% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@PHDTHESIS{SowmyaSpandana:972605,
author = {Sowmya Spandana, Somanchi},
othercontributors = {Stampfer, Christoph and Morgenstern, Markus},
title = {{T}ransport through graphene quantum point contacts},
school = {RWTH Aachen University},
type = {Dissertation},
address = {Aachen},
publisher = {RWTH Aachen University},
reportid = {RWTH-2023-10276},
pages = {1 Online-Ressource : Illustrationen, Diagramme},
year = {2023},
note = {Veröffentlicht auf dem Publikationsserver der RWTH Aachen
University; Dissertation, RWTH Aachen University, 2023},
abstract = {This thesis investigates low temperature transport through
graphene quantum point contacts (QPCs) encapsulated in
hexagonal boron nitride (hBN) using the van der Waals
technique. Single layer graphene (SLG)QPCs are fabricated
using electron beam lithography followed by SF6 based
reactive ion etching to define the shape and the width of
the QPC. In such devices, we observe that the rough edges
due to physical etching play an important role in the
quantized conductance characteristics of the QPC
particularly around the charge neutrality point (CNP). In
order to be able to achieve better control over the charging
of these localized edge states, we fabricate local top gates
to see if it is possible to control the edges independently
from the rest of the QPC channel. In one of the two devices
measured, we use a pair of top gates in a split gate
geometry that cover only the edges on either side of the
QPC. Here, we not only observe quantized conductance kinks
on the order of 2 - 3 e2/h but also a non-linear relative
gate lever arm. This can be explaine dusing the fact that
the edges are very likely to be terminated by Fluorine atoms
after etching with SF6 whichresults in higher charge
accumulation along the edges and consequently, a gate
voltage dependent gate lever arm. In the other device, we
employ a single top gate spanning the entire channel of the
QPC except the edges. We measure the conductance as a
function of both the top and the back gate voltages and
observe conductance kinks that result in a linear relative
gate lever arm. This dominant linear line denotes the charge
neutrality of each individually measured conductance trace
and its slope is referred to as the "major" slope. However,
interestingly, we also observe several other features that
evolve with a smaller "minor" slope. In both experiment and
theoretical calculations using the tight binding model, we
notice that sweeping the gates simultaneously along a
direction with the minor slope results in a much cleaner
conductance trace especially around the CNP where the edge
disorder is the maximum. This suggests that the features
corresponding to the minor slope are due to the effect of
the electric field lines of the top gate on the edge states.
Since these localized edge states are farther from the top
gate as compared to the channel, they are tuned less
strongly as compared to the Bloch states in the channel
right under the top gate. This is further corroborated by
the Landau fan measurements along both the directions with
major and minor slope. Here,we observe that (i) the larger
Landau level features at higher magnetic field appear to be
unaffected by the direction of sweep. (ii) There are number
of vertical straight lines that are unaffected by the
magnetic field in the low magnetic field, low charge carrier
density region around the CNP. These are the localized
states due to the edges. (iii) The number of such vertical
straight line features is lesser along the minor line than
any other direction of sweep of the gates. (iv) In general,
the evolution of the conductance kinks from the size
quantization to their respective Landau levels is much
cleaner along the direction of the minor line without a lot
of interference from localized states. Thus, we have been
able to use the top gate as a knob to disentangle the
features related to edge disorder from size quantization. We
then move to bilayer graphene (BLG) where we apply voltage
on a pair of split gates to define the width of the QPC.We
create a displacement field using the combination of the
split gates and a graphite back gate. This depletes the
charge carriers underneath the side gates, thereby creating
a 250 nm wide channel in between the source and the drain.
Using a layer of graphite as the back gate instead of the
doped Si as in the case of the single layer QPCs ensures
that the gate is much closer to BLG resulting in a far
better tuning besides also screening any impurities from the
surrounding SiO2 or hBN. We include three other finger gates
along the length of the QPC channel to tune the charge
carrier density locally. Conductance traces exhibit clear 4
e2/h steps that split into intermediate kinks at higher
values of parallel magnetic field indicating spin degeneracy
lifting. From the crossing points of spin-up and spin-down
branches of successive sub-bands, we extract the values of
sub-band spacing. More importantly, in the transconductance
plots as a function of the finger gate voltage and the
magnetic field, we observe discontinuities in the applied
voltage at (i) 0 T between the spin-up and spin-down levels
of the first sub-band. This is manifested in the form of a
step at 2e2/h that remains unaffected by the magnetic field.
(ii) Another gap in voltage is observed at a higher value of
magnetic field at the crossing point of the spin up level of
the first sub-band and the spin-down level of the second
sub-band. This is evident in the form of step at around 1.5
× 4 e2/h that travels down 4 e2/h which was observed
earlier in GaAs heterostructures and referred to as the 0.7
analog, similar to the 0.7 anomaly at 0 T as a result of
exchange/electron - electron (e-e) interactions. In our
device, we attribute the voltage gap at 0 T to a spin-orbit
(SO) coupling of the Kane - Mele type that dominates the e-e
interactions. While at higher magnetic field, this situation
is reversed and the Zeeman effects quenches the SO
interaction. Both these voltage gaps seem to evolve linearly
with the displacement field.},
cin = {132110 / 130000},
ddc = {530},
cid = {$I:(DE-82)132110_20140620$ / $I:(DE-82)130000_20140620$},
pnm = {SPINOGRAPH - Spintronics in Graphene (607904)},
pid = {G:(EU-Grant)607904},
typ = {PUB:(DE-HGF)11},
doi = {10.18154/RWTH-2023-10276},
url = {https://publications.rwth-aachen.de/record/972605},
}
guest ::