This paper describes the construction of a cryostat and an optical system with a free-space coupling efficiency of 56.5% ± 3.4% to a superconducting nanowire single-photon detector (SNSPD) for infrared quantum communication and spectrum analysis. A 1K pot decreases the base temperature to T = 1.7 K from the 2.9 K reached by the cold head cooled by a pulse-tube cryocooler. The minimum spot size coupled to the detector chip was 6.6 ± 0.11 µm starting from a fiber source at wavelength, λ = 1.55 µm. We demonstrated photon counting on a detector with an 8 × 7.3 µm2 area. We measured a dark count rate of 95 ± 3.35 kcps and a system detection efficiency of 1.64% ± 0.13%. We explain the key steps that are required to improve further the coupling efficiency.
© 2016 Optical Society of America
Free-space quantum optical communication in the infrared (IR)  is an important technology for applications such as naval operations that cannot rely on optical fibers. The mid-infrared (mid-IR) range is particularly interesting because of a window of transmission in the atmosphere at wavelength λ = 10 µm. These communications require high-speed single-photon detectors sensitive to mid-IR radiation. At present, superconducting nanowire single-photon detectors (SNSPDs) [2,3] represent one of the best detectors for this application, due to their single-photon sensitivity, their high speed (few-ns reset time), and their high time resolution (few tens of ps time-jitter). System detection efficiencies (SDE) greater than 67% [4–6] were demonstrated for SNSPDs at near-infrared (near-IR) wavelength (λ = 1.55 µm). This demonstration was achieved not only by increasing the devices detection efficiency but also by maximizing the coupling efficiency, i.e. the fraction of photons emitted by the source that are coupled to the SNSPD. In all the cited cases, an optical fiber was aligned to an SNSPD, either passively  or actively . We identified three main reasons why we cannot apply the same alignment method to mid-IR optical communications. First, the dimensions of the detector depend on the fiber used. An optical fiber has a fixed core diameter that consequently limits the minimum diameter of the beam emitted by the fiber. To ensure high coupling efficiency, the minimum dimension of the detector has to be larger than the beam diameter; thus, this dimension is also limited. This requirement becomes a non-trivial issue for experiments in the mid-IR, which require optical fibers with a mode-field diameter MFD > 20 µm versus the typical active area of an SNSPD which is ≤ 15 × 15 µm2. SNSPDs with active area > 15 × 15 µm2 have been successfully fabricated using parallel nanowire configuration to compensate for the loss in reset speed [7,8]. However, the reset time for large-area SNSPDs reported is still > 20 ns, which limits the maximum count rate of the detector. Second, using mid-IR fibers would limit the scalability of the system to multiple channels. Every channel in the system requires an optical fiber; thus, for a large array of detectors a bundle of fibers is required, and integrating this bundle into a cryostat with tight spaces can make the design challenging. Finally, mid-IR optical fibers are more rigid and fragile than near-IR fibers, and the only commercially available prototypes are multi-mode: there is a long-term bend radius of > 40 mm for mid-IR fibers vs a long-term bend radius of > 13 mm for near-IR fibers. Thus, it would be more difficult to thermalize, cleave, or integrate a mid-IR fiber with components affected by thermal contraction. In addition, the lack of single-mode 10-µm-wavelength fibers makes them less desirable for optical communication.
An optical system based on free-space optics solves these three problems. In diffraction optics, an optical beam can be focused on a spot whose minimum diameter is smaller than the wavelength of the beam itself. In addition, a free-space optical system can host several channels in a single optical path as long as there is no cross talk at the receiver. Depending on the spot dimension achieved and the diffraction limit of the system, we can determine how many channels in parallel the system can accommodate.
To date we could not find any demonstration of a cryogenic system with free-space high-coupling efficiency to a single SNSPD at 1.55-µm-wavelength or higher. Free-space coupling to cryogenic detectors was proposed in the past for astronomical imaging [9,10] and spectroscopy [11,12], and for quantum communications [13,14]. Some of these proposals used semiconductor devices. These detectors could be fabricated in arrays with an area four orders of magnitude larger than an SNSPD; however, they were not single-photon sensitive. In addition, the base temperature needed for these devices to operate (~6 K , ~50 K ), allowed the use of cryostats with a cooling power that is not available below 4.2 K. Other systems used cryogenic detectors to receive quasi-optical millimeter and sub-millimeter radiation . In this case, it is possible to build in-plane antennas that can efficiently focus the radiation onto the detector. Thus, the active area of the receiver is effectively mm2-scale. Mazin et al.  achieved free-space coupling to superconducting detectors at a base temperature of 100 mK using an adiabatic demagnetization refrigerator, but they limited the wavelength detection range up to 1.1 µm. Verevkin et al.  used free-space coupling for SNSPDs, but they state that “the working area of our detectors is always smaller and often much smaller than the incident photon beam size”. Finally, Allman et al.  coupled a 1.55-µm-wavelength beam to 64-pixel SNSPD array for imaging purposes, while we were interested in coupling to single SNSPDs for optical communications. We propose here a cryogenic setup for superconducting single-photon detectors built to achieve high-efficiency (> 50%) free-space coupling on a single device.
We designed and built a vibration-isolating cryostat with free-space optical access able to reach a base temperature T = 2.9 K, with an additional stage that can cycle the sample stage to T = 1.7 K for 1.5 hours; we measured vibrations amplitudes of 498 ± 98 nm at the sample stage. The optical system, composed of two lenses (see Fig. 1(b)), was able to focus light on a detector with a minimum spot waist of 6.6 ± 0.11 µm at λ = 1.55 µm. One of the two lenses was mounted and thermalized inside the cryostat, and it was aligned to the SNSPD chip before the cooldown.
We used an 8 × 7.3 µm2 area NbN-on-sapphire SNSPD based on 100-nm-wide nanowires 50%-fill-factor to calibrate our set-up. We biased the detector at 97% of its switching current (the current at which the detector stops being superconducting and switches to a resistive state). At this set point, we measured a dark count rate of 95 ± 3.35 kcps. At the same bias current we measured an SDE of 1.64% ± 0.13%. By characterizing the dimension of the beam at the detector, we estimated a coupling efficiency of 56.5% ± 3.4%. From the ratio between the two efficiencies, we calculated that the SNSPD’s device detection efficiency (DDE) was 2.9% at the same bias current.
This document is divided as follows: in Sections 2 and 3, we present the design of the experimental setup divided into the optical system and the cryogenic system, respectively. In Section 4, we present the results of the mechanical vibrations measurements. In Section 5, we analyze the results obtained from the measurements of SDE on large area SNSPDs. Finally, we conclude by discussing the impact of this report on future work.
2. Optical system
We designed a three-lens optical system to image the surface of the SNSPD chip and to focus a test beam from a laser source on a detector. A picture and schematic of the system are shown in Fig. 1. The imaging system was designed to achieve a field of view of ~200 µm. Imaging the SNSPD chip allowed us to align the optical beam to the detector and to place Lens 3 at the correct focal distance. Even though the ultimate goal was to use the optical system at mid-IR wavelengths, for this first demonstration we created an optical system able to focus a beam spot of 12.7 µm in diameter at λ = 1.55 µm.
We designed a two-lens telescope to couple > 90% of the light coming from the source onto the active area of the detector; because of the large numerical aperture required to couple λ = 10 µm light on SNSPD, we had to mount Lens 3 inside the cryostat. The yield in the fabrication process and the constraints in the detector’s speed due to the kinetic inductance of the NbN [15, 16], limit our SNSPDs maximum dimensions. An NbN SNSPD with a 15 × 15 µm2 active area, 80-nm-wide nanowires and 40% fill factor typically showed a reset time of 9 ns, which is as high as we wanted to go for 100 MHz optical communications. We determined from Gaussian optics that in order to have 90% of the source power impinging on an active area of 15 × 15 µm2, the beam waist had to be no larger than 7.5 µm. As a consequence, we determined that at λ = 10 µm we needed a numerical aperture of NA = 0.41 at Lens 3; thus, for a 25-mm-diameter lens, rather than a larger lens, the maximum acceptable focal length is fLens3 ~28 mm. We chose a 25-mm-diameter lens because increasing the aperture of the optical system not only would have increased the stray light impinging on the detector, but it would have also compromised the cooling ability of the cryostat, due to the increased incoming thermal radiation. In addition, because of the short focal length of Lens 3, we mounted it inside the cryostat, as shown in Fig. 1(b). From the commercially available lenses, we selected an aspheric lens for Lens 3 with f = 20 mm with IR anti-reflective coating (1050 – 1700 nm).
Starting from our selection for Lens 3 and design wavelength, we picked Lens 1 from commercially available lenses. As we mentioned earlier, for the test described in this document we used a fiber-coupled coherent light source at λ = 1.55 µm, as a signal. We characterized the beam profile at the output of the optical fiber, and we observed a Gaussian beam with a beam quality M2 = 1.35 and a beam waist of 8.3 ± 0.05 µm. Thus, we chose an aspheric lens with fLens1 = 26 mm, so that the demagnification of the telescope was 1.3 × . In Fig. 2(a), we show the profile of the Gaussian beam at the output of the telescope characterized with a beam profiler. We obtained a minimum Gaussian beam waist of 4.82 ± 0.04 µm, which was close to the 4.8 µm waist that we obtained from theoretical calculations .
The fLens4 was selected to obtain a field of view of 200 µm. In Fig. 1(b), the green lines represent the path followed by the incoherent visible light (λ = 635 nm) in the imaging system. The light back-reflected by the SNSPD chip is focused through a telescope formed by lenses 2 and 3 on a CCD camera. The active area of the camera chip is 12.5 × 12.5 mm2; thus, for a field of view of 200 µm we needed a magnification of 25 × . Because fLens3 = 20 mm, we used fLens4 = 500 mm. Figure 2(b) shows an image acquired with the optical setup and centered on an SNSPD.
Because of the imaging system, we are able to align the focusing system to the SNSPD in two separate steps . First, Lens 3 was aligned to center the image on the selected SNSPD. In a second step, we aligned the optical source and Lens 1 so that the beam spot was centered on the detector. The use of two separate optical systems allowed us also to verify that Lens 1 was at the focal distance from the SNSPD chip; in particular, we were able to verify that the image was focused on the same plane where the beam from Source 2 was focused.
3. Cryogenic set-up
For single-photon optical communications in the mid-IR, we needed a cryostat able to reach a base temperature T < 2 K, even with an optical opening. The two main reasons for needing that temperature are dark counts and material used for the SNSPDs. It has been demonstrated in the past that reducing the base temperature of an SNSPD even by a few tenths of K can significantly change the detector’s dark count rate ; thus, the temperature should be as low as possible. In addition, we wanted to be able to operate SNSPDs based on WSi  and not NbN, which are more sensitive to mid-IR photons.
We built a cryostat able to reach a base temperature T = 1.7 K at the sample stage; the system was precooled to T = 2.9 K with a Cryomech pulse-tube cryocooler (PT415), and then reached base temperature T = 1.7 K using a sorption fridge from PhotonSpot Inc. The system was also designed to isolate the mechanical vibrations generated by the pulse-tube from the sample stage (see Section 4).
In Fig. 3(a), we show a schematic of the cryostat. The system inside the external chassis can be physically separated into two assemblies. The top assembly is shown in Fig. 3(b), and it is mainly responsible for cooling the system: the pulse-tube cryocooler (Cryomech PT415) has a first stage (Stage 1) with a cooling power Pc = 36 W at T = 45 K, and a second stage (Stage 2) with Pc = 1.5 W at T = 4.2 K; the cold head of the sorption fridge (PhotonSpot Freeze 4) can reach a temperature T < 1 K, for a time that depends on the heat load (Qc) incident on it and on the base temperature reached by the pulse-tube cryocooler. The bottom assembly is shown in Fig. 3(c). This assembly hosts the SNSPD chip and Lens 3, and is mounted directly on the optical table; thus, we can perform optical alignment to the detector before cooling down the entire system. The components in Fig. 3(b) and 3(c) with the same label have been thermally connected with oxygen-free high thermal conductivity (OFHC) copper. The copper parts are bolted together, with a layer of Apiezon vacuum grease in between to improve the heat conduction. This flexible thermal coupling allowed us to detach the bottom and the top assemblies, and easily access the cooling stages of the system while keeping the optics in the cryostat on the optical table. Lens 3 is placed in an aluminum lens mount mounted to a 25-mm-diameter stainless steel post and bolted to the bottom plate of the radiation shield with a clamp. The metal post guarantees that the lens is thermalized to the temperature of the radiation shield.
The materials and the geometry of the system were selected with the goal to minimize Qc on the pulse-tube cryocooler and on the sorption fridge . Oxygen-free high-conductivity (OFHC) copper is one of the most widely used materials for cryogenic thermal connections because of its unmatched thermal conductivity. G-10 is another widely used material in cryogenics; because of its low thermal conductivity and high rigidity, G-10 is an ideal material for rigid connections between parts at different temperatures.
Table 1 shows the heat load (Qc) budget for the top and bottom assemblies. For the top assembly, as shown in Fig. 3(b), we used G-10 bars to hang the stage at T = 35 K (Stage 1) from the top flange of the cryostat (T = 300 K), and the stage at T = 3 K (Stage 2) from Stage 1. The 35K stage and the 3K stage are cooled by two stages of the cryocooler. The sorption fridge is mounted on and precooled by the Stage 2. In the bottom assembly, we have used bars made of G-10 and knuckles made of aluminum glued with Stycast Epoxy to mount Stage 1 on the bottom of the cryostat, and then Stage 2 on top of Stage 1. We have also used a G-10 post to mount the chip holder on Stage 2. The heat load and the imperfect connection with the cold head of the sorption fridge allowed the sample stage to reach a base temperature of 1.7 K.
We estimated the contribution to Qc from thermal radiation . We calculated that the heat radiated from the room temperature stage to Stage 1 is ~2.5 W. We therefore constructed a radiation shield mounted directly to Stage 1 both at the top and at the bottom of the cryostat. In addition to the radiation shield, we used a 10 layer superinsulation made of aluminum-coated Mylar between the radiation shield and the outer walls (at room temperature), between the radiation shield and Stage 2, and around the copper braid connecting the cold head of the sorption fridge to the chip holder. Each layer of Mylar is expected to reduce the amount of incident heat due to radiation up to a factor 2. We cut a 25-mm-diameter window in the radiation shield to allow the light from Source 2 to access the cryostat and used two optical filters to get attenuate the stray radiation (i.e. not coming from the source). We calculated that 225 mW of power would pass through the window without filters. This amount of power would have compromised the cooling performance of the cryostat and the operation of the SNSPD by adding optical noise. Thus, we mounted a 1.05 µm IR longpass filter (Edmund Optics #68-651) on the window of the radiation shield and a bandpass 1550 nm interference filter with 10 nm FWHM bandwidth (Edmund Optics #86-091) on the same optical mount of Lens 3.
4. Mechanical vibrations
Another requirement for our cryostat was to reduce the mechanical vibrations between the optical system and the SNSPD below 3 µm. Mechanical vibrations with amplitude > 3 µm can reduce the average coupling efficiency of the optical setup by more than 10%. When the Cryomech pulse-tube was operated without modification, the 3K stage vibrated with an RMS amplitude of 10 µm. As shown in Fig. 3(a), the cryocooler was rigidly fixed at the top to the external chassis of the cryostat and to the optical table. Because of the large mass of the entire system (600 kg including the optical table) the vibrations were damped at the top of the cryostat. However, the cryocooler acted like a vertical cantilever, so that the two pulse-tube stages still vibrated. We further isolated the vibrations by using a combination of OFHC copper bars and braids to connect the pulse-tube stages to Stage 1 and Stage 2 at the top and bottom of the cryostat. The copper bars guaranteed high heat conduction, while the soft braids decoupled the detector from the cantilevered vibrations from the pulse-tube.
We determined that the amplitude of the vibrations at the sample stage was 498 ± 98 nm by using the time-dependence of the count rate from the detector. Figure 4(a) shows the count rate measured on an 8 × 7.3 µm2 area NbN SNSPD, when the beam was shifted along the longer axis of the SNSPD by 4.6 ± 0.5 µm from the center of the detector. When the center of the beam and the center of the detector were coincident, we could not see the same oscillations because of the small effect on the light coupling. On the edge of the detector, a change in the beam position produced a larger relative change in the impinging power. Figure 4(b) shows the Fourier Transform of the signal on Fig. 4(a). A peak was present at a frequency of 1.5 Hz, which was the frequency of the piston of the pulse-tube’s engine. Figure 4(c) shows the average and the standard deviation (error bar) of the count rate measured as a function of the distance between the beam’s center and the SNSPD’s center. The curve was fitted to extrapolate a beam waist of 6.6 ± 0.11 µm. From the beam waist and the standard deviation of the count rate, we determined a vibration amplitude of 498 ± 98 nm at 1.5 Hz. We measured that the beam power changed with a standard deviation of 5.25%, and we included it in the vibration amplitude error.
We measured the vibrations of the sample stage at higher frequencies by using the oscillations in the power of the light reflected by a highly reflective chip, shown in Fig. 4(d). We focused the infrared beam on the edge of a Si chip coated with a 50 nm Au layer mounted on the sample holder. The light reflected by the chip was focused on a fast free-space photodetector (bandwidth DC – 460 kHz), which was connected to an oscilloscope. From the oscillations traces we observed vibrations at ~1.5 Hz and 19 Hz with amplitudes of 170 ± 50 nm and 91 ± 50 nm, respectively, which can be seen also in the FFT graph in Fig. 4(e). Figure 4(f) shows a zoom-in of the FFT graph to highlight the peaks. Although this measurement shows a lower vibration amplitude compared to the measurement described in the previous paragraph, the frequency is the same that we observed in the count rate measurement. We were not able to explain the origin of the vibrations observed at 19 Hz. We did not observe oscillations at higher frequencies.
5. Free-space coupling demonstration
In Section 4 we proved that we could couple near-IR light with free-space optics on an 8 × 7.3 µm2 area NbN SNSPD based on 100 nm wide nanowires and 50% fill factor. We tested the SDE using the same detector. The SDE is defined as the ratio between the photon count rate (PCR) registered by the detector (excluding dark counts) and the photon flux measured in fiber at the optical source. The count rate measured when the optical source is off is the system dark count rate (SDCR).
In Fig. 5(a), we plotted SDCR (blue squares) and PCR (red triangles) as a function of the bias current, Ibias, applied to the detector normalized by its switching current (Isw). As we can see from the graph, SDCR < PCR for Ibias up to 97% of Isw at 1.7 K. Thus, we can reliably extract PCR from the total counts and the SDE, which we plotted in Fig. 5(b) as a function of Ibias applied to the detector normalized by Isw. The system reached a maximum SDE of 1.64% ± 0.13%. For this test, we used a coherent light source with λ = 1.55 µm with a power of 710 ± 37 fW measured in fiber outside the cryostat, which corresponds to a total photon rate of 5.53 ± 0.29 Mphoton/s. From the beam waist measurement described in Section 4 (w = 6.6 ± 0.11 µm) and the active area of the detector, we calculated that the coupling efficiency (CE) of the system was 56.5% ± 3.4%. Thus, we can calculate the maximum device detection efficiency DDE = SDE/CE = 2.9%.
The scope of this work was to create a free-space coupled cryogenic system and demonstrate coupling efficiency CE > 50% at a base temperature < 1.7 K. The use of free-space optics allows us to adapt the optical setup for a different wavelength range by replacing the lenses. Our demonstration showed promising results of obtaining SDE > 50%. In particular, if we were to use a 10 × 10 µm2 active area detector, we would achieve a CE = 76%, and with a DDE > 65% (already demonstrated for optical cavity integrated NbN SNSPDs) we would obtain a receiver with SDE > 70%. In addition, our measurements showed that we can reliably characterize the DDE of an SNSPD.
We could improve the performance of the cryostat if we were to use e-beam welded copper braids and if we were to mount the chip directly on the cold head of the sorption fridge. The OFHC copper braids that we use at the moment are soldered with silver based solder to copper plates for easy mounting. If we were to e-beam weld those parts instead, we could obtain a conductivity twice as high as the present value. In addition, mounting the chip directly to the sorption fridge would reduce the temperature gradient between the two. This change requires a substantial redesign of the cryostat and thus measuring that the vibrations amplitude of the chip holder are unchanged.
Our ultimate goal is to demonstrate high coupling efficiency with mid-IR optics at λ = 10 µm. The use of free-space optics allows us to adapt the optical setup for a different wavelength range by replacing the lenses. For future experiments at 10-µm-wavelength, we will replace the optical components with materials compatible with the mid-IR, such as germanium or zinc-selenide. In addition, we will replace Lens 1 with a larger focal length lens, while keeping Lens 3 at the same focal length, because we will need a stronger demagnification at mid-IR wavelength. We estimate that losses due to these components to be around 9% due to reflections. Furthermore, we will replace the CCD camera with an near-IR camera to image the SNSPD chip. It is important to point out that moving to mid-IR wavelengths will require a substantial engineering effort to filter parasitic radiation. We calculated that even if we were to use a spatial filter on the radiation shield to cut off thermal radiation at T = 300 K, we could still observe > 109 photon/s due to stray radiation. Commercial 10-µm-wavelength, bandpass filters have a bandwidth of 1 µm at best, which can only reduce the stray radiation to 108 photon/s. Thus, we are designing a double-monochromator setup to integrate inside the cryostat that will act as a bandpass filter. From an initial design, we calculated that this kind of set-up could reduce the bandwidth of the stray radiation < 0.2 µm. Once we can prove high coupling efficiency at the mid-IR wavelength, we will attempt to couple multiple sources to an SNSPD array.
Finally, we plan to operate SNSPDs based on WSi, instead of NbN. The reason for this consideration is that WSi is a material with a lower bandgap than NbN, so it is more sensitive to low energy photons. As the bandgap in WSi is lower than in NbN, proportionally the critical temperature of WSi is lower. A temperature T < 2 K is required to operate thin-film WSi SNSPDs. This choice of material comes with a significant drawback in the speed of the devices. SNSPDs based on WSi have a much larger kinetic inductance compared to SNSPDs based on NbN. Marsili et al. in  demonstrated high-detection-efficiency WSi SNSPD with 15 × 15 µm2 active area with 120 ns reset time. We expect that finding a compromise between active area and reset time will represent the biggest engineering challenge on the detectors.
Our system represents the first step in the realization of a mid-IR single-photon receiver for free-space optical communication. Upgrading this receiver to a multi-channel system can allow single-photon communication at 100 Mbit/s. This technology could not only allow low-power secured maritime communications, but it could also be applied to space-to-earth communication.
Outside of the secured optical communication framework, our free-space coupled system could be used for several other applications. A version of our system with visible-wavelength optics could be used to study the time-resolved emission of NV-centers, which are a key technology for quantum photonics. Mid-IR spectroscopy has become important for gas-sensing, and we could use a multi-channel version of our system to study the evolution of µs-time-scale chemical reactions.
The authors thank J. Daley, M. Mondol, and Dr. V. Anant for technical support and Dr. J. Mower, Dr. F. Marsili, and Dr. Y. Ivry. This work was sponsored by the Office of Naval Research (ONR) BAA 13-001 and by the Intelligence Advanced Research Projects Activity (IARPA) via Air Force Research Laboratory (AFRL) contract number FA8650-11-C7-105. A.E.D. and A.M.C. were supported by the iQuISE fellowship.
References and links
2. G. N. Gol’tsman, O. Okunev, G. Chulkova, A. Lipatov, A. Semenov, K. Smirnov, B. Voronov, A. Dzardanov, C. Williams, and R. Sobolewski, “Picosecond superconducting single-photon optical detector,” Appl. Phys. Lett. 79(6), 705 (2001). [CrossRef]
3. F. Marsili, F. Bellei, F. Najafi, A. E. Dane, E. A. Dauler, R. J. Molnar, and K. K. Berggren, “Efficient Single Photon Detection from 500 nm to 5 μm Wavelength,” Nano Lett. 12(9), 4799–4804 (2012). [CrossRef] [PubMed]
4. F. Marsili, V. B. Verma, J. Stern, S. Harrington, E. Lita, T. Gerrits, I. Vayshenker, B. Baek, M. D. Shaw, R. P. Mirin, and S. W. Nam, “Detecting single infrared photons with 93% system efficiency,” Nat. Photonics 7(3), 210–214 (2013). [CrossRef]
5. D. Rosenberg, A. J. Kerman, R. J. Molnar, and E. A. Dauler, “High-speed and high-efficiency superconducting nanowire single photon detector array,” Opt. Express 21(2), 1440–1447 (2013). [CrossRef] [PubMed]
6. T. Yamashita, S. Miki, H. Terai, and Z. Wang, “Low-filling-factor superconducting single photon detector with high system detection efficiency,” Opt. Express 21(22), 27177–27184 (2013). [CrossRef] [PubMed]
7. F. Mattioli, M. Ejrnaes, A. Gaggero, A. Casaburi, R. Cristiano, S. Pagano, and R. Leoni, “Large area single photon detectors based on parallel configuration NbN nanowires,” J. Vac. Sci. Technol. B Microelectron. Nanom. Struct. 30(3), 031204 (2012).
8. A. Casaburi, A. Pizzone, and R. H. Hadfield, “Large area superconducting nanowire single photon detector arrays,” Proc. Fotonica AEIT Italian Conf. Photon. Technol. (2014), pp. 1–4. [CrossRef]
9. T. Onaka, Y. Sugiyama, and S. Miura, “Telescope system of the infrared imaging surveyor (IRIS),” Infr. Astron. Instr. 3354, 900–904 (1998). [CrossRef]
10. M. S. Allman, V. B. Verma, M. Stevens, T. Gerrits, R. D. Horansky, A. E. Lita, F. Marsili, A. Beyer, M. D. Shaw, D. Kumor, R. Mirin, and S. W. Nam, “A near-infrared 64-pixel superconducting nanowire single photon detector array with integrated multiplexed readout,” Appl. Phys. Lett. 106(19), 192601 (2015). [CrossRef]
11. K. H. Hinkle, R. Cuberly, N. Gaughan, J. Heynssens, R. Joyce, S. Ridgway, P. Schmitt, and J. E. Simmons, “Phoenix : a cryogenic high-resolution 1-5 micron infrared spectrograph,” Proc. SPIE 3354, 810–821 (1998). [CrossRef]
12. B. A. Mazin, S. R. Meeker, M. J. Strader, P. Szypryt, D. Marsden, J. C. van Eyken, G. E. Duggan, A. B. Walter, G. Ulbricht, M. Johnson, B. Bumble, K. O’Brien, and C. Stoughton, “ARCONS: A 2024 pixel optical through near-IR cryogenic imaging spectrophotometer,” Publ. Astron. Soc. Pac. 125(933), 1348–1361 (2013). [CrossRef]
13. T. H. Buttgenbach and S. Member, “An improved solution for integrated array optics in quasi-optical mm and submm receivers : the hybrid antenna,” IEEE Trans. Microw. Theory Tech. 41(10), 1750–1760 (1993). [CrossRef]
14. A. A. Verevkin, J. Zhang, W. Slysz, R. Sobolewski, A. P. Lipatov, O. Okunev, G. Chulkova, A. Korneev, and G. N. Gol’tsman, “Superconducting single-photon detectors for GHz-rate free-space quantum communications,” Proc. SPIE 4821, 447–454 (2002).
15. A. J. Kerman, E. Dauler, W. E. Keicher, J. K. W. Yang, K. K. Berggren, G. Gol’tsman, and B. Voronov, “Kinetic-inductance-limited reset time of superconducting nanowire photon counters,” Appl. Phys. Lett. 88(11), 111116 (2006). [CrossRef]
16. A. J. Kerman, E. Dauler, J. K. W. Yang, K. M. Rosfjord, V. Anant, K. K. Berggren, G. N. Gol’tsman, and B. M. Voronov, “Constriction-limited detection efficiency of superconducting nanowire single-photon detectors,” Appl. Phys. Lett. 90(10), 101110 (2007). [CrossRef]
18. T. W. J. Geist, N. Foerster, D. Hengstler, S. Kempf, E. Pavlov, C. Pies, P. Ranitzsch, S. Schäfer, V. Schultheiss and C. E. L. Gastaldo, A. Fleischmann, “Low temperature particle detectors with magnetic penetration depth thermometers.” LTD-15 Abstract (2013).
19. J. Kitaygorsky, I. Komissarov, A. Jukna, D. Pan, O. Minaeva, N. Kaurova, A. Divochiy, A. Korneev, M. Tarkhov, B. Voronov, I. Milostnaya, G. Gol, and R. R. Sobolewski, “Dark Counts in Nanostructured NbN Superconducting Single-Photon Detectors and Bridges,” IEEE Trans. on Appl. Supercon. 17(2), 275–278 (2007).
20. B. Baek, A. E. Lita, V. Verma, and S. W. Nam, “Superconducting a-WxSi1-x nanowire single-photon detector with saturated internal quantum efficiency from visible to 1850 nm,” Appl. Phys. Lett. 98(25), 251105 (2011). [CrossRef]
21. J. W. Ekin, Experimental Techniques for Low-Temperature (Oxford University Press Inc., 2006)