Posted in | News | Semiconductor | Business

VLSI Symposium 2014 to Feature UCSB Team’s III-V MOSFETs

Researchers from the University of California, Santa Barbara (UCSB) will introduce today the highest performing III-V metal-oxide semiconductor (MOS) field-effect transistors (FETs) at the 2014 Symposium on VLSI Technology.

The UCSB research promises to help deliver higher semiconductor performance at lower power consumption levels for next-generation, high-performance servers. The research is supported by the Semiconductor Research Corporation (SRC), the world’s leading university-research consortium for semiconductors and related technologies.

The UCSB team’s III-V MOSFETs, for the first time in the industry, exhibit on-current, off-current and operating voltage comparable to or exceeding production silicon devices — while being constructed at small dimensions relevant to the VLSI (very-large-scale integration) industry.

For the past decade, III-V MOSFETs have been widely studied by a large number of research groups, but no research group had reported a III-V MOSFET with a performance equal to, let alone surpassing, that of a silicon MOSFET of similar size. In particular, UCSB’s transistors possess 25 nanometer (nm) gate lengths, an on-current of 0.5mA and off-current of 100nA per micron of transistor width and require only 0.5 volt to operate.

“The goal in developing new transistors is to reach or beat performance goals while making the transistor smaller—it is no good getting high performance in a big transistor,” said Mark Rodwell, professor of Electrical and Computer Engineering, UCSB. “In time, the UCSB III-V MOSFET should perform significantly better than silicon FinFETs of equal size.”

To reach this breakthrough in performance, the UCSB team made three key improvements to the III-V MOSFET structure. First, the transistors use extremely thin semiconductor channels, some 2.5nm (17 atoms) thick, with the semiconductor being indium arsenide (InAs). Making such thin layers improves the on-current and reduces the off-current. These ultra-thin layers were developed by UCSB Ph.D student Cheng-Ying Huang under the guidance of Professor Arthur Gossard.

Next, the UCSB transistors use very-high-quality gate insulators, dielectrics between the gate electrode and the semiconductor. These layers are a stack of alumina (Al2O3, on InAs) and zirconia (ZrO2), and have a very high capacitance density. This means that when the transistor is turned on, a large density of electrons can be induced into the semiconductor channel. Development of these dielectric layers was led by UCSB Ph.D student Varista Chobpattana under the guidance of Professor Susanne Stemmer.

Third, the UCSB transistors use a vertical spacer layer design. This vertical spacer more smoothly distributes the field within the transistor, avoiding band-to-band tunneling. As with the very thin InAs channel design, the vertical spacer makes the leakage currents smaller, allowing the transistor's off-current to rival that of silicon MOSFETs. The overall design, construction and testing of the transistor was led by UCSB Ph.D student Sanghoon Lee under Rodwell’s guidance.

“The UCSB team’s result goes a long way toward helping the industry address more efficient computing capabilities, with higher performance but lower voltage and energy consumption,” said Kwok Ng, Senior Director of Device Sciences at SRC. “This research is another critical step in helping ensure the continuation of Moore’s Law — the scaling of electronic components.”

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type
Submit

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.