Ultra-broadband Mid to Long-wave Infrared Spintronic Poisson Bolometer

TL;DR

The paper tackles the need for true ultra-broadband infrared detection across MWIR–LWIR in a single, uncooled device. It introduces the Spintronic Poisson bolometer, which uses Poisson-distributed discrete events generated by thermally driven spin transitions, transduced by a spintronic layer and enhanced by nanoplasmonic absorbers to achieve 3–14 μm sensitivity at room temperature. With an NETD of 80–100 mK and operation at 0 Oe bias, the device demonstrates performance competitive with cooled InSb detectors and superior to VOx uncooled imagers, while offering CMOS‑level compatibility and scalable potential for arrays. This work establishes a new detection paradigm that leverages thermal fluctuations as informative signals and points toward on-chip, ultra-broadband infrared imaging for autonomous systems, environmental monitoring, and HADAR applications.

Abstract

Infrared detectors have traditionally been divided into two fundamental classes, mid-wave (MWIR, 3-5 um) and long-wave (LWIR, 8-14 um). Integrating MWIR and LWIR within a single device is challenging due to distinct materials, cooling needs, and detection mechanisms, while such integration is critical for improved object recognition, temperature estimation, and environmental sensing. In this work, we demonstrate a Spintronic Poisson (SP) bolometer enabling room-temperature ultra-broadband sensing across 3-14 um. Unlike conventional bolometers that rely on continuous analog signals, the SP bolometer implements a Poisson-counting detection paradigm, encoding temperature in discrete stochastic events, which turns thermal noise from a limitation into the basis of the estimator itself. We fabricate the SP bolometer using a spintronic transduction layer integrated with a plasmonic nanoantenna array to enhance broadband infrared absorption. Using spintronic transduction, the device achieves the noise-equivalent temperature difference (NETD, thermal sensitivity metric) of 80-100 mK at 300 K, surpassing uncooled detectors and approaching cooled technologies. This work establishes a statistical detection paradigm for room-temperature infrared sensing with broad application potential.

Figures (5)

• Figure 1: Ultra-broadband SP-bolometer performance and comparison with state-of-the-art infrared detectors. (a) Ultra-broadband sensed spectrum of the SP-bolometer, where light green is at a scene temperature of 300K, and dark green is at a scene temperature of 600K. (b) Temperature estimation standard deviation (square root of Cramer-Rao bound (CRB)) for the SWIR, MWIR, LWIR, and the SWIR-MWIR-LWIR combined cases. Ultra-broadband combined achieves the lowest CRB (i.e., the lowest standard deviation), indicating superior accuracy compared to single-band configurations. (c) Performance comparison of infrared detectors showing noise-equivalent temperature difference (NETD) versus wavelength for different technologies. Lower NETD means higher thermal sensitivity. Colored bars indicate operating temperatures (see legend), with labels showing detector name, SWaP, and Sensor temperature. SWaP: size, weight, and power. The Spintronic Poisson bolometer (SP, this work) achieves room-temperature operation across 0.8-14 $\mu m$ with 80-100 mK NETD. CNT: Carbon nanotube liu2018high, BP/$MoS_2$: Black Phosphorus on top of Molybdenum Disulfide shu2024high, a-Si: Amorphous Silicon rogalski2016challengeslynredAthena1920, CQD: colloidal quantum dots tang2019dual, VOx: Vanadium Oxide lynredPico1024Gen2, SLS: Strained Layer Superlattice lee2023designirc906sls, MCT: Mercury Cadmium Telluride leonardoCondorHD, InSb: Indium Antimonide flirA6750sc. *MWIR and LWIR are experimentally characterized in this work, while SWIR performance is estimated.
• Figure 2: Schematic of SP-bolometer operation and spectral response. (a) Incident light is absorbed in the transduction layer, creating a thermal hotspot that diffuses through the bolometer. A synthetic antiferromagnet (SAF) adjacent to the sensing layer stabilizes the magnetic orientation. (b) Heat absorption increases the transition probability in the sensing layer, producing higher count rates. (c) The readout is obtained via resistance changes in the sensing layer. (d) SEM image of the SP-bolometer device with a plasmonic nanoantenna array atop a transduction layer (Ge/Ti). The inset on the right shows a top view of the nanoantenna array atop the SP-bolometer nanopillar, overlaid with a COMSOL-simulated electric field distribution. The left inset shows a side view of the SP-bolometer. (e) Spectral absorptance (purple), lens transmittance (cyan), blackbody radiation (red), and sensed spectrum (green) at a scene temperature of 23 °C. The nanoantenna array provides strong MWIR absorption and significant LWIR absorption, supporting broadband sensing despite the blackbody distribution.
• Figure 3: Broadband NETD measurement setup and response of the SP-bolometer. (a) Schematic of the broadband NETD measurement setup for the SP-bolometer, including ultra-stable blackbody source and broadband ZnSe lens (f/0.5). (b) Broadband NETD response of the SP-bolometer without optical filtering, measured at varying blackbody temperatures, demonstrating a NETD of approximately 81mK under 0Oe magnetic field and 575µA bias.
• Figure 4: Consistent MWIR–LWIR NETD performance of the SP-bolometer compared with state-of-the-art infrared cameras. (a) Experimental setup for ultra-broadband MWIR–LWIR NETD measurements using the Spintronic Poisson (SP) bolometer with a high-speed filter wheel. (b) Spectral transmittance of the optical filters used in the NETD measurements. (c) Spectral comparison of NETD for a representative state-of-the-art uncooled analog $VO_x$ microbolometer (analog, LWIR, 300K, FLIR A325sc) in blue color, InSb photodiode (analog, MWIR, 77K, Telops Spark M150) in orange color, and broadband SP-bolometer (purple). The InSb photodiode performs better with lower NETD across the $3 \mu \text{m}$ to $13.5 \mu \text{m}$ range. However, the SP-bolometer's NETD is comparable to the InSb photodiode and notably outperforms the $VO_x$ microbolometer across all bands. (d) Zoomed-in view of (c), highlighting differences in NETD between the SP bolometer and InSb photodiode. Error bars for the $VO_x$ microbolometer increase at wavelengths farther from $10.6 \mu \text{m}$, likely due to the LWIR lens’s optical bandwidth limitations. $T_s$: Sensor Temperature. The NETD values are normalized to an $f$-number of $f/1.0$, where $\text{NETD}_{f/1} = \text{NETD}/(f/\#)^2$. The $f$-numbers are $f/\# = 0.5$ for the SP-bolometer, $f/\# = 0.6$ for the $VO_x$ microbolometer, and $f/\# = 2.3$ for the InSb photodiode.
• Figure 5: Consistent MWIR-LWIR sensor response of the SP-bolometer compared with state-of-the-art cameras. (a) SP-bolometer response (count rate) to the MWIR band ($3-5~\mu\text{m}$). (b) $VO_x$ microbolometer (analog, LWIR, 300K, FLIR A325sc) response (count rate) to the MWIR band ($3-5~\mu\text{m}$). (c) InSb photodiode (analog, MWIR, 77K, Telops Spark M150) response (count rate) to the MWIR band ($3-5~\mu\text{m}$). (d) MWIR filter transmittance ($3-5~\mu\text{m}$). (e) SP-bolometer response (count rate) to the LWIR band ($10-11.2~\mu\text{m}$). (f) $VO_x$ microbolometer (analog, LWIR, 300K, FLIR A325sc) response (count rate) to the LWIR band ($10-11.2~\mu\text{m}$). (g) InSb photodiode (analog, MWIR, 77K, Telops Spark M150) response (count rate) to the LWIR band ($10-11.2~\mu\text{m}$). (h) LWIR filter transmittance ($10-11.2~\mu\text{m}$). $T_s$: sensor temperature. The SP-bolometer exhibits a consistent response across MWIR and LWIR bands, performing comparably to the cooled InSb photodiode in MWIR and outperforming the $VO_x$ microbolometer across LWIR. Error bars represent the standard deviation from five measurements with 33 ms integration time.