betaFeedback

WorksPhysics and Astronomy

Search for Dark Photon Dark Matter at 74–159 μμeV/c2c^2 Using Broadband Cryogenic Receivers

Abstract

Dark photons are hypothetical massive vector bosons that couple to ordinary photons through kinetic mixing χχ. Dark photon dark matter (DPDM) converts into photons at a metal surface, with a frequency set by its mass. Using cryogenic K- and Ka-band receivers with a broadband spectrometer, we searched for this signal over 74–159 μμeV/c2c^2 (18–38.5 GHz) with an integration time of 4.8×1054.8\times10^5 s per frequency bin. We found no statistically significant signal and set upper limits of χ<(3.1χ<(3.1–17)×10−1217)\times10^{-12} at the 95% confidence level over 99.8% of the search range, assuming DPDM constitutes all local dark matter. These results improve on previous limits by more than an order of magnitude over most of 74–110 μμeV/c2c^2 and provide the first direct limits over most of 110–159 μμeV/c2c^2, approximately two orders of magnitude below cosmological bounds.

Astrophysical and cosmological observations provide evidence for dark matter through its gravitational effects, but its microscopic nature remains unknown. The dark photon, a hypothetical massive vector boson that mixes kinetically with the photon [1, 2], is a candidate for dark matter. Dark photon dark matter (DPDM) produced by quantum fluctuations during inflation can account for the full dark matter abundance [3]. For inflationary scales near the upper limit from cosmic microwave background observations [4], this requires masses of order 100 μeV/c2\mu\mathrm{eV}/c^{2}.

The kinetic mixing of DPDM is constrained over a broad mass range by cosmological observations [5] [2]. Direct searches set stronger limits [6, 7, 8, 9, 10, 11], but only over narrow mass ranges, leaving most masses above 110 μeV/c2\mu\mathrm{eV}/c^{2} without direct limits. Resonant enhancement improves sensitivity but narrows the bandwidth of each configuration, so that broad coverage requires many tuning steps [11, 12].

We use a broadband detection method based on the conversion of DPDM into ordinary photons at a metal surface through kinetic mixing χ\chi [13]. The photons are emitted nearly perpendicular to the surface [14], which in our setup is a metal plate facing a horn antenna. The photon frequency ν\nu satisfies hν≃mDPc2h\nu\simeq m_{\mathrm{DP}}c^{2}, where mDPm_{\mathrm{DP}} is the dark photon mass, hh is Planck’s constant, and cc is the speed of light. The signal has a fractional linewidth Δν/ν∼10−6\Delta\nu/\nu\sim10^{-6} due to the velocity dispersion of a virialized Galactic halo [6].

The total signal power PDPP_{\mathrm{DP}} received by the antenna is given by [13]

PDP=χ2cρAeffα2.(1)P_{\mathrm{DP}}=\chi^{2}c\rho A_{\mathrm{eff}}\alpha^{2}. \tag*{(1)}

We assume DPDM constitutes all local dark matter, with energy density ρ=(0.39±0.03) GeV/cm3\rho=(0.39\pm0.03)\,\mathrm{GeV}/\mathrm{cm}^{3} [15]. For randomly oriented DPDM polarization, we take the polarization factor α=1/3\alpha=\sqrt{1/3} for a single-polarization receiver [2]. The effective detection area Aeff(ν)A_{\mathrm{eff}}(\nu) is the ratio of received signal power to power emitted per unit area by a perfectly conducting plane in the detected polarization.

In this Letter, we search for DPDM over 74–159 μeV/c2\mu\mathrm{eV}/c^{2} (18–38.5 GHz) by extending the cryogenic receiver system of our previous search [7] to higher frequencies and employing DSpec, a broadband spectrometer developed in Ref. [16]. Compared with the spectrum analyzer used in our previous searches [7, 8], DSpec increases the integration time per frequency bin by a factor of 2000 for the same observing time and frequency coverage. We improve on the previous limits by more than an order of magnitude over most of 74–110 μeV/c2\mu\mathrm{eV}/c^{2} and obtain the first direct limits across 110–159 μeV/c2\mu\mathrm{eV}/c^{2}.

Figure 1 shows the apparatus. Alongside the K-band (18–26 GHz) receiver used previously [7], we employ a new Ka-band (26–38.5 GHz) receiver (Table 1). The Ka-band system noise temperature is typically 60–120 K, comparable to 40–100 K in the K band. The measurements were performed at (35.03° N, 135.78° E), with both horns pointing to the zenith and their polarization axes at an azimuth of approximately 165∘165^\circ, measured clockwise from true north.

Schematic cross section of the cryostat and the warm receiver chain

Figure 1. Schematic cross section of the cryostat (left) and the warm receiver chain in the dashed box (right). The horn antenna collects photons from the conversion of dark photon dark matter (DPDM) at the metal plate. C-LNA/W-LNA: cold/warm low-noise amplifiers; RT-MLI: radio-transparent multilayer insulation [17]; SG: signal generator (FSL-0020); ×2: frequency doubler (SFA-203403205-KFSF-S1); LO: local oscillator; I/Q: in-phase/quadrature components; LPF: low-pass filter. Switches select between signal paths and 50-Ω\Omega terminations.

ReceiverK bandKa band
Range (GHz)18–2626–3232–38.5
Horn antennaSGH-42-SC000SAR-2507-28-S2
C-LNALNF-LNC15-29B
W-LNAALM-1826S210ZVA-24443G1+
MixerMMIQ-1040LS
IF amplifierZX60-43-S+ZX60-43-S+ZKL-2R7+

Table 1. Receiver components and frequency ranges. IF: intermediate frequency.

The downconverted intermediate-frequency (IF) signal is recorded by dSpec, which continuously acquires spectra over 4 GHz with 15.6-kHz bin spacing and a 0.983-s integration time per spectrum. The observed band, centered on the local oscillator (LO) frequency, is selected by tuning the LO.

We evaluated Aeff(ν)A_{\mathrm{eff}}(\nu) by numerically propagating the conversion fields through the apparatus geometry and calculating their coupling to each horn, whose aperture fields were obtained from electromagnetic simulations. Conversion photons from other metal parts inside the apparatus interfere with those from the plate with uncertain relative phases. We conservatively adopted the minimum AeffA_{\mathrm{eff}} over these phases and over the uncertain geometry below the window, which ranges over 31.6–63.7 cm2^2 and 24.5–52.7 cm2^2 for the K- and Ka-band receivers, respectively. The calculation, boundary conditions, and systematic uncertainties are described in Appendix A.

The search covered 18–38.5 GHz using the LO frequency pairs listed in Table II. For each pair, we alternated 600-s measurements between the two LO frequencies, separated by 20 MHz, to investigate spurious signals and to retain usable data where only one of them was contaminated. Between these measurements, we switched the spectrometer inputs to 50-Ω\Omega terminations for 10 s to probe spurious signals originating downstream of the switches. The total integration time for the search was 2.4×1052.4 \times10^{5} s at each LO frequency.

ReceiverK bandKa band
Range (GHz)18–2626–3232–38.5
LO pairs (GHz)(20.00, 20.02)(27.50, 27.52)(33.50, 33.52)
(22.25, 22.27)(30.50, 30.52)(36.50, 36.52)
(24.50, 24.52)
PeriodSep. 27–Oct. 22 2025Dec. 18–30 2025Jan. 16–28 2026

Table 2. Search frequency ranges, LO frequency pairs, and measurement periods.

At the lower frequency of each LO pair, we also performed a separate measurement with the metal plate replaced by a room-temperature absorber, with an integration time exceeding 2.4×1052.4 \times10^{5} s. These data were used as mock search data and for characterizing spurious signals.

To calibrate the search data, we measured the receiver gain using the hot–cold method [18]. Radio-wave absorbers at room and liquid-nitrogen temperatures were placed above the cryostat window, and spectra were recorded for 15 s at each temperature. During the search, we performed these measurements daily at each LO frequency. Separately, we repeated the hot–cold measurement hourly over 24 hours to assess gain stability. In the analysis, we used the gains from the daily measurements, averaged in linear units at each frequency.

For each LO frequency, we averaged the spectra over the full measurement period and estimated the statistical fluctuations of this time-averaged spectrum using null spectra, as in our previous searches [7, 8]. We generated 10,000 null spectra by pairing consecutive spectra, reversing the sign of one randomly chosen spectrum in each pair, and averaging all spectra. This cancels stable spectral features while preserving the statistical noise level of the time-averaged spectrum. At each frequency bin, the standard deviation across the null spectra was used as the statistical uncertainty for both the time-averaged spectrum and every null spectrum.

As in our previous searches [7, 8], we extracted the signal power for each trial dark photon mass, expressed as ν0≡mDPc2/h\nu_0 \equiv m_{\mathrm{DP}}c^2/h, by fitting the expected signal shape with a linear background:

f(ν;P,a,b)=Pfsig(ν;ν0)+a(ν−ν0)+b.(2)f(\nu; P, a, b) = P f_{\mathrm{sig}}(\nu; \nu_0) + a(\nu- \nu_0) + b. \tag*{(2)}

Here, ν\nu is the center frequency of each FFT bin, PP is the fitted signal power with statistical uncertainty ΔP\Delta P, and aa and bb describe the background. The signal function fsigf_{\mathrm{sig}} gives the fraction of total signal power expected in each bin. It was calculated from the Maxwell–Boltzmann halo spectrum with Galactic circular speed and Earth speed both set to 220 km/s [6, 19], convolved with the power response of the rectangular-window FFT in dSpec [16]. We evaluated PP at ν0\nu_0 values spaced by 1.95 kHz, using χ2\chi^2 fits to the time-averaged and null spectra over ν0−50\nu_0 - 50 to ν0+200\nu_0 + 200 kHz.

We applied the gain calibration to PP and ΔP\Delta P. We then combined the results from the two LO frequencies with inverse-variance weights to obtain the received signal power PDPP_{\mathrm{DP}} and its statistical uncertainty ΔPDP\Delta P_{\mathrm{DP}}. Data affected by spurious signals were omitted from this combination.

We evaluated two classes of spurious signals that could mimic a DPDM signal: those originating in the mixer, LO, or downstream receiver chain (IF spurious signals), and those appearing at fixed radio frequencies (RF spurious signals). The former were identified using absorber and termination measurements and comparisons of spectra at different LO frequencies. The latter were traced to harmonics of the dSpec clocks in follow-up measurements with an independent spectrum analyzer: the tested peaks appeared only when dSpec was powered on. Data affected by either class were excluded from the search. The

identification procedures and exclusion criteria are described in Appendix § 3. The same selection, calibration, and combination were applied to the null spectra.

After the exclusion of frequency intervals contaminated by identified spurious signals, 99.8% of the search range remained. We evaluated the statistical significance following the procedure used in our previous searches [7, 8]. The fits to the null spectra provided the distributions of the ratio PDP/ΔPDPP_{\mathrm{DP}}/\Delta P_{\mathrm{DP}} under the zero-signal hypothesis. These distributions were consistent with a standard normal distribution. Using these distributions, we calculated the local pp value at each frequency as the probability of obtaining a ratio at least as large as the observed value.

We accounted for the look-elsewhere effect [20] using an effective number of independent trials, Neff=1.94×106N_{\mathrm{eff}} = 1.94 \times10^6, estimated from the null spectra with correlations between neighboring frequencies included. The probability of an equally or more significant fluctuation anywhere in the search range was

pglobal=1−(1−plocalmin⁡)Neff,(3)p_{\mathrm{global}} = 1 - \left(1 - p_{\mathrm{local}}^{\min}\right)^{N_{\mathrm{eff}}}, \tag*{(3)}

where plocalmin⁡p_{\mathrm{local}}^{\min} is the smallest local pp value.

Figure 3 shows the frequency dependence and distribution of PDP/ΔPDPP_{\mathrm{DP}}/\Delta P_{\mathrm{DP}} in the search data. The smallest local pp value was 2.43×10−72.43 \times10^{-7} at ν0=37.584939\nu_0 = 37.584939 GHz (PDP/ΔPDP=4.96P_{\mathrm{DP}}/\Delta P_{\mathrm{DP}} = 4.96), corresponding to a global pp value of 0.38. Figure 2 shows the spectra and fits for both LO frequencies at this ν0\nu_0. No statistically significant DPDM signal was found.

Measured spectra near $\nu_0 = 37.584939$ GHz and fits for signal extraction

Figure 2. Measured spectra near ν0=37.584939\nu_0 = 37.584939 GHz and fits for signal extraction (solid lines). The upper and lower panels correspond to LO frequencies of 36.50 and 36.52 GHz, respectively. Error bars indicate 1σ1\sigma statistical uncertainties. At this frequency, we obtained the smallest local pp value in the search, while the corresponding global pp value indicated no significant excess.

Ratio of extracted signal power to its statistical uncertainty plotted versus frequency and as a histogram

Figure 3. The ratio PDP/ΔPDPP_{\mathrm{DP}}/\Delta P_{\mathrm{DP}} of extracted signal power to its statistical uncertainty is plotted versus frequency (left) and as a histogram (right). Light gray indicates results from intervals excluded from the search. Dashed lines indicate global significances of 2σ2\sigma and 3σ3\sigma.

The systematic uncertainties in χ\chi are summarized in Table III. The uncertainties in AeffA_{\mathrm{eff}} were evaluated by recalculating AeffA_{\mathrm{eff}} with varied window geometry and photon direction. We compared the frequency-averaged linear gain with its time average at each LO frequency, separately for the daily calibrations during the search and the 24-hour measurements. We adopted the largest relative deviations, 9.5% (K band) and 5.3% (Ka band), as the gain uncertainties; these contribute 5.1% and 2.8%, respectively, to the uncertainty in χ\chi. The uncertainty in the absorber temperature contributes 0.5%.

ReceiverK bandKa band
Effective detection area (AeffA_{\mathrm{eff}})1.6–23.62.2–24.9
Window geometry1.5–23.62.0–24.9
Direction of conversion photons0.50.8
Gain5.12.8
Time variation5.12.8
Absorber temperature0.5
ν0\nu_0 spacing0.8
Earth speed variation (vEv_{\mathrm{E}})0.9
Dark matter density (ρ\rho)3.9
Total6.8–24.55.4–25.4

Table III. Systematic uncertainties in χ\chi. Interference with conversion photons from other metal parts and the geometry below the window are accounted for by adopting the minimum Aeff(ν)A_{\mathrm{eff}}(\nu) and are not listed.

We evaluated the bias in PP due to the discrete spacing of ν0\nu_0 by fitting noiseless signals at frequencies between adjacent ν0\nu_0 values and comparing the fitted and injected powers. We evaluated the effect of Earth speed variation by varying vEv_{\mathrm{E}} in the signal model, following our previous analysis [8]. The uncertainty in the local dark matter density [15] contributes 3.9% to that in χ\chi.

We set the 95% confidence-level upper limit on χ\chi at each ν0\nu_0 to the value for which the probability of PDP/ΔPDP≤max⁡[(PDP/ΔPDP)obs,0]P_{\mathrm{DP}}/\Delta P_{\mathrm{DP}} \le\max[(P_{\mathrm{DP}}/\Delta P_{\mathrm{DP}})_{\mathrm{obs}},0] is 5%. In this calculation, the expected signal power was given by Eq. (1) with the minimum Aeff(ν)A_{\mathrm{eff}}(\nu), the statistical fluctuation was treated as Gaussian with standard deviation ΔPDP\Delta P_{\mathrm{DP}}, and the total systematic uncertainty in Table III as an independent Gaussian factor in χ\chi. Figure 4 shows the resulting 95% confidence-level upper limits, χ<(3.1\chi< (3.1–17)×10−1217) \times10^{-12} over 74–159 μeV/c2\mu\mathrm{eV}/c^2, together with previous constraints. We also set limits in intervals excluded from the search, without subtracting identified spurious signals. These conservative upper limits are as high as 3.8×10−103.8 \times10^{-10}.

Upper limits on $\chi$ at 95% confidence level

Figure 4. Upper limits on χ\chi at 95% confidence level from this work (red) and our previous searches [7, 8] (blue). The limits from this work include systematic uncertainties. Shading above the curves indicates excluded regions. Other experimental [6, 9, 10, 11, 12, 21, 22, 23, 24] [2] and cosmological [5, 25] constraints are shown in gray, using data from AxionLimits [26].

We searched for DPDM over 74–159 μeV/c2\mu\mathrm{eV}/c^2 using cryogenic K- and Ka-band receivers with a broadband spectrometer. We found no statistically significant signal and set 95% confidence-level upper limits of χ<(3.1\chi< (3.1–17)×10−1217) \times10^{-12}. These limits improve on our previous limits [7] by more than an order of magnitude over most of 74–110 μeV/c2\mu\mathrm{eV}/c^2. Over most of 110–159 μeV/c2\mu\mathrm{eV}/c^2, they are the first direct limits, approximately two orders of magnitude below cosmological bounds.

Acknowledgments

This work was supported by JSPS KAK-ENHI Grants No. JP20K14486, No. JP20K20427, No. JP21H05460, No. JP22K18713, No. JP24KJ1452, No. JP24KK0066, and No. JP25H02181; JST CREST Grant No. JPMJCR2313, Japan; JST FOREST Program Grant No. JPMJFR25K, Japan; and a research grant from the Murata Science and Education Foundation. H.T. was supported by a JSPS Research Fellowship for Young Scientists. S.A. acknowledges support from the Hakubi Project of Kyoto University.

Effective detection area

Figure 5 shows the geometry used in the calculation. The vacuum window of our previous search [7], made of ultrahigh-molecular-weight polyethylene, was modeled with its measured profile, which is bowed inward by atmospheric pressure. We calculated the conversion fields at the metal plate and window interfaces from the electromagnetic boundary conditions in the presence of DPDM [27]. The plate was treated as a perfect conductor, and propagation between these surfaces, including multiple reflections, was evaluated using equivalent electric and magnetic surface currents [28]. A perfectly matched layer (PML) provided a nonreflecting radial boundary outside the window opening. Below the window, radio-wave absorbers suppress multiple reflections and limit the horn’s view of surrounding structures. From the window exit aperture A1A_1, we propagated the field through A2A_2 and A3A_3, modeled as openings in perfectly absorbing screens at the 30-K shield and the absorber-lined cylinder entrance, respectively. Using electromagnetic reciprocity [29], we calculated the received power from the conversion fields propagated to the horn aperture plane and the aperture fields of the horn in transmitting mode. The latter were simulated with Ansys HFSS [30] for unit incident power at the waveguide port. The conversion fields, including those from the three exposed annular metal surfaces S1S_1–S3S_3 [Figure 5(b)], were combined coherently because the apparatus lies within the spatial coherence length of Galactic DPDM [2]. Since the annuli belong to separate cryostat components at different temperatures, their positions relative to the plate cannot be determined with a precision much better than the photon wavelengths of 8–17 mm. We therefore obtained conservative bounds on AeffA_{\rm eff} by independently varying each annular contribution’s phase over 00–2π2\pi relative to the plate–window field. The minimum AeffA_{\rm eff} is lower than that without the annular contributions by up to 44.4% and 41.5% for the K- and Ka-band receivers, respectively. Similarly, for the geometry below the window, namely the horn position and the dimensions of the apertures A2A_2 and A3A_3, we adopted the lowest AeffA_{\rm eff} within the uncertainty of each parameter. We used the resulting minimum AeffA_{\rm eff} for the reported limits. AeffA_{\rm eff} and its bounds were evaluated at 1-GHz intervals, at which the horn fields were simulated, and linearly interpolated in between.

Figure 5. Geometry for the effective detection area calculation: (a) schematic section of the cryostat and (b) computational model, not to scale. The gray layer is the vacuum window. S1S_1–S3S_3 denote the exposed annular metal surfaces facing the horn on the vacuum-chamber ceiling, the 30-K shield ceiling, and the joint of the 3-K shield, respectively.

The remaining uncertainties in AeffA_{\rm eff}, from the window geometry and the direction of the conversion photons, were evaluated by recalculating AeffA_{\rm eff} with these varied; the resulting uncertainties in χ\chi are summarized in Table III.

ble III.

Identification and exclusion of spurious signals

In dSpec, a 122.88-MHz crystal oscillator generates the 409.6-MHz ADC reference clock. IF spurious signals were identified from the absorber measurements by

examining features with ∣P/ΔP∣≥5.5|P/\Delta P| \ge5.5 and comparing them with 50-Ω\Omega termination spectra and measurements at different LO frequencies. The dominant components occurred at integer multiples of 40.96 MHz, as previously observed [16], and included harmonics of 409.6, 122.88, 655.36, 1146.88, and 1802.24 MHz. For each LO pair, we also excluded 40.96-MHz multiples identified in the corresponding absorber data.

RF peaks at the known clock harmonics with ∣P/ΔP∣≥5.5|P/\Delta P| \ge5.5 were examined in follow-up measurements using an Anritsu MS2840A spectrum analyzer in place of dSpec, comparing spectra taken with dSpec powered on and off at a 300-Hz resolution bandwidth over a 2.5-MHz span. The search data contained 19 such peaks at harmonics of 409.6 MHz and one at the 168th harmonic of 122.88 MHz (20.64384 GHz). The tested 409.6-MHz harmonics, n=44n = 44–49 in the K band and n=72n = 72 (29.4912 GHz) in the Ka band, and the 122.88-MHz peak appeared only with dSpec powered on and were thus identified as spurious. For the 409.6-MHz series, which showed this behavior at consecutive harmonics, we excluded all harmonics in the search range, including the other 12 peaks and those below the threshold. The 122.88-MHz peak was excluded individually.

Intervals within 10 MHz of each LO frequency were excluded because of the low IF gain near DC. Using the calculated response to monochromatic signals, we excluded intervals where PP would change by at least ΔP\Delta P (Table IV). Where data at only one LO frequency were affected, those at the other were retained; intervals excluded at both were omitted from the search. The remaining intervals cover 99.8% of the search range.

∣P/ΔP∣|P/\Delta P|Frequency offset (MHz)
[0, 10)[−0.1, +0.3)
[10, 10^3)[−0.5, +0.5)
[10310^3, 10410^4)[−1, +1)
≥104\ge10^4[−10, +10)

Table IV. Frequency intervals excluded around identified spurious signals.

References

  1. [4]P. A. R. Ade et al. (BICEP/Keck Collaboration), Improved constraints on primordial gravitational waves using Planck, WMAP, and BICEP/Keck observations through the 2018 observing season, Phys. Rev. Lett. 127, 151301 (2021).DOI
  2. [5]P. Arias, D. Cadamuro, M. Goodsell, J. Jaeckel, J. Redondo, and A. Ringwald, WISPy cold dark matter, J. Cosmol. Astropart. Phys. 2012 (06), 013.DOI
  3. [6]N. Tomita, S. Oguri, Y. Inoue, M. Minowa, T. Nagasaki, J. Suzuki, and O. Tajima, Search for hidden-photon cold dark matter using a K-band cryogenic receiver, J. Cosmol. Astropart. Phys. 2020 (09), 012.
  4. [7]S. Kotaka, S. Adachi, R. Fujinaka, S. Honda, H. Nakata, Y. Seino, Y. Sueno, T. Sumida, J. Suzuki, O. Tajima, and S. Takeichi (DOSUE-RR Collaboration), Search for dark photon dark matter in the mass range 74–110 μeV/c² with a cryogenic millimeter-wave receiver, Phys. Rev. Lett. 130, 071805 (2023).
  5. [8]S. Adachi, R. Fujinaka, Y. Muto, H. Nakata, Y. Sueno, T. Sumida, J. Suzuki, O. Tajima, H. Takeuchi, and S. Honda (DOSUE-RR Collaboration), Search for dark photon dark matter in the mass range 41–74 μeV using millimeter-wave receiver and radioshielding box, Phys. Rev. D 109, 012008 (2024).DOI
  6. [9]F. Bajjali, S. Dornbusch, M. Ekmedžić, D. Horns, C. Kasemann, A. Lobanov, A. Mkrtchyan, L. H. Nguyen, M. Tluzckont, G. Tuccari, J. Ulrichs, G. Wieching, and A. Zensus, First results from BRASS-p broadband searches for hidden photon dark matter, J. Cosmol. Astropart. Phys. 2023 (08), 077.DOI
  7. [10]S. Knirck et al. (BREAD Collaboration), First results from a broadband search for dark photon dark matter in the 44 to 52 μeV range with a coaxial dish antenna, Phys. Rev. Lett. 132, 131004 (2024).DOI
  8. [11]J. Egge et al. (MADMAX Collaboration), First search for dark photon dark matter with a MADMAX prototype, Phys. Rev. Lett. 134, 151004 (2025).
  9. [12]R. Cervantes, G. Carosi, C. Harnetty, S. Kimes, B. H. LaRocque, G. Leum, P. Mohapatra, N. S. Oblath, R. Ottens, Y. Park, G. Rybka, J. Sinis, and J. Yang, Search for 70 μeV dark photon dark matter with a dielectrically loaded multiwavelength microwave cavity, Phys. Rev. Lett. 129, 201301 (2022).DOI
  10. [13]D. Horns, J. Jaeckel, A. Lindner, A. Lobanov, J. Redondo, and A. Ringwald, Searching for WISPy cold dark matter with a dish antenna, J. Cosmol. Astropart. Phys. 2013 (04), 016.DOI
  11. [14]J. Jaeckel and S. Knirck, Directional resolution of dish antenna experiments to search for WISPy dark matter, J. Cosmol. Astropart. Phys. 2016 (01), 005.DOI
  12. [15]R. Catena and P. Ullio, A novel determination of the local dark matter density, J. Cosmol. Astropart. Phys. 2010 (08), 004.DOI
  13. [16]H. Takeuchi, J. Suzuki, S. Adachi, and O. Tajima, Broadband, dead-time-free spectrometer using RFSoC for WISP dark matter searches, Prog. Theor. Exp. Phys. 2025, 053H02 (2025).arxiv.org/abs/2501.12003
  14. [17]J. Choi, H. Ishitsuka, S. Mima, S. Oguri, K. Takahashi, and O. Tajima, Radio-transparent multi-layer insulation for radiowave receivers, Rev. Sci. Instrum. 84, 114502 (2013).
  15. [18]R. B. Partridge, 3K: The Cosmic Microwave Background Radiation (Cambridge University Press, Cambridge, 1995).
  16. [19]A. K. Drukier, K. Freese, and D. N. Spergel, Detecting cold dark-matter candidates, Phys. Rev. D 33, 3495 (1986).
  17. [20]J. W. Foster, N. L. Rodd, and B. R. Safdi, Revealing the dark matter halo with axion direct detection, Phys. Rev. D 97, 123006 (2018).DOI
  18. [21]A. P. Quiskamp, B. T. McAllister, P. Altin, E. N. Ivanov, M. Goryachev, and M. E. Tobar, Direct search for dark matter axions excluding ALP cogenesis in the 63- to 67-μeV range with the ORGAN experiment, Sci. Adv. 8, eabq3765 (2022).DOI
  19. [22]A. Quiskamp, B. T. McAllister, P. Altin, E. N. Ivanov, M. Goryachev, and M. E. Tobar, Exclusion of axionlike-particle cogenesis dark matter in a mass window above 100 μeV, Phys. Rev. Lett. 132, 031601 (2024).DOI
  20. [23]M. Di Vora et al., Search for invisible axion dark matter of mass m_a = 43 μeV with the QUAX-γ experiment, Phys. Rev. D 103, 102004 (2021).
  21. [24]R. Di Vora et al. (QUAX Collaboration), Search for galactic axions with a traveling wave parametric amplifier, Phys. Rev. D 108, 062005 (2023).DOI
  22. [25]S. J. Witte, S. Rosauro-Alcaraz, S. D. McDermott, and V. Poulin, Dark photon dark matter in the presence of inhomogeneous structure, J. High Energy Phys. 2020 (06), 132.DOI
  23. [26]C. O’Hare, cajohare/axionlimits: Axionlimits, https://cajohare.github.io/AxionLimits/ (2020).
  24. [27]J. Jaeckel and J. Redondo, Resonant to broadband searches for cold dark matter consisting of weakly interacting slim particles, Phys. Rev. D 88, 115002 (2013).DOI
  25. [28]J. D. Jackson, Classical Electrodynamics, 3rd ed. (Wiley, New York, 1999) secs. 10.6–10.7.
  26. [29]V. H. Rumsey, Reaction concept in electromagnetic theory, Phys. Rev. 94, 1483 (1954).
  27. [30]Ansys, Inc., Ansys HFSS, Canonsburg, PA, computer code, versions 2022 R2 (K band) and 2023 R1 (Ka band).

Paper details

vlnc:W_ja2n3c7v
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 9

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