The major task of detecting axions or axion-like particles has two challenges. On the one hand, the ultimate sensitivity is required, down to the energy of a single microwave photon of the yoctojoule range. On the other hand, since the detected events are supposed to be rare, the dark count rate of the detector must be extremely low. We show that this trade-off can be approached due to the peculiar switching dynamics of an underdamped Josephson junction in the phase diffusion regime. The detection of a few photons’ energy at 10 GHz with dark count time above 10 s and the efficiency close to unity was demonstrated. Further enhancements require a detailed investigation of the junction switching dynamics.

An aluminium Josephson junction (JJ), with a critical current suppressed by a factor of three compared with the maximal value calculated from the gap, is experimentally investigated for application as a threshold detector for microwave photons. We present the preliminary results of measurements of the lifetime of the superconducting state and the probability of switching by a 9 GHz external signal. We found an anomalously large lifetime, not described by the Kramers’ theory for the escape time over a barrier under the influence of fluctuations. We explain it by the phase diffusion regime, which is evident from the temperature dependence of the switching current histograms. Therefore, phase diffusion allows for a significant improvement of the noise immunity of a device, radically decreasing the dark count rate, but it will also decrease the single-photon sensitivity of the considered threshold detector. Quantization of the switching probability tilt as a function of the signal attenuation for various bias currents through the JJ is observed, which resembles the differentiation between N and N + 1 photon absorption.

Axions and axion-like particles appear in well motivated extensions of the Standard Model of particle physics and may be the solution to the long-standing puzzle of the Dark Matter in our Universe. Several new experiments are forseen in the next decade searching them in a wide range of the parameter space. In the mass region from few to several tens of microelectronvolt, detector sensitivity will be limited by the Standard Quantum Limit of linear amplifiers and a new class of single microwave-photon detector will be needed. We have developed a single photon counter based on the voltage switching of an underdamped Josephson junction that is coupled to a coplanar waveguide. By measuring the switching voltage we can register single photons at 14 GHz with the rate less than 1 photon per 3000 sec.

We have observed the photon noise by measuring a response to the black body 350 GHz radiation and noise of the cold-electron bolometers (CEBs). The experimental results have been fit to the theoretical model of CEBs with two heat-balance equations. The measured noise has been decomposed into several terms with the help of theory. It is demonstrated that the photon noise exceeds any other noise components, which allows us to conclude that the bolometers measure the photon noise. Moreover, a peculiar shape of the noise dependence on the absorbed power originates completely from the photonic component according to the theory. In the additional experiment on heating of the cryostat plate together with the sample holder, we have observed the near independence of the noise on the electron temperature of the absorber, which has provided another proof of the presence of the photon noise in the first experiment. The least ratio between internal and photon noise equivalent powers, observed in our experiments, is 1.1 for the absorbed power of 1–2 pW.

The Cosmic Microwave Background (CMB) radiation is the only observable that allows studying the earliest stage of the Universe. Radioastronomy instruments for CMB investigation require low working temperatures around 100 mK to get the necessary sensitivity. On-chip electron cooling of receivers is a pathway for future space missions due to problems of dilution fridges at low gravity. Here, we demonstrate experimentally that in a Cold-Electron Bolometer (CEB) a theoretical limit of electron cooling down to 65 mK from phonon temperature of 300 mK can be reached. It is possible due to effective withdrawing of hot electrons from the tunnel barrier by double stock, special traps and suppression of Andreev Joule heating in hybrid Al/Fe normal nanoabsorber.

A novel type of resonant cold-electron bolometer (RCEB) is proposed. In this sensor, the internal resonance is organized by the kinetic inductance of an ultrasmall NbN superconducting nanostrip and the capacitance of the nanoscale superconductor-insulator-normal (SIN) tunnel junction. The size of kinetic inductance is about 300 times smaller than the size of geometrical inductance of the same value. This reduction of size gives an opportunity to create nanofilters with a total size considerably smaller than the wavelength. This internal resonance acts as a bandpass filter (instead of external filters) with a bandwidth of 3%-50%, something needed for radioastronomy applications. The use of NbN as Andreev contact material against escaping hot quasiparticles provides a unique opportunity to create an optimal CEB with only one SIN junction and an Andreev SN contact for thermal protection, in contrast to a classical CEB with two SIN junctions. The problem of hot quasiparticles escaping over the Andreev barrier (Δ) for Al electrode is overcome here by creating a barrier higher than quantum energy Δ > hf to have working kinetic inductance. Replacing two SIN junctions by one junction, we could use this junction of half the area to keep the same capacitive coupling. Therefore, a single-junction RCEB would have four times less area of junctions (and volume of absorber) in comparison with a two-junction RCEB and should show better noise performance due to the proportionally decreased e-ph (electron-phonon) noise component. The RCEB can be effectively used to create multiband elements that are actual tasks in submillimeter astronomy due to the benefit that comes from its ability to use co-located data, and problems with the dramatic increase of the size of the focal plane.