---
title: Superconducting Microwire Single-Photon Detector
url: https://www.emergentmind.com/topics/superconducting-microwire-single-photon-detector-smspd
type: topic
---

# Superconducting Microwire Single-Photon Detector

to=arxiv_search.search  彩神争霸提现ి ಲ್ಲಿjson
{"query":"superconducting microwire single photon detector SMSPD MgB2 single-photon detection 20 K 2308.15228", "max_results": 5}
to=arxiv_search.search 公众号天天中彩票json
{"query":"Superconducting microstrip single-photon detector with system detection efficiency over 90% at 1550 nm 2101.05407", "max_results": 5}
A superconducting microwire single-photon detector (SMSPD) is a current-biased superconducting single-photon detector in which the photon-sensitive element is a micrometer-scale strip or meander rather than the $\sim 100$ nm wire typical of a conventional superconducting nanowire single-photon detector (SNSPD). In the cited literature, both “superconducting microwire single-photon detector” and “superconducting microstrip single-photon detector” are used, with the same acronym, SMSPD. The defining proposition of the platform is that single-photon sensitivity can persist in wires with widths from submicrometer to several micrometers, enabling larger active areas, reduced lithographic stringency, higher switching currents, and, in some material systems, elevated operating temperatures. Reported implementations span WSi, MoSi, NbN, and MgB$_2$, with demonstrated operating modes that include saturated internal detection efficiency at telecom wavelengths, broadband polarization-insensitive coupling, millimeter-scale active areas, and operation up to $20$ K [2002.12858] [2101.05407] [2202.05942] [2308.15228].

## 1. Historical emergence and nomenclature

The modern SMSPD literature emerged from the observation that micrometer-scale superconducting wires can remain single-photon sensitive in the near-infrared. In silicon-rich WSi, wire widths from $1~\mu$m to $3~\mu$m achieved saturated internal detection efficiency at $1.55~\mu$m, and a meandered detector with $2.0~\mu$m wire width was demonstrated over a surface area of $362 \times 362~\mu\mathrm{m}^2$ [2002.12858]. In MoSi, meander-shaped detectors with widths of $1~\mu$m and $3~\mu$m and active areas up to $400 \times 400~\mu\mathrm{m}^2$ also showed saturated internal detection efficiency at $1550$ nm for $3$ nm films, while $5$ nm films with the same geometry were insensitive to single near-infrared photons [2002.09088].

Subsequent work established that SMSPDs are not restricted to proof-of-principle operation. NbN devices reached a saturated system detection efficiency of $\sim 92.2\%$ at $1550$ nm with a dark count rate of $\sim 200$ cps and a minimum timing jitter of $\sim 48$ ps at $0.84$ K [2101.05407]. Silicon-rich WSi devices with a high-fill-factor “candelabra meander” demonstrated simultaneous low-polarization sensitivity and high detection efficiency in a fiber-coupled configuration [2202.05942]. Large-area and array formats then followed, including $1~\mathrm{mm}^2$ NbN devices fabricated by ultraviolet photolithography and $8$-pixel $1~\mathrm{mm}^2$ WSi/MoSi arrays with single-photon sensitivity at $1064$ nm [2304.06921] [2303.10739].

A central point of terminology is that the field uses both “microwire” and “microstrip” for substantially similar architectures. The distinction in practice is often geometric rather than conceptual: the photosensitive element is a lithographically wide superconducting strip, current biased close to its switching or depairing limit, and patterned into a straight bridge, spiral, meander, or related topology.

## 2. Device architectures, materials, and fabrication strategies

SMSPD architectures are unusually diverse because the wider strip relaxes some nanofabrication constraints while making geometry, optical absorption, and current distribution more explicitly co-optimized. In NbN, one representative telecom detector used a $1.0~\mu$m-wide, $7$ nm-thick microstrip in a $50~\mu$m-diameter double spiral with filling factor $f \simeq 0.8$, integrated above $13$ pairs of SiO$_2$/Ta$_2$O$_5$ quarter-wave layers on Si; simulated peak absorptance was $A_p \simeq 97$–$100\%$ for $w=1~\mu$m and $f=0.8$ [2101.05407]. In WSi, a $0.51~\mu$m-wide, $2.2$ nm-thick device replaced the traditional parallel-strip active region with a “candelabra meander,” moving optimized $90^\circ$ and $180^\circ$ bends outside the active area and reaching a fill factor of $\simeq 0.91$ [2202.05942]. In MgB$_2$, meander-shaped wires with widths $W=1$–$5~\mu$m, lengths $L=1.2$–$10.2$ mm, filling factor $\approx 0.28$, and active area up to $400 \times 400~\mu\mathrm{m}^2$ were fabricated from thin films grown by Hybrid Physical–Chemical Vapor Deposition on $(0001)$ 6H–SiC [2308.15228].

Large-area scalability is a recurring design objective. Standard UV i-line photolithography was used to fabricate NbN SMSPDs on $4$-inch wafers, including a device with an active area of $1~\mathrm{mm} \times 1~\mathrm{mm}$ [2304.06921]. WSi and MoSi arrays extended the single-pixel concept to $8$-channel $1~\mathrm{mm}^2$ formats, with each pixel formed by a $1.0~\mu$m-wide meander stripe and individually read out [2303.10739]. A free-space coupled NbN device for time-of-flight imaging adopted a circular candelabra-style meander of diameter $\approx 260~\mu$m, with straight sections of nominal width $1.98$–$2.00~\mu$m and gap $0.5~\mu$m [2402.02311].

He$^+$ irradiation appears repeatedly as a materials-engineering tool. NbN work used $20$ keV He$^+$ irradiation with fluence $\simeq 5 \times 10^{16}$ ions/cm$^2$ to engineer defect density and improve performance at $2.1$ K [2101.05407]. MgB$_2$ devices used a $30$ keV He$^+$ beam at a dose of $5 \times 10^{15}$ ions/cm$^2$, with post-irradiation changes in $T_c$, $I_{sw}$, and $R_s<5\%$ [2308.15228]. This suggests that controlled disorder is being used not merely to compensate fabrication nonuniformity, but to tune the detection threshold itself.

| Platform | Representative geometry | Representative result |
|---|---|---|
| NbN [2101.05407] | $1.0~\mu$m wire, $50~\mu$m diameter double spiral | SDE $\sim 92.2\%$ at $1550$ nm |
| WSi [2202.05942] | $0.51~\mu$m candelabra meander, $f \simeq 0.91$ | $96.5$–$96.9\% \pm 0.5\%$ at $1550$ nm |
| MgB$_2$ [2308.15228] | $1$–$5~\mu$m meanders, area up to $400 \times 400~\mu\mathrm{m}^2$ | single-photon detection up to $20$ K |
| NbN [2304.06921] | UV-photolithographic $1~\mathrm{mm} \times 1~\mathrm{mm}$ meander | near-saturated IDE up to $800$ nm |

## 3. Detection physics and circuit models

The physical picture used across the literature remains the hotspot or resistive-domain framework, with several refinements. In MgB$_2$, a photon is absorbed, locally breaking Cooper pairs and creating a normal-resistive hotspot; self-heating through Joule power $I_b^2 R_{\mathrm{hot}}$ competes with electron cooling to sustain the normal domain [2308.15228]. In NbN, a simplified diffusion-hotspot criterion is written as
$$
E_{\min}=\frac{N_0 \Delta^2 w d}{\zeta \,[1-I_b/I_{\mathrm{dep}}]},
$$
with $\zeta \simeq 0.25$ the quasiparticle multiplication efficiency [2101.05407]. In MoSi, a kinetic-equation approach relates saturation to whether the hotspot radius $R_{NS}$ becomes large enough that $W \le 2R_{NS}(I_b)$ [2002.09088].

Several papers use compact empirical forms for bias-dependent efficiency. MgB$_2$ adopts
$$
\eta(I_b)=1-\exp[-\alpha\cdot (I_b/I_{sw}-1)],
$$
while WSi array modeling uses
$$
\mathrm{IDE}(I)\approx 1-\exp[-\alpha\,(I-I_{\mathrm{th}})]
\quad (I>I_{\mathrm{th}}).
$$
These expressions are not universal microscopic laws; they are fitting forms for the turn-on and saturation regime [2308.15228] [2303.10739].

Electrical behavior is controlled by kinetic inductance and current redistribution. For MgB$_2$,
$$
L_k=\mu_0 \lambda_L^2/A,\qquad A=W\cdot d,
$$
and the recovery time is
$$
\tau = L_k/R_{\mathrm{load}},
$$
with measured $\tau \approx 1.3$ ns for the $30\%$-peak decay time [2308.15228]. In NbN, the same scaling underlies the lower inductance of micrometer-wide strips relative to nanowires; for $\ell \approx \pi \cdot 50~\mu$m, $w=1~\mu$m, $d=7$ nm, and $\lambda_L \approx 450$ nm, $L_k$ is on the order of $10$–$20$ nH [2101.05407]. In large-area formats the total wire length dominates. A $1~\mathrm{mm}\times 1~\mathrm{mm}$ NbN meander had total wire length $L_{\mathrm{tot}} \approx 1$ m, and the resulting kinetic inductance limited the maximum count rate [2304.06921].

The external shunt resistor is a specific circuit-level issue for SMSPDs because wide strips can have high current-carrying capacity and low intrinsic kinetic inductance. Shunting suppresses latching, enlarges the usable bias range, and can improve internal detection efficiency, but reducing $R_s$ also lowers pulse amplitude and increases pulse decay time, thereby degrading timing jitter and count-rate performance [2304.08762]. That trade-off is one of the most practical distinctions between SMSPD optimization and conventional SNSPD optimization.

## 4. Performance envelope

The reported performance landscape is broad because different SMSPDs target different operating corners rather than a single optimum. At the high-efficiency telecom extreme, NbN achieved a saturated system detection efficiency of $\sim 92.2\%$ at a dark count rate of $\sim 200$ cps, with polarization sensitivity of $\sim 1.03$ and minimum timing jitter of $\sim 48$ ps at $1550$ nm [2101.05407]. High-fill-factor WSi pushed fiber-coupled performance further, reporting $\eta=(96.5$–$96.9)\%\pm 0.5\%$ at $1550$ nm and $1\times 10^5$ counts/s, together with polarization sensitivity $S=1.02\pm 0.008$ and $\eta>91.8\%$ over a $40$ nm bandwidth centered at $1550$ nm [2202.05942].

At the large-area and high-rate extreme, MgB$_2$ devices demonstrated single-photon sensitivity at $1.55~\mu$m up to $20$ K, saturation of internal detection efficiency for $1~\mu$m-wide devices at $3.7$ K, timing jitter of $\sim 50$ ps, active area up to $400 \times 400~\mu\mathrm{m}^2$, reset time as low as $\sim 1$ ns, and linearity of detection rate versus incident power up to at least $\sim 100$ Mcps [2308.15228]. This combination is notable because large area is usually associated with long reset time.

Millimeter-scale devices emphasize a different balance. A $1~\mathrm{mm}\times 1~\mathrm{mm}$ NbN SMSPD fabricated by ultraviolet photolithography showed near-saturated internal detection efficiency at wavelengths up to $800$ nm at $0.85$ K, while at $1550$ nm it exhibited a system detection efficiency of $\sim 5\%$ ($7\%$) and timing jitter of $102$ ($144$) ps under $\sim 18~\mu$m ($600~\mu$m) illumination [2304.06921]. An $8$-pixel $1~\mathrm{mm}^2$ array in WSi reached saturated internal detection efficiency in $6/8$ pixels at $1064$ nm, with measured system detection efficiency $\approx 4.2\%$, plateau dark count rate $\approx 0.1$ cps, instrument-response-function FWHM $=325$ ps, and reset time $\tau_R \simeq 350$ ns from $L_K \approx 17.5~\mu$H and $R_L \approx 50~\Omega$ [2303.10739].

A free-space coupled large-active-area NbN SMSPD showed the other side of the architecture spectrum: a $260~\mu$m-diameter active area, free-space-coupled SDE $\approx 6\%$, SMF-coupled SDE saturating at $\approx 16\%$, DCR $\approx 4.8$ kcps with cryolens at $I_b=139~\mu$A, and system jitter $171$ ps at $1550$ nm [2402.02311]. The paper explicitly contrasts this with the simpler alignment afforded by the larger area.

## 5. Design trade-offs and comparison with conventional SNSPDs

The main SMSPD trade-off is between optical fill factor, kinetic inductance, current crowding, and the bias margin needed for saturation. Increasing fill factor improves absorptance, but it generally increases total wire length and therefore $L_k$, slowing reset. This is stated directly for periodic meanders as $F=w/p$ in NbN large-area devices and as $f=w/(w+s)$ in WSi high-fill-factor devices [2304.06921] [2202.05942]. Current crowding at bends is correspondingly more important as fill factor rises. One route is geometric: the candelabra meander removes sharp turns from the active area [2202.05942]. Another route is materials-topographic: locally thickening the bends by depositing a secondary superconducting film increased $I_{sw}$ and improved detection efficiency, intrinsic dark count rate, and timing jitter in both SNSPDs and SMSPDs [2112.08009].

Relative to conventional SNSPDs, SMSPDs can offer larger active area, higher switching current, lower lithographic precision requirements, and potentially lower kinetic inductance per unit active area, but these advantages are not automatic. The NbN telecom device explicitly compared a $1.0~\mu$m SMSPD with $\sim 100$ nm SNSPDs and reported $66$–$80~\mu$A switching current, $\sim 10$–$20$ nH kinetic inductance, and yield $\sim 68\%$ versus $<40\%$ for NbN SNSPDs with $I_{sw}>60~\mu$A [2101.05407]. MgB$_2$ broadens the comparison further by shifting the operating temperature from the $2$–$4$ K range typical for NbN, WSi, and MoSi to $20$ K while retaining jitter $\sim 50$ ps and reset time $\sim 1$ ns [2308.15228].

A common misconception is that microwire detectors are inherently low-efficiency or only visible-wavelength devices. The literature does not support that generalization: NbN and WSi SMSPDs report $>90\%$ system detection efficiency at $1550$ nm [2101.05407] [2202.05942]. A second misconception is that wide superconducting strips are simply easier SNSPDs. The literature instead shows a different design regime, with distinct concerns about Pearl length, current uniformity across micrometer widths, shunt-resistor optimization, and hotspot extension across a much larger transverse dimension [2308.15228] [2304.08762].

## 6. Applications, arrays, and open questions

The application space follows directly from the combination of large area, relaxed alignment, and competitive timing. The cited work names quantum communications at $1550$ nm, lightweight cryogenics for space-based links, high-throughput LIDAR, biomedical imaging, Boson sampling, dark-matter searches, and free-space time-of-flight imaging [2308.15228] [2101.05407] [2402.02311]. The time-of-flight demonstration used a single-pixel galvanometer scanning system and reconstructed depth and intensity, with depth resolution $\Delta z \approx 2.6$ cm derived from $\Delta t_{\mathrm{sys}}=171$ ps [2402.02311].

Array scaling is already established at the millimeter scale. An $8$-pixel $1~\mathrm{mm}^2$ NIR-sensitive array was described as the first report of an SMSPD array [2303.10739]. Later WSi arrays were characterized for charged-particle detection at Fermilab and CERN, with fill-factor-normalized efficiencies of about $50\%$ for $120$ GeV protons, electrons, and pions in a $2\times 2~\mathrm{mm}^2$ device and about $75\%\pm 3\%$ for $120$ GeV hadrons and muons in a $1\times 1~\mathrm{mm}^2$ device, together with time resolution of $1.15$ ns and about $130$ ps, respectively [2410.00251] [2510.11725]. These are not single-photon metrics, but they indicate that the microwire architecture is now being evaluated as a broader superconducting sensing platform.

Open questions remain fundamental as well as practical. MgB$_2$ work explicitly raises the role of two-band superconductivity, the impact of He$^+$-induced disorder on the local superconducting gap and hotspot dynamics, and the general mechanism of single-photon detection in high-critical-temperature superconductors [2308.15228]. Shunted NbN devices showed a nonlinear relation between detection current and photon energy and weak internal-detection-efficiency saturation up to $2000$ nm after helium-ion irradiation [2304.08762]. In low-background WSi arrays, excess correlated dark counts across pixels were observed at rates above accidental expectations; a plausible source is cosmic muons [2506.04533].

Taken together, these results place SMSPDs in a distinct regime of superconducting photodetection: micrometer-scale superconducting strips that can preserve single-photon sensitivity while enlarging active area, relaxing alignment and fabrication constraints, and, depending on the material system, supporting either ultrahigh telecom efficiency, broadband polarization insensitivity, millimeter-scale coverage, or elevated-temperature operation.

Source: https://www.emergentmind.com/topics/superconducting-microwire-single-photon-detector-smspd