
Spin polarisation and spin dependent scattering of holes in transverse magnetic
focussing
M. J. Rendell,1, ∗S. D. Liles,2A. Srinivasan,2O. Klochan,3, 1 I. Farrer,4, 5 D. A. Ritchie,5and A. R. Hamilton1, †
1School of Physics and Australian Research Council Centre of Excellence in Future Low-Energy Electronics Technologies,
University of New South Wales, Sydney, NSW 2052, Australia
2School of Physics, University of New South Wales, Sydney, NSW 2052, Australia
3University of New South Wales Canberra, Canberra, ACT 2600, Australia
4Department of Electronic and Electrical Engineering, University of Sheffield, Sheffield, S1 3JD, UK
5Cavendish Laboratory, University of Cambridge, Cambridge, CB3 0HE, UK
(Dated: October 10, 2022)
In 2D systems with a spin-orbit interaction, magnetic focussing can be used to create a spatial
separation of particles with different spin. Here we measure hole magnetic focussing for two different
magnitudes of the Rashba spin-orbit interaction. We find that when the Rashba spin-orbit magni-
tude is large there is significant attenuation of one of the focussing peaks, which is conventionally
associated with a change in the spin polarisation. We instead show that in hole systems with a k3
spin-orbit interaction, this peak suppression is due to a change in the scattering of one spin state,
not a change in spin polarisation. We also show that the change in scattering length extracted from
magnetic focussing is consistent with results obtained from measurements of Shubnikov-de Haas
oscillations. This result suggests that scattering must be considered when relating focussing peak
amplitude to spin polarisation in hole systems.
I. INTRODUCTION
In a magnetic focussing experiment, a collimated beam
of charge is focussed into a circular orbit by a transverse
magnetic field, analogous to a mass spectrometer. Mag-
netic focussing was originally proposed as a method of
studying the Fermi surface of metals [1, 2], and has also
been used to measure band structures in graphene [3],
and electron-electron scattering lengths in GaAs/AlGaAs
[4].
In systems with a spin-orbit interaction (SOI), the
magnetic focussing trajectories become spin dependent
as the spin states are now coupled to momentum. If the
SOI is sufficiently large, magnetic focussing can spatially
separate the spin states and create a spin-dependent mass
spectrometer [5–12]. The high mobility and large SOI
of 2D hole systems in GaAs has made them an ideal
candidate for spin-dependent magnetic focussing experi-
ments. Experimental work has used magnetic focussing
to measure spatial separation of spin [5], spin filtering
by quantum point contacts (QPCs) [7] and interactions
between 1D subbands in a QPC [13]. Magnetic focussing
of holes has also been proposed as a way to measure
g-factor anisotropies [14], and complex spin dynamics
[9, 15], which are not visible in other measurements of
2D systems such as Shubnikov-de Haas oscillations.
Here, we concentrate on the use of magnetic focussing
peak amplitude as a measure of the spin polarisation
[5, 7, 16–18]. It has been proposed that the relative am-
plitudes of the spin-split magnetic focussing peaks is de-
termined by the spin polarisation of the injected charge.
∗M. J. Rendell and S. D. Liles contributed equally to this work
†alex.hamilton@unsw.edu.au
This technique has been used in hole systems to observe
spontaneous polarisation in QPC transmission [7] and
spin-dependent transmission of QPCs [5, 13]. Despite
magnetic focussing being used for these techniques, there
has been limited study of the effect of changing the mag-
nitude of the Rashba SOI on hole magnetic focussing.
A recent study investigated magnetic focussing using a
device where the Rashba SOI magnitude could be tuned
in situ using a top gate voltage (VTG) [19]. This tech-
nique revealed an increase in the spatial separation of
the spin-split focussing trajectories as the Rashba SOI
was increased. However, there is a limit to the amount
the Rashba SOI can be changed using this method. In
addition, any change to VT G will also change the 2D hole
density and confining potential in addition to the Rashba
SOI magnitude. As such, further study requires a differ-
ent method of changing the Rashba SOI.
In this work we study magnetic focussing in two litho-
graphically identical samples which differ only in the
magnitude of the Rashba SOI. We change the Rashba
SOI by changing the heterostructure used to confine the
2D system, allowing us to create a large change in the
magnitude of the Rashba SOI for a similar VTG and 2D
density. By comparing the two samples, we observe a
change in the amplitude of the magnetic focussing peaks,
which is typically associated with a change in the spin po-
larisation. However, we instead find that the change in
peak amplitude is consistent with an increase in scatter-
ing of one spin state rather than a change in spin polar-
isation. We measure the scattering length of each spin
state from the focussing peak amplitude, and find good
agreement with scattering lenghts found from Shubnikov-
de Haas measurements. We conclude that the change in
focussing peak amplitude is due to the k3Rashba term
causing a different effective mass and hence scattering
length of each spin state, rather than a change in spin po-
arXiv:2210.03383v1 [cond-mat.mes-hall] 7 Oct 2022