World’s first microsystem fabricated on 300 mm wafers
This summer, researchers at Leti fabricated the world’s first M&NEMS micro-accelerometers on 300 mm wafers, sending three crucial messages to the academic and industrial research and development communities.
First, the research proved that MEMS, or microelectromechanical systems, can be fabricated on 300 mm wafers, the largest format used in microelectronics. This advance will give MEMS access to the benefi ts associated with being among the most advanced technologies, from lower costs and energy consumption to more functions and smart capabilities.
97% of the fabrication process completed in house with a yield of 90%
Second, the future of Leti’s M&NEMS technology looks bright. It could ultimately enable sensors—accelerometers, gyrometers, magnetometers, pressure sensors, and microphones—to be fabricated using a single technology. The technology could also be used to build combination sensors—with three accelerometers and three gyrometers, for instance—on a single chip.
The technology appears to be highly compatible with miniaturization, and fabrication on 300 mm wafers will make it even more attractive to manufacturers.
Third, the achievement cements Leti’s position as the world’s leading center for MEMS R&D. The fabrication process used this summer counted around a hundred steps, 97% of which Leti completed in house. The accelerometers are fully functional and some structures reached yields of 90%.
The technology has passed the proof-of-concept hurdle. Interest from manufacturers will be the prerequisite for further development work. However, Leti has successfully increased its lead over its major competitors: IME in Singapore, C2MI in Canada, and IMEC in Belgium.
Spintec’s new STT-MRAM gives a sub-nanosecond performance
Spintec is developing spin-transfer torque magnetic random-access memory, or STT-MRAM, with write speeds under one nanosecond. That’s ten times faster than what Samsung and Intel are announcing for their 2016 product releases. Habitually, write speeds are around fi ve to ten nanoseconds. But Spintec’s memory uses a diff erent—and faster—process to trigger the write pulse. The secret is two orthogonal polarization vectors placed on each side of the memory layer that maintain a non-zero spin-transfer torque (STT).
The ultra-fast speeds could make Grenoble’s STT-MRAM compatible with very-high-potential applications like SRAM cache. The researchers are currently fi ne-tuning the concept, making improvements to the shape of the memory points and lowering switching energy as much as possible. They have fi led several patent applications.
Interview: Nanowirebased flexible LEDs a first step toward flexible displays
Interview: Christophe Durand, Research Scientist, INAC
« Nanowirebased flexible LEDs a first step toward flexible displays»
INAC and three other institutes (CNRS, Paris-Sud University, Grenoble-Alpes University Joseph Fourier School of Science) have developed the world’s First two-color, nanowire-based Flexible LEDs.
What does this advance mean?
Flexible LEDs are a prerequisite to flexible, bendable displays. They are currently being made from organic compounds at the cost of less-brilliant blues and shorter lifespans.
Our nitride nanowire-based LEDs are much longerlasting and can emit blue and green light. Once we add red, we will be able to emit white light and play videos.
How did INAC contribute to the research?
First, we used MOCVD to grow nitride nanowires typically measuring 1 micron in diameter by 25 microns in length. This is a topic we have been working on since 2010, and some of our research has earned grants from the French National Research Agency. We also transferred some of the technology we developed to startup Aledia.
Second, we developed a core-shell sheath around the nanowire. It is the amount of indium in the sheath’s indium gallium nitride quantum wells that determines the color of the emitted light. In this research we successfully increased the amount of indium by 25% to 30% to get green light.
And you’re able to bend the nanowires without breaking them?
Yes. They are coated with a polymer, and then detached from their substrate to obtain a nanowire carpet, which is a sort of fl exible membrane. It is the polymer that enables the material to bend.
Miniaturization record for single-electron CMOS transistor
Single-electron transistors (SETs) measuring just a few nanometers have been made by several teams of researchers around the world. However, none had managed to do it on CMOS and achieve operation at ambient temperature. Researchers from Leti and INAC have been successful on both fronts, with a transistor inside a 3.4 nm diameter silicon nanowire that operates from low up to ambient temperatures.
The research, which was published in Nanoletters, is crucial in that single-electron transistors have the potential to slash circuit energy consumption and could even be used for quantum processors. The researchers are currently working on reducing the variability of their transistor, even if that means compromising on miniaturization.
FDSOI design: Promising start for Silicon Impulse
Where can you design, prototype, and actually manufacture small runs of FDSOI circuits? Only at Leti’s Silicon Impulse design center! Whether you want to test a technology or scale it up for manufacturing, the center has everything you need.
Founded in March 2015, Silicon Impulse has already drawn attention from numerous manufacturers and academic laboratories in France and worldwide. Next February, Silicon Impulse’s first multi-project wafer will be made using STMicroelectronics’ 28 nm FDSOI technology; the circuits of various different partners will be found on the wafer.
FDSOI 28 nm is gaining traction on the Internet of Things and in low- and very-low-energy trends. Silicon Impulse, which is already backed by STMicroelectronics, is also in advanced-stage negotiations with Global Foundries, another international foundry giant, in an effort to reach the 22 nm FDSOI.
First-ever analysis of neutral particles by mass spectrometry
Researchers from INAC, Leti, and IRTSV recently completed the first-ever successful analysis of neutral particles by mass spectrometry. The scientific achievement, which took place right here in Grenoble, was praised by an article in Nature Communications. The researchers chose to bypass traditional spectrometry methods, preferring to use nano-electromechanical systems (NEMS), whose resonance frequency varies depending on the mass being supported. In this particular case, the mass was made up of nanometric copper and tantalum aggregates.
Biology a top-priority application
NEMS mass spectrometry has the advantage of being fully operational for non-ionized particles and/or very heavy particles—for which traditional spectrometry is not ideal. These advantages make the technique particularly attractive for biological research and clinical applications, where it can be used to detect things like viruses and protein complexes.
The method also marks a leap forward in the analysis of certain objects, previously hindered by the fact that particles were ionized prior to analysis—with NEMS the particles are not ionized.
The ten scientists who worked on this project filed for four patents and are pursuing their work on two NEMS spectrometers, one at INAC and one at Nanobio. They have also started work on a new system dedicated to neutral spaces and to NEMS for specific applications, with the goal of ultimately transferring the system to a spectrometry equipment manufacturer.
The next step will be to carry out an initial demonstration of the method on a biological species never before measured. The demonstration—slated for 2017—will validate the effectiveness of the technology.
MIT chooses Leti to supply suspended microresonators
The news dates back to early 2014 and has been kept under wraps for almost a year. A Massachusetts Institute of Technology biology lab is turning to Leti to develop some highly complex suspended microresonators. After receiving the initial components, which delivered satisfactory performance, the lab decided to switch foundries, and has now placed an order for 1,500 of the resonators with Leti for June.
The resonators boast a beam that vibrates at several hundred kHz and microfluidic channels through which cells flow. The frequency varies depending on the cells’ mass, which can be measured accurately down to the femtogram (10-15 g). This level of precision makes it possible, for instance, to separate cancerous cells from healthy ones.
One of the most technically-complex devices ever made at Leti
According to the researchers at Leti, the microresonator is one of the most technically-complex devices they have ever made. It packs in three substrates (silicon, SOI, and glass) and three sealing techniques (molecular, anode, and eutectic). The width of the microfluidic channels varies from 10 microns to 3 microns all along their path.
The fabrication process includes an impressive 150 steps (compared to just 50 for a traditional MEMS process). The quality factor, at more than 13,000, is very close to the state of the art for this type of resonator.
MIT generally does not initiate this type of partnership with other research organizations. The fact that the school turned to a lab halfway around the world is something Leti can be proud of. In the meantime, the researchers hope that this initial collaboration will lead to new partnerships.
Artificial pancreas tested on 15 patients
An artificial pancreas developed by CEA-Leti and CERTID (a private-sector diabetes research center) was recently tested on 15 patients. The system consists of an insulin pump, a blood-sugar sensor, and a dedicated algorithm installed on a smartphone. The algorithm uses the patient’s profile and activity (meals, physical exercise) to calculate the insulin dosage, which is automatically delivered by the pump.
Clinical testing demonstrated that the artificial pancreas was very effective at maintaining patients’ blood sugar at the right levels. A joint lab between Leti and CERTID is currently being set up to pursue the development work. An initial prototype should be available by the end of 2015.
PCR could soon be used in hospital emergency rooms
Hospitals currently perform their PCR (polymerase chain reaction) analyses at their central labs. It will soon be possible to complete the analyses right in hospital units like the emergency room. Researchers from CEA-Leti and Dr. Didier Raoult’s team at IHU Méditerranée Infection in Marseille have joined forces to tackle the problem.
The researchers from CEA-Leti are adapting their Flowpad microfluidics platform to automate PCR analysis of human samples like spinal fluid, blood, and saliva so that clinicians can perform the analyses themselves right in their hospital units.
The innovation will first target STDs, meningitis, and encephalitis, pathologies that represent more than a million tests a year in France alone. The researchers have filed for a patent for the system, which will be evaluated in late 2015.