Tag Archives: letter-mems-tech

The nanoelectronics research centers imec and Holst Centre (set up by imec and TNO), presented a low-power wide-area (LPWA) multi-standard radio chip today at imec’s annual technology forum in Brussels (ITF Brussels 2016). The new radio chip is a best-in-class product, which can operate with a lower level of power consumption than any other radio chip technology released to date for long range connectivity in sensor networks. The sub-GHz radio chip’s technology can serve a multitude of protocols including IEEE 802.15.4g/k, W-MBUS, KNX-RF, as well as the popular LoRa and SIGFOX networks, and future cellular IoT for applications such as smart metering, smart home, smart city and critical infrastructure monitoring.

The radio chip operates in industrial, scientific, medical (ISM) and short-range devices (SRD) bands, covering a frequency range from 780MHz to 930MHz. The robust, low-power design combines a large link budget, with state-of-the-art interference rejection and lowest bill of materials by minimizing external components as compared to of-the-shelf available chips. The radio is implemented as a complete System-on-Chip (SoC) including the RF front end, power management, an ARM processor,160kBytes of SRAM and peripherals like SPI, I2C and UART. It features a targeted sensitivity of -120dBm at 0.1% BER (1kbps) and ultra-low power consumption of 8mW (Rx) and 113mW (Tx) for 13.5dBm output power. The receiver supports a wide gain range to handle input signals from -120dBm to -15dBm, corresponding to a large dynamic range of 105dB. The PA features automatic ramp-up and ramp-down for ARIB spectral mask compliancy. Furthermore, the output power is controllable from <-40dBm up to 15dBm.

“With the foreseen release of the NB-IoT protocol in June 2016 by the 3GPP, it is clear that protocols such as NB-IoT, SigFox and LoRA are here to stay for the coming years,” stated Kathleen Philips, program director perceptive systems at imec/Holst Centre. “Our novel sub-GHz radio chip can serve multiple of these protocols and is an ideal solution for long-range wireless connectivity  for IoT applications.”

Imec’s Industrial Affiliation program on the Intuitive Internet-of-Things (IoT) focuses on developing the building blocks for the future. The program explores an intuitive IoT, with sensor systems that can detect and assist with the needs and wants of people in an unobtrusive way, and can take into account their varied perspectives and surrounding environment. Along with low-power radio chips, imec also develops ultra-small, low-cost, intelligent, and ultra–low power sensors and heterogeneous sensor networks. Interested companies are invited to partner with imec on its varied research initiatives. Companies can also connect with imec to request access to imec’s technological advances to further develop their projects through licensing programs with imec.

TowerJazz, the global specialty foundry, today announced volume production of a new RF technology capable of integrating a wireless front-end module (FEM) on a single chip, tailored to meet the challenges of Internet of Things (IoT) applications. Analysts estimate that the number of IoT connected devices will grow at a 15-20% growth rate annually, reaching up to 30 billion units by 2020. McKinsey Global Institute recently estimated that IoT could generate up to $11 trillion in global value by 2025.

The TowerJazz process enables integration of power amplifiers, switches, and low noise amplifiers as well as CMOS digital and power control on a single die. TowerJazz is delivering this product today for smartphones, tablets and wearables, and this technology also meets the more universal requirements of IoT applications by providing cost, power, performance, and form factor benefits vs. competing solutions.

As an example, TowerJazz has partnered with Skyworks Solutions, Inc., an innovator of high performance analog semiconductors connecting people, places and things, to deliver a first of its kind integrated wireless FEM using this technology.

“We are pleased that our long partnership with TowerJazz on SiGe BiCMOS for PA based products is now in volume production for key customers of Skyworks Solutions,” said Bill Vaillancourt, GM/VP Skyworks Connectivity Solutions.

TowerJazz’s new RF technology includes a 0.18um SiGe PA device with best in class silicon-based performance, a low Ron-Coff switch device, a SiGe low noise amplifier device, 5V CMOS for power control, 0.18um CMOS for integrating MIPI or other digital content as well as thick Cu metal layers for low-loss inductors and matching components. By offering all active components typically required for a wireless FEM, this technology enables a new family of products that can integrate multiple communication standards (WiFi, Bluetooth, 802.15.4 or NFC) that form the backbone of the IoT fabric today onto the same chip.

“This new technology complements our existing suite of SiGe PA and RF SOI switch technology offerings and provides customers new architectural options by enabling the combination of these elements on a single die while offering best in class silicon-based PA performance,” said Marco Racanelli, Sr. VP and GM of RF/High Performance Analog and US Aerospace & Defense Business Groups, and Newport Beach Site Manager, TowerJazz.

TowerJazz will exhibit and demonstrate its advanced process technologies for specialty IC manufacturing in booth #1532 at IMS2016, the premier conference in the RF and microwave industry. Please visit the company website for more information on TowerJazz’s RF and high performance analog technology offerings.

STMicroelectronics announced advanced high-efficiency power semiconductors for Hybrid and Electric Vehicles (EVs) with a timetable for qualification to the automotive quality standard AEC-Q101.

In EVs and hybrids, where better electrical efficiency means greater mileage, ST’s latest silicon-carbide (SiC) technology enables auto makers to create vehicles that travel further, recharge faster, and fit better into owners’ lives. A leader in silicon carbide, ST is among the first to present new-generation rectifiers and MOSFETs for high-voltage power modules and discrete solutions addressing all the vehicle’s main electrical blocks. These include the traction inverter, on-board battery charger, and auxiliary DC-DC converter.

Today’s power modules typically rely on standard silicon diodes and Insulated Gate Bipolar Transistors (IGBTs). Silicon carbide is a newer, wide-bandgap technology that allows smaller device geometries capable of operating well above the 400V range of today’s electric and hybrid drivetrains. The smaller SiC diode and transistor structures present lower internal resistance and respond more quickly than standard silicon devices, which minimize energy losses and allow associated components to be smaller, saving even more size and weight.

“Major carmakers and automotive Tier-1s are now committing to silicon-carbide technology for future product development to leverage its higher aggregate efficiency compared to standard silicon in a wide range of operating scenarios,” said Mario Aleo, Group Vice President and General Manager, Power Transistor Division, STMicroelectronics. “Our SiC devices have demonstrated superior performance and reached an advanced stage of qualification as we support customers preparing to launch new products in the 2017 timeframe.”

ST has been among the first companies to produce silicon-carbide high-voltage MOSFETs, with its first 1200V SiC MOSFET introduced back in 2014, achieving industry-leading 200degreesC rating for more efficient and simplified designs.

The Company is using the industry’s most advanced processes to fabricate SiC MOSFETs and diodes on 4-inch wafers. In order to drive down the manufacturing costs, improve the quality, and deliver the large volumes demanded by the auto industry, ST is scaling up its production of SiC MOSFETs and diodes to 6-inch wafers, and is on schedule to complete both conversions by the end of 2016.

ST has already qualified its 650V SiC diodes to AEC-Q101, and will complete qualification of the latest 650V SiC MOSFETs and 1200V SiC diodes in early 2017. The qualification of the new-generation 1200V SiC MOSFETs will be completed by the end of 2017.

The STPSC20065WY 650V SiC diode is in full production now in DO-247. The range also includes lower current ratings and smaller form-factor TO-220 package options. The STPSC10H12D 1200V SiC diode is sampling now to lead customers in the TO-220AC package and goes to production this month, with volume production of the automotive-grade version planned for Q4 2016. Multiple current ratings from 6A to 20A and packaging options will also be available.

The SCTW100N65G2AG 650V SiC MOSFET is sampling now to lead customers in the HiP247 package. It will ramp up in volumes in H1 2017. To enable more compact designs, a 650V SiC MOSFET in the surface-mount H2PAK will also be qualified to AEC-Q101 in H1 2017.

Silicon Integration Initiative (Si2), an Austin-based integrated circuit research and development joint venture, has launched a project to help designers reduce power consumption, a growing challenge for most system-on-chip designs. The project will develop new power modeling technology to estimate power consumption more easily and more accurately throughout the design process, especially during the earliest stages.

The end result will be a new power modeling standard to reduce resources and costs needed to develop virtually every type of SoC. Jerry Frenkil, director of OpenStandards, said that the Si2 Low Power Working Group, part of the newly restructured Si2 OpenStandards program, will lead this industry-wide effort.

“Every SoC design team is grappling with the continued need to reduce power consumption,” Frenkil said. “That’s especially true for mobile devices, but it’s also a concern throughout the electronics industry. One way to accomplish this is through improved multi-level power modeling techniques that better predict SoC power and performance. Right now there’s no commonly accepted way to develop an accurate estimation of power consumption early in the design phase. This often leads to designs being power inefficient, performance constrained, or both.”

Frenkil said the standard will also “enable more efficient and reliable power analyses and optimizations since the same model will be used from system-level design through gate level implementation and all phases in between.”

The approved specification will be contributed to the IEEE P2416 Standards Working Group for industry-wide distribution. Nagu Dhanwada, senior R&D engineer at IBM, chairs both the IEEE P2416 and Si2 Low Power Modeling Working Groups. “Since Si2 is an R&D joint venture, its members can work together to develop specifications, tests and proof-of-concepts with anti-trust protection. This specification will greatly accelerate standardization efforts within P2416, and testing prior to IEEE standardization will enable us to rapidly prove out the use of the new standard before it hits the street,” Dhanwada explained.

IEEE P2416 is an essential component of a coordinated IEEE effort focusing on system-level design. The IEEE 1801 standard currently expresses design intent. It’s latest update, IEEE 1801-2015, includes support for power modeling.

John Biggs, chair of the IEEE 1801 Working Group said, “Efforts of the Si2 Low Power Working Group will help the IEEE P2416 Working Group standardize the representation of power consumption data. The fruits of this work, in combination with the new power modeling capability in IEEE 1801-2015, should greatly ease the challenging task of energy aware system level design.”

The new Si2 model specification is expected to be completed in October. For more information about this project, contact Jerry Frenkil at [email protected]. For information about the Low Power Working Group and other OpenStandards programs, visit http://www.si2.org/openstandards/.

Founded in 1988, Si2 is a research and development joint venture that provides standard interoperability solutions for integrated circuit design tools. All Si2 activities are carried out under the auspices of the The National Cooperative Research and Production Act of 1993, the fundamental law that defines R&D joint ventures and offers them a large measure of protection against federal antitrust laws. Si2’s international membership includes semiconductor foundries, fabless manufacturers, and EDA companies.

Nanoelectronics research center imec has announced that Dr. Gordon E. Moore, creator of the famous Moore’s law theory and co-founder of Intel, is the recipient of its lifetime of innovation award. Imec’s annual award recognizes Dr. Moore’s visionary view, unrivalled innovation, and his profound impact on the global electronics industry.

In 1965, Dr. Moore predicted that the number of components on an integrated circuit (IC) would double every year for the coming 10 years, thereby making ICs and computer processing simultaneously faster, cheaper, and more powerful. In 1975, Dr. Moore revised the forecast rate to approximately every two years. Moore’s law turned out to be incredibly accurate, growing beyond its predictive character to become an industry driver that holds true today, 50 years later. Keeping up with Moore law’s progression has required a tremendous amount of engineering and commitment from the global semiconductor industry. While its meaning has evolved over generations, it has had a profound impact in many areas of technological change and progress.

“It is truly an honor to present imec’s lifetime innovation award to Dr. Moore, on behalf of all our global partners and our researchers,” stated Luc Van den hove, president and CEO of imec. “Dr. Moore’s name is synonymous with progress, and his vision has inspired and given direction to the entire semiconductor industry, which has revolutionized the way we compute, communicate, and interact. As the industry upholds this prediction and brings forth new innovations in chip technology, the future of Moore’s law will impact such things as healthcare, a sustainable climate, and safer transport all for the better.”

Dr. Moore began his career at Johns Hopkins University. He cofounded Fairchild Semiconductor in 1957 and launched Intel in 1968 together with Robert Noyce and Andy Grove. Today, Intel is a world leader in the design and manufacturing of integrated circuits and is the largest semiconductor company. Dr. Moore served as Intel CEO from 1975-1987, and then became its chairman of the board until his retirement in 1997.

“Although Moore’s law was created more than 50 years ago, it remains extremely valid and serves as a guide to what we innovate at imec,” continued Van den hove. “Throughout our organizations’ 32-year existence, we’ve worked at enabling Moore’s law and helping our partners innovate and develop the modern technology that society has embraced and demands. Dr. Moore’s legacy continues to be our mission and we are privileged to honor him.” 

Imec’s Lifetime of Innovation award is awarded to Dr. Moore on May 24, 2016 at its annual ITF Brussels, the flagship of imec’s worldwide ITF events.

One secret to creating the world’s fastest silicon-based flexible transistors: a very, very tiny knife.

Working in collaboration with colleagues around the country, University of Wisconsin-Madison engineers have pioneered a unique method that could allow manufacturers to easily and cheaply fabricate high-performance transistors with wireless capabilities on huge rolls of flexible plastic.

The researchers — led by Zhenqiang (Jack) Ma, the Lynn H. Matthias Professor in Engineering and Vilas Distinguished Achievement Professor in electrical and computer engineering, and research scientist Jung-Hun Seo — fabricated a transistor that operates at a record 38 gigahertz, though their simulations show it could be capable of operating at a mind-boggling 110 gigahertz. In computing, that translates to lightning-fast processor speeds.

It’s also very useful in wireless applications. The transistor can transmit data or transfer power wirelessly, a capability that could unlock advances in a whole host of applications ranging from wearable electronics to sensors.

The team published details of its advance April 20 in the journal Scientific Reports.

The researchers’ nanoscale fabrication method upends conventional lithographic approaches — which use light and chemicals to pattern flexible transistors — overcoming such limitations as light diffraction, imprecision that leads to short circuits of different contacts, and the need to fabricate the circuitry in multiple passes.

Using low-temperature processes, Ma, Seo and their colleagues patterned the circuitry on their flexible transistor — single-crystalline silicon ultimately placed on a polyethylene terephthalate (more commonly known as PET) substrate — drawing on a simple, low-cost process called nanoimprint lithography.

In a method called selective doping, researchers introduce impurities into materials in precise locations to enhance their properties — in this case, electrical conductivity. But sometimes the dopant merges into areas of the material it shouldn’t, causing what is known as the short channel effect. However, the UW-Madison researchers took an unconventional approach: They blanketed their single crystalline silicon with a dopant, rather than selectively doping it.

Then, they added a light-sensitive material, or photoresist layer, and used a technique called electron-beam lithography — which uses a focused beam of electrons to create shapes as narrow as 10 nanometers wide — on the photoresist to create a reusable mold of the nanoscale patterns they desired. They applied the mold to an ultrathin, very flexible silicon membrane to create a photoresist pattern. Then they finished with a dry-etching process — essentially, a nanoscale knife — that cut precise, nanometer-scale trenches in the silicon following the patterns in the mold, and added wide gates, which function as switches, atop the trenches.

With a unique, three-dimensional current-flow pattern, the high performance transistor consumes less energy and operates more efficiently. And because the researchers’ method enables them to slice much narrower trenches than conventional fabrication processes can, it also could enable semiconductor manufacturers to squeeze an even greater number of transistors onto an electronic device.

Ultimately, says Ma, because the mold can be reused, the method could easily scale for use in a technology called roll-to-roll processing (think of a giant, patterned rolling pin moving across sheets of plastic the size of a tabletop), and that would allow semiconductor manufacturers to repeat their pattern and mass-fabricate many devices on a roll of flexible plastic.

“Nanoimprint lithography addresses future applications for flexible electronics,” says Ma, whose work was supported by the Air Force Office of Scientific Research. “We don’t want to make them the way the semiconductor industry does now. Our step, which is most critical for roll-to-roll printing, is ready.”

Converting a single photon from one color, or frequency, to another is an essential tool in quantum communication, which harnesses the subtle correlations between the subatomic properties of photons (particles of light) to securely store and transmit information. Scientists at the National Institute of Standards and Technology (NIST) have now developed a miniaturized version of a frequency converter, using technology similar to that used to make computer chips.

False-color scanning electron micrograph of a nanophotonic frequency converter, consisting of a ring-shaped resonator (shaded blue) into which light is injected using a waveguide (shaded red). The input signal, depicted as a purple arrow, is converted to a new frequency (blue arrow) through the application of two pump lasers (light and dark red arrows). Credit: K. Srinivasan et al./NIST

False-color scanning electron micrograph of a nanophotonic frequency converter, consisting of a ring-shaped resonator (shaded blue) into which light is injected using a waveguide (shaded red). The input signal, depicted as a purple arrow, is converted to a new frequency (blue arrow) through the application of two pump lasers (light and dark red arrows). Credit: K. Srinivasan et al./NIST

The tiny device, which promises to help improve the security and increase the distance over which next-generation quantum communication systems operate, can be tailored for a wide variety of uses, enables easy integration with other information-processing elements and can be mass produced.

The new nanoscale optical frequency converter efficiently converts photons from one frequency to the other while consuming only a small amount of power and adding a very low level of noise, namely background light not associated with the incoming signal.

Frequency converters are essential for addressing two problems. The frequencies at which quantum systems optimally generate and store information are typically much higher than the frequencies required to transmit that information over kilometer-scale distances in optical fibers. Converting the photons between these frequencies requires a shift of hundreds of terahertz (one terahertz is a trillion wave cycles per second).

A much smaller, but still critical, frequency mismatch arises when two quantum systems that are intended to be identical have small variations in shape and composition. These variations cause the systems to generate photons that differ slightly in frequency instead of being exact replicas, which the quantum communication network may require.

The new photon frequency converter, an example of nanophotonic engineering, addresses both issues, Qing Li, Marcelo Davanço and Kartik Srinivasan write in Nature Photonics. The key component of the chip-integrated device is a tiny ring-shaped resonator, about 80 micrometers in diameter (slightly less than the width of a human hair) and a few tenths of a micrometer in thickness. The shape and dimensions of the ring, which is made of silicon nitride, are chosen to enhance the inherent properties of the material in converting light from one frequency to another. The ring resonator is driven by two pump lasers, each operating at a separate frequency. In a scheme known as four-wave-mixing Bragg scattering, a photon entering the ring is shifted in frequency by an amount equal to the difference in frequencies of the two pump lasers.

Like cycling around a racetrack, incoming light circulates around the resonator hundreds of times before exiting, greatly enhancing the device’s ability to shift the photon’s frequency at low power and with low background noise. Rather than using a few watts of power, as typical in previous experiments, the system consumes only about a hundredth of that amount. Importantly, the added amount of noise is low enough for future experiments using single-photon sources.

While other technologies have been applied to frequency conversion, “nanophotonics has the benefit of potentially enabling the devices to be much smaller, easier to customize, lower power, and compatible with batch fabrication technology,” said Srinivasan. “Our work is a first demonstration of a nanophotonic technology suitable for this demanding task of quantum frequency conversion.”

This work was performed by researchers at NIST’s Center for Nanoscale Science and Technology.

“Efficient and low-noise single-photon-level frequency conversion interfaces using silicon nanophotonics.” Q. Li, M. Davanço and K. Srinivasan.  Nature Photonics, 18 April 2016. DOI: 10.1038/nphoton.2016.64

MU, a medical-device manufacturer, and STMicroelectronics today announced that MU’s US-304 portable ultrasound imager, powered by ST’s STHV800 pulser, is aiming to increase the quality of point-of-care medical diagnostics in remote rural areas of Africa.

MU’s device has been developed for the “Doctor Car” mobile-clinic project to provide medical care in remote rural areas of Africa. In this project, medical workers use a special vehicle equipped with remote-healthcare systems to diagnose residents in remote rural areas where medical facilities are unavailable. The data obtained by the portable ultrasound device is transferred via mobile networks to healthcare entities in urban areas for detailed diagnosis and proper treatment. MU will start shipping ultrasound imagers to Doctor Cars and clinics in Africa this year.

The MU US-304 is a convex-type ultrasonic imager (3.5MHz) capable of performing abdominal diagnosis up to 15cm under the skin. It can be carried anywhere and simply connected via USB to a laptop or tablet. The MU device integrates ST’s high-voltage, high-speed ultrasonic-pulser IC (integrated circuit) with an 8-channel transducer driver circuit manufactured in ST’s proprietary 200V SOI-BCD semiconductor process. This process enables the integration of high-voltage CMOS technology, precise analog circuitry, and robust power stages on the same chip.

The industry’s most highly integrated ultrasonic pulser, ST’s STHV800 also offers low noise and tiny size to help produce accurate diagnostic images at a much lower cost and power consumption compared with stationary ultrasound equipment.

“The challenge in developing point-of-care ultrasound diagnostic devices is to achieve high portability and low cost without sacrificing performance. ST technology has proven an ideal solution to this problem,” said Yasuhiro Tamura, President, MU. “As we continue to create products for medical care in developing regions, in cooperation with ST, we hope to expand our application scope to new areas including livestock care.”

“MU’s newest portable ultrasound device is on course to improve the quality of medical diagnostics in remote rural areas, where the need is great,” said Hiroshi Noguchi, Director, Analog, MEMS and Sensors Group, STMicroelectronics Japan. “The selection of ST technology confirms our commitment to providing ultrasound-equipment makers with the highest performing ICs in the market and positions ST as the go-to partner for creating innovative applications that make positive contributions to people’s health and quality of life.”

ST offers a cost-effective evaluation board (STEVAL-IME013V1) that integrates the STHV800 pulser IC with an STM32F4 ARM Cortex-M microcontroller. The board’s graphical user interface and preset waveforms make it simple for designers to test the pulser under different conditions.

Imec, the nanoelectronics research center, today announced that its annual Imec Technology Forum (ITF) in Brussels will take place May 24-25, 2016 in Brussels, Belgium at SQUARE, Brussels Meeting Centre. ITF Brussels is the flagship of imec’s worldwide series of technology forums that brings experts and visionaries together to discuss the future of technology and tech-innovation to market. This year’s theme is “Daring to Take a Different View—Nanotechnology in the Hot Seat,” which will explore nanotechnology from all angles, question its future course, and identify new applications and paths for its use.

“The heart of imec is innovation and collaboration, and ITF Brussels will demonstrate that. Innovation is the result of hard work, endless questions, challenges to the status quo. Attendees will experience first-hand how constantly pushing these boundaries is essential to come up with groundbreaking solutions and stimulate innovation,” stated Luc Van den hove, president and CEO of imec. “The recent events in Brussels have deeply touched all of us, however the city is open for business and travel. Imec is privileged to bring our partners and international guests together in Brussels to focus on this year’s theme.”

Expecting to draw more than 1,000 attendees, ITF Brussels will offer numerous expert speakers from within the imec organization such as An Steegen, senior vice president process technology, and Wim Van Thillo, director perceptive systems for IoT, automotive and wireless. Industry speakers will also headlineincluding C-level executives from Samsung, Mentor Graphics, ASM International, Infineon Technologies, GlobalFoundries, J&J Pharmaceuticals, Audi, Microsoft, to name just a few.

ITF Brussels will introduce two new additions to the conference line up: interactive panel discussions and imec hot seats. Hot topics in today’s technology discussion will be explored such as “Scaling is dead. Long live scaling.”; “How close are we to precision medicine?’; “It’s a software world but it would be nothing without hardware?”, and “Combining ecological and economical sustainability.” Panelists will comprise both imec and guest partner executives and offer their various perspectives, and attendees will have an interactive role with panel voting and Q&A. Imec hot seats will place imec experts front and center to answer attendee questions, exchange ideas, investigate collaboration opportunities and consider different views. Spanning its vast R&D focus, topics such as energy, IoT, healthcare, services, and CMOS will be analyzed.

ITF Brussels 2016 takes place May 24-25, 2016 at the SQUARE, Brussels Meeting Centre. For the full list of speakers, conference program, registration and more information, please visit: http://www.itf2016.be/Homepage/page.aspx/2098

In a new study recently published in Nature Nanotechnology, researchers from Columbia Engineering, Cornell, and Stanford have demonstrated heat transfer can be made 100 times stronger than has been predicted, simply by bringing two objects extremely close–at nanoscale distances–without touching. Led by Columbia Engineering’s Michal Lipson and Stanford Engineering’s Shanhui Fan, the team used custom-made ultra-high precision micro-mechanical displacement controllers to achieve heat transfer using light at the largest magnitude reported to date between two parallel objects.

This is a still shot from a video of the high-precision micro-electromechanical system (MEMS) used to control the distance between two beams at different temperatures. The video is taken under a high magnification microscope. The whole video frame dimension is comparable to the diameter of a strand of human hair. Credit: Raphael St-Gelais, Lipson Nanophotonics Group, Columbia Engineering

This is a still shot from a video of the high-precision micro-electromechanical system (MEMS) used to control the distance between two beams at different temperatures. The video is taken under a high magnification microscope. The whole video frame dimension is comparable to the diameter of a strand of human hair. Credit: Raphael St-Gelais, Lipson Nanophotonics Group, Columbia Engineering

“At separations as small as 40 nanometers, we achieved almost a 100-fold enhancement of heat transfer compared to classical predictions,” says Lipson, Eugene Higgins Professor of Electrical Engineering and professor of applied physics. “This is very exciting as it means that light could now become a dominant heat transfer channel between objects that usually exchange heat mostly through conduction or convection. And, while other teams have demonstrated heat transfer using light at the nanoscale before, we are the first to reach performances that could be used for energy applications, such as directly converting heat to electricity using photovoltaic cells.”

All objects in our environment exchange heat with their surroundings using light. This includes the light coming at us from the sun, the glowing red color of the heating element inside our toaster ovens, or the “night vision” cameras that enable image recording even in complete darkness. But heat exchange using light is usually very weak compared to what can be achieved by conduction (i.e., by simply putting two objects in contact with each other) or by convection (i.e., using hot air). Radiative heat transfer at nanoscale distances, while theorized, has been especially challenging to achieve because of the difficulty of maintaining large thermal gradients over nanometer-scale distances while avoiding other heat transfer mechanisms like conduction.

Lipson’s team was able to bring objects at different temperatures very close to each other–at distances smaller than 100 nanometers, or 1/1000th of the diameter of a strand of human hair. They were able to demonstrate near-field radiative heat transfer between parallel SiC (silicon carbide) nanobeams in the deep sub-wavelength regime. They used a high-precision micro-electromechanical system (MEMS) to control the distance between the beams and exploited the mechanical stability of nanobeams under high tensile stress to minimize thermal buckling effects, thus keeping control of the nanometer-scale separation even at large thermal gradients.

Using this approach, the team was able to bring two parallel objects at different temperatures to distances as small as 42 nm without touching. In this case they observed that the heat transfer between the objects was close to 100 times stronger that what is predicted by conventional thermal radiation laws (i.e. “blackbody radiation”). They were able to repeat this experiment for temperature differences as high as 260oC (500oF) between the two objects. Such high temperature difference is especially important for energy conversion applications since, in these cases, the conversion efficiency is always proportional to the thermal difference between the hot and the cold objects involved.

“An important implication of our work is that thermal radiation can now be used as a dominant heat transfer mechanism between objects at different temperatures,” explains Raphael St-Gelais, the study’s lead author and postdoctoral fellow working with Lipson at Columbia Engineering. “This means that we can control heat flow with a lot of the same techniques we have for manipulating light. This is a big deal since there are a lot of interesting things we can do with light, such as converting it to electricity using photovoltaic cells.”

St-Gelais and Linxiao Zhu, who co-authored the study and is a PhD candidate in Fan’s group at Stanford, note that the team’s approach can be scaled up to a larger effective area by simply arraying several nanobeams–on top of a photovoltaic cell, for example–and by individually controlling their out-of-plane displacement using MEMS actuators. The researchers are now looking at applying their same approach for ultra-high-precision displacement control, this time with an actual photovoltaic cell to generate electricity directly from heat.

“This very strong, non-contact, heat transfer channel could be used for controlling the temperature of delicate nano devices that cannot be touched, or for very efficiently converting heat to electricity by radiating large amounts of heat from a hot object to a photovoltaic cell in its extreme proximity,” Lipson adds. “And if we can shine a large amount of heat in the form of light from a hot object to a photovoltaic cell, we could potentially create compact modules for direct conversion of heat to electrical power. These modules could be used inside cars, for instance, to convert wasted heat from the combustion engine back to useful electrical power. We could also use them in our homes to generate electricity from alternative energy sources such as biofuels and stored solar energy.”