Researchers at Kyoto Institute of Technology (KIT) have achieved a major breakthrough in infrared sensing technology with the development of a cooling-free, high-sensitivity infrared sensor based on semiconducting single-walled carbon nanotubes (SWCNTs). Announced on March 31, 2026, this innovation promises to transform non-destructive testing, security screening, and medical diagnostics by delivering performance comparable to cooled systems without the bulk and cost of cryogenic equipment.

The sensor leverages the unique photothermoelectric (PTE) effect in precisely engineered SWCNT films, achieving an 11-fold increase in sensitivity over previous carbon nanotube-based designs. This advancement stems from a collaborative effort involving KIT's Materials Science & Engineering faculty, Chuo University, and Japan's National Institute of Advanced Industrial Science and Technology (AIST), highlighting the strength of interdisciplinary university research in Japan.

Challenges in Traditional Infrared Detection

Infrared (IR) sensors detect heat radiation in wavelengths longer than visible light, typically from 700 nanometers to 1 millimeter. They are essential for night vision, thermal imaging, and detecting hidden objects since IR penetrates clothing, plastics, and packaging materials. However, high-sensitivity IR detection has historically required cooled detectors, such as those using mercury cadmium telluride (MCT) or indium antimonide (InSb), which operate at cryogenic temperatures (around 77 Kelvin) to reduce thermal noise.

Cooled systems excel in sensitivity but suffer from high power consumption, large size, and expense—limiting them to military or laboratory use. Uncooled alternatives, like microbolometers based on vanadium oxide, trade off sensitivity for portability but still fall short for demanding applications. The global uncooled IR imaging market, valued at approximately $5.14 billion in 2026, is projected to grow at 8.21% CAGR to $7.63 billion by 2031, driven by demand for compact, affordable solutions in consumer electronics, automotive, and industry.

In Japan, where manufacturing precision is paramount, there's acute need for advanced uncooled sensors to enhance quality control without halting production lines.

The Power of Semiconducting Carbon Nanotubes

Carbon nanotubes (CNTs) are cylindrical carbon structures with diameters of 1-2 nanometers and lengths up to millimeters. Single-walled CNTs (SWCNTs) exhibit metallic or semiconducting behavior depending on chirality—the atomic arrangement defining their 'handedness'. Semiconducting SWCNTs, comprising about two-thirds of as-produced tubes, offer broadband IR absorption due to their tunable bandgap (0.4-1.5 eV).

Prior CNT IR sensors used unsorted mixtures, diluting performance with metallic tubes that shunt photocurrents. KIT's team sorted high-purity (>98%) semiconducting SWCNTs (diameter 1.35-1.45 nm) using ethyl cellulose in tetrahydrofuran, followed by centrifugation and filtration into ~500 nm thin films. Chemical doping created p-type (holes dominant, using AgTFSI) and n-type (electrons dominant, KOH complex) regions, forming an in-plane p-n junction. Parylene C passivation (~500 nm) ensured doping stability against oxidation.

How the Sensor Works: Step-by-Step

The device's operation hinges on the photothermoelectric (PTE) effect, combining plasmonic absorption and thermoelectric conversion:

• Step 1: IR Absorption - Incoming IR photons excite plasmon resonance—collective electron oscillations—in semiconducting SWCNTs, achieving >95% absorptance at 2.52 THz (118 μm wavelength).
• Step 2: Localized Heating - Plasmon decay converts photon energy to heat, creating a temperature gradient (ΔT) across the p-n junction since doped regions have suppressed thermal conductivity.
• Step 3: Carrier Diffusion - Hot carriers diffuse from high-T to low-T regions; Seebeck effect generates voltage (Seebeck coefficient S = 270 μV/K, 3.6x higher than unsorted CNTs).
• Step 4: Signal Readout - Photovoltage (>2 mV at low power, responsivity ~100 mV/W) is measured via electrodes, enabling imaging in transmission mode.

This self-absorbing architecture outperforms separate absorber-thermoelectric designs by minimizing thermal resistance.

Photo by Marcus Loke on Unsplash

Performance Benchmarks and Validation

Testing showed responsivity of 100 mV/W at 218 GHz and 2.52 THz—11 times that of unsorted CNT devices (11.1 mV/W). Effective Seebeck enhancement and doping-optimized power factor (α²σ/κ) drove gains. Noise equivalent power (NEP) is low due to high signal-to-noise from broadband response.

Demonstrations included transmission imaging of metal foil behind opaque sheets using 0.26 THz waves, proving non-destructive package inspection. Devices on flexible PET substrates operated uncooled at room temperature, with stable response over time thanks to passivation.

For context, commercial uncooled microbolometers achieve ~10-50 mK NETD (noise equivalent temperature difference); this CNT sensor approaches cooled performance while being scalable and low-cost.

Key Players: KIT's Nonoguchi Lab and Collaborators

Leading the effort is Associate Professor Yoshiyuki Nonoguchi at KIT's Faculty of Materials Science & Engineering. His lab specializes in organic thermoelectrics and CNT dispersion, with over 2,200 citations on Google Scholar. Nonoguchi noted, "We achieved highly sensitive IR detection without cooling, enabling portable devices with high voltage output." Co-authors include Prof. Takeshi Yamao and Asst. Prof. Yuhi Inada (KIT), Prof. Yukio Kawano and Asst. Prof. Kou Li (Chuo University), and Daichi Suzuki (AIST).

Chuo's Kawano lab excels in broadband THz/IR sensors; AIST provided nano-evaluation expertise. KIT, founded 1949, emphasizes applied materials science, ranking among Japan's top for nanotechnology patents.

Transformative Applications

This sensor excels in non-destructive testing (NDT), imaging internals of sealed packages—critical for Japan's electronics and pharma sectors. Security applications include airport screening; medical uses span fever detection to bio-imaging. In 6G communications, it could detect THz signals for ultra-fast data.

Japan's manufacturing, producing 50% of global semiconductors, benefits immensely. Integration into wearables or drones enables portable NDT, reducing downtime in auto and aerospace.

Read the full study for technical depth: Synergistic Plasmonic‐Thermoelectric Enhancement in Semiconducting CNTs.

Japan's Higher Education in Nano-Innovation

Japan invests heavily in nanotechnology, with R&D spending >3% GDP. Universities like KIT contribute via MEXT/JSPS grants; 2026 nano tech expo showcases such advances. KIT-Chuo-AIST ties exemplify 'industry-academia-government' model, fostering tech transfer. Japan's nano research output remains top-tier, with U Tokyo and Kyoto U leading AI-ranked lists.

This sensor underscores universities' role in addressing societal challenges like aging infrastructure inspection.

Photo by Perry Merrity II on Unsplash

Future Outlook and Challenges

Next steps: Flexible substrates for wearables, array scaling for imaging arrays, commercialization via AIST ventures. Challenges include yield of sorted CNTs and integration with CMOS readout.

Nonoguchi's team eyes social implementation through partnerships. As uncooled IR demand surges, this positions Japanese higher ed as global leaders.

Explore KIT's press release: KIT Announcement. JST summary: JST Press.