Articles tagged with "semiconductor-materials"
One-way heat diode could help batteries run cooler and last longer
University of Houston engineers have developed a novel thermal rectification technique that enables heat to flow in only one direction, akin to an electrical diode but for heat. This breakthrough allows unprecedented control over device temperatures by permitting forward heat transfer while completely blocking reverse flow. Achieved through semiconductor materials subjected to a magnetic field, this innovation precisely controls radiative heat at the microscopic level, addressing the longstanding challenge of managing heat that traditionally spreads in all directions and causes overheating in electronics. The technology promises to extend battery life in cell phones, electric vehicles, satellites, and improve efficiency in AI data centers by maintaining optimal operating temperatures even under extreme conditions. Beyond unidirectional heat flow, the research team is also developing a thermal circulator device that moves radiative heat in a continuous loop among multiple surfaces, further enhancing thermal management capabilities. Additionally, a companion study demonstrates that asymmetric thermal conductivity can enable conduction heat rectification, bridging theoretical concepts with practical applications for microchips and high-performance batteries. This technology holds particular promise for space
energythermal-managementbattery-technologysemiconductor-materialsheat-dioderadiative-heat-flowelectronics-coolingAI chip startup Ricursive hits $4B valuation two months after launch
Ricursive Intelligence, an AI chip startup founded by former Google researchers Anna Goldie (CEO) and Azalia Mirhoseini (CTO), has rapidly achieved a $4 billion valuation just two months after its formal launch. The company raised $300 million in a Series A round led by Lightspeed, bringing its total funding to $335 million. Ricursive is developing an AI system capable of designing and autonomously improving AI chips, including creating its own silicon substrate layer to accelerate chip advancements. The founders’ prior work on reinforcement learning for chip layout design has been instrumental in four generations of Google’s TPU chips. Ricursive is part of a broader trend of startups focused on AI systems that self-improve hardware. Notably, it should not be confused with Recursive, another AI startup working on similar self-improving AI systems and reportedly also targeting a $4 billion valuation. Additionally, Naveen Rao’s Unconventional AI recently raised a $475 million seed round at a $4.5
AI-chipssemiconductor-materialschip-design-automationsilicon-substratereinforcement-learningAI-hardwarestartup-fundingChemical method converts toxic arsenic sludge into chip material
Researchers in Denmark have developed an innovative chemical process that transforms toxic arsenic waste—commonly found in sludge from groundwater treatment and mining operations—into a valuable metallic material suitable for use in semiconductors, batteries, and clean energy technologies. This breakthrough addresses the persistent environmental and disposal challenges posed by arsenic residues, which remain hazardous even after removal from water or ore. The process converts arsenic into a glassy metal form, which exhibits unique electrical and mechanical properties ideal for advanced industrial applications. Detailed atomic-level analysis conducted at the Canadian Light Source confirmed that the upcycled arsenic meets the technical standards required for electronics and energy systems. Led by scientists at the Geological Survey of Denmark and Greenland, the research marks a paradigm shift in arsenic waste management by turning a long-standing environmental liability into a strategic industrial resource. Given arsenic’s growing importance as a critical mineral in the transition to clean energy, this method could reduce pollution while supplying essential materials. The team now aims to scale the process beyond laboratory
materialsclean-energyarsenic-recyclingsemiconductor-materialsbattery-technologyenvironmental-sustainabilitychemical-processingTin-based perovskites could be used to make stable, eco-friendly solar cells
Researchers from the Helmholtz-Zentrum Berlin (HZB) and the University of Potsdam have investigated ion densities in four common perovskite compounds used for solar cells, discovering that tin-based perovskites produced with an alternative solvent exhibit significantly lower ion density—only one tenth that of lead-based perovskites. This lower ion density correlates with enhanced stability, as mobile halide ions are a primary cause of degradation in perovskite solar cells. The team found that tin perovskites degrade five times slower than lead-based ones, with one tin perovskite variant showing excellent operational stability for over 600 hours. The tin perovskites were synthesized using different solvents, including dimethyl sulfoxide (DMSO) and a DMF-DMI solvent mixture, the latter helping to avoid tin oxidation and reduce ion migration. Lead-based perovskites exhibited the highest ion density, while tin-lead mixtures and tin-only perovskites
energyperovskite-solar-cellstin-based-perovskitesphotovoltaic-technologysemiconductor-materialssolar-energymaterial-stabilityScientists uncover what's truly limiting silicon solar cell efficiency
Korean researchers from the Korea Institute of Energy Research (KIER) and Chungbuk National University have identified two distinct microscopic defects that limit the efficiency of silicon heterojunction (SHJ) solar cells, currently the most efficient silicon-based solar technology. SHJ cells combine crystalline silicon with thin amorphous silicon layers and are key components in next-generation tandem solar architectures aimed at surpassing conventional silicon cell performance. The team improved upon the traditional Deep Level Transient Spectroscopy (DLTS) method by developing a new analysis approach that tracks the full transient response of defects, revealing that what was previously thought to be a single defect signature is actually a superposition of two independent defect types with different energy levels and behaviors. The study found one defect to be a slow, deep-level component and the other a fast, shallow-level component, each with distinct spatial locations and atomic bonding configurations within the device. Importantly, these defects can switch bonding configurations depending on manufacturing conditions and device operation, with hydrogen playing a
energysolar-cellssilicon-heterojunctionphotovoltaicsdefect-analysisrenewable-energysemiconductor-materialsHow the world’s most efficient silicon solar cell reached a record 27.81% efficiency
Chinese solar manufacturer Longi has detailed the technical innovations behind its world-record silicon solar cell efficiency of 27.81%, certified by Germany’s ISFH and published in Nature. The cell employs a hybrid interdigitated back-contact (HIBC) design combining n-type and p-type contacts exclusively on the rear, eliminating front-side metal shading and enhancing light absorption. Key advancements include the use of passivated tunneling contacts, dielectric passivation layers, and edge passivation on high-resistivity half-cut M10 wafers to reduce carrier recombination and losses. A novel high-low temperature fabrication process enables simultaneous diffusion and deposition while passivating wafer edges in situ (iPET technology). Additional improvements involve a multilayer front surface stack of aluminum oxide and silicon nitride to optimize optical properties and suppress recombination, alongside an amorphous silicon layer whose thickness is carefully controlled and crystallized by pulsed green nanosecond laser to balance passivation and conductivity. The doping of the n-type poly
energysolar-cellsilicon-photovoltaicssolar-energymaterials-sciencesemiconductor-materialsrenewable-energyPhoton teleportation achieved between two independent quantum dots
Researchers at the University of Stuttgart, in collaboration with partners from the Leibniz Institute for Solid State and Materials Research in Dresden and Saarland University, have successfully demonstrated quantum teleportation between photons emitted by two different quantum dots. This breakthrough addresses a critical challenge in building quantum repeaters, which are essential for extending secure quantum communication over long distances via fiber networks. By using nearly identical quantum dots to generate single photons and entangled photon pairs, and employing quantum frequency converters to align their frequencies, the team achieved a transfer of polarization states with a success rate slightly above 70 percent. This achievement marks the first time quantum information has been teleported between photons from independent quantum dots, a feat previously hindered by the difficulty of producing indistinguishable photons from separate sources. The experiment involved sending one photon through a 10-meter optical fiber to interfere with another, enabling the teleportation process. The work is part of the Quantenrepeater.Net project, a large German research consortium aiming to develop quantum repeat
quantum-dotsquantum-teleportationquantum-communicationquantum-networksquantum-repeaterssemiconductor-materialsphotonicsWorld’s smallest neural implant tracks brain signals through light
Cornell researchers have developed the world’s smallest neural implant, called the microscale optoelectronic tetherless electrode (MOTE), which is about the size of a grain of salt (approximately 300 microns long and 70 microns wide). This implant is capable of wirelessly recording brain activity in living animals for over a year by using harmless red and infrared laser beams to power the device and transmit data through tiny pulses of infrared light. The implant’s semiconductor diode, made from aluminum gallium arsenide, captures light energy to power the circuit and sends encoded brain signals optically, employing pulse position modulation to minimize power consumption while maintaining effective data communication. The MOTE was tested in mice by implanting it in the barrel cortex, where it successfully recorded neuron spikes and synaptic activity continuously for a year without causing adverse effects or immune responses. Its extremely small size reduces brain tissue irritation and avoids the complications associated with traditional electrodes and optical fibers, which often provoke immune reactions due to tissue movement
IoTneural-implantoptoelectronicssemiconductor-materialswireless-brain-monitoringbio-integrated-sensinglow-power-communicationScientists discover boron arsenide beats diamond in heat transfer
Researchers at the University of Houston have discovered that boron arsenide (BAs), a synthetic crystal, surpasses diamond in thermal conductivity, achieving values above 2,100 W/mK at room temperature. This finding challenges the long-held belief that diamond is the best isotropic heat conductor and suggests that existing theoretical models need revision, as previous calculations—factoring in four-phonon scattering—had capped BAs’s conductivity at 1,360 W/mK. The breakthrough was made possible by producing ultra-pure BAs crystals through refined synthesis techniques, overcoming limitations caused by impurities in earlier samples. Beyond its record-breaking heat conduction, boron arsenide also exhibits promising semiconductor properties, including a wider band gap, higher carrier mobility, and compatibility with chip integration due to its thermal expansion coefficient. These combined attributes make BAs a strong candidate to outperform silicon in electronics, offering potential improvements in thermal management for devices ranging from smartphones to data centers and high-performance computing systems. Supported by a National
materialsboron-arsenidethermal-conductivitysemiconductor-materialsheat-transferelectronics-coolingadvanced-materialsWall Street analysts explain how AMD’s own stock will pay for OpenAI’s billions in chip purchases
AMD and OpenAI have announced an expanded partnership in which OpenAI will assist AMD in refining its Instinct GPUs—AMD’s competitor to Nvidia chips—and commit to purchasing 6 gigawatts of compute capacity over several years. The deal is valued in the billions, but rather than paying with cash, OpenAI will use AMD stock to finance its purchases. AMD has granted OpenAI up to 160 million stock warrants, which vest as certain milestones are met, including significant increases in AMD’s stock price. For example, the final tranche requires AMD’s market cap to reach around $1 trillion, implying a potential value of about $100 billion for OpenAI’s stake if all conditions are met and shares are held without selling. UBS analyst Timothy Arcuri suggests that OpenAI will likely sell portions of its AMD stock over time to cover its GPU purchases, effectively making this a financing arrangement for AMD. Despite the unconventional structure, the deal serves as a strong validation of AMD’s AI GPU capabilities,
energyAI-chipsAMDOpenAIGPUssemiconductor-materialscompute-capacityNew molecular coating method improves quantum photon purity by 87%
Researchers at Northwestern University have developed a novel molecular coating technique that significantly enhances the purity and reliability of single-photon sources critical for quantum technologies. By applying a layer of PTCDA molecules onto tungsten diselenide, an atomically thin semiconductor known for its single-photon emission at atomic defects, the team achieved an 87% improvement in photon spectral purity. This coating protects the fragile quantum emitters from atmospheric contaminants like oxygen, which previously caused variability and noise in photon production, without altering the semiconductor’s intrinsic electronic properties. The PTCDA coating not only stabilizes the photon emission but also uniformly shifts the photon energy to lower levels, beneficial for quantum communication devices. This uniformity and improved control over photon emission are essential for developing scalable, tunable, and stable single-photon sources, which are foundational for quantum computing, sensing, and secure quantum communication networks. The researchers plan to extend this approach to other semiconductor materials and explore electrically driven photon emission, aiming to advance toward interconnected quantum networks and
quantum-materialsmolecular-coatingtungsten-diselenidesingle-photon-emittersquantum-communicationsemiconductor-materialsquantum-technologyMicrosoft in-chip cooling breakthrough cuts GPU heat rise by 65%
Microsoft has developed a breakthrough in-chip microfluidic cooling technology that channels liquid coolant directly inside GPU chips to remove heat more efficiently. This approach carves microscopic grooves into the silicon, enabling coolant to flow in direct contact with hot spots, which reduces the maximum GPU temperature rise by up to 65% and outperforms traditional cold plate cooling systems by as much as three times. The system also leverages AI to identify heat patterns and direct cooling precisely where needed. Microsoft successfully demonstrated this technology by cooling a server running simulated Teams meetings and is now prioritizing reliability testing. The design was inspired by natural vein structures, with Microsoft collaborating with Swiss startup Corintis to create bio-inspired coolant channels that improve heat dissipation compared to straight channels. The engineering challenge involved balancing channel depth for effective coolant flow without compromising silicon strength, developing leak-proof chip packaging, and integrating etching processes into chip manufacturing. Beyond individual chips, Microsoft envisions microfluidics playing a major role in datacenters by enabling more
energycooling-technologymicrofluidicsGPU-coolingAI-hardwaresemiconductor-materialsthermal-managementFlorida team builds chip to run AI tasks 100-fold at lower power cost
Researchers at the University of Florida have developed a novel silicon photonic chip that uses light, rather than solely electricity, to perform convolution operations—key computations in AI tasks such as image and pattern recognition. By integrating optical components like laser light and microscopic Fresnel lenses directly onto the chip, the device can execute these operations much faster and with significantly lower energy consumption. Tests demonstrated that the prototype achieved about 98% accuracy in classifying handwritten digits, comparable to conventional electronic chips, while operating at near-zero energy for these computations. A notable innovation of this chip is its ability to process multiple data streams simultaneously through wavelength multiplexing, using lasers of different colors passing through the lenses concurrently. This parallel processing capability enhances efficiency and throughput. The project, involving collaboration with UCLA and George Washington University, aligns with trends in the industry where companies like NVIDIA are already incorporating optical components into AI hardware. The researchers anticipate that chip-based optical computing will become integral to future AI systems, potentially enabling more sustainable scaling of AI technologies
energyAI-chipoptical-computingsilicon-photonicsenergy-efficiencymachine-learningsemiconductor-materialsChina achieves 99.99997% helium purity using natural gas feedstock
China has developed a groundbreaking helium extraction device capable of producing ultra-pure helium at 99.99997% purity (6N9 grade) from natural gas fields with very low helium content. Created by Yan’an-based Vacree Technologies after six years of research led by scientist Rong Chengxu, the system employs a multi-step process combining catalytic dehydrogenation, membrane separation, pressure swing adsorption, and ultra-low-temperature refining to remove impurities such as hydrogen, nitrogen, methane, and neon. This innovation allows continuous production of up to 400,000 cubic meters of helium annually and represents a significant technological leap, as achieving such purity levels has been a major challenge worldwide, especially given China’s natural gas helium concentrations of only 0.03 to 0.05 percent compared to 1 to 7 percent in international fields. The advancement addresses both scientific and strategic needs, securing China’s supply of helium critical for applications in particle accelerators, MRI machines, semiconductor manufacturing, spaceflight, and
energyhelium-extractionnatural-gascryogenic-technologyvacuum-technologysemiconductor-materialsstrategic-resourcesJapan's firm develops new plasma method for perovskite manufacturing
Japan’s Sumitomo Heavy Industries (SHI) has developed a novel plasma-based method called Reactive Plasma Deposition (RPD) to manufacture the electron transport layer (ETL) in perovskite solar cells. This layer is crucial as it facilitates electron flow from the perovskite to the electrode. Unlike traditional ETL fabrication methods that involve high temperatures, harsh particles, or toxic gases—often damaging the fragile perovskite material—SHI’s RPD deposits ultra-thin tin oxide (SnO₂) films at low temperatures using non-hazardous gases. This approach is reportedly 200 times faster and costs only about 0.5% of current ETL manufacturing expenses, making it safer, more environmentally friendly, and highly suitable for mass production. While the technology represents a significant breakthrough by enabling low-cost, scalable, and damage-minimized ETL production, SHI acknowledges challenges remain, such as high material costs and handling flammable or toxic precursor gases,
energyperovskite-solar-cellsplasma-depositiontin-oxidephysical-vapor-depositionsolar-energysemiconductor-materialsHot-cold design supercharges solar thermoelectric efficiency by 15x
Researchers at the University of Rochester have developed a novel "hot-cold" design that increases the efficiency of solar thermoelectric generators (STEGs) by 15 times, potentially bridging the gap between STEGs and conventional photovoltaic solar panels. Unlike traditional solar panels that rely on photovoltaic cells, STEGs generate electricity by exploiting the temperature difference between a hot side and a cold side through the Seebeck effect. Historically, STEGs have been limited by low efficiency, converting less than 1% of sunlight into electricity compared to about 20% for residential solar panels. The new approach improves efficiency not by altering semiconductor materials but by enhancing thermal management on both the hot and cold sides of the device. The innovations include using femtosecond laser pulses to etch nanoscale structures into tungsten on the hot side, creating a "black metal" surface that selectively absorbs solar wavelengths while minimizing heat loss. This surface is covered with a plastic layer that traps heat by limiting convection and conduction, effectively raising
energysolar-energythermoelectric-generatorsrenewable-energythermal-managementsemiconductor-materialsnanostructuresInstead of selling to Meta, AI chip startup FuriosaAI signed a huge customer
South Korean AI chip startup FuriosaAI recently announced a partnership to supply its AI chip, RNGD, to enterprises using LG AI Research’s EXAONE platform, a next-generation hybrid AI model optimized for large language models (LLMs). This collaboration targets multiple sectors including electronics, finance, telecommunications, and biotechnology. The deal follows FuriosaAI’s decision to reject Meta’s $800 million acquisition offer three months prior, citing disagreements over post-acquisition strategy and organizational structure rather than price. FuriosaAI’s CEO June Paik emphasized the company’s commitment to remaining independent and advancing sustainable AI computing. The partnership with LG AI Research is significant as it represents a rare endorsement of a competitor to Nvidia by a major enterprise. FuriosaAI’s RNGD chip demonstrated 2.25 times better inference performance and greater energy efficiency compared to competitive GPUs when running LG’s EXAONE models. Unlike general-purpose GPUs, FuriosaAI’s hardware is specifically designed for AI computing, lowering total cost of ownership while
AI-chipsFuriosaAILG-AI-Researchenergy-efficiencyAI-computingsemiconductor-materialsAI-hardwareSolar cells on ultra-thin glass to transform energy technology for space
Researchers from Loughborough and Swansea universities are developing lightweight cadmium telluride (CdTe) solar cells deposited on ultra-thin glass to revolutionize energy systems for satellites and space manufacturing. This CdTe-on-glass technology offers a lighter, cheaper, and highly radiation-resistant alternative to the conventional silicon and multi-junction solar cells currently used in space missions. While multi-junction cells dominate due to their high efficiency, their complex manufacturing and high costs limit scalability. The new technology targets 20% efficiency in space and has already achieved 23.1% efficiency on Earth, with initial space testing conducted aboard the AlSat-Nano CubeSat. The collaboration aligns with the UK’s strategic vision to capture a significant share of the growing global space technology market, valued at £17.5 billion in the UK alone. The European Space Agency forecasts a surge in space solar demand from 1 MWp/year to 10 GWp/year by 2035, driven by satellite constellations and
energysolar-cellsspace-technologycadmium-telluridephotovoltaicssemiconductor-materialssatellite-power-systemsMIT’s AI-powered robot speeds up search for better solar materials
MIT researchers have developed an AI-powered autonomous robotic system that dramatically accelerates the measurement of photoconductivity—a key electrical property influencing the performance of semiconductor materials used in solar cells and electronics. The robot uses a probe to make contact-based measurements, guided by machine learning models imbued with domain knowledge from chemists and materials scientists. This enables it to identify optimal contact points on perovskite samples, a class of semiconductors relevant to photovoltaics, and efficiently plan the probe’s path to maximize data collection speed and accuracy. In a 24-hour test, the robot completed over 3,000 photoconductivity measurements, outperforming existing AI models in both precision and throughput by taking 125 unique measurements per hour. This rapid, autonomous approach allows scientists to quickly characterize new materials, potentially leading to the discovery of more efficient solar panel components. The research team, led by Professor Tonio Buonassisi, envisions creating fully autonomous laboratories that can accelerate materials discovery by combining fast
robotAIsolar-energysemiconductor-materialsphotoconductivityautonomous-systemsmaterials-scienceUS scientists shatter 165-year-old physics rule for energy breakthroughs
A research team at Penn State has reported a groundbreaking violation of Kirchhoff’s law of thermal radiation, a fundamental physics principle established 165 years ago. Kirchhoff’s law states that a material’s ability to absorb energy at a specific wavelength and angle must equal its ability to emit energy under the same conditions. However, the Penn State scientists demonstrated a significant non-reciprocity contrast of 0.43—indicating a notable difference between absorptivity and emissivity—using a specially designed five-layer thin film semiconductor structure. This effect was observed over a broad 10-micrometer wavelength range, marking the strongest recorded deviation from the law’s reciprocity principle. This breakthrough has important implications for energy harvesting and thermal management technologies. For example, conventional solar cells must emit some absorbed energy back to the sun, which limits their efficiency. The new non-reciprocal material can direct emitted energy away from the sun, potentially allowing that energy to be captured by additional solar cells and thus improving overall
energythermal-radiationKirchhoff's-lawnon-reciprocitythin-film-materialsenergy-harvestingsemiconductor-materialsUS, India, Japan, Australia unite to cut China’s mineral dominance
The United States, India, Japan, and Australia, collectively known as the Quad, have launched the Quad Critical Minerals Initiative aimed at reducing global dependence on China for critical minerals essential to modern technologies like electric vehicles, batteries, and semiconductors. The initiative seeks to secure and diversify supply chains amid concerns over China’s dominant position in key minerals, particularly graphite, which is vital for electric vehicle batteries. The announcement coincides with a Quad foreign ministers’ meeting in Washington hosted by U.S. Secretary of State Marco Rubio, signaling a strategic pivot toward Asia by the U.S. administration and emphasizing the need for supply chain diversification to counter economic coercion and supply disruptions. While the Quad’s joint statement did not explicitly name China, it expressed serious concerns about provocative actions in the South and East China Seas that threaten regional peace and stability, underscoring the group’s commitment to a “free and open Indo-Pacific.” Indian Foreign Minister Subrahmanyam Jaishankar stressed the importance of regional autonomy and freedom
critical-mineralssupply-chain-diversificationelectric-vehicle-batteriesQuad-alliancematerials-securityenergy-materialssemiconductor-materials11 startups from YC Demo Day that investors are talking about
At Y Combinator’s Spring 2025 Demo Day, the majority of startups showcased were focused on AI, either developing AI agents or tools to support AI development. A notable trend was the emergence of startups creating specialized AI assistants modeled as “Cursor for X,” targeting specific professional domains such as knowledge workers (Den) and lawyers (Vessence). Besides AI, robotics also featured as a resurging area of interest. Among the startups attracting significant investor attention were Anvil, which offers SEO optimization tailored for AI-driven content discovery platforms; Atum Works, innovating 3D-stacked chips to overcome transistor scaling limits with potential to rival NVIDIA; and Auctor, which automates enterprise software implementation and has already drawn interest from major vendors like SAP and AWS. Other highlighted startups include Cactus, providing an AI copilot to help solopreneurs manage calls and payments; Den, considered one of the hottest companies for its AI agents designed to replace tools like Slack and Notion for enterprise
robotAI-startupsrobotics-revival3D-chipssemiconductor-materialsenterprise-automationAI-agents