Spin and Electron Coherence in 2D, 1D, and 0D Systems
C.M. Marcus1, S.M. Cronenwett1,2, H.J. Lynch1, D. Zumbühl1, J.B. Miller1,3
1
Department of Physics, Harvard University, 2Department of Physics, Stanford
University, 3Division of Engineering and Applied Sciences, Harvard University,
As a possible qubit for the storage and manipulation of quantum information, the spin of
an electron appears to have a number of favorable features, particularly its relative
immunity to environmental influence, compared, for instance, to the electron phase,
more commonly investigated in mesoscopic physics. For this reason, as well as
motivations concerning the use of spin for information storage in more classical
settings, a great effort has been focused lately on controlling electron spin in electronic
structures.
This talk will review three aspects of our recent work on spin effects in mesoscopic
semiconductor structures. Topics are arranged by the dimensionality of the system:
Two Dimensions: We have revisited the problem of quantum coherence signatures of
spin in clean (i.e. high mobility) two dimensional electron gasses (2DEGs) [1]. A well
investigated signature of spin physics in quantum coherent systems is antilocalization,
an enhancement of the conductivity zero magnetic field due to the destructive
interference of counterpropagating diffusive trajectories (“anti” here is in contrast to the
situation in the absence of spin-orbit coupling, where interference always produces
constructive backscattering and hence a reduction in conductivity at zero field, known as
weak localization). Surprisingly, in the domain of high mobility, where one might hope
to implement spintronic devices, both theory and experiment are not well explored. We
have investigated antilocalization in a high-density, high mobility gated GaAs/AlGaAs
2DEG. We are able to cross over from weak localization to antilocalization by gating
the device. By fitting to a new theory of magnetotransport in the quasi-ballistic regime
are able to extract the strength and angular dependence of spin-orbit coupling from
transport.
One dimension: Besides the well understood effect of mode quantization, giving rise to
plateaus in conductance at 2e2/h, semiconductor quantum point contacts commonly
show an extra plateau-like feature in conductance as a function of gate voltage at around
0.7(2e2/h). Mounting evidence suggests that this feature is spin related. We have
explored an interesting aspect of this so-called 0.7 feature, which is the way that it
disappears as the temperature is lowered [2]. We suggest a Kondo-like origin of the
disappearance, and demonstrate that the extracted Kondo temperature can be varied by
more than an order of magnitude by changing the confining gate voltage. If this picture
is correct, then creating a single spin at an arbitrary location in an otherwise
homogeneous, nonmagnetic material becomes rather easy: one simply makes a
constriction in a clean 2DEG, adjusts the gates to the vicinity of the 0.7 feature, and a
free spin appears. To entangle that spin with a reservoir, one can then lower the
temperature below the (gate-controlled) Kondo temperature. Experimental details
leading to this picture will be presented.
Zero Dimension: The intuitive picture of an electron “standing wave” in a zero-
dimensional quantum dot becomes somewhat hard to visualize when the additional
influences of Zeeman splitting and spin-orbit interactions become significant. The
picture of a spin-resolved standing wave offers a number of interesting possibilities,
from filtering and detection of spin to novel quantum interference effects such as
ballistic-chaotic antilocalization. Recent theory has addressed quantum dots with both
spin-orbit and Zeeman splitting at the level of basic symmetries of the problem, as well
as a full random matrix theory of quantum magnetotransport [4, 5]. We describe recent
experiments carried out in a GaAs quantum dot with sufficiently high density that
antilocalization appears. Experimental results [3] are found to be in good agreement
with theory. These measurements are the first investigation of antilocalization in a
quantum dot, and the first demonstration of the predicted quenching of antilocalization
due to confinement, to our knowledge.
We gratefully acknowledge support from ARO-MURI, DAPRA, the National Science
Foundation, the NDSEG program (JBM), and the Lucent GRPW Program.
[1] J.B. Miller, D.M. Zumbühl, D. Goldhaber-Gordon, C.M. Marcus, Y. Lyanda-Geller, K. Campman,
A.C. Gossard, Gate-Controlled Spin-Orbit Quantum Interference Effects in Lateral Transport (in
preparation).
[2] .S.M. Cronenwett, H.J. Lynch, L.P. Kouwenhoven, D. Goldhaber-Gordon, N. S. Wingreen, K. Hirose,
V. Umansky, Low-Temperature Fate of the 0.7 Structure in a Point Contact: A Kondo-like Correlated
State in an Open System, Physical Review Letters (in press.)
[3] D.M. Zumbühl, J.B. Miller, C.M. Marcus, K. Campman, A.C. Gossard, Gate-Controlled Spin-Orbit
Quantum Interference Effects in Lateral Transport (in preparation).
[4] I. L. Aleiner and V.I. Falko, Physical Review Letters 87, 256801 (2001).
[5] J. H. Cremers, P. W. Brouwer, B.I. Halperin, I.L. Aleiner, V.I. Falko, (in preparation).