Nov 11, 2011
InAs membranes could compete with Si MOSFETs
A novel class of 2D semiconductor has been developed by researchers in the US. The free-standing quantum membranes, which are made of indium arsenide and are just a few layers thick, have properties that are in stark contrast to those of conventional MOSFETs. Besides the fundamental science implications, the new work could help us better understand how so-called structurally confined semiconductor devices function.
Two-dimensional semiconductors can be used in high-performance electronics or quantum optics and scientists have recently turned their attention to various layered materials, such as graphene and MoS2 because of their intriguing properties. Now, Ali Javey of the University of California at Berkeley and colleagues have developed another class of 2D semiconductors made of indium arsenide (InAs), dubbed quantum membranes (QMs), whose band structures can be tuned from bulk to 2D simply by changing the thickness of the material. Bulk InAs has a large Bohr radius of 34 nm, which means that the electrons can be readily quantum confined in the material – even when layers in the sample are as big as tens of nanometres across. It also means that size effects in this material are fundamentally different to those seen in other layered semiconductors.
Quantum confinement describes how the electronic and optical properties of a material change as the sample size becomes smaller – typically to 10 nm or less. The effect comes thanks to electrons and holes being squeezed into a space that approaches a critical quantum measurement, called the exciton Bohr radius.
Unlike conventional III-V quantum-well structures that are confined by growing a large band gap semiconductor layer on an epitaxial growth substrate, InAs QMs are confined at both top and bottom surfaces by either an insulator layer and/or a vacuum. These structures are thus free standing and can be placed on a variety of substrates. What is more, the structure of the QM is such that the active semiconductor layer in the material can be placed in direct contact with the gate stack – something that allows for high-performance devices.
Javey's team employed an epitaxial layer transfer technique that involves growing 5–50 nm thick InAs on GaSb/AlGaSb substrates. The top InAs layer is then patterned to the desired shape and structure and this is followed by a selective wet etch of the sacrificial AlGaSb layer. The InAs layer is then picked up and transferred onto a substrate like Si/SiO2 or CaF2 with the help of a PDMS stamp.
"The result is a free-standing InAs QM on a substrate of our choice that is bonded through van der Waals interactions without the constraints of the original growth substrate," Javey told nanotechweb.org. "This allows us to explore the fundamental material and device properties of QMs of varying thicknesses."
Thanks to optical absorption studies, the researchers succeeded in mapping out the sub-band energies of InAs QMs while varying the thickness of the structures. They also explored the effect of quantized sub-bands on the electrical properties of field-effect transistors made from the QMs and found that electron mobility in the material did not depend on the applied field, except at very high fields. Such behaviour is very different to that seen in conventional MOSFETs.
High carrier mobility
"This study reveals the basic transport physics and performance limitations in QM field-effect transistors (QMFETs) by studying the quantum confinement effect in transport properties such as effective mobility and quantum resistance, thus providing guidelines for future device designs, especially for III-V nanoelectronics," said Javey. "III-V devices on Si substrates are particularly attractive candidates for replacing conventional Si MOSFETs, thanks to their high carrier mobility, which can enable high-speed, low-power devices."
The team now plans to use its platform to further explore various other fundamental material properties in these semiconductors and discover how good III-V QMFETs really are when compared to state-of-the-art Si MOSFETs.
The work is detailed in Nano Letters.
About the author
Belle Dumé is contributing editor at nanotechweb.org