{"project":{"acronym":"","projectId":88532,"title":"Modeling and Development of Superconducting Nanowire Single Photon Detectors","primaryTaxonomyNodes":[{"taxonomyNodeId":10647,"taxonomyRootId":8816,"parentNodeId":10646,"level":3,"code":"TX05.1.1","title":"Detector Development","definition":"Detector development includes the development of high detection efficiency, low-dark-count, low-jitter photon counting detectors and readout systems for both ground and flight applications.","exampleTechnologies":"Tungsten silicide (WSi) superconducting arrays, high T superconducting arrays, e.g., MgB2, indium gallium arsenide (InGaAs) flight arrays","hasChildren":false,"hasInteriorContent":true}],"startTrl":2,"currentTrl":3,"endTrl":3,"benefits":"JPL is currently developing 64-pixel SNSPD arrays for use as the ground receiver of tests of deep space optical communication (DSOC) links. This research project would provide an opportunity to research this device physics in a level of detail not practical for the JPL team, while providing feedback that can improve the performance of this crucial space technology.","description":"This proposal outlines a research project as the central component of a Ph.D. program focused on the device physics of superconducting nanowire single photon detectors (SNSPDs). This program would be carried out by a profile 4 candidate at the California Institute of Technology in collaboration with the JPL and seeks 3.5 years of fellowship support. The research of this proposal aims to use a balance of theory, simulation, and experimental methods to investigate the underlying device physics of SNSPDs. SNSPDs are a recently developed class of optical detector with single photon sensitivity and impressive performance characteristics including high count rates, high efficiency, and low dark count rates. Despite the rapid progress in their development since their first demonstration fifteen years ago, the specific details of certain aspects of their operation are a topic of discussion in the literature. In particular, the detection mechanism of SNSPDs and the evolution of quasiparticles within a hotspot are not fully understood. Through first principles modeling and corresponding experimental validation, this research aims to expand the current knowledge of both engineering related design challenges and the fundamental physics that dictates SNSPD behavior. Initial investigations into the hotspot evolution within amorphous superconductors and heat dissipation within these devices are expected to yield design criteria for optimizing SNSPD arrays. These initial questions will lead to a more fundamental analysis of the dynamics of quasiparticles within a superconductor which are expected to provide the knowledge necessary to design higher performance SNSPDs. JPL is currently developing 64-pixel SNSPD arrays for use as the ground receiver of tests of deep space optical communication (DSOC) links. This research project would provide an opportunity to research this device physics in a level of detail not practical for the JPL team, while providing feedback that can improve the performance of this crucial space technology. This makes the work directly relevant for TABS Element 5.1.1 (optical communication detector development) while carrying out the investigation in a way both not practical for the current NASA team and substantial enough to be a Ph.D. project.","startYear":2016,"startMonth":9,"endYear":2020,"endMonth":5,"statusDescription":"Completed","principalInvestigators":[{"contactId":266007,"canUserEdit":false,"firstName":"Keith","lastName":"Schwab","fullName":"Keith Schwab","fullNameInverted":"Schwab, Keith","publicEmail":false,"nacontact":false}],"programDirectors":[{"contactId":84634,"canUserEdit":false,"firstName":"Claudia","lastName":"Meyer","fullName":"Claudia M Meyer","fullNameInverted":"Meyer, Claudia M","middleInitial":"M","primaryEmail":"claudia.m.meyer@nasa.gov","publicEmail":true,"nacontact":false}],"programExecutives":[{"contactId":84634,"canUserEdit":false,"firstName":"Claudia","lastName":"Meyer","fullName":"Claudia M Meyer","fullNameInverted":"Meyer, Claudia M","middleInitial":"M","primaryEmail":"claudia.m.meyer@nasa.gov","publicEmail":true,"nacontact":false}],"programManagers":[{"contactId":183514,"canUserEdit":false,"firstName":"Hung","lastName":"Nguyen","fullName":"Hung D Nguyen","fullNameInverted":"Nguyen, Hung D","middleInitial":"D","primaryEmail":"hung.d.nguyen@nasa.gov","publicEmail":true,"nacontact":false}],"projectManagers":[{"contactId":323559,"canUserEdit":false,"firstName":"Matthew","lastName":"Shaw","fullName":"Matthew D Shaw","fullNameInverted":"Shaw, Matthew D","middleInitial":"D","primaryEmail":"matthew.d.shaw@jpl.nasa.gov","publicEmail":true,"nacontact":false}],"coInvestigators":[{"contactId":205037,"canUserEdit":false,"firstName":"Jason","lastName":"Allmaras","fullName":"Jason P Allmaras","fullNameInverted":"Allmaras, Jason P","middleInitial":"P","primaryEmail":"jason.p.allmaras@jpl.nasa.gov","publicEmail":true,"nacontact":false}],"website":"https://www.nasa.gov/strg#.VQb6T0jJzyE","libraryItems":[],"transitions":[{"transitionId":75921,"projectId":88532,"partner":"Other","transitionDate":"2016-09-01","path":"Advanced From","infoText":"Advanced from another project within the program","infoTextExtra":"Another project within the program","dateText":"September 2016"},{"transitionId":75922,"projectId":88532,"transitionDate":"2020-05-01","path":"Closed Out","details":"Superconducting nanowire single-photon detectors (SNSPDs) [1] have demonstrated remarkable efficiency, timing resolution, and intrinsic dark count rate properties, making them an appealing technology for low phonon flux optical communication detectors. During my time with the NSTRF program, I worked the with SNSPD development team at JPL to explore the physics and optimization of SNSPDs through experimental, computational, and theoretical methods. The combination of the fellowship and integration of the JPL team gave me the flexibility to investigate a range of topics with the broad goal of improving SNSPDs for optical communication and imaging applications. The main topics of research were (1) optimization and characterization of SNSPD arrays for the Deep Space Optical Communication (DSOC) ground detector, (2) design, fabrication, and testing of a new multiplexing technique for imaging arrays known as the thermally-coupled row-column array (thermal row-column), and (3) investigation of the detection mechanism physics through experimental and modeling efforts. This work culminated in the completion of my Ph.D. thesis “Modeling and Development of Superconducting Nanowire Single-Photon Detectors.” My work with the SNSPD team began with optimization and characterization of 64-pixel arrays of WSi nanowires for the DSOC ground detector. Optimizing detectors for free-space optical communication presents a unique set of challenges. In order to couple to a large telescope, detectors must have a relatively large active area. It is beneficial to achieve uniform illumination over the active pixels in order to permit the highest possible count rate over the array, but for an SNSPD array, this requires co-winding the pixels as shown in Figure 1(b). Unfortunately, if the nanowires are too close together, this leads to crosstalk. Crosstalk can be avoided by increasing the separation between wires, but this reduces the optical coupling efficiency. To investigate the possibility of achieving high optical efficiency with low nanowire fill factor, I improved upon existing rigorous coupled-wave analysis (RCWA) modeling software in order to calculate the optical efficiency of the device with different anti-reflection (AR) coating designs. By adopting a multi-layer AR design using TiO2 and SiO2, modeling predicted that the array could achieve meet the efficiency requirements of the DSOC program while keeping the nanowires sufficiently far apart to prevent crosstalk. When fabricated, arrays using this optical design demonstrated 75% optical efficiency, which was sufficient for meeting the DSOC requirements. The optical design of the DSOC ground system requires that the detector is coupled to the telescope with a large numerical aperture. Typically, optical designs for SNSPDs only optimize for illumination from the direction perpendicular to the detector, but with a large numerical aperture, one must consider illumination from other angles of incidence. To handle this, I used the improved RCWA model to calculate the effect of off-axis illumination on the detector efficiency. To validate the model, I devised and performed a set of measurements on a prototype DSOC detector using a cryogenic lens which permitted measuring a range of incidence angles on the detector with only a limited window through the radiation shields of the cryostat and without requiring any cryogenic moving parts. The results of the model were in good qualitative agreement with the experimental measurements, demonstrating the accuracy of the optical model. One of the challenges of SNSPD development is scaling the performance of single pixels to the kilopixel and megapixel scale. With traditional readout, each pixel of an array has a dedicated low noise RF amplifier and time-to-digital converter. However, it is not practical to scale this readout to the kilopixel or megapixel scale due to the heat load place on the cryogenic stage and the complexity of the readout electronic. To overcome this challenge, various multiplexing schemes have been proposed and demonstrated. In a row-column biasing architecture [2-4], correlations between detections on row and column readout channels are used to determine the detecting pixel of an array. SFQ circuits have been demonstrated for both direct [5] and row-column [4] readout of SNSPDs. A third approach treats the active nanowire as a delay line and uses differential readout to determine the location of absorption along the length of the nanowire [6,7]. Nanowires have been biased in series with different resistive shunts in a pulse amplitude multiplexing scheme [8], but this approach is not scalable beyond a few tens of channels. Frequency multiplexing has been demonstrated using both DC [9] and AC [10] nanowire biasing, but reaching the kilopixel scale as in the microwave kinetic inductance detector (MKID) community [11-13] remains challenging. During this fellowship, I developed a new multiplexing architecture which was proposed by the JPL SNSPD team. This architecture, known as the thermal row-column (TRC), uses the heat generated during the photon detection process to couple two photosensitive layers of an array. After the initial formation of a normal domain in an SNSPD during photon detection, the kinetic inductance of the nanowire combined with Joule heating leads to dissipation of thermal energy. The TRC utilizes this thermal energy to couple row and column channels of an array, with the overlapping area of a row and a column channel forming a single pixel. With N nanowires patterned as rows on one layer and N nanowires patterned as columns on a second with each channel having its own dedicated readout, an N x N array can be operated with only 2N readout channels. Figure 2 shows a representation of the TRC layout for a 4 x 4 device with the inset showing a schematic of the detection process. For an initial demonstration of the TRC architecture, I fabricated and tested two 4 x 4 arrays with different nanowire designs. These devices showed saturated internal efficiency at 1550 nm and proper correlations between counts on the two layers to enable imaging. These results were summarized and published in Nano Letters [14]. Following this initial demonstration, I focused on scaling to larger format arrays with 1024 effective pixels. This only requires 64 readout channels, making it compatible with the array readout system at JPL. This array was integrated with an optical cavity to enhance absorption, leading to 67% detection efficiency and 59 of 64 channels showing correct operation and saturated internal efficiency. Figure 3 shows the photoresponse count rate for the prototype kilopixel array along with a laser spot imaged with the detector. While the yield must be improved to enable true kilopixel resolution, this prototype demonstrated the feasibility of high efficiency kilopixel scale TRC arrays. When combined with other multiplexing technologies such as time domain or row-column biasing, the TRC has the capability of enabling megapixel scale SNSPD arrays for low photon-flux applications. Despite the remarkable demonstrated performance of SNSPDs, the community still lacks a quantitatively accurate model for several aspects of the detector physics. Figure 4 shows the current understanding of the qualitative process of single-photon detection in an SNSPD, which involves the physics of high energy excitations, evolution of the superconducting state, and the coupled electrical and thermal (electrothermal) behavior of the nanowire and readout circuitry. While the present description is capable of describing the main qualitative features of the detector behavior, there are still discrepancies when comparing models to experimental measurements. In particular, the detection mechanism is not fully understood, and models of the evolution of the superconducting state following the absorption of a photon only provide a qualitative picture of the detection process. In addition, the electrothermal evolution of the combined superconductor and electric circuit is not well described by existing models for low Tc superconductors, such as the technologically important WSi material system. The work performed during this fellowship examined both of these problems and made strides toward a more complete description of the physics of SNSPDs. Early in the fellowship, performed a number of experiments to better understand the electrothermal evolution of the WSi material system with a particular interest in the reset dynamics of WSi nanowires. The reset time of SNSPDs is typically limited by the kinetic inductance of the nanowire. By shortening a device, a faster reset time can be achieved. However, if the reset time is too fast, the detector fails to return to the superconducting state after a detection event, in a process known as latching [15]. Latching poses an ultimate limit on the maximum count rate which can be achieved by an SNSPD and is therefore of interest when optimizing device design. The experiments and modeling based on the two-temperature model [16] suggested that the latching behavior in WSi nanowires is limited by the heat dissipation in the thin dielectric layers used to enhance optical absorption in high efficiency devices. Inspired by this observation, I also performed experimental measurements to better understand the lateral heat transport in these thin dielectrics using a new method known as nanowire thermometry [17]. While these measurements helped to quantify the heat transport properties of WSi devices, the quasi-equilibrium two-temperature model is insufficient to describe all of the experimental observations, indicating that more advanced non-equilibrium formulations of the energy transfer processes are needed to quantitatively describe SNSPD electrothermal behavior. In a collaboration with Dr. Alexander Kozorezov of Lancaster University, I also studied the downconversion and suppression of superconductivity through extensive numerical modeling. New experimental measurements performed as part of a collaboration between MIT, JPL, and NIST Boulder demonstrated record low timing jitter and enabled a new measurement technique where the delay time between the detection process of photons of different energies can be measured [18]. This new technique, known as the relative latency, provides a window into the timescale of the photon detection process, and for the first time, enables comparisons of experiment and theory in the timing of the detection process. Inspired by these experimental results, we formulated a theory of intrinsic timing jitter based on energy fluctuations during the downconversion process [19] and the characteristic response time of the detector [20]. The qualitative connection between these fluctuations and timing jitter is shown in Figure 5. If the response time of the detector is a deterministic function of the energy deposited in the superconductor, the fluctuations in the energy retained by the superconductor due to phonon escape during downconversion directly translate to fluctuations in the response time of the detector. This concept is capable of describing the main qualitative features of experimental measurements [18], but a microscopic model is required to determine the characteristics of this characteristic latency curve. To formulate the microscopic model of the detection process, we used the most advanced existing model in the literature [21]. This model uses time-dependent Ginzburg-Landau (TDGL) formulation to describe the superconducting state, coupled to a two temperature model to describe the energy evolution of the system and a current conservation equation. These five equations are solved simultaneously to describe the evolution of the system in 2D. The main feature of this formulation is the formation and motion of vortices in the superconductor as the primary means of suppressing superconductivity. This is shown qualitatively in Figure 6, which shows the simulated evolution of the magnitude of the order parameter for a photon absorbed near the edge of the nanowire. The 2D TDGL model provides a picture of the detection process which is able to qualitatively describe he main features of SNSPD detection. It accurately captures the shapes of the internal efficiency curves as a function of bias current and also reproduces the correct polarization dependent response of the detectors. However, the model is unable to reproduce the intrinsic timing jitter or relative latency. Instead, the standard TDGL model underestimates the timing jitter and relative latency, suggesting that the existing model predicts suppression of superconductivity which is faster than observed experimentally. To address this shortcoming, we modify the model to use the generalized TDGL model [22,23] which relaxes some of the assumptions of the standard TDGL model which are not satisfied during the single photon detection process. Using this more advanced model, we are able to simultaneously fit the polarization-dependent internal efficiency and the order of magnitude of the relative latency, as shown in Figures 7(a) and 7(b), respectively, with reasonable estimates of the material properties and values of the fitting parameters. Overall, the work performed during this fellowship has provided new experimental and theoretical insight into the operation of SNSPDs. Modeling and optimization of the DSOC ground detector has helped prepare the JPL team for the upcoming technology demonstration mission. The investigation into the heat transport in thin dielectrics directly enabled the design and fabrication of the thermal row-column multiplexing architecture, which is expected to be of considerable technological interest for low count rate single-phonon imaging applications. Finally, the experimental and modeling work to better understand the detection process in SNSPDs has improved the understanding of the technology to help inform the design of nanowires which can achieve high efficiency at longer wavelengths and in wider nanowire structure. While considerable work is still needed to provide a fully quantitative and predictive model of SNSPD performance, the strides made during this fellowship have significantly advance the accuracy of state of the art SNSPD modeling. ","infoText":"Closed out","infoTextExtra":"","dateText":"May 2020"}],"responsibleMd":{"acronym":"STMD","canUserEdit":false,"city":"","external":false,"linkCount":0,"organizationId":4875,"organizationName":"Space Technology Mission Directorate","organizationType":"NASA_Mission_Directorate","naorganization":false,"organizationTypePretty":"NASA Mission Directorate"},"program":{"acronym":"STRG","active":true,"description":"
\tThe Space Technology Research Grants Program will accelerate the development of "push" technologies to support the future space science and exploration needs of NASA, other government agencies and the commercial space sector. Innovative efforts with high risk and high payoff will be encouraged. The program is composed of two competitively awarded components.
","programId":69,"responsibleMd":{"acronym":"STMD","canUserEdit":false,"city":"","external":false,"linkCount":0,"organizationId":4875,"organizationName":"Space Technology Mission Directorate","organizationType":"NASA_Mission_Directorate","naorganization":false,"organizationTypePretty":"NASA Mission Directorate"},"responsibleMdId":4875,"stockImageFileId":36658,"title":"Space Technology Research Grants"},"leadOrganization":{"acronym":"CalTech","canUserEdit":false,"city":"Pasadena","country":{"abbreviation":"US","countryId":236,"name":"United States"},"countryId":236,"external":true,"linkCount":0,"organizationId":3115,"organizationName":"California Institute of Technology","organizationType":"Academia","stateTerritory":{"abbreviation":"CA","country":{"abbreviation":"US","countryId":236,"name":"United States"},"countryId":236,"name":"California","stateTerritoryId":59},"stateTerritoryId":59,"murepUnitId":110404,"naorganization":false,"organizationTypePretty":"Academia"},"supportingOrganizations":[{"acronym":"JPL","canUserEdit":false,"city":"Pasadena","country":{"abbreviation":"US","countryId":236,"name":"United States"},"countryId":236,"external":true,"linkCount":0,"organizationId":4946,"organizationName":"Jet Propulsion Laboratory","organizationType":"FFRDC_2fUARC","stateTerritory":{"abbreviation":"CA","country":{"abbreviation":"US","countryId":236,"name":"United States"},"countryId":236,"name":"California","stateTerritoryId":59},"stateTerritoryId":59,"naorganization":false,"organizationTypePretty":"FFRDC/UARC"}],"statesWithWork":[{"abbreviation":"CA","country":{"abbreviation":"US","countryId":236,"name":"United States"},"countryId":236,"name":"California","stateTerritoryId":59}],"lastUpdated":"2024-2-6","releaseStatusString":"Released","viewCount":670,"endDateString":"May 2020","startDateString":"Sep 2016"}}