Search for Invisible Axion Dark Matter in the $3.3–4.2\mu \mathrm{eV}$ Mass Range

We report the results from a haloscope search for axion dark matter in the $3.3–4.2\mu \mathrm{eV}$ mass range. This search excludes the axion-photon coupling predicted by one of the benchmark models of “invisible” axion dark matter, the Kim-Shifman-Vainshtein-Zakharov model. This sensitivity is achieved using a large-volume cavity, a superconducting magnet, an ultra low noise Josephson parametric amplifier, and sub-Kelvin temperatures. The validity of our detection procedure is ensured by injecting and detecting blind synthetic axion signals.

Figure 1

The ADMX rf diagram. $C_1$, $C_2$, and $C_3$ are circulators. The JPA is connected with a pump power line via a directional coupler. The gray-colored rectangular boxes denote cryogenic attenuators. The switch is connected to the cavity during data taking, the bypass for the hot load is used for system noise calibrations.

Figure 2

Digitization spectra including a blinded SAG for ${\mathrm{TM}}_{010}$ (black) and ${\mathrm{TM}}_{011}$ (red) mode after removing the receiver shape and JPA standing wave distortion. The bottom small plot shows combined SNR with respect to DFSZ power. The right small plot shows the integrated signal power and how it varies according to the Lorentzian cavity enhancement.

Figure 3

Electric field distribution as a heat map for ${\mathrm{TM}}_{010}$ (left) and ${\mathrm{TM}}_{011}$ (right) modes simulated with COMSOL Multiphysics [42]. Electric field for the mode and impressed magnetic field are shown as red and black arrows, respectively. Calculated form factors with CST Magnetic Field Solver are 0.455 and $<0.001$ for ${\mathrm{TM}}_{010}$ and ${\mathrm{TM}}_{011}$ modes, respectively. The overlap between the ${\mathrm{TM}}_{010}$ electric field and the magnetic field is large and consistent across the volume, while the overlap between the ${\mathrm{TM}}_{011}$ electric field is of opposite sign in the top and bottom of the cavity, leading to cancellations in the cavity response to the spatially uniform dark matter axion field.

Figure 4

90% C.L. upper limits on $g_{a\gamma \gamma }$ as a function of axion mass. The gray-, blue-, and yellow-colored areas represent previous ADMX limits reported in Refs. [28, 32, 33]. The red-colored area shows the limits of this work. We ruled out KSVZ (DFSZ) axions in the $3.3–4.2(3.9–4.1)\mu \mathrm{eV}$ mass range.

Figure 5

90% C.L. upper limits on the dark matter energy density assuming the KSVZ model for axion coupling. The blue- and yellow-colored areas represent previous ADMX limits reported in Refs. [32, 33], respectively. The red-colored area shows the limits of this work. The KSVZ axions are excluded even though the axion density is $0.1\mathrm{GeV}/\mathrm{cc}$ (20%) of the total dark matter density.