
ALBERTA-THY-7-22
Majorana Modes of Giant Vortices
Logan Gates1, ∗and Alexander A. Penin1, †
1Department of Physics, University of Alberta, Edmonton, Alberta T6G 2J1, Canada
We study Majorana zero modes bound to giant vortices in topological superconductors or topo-
logical insulator/normal superconductor heterostructures. By expanding in inverse powers of a large
winding number n, we find an analytic solution for asymptotically all nzero modes required by the
index theorem. Contrary to the existing estimates, the solution is not pinned to the vortex boundary
and is composed of the warped lowest Landau level states. While the dynamics which shapes the
zero modes is a subtle interference of the magnetic effects and Andreev reflection, the solution is
very robust and is determined by a single parameter, the vortex radius. The resulting local density
of states has a number of features which give remarkable signatures for an experimental observation
of the Majorana fermions in two dimensions.
Majorana quasiparticle excitations in various con-
densed matter systems are in a spotlight of theoretical
and experimental studies for over a decade [1–3]. A
renown example of the Majorana quasiparticles are the
zero-energy states bound to the vortices in a topological
superconductor [4, 5] or on the interface between a topo-
logical insulator and a normal superconductor [6]. The
giant vortices of large winding number nare of particular
interest since they host multiple zero modes [7] and can
be used to study highly nontrivial systems of interacting
Majorana states such as the Sachdev-Ye-Kitaev model
[8]. The vortices with n > 1 have already been observed
in mesoscopic semiconductors [9] and can be engineered
in the specially designed heterostructures. Unambiguous
identification of the zero modes in the vortex core poses a
great challenge for the modern experimental techniques
[10, 11], and the measurement of their spatial distribu-
tion is a promising method to distinguish the true Ma-
jorana states [12]. Thus, the search and design of the
systems with the signature spatial properties of the zero
modes as well as the theoretical evaluation of their shape
and the local density of states are of primary interest.
Though the existence and stability of the zero modes are
predicted by the index theorem [13, 14] and can be veri-
fied through a qualitative analysis of the field equations,
such an approach is too coarse to catch subtle dynami-
cal effects which significantly affect the structure of the
solution. On the other hand, the brute-force numerical
simulations may be insufficient to identify the univer-
sal properties and characteristic features of the solution,
hence a form of quantitative analytic approach is manda-
tory. A systematic analysis in this case is complicated by
nonintegrable nonlinear nature of the vortex dynamics.
In this Letter we present such an analysis for the case
of the giant vortices. It is based on a novel method of
the asymptotic expansion in inverse powers of the vortex
winding number [15, 16] and provides the analytic solu-
tion for almost all of the nzero modes required by the
index theorem. Our prediction for the density of states
has a number of remarkable properties which have been
overlooked before and can be used to get compelling ex-
perimental evidence of the Majorana vortex states.
The equations for the Majorana zero modes can be
inferred from the Jackiw-Rossi theory of charged mass-
less two-component Dirac fermion in 2 + 1 dimensions
described by the Lagrange density [7]
LJR =i¯
ψ/
Dψ +1
2¯
ψψcφ+¯
ψcψφ∗,(1)
where /
D=γµDµ,Dµ=∂µ+iAµis the gauge covariant
derivative, the Dirac matrices reduce to the Pauli ma-
trices γµ= (σ3, iσ2,−iσ1), ψc=−iσ1ψ∗is the charge
conjugate spinor, and φis a scalar field of charge 2 rep-
resenting the pair potential. For the static zero-energy
states the field equations for the spinor components read
D±ψ±+φψ∗±= 0 ,(2)
where the chiral derivatives are D±=D1±iD2. We
are interested in the solution of Eq. (2) in the back-
ground of the axially symmetric Abrikosov vortex [17] of
the winding number n, which implies the following field
configuration in polar coordinates φ(r, θ) = f(r)einθ ,
Aθ=−na(r)/2, Ar= 0, with f(0) = a(0) = 0 and
f(∞) = f∞,a(∞) = 1. Then the negative chirality
equation does not have a normalizable solution and the
nzero modes of positive chirality can be written as fol-
lows
ξ+
l=1
√2eilθψ+
l+ei(n−1−l)θψ+
n−1−l,
η+
l=i
√2eilθψ+
l−ei(n−1−l)θψ+
n−1−l,
(3)
where 0 ≤l≤n/2−1 for even nand 0 ≤l≤(n−1)/2,
η+
(n−1)/2= 0 for odd n. The partial wave amplitudes
satisfy the following equations
d
dr −l
r+na
2rψ+
l+fψ+
n−1−l= 0 ,
d
dr −n−l−1
r+na
2rψ+
n−1−l+fψ+
l= 0 .
(4)
After identification of ψland ψn−1−lwith the compo-
nents of the Nambu spinor, and of fwith the pair po-
tential the above system reproduces the Bogoliubov-de-
Gennes equations for the Majorana vortex zero modes of
arXiv:2210.04908v1 [cond-mat.mes-hall] 10 Oct 2022