and quasiparticle generation, with the precise dependence determined by the thermal anchoring of the device. In practice, current on-chip demonstrations are limited by fiber-to-chip losses and by the pump power they can sustain.

Figure 3. Scan-rate enhancement compared to standard homodyne detection under continuous-wave pumping. Dashed lines: the corresponding enhancement EE. (a) Contour map of the optimal enhancement as a function of the geog_{\mathrm{eo}} and QmQ_{\mathrm{m}}, at a fixed temperature Tm=100T_{\mathrm{m}} = 100 mK and pump power Ppump=400 μWP_{\mathrm{pump}} = 400\,\mu\mathrm{W}. (b) Optimal enhancement versus TmT_{\mathrm{m}} and PpumpP_{\mathrm{pump}}, at a fixed Qm=5×104Q_{\mathrm{m}} = 5 \times10^{4} and geo/(2π)=90g_{\mathrm{eo}}/(2\pi) = 90 Hz.

Pulsed pump operation

While CW operation is constrained by steady-state heating, pulsed pumping allows high pump photon number for high cooperativity. Pulsed operation relaxes the thermal budget of the device: a peak pump power of ∼100\sim100 mW, applied at a duty cycle of 1–5%, dissipates only ∼1\sim1 mW of average optical power at the mixing chamber while maintaining a high intra-cavity pump photon number npn_{\mathrm{p}}, sufficient for larger cooperativity C≥1C \ge1 and hence near unity internal efficiency ηint=4C/(1+C)2\eta_{\mathrm{int}} = 4C/(1 + C)^2. Pump-induced heating means the mode occupancy relaxes only partially within one cycle and does not reach its steady-state (∼100\sim100 mK) value. In experiments, the thermal photon number nmn_{\mathrm{m}} is typically on the order of 0.1 at frequencies around 3–9 GHz [38, 39, 46, 47, 48, 49, 52, 53]. Specifically, evaluated at 9 GHz, the effective input noise photon number falls within the range nm=0.01n_{\mathrm{m}} = 0.01–0.1. Since the bandwidth of transducer (∼1\sim1–10 MHz) is much larger than the signal bandwidth Δfa\Delta f_{\mathrm{a}}, the transducer responds essentially instantaneously. The pulse duration and detection time window is 1/Δfa≈111 μs1/\Delta f_{\mathrm{a}} \approx111\,\mu\mathrm{s}, with dark photon number nD=1/Δfan_{\mathrm{D}} = 1/\Delta f_{\mathrm{a}} for 1 Hz dark photon counting rate. For the numerical evaluation for pulsed pump operation, we fix the dark photon number nDn_{\mathrm{D}} within this detection window and the repetition rate at fr=100f_{\mathrm{r}} = 100 Hz.

We now repeat the optimization for pulsed operation,

using the parameters of Table I. With intrinsic cooperativity C0=4C_{0} = 4 and noise photon number nm=0.01n_{\mathrm{m}} = 0.01, Figure 4(a) shows the enhancement as a function of βm\beta_{\mathrm{m}}. We find a peak enhancement of E=10.1E = 10.1 at βo=1.41\beta_{\mathrm{o}} = 1.41 and βm=8.67\beta_{\mathrm{m}} = 8.67, for which the total transducer efficiency is ηT=26.2%\eta_{\mathrm{T}} = 26.2\%. The optimum again lies in the overcoupled regime. The large pump power lifts the internal efficiency, leaving more efficiency to trade away; the optimal βm\beta_{\mathrm{m}} therefore moves deeper into the overcoupled regime, where nTn_{\mathrm{T}} is further suppressed. Figure 4(b) shows the enhancement EE obtained by maximizing over the coupling ratios at each point of the (nm,C0)(n_{\mathrm{m}}, C_{0}) plane. The bulk-LiNbO3_{3} transducer of [38] reaches cooperativities up to C=0.92C = 0.92, corresponding to C0≈4C_{0} \approx4. For a repetition rate of 100 Hz, the enhancement ranges from E=1.7E = 1.7 for nm=0.1n_{\mathrm{m}} = 0.1 to E=10.1E = 10.1 for nm=0.01n_{\mathrm{m}} = 0.01. Such an occupancy has been demonstrated on the same platform at lower pump power in Ref. [53].

Trade-off between transduction efficiency and noise in the pulsed-pump regime, and contour map of scan-rate enhancement

Figure 4. (a) Trade-off between transduction efficiency and noise in the pulsed-pump regime. Transduction efficiency ηT\eta_{\mathrm{T}} (red solid), transducer noise nTn_{\mathrm{T}} (blue dashed), and scan-rate enhancement EE (black solid) versus the microwave coupling ratio βm\beta_{\mathrm{m}}, with the optical coupling fixed at its optimal value βo=1.41\beta_{\mathrm{o}} = 1.41. Each curve is normalized to its own maximum (ηTmax⁡=36.5%\eta_{\mathrm{T}}^{\max} = 36.5\%, nTmax⁡=5.3×10−3n_{\mathrm{T}}^{\max} = 5.3 \times10^{-3}, Emax⁡=10.1E^{\max} = 10.1). Here C0=4C_{0} = 4 and nm=0.01n_{\mathrm{m}} = 0.01. (b) Contour map of the enhancement as a function of the intrinsic cooperativity C0C_{0} and transducer noise nmn_{\mathrm{m}}, at a fixed repetition rate fr=100f_{\mathrm{r}} = 100 Hz.

Within the quasi-steady-state approximation, the cooperativity CC can be expressed in terms of the single-photon coupling rate geog_{\mathrm{eo}}, the pump power, and the intrinsic loss rates (i.e., the quality factors). Over the parameter ranges listed in Table I, we evaluate the scan-rate enhancement. Figure 5(a) maps the optimal enhancement over the (geo,Qm)(g_{\mathrm{eo}}, Q_{\mathrm{m}}) plane, while Figure 5(b) shows it over the (nm,Ppump)(n_{\mathrm{m}}, P_{\mathrm{pump}}) plane, with (βm,βo)(\beta_{\mathrm{m}}, \beta_{\mathrm{o}}) optimized independently at each point. The enhancement increases with a higher single-photon coupling rate geog_{\mathrm{eo}} and quality factor QmQ_{\mathrm{m}}, exceeding E=50E = 50 in Figure 5(a) when geo/(2π)=90g_{\mathrm{eo}}/(2\pi) = 90 Hz and Qm=5×104Q_{\mathrm{m}} = 5 \times10^{4}. Similarly, it exceeds 50 in Figure 5(b) when nm≲0.02n_{\mathrm{m}} \lesssim0.02 and Ppump≳250P_{\mathrm{pump}} \gtrsim250 mW.

Figure 5. Scan-rate enhancement compared to standard homodyne detection under pulsed pumping. Dashed lines: the corresponding enhancement EE. (a) Contour map of the optimal enhancement as a function of the geog_{\mathrm{eo}} and QmQ_{\mathrm{m}}, at a fixed noise photon nm=0.01n_{\mathrm{m}} = 0.01 and pump power Ppump=100P_{\mathrm{pump}} = 100 mW. (b) Optimal enhancement versus nmn_{\mathrm{m}} and PpumpP_{\mathrm{pump}}, at a fixed Qm=5×104Q_{\mathrm{m}} = 5 \times10^{4} and geo/(2π)=90g_{\mathrm{eo}}/(2\pi) = 90 Hz.

Discussion

We have shown that converting axion-induced microwave photons to the optical domain and counting them can substantially outperform conventional homodyne readout. The scan-rate enhancement originates from that photon counting scheme avoids the standard quantum limit of phase-preserving amplification. A central result of our analysis is that the optimal operating point is not the one that maximizes the transduction efficiency. Because the figure of merit is set by the balance between converted signal and transducer noise, the optimum lies in the overcoupled regime, at a microwave coupling ratio βm\beta_{\mathrm{m}} well above the value that maximizes ηT\eta_{\mathrm{T}}. The efficiency is traded away in exchange for a stronger suppression of the transducer noise. Current transduction experiments have not been developed for this specific application and operate in the near critical coupling regime to maximize efficiency. Still, the key ingredients for photon counting applications have been demonstrated on various platforms, both on-chip and 3D designs. The main challenge is to combine high cooperativity with low microwave occupancy, and optimal optical coupling to sustain this regime over long pump pulses.

Looking ahead, the scheme is attractive in high frequency, where broadband quantum-limited phase-preserving amplification remains challenging above 10 GHz. The electro-optic interaction itself is not a limitation in this range. The quantum transducers can be engineered to address microwave modes ranging from a few to ten GHz, provided that the microwave resonance matches the free spectral range of the optical cavity. The transducer bandwidth of 1–10 MHz set by the microwave linewidth then fully covers the signal linewidth. And the central conversion frequency can be tuned over 100 MHz by tuning the microwave cavity itself, allowing the transducer to follow the haloscope cavity as it is stepped.

While our model is specific to electro-optic transduction, the framework may be extended to other platforms, such as piezo-optomechanical transducers, by including the corresponding noise and efficiency contributions of the intermediate mechanical mode. Our proposed scheme could be further enhanced using two-mode squeezing input state and joint detection receiver as proposed in [54, 55, 56, 57]. Combining the readout with multi-cavity arrays or higher-order-mode cavities would allow the enhancement demonstrated here to multiplicatively compound with these architectures, translating into a massive boost in the scan rate and providing a concrete pathway toward high-speed axion searches at the MHz/day level in the 5–12 GHz range.

Acknowledgments

This manuscript has been authored by the Fermi Forward Discovery Group, LLC under Contract No. 89243024CSC000002 with the U.S. Department of Energy, Office of Science, Office of High Energy Physics. The authors acknowledge support from the DOE Office of Science Early Career Research Program, Office of High Energy Physics (HEP), Detector R&D and QIS programs, and the Fermilab Laboratory Directed R&D (LDRD) program. R.C. acknowledges support from the U.S. Department of Energy, Office of Science, National Quantum Information Science Research Centers, Superconducting Quantum Materials and Systems Center (SQMS).

Data Availability

The results presented in this work can be reproduced using the model and equations provided in the main text and Appendices, together with the parameters listed in Table I. The code used to generate the figures is available from the authors upon reasonable request.

Axion state in sensing cavity

The axion-induced field amplitude b(r,t)b(\mathbf{r},t) is expressed as

b(r,t)=12[b(r)e−iωat+b∗(r)eiωat],(27)b(\mathbf{r},t) = \frac{1}{2}\left[b(\mathbf{r})e^{-i\omega_{\mathrm{a}}t} + b^{*}(\mathbf{r})e^{i\omega_{\mathrm{a}}t}\right], \tag*{(27)}

where ωa\omega_{\mathrm{a}} is the eigenfrequency of the axion field. The coupling between the axion field and the electromagnetic field is described by the axion-photon interaction term in the Lagrangian

Laγγ=−14gaγγb(r,t)FμνF~μν=gaγγb(r,t)E⋅B,(28)\mathcal{L}_{a\gamma\gamma} = -\frac{1}{4}g_{a\gamma\gamma}b(\mathbf{r},t)F_{\mu\nu}\widetilde{F}^{\mu\nu} = g_{a\gamma\gamma}b(\mathbf{r},t)\mathbf{E}\cdot\mathbf{B}, \tag*{(28)}

where gaγγg_{a\gamma\gamma} is the axion-photon coupling constant, FμνF_{\mu\nu} is the electromagnetic field tensor, and F~μν\widetilde{F}^{\mu\nu} is its dual. Under the background magnetic field approximation where B≈B0(r)\mathbf{B} \approx\mathbf{B}_{0}(\mathbf{r}), keeping the first-order part of expression, the interaction Hamiltonian is given by

Hint≈−∫Vgaγγb(r,t)B0(r)⋅E(r,t) d3r.(29)H_{\mathrm{int}} \approx-\int_{V}g_{a\gamma\gamma}b(\mathbf{r},t)\mathbf{B}_{0}(\mathbf{r})\cdot\mathbf{E}(\mathbf{r},t)\,\mathrm{d}^{3}r. \tag*{(29)}

The cavity mode expansion and the standard second quantization in natural units can be written as

E(r,t)=iωa2(a^e−iωat−a^†eiωat)e(r),(30)\mathbf{E}(\mathbf{r},t)=i\sqrt{\frac{\omega_a}{2}}\left(\hat{a}e^{-i\omega_a t}-\hat{a}^{\dagger}e^{i\omega_a t}\right)\mathbf{e}(\mathbf{r}), \tag*{(30)}
b(r,t)=12ωaVa(b^e−iωat+b^†eiωat)ψ(r),(31)b(\mathbf{r},t)=\sqrt{\frac{1}{2\omega_a V_a}}\left(\hat{b}e^{-i\omega_a t}+\hat{b}^{\dagger}e^{i\omega_a t}\right)\psi(\mathbf{r}), \tag*{(31)}

where ∫∣ψ(r)∣2d3r=Va\int\lvert\psi(\mathbf{r})\rvert^2d^3r=V_a is the effective volume of the mode and ∫∣e(r)∣2d3r=1\int\lvert\mathbf{e}(\mathbf{r})\rvert^2d^3r=1. Applying the rotating-wave approximation by neglecting the rapidly oscillating counter-rotating terms (b^a^(\hat{b}\hat{a} and b^†a^†)\hat{b}^{\dagger}\hat{a}^{\dagger}), the interaction Hamiltonian can be simplified to

Hint=κa(b^†a^+b^a^†),(32)H_{\mathrm{int}}=\kappa_a\left(\hat{b}^{\dagger}\hat{a}+\hat{b}\hat{a}^{\dagger}\right), \tag*{(32)}

where the effective coupling strength κa\kappa_a is given by

κa=gaγγ21Va∫Vd3r B0(r)⋅e(r)ψ(r).(33)\kappa_a=\frac{g_{a\gamma\gamma}}{2}\sqrt{\frac{1}{V_a}}\int_V d^3r\,\mathbf{B}_0(\mathbf{r})\cdot\mathbf{e}(\mathbf{r})\psi(\mathbf{r}). \tag*{(33)}

Here, κa\kappa_a quantifies the effective spatial overlap governing the axion-photon conversion. In natural units, taking typical values of gaγγ∼10−15 GeV−1g_{a\gamma\gamma}\sim10^{-15}\,\mathrm{GeV}^{-1} and a background magnetic field of B0≈1.95×10−15 GeV2B_0\approx1.95\times10^{-15}\,\mathrm{GeV}^2 (corresponding to 10 T), the coupling rate evaluates to κa∼10−31 GeV\kappa_a\sim10^{-31}\,\mathrm{GeV} (approximately 10−6 s−110^{-6}\,\mathrm{s}^{-1}). From da^dt=−i[a^,Hint]\frac{d\hat{a}}{dt}=-i[\hat{a},H_{\mathrm{int}}], the coupling between the axion field b^\hat{b} and the cavity mode a^\hat{a} is given by −i[a^,Hint]=−iκab^-i[\hat{a},H_{\mathrm{int}}]=-i\kappa_a\hat{b}. Now, we consider a single-port (reflective) sensing cavity. The dynamics of the cavity mode a^(t)\hat{a}(t) is governed by the Heisenberg-Langevin equation in the time domain:

da^dt=−iωca^−κtotal2a^−κexta^in−κia^i−iκab^,(34)\frac{d\hat{a}}{dt}=-i\omega_c\hat{a}-\frac{\kappa_{\mathrm{total}}}{2}\hat{a}-\sqrt{\kappa_{\mathrm{ext}}}\hat{a}_{\mathrm{in}}-\sqrt{\kappa_i}\hat{a}_i-i\kappa_a\hat{b}, \tag*{(34)}

where ωc\omega_c is the resonant frequency of the cavity, a^in\hat{a}_{\mathrm{in}} and a^i\hat{a}_i represent the input field and intrinsic noise operators, respectively, and κtotal=κi+κext\kappa_{\mathrm{total}}=\kappa_i+\kappa_{\mathrm{ext}}, with κi\kappa_i and κext\kappa_{\mathrm{ext}} denoting the intrinsic and extrinsic decay rates. To derive the steady-state transfer function, we transform the equations into the frequency domain using the Fourier convention A^(ω)=12π∫−∞∞dt A^(t)eiωt\hat{A}(\omega)=\frac{1}{\sqrt{2\pi}}\int_{-\infty}^{\infty}dt\,\hat{A}(t)e^{i\omega t}. Hereafter, we work exclusively in the frequency domain unless stated otherwise. The resulting equation of motion and the corresponding input-output relation are given by:

−iΔωa^=−κtotal2a^−κexta^in−κia^i−iκab^,(35)-i\Delta\omega\hat{a}=-\frac{\kappa_{\mathrm{total}}}{2}\hat{a}-\sqrt{\kappa_{\mathrm{ext}}}\hat{a}_{\mathrm{in}}-\sqrt{\kappa_i}\hat{a}_i-i\kappa_a\hat{b}, \tag*{(35)}
a^out=a^in−κexta^,(36)\hat{a}_{\mathrm{out}}=\hat{a}_{\mathrm{in}}-\sqrt{\kappa_{\mathrm{ext}}}\hat{a}, \tag*{(36)}

where Δω=ω−ωc\Delta\omega=\omega-\omega_c. From the above equations, the output mode in frequency domain is given by

a^out=Cin(ω)a^in+Ci(ω)a^i+Ca(ω)b^,(37)\hat{a}_{\mathrm{out}}=C_{\mathrm{in}}(\omega)\hat{a}_{\mathrm{in}}+C_i(\omega)\hat{a}_i+C_a(\omega)\hat{b}, \tag*{(37)}

where the transfer coefficients are given by

Cin(ω)=1−κextκtotal2−iΔω,(38)C_{\mathrm{in}}(\omega)=1-\frac{\kappa_{\mathrm{ext}}}{\frac{\kappa_{\mathrm{total}}}{2}-i\Delta\omega}, \tag*{(38)}
Ci(ω)=−κextκiκtotal2−iΔω,(39)C_i(\omega)=-\frac{\sqrt{\kappa_{\mathrm{ext}}\kappa_i}}{\frac{\kappa_{\mathrm{total}}}{2}-i\Delta\omega}, \tag*{(39)}
Ca(ω)=−iκaκextκtotal2−iΔω.(40)C_a(\omega)=-\frac{i\kappa_a\sqrt{\kappa_{\mathrm{ext}}}}{\frac{\kappa_{\mathrm{total}}}{2}-i\Delta\omega}. \tag*{(40)}

Treating the axion field as a classical driving source within the cavity, the resulting steady-state photon field forms a coherent state satisfying b^∣αa⟩=αa∣αa⟩\hat{b}\lvert\alpha_a\rangle=\alpha_a\lvert\alpha_a\rangle. The input noise modes are characterized by the vacuum state for a^in\hat{a}_{\mathrm{in}} and the thermal state for a^i\hat{a}_i with the correlation function:

⟨a^i†(ω)a^i(ω′)⟩=δ(ω−ω′)ns(ω),(41)\langle\hat{a}_i^{\dagger}(\omega)\hat{a}_i(\omega')\rangle=\delta(\omega-\omega')n_s(\omega), \tag*{(41)}

where the mean thermal photon number is given by the Bose-Einstein distribution

ns=1eℏω/(kBTs)−1.(42)n_s=\frac{1}{e^{\hbar\omega/(k_{\mathrm{B}}T_s)}-1}. \tag*{(42)}

In the absence of a thermal background, the power of axion signal can be derived from P=ℏωa⟨a^out†(t)a^out(t)⟩P=\hbar\omega_a\langle\hat{a}_{\mathrm{out}}^{\dagger}(t)\hat{a}_{\mathrm{out}}(t)\rangle. From a^(t)=∫−∞∞dω2πa^(ω)e−iωt\hat{a}(t)=\int_{-\infty}^{\infty}\frac{d\omega}{\sqrt{2\pi}}\hat{a}(\omega)e^{-i\omega t},

⟨a^†(t)a^(t)⟩=∫∫dω′dω2π⟨a^†(ω′)a^(ω)⟩ei(ω′−ω)t.(43)\langle\hat{a}^{\dagger}(t)\hat{a}(t)\rangle=\int\int\frac{d\omega' d\omega}{2\pi}\langle\hat{a}^{\dagger}(\omega')\hat{a}(\omega)\rangle e^{i(\omega'-\omega)t}. \tag*{(43)}

For stationary process ⟨b^†(ω′)b^(ω)⟩=∣αa(ω)∣2δ(ω−ω′)\langle\hat{b}^{\dagger}(\omega')\hat{b}(\omega)\rangle=\lvert\alpha_a(\omega)\rvert^2\delta(\omega-\omega'), where αa\alpha_a is the amplitude of coherent state. The power

Pℏωa=∫−∞∞⟨a^out†(ω)a^out(ω)⟩dω2π=∫−∞∞∣Ca(ω)∣2∣αa(ω)∣2dω2π≈∣Ca(ωa)∣2∣α0∣2Δfa.(44)\begin{aligned} \frac{P}{\hbar\omega_a}=\int_{-\infty}^{\infty}\langle\hat{a}_{\mathrm{out}}^{\dagger}(\omega)\hat{a}_{\mathrm{out}}(\omega)\rangle\frac{d\omega}{2\pi} \\ &=\int_{-\infty}^{\infty}\lvert C_a(\omega)\rvert^2\lvert\alpha_a(\omega)\rvert^2\frac{d\omega}{2\pi} \\ &\approx\lvert C_a(\omega_a)\rvert^2\lvert\alpha_0\rvert^2\Delta f_a. \tag*{(44)} \end{aligned}

Here, we treat the axion signal αa(ω)\alpha_a(\omega) as a narrow-band process centered around ωa\omega_a with an effective bandwidth Δfa\Delta f_a and define

∫−∞∞∣αa(ω)∣2dω2π=∣α0∣2Δfa,(45)\int_{-\infty}^{\infty}\lvert\alpha_a(\omega)\rvert^2\frac{d\omega}{2\pi}=\lvert\alpha_0\rvert^2\Delta f_a, \tag*{(45)}

where ∣α0∣2\lvert\alpha_0\rvert^2 represents the average spectral density. In experiments, the relevant parameters are typically defined as

QL≡ωcκtotal,(46)Q_L\equiv\frac{\omega_c}{\kappa_{\mathrm{total}}}, \tag*{(46)}
ββ+1=κext/κiκext/κi+1=κextκtotal,(47)\frac{\beta}{\beta+1}=\frac{\kappa_{\mathrm{ext}}/\kappa_i}{\kappa_{\mathrm{ext}}/\kappa_i+1}=\frac{\kappa_{\mathrm{ext}}}{\kappa_{\mathrm{total}}}, \tag*{(47)}
Δ≡QLfa−fcfc=fa−fcΔfc,(48)\Delta\equiv Q_L\frac{f_a-f_c}{f_c}=\frac{f_a-f_c}{\Delta f_c}, \tag*{(48)}

where β=κext/κi\beta=\kappa_{\mathrm{ext}}/\kappa_i is the coupling coefficient, QLQ_L is the loaded quality factor, and Δ\Delta represents the normalized detuning. Substituting these definitions into Eq. (44), we obtain the detected power:

P=4ℏQLββ+111+4Δ2∣κaα0∣2Δfa.(49)P = 4\hbar Q_L \frac{\beta}{\beta+1}\frac{1}{1+4\Delta^2}|\kappa_a\alpha_0|^2\Delta f_a. \tag*{(49)}

Comparing Eq. (49) with the axion haloscope power formula,

P=gaγγ2ρamaB02VeffQLββ+111+4Δ2,(50)P = g_{a\gamma\gamma}^2\frac{\rho_a}{m_a}B_0^2V_{\mathrm{eff}}Q_L\frac{\beta}{\beta+1}\frac{1}{1+4\Delta^2}, \tag*{(50)}

we establish the relation for the signal amplitude:

∣κaα0∣2=14ℏΔfagaγγ2ρamaB02Veff.(51)|\kappa_a\alpha_0|^2 = \frac{1}{4\hbar\Delta f_a}g_{a\gamma\gamma}^2\frac{\rho_a}{m_a}B_0^2V_{\mathrm{eff}}. \tag*{(51)}

The real quadrature operators are defined as q^=12(a^+a^†)\hat{q} = \frac{1}{\sqrt{2}}(\hat{a}+\hat{a}^{\dagger}) and p^=i2(a^†−a^)\hat{p} = \frac{i}{\sqrt{2}}(\hat{a}^{\dagger}-\hat{a}), satisfying the commutation relation [q^,p^]=i[\hat{q},\hat{p}] = i. For a coherent state ∣αa⟩|\alpha_a\rangle, the corresponding displacement vector d\boldsymbol{d} (mean values) and the corresponding covariance matrix V\mathbf{V} can be obtained from above quadrature definitions:

da≡(⟨q^⟩,⟨p^⟩)=2∣αa∣(cos⁡ϕa,sin⁡ϕa),(52)\boldsymbol{d}_a \equiv(\langle\hat{q}\rangle,\langle\hat{p}\rangle) = \sqrt{2}|\alpha_a|(\cos\phi_a,\sin\phi_a), \tag*{(52)}
Va≡(⟨Δq^2⟩12⟨{Δq^,Δp^}⟩12⟨{Δq^,Δp^}⟩⟨Δp^2⟩)=12I,(53)\mathbf{V}_a \equiv\begin{pmatrix} \langle\Delta\hat{q}^2\rangle& \frac{1}{2}\langle\{\Delta\hat{q},\Delta\hat{p}\}\rangle\\ \frac{1}{2}\langle\{\Delta\hat{q},\Delta\hat{p}\}\rangle& \langle\Delta\hat{p}^2\rangle\end{pmatrix} = \frac{1}{2}\mathbf{I}, \tag*{(53)}

where αa=∣αa∣eiϕa\alpha_a = |\alpha_a|e^{i\phi_a}, {⋅,⋅}\{\cdot,\cdot\} denotes the anti-commutator, and I\mathbf{I} is the 2×22 \times2 identity matrix. From Eq. (37), we can derive the transformation of the mean vector and covariance matrix of the output state. Since din=0\boldsymbol{d}_{\mathrm{in}} = 0, di=0\boldsymbol{d}_i = 0, Vin=12I\mathbf{V}_{\mathrm{in}} = \frac{1}{2}\mathbf{I} and Vi=(12+ns)I\mathbf{V}_i = (\frac{1}{2}+n_s)\mathbf{I}, then

dout=2∣αa(ω)Ca(ω)∣(cos⁡ϕa,sin⁡ϕa),(54)\boldsymbol{d}_{\mathrm{out}} = \sqrt{2}|\alpha_a(\omega)C_a(\omega)|(\cos\phi_a,\sin\phi_a), \tag*{(54)}
Vout=(1/2+∣Ci∣2ns+∣Ca∣2/2)I.(55)\mathbf{V}_{\mathrm{out}} = (1/2+|C_i|^2n_s+|C_a|^2/2)\mathbf{I}. \tag*{(55)}

Since αa(ω)\alpha_a(\omega) is narrow-band process compared with the sensing cavity, ∣αa(ω)Ca(ω)∣≈∣α0Ca(ωa)∣|\alpha_a(\omega)C_a(\omega)| \approx|\alpha_0C_a(\omega_a)|. By Eq. (44),

dout=2PℏωaΔfa(cos⁡ϕa,sin⁡ϕa)≡2na(cos⁡ϕa,sin⁡ϕa),(56)\begin{aligned} \boldsymbol{d}_{\mathrm{out}} = \sqrt{2}\sqrt{\frac{P}{\hbar\omega_a\Delta f_a}}(\cos\phi_a,\sin\phi_a) \\ &\equiv\sqrt{2n_a}(\cos\phi_a,\sin\phi_a), \tag*{(56)} \end{aligned}

where na=P/(ℏωaΔfa)n_a = P/(\hbar\omega_a\Delta f_a). Furthermore, since κa≈10−6≪κi\kappa_a \approx10^{-6} \ll\sqrt{\kappa_i}, the output covariance matrix becomes

Vout=(1/2+∣Ci∣2ns)I=[1/2+4β(1+β)2L(Δ)ns]I,(57)\begin{aligned} \mathbf{V}_{\mathrm{out}} = (1/2+|C_i|^2n_s)\mathbf{I} \\ &= \left[1/2+\frac{4\beta}{(1+\beta)^2}\mathcal{L}(\Delta)n_s\right]\mathbf{I}, \tag*{(57)} \end{aligned}

where L(Δ)≡1/(1+4Δ2)\mathcal{L}(\Delta) \equiv1/(1+4\Delta^2) is the Lorentzian shape function.

State transfer

Phase-Insensitive Amplifier and Noise Performance

The input-output relation for a general linear amplifier and a Josephson parametric amplifier (JPA) can be expressed as [23, 24]

a^out=Ga^in+F^,(58)\hat{a}_{\mathrm{out}} = \sqrt{G}\hat{a}_{\mathrm{in}}+\hat{F}, \tag*{(58)}
a^out=Ga^in+G−1a^e†,(59)\hat{a}_{\mathrm{out}} = \sqrt{G}\hat{a}_{\mathrm{in}}+\sqrt{G-1}\hat{a}_e^\dagger, \tag*{(59)}

where [F^,F^†]=1−G[\hat{F},\hat{F}^{\dagger}] = 1-G and a^e\hat{a}_e denotes the environmental annihilation operator. The thermal state of the environment is characterized by its mean thermal photon number,

nJPA=⟨a^e†a^e⟩=1eℏω/kBT−1.(60)n_{\mathrm{JPA}} = \langle\hat{a}_e^\dagger\hat{a}_e\rangle= \frac{1}{e^{\hbar\omega/k_BT}-1}. \tag*{(60)}

For a phase-insensitive broadband JPA, the covariance matrix of the environment state is given by VJPA=(1/2+nJPA)I\mathbf{V}_{\mathrm{JPA}} = (1/2+n_{\mathrm{JPA}})\mathbf{I}. Consequently, the input-output mapping constitutes a Gaussian channel, yielding the output covariance matrix and displacement vector:

Vout=GJPAVin+(GJPA−1)VJPA=[GJPA(1+∣Ci∣2ns+nJPA)−(12+nJPA)]I,(61)\begin{aligned} \mathbf{V}_{\mathrm{out}} = G_{\mathrm{JPA}}\mathbf{V}_{\mathrm{in}}+(G_{\mathrm{JPA}}-1)\mathbf{V}_{\mathrm{JPA}} \\ &= \left[G_{\mathrm{JPA}}\left(1+|C_i|^2n_s+n_{\mathrm{JPA}}\right)-\left(\frac{1}{2}+n_{\mathrm{JPA}}\right)\right]\mathbf{I}, \tag*{(61)} \end{aligned}
dout=GJPAdin.(62)\boldsymbol{d}_{\mathrm{out}} = \sqrt{G_{\mathrm{JPA}}}\boldsymbol{d}_{\mathrm{in}}. \tag*{(62)}

Defining the input-referred noise via the relation Vout=(1/2+GJPAnin,equiv)I\mathbf{V}_{\mathrm{out}} = (1/2+G_{\mathrm{JPA}}n_{\mathrm{in,equiv}})\mathbf{I}, we obtain

nin,equiv=1+∣Ci∣2ns+nJPA−1+nJPAGJPAn_{\mathrm{in,equiv}} = 1+|C_i|^2n_s+n_{\mathrm{JPA}}-\frac{1+n_{\mathrm{JPA}}}{G_{\mathrm{JPA}}}
≈1+∣Ci∣2ns+nJPA.(63)\approx1+|C_i|^2n_s+n_{\mathrm{JPA}}. \tag*{(63)}

To account for the subsequent amplification stages, we consider a concatenated high-electron-mobility transistor (HEMT) amplifier. The resulting output covariance matrix and displacement vector become

Vout=[GHEMTGJPA(1+∣Ci∣2ns+nJPA)+GHEMT(nHEMT−nJPA)−(12+nJPA)]I,(64)\mathbf{V}_{\mathrm{out}} = \left[G_{\mathrm{HEMT}}G_{\mathrm{JPA}}\left(1+|C_i|^2n_s+n_{\mathrm{JPA}}\right)+G_{\mathrm{HEMT}}(n_{\mathrm{HEMT}}-n_{\mathrm{JPA}})-\left(\frac{1}{2}+n_{\mathrm{JPA}}\right)\right]\mathbf{I}, \tag*{(64)}
dout=GHEMTGJPAdin.(65)\boldsymbol{d}_{\mathrm{out}} = \sqrt{G_{\mathrm{HEMT}}G_{\mathrm{JPA}}}\boldsymbol{d}_{\mathrm{in}}. \tag*{(65)}

In the limit where nHEMT≫nJPAn_{\mathrm{HEMT}} \gg n_{\mathrm{JPA}}, the input-referred noise simplifies to

nin,equiv≈1+∣Ci∣2ns+nJPA+nHEMTGJPA.(66)n_{\mathrm{in,equiv}} \approx1+|C_i|^2n_s+n_{\mathrm{JPA}}+\frac{n_{\mathrm{HEMT}}}{G_{\mathrm{JPA}}}. \tag*{(66)}

By induction, the Friis formula for cascaded linear amplifiers, including the JPA, HEMT, and subsequent mixer, yields the comprehensive input-referred noise:

nin,equiv=1+∣Ci∣2ns+nJPA+nHEMTGJPA+nmixerGJPAGHEMT.(67)n_{\mathrm{in,equiv}} = 1 + \lvert C_i\rvert^2 n_s + n_{\mathrm{JPA}} + \frac{n_{\mathrm{HEMT}}}{G_{\mathrm{JPA}}} + \frac{n_{\mathrm{mixer}}}{G_{\mathrm{JPA}}G_{\mathrm{HEMT}}}. \tag*{(67)}

The total gain G=GJPAGHEMTGmixerG = G_{\mathrm{JPA}}G_{\mathrm{HEMT}}G_{\mathrm{mixer}}. Following this cascaded amplification chain, the final amplified signal state is characterized by the mean displacement vector and covariance matrix:

damp=2Gna(cos⁡ϕa,sin⁡ϕa),(68)\mathbf{d}_{\mathrm{amp}} = \sqrt{2G}n_a(\cos\phi_a,\sin\phi_a), \tag*{(68)}
Vamp=(12+Gnin,equiv)I.(69)\mathbf{V}_{\mathrm{amp}} = \left(\frac{1}{2}+Gn_{\mathrm{in,equiv}}\right)I. \tag*{(69)}

Finally, we note that the quantum limit nin,equiv>1n_{\mathrm{in,equiv}} > 1 arises from two contributions in Eq. (61): the input quantum fluctuations VinV_{\mathrm{in}} and the amplifier fluctuations VJPAV_{\mathrm{JPA}}. By employing a squeezed input and a corresponding phase-sensitive amplifier, one can achieve ⟨Δqin,equiv2⟩<1\langle\Delta q_{\mathrm{in,equiv}}^2\rangle< 1 but ⟨Δpin,equiv2⟩>1\langle\Delta p_{\mathrm{in,equiv}}^2\rangle> 1. This approach is known as squeezing-enhanced sensing.

Quantum transduction

Assuming a steady-state conversion process, the input-output relations for electro-optical transduction in the frequency domain are given by:

0=iΔoa^−igm^−γo2a^+γoea^in+γoia^i,(70)0 = i\Delta_{\mathrm{o}}\hat{a} - ig\hat{m} - \frac{\gamma_{\mathrm{o}}}{2}\hat{a} + \sqrt{\gamma_{\mathrm{oe}}}\hat{a}_{\mathrm{in}} + \sqrt{\gamma_{\mathrm{oi}}}\hat{a}_{i}, \tag*{(70)}
0=iΔmm^−iga^−γm2m^+γmem^in+γmim^i,(71)0 = i\Delta_{\mathrm{m}}\hat{m} - ig\hat{a} - \frac{\gamma_{\mathrm{m}}}{2}\hat{m} + \sqrt{\gamma_{\mathrm{me}}}\hat{m}_{\mathrm{in}} + \sqrt{\gamma_{\mathrm{mi}}}\hat{m}_{i}, \tag*{(71)}
a^out=−γoea^+a^in,(72)\hat{a}_{\mathrm{out}} = -\sqrt{\gamma_{\mathrm{oe}}}\hat{a}+\hat{a}_{\mathrm{in}}, \tag*{(72)}
m^out=−γmem^+m^in,(73)\hat{m}_{\mathrm{out}} = -\sqrt{\gamma_{\mathrm{me}}}\hat{m}+\hat{m}_{\mathrm{in}}, \tag*{(73)}

where the variables and coefficients characterizing the transducer model are comprehensively summarized in Table II, categorized by the optical mode, microwave mode, and electro-optic coupling. Solving the quantum Langevin equations yields the optical output mode as

a^out=(1+γoeXmXoXm+g2)a^in+γoeγoiXmXoXm+g2a^i+igγoeγmeXoXm+g2m^in+igγoeγmiXoXm+g2m^i,(74)\hat{a}_{\mathrm{out}} = \left(1+\frac{\gamma_{\mathrm{oe}}X_{\mathrm{m}}}{X_{\mathrm{o}}X_{\mathrm{m}}+g^2}\right)\hat{a}_{\mathrm{in}}+\frac{\sqrt{\gamma_{\mathrm{oe}}}\sqrt{\gamma_{\mathrm{oi}}}X_{\mathrm{m}}}{X_{\mathrm{o}}X_{\mathrm{m}}+g^2}\hat{a}_{i}+\frac{ig\sqrt{\gamma_{\mathrm{oe}}}\sqrt{\gamma_{\mathrm{me}}}}{X_{\mathrm{o}}X_{\mathrm{m}}+g^2}\hat{m}_{\mathrm{in}}+\frac{ig\sqrt{\gamma_{\mathrm{oe}}}\sqrt{\gamma_{\mathrm{mi}}}}{X_{\mathrm{o}}X_{\mathrm{m}}+g^2}\hat{m}_{i}, \tag*{(74)}

where Xo=iΔo−γo2X_{\mathrm{o}}=i\Delta_{\mathrm{o}}-\frac{\gamma_{\mathrm{o}}}{2}, and Xm=iΔm−γm2X_{\mathrm{m}}=i\Delta_{\mathrm{m}}-\frac{\gamma_{\mathrm{m}}}{2}. The conversion from m^in\hat{m}_{\mathrm{in}} to a^out\hat{a}_{\mathrm{out}} is determined by

igγoeγmeXoXm+g2m^in.(75)\frac{ig\sqrt{\gamma_{\mathrm{oe}}}\sqrt{\gamma_{\mathrm{me}}}}{X_{\mathrm{o}}X_{\mathrm{m}}+g^2}\hat{m}_{\mathrm{in}}. \tag*{(75)}

To maximize the conversion efficiency, the detunings are set to Δo=Δm=0\Delta_{\mathrm{o}}=\Delta_{\mathrm{m}}=0, yielding an efficiency of

ηT=∣igγoeγmeγoγm/4+g2∣2=ζmζo4C(1+C)2,(76)\eta_{\mathrm{T}}=\left\lvert\frac{ig\sqrt{\gamma_{\mathrm{oe}}}\sqrt{\gamma_{\mathrm{me}}}}{\gamma_{\mathrm{o}}\gamma_{\mathrm{m}}/4+g^2}\right\rvert^2=\zeta_{\mathrm{m}}\zeta_{\mathrm{o}}\frac{4C}{(1+C)^2}, \tag*{(76)}

where C=4g2/(γoγm)C=4g^2/(\gamma_{\mathrm{o}}\gamma_{\mathrm{m}}), ζo=γoeγo\zeta_{\mathrm{o}}=\frac{\gamma_{\mathrm{oe}}}{\gamma_{\mathrm{o}}} and ζm=γmeγm\zeta_{\mathrm{m}}=\frac{\gamma_{\mathrm{me}}}{\gamma_{\mathrm{m}}}. Setting Δo=0\Delta_{\mathrm{o}}=0 and Δm=0\Delta_{\mathrm{m}}=0 respectively, the conversion bandwidths for the microwave and optical channels are given by

ΔνFWHM(Δm)=∣C−1∣γm,(77)\Delta\nu_{\mathrm{FWHM}}(\Delta_{\mathrm{m}})=\lvert C-1\rvert\gamma_{\mathrm{m}}, \tag*{(77)}
ΔνFWHM(Δo)=∣C−1∣γo.(78)\Delta\nu_{\mathrm{FWHM}}(\Delta_{\mathrm{o}})=\lvert C-1\rvert\gamma_{\mathrm{o}}. \tag*{(78)}

Since these conversion bandwidths are typically greater than 1 MHz1\ \mathrm{MHz}—much larger than the typical signal bandwidth—the conversion efficiency can be treated as constant for the signal.

At optical frequencies and room temperature, the thermal occupation of the optical input and intrinsic noise modes is negligible (no,in=no,i≈0n_{\mathrm{o,in}}=n_{\mathrm{o,i}}\approx0), whereas for the microwave noise mode, the correlation is given by

⟨m^i†(ω′)m^i(ω)⟩=nmδ(ω′−ω),(79)\langle\hat{m}_{i}^{\dagger}(\omega')\hat{m}_{i}(\omega)\rangle=n_{\mathrm{m}}\delta(\omega'-\omega), \tag*{(79)}

where nmn_{\mathrm{m}} represents the thermal photon number. The add noise photon from m^i\hat{m}_{i} is given by

nadd=∣igγoeγmiγoγm/4+g2∣2∫−∞∞dω ⟨m^i†(ω)m^i(ω)⟩=(1−ζm)ζo4C(1+C)2nm.(80)\begin{aligned} n_{\mathrm{add}}=\left\lvert\frac{ig\sqrt{\gamma_{\mathrm{oe}}}\sqrt{\gamma_{\mathrm{mi}}}}{\gamma_{\mathrm{o}}\gamma_{\mathrm{m}}/4+g^2}\right\rvert^2\int_{-\infty}^{\infty}d\omega\ \langle\hat{m}_{i}^{\dagger}(\omega)\hat{m}_{i}(\omega)\rangle\\ &=(1-\zeta_{\mathrm{m}})\zeta_{\mathrm{o}}\frac{4C}{(1+C)^2}n_{\mathrm{m}}. \tag*{(80)} \end{aligned}

Given that the input microwave mode m^in\hat{m}_{\mathrm{in}} has the mean and covariance matrix specified in Eqs. (56) and (57), the mean and covariance matrix of the output quantum state a^out\hat{a}_{\mathrm{out}} follow directly from their definitions and are given by:

dT=2ηTna(cos⁡ϕa,sin⁡ϕa),(81)\mathbf{d}_{\mathrm{T}}=\sqrt{2\eta_{\mathrm{T}}}n_a(\cos\phi_a,\sin\phi_a), \tag*{(81)}
VT=(12+nT)I.(82)\mathbf{V}_{\mathrm{T}}=\left(\frac{1}{2}+n_{\mathrm{T}}\right)I. \tag*{(82)}

where the noise photon number is given by

nT=ζo4C(1+C)2[(1−ζm)nm+ζm∣Ci(Δ=0)∣2ns]=ζo4C(1+C)2[(1−ζm)nm+ζm4β(1+β)2ns].(83)\begin{aligned} n_{\mathrm{T}}=\zeta_{\mathrm{o}}\frac{4C}{(1+C)^2}\left[(1-\zeta_{\mathrm{m}})n_{\mathrm{m}}+\zeta_{\mathrm{m}}|C_{\mathrm{i}}(\Delta=0)|^2n_{\mathrm{s}}\right] \\ &=\zeta_{\mathrm{o}}\frac{4C}{(1+C)^2}\left[(1-\zeta_{\mathrm{m}})n_{\mathrm{m}}+\zeta_{\mathrm{m}}\frac{4\beta}{(1+\beta)^2}n_{\mathrm{s}}\right]. \tag*{(83)} \end{aligned}

Signal-to-noise ratio and scan rate

Homodyne detection

In homodyne detection, a single frame of the power spectrum q2(ω)q^2(\omega) is obtained via an analog-to-digital converter (ADC) and a fast Fourier transform (FFT). Each spectrum has a time duration of 1/Δν1/\Delta\nu, where Δν\Delta\nu is the spectral bin resolution. To suppress statistical noise, NN independent frames are collected, yielding a total integration time of

tint=NΔν.(84)t_{\mathrm{int}}=\frac{N}{\Delta\nu}. \tag*{(84)}

According to the central limit theorem, the sample mean of the power spectrum, X=1N∑i=1Nqi2X=\frac{1}{N}\sum_{i=1}^{N}q_i^2, approximately follows a normal distribution:

X∼N(μ=⟨q^2⟩, σ2=⟨q^4⟩−⟨q^2⟩2N).(85)X\sim\mathcal{N}\left(\mu=\langle\hat{q}^2\rangle,\ \sigma^2=\frac{\langle\hat{q}^4\rangle-\langle\hat{q}^2\rangle^2}{N}\right). \tag*{(85)}

When axion signal presents, the output state is a Gaussian state characterized by the mean displacement vector and covariance matrix:

damp=2Gna(cos⁡ϕa,sin⁡ϕa),(86)\mathbf{d}_{\mathrm{amp}}=\sqrt{2G n_{\mathrm{a}}}(\cos\phi_{\mathrm{a}},\sin\phi_{\mathrm{a}}), \tag*{(86)}
Vamp=(12+Gnin,equiv)I.(87)\mathbf{V}_{\mathrm{amp}}=\left(\frac{1}{2}+G n_{\mathrm{in,equiv}}\right)\mathbf{I}. \tag*{(87)}

Assume that the phase ϕa\phi_{\mathrm{a}} is uniformly distributed over the averaging interval, the moments of the quadratures are given by

⟨q^2⟩=(12+Gnin,equiv)+Gna,(88)\langle\hat{q}^2\rangle=\left(\frac{1}{2}+G n_{\mathrm{in,equiv}}\right)+G n_{\mathrm{a}}, \tag*{(88)}
⟨q^4⟩−⟨q^2⟩2=4(12+Gnin,equiv)Gna+2(12+Gnin,equiv)2.(89)\langle\hat{q}^4\rangle-\langle\hat{q}^2\rangle^2=4\left(\frac{1}{2}+G n_{\mathrm{in,equiv}}\right)G n_{\mathrm{a}}+2\left(\frac{1}{2}+G n_{\mathrm{in,equiv}}\right)^2. \tag*{(89)}

Since nin,equiv≫nan_{\mathrm{in,equiv}}\gg n_{\mathrm{a}} and Gnin,equiv≫1G n_{\mathrm{in,equiv}}\gg1, the signal-to-noise ratio (SNR) simplifies to

SNR=μaxion−μno-axionσno-axion≈na2nin,equivtintΔν.(90)\mathrm{SNR}=\frac{\mu_{\mathrm{axion}}-\mu_{\mathrm{no\text{-}axion}}}{\sigma_{\mathrm{no\text{-}axion}}}\approx\frac{n_{\mathrm{a}}}{\sqrt{2}n_{\mathrm{in,equiv}}}\sqrt{t_{\mathrm{int}}\Delta\nu}. \tag*{(90)}

According to (67), the system’s input-referred noise exhibits Lorentzian frequency dependence due to ∣Ci∣2ns|C_{\mathrm{i}}|^2n_{\mathrm{s}}. Therefore, we can write the noise as

nin,equiv=nbase+(npeak−nbase)L(Δ),(91)n_{\mathrm{in,equiv}}=n_{\mathrm{base}}+(n_{\mathrm{peak}}-n_{\mathrm{base}})\mathcal{L}(\Delta), \tag*{(91)}

where L(Δ)=1/(1+4Δ2)\mathcal{L}(\Delta)=1/(1+4\Delta^2) is the function of Lorentzian line shape and nbase>1n_{\mathrm{base}}>1. Consider a scan process with a frequency step size of δ=Δfscan/N2\delta=\Delta f_{\mathrm{scan}}/N_2, where N2N_2 is the total number of scanning steps. The total SNR is given by

SNRtotal2=∑k=1N2na2(Δk)2nin,equiv2(Δk)tintΔν=tintΔν1δ∑k=1N2na2(Δk)2nin,equiv2(Δk)δ≈tintΔν1δ∫−Δfscan/2Δfscan/2dfana2(Δ)2nin,equiv2(Δ)≈tintΔνΔfcδ∫−Δfscan/(2Δfc)Δfscan/(2Δfc)dΔna2(Δ)2nin,equiv2(Δ)≈tintΔνΔfcδ∫−∞∞dΔna2(Δ)2nin,equiv2(Δ).(92)\begin{aligned} \mathrm{SNR}_{\mathrm{total}}^2=\sum_{k=1}^{N_2}\frac{n_{\mathrm{a}}^2(\Delta_k)}{2n_{\mathrm{in,equiv}}^2(\Delta_k)}t_{\mathrm{int}}\Delta\nu\\ &=t_{\mathrm{int}}\Delta\nu\frac{1}{\delta}\sum_{k=1}^{N_2}\frac{n_{\mathrm{a}}^2(\Delta_k)}{2n_{\mathrm{in,equiv}}^2(\Delta_k)}\delta\\ &\approx t_{\mathrm{int}}\Delta\nu\frac{1}{\delta}\int_{-\Delta f_{\mathrm{scan}}/2}^{\Delta f_{\mathrm{scan}}/2}\mathrm{d}f_{\mathrm{a}}\frac{n_{\mathrm{a}}^2(\Delta)}{2n_{\mathrm{in,equiv}}^2(\Delta)} \\ &\approx t_{\mathrm{int}}\Delta\nu\frac{\Delta f_{\mathrm{c}}}{\delta}\int_{-\Delta f_{\mathrm{scan}}/(2\Delta f_{\mathrm{c}})}^{\Delta f_{\mathrm{scan}}/(2\Delta f_{\mathrm{c}})}\mathrm{d}\Delta\frac{n_{\mathrm{a}}^2(\Delta)}{2n_{\mathrm{in,equiv}}^2(\Delta)} \\ &\approx t_{\mathrm{int}}\Delta\nu\frac{\Delta f_{\mathrm{c}}}{\delta}\int_{-\infty}^{\infty}\mathrm{d}\Delta\frac{n_{\mathrm{a}}^2(\Delta)}{2n_{\mathrm{in,equiv}}^2(\Delta)}. \tag*{(92)} \end{aligned}

Here, we have converted the discrete summation into an integral approximation, which is valid for δ≪Δfc\delta\ll\Delta f_{\mathrm{c}}. And we assume a broad scan range such that Δfscan/Δfc≫1\Delta f_{\mathrm{scan}}/\Delta f_{\mathrm{c}}\gg1. Given that the scanning rate is defined as R=δ/tintR=\delta/t_{\mathrm{int}}, the total squared SNR can be rewritten as

SNRtotal2=ΔνΔfcR∫−∞∞dΔna2(Δ)2nin,equiv2(Δ).(93)\mathrm{SNR}_{\mathrm{total}}^2=\frac{\Delta\nu\Delta f_{\mathrm{c}}}{R}\int_{-\infty}^{\infty}\mathrm{d}\Delta\frac{n_{\mathrm{a}}^2(\Delta)}{2n_{\mathrm{in,equiv}}^2(\Delta)}. \tag*{(93)}

Substituting na=na,0L(Δ)n_{\mathrm{a}}=n_{\mathrm{a},0}\mathcal{L}(\Delta) and (91), we obtain

SNRtotal2=π8Rna,02npeak3/2nbase1/2ΔfcΔν.(94)\mathrm{SNR}_{\mathrm{total}}^2=\frac{\pi}{8R}\frac{n_{\mathrm{a},0}^2}{n_{\mathrm{peak}}^{3/2}n_{\mathrm{base}}^{1/2}}\Delta f_{\mathrm{c}}\Delta\nu. \tag*{(94)}

Photon detection

A superconducting nanowire single-photon detector is not photon-number resolving: it registers a click whenever one or more photons are absorbed. The natural quantity is therefore the complementary no-click probability. Since the signal absorption and the dark-count process are statistically independent, then

P(no click)=Psig(0)Pdark(0).(95)P(\mathrm{no\ click})=P_{\mathrm{sig}}(0)P_{\mathrm{dark}}(0). \tag*{(95)}

The signal is a Gaussian state with displacement dT\mathbf{d}_{\mathrm{T}} and covariance matrix VT=(1/2+nT)I\mathbf{V}_{\mathrm{T}}=(1/2+n_{\mathrm{T}})\mathbf{I}. A detector of efficiency ηD\eta_{\mathrm{D}} is modeled as a pure loss channel, which maps dT→ηDdT\mathbf{d}_{\mathrm{T}}\to\sqrt{\eta_{\mathrm{D}}}\mathbf{d}_{\mathrm{T}} and nT→ηDnTn_{\mathrm{T}}\to\eta_{\mathrm{D}}n_{\mathrm{T}}. Then

Psig(0)=11+ηDnTexp⁡(−ηD∣dT∣22(1+ηDnT)).(96)P_{\mathrm{sig}}(0)=\frac{1}{1+\eta_{\mathrm{D}}n_{\mathrm{T}}}\exp\left(-\frac{\eta_{\mathrm{D}}|\mathbf{d}_{\mathrm{T}}|^2}{2(1+\eta_{\mathrm{D}}n_{\mathrm{T}})}\right). \tag*{(96)}

Dark counts form an independent Poisson process with

Pdark(k)=λke−rDτk!.(97)P_{\mathrm{dark}}(k)=\frac{\lambda^k e^{-r_{\mathrm{D}}\tau}}{k!}. \tag*{(97)}

where rDr_{\mathrm{D}} is the dark counting rate and τ\tau is the time window. A mode of bandwidth Δfa\Delta f_{\mathrm{a}} is resolved in a time window τ=1/Δfa\tau= 1/\Delta f_{\mathrm{a}}, so the number of dark counts per measurement window is nD=rD/Δfan_{\mathrm{D}} = r_{\mathrm{D}}/\Delta f_{\mathrm{a}} and Pdark(0)=e−nDP_{\mathrm{dark}}(0) = e^{-n_{\mathrm{D}}}. Combining these two factors,

P(no click)=11+ηDnTe−nDexp⁡(−ηDηTna1+ηDnT).(98)P(\text{no click}) = \frac{1}{1+\eta_{\mathrm{D}}n_{\mathrm{T}}}e^{-n_{\mathrm{D}}}\exp\left(-\frac{\eta_{\mathrm{D}}\eta_{\mathrm{T}}n_{\mathrm{a}}}{1+\eta_{\mathrm{D}}n_{\mathrm{T}}}\right). \tag*{(98)}

With na≪nTn_{\mathrm{a}} \ll n_{\mathrm{T}} and nD≪1n_{\mathrm{D}} \ll1, the probability of click is given by

λ≡P(click)≈ηDnT+nD1+ηDnT+ηDηTna(1+ηDnT)2.(99)\lambda\equiv P(\text{click}) \approx\frac{\eta_{\mathrm{D}}n_{\mathrm{T}}+n_{\mathrm{D}}}{1+\eta_{\mathrm{D}}n_{\mathrm{T}}}+\frac{\eta_{\mathrm{D}}\eta_{\mathrm{T}}n_{\mathrm{a}}}{(1+\eta_{\mathrm{D}}n_{\mathrm{T}})^2}. \tag*{(99)}

For a sufficiently long integration time tintt_{\mathrm{int}}, the number of independent modes is N=tintfrN=t_{\mathrm{int}}f_{\mathrm{r}}, where fr=Δfaf_{\mathrm{r}}=\Delta f_{\mathrm{a}} in the continuous-pump regime (set by the signal coherence bandwidth), and frf_{\mathrm{r}} is the pulse repetition rate in the pulsed-pump regime. The random variable X=∑i=1Nλi/NX=\sum_{i=1}^{N}\lambda_i/N has mean and variance

μ=λ,(100)\mu= \lambda, \tag*{(100)}
σ2=λ(1−λ)/N.(101)\sigma^2 = \lambda(1-\lambda)/N. \tag*{(101)}

The signal-to-noise ratio is given by

SNR=μaxion−μno-axionσno-axion≈ηDηTnaηDnT+nDtintfr.(102)\mathrm{SNR} = \frac{\mu_{\mathrm{axion}}-\mu_{\mathrm{no\text{-}axion}}}{\sigma_{\mathrm{no\text{-}axion}}} \approx\frac{\eta_{\mathrm{D}}\eta_{\mathrm{T}}n_{\mathrm{a}}}{\sqrt{\eta_{\mathrm{D}}n_{\mathrm{T}}+n_{\mathrm{D}}}}\sqrt{t_{\mathrm{int}}f_{\mathrm{r}}}. \tag*{(102)}

Consider a scan process with a frequency step size of δ=Δfscan/N2\delta=\Delta f_{\mathrm{scan}}/N_2, where N2N_2 is the total number of scanning steps. Similar to Eq. (92), the total SNR is given by

SNRtotal2=∑k=1N2ηT2ηD2na2(Δk)ηDnT+nDtintfr≈ηT2ηD2na2ηDnT+nDtintfrδ∫−∞∞dfa na2(Δ)≈π4RηD2ηT2na,02ηDnT+nDΔfcfr.(103)\begin{aligned} \mathrm{SNR}_{\mathrm{total}}^2 = \sum_{k=1}^{N_2}\frac{\eta_{\mathrm{T}}^2\eta_{\mathrm{D}}^2n_{\mathrm{a}}^2(\Delta_k)}{\eta_{\mathrm{D}}n_{\mathrm{T}}+n_{\mathrm{D}}}t_{\mathrm{int}}f_{\mathrm{r}} \\ &\approx\frac{\eta_{\mathrm{T}}^2\eta_{\mathrm{D}}^2n_{\mathrm{a}}^2}{\eta_{\mathrm{D}}n_{\mathrm{T}}+n_{\mathrm{D}}}\frac{t_{\mathrm{int}}f_{\mathrm{r}}}{\delta}\int_{-\infty}^{\infty}df_{\mathrm{a}}\,n_{\mathrm{a}}^2(\Delta) \\ &\approx\frac{\pi}{4R}\frac{\eta_{\mathrm{D}}^2\eta_{\mathrm{T}}^2n_{\mathrm{a},0}^2}{\eta_{\mathrm{D}}n_{\mathrm{T}}+n_{\mathrm{D}}}\Delta f_{\mathrm{c}}f_{\mathrm{r}}. \tag*{(103)} \end{aligned}

References

  1. [1]R. Workman, M. Moskovic, M. Carena, D. Robinson, T. DeGrand, J. Anderson, G. Venanzoni, K. Black, R. Mitchell, T. Sjöstrand, et al., Progress of Theoretical and Experimental Physics 2022, 083C01 (2022), https://academic.oup.com/ptp/article-pdf/2022/8/083C01/49175539/ptac097.pdf.
  2. [2]V. C. Rubin, W. K. Ford, Jr., N. Thonnard, and D. Burstein, Astrophys. J. 261, 439 (1982).
  3. [3]K. G. Begeman, A. H. Broeils, and R. H. Sanders, Monthly Notices of the Royal Astronomical Society 249, 523 (1991), https://academic.oup.com/mnras/article-pdf/249/3/523/18160929/mnras249-0523.pdf.
  4. [4]A. N. Taylor, S. Dye, T. J. Broadhurst, N. Benitez, and E. van Kampen, The Astrophysical Journal 501, 539 (1998).
  5. [5]P. Natarajan, U. Chadayammuri, M. Jauzac, J. Richard, J.-P. Kneib, H. Ebeling, F. Jiang, F. van den Bosch, M. Limousin, E. Jullo, H. Atek, A. Pillepich, C. Popa, F. Marinacci, L. Hernquist, M. Meneghetti, and M. Vogelsberger, Monthly Notices of the Royal Astronomical Society 468, 1962 (2017), https://academic.oup.com/mnras/article-pdf/468/2/1962/11210742/stw3385.pdf.
  6. [6]M. Markevitch, A. H. Gonzalez, D. Clowe, A. Vikhlinin, W. Forman, C. Jones, S. Murray, and W. Tucker, The Astrophysical Journal 606, 819 (2004).
  7. [7]N. Aghanim, Y. Akrami, M. Ashdown, J. Aumont, C. Baccigalupi, M. Ballardini, A. J. Banday, R. B. Barreiro, N. Bartolo, and et al., Astronomy & Astrophysics 641, A6 (2020).
  8. [8]R. D. Peccei and H. R. Quinn, Phys. Rev. Lett. 38, 1440 (1977).
  9. [9]S. Weinberg, Phys. Rev. Lett. 40, 223 (1978).
  10. [10]F. Wilczek, Phys. Rev. Lett. 40, 279 (1978).
  11. [11]P. Sikivie, Phys. Rev. Lett. 51, 1415 (1983).
  12. [12]N. Du, N. Force, R. Khatiwada, E. Lentz, R. Ottens, L. J. Rosenberg, G. Rybka, G. Carosi, N. Woollett, D. Bowring, A. S. Chou, A. Sonnenschein, W. Wester, C. Boutan, N. S. Oblath, R. Bradley, E. J. Daw, A. V. Dixit, J. Clarke, S. R. O’Kelley, N. Crisosto, J. R. Gleason, S. Jois, P. Sikivie, I. Stern, N. S. Sullivan, D. B. Tanner, and G. C. Hilton (ADMX Collaboration), Phys. Rev. Lett. 120, 151301 (2018).
  13. [13]B. Ary dos Santos Garcia, D. Bergermann, A. Caldwell, V. Dabhi, C. Diaconu, J. Diehl, G. Dvali, J. Egge, E. Garutti, S. Heyminck, F. Hubaut, A. Ivanov, J. Jochum, S. Knirck, M. Kramer, D. Kreike­meyer-Lorenzo, C. Krieger, C. Lee, D. Leppla-Weber, X. Li, A. Lindner, B. Majorovits, J. P. A. Maldonado, A. Martini, A. Miyazaki, E. Öz, P. Pralavorio, G. Raffelt, J. Redondo, A. Ringwald, J. Schaffran, A. Schmidt, F. Steffen, C. Strandhagen, I. Usherov, H. Wang, and G. Wieching (MADMAX Collaboration), Phys. Rev. Lett. 135, 041001 (2025).
  14. [14]R. Cervantes, G. Carosi, S. Kimes, C. Hanretty, B. H. LaRocque, G. Leum, P. Mohapatra, N. Oblath, R. Ottens, Y. Park, et al., Physical Review D 106, 102002 (2022).
  15. [15]R. Cervantes, G. Carosi, C. Hanretty, and et al., Phys. Rev. Lett. 129, 201301 (2022).
  16. [16]A. J. Millar, S. M. Anl͟age, R. Balafendiev, P. Belov, K. van Bibber, J. Conrad, M. Demarteau, A. Droster, K. Dunne, A. G. Rosso, J. E. Gudmundsson, H. Jackson, G. Kaur, T. Klaesson, N. Kowitt, M. Lawson, A. Leder, A. Miyazaki, S. Morampudi, H. V. Peiris, H. S. Røising, G. Singh, D. Sun, J. H. Thomas, F. Wilczek, S. Withington, M. Wooten, J. Dilling, M. Febbraro, S. Knirck, and C. Marvinney (Endorsers), Phys. Rev. D 107, 055013 (2023).
  17. [17]T. A. Dyson, C. L. Bartram, A. Davidson, J. B. Ezekiel, L. M. Futuraura, T. Liu, and C.-L. Kuo, Phys. Rev. Appl. 21, L041002 (2024).
  18. [18]J. Chiles, I. Charaev, R. Lasenby, M. Baryakhtar, J. Huang, A. Roshko, G. Burton, M. Colangelo, K. Van Tilburg, A. Arvanitaki, S. W. Nam, and K. K. Berggren, Phys. Rev. Lett. 128, 231802 (2022).
  19. [19]P. Sikivie, D. B. Tanner, and Y. Wang, Phys. Rev. D 50, 4744 (1994).
  20. [20]C. Hagmann, P. Sikivie, N. S. Sullivan, and D. B. Tanner, Phys. Rev. D 42, 1297(R) (1990).
  21. [21]C. Hagmann, D. Kinion, W. Stoeffl, K. van Bibber, E. Daw, H. Peng, L. J. Rosenberg, J. LaVigne, P. Sikivie, N. S. Sullivan, D. B. Tanner, F. Nezrick, M. S. Turner, D. M. Moltz, J. Powell, and N. A. Golubev, Phys. Rev. Lett. 80, 2043 (1998).
  22. [22]L. Zhong, S. Al Kenany, K. M. Backes, B. M. Brubaker, S. B. Cahn, G. Carosi, Y. V. Gurevich, W. F. Kindel, S. K. Lamoreaux, K. W. Lehnert, S. M. Lewis, M. Malnou, R. H. Maruyama, D. A. Palken, N. M. Rapidis, J. R. Root, M. Simanovskaia, T. M. Shokair, D. H. Speller, I. Urdinaran, and K. A. van Bibber, Phys. Rev. D 97, 092001 (2018).
  23. [23]A. A. Clerk, M. H. Devoret, S. M. Girvin, F. Marquardt, and R. J. Schoelkopf, Rev. Mod. Phys. 82, 1155 (2010).
  24. [24]C. M. Caves, Phys. Rev. D 26, 1817 (1982).
  25. [25]M. Hatridge, R. Vijav, D. H. Shictter, J. Clarke, and I. Siddiqi, Phys. Rev. B 83, 134501 (2011).
  26. [26]S. J. Astalos, G. Carosi, C. Hagmann, D. Kinion, K. van Bibber, M. Hotz, L. J. Rosenberg, G. Rybka, J. Hoskins, J. Hwang, P. Sikivie, D. B. Tanner, R. Bradley, and J. Clarke, Phys. Rev. Lett. 104, 041301 (2010).
  27. [27]M. M. Brubaker, L. Zhong, Y. V. Gurevich, S. B. Cahn, S. K. Lamoreaux, M. Simanovskaia, J. R. Root, S. M. Lewis, S. Al Kenany, K. M. Backes, I. Urdinaran, N. M. Rapidis, T. M. Shokair, K. A. van Bibber, D. A. Palken, M. Malnou, W. F. Kindel, M. A. Anil, K. W. Lehnert, and G. Carosi, Phys. Rev. Lett. 118, 061302 (2017).
  28. [28]M. Malnou, D. A. Palken, B. M. Brubaker, L. R. Vale, G. C. Hilton, and K. W. Lehnert, Phys. Rev. X 9, 021023 (2019).
  29. [29]K. M. Backes, D. A. Palken, S. A. Kenany, B. M. Brubaker, S. B. Cahn, A. Droster, G. C. Hilton, S. Ghosh, H. Jackson, S. K. Lamoreaux, A. F. Leder, K. W. Lehnert, S. M. Lewis, M. Malnou, R. H. Maruyama, N. M. Rapidis, M. Simanovskaia, S. Singh, D. H. Speller, I. Urdinaran, L. R. Vale, E. C. van Assendelft, K. van Bibber, and H. Wang, Nature 590, 238 (2021).
  30. [30]X. Bai, M. J. Jewell, J. Echevers, K. van Bibber, A. Droster, M. H. Esmat, S. Ghosh, E. Graham, H. Jackson, C. Laffan, S. K. Lamoreaux, A. F. Leder, K. W. Lehnert, S. M. Lewis, R. H. Maruyama, R. D. Nath, N. M. Rapidis, E. P. Ruddy, M. Silva-Feaver, M. Simanovskaia, S. Singh, D. H. Speller, S. Zacarias, and Y. Zhu (HAYSTAC Collaboration), Phys. Rev. Lett. 134, 151006 (2025).
  31. [31]C. Braggio, L. Balembois, R. Di Vora, Z. Wang, J. Traverso, P. Pallegioix, G. Carugno, A. Ortolan, G. Ruoso, U. Gambardella, D. D’Agostino, P. Bertet, and E. Flurin, Phys. Rev. X 15, 021031 (2025).
  32. [32]A. V. Dvïtx, S. Chakram, K. He, A. Agrawal, R. K. Naik, D. I. Schuster, and A. Chou, Phys. Rev. Lett. 126, 141302 (2021).
  33. [33]K. Inomata, Z. Lin, K. Koshino, W. D. Oliver, J.-S. Tsai, T. Yamamoto, and Y. Nakamura, Nature Communications 7, 12303 (2016).
  34. [34]A. L. Pankratov, L. S. Revin, A. V. Gordeeva, A. A. Yablokov, L. S. Kuzmin, and E. Il’ichev, npj Quantum Information 8, 61 (2022).
  35. [35]A. L. Pankratov, A. V. Gordeeva, A. V. Chiginev, L. S. Revin, A. V. Blagodatskin, N. Crescini, and L. S. Kuzmin, Nature Communications 16, 3457 (2025).
  36. [36]N. Lauk, N. Sinclair, S. Barzanjeh, J. P. Covey, M. Saffman, M. Spiropulu, and C. Simon, Quantum Science and Technology 5, 020501 (2020).
  37. [37]J. Holzgrafe, N. Sinclair, D. Zhu, A. Shams-Ansari, M. Colangelo, Y. Hu, M. Zhang, K. K. Berggren, and M. Lončar, Optica 7, 1714 (2020).
  38. [38]R. Sahu, W. Hease, A. Rueda, G. Arnold, L. Qiu, and J. M. Fink, Nature Communications 13, 1276 (2022).
  39. [39]L. Qiu, R. Sahu, W. Hease, G. Arnold, and J. M. Fink, Nature Communications 14, 3784 (2023).
  40. [40]C. Wang, I. Gonin, A. Grassellino, S. Kazakov, A. Romanenko, V. P. Yakovlev, and S. Zorzetti, npj Quantum Information 8, 149 (2022).
  41. [41]C. M. Natarajan, M. G. Tanner, and R. H. Hadfield, Superconductor science and technology 25, 063001 (2012).
  42. [42]P. Sikivie, Phys. Rev. D 32, 2988 (1985).
  43. [43]C. W. Gardiner and M. J. Collett, Phys. Rev. A 31, 3761 (1985).
  44. [44]M. Tsang, Phys. Rev. A 84, 043845 (2011).
  45. [45]M. Aspelmeyer, T. J. Kippenberg, and F. Marquardt, Rev. Mod. Phys. 86, 1391 (2014).
  46. [46]W. Hease, A. Rueda, R. Sahu, M. Wulf, G. Arnold, H. G. Schwefel, and J. M. Fink, PRX Quantum 1, 020315 (2020).
  47. [47]Y. Xu, A. A. Sayem, L. Fan, C.-L. Zou, S. Wang, R. Cheng, W. Fu, L. Yang, M. Xu, and H. X. Tang, Nature Communications 12, 4453 (2021).
  48. [48]H. K. Warner, J. Holzgrafe, B. Yankelevich, D. Barton, S. Poletto, C. J. Xin, N. Sinclair, D. Zhu, E. Sete, B. Langley, E. Batson, M. Colangelo, A. Shams-Ansari, G. Joe, K. K. Berggren, L. Jiang, M. J. Reagor, and M. Lončar, Nature Physics 21, 831 (2025).
  49. [49]W. Fu, M. Xu, X. Liu, C.-L. Zou, C. Zhong, X. Han, M. Shen, Y. Xu, R. Cheng, S. Wang, L. Jiang, and H. X. Tang, Phys. Rev. A 103, 053504 (2021).
  50. [50]S. Zorzetti, C. Wang, I. Gonin, S. Kazakov, T. Khabiboulline, A. Romanenko, V. P. Yakovlev, and A. Grassellino, Physical Review B 107, L220302 (2023).
  51. [51]A. Reineri, F. Crisa, A. Murthy, M. Shinde, D. Bafia, C. Wang, T. Roy, A. Romanenko, J. Zasadzinski, A. Grassellino, et al., EPJ Quantum Technology 13, 104 (2026).
  52. [52]A. Sekine, B. Murakami, and Y. Doi, npj Nanophotonics 3, 42 (2026).
  53. [53]T. Werner, E. Riyazi, S. Hawaldar, R. Sahu, G. Arnold, P. Falthansl-Scheinecker, J. A. S. Naranjo, D. Loi, L. N. Kapoor, M. Zemlicka, et al., arXiv preprint arXiv:2602.00928 (2026).arxiv.org/abs/2602.00928
  54. [54]H. Shi and Q. Zhuang, npj Quantum Information 9, 27 (2023).
  55. [55]M. Rosati and A. Solana, Optica Quantum 2, 390 (2024).
  56. [56]H. Shi, A. J. Brady, W. Górecki, L. Maccone, R. Di Candia, and Q. Zhuang, npj Quantum Information 11, 48 (2025).
  57. [57]C. Wang, D. Hu, S. Zorzetti, A. Grassellino, A. Romanenko, and Z. Zhang, Physical Review Letters 136, 250801 (2026).

Paper details

vlnc:W_f2c8ugr9
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents