Superconducting Nanowire Single Photon Detectors (SNSPDs) are fast and efficient detectors with important applications in optical quantum computers and quantum networks. Current SNSPD implementations lack intrinsic single-wire photon-number resolution. This thesis discusses the electronic and optical characterization of SNSPDs with the goal of obtaining high-speed, high-efficiency photon-number-resolving detectors. Novel laser-induced relaxation oscillations are observed, caused by laser pulses triggering the detector and changing its biasing condition. This leads to discrete steps in the count rate at integer multiples of the pulsed laser repetition rate and limits the maximum achievable count rate for SNSPDs. A Monte Carlo Markov Chain method is applied to detector tomography to characterize single-photon detectors with unprecedented accuracy, and can be applied to detectors regardless of their detection efficiency. Photon number resolution is achieved in (commercial) SNSPDs by projecting measured voltage traces onto the derivative of the average voltage trace. This method enables discrimination of photon numbers by measuring the time difference between traces in a noise-resilient way. The final chapters discuss the design and characterization of impedance-tapered SNSPDs. These devices show both differences in pulse heights and timing depending on the photon number.
Superconducting Nanowire Single Photon Detectors (SNSPDs) are fast and efficient detectors with important applications in optical quantum computers and quantum networks. Current SNSPD implementations lack intrinsic single-wire photon-number resolution. This thesis discusses the electronic and optical characterization of SNSPDs with the goal of obtaining high-speed, high-efficiency photon-number-resolving detectors. Novel laser-induced relaxation oscillations are observed, caused by laser pulses triggering the detector and changing its biasing condition. This leads to discrete steps in the count rate at integer multiples of the pulsed laser repetition rate and limits the maximum achievable count rate for SNSPDs. A Monte Carlo Markov Chain method is applied to detector tomography to characterize single-photon detectors with unprecedented accuracy, and can be applied to detectors regardless of their detection efficiency. Photon number resolution is achieved in (commercial) SNSPDs by projecting measured voltage traces onto the derivative of the average voltage trace. This method enables discrimination of photon numbers by measuring the time difference between traces in a noise-resilient way. The final chapters discuss the design and characterization of impedance-tapered SNSPDs. These devices show both differences in pulse heights and timing depending on the photon number.
Simulations are run to optimize the ability to resolve the number of photons detected from the pulse height by a Superconducting Nanowire Single-Photon Detector (SNSPD). This is set up in a manner that features an impedance-matching transmission line taper that provides a characteristic impedance which transitions from kΩ to 50 Ω, with the taper providing an effective load impedance that outputs pulses with not only larger amplitudes but also showed a distinct separation for multi-photon events. The first part of this project tries to computationally match the experimental results obtained by our collaborators at MIT. Once these results are achieved, we try to modify the device geometry to improve the key detector metrics which include timing jitter, dark count rate, reset time, and overall detection efficiency.
Single-photon detectors based on superconducting nanowires (SNSPDs) are promising technologies with good comprehensive performance, especially their excellent temporal resolution for applications in quantum precision measurements and time-varying ultimate detection. The response time of an SNSPD can be characterized by aspects of hotspot thermal relaxation and electrical relaxation, which are also responsible for the time jitter of the SNSPD. The factors affecting the temporal resolution of an SNSPD, the underlying physical mechanism, the intrinsic limit, and possible optimization methods are reviewed. The poor energy resolution and lack of photon number resolution for conventional SNSPDs can be overcome by using high-impedance readouts, novel device structures, characterization methods and, most importantly, SNSPD arrays, which can simultaneously further improve the temporal resolution. The study of the temporal and photon number resolution of SNSPDs helps reveal the underlying physics and is valuable for their new application in various practical engineering areas.
Superconducting series nanowire single-photon detectors (SNDs) have been used for resolving photon numbers. But it can also exhibit a higher counting rate than a single superconducting nanowire single-photon detector (SNSPD), because the serial architecture allows multiple detectors working in parallel and the inductance of each pixel is reduced. In this paper, we found each pixel of the SND could reset fast although the whole detector took a much longer time for recovering. Therefore, the speed of an SND with a large kinetic inductance of 1.25 μH was characterized and compared to a standard SNSPD with the same total inductance. By measuring the detected photon rate over the input photon rate, we obtained a photon counting rate enhancement by 7.5 times. Meanwhile, the timing jitter of the SND was 63 ps at single photon detection regime. Our results show that the series structure and the readout scheme are quite practical in realizing high speed single-photon detectors besides resolving photon numbers.
Abstract We experimentally demonstrate high-fidelity detection of 0, 1- and 2-photon states in a commercial superconducting nanowire single photon detector (SNSPD) under 20 MHz pulsed coherent illumination. The arrival of a different number of photons induces distinct temporal dynamics in the disruption of the superconducting state of the detector. Therefore, specific features emerge in the arrival-time histograms, that we show are directly correlated with the photon-number state. We show that, across a broad range of time-averaged light intensities, the probabilities of detecting 1- and 2 + -photon events collapse onto the Poisson distribution of coherent light. We identify the voltage detection threshold as a critical factor for photon number resolving (PNR) capabilities. Even under demanding conditions of high repetition rate illumination, we found that a high trigger level significantly enhances total detection fidelity, achieving 0.920 and 0.845 for 1-photon and for 2-photon events. These results correspond to remarkably low conditional misidentification probabilities, 0.000 02 and 0.0013, describing the classification error among detected 1- and 2-photon events. We identify an inherent 39 ps latency difference between single- and multi-photon detection events. We show that the measured 2-photon conditional misidentification probability corresponds to the one of a multiplexed PNR detector with 730 pixels. We believe that our findings provide a framework for developing high/fidelity spatially multiplexed SNSPD by exploiting the intrinsic PNR capabilities of individual constituting elements.
Superconducting nanowire single photon detectors (SNSPDs) have separately demonstrated high efficiency, low noise, and extremely high speed when detecting single photons. However, achieving all of these simultaneously has been limited by detector subtleties and tradeoffs. Here, we report an SNSPD system with <80 ps timing resolution, kHz noise count rates, and 76% fiber-coupled system detection efficiency in the low-flux limit at 1550 nm. We present a model for determining the detection efficiency penalty due to the detection recovery time, and we validate our method using experimental data obtained at high count rates. We demonstrate improved performance tradeoffs, such as 68% system detection efficiency, including losses due to detector recovery time, when coupled to a Poisson source emitting 100 million photons per second. Our system can provide limited photon number resolution, continuous cryogen-free operation, and scalability to future imaging and GHz-count-rate applications.
Superconducting nanowire single-photon detectors (SNSPDs) are renowned for their high detection efficiency, low dark count rate, and fast timing resolution, which have emerged as an essential device in quantum technologies, including quantum computing and communication, which also significantly benefits classical applications such as deep-space communication, neuromorphic computing, and light detection and ranging (LiDAR). However, designing and customizing SNSPD into versatile functionalities to satisfy these intensive requirements poses significant challenges, including thin-film deposition, nanofabrication, electrical integration, and light delivery strategies. While prior work has typically optimized individual steps – such as fabrication or characterization – there is a lack of reported systematic, integrated frameworks that span all stages from device design to performance benchmarking. This gap hinders broader adoption of the technology across scientific and industrial communities and limits the rigorous exploration of advanced detection concepts. This thesis aims to establish an end-to-end workflow for the customizable fabrication and comprehensive characterization of superconducting nanowire single-photon detectors, and to demonstrate that such a platform is essential for the investigation of diverse single-photon detection architectures. It begins with the optimization of NbTiN thin-film deposition utilizing an ultrahigh vacuum sputtering system. By systematically identifying and optimizing key deposition parameters, we achieve NbTiN films with targeted superconducting properties along with high uniformity and reproducibility. This fast and reliable workflow establishes a solid foundation for detector fabrication and provides instructions for flexible control over film properties. Furthermore, effective nanofabrication protocols for the high-quality SNSPDs were elaborated, along with the specialized cryogenic packaging techniques for reliable electrical and optical integration. The achieved devices demonstrated performance metrics comparable to commercial systems. The implementation of this full fabrication—characterization cycle also enabled flexible exploration of advanced detector architectures. For example, we investigated superconducting microbridge single-photon detectors (SMSPDs) as a complementary platform, demonstrating a fivefold enhancement in signal-to-noise ratio and a tenfold reduction in recovery time relative to conventional SNSPDs. These results provide quantitative insight into detector design trade-offs. Within this framework, photon-number-resolving (PNR) detection using SNSPDs was further investigated as an essential function for advanced quantum and classical applications. We presented innovative detector architectures—interleaved multi-pixel arrays and multilayer coherent absorption structures—to overcome limitations inherent in conventional multiplexing methods, such as nonuniform detection, temporal degradation, and crosstalk. An interleaved multi-pixel SNSPD array with up to four-photon resolution was achieved with uniform detection, as well as a multilayer architecture based on coherent absorption with uniform and near-unity absorption efficiency. Importantly, a theoretical comparison between these two architectures was conducted to evaluate their absorption efficiency and PNR performance, offering insights into structural optimization and device scalability for advanced PNR detectors. Overall, this thesis establishes a fully in-house SNSPD fabrication and characterization capability, providing a robust foundation for the development, customization, and evaluation of high-performance superconducting single-photon detectors. This platform enables reproducible design optimization, quantitative performance benchmarking, and systematic exploration of detector architectures. By integrating fabrication, characterization, and architectural innovation within a consistent workflow, it facilitates both fundamental investigation and application-driven advancement of single-photon detection technologies in our laboratory and the broader research community in Singapore.