|
|
|
|
|
Electron Transport in Strained Silicon
Two-dimensional electron systems (2DES) in semiconductors have been the primary vehicle
for research on quantum mechanics of the large number of interacting electrons as identical
particles in an ordinary solid, a subject that has been a major source for new physics
challenging and fascinating the human mind, and an incubator for revolutionary technologies.
From the perspective of applications, fractional quantum Hall effect observed in correlated
2DES in semiconductor quantum wells has been considered as the ideal physical embodiment
for topological quantum computing. Topological quantum computing is ostensibly immune to
the problem that plagues the other quantum computational schemes, namely, quantum
de-coherence and allows for large, many qubit quantum computers to be realized.
To date, 2DES in GaAs-AlGaAs heterostructures has spearheaded the understanding in this
field because of the availability of samples with extremely high mobility, homogeneity and
low electron density, the three characteristics necessary for detailed study of ultra-low
energy processes. Unlike GaAs with a single conduction band minimum, 2DES in Si under
biaxial tensile stress has two valleys to its conduction band thereby offers the additional
valley degree of freedom that is expected (and has been shown in some cases) to lead to a
plethora of exciting phenomena that will undoubtedly advance the understanding in this field.
The one hurdle that has been impeding the study of 2DES in strained Si is the availability
of high quality samples comparing to that in GaAs-AlGaAs materials systems. For example,
the highest electron mobility ever reported in strained Si is a factor of 50 lower than the
31,000,000 cm2/V-s value reported in GaAs, with the actual reproducible mobility values
being around 300,000 cm2/V-s. Our group has the technical know-how for reproducibly fabricate
highly homogeneous 2DES in strained Si with mobility values ~350,000 cm2/V-s.
In addition to continued effort in improving the mobility, we are also working on reducing
the lower limit achievable for the density of 2DES. Our approaches include two parallel ways
for further minimizing the background impurity including the neutral ones and for eliminating
the potential undulation due to deformation potential from the buried dislocations in
conventional relaxed SiGe buffer layers. We are also exploring the fabrication of 2DES in
Si under uniaxial strain along different directions with respect to that of the current flow
and the crystallographic orientation of Si. Such device structure is expected to permit very
large density parameter rs to be achieved leading to high interaction strength between
electrons and accentuating correlated behaviors.
It is important to note the fundamental difference between the “ideal” samples for
transport physics and those for modern field effect transistor technologies, especially
when strained Si is concerned. For transistor applications, the most important figure of
merit is high transconductance at room temperature that requires simultaneously high
electron density, mobility, and saturated velocity under high electric field. In contrast,
the ideal sample for transport study should have high mobility, homogeneity and low
electron density at LHe temperature. The governing physics for the two sets of requirements
are fundamentally different, and so are the optimized device structures.
* The research is in collaboration with the research group of Prof. Daniel C. Tsui of Princeton University.
Figure 1. TEM picture of a typical sample of 2DES in strained Si (of 14 nm thickness) with n-type delta-doped layer 22 nm above the strained Si channel.
Figure 2. Magneto-transport in 2DES in strained Si.
Figure 3. Schematics of the process flow in fabricating strained Si on porous Si substrates.
Figure 4. The measured values of biaxial strain in a 100 nm thick single crystalline Si film on
partially oxidized microporous Si substrate. Strain relaxation in some cases is via crack
formation of dislocation half-loop introduction.
|
|
|
|
Semiconductor Materials Research Laboratory, Department of Materials Science and Engineering, University of California at Los Angeles
Box 951595, Los Angeles, CA 90095-1595 (Tel) +1 310 825 2971 UCLA SMRL © 2006 | All Rights Reserved
|
|
|
|
|
|
