Phase-controlled quantum transport signatures in a quantum dot-Majorana hybrid ring system
Extracted source text · arXiv 2510.07826
%\documentclass[twocolumn,superscriptaddress,showpacs,aps,prb,amsmath,amssymb]{revtex4}
\documentclass[twocolumn, superscriptaddress,showpacs,amsmath,amssymb]{revtex4}
% \documentclass[superscriptaddress,aps,amsmath,amssymb,pra]{revtex4}
%\documentclass[aps,amsmath,amssymb,pra]{revtex4}
%\documentclass[twocolumn,superscriptaddress,showpacs,aps]{revtex4}
%\documentclass[preprint,superscriptaddress,showpacs,aps]{revtex4}
%
%\usepackage{CJK} %%%%可以加入中文字%Li-version
%\usepackage{CJKutf8} %%jsjin
%
\usepackage{mathrsfs}
\usepackage{hyperref}
\usepackage[dvips]{graphics,epsfig}
\usepackage{amstext}
\usepackage{amssymb}
\usepackage{amsmath}
\usepackage{graphicx,color}
\usepackage{bm}
\newcommand{\quan}[1]{\textcircled{\footnotesize{#1}}}%%% 圆圈1的标记
\newcommand{\leftact}{\overset{\rightarrow}}
\newcommand{\rightact}{\overset{\leftarrow}}
\newcommand{\leftrightact}{\overset{\leftrightarrow}}
\newcommand{\s}{\mbox{\tiny S}}
\newcommand{\tS}{\mbox{\tiny S}}
\newcommand{\w}{\omega}
\newcommand{\B}{\mbox{\tiny B}}
\newcommand{\U}{\mbox{\tiny U}}
\newcommand{\SB}{\mbox{\tiny SB}}
\newcommand{\T}{\mbox{\tiny T}}
\newcommand{\M}{\mbox{\tiny M}}
\newcommand{\D}{\mbox{\tiny D}}
\newcommand{\tL}{\mbox{\tiny L}}
\newcommand{\tR}{\mbox{\tiny R}}
\newcommand{\ti}{\Tilde}
\newcommand{\wti}{\widetilde}
\newcommand{\nl}{\nonumber \\}
\newcommand{\nla}{\nl&\quad}
\newcommand{\Sch}{Schr\"{o}dinger\ }
\newcommand{\up}{\uparrow}
\newcommand{\down}{\downarrow}
\newcommand{\Sec}[1]{Sec.\,\ref{#1}}
\newcommand{\App}[1]{Appendix\,\ref{#1}}
\newcommand{\be}{\begin{equation}}
\newcommand{\ee}{\end{equation}}
\newcommand{\bea}{\begin{eqnarray}}
\newcommand{\eea}{\end{eqnarray}}
\newcommand{\bsube}{\begin{subequations}}
\newcommand{\esube}{\end{subequations}}
\newcommand{\ben}{\begin{equation*}}
\newcommand{\een}{\end{equation*}}
\newcommand{\Eq}[1]{Eq.\,(\ref{#1})}
\newcommand{\Eqs}[1]{Eqs.\,(\ref{#1})}
\newcommand{\Fig}[1]{Fig.\,\ref{#1}}
\newcommand{\Tab}[1]{Table \ref{#1}}
\newcommand{\dg}{\dagger}
\newcommand{\la}{\langle}
\newcommand{\ra}{\rangle}
\newcommand{\kb}{\rangle\langle}
\newcommand{\mb}{\mbox}
\newcommand{\cf}{cf.\ }
\newcommand{\lla}{\langle\langle}
\newcommand{\rra}{\rangle\rangle}
\newcommand{\La}{\big\la}
\newcommand{\Ra}{\big\ra}
\newcommand{\tprime}{\mbox{\tiny $^\prime$}}
\newcommand{\greater}{\mbox{\tiny $>$}}
\newcommand{\lesser}{\mbox{\tiny $<$}}
\newcommand{\lgter}{\mbox{\tiny $\lessgtr$}}
\newcommand{\Opm}{\hat O_{\mbox{\tiny $\pm$}}}
%\pagestyle{plain}
\begin{document}
%\begin{CJK*}{GBK}{Song} %%XQLi
%\begin{CJK*}{UTF8}{gbsn} %%%jsjin
\title
{Phase-controlled quantum transport signatures in a quantum dot-Majorana hybrid
ring system}
\author{Sirui Yu}
\affiliation{School of Physics, Hangzhou Normal University,
Hangzhou, Zhejiang 311121, China}
\author{Junrong Wang}
\affiliation{School of Physics, Hangzhou Normal University,
Hangzhou, Zhejiang 311121, China}
\author{Huajin Zhao}
\affiliation{School of Physics, Hangzhou Normal University,
Hangzhou, Zhejiang 311121, China}
\author{Hong Mao} %\email{mao@hznu.edu.cn}
\affiliation{School of Physics, Hangzhou Normal University,
Hangzhou, Zhejiang 311121, China}
\author{Jinshuang Jin} \email{jsjin@hznu.edu.cn}
\affiliation{School of Physics, Hangzhou Normal University,
Hangzhou, Zhejiang 311121, China}
\date{\today}
\begin{abstract}
We investigate the quantum transport in a hybrid ring system
consisting of a quantum dot (QD) coupled to two Majorana bound states (MBSs) hosted
in a topological superconducting nanowire, threaded by a magnetic flux.
%
Utilizing the dissipaton equation-of-motion approach,
we demonstrate that the differential conductance shows periodic behavior and
its periodicity depends on both the QD energy level and the MBS overlapping.
%
A zero-bias peak (ZBP) emerges as a result of
the balance between normal
and anomalous tunneling processes,
associated with the presence of
a single MBS.
%
Beyond the phase-dependent periodic behavior,
the shot noise exhibits voltage-dependent transitions
between sub-Poissonian ($F = 0.5$),
Poissonian ($F = 1$),
and super-Poissonian ($F > 1$) regimes.
%
Strikingly, we find a giant Fano factor ($F\gg1$)
emerging at the balance point, accompanied by a peak in the shot noise.
%
This distinctive feature may serve as a supplementary signature for MBS detection.
%
However, both ZBP in the differential conductance and shot noise peak
are degraded by thermal effects.
\end{abstract}
%\pacs{05.40.Ca, 73.63.-b, 73.23.Hk}
\maketitle
\section{Introduction}
%
Inspired by the pioneering theoretical contributions of Kitaev in 2001 \cite{Kit01131}, Majorana bound
states (MBSs) have garnered widespread attention in condensed matter physics
and quantum information science.
%
On the one hand, MBSs possess distinctive physical attributes,
characterized by their non-Abelian braiding statistics \cite{Kit032}.
On the other hand, owing to these exceptional properties,
MBSs hold promise for applications in topological quantum computers
\cite{Kit01131,Kit032,Nay081083,Ali10125318,Lei11210502,Aas16031016}.
%
%%%%%
Various physical platforms have been proposed to realize MBSs
\cite{Lut10077001,Ore10177002,
Fle10180516,Sar16035143,Moo18165302,Law09237001,
Dan20036801,Men20036802}.
%
Among these, a popular scheme is
the spin--orbit coupling
semiconductor nanowire proximity coupled to a superconductor.
In this platform, %
zero-energy MBSs, also referred to as Majorana zero-energy modes (MZMs),
are expected to reside at the nanowire ends
\cite{Lut10077001,Ore10177002}.
%%
Current experimental efforts focus on
detecting and verifying MBS signatures,
with transport measurements serving as a key diagnostic tool.
%
For instance, distinct transport signatures have been demonstrated in setups where
MBSs are connected to either single lead (two-terminal)
\cite{Fle10180516,Sar16035143,Moo18165302,Law09237001}, %Vui19061,
or two leads (three-terminal) \cite{Dan20036801,Men20036802}, typically via
the observation of zero-bias peak in the tunneling conductance spectroscopy.
%
In addition, hybrid architectures incorporating quantum dots (QDs) as intermediaries between MBSs
and transport leads,
have been proposed \cite{Cao12115311,Zoc13036802,Fen21123032,Liu15081405,Smi17063020,
Fen22035148,Smi22205430,Smi19165427,Cao23121407,Zha25023006}.
These studies suggested that
Majorana signatures may also be inferred through
current noise measurements.
%
Although these measurements provide encouraging hints,
the conclusive evidence for MBSs have not yet been obtained,
and the distinctive transport characteristics associated with their presence
require further exploration.
%%%%
In this context, the study of the nonequilibrium transport through a
hybrid ring system involving MBSs
represents an exciting and promising avenue of research.
The system consists of a QD coupled
to two Majorana modes located at the ends of a topological
superconducting nanowire (TNW), with a magnetic flux $\Phi$ through the loop,
as schematically shown in \Fig{fig1}.
%
This hybrid ring configuration was initially proposed by Flensberg \cite{Fle11090503} for
performing the non-Abelian operations on MBSs through magnetic flux tuning.
%
Subsequently, Liu and Baranger \cite{Liu11201308}
predicted that this system can be used to tune Flensberg’s
qubit to the required degeneracy point (the presence of single MBS), where
the zero-temperature peak conductance reaches half of the quantized value,
$G=e^2/h$, while vanishing elsewhere.
This half-quantized conductance has been demonstrated to be
robust against variations in the QD energy level \cite{Ver14165314}
and has been measured in a QD coupled to an InAs--Al nanowire heterostructure \cite{Den161557}.
The QD--TNW interferometer can also serve as a mesoscopic probe of interwire couplings
\cite{Ram18045301}.
%
In recent years, the hybrid QD--TNW ring system has garnered significant interests
\cite{Wan181,Zen17127302, Cal201900409,Smi21205406, Smi24195410,Gon22122,Mat22155409,Mat231616,Bah25045920}.
A number of signatures associated with MBSs have been identified in the strong coupling regime, such as
Fano resonances \cite{Wan181,Zen17127302, Cal201900409},
fractional entropy \cite{Smi21205406}, and differential finite-frequency quantum noise \cite{Smi24195410}.
%
Furthermore, studies have addressed Coulomb-interaction-enhanced
double-resonant tunneling \cite{Gon22122},
and the influence of electron–phonon coupling on conductance
periodicity \cite{Mat22155409,Mat231616}.
%
More recently, a four-MBS configuration with two TNW loops has been investigated,
uncovering periodic structures in both linear and differential conductance \cite{Bah25045920}.
%
Despite these advances, the transport properties originating purely MBSs
are still not fully understood, even in the simple geometry of \Fig{fig1}
where electron–electron and electron–phonon interactions are neglected.
%%%
In this work, we present a comprehensive study
of the magnetic-flux-dependent differential conductance and shot noise
in the hybrid QD--TNW ring interferometer (\Fig{fig1}).
%
We focus on the weak coupling between the QD and electrodes
at finite temperature---a scenario readily accessible in current experiments.
%
Our results show that the signatures originating from MBSs are modulated by the flux.
Accordingly, the differential conductance and shot noise display
the periodic structures.
%
We provide a general condition governing the periodicity.
Notably, the shot
noise reveals voltage-dependent statistical regimes:
sub-Poissonian (Fano factor $F=0.5$), Poissonian ($F=1$), and super-Poissonian ($F>1$).
A pronounced Fano factor emerges at zero bias, originating purely from MBSs.
%
All these transport properties are thoroughly
investigated in our study.
We believe these results contribute to deeper understanding
of MBS-mediated quantum transport
through mesoscopic systems.
%%%%%%%%%%%%%%%%%%%
The present numerical calculations are performed using
the fermionic dissipaton equations-of-motion (DEOM)
approach, which provides accurate evaluations on the
transport current and its noise spectrum \cite{Yan14054105,Jin15234108,Yan16110306}.
%
DEOM is a quasi-particle generalization of
the well--established hierarchical equation of motion
(HEOM) \cite{Tan89101,Tan906676,Tan06082001,Tan20020901,
Xu05041103,Jin08234703}. HEOM rooted in the path integral influence functional theory
\cite{Fey63118,Wei08,Kle09}, to efficiently simulate interacting open quantum systems.
Its hierarchical construction captures dissipation, memory time,
Columbic interactions, and cotunneling.
%
The key features for fermionic HEOM/DEOM are as follows
\cite{Jin08234703,Zhe121129,Li12266403,Ye16608}:
(1) For noninteracting systems,
the second-tier truncation
is exact and agrees with Landauer–Büttiker scattering
\cite{Dat95} and nonequilibrium Green function methods \cite{Hau08} and
%%%
(2) for interacting systems, convergence is rapid with finite tiers,
which depends on both the system and bath configurations.
%
HEOM/DEOM has been widely applied transient
transport \cite{Zhe08184112,Zhe08093016,Zhe09164708,Zhe13086601},
thermopower \cite{Ye14165116}, and the spectral density of local impurity system in
the Kondo regime \cite{Li12266403,Wan13035129}.
%
Via the dissipaton decomposition for the hybridizing bath (electrodes),
DEOM theory comprises
both the HEOM formalism \cite{Jin08234703}
and the dissipaton algebra that addresses
the hybrid bath dynamics.
%
Consequently, DEOM serves as a unified framework for evaluating various noise-related
spectra \cite{Zha18780,Wan20041102,Zha16237,Zha16204109},
%
including the quantum current noise \cite{Jin15234108,Jin20235144}.
%
For further detail, we refer to Refs.\cite{Yan14054105,Jin15234108,Yan16110306}.
%
The remainder of this paper is organized as follows. In \Sec{thmeth}, we give
the model description and introduce the definitions
of the current and its power spectrum. Based on the DEOM method
%
which is briefly introduced
in tbe \App{appdeom},
%
we carry out the precise
numerical results in \Sec{thresu}.
%
First,
we analyze the QD--wire Hamiltonian eigenspectrum in \Sec{thham}.
Then we demonstrate the detailed results
for the differential conductance and shot noise \Sec{thcond}
and \Sec{thshot}, respectively. Finally,
we give the summary in \Sec{thsum}.
%
\section{Methodology}
\label{thmeth}
\begin{figure}
\includegraphics[width=0.65\columnwidth]{fig1-model.png} %,clip=true
\caption{
Schematic diagram for the transport through the QD-wire hybrid ring system.
The QD is contacted by the two electron reservoirs under the bias voltage
($V=\mu_{\tL}-\mu_{\tR}$) with the tunneling rates $\Gamma_{\rm L}$ and $\Gamma_{\rm R}$.
It is further side coupled to the two Majora modes ($\gamma_1$ and $\gamma_2$)
at the two ends of a 1D topological superconductor nanowire
with the coupling coefficients $\lambda_1$ and $\lambda_2$, respectively.
$\Phi$ is the magnetic flux through the
loop.
}
\label{fig1}
\end{figure}
%
We investigate quantum transport through
a hybrid ring system comprising a quantum dot (QD) coupled to two Majorana
modes ($\gamma_1$ and $\gamma_2$), located at the ends of
a topological superconducting nanowire (TNW),
as illustrated in \Fig{fig1}.
%
The total Hamiltonian, $H_{\rm tot}=H_{\s}+H_{\B}+H_{\s \B}$, consists of three
components.
%%
The central QD-TNW hybrid system is described by the
low energy Hamiltonian \cite{Fle11090503,Liu11201308},
\begin{align}\label{HS0}
H_{\s}=\varepsilon_{\D}\hat d^\dg \hat d \!+\! \frac{i}{2} \varepsilon_{\M}\hat\gamma_{1}\hat\gamma_{2}
\!+\! (\lambda_{1}\hat d^\dg\hat\gamma_{1}\!+\! i\lambda_{2}\hat d^\dg\hat\gamma_{2}+{\rm H.c.}).
\end{align}
The first term represents the Hamiltonian of the QD,
characterized by a single fermionic mode with creation
(annihilation) operator $\hat d^\dg$ ($\hat d$)
and the electronic energy level $\varepsilon_{\D}$.
The second term describes the TNW with
a pair of Majorana
modes $\hat\gamma_1$ and $\hat\gamma_2$, and their coupling energy $\varepsilon_{\M}$,
which depends on the wire length.
%
The last term accounts for the coupling Hamiltonian between the QD and two Majorana modes,
with coupling coefficients $\lambda_{1}$ and $\lambda_{2}$.
%
Without loss of generality,
we set
$\lambda_{1}=\lambda $ to be real and $\lambda_{2}=\lambda e^{i\phi}$, assuming
symmetrical coupling strengths.
%
The phase difference between the two couplings ($\phi=\arg(\lambda_{2}/\lambda_{1})$),
is related to the magnetic flux $\Phi$ via $\phi=\Phi/\Phi_{0}$,
where $\Phi_{0} = h/e$.
%
%%%
By expressing the Majorana operators in terms of regular fermionic operators via
$\hat\gamma_{1}=\hat f +\hat f^\dg $ and $\hat\gamma_{2} = -i(\hat f -\hat f^\dg)$,
where $\hat f^\dg$ ($\hat f $) is the fermionic creation (annihilation) operator,
the Hamiltonian \Eq{HS0} can be rewritten as
\begin{align}\label{HS1}
H_{\s}&=\varepsilon_{\D}\hat d^\dg \hat d + \varepsilon_{\M}\hat f^\dg \hat f&\nl
&\!+ \!\lambda\big[(1\!+\!e^{i\phi})\hat d^\dg \hat f \!+\! {\rm H.c.}\big]
\!+\! \lambda\big[(1\! -\! e^{i\phi})\hat d^\dg \hat f^\dg \!+ \!{\rm H.c.}\big].
\end{align}
%
In this form,
the coupling between QD and MBSs consists of two distinct processes:
the normal tunneling (NT) term, given by
$\lambda\big[(1\!+\!e^{i\phi})\hat d^\dg \hat f \!+\! {\rm H.c.}\big]$,
and anomalous tunneling (AT) term, given by
$\lambda\big[(1\! -\! e^{i\phi})\hat d^\dg \hat f^\dg \!+ \!{\rm H.c.}\big]$.
%
These two tunneling mechanisms are associated with local and crossed Andreev reflection processes,
respectively, though a detailed analysis of these
is beyond the scope of this work.
%
Here, we focus on the distinct transport behavior arising from the competition between NT and AT,
which can be tuned by the magnetic flux phase $\phi$. The amplitude of these
processes are modulated by factors proportional to $\propto 1\! \pm\! e^{i\phi}$.
%
For clarity, we classify
the phase-dependent
behavior into three cases: (i) $\phi=2m\pi$,
where only one tunneling process is present,
either NT (ia) or AT (ib);
(ii) $\phi= (2m+1)\pi/2$, where NT and AT are equally weighted;
and (iii) other phase values, where NT and AT coexist with different amplitudes.
We will explore these transport characteristics through
the differential conductance and shot noise.
%%%
The bath Hamiltonian for the two electron reservoirs
is given by $H_{\B}=\sum_{\alpha}(\varepsilon_{\alpha k}-\mu_\alpha)
\hat c^\dg_{\alpha k}\hat c_{\alpha k}$, with
the applied bias voltage $V=\mu_{\tL}-\mu_{\tR}$ and $\alpha={ L,R}$.
%
The system-bath coupling Hamiltonian ($ H_{\s\B}$) describes the
standard electrons tunneling between the QD and electrodes:
\begin{align}\label{Hcoup}
H_{\s\B}&= \sum_{\alpha k}\left( t_{\alpha k} \hat d^\dg \hat c_{\alpha k}+{\rm H.c.}\right).
%{\rm H.c.}
\end{align}
%
The hybridization spectral function is assumed to be Lorentzian,
\be\label{Jw}
J_{\alpha }(\omega)
\equiv\pi\sum_k t_{\alpha k}t^\ast_{\alpha k}\delta(\omega-\varepsilon_{\alpha k})
=\frac{\Gamma_{\alpha}W^2}{\omega^2+W^2}.
\ee
%
Throughout this work, we adopt units of $e=\hbar=1$
for the electron charge and the Planck constant.
%%%%%%%%%%%%%%%%%%%
%%%
The current operator for the electron
transfer from $\alpha$-reservoir to
the hybrid system is
$\hat I_{\alpha}\equiv -\dot{\hat N}_{\alpha}
=i[\hat N_{\alpha},H_{\rm tot}] =-i \big(\hat d^{\dg} \hat F_{\alpha }
-\hat F^{\dg}_{\alpha } \hat d \big)$,
where $\hat N_{\alpha}=\sum_{k}\hat c^{\dg}_{\alpha k}\hat c_{\alpha k}$
is the electron number operator in
the $\alpha$-reservoir.
%
The mean steady--state transport current is denoted by $\bar I_{\alpha}
= {\rm Tr}_{\rm tot}\big(\hat I_{\alpha}\rho^{\rm st}_{\rm tot}\big)
\equiv \la \hat I_{\alpha}\ra$
and its fluctuation is given by,
\be\label{corr-curr}
\La \delta{\hat I}_\alpha(t)\delta{\hat I}_{\alpha'}(0)\Ra
=\La [{\hat I}_\alpha(t)-\bar I_{\alpha}]
[{\hat I}_{\alpha'}(0)-\bar I_{\alpha'}]\Ra.
\ee
%
The lead-specified current noise spectrum is the Fourier transformation,
\be\label{Sw_alp}
S_{\alpha\alpha'}(\omega)=\int_{-\infty}^{\infty} \!dt\,
e^{i\omega t} \La \delta{\hat I}_\alpha(t)\delta{\hat I}_{\alpha'}(0)\Ra.
\ee
%%
In contrast to the symmetrized version,
$S^{{\rm sym}}_{\alpha\alpha'}(\omega)
=S_{\alpha\alpha'}(\omega)
+S_{\alpha'\alpha}(-\omega)$,
%%
the asymmetric $S_{\alpha\alpha'}(\omega)$
is directly related to the absorption and emission processes
with $\omega>0$ and $<0$,
respectively
%%%
\cite{Eng04136602,Rot09075307,Yan14115411,
Bas10166801,Bas12046802,Del18041412}.
Here, we focus on the zero-frequency counterpart, referred to as shot noise,
especially the auto-correlation noise,
$S^{\rm sym}_{\alpha\alpha}(0)=2S_{\alpha\alpha}(0)$.
The Fano factor \cite{Bla001} is defined as $F_{\alpha}=S^{\rm sym}_{\alpha\alpha}(0)/2I=S_{\alpha\alpha}(0)/I$.
The value of this factor can be $F_{\alpha}>1$, $F_{\alpha}=1$, and $F_{\alpha}<1$,
describing super-Poissonian, Poissonian, and sub-Poissonian distributions, respectively.
%==============method===============
The model under consideration can be studied exactly using several well-established methods, including
the Landauer–Büttiker scattering
theory \cite{Dat95}, nonequilibrium Green function technique \cite{Hau08}, and
the dissipaton equation-of-motions (DEOM) approach \cite{Yan14054105,Jin15234108,Yan16110306}.
%
In this study, we employ the fermionic DEOM method
\cite{Yan14054105,Jin15234108,Yan16110306} to perform numerically accurate evaluations of
both the differential conductance
and the shot noise at finite temperature. A brief overview of the DEOM
methodology is provided in the \App{appdeom}.
\begin{figure*}
\centering
\includegraphics[width=0.680\textwidth,clip=true]{fig2.pdf}
%\centerline{\includegraphics*[width=1.0\columnwidth,angle=0]{fig2.eps}}
\caption{ Differential conductance $G$ as a function of bias voltage
tuned by the flux phase $\phi$
for different $\varepsilon_{\D}$ and $\varepsilon_{\M}$.
%The other parameters are $\Gamma=0.1$ and $k_{\B}T=0.1$.
%
The conductance $G$ exhibits
a period of $\pi$ when $\varepsilon_{\D}\varepsilon_{\M}=0$, whereas its period is $2\pi$
when
$\varepsilon_{\D}\varepsilon_{\M}\neq0$.
We classify the phase-dependent
behavior into four cases: (ia) $\phi=2m\pi$, (ib) $\phi=(2m+1)\pi$, (ii) $\phi= (2m+1)\pi/2$,
and (iii) other phase values.
\label{fig2}}
\end{figure*}
\section{Results}
\label{thresu}
\subsection{Hamiltonian eigenspectrum}
\label{thham}
Before presenting the results for differential conductance and shot noise,
we will first analyze the Hamiltonian eigenspectrum for the QD-wire hybrid system,
following a procedure similar to that described in our previous work \cite{Zha25023006}.
%
In the Fock states basis $\{|n_{\D}n_{\M}\ra=|01\ra,|10\ra,|00\ra,|11\ra\}$,
associated with
$ \hat n_{\D}= \hat d^\dg \hat d $
and $ \hat n_{\M}= \hat f^\dg \hat f $,
the Hamiltonian in \Eq{HS1} becomes block diagonal within the
subspaces of odd parity $\{ |01\ra, |10\ra \}$ (${ p}= -1$)
and even parity $\{|00\ra, |11\ra\}$ (${ p}= +1$),
\begin{align}\label{Hsm}
H_{\s}={
\left( \begin{array}{cc}
H_{-} & 0 \\
0 & H_{+}
\end{array}
\right)},
\end{align}
where
\bsube
\begin{align}
H_{-}={
\left( \begin{array}{cc}
\varepsilon_{\M} & \lambda(1+e^{-i\phi})\\
\lambda(1+e^{i\phi}) & \varepsilon_{\D}
\end{array}
\right)},
\end{align}
%
\begin{align}
\label{HevenU0}
H_{+}={
\left( \begin{array}{cc}
0 & \lambda(1-e^{-i\phi})\\
\lambda(1-e^{i\phi}) & \varepsilon_{\D}+\varepsilon_{\M}
\end{array}
\right)}.
\end{align}
\esube
%%%%%%%%%
They indicate the intrinsic coherent Rabi oscillations in the odd and even parity subspaces
induced by the electron NT and AT processes, respectively.
%
The corresponding parity-dependent Rabi frequency reads
\begin{align}\label{Delta}
\Delta_{\rm p}
&=\sqrt{(\varepsilon_{\D} + p\varepsilon_{\M})^{2}+8\lambda^{2} (1-p \cos\phi)}.
\end{align}
%
It is the energy difference between the two eigenstates of
$H_{\rm p}$ ($|e_{\rm p}\ra$ and $| g_{\rm p}\ra$),
i.e., $\Delta_{\rm p}= \varepsilon_{e_{\rm p}}- \varepsilon_{g_{\rm p}}$.
%
The eigen-energies are $\varepsilon_{e_{\rm p}}=\varepsilon_0+\frac{ \Delta_{\rm p}}{2}$
and $\varepsilon_{g_{\rm p}}=\varepsilon_0-\frac{ \Delta_{\rm p}}{2}$,
with $\varepsilon_0=\frac{1}{2}(\varepsilon_{\D}+\varepsilon_{\M})$.
Apparently, the Rabi frequency is periodically modulated by the threading magnetic flux
with $2\pi$ periodicity.
%
%%%
During the tunneling processes,
the conductance peaks
occur whenever the Fermi surface of the electrode is on resonance with the allowed transitions
between the ground state of the QD-wire system and
an excited state with opposite parity \cite{Zha25023006,Cla17201109}.
%
The corresponding peak positions are given by
\begin{align}\label{peaks}
E_{1} &=-E_4=\frac{1}{2}(\Delta_{+}+\Delta_{-}),~~
\nl
E_{2}&=-E_3=\frac{1}{2}(-\Delta_{+}+\Delta_{-}).
\end{align}
%%%
This leads to the zero-bias peak (ZBP) at $\mu_{\tL} = V/2 = E_2
= E_3 = 0$, which requires $\Delta_+=\Delta_-$. This condition leads to
energy-level degeneracy, as demonstrated in Ref.\,\cite{Liu11201308,Gon22122}.
In terms of \Eq{Delta},
the condition can be written as
%
\begin{align}
\label{zbpc}
\varepsilon_{\D}\varepsilon_{\M}=4\lambda^{2}\cos\phi.
\end{align}
Here, considering the Majorana zero-energy modes (MZMs with $\varepsilon_{\M}\rightarrow0$),
the ZBP appears at $\phi= (2m+1)\pi/2$ [case (ii)] where the NT and AT processes
contribute equally [c.f.\Eq{HS1}].
%
Consequently, the condition for the ZBP in \Eq{zbpc} reduces to
$\varepsilon_{\D}\varepsilon_{\M}=0$.
This explains why the ZBP remains robust against the
fluctuations in the QD energy level $\varepsilon_{\D}$ when
$\varepsilon_{\M}=0$, as shown in Ref.\,\cite{Ver14165314}.
%%%%
In the subsequent numerical calculations,
we consider a symmetrical-lead configuration
where $\mu_{\tL}=-\mu_{\tR}=V/2$ and $\Gamma_{\tL}=\Gamma_{\tR}= \Gamma/2$.
Consequently, we denote the differential conductance as
$G=G_{\rm L}=-G_{\rm R}=\frac{{\rm d}{\bar I}_{\rm L}}{{\rm d}V}$, the steady current as
$\bar I=\bar I_{\rm L}=-\bar I_{\rm R}$,
and the shot noise as $S= S_{\rm LL}(0)=S_{\rm RR}(0)$.
%
We focus on the scenario of weak coupling between the QD and electrodes
at finite temperature, i.e.,
$\Gamma\lesssim k_{\B}T$.
%
Although the master equation approaches based on second-order perturbation
theory can yield at least correct qualitative results in this regime \cite{Xu22064130},
we utilize the numerically accurate DEOM method to obtain precise outcomes,
particularly for the shot noise.
%
Unless stated otherwise,
we set $\Gamma= k_{\B}T=0.1\lambda$, use
$\lambda=1$ as the energy unit, and
assume the wide bandwidth limit $W=50
\Gamma\gg\Gamma$ for the electron reservoirs.
%
\subsection{The differential conductance}
\label{thcond}
Let us first examine the characteristic of the differential conductance.
As a function of bias voltage, the differential conductance exhibits
periodic modulation with respect to
the phase $\phi$, as numerically illustrated in \Fig{fig2}.
%
We observe that the periodicity of $G(V)$
is $\pi$ when $\varepsilon_{\D}\varepsilon_{\M}=0$ and $2\pi$
when
$\varepsilon_{\D}\varepsilon_{\M}\neq0$.
%
The underlying physical mechanism is fundamentally
rooted in the distinct tunneling processes
that occur between the QD and the wire.
These processes are precisely tuned by the flux,
as described by \Eq{HS1}, and can be further elucidated as follows.
%%%%
(ia) For $\phi=2m\pi$, only electron NT processes exist
between the QD and wire.
This leads to the intrinsic Rabi oscillation
via the electron channels associated with
$\varepsilon_{\D}$ and $\varepsilon_{\M}$
in the odd parity subspace (\{$|10\ra$ and $|01\ra$\}), characterized by
the Rabi frequency $\Delta_-$ [see \Eq{Delta}].
%
Meanwhile, $\Delta_+=\varepsilon_{\D}+\varepsilon_{\M}$
does not represent the Rabi frequency; rather, it denotes
the energy difference between the states $|11\ra$ and $|00\ra$.
%
Consequently, the two conductance peaks emerge at $E_{1}=E_{2}=\Delta_-/2=2\lambda$ and $E_3=E_4=
-\Delta_-/2=-2\lambda$, symmetrically positioned around zero-bias voltage
for $\varepsilon_{\D}=\varepsilon_{\M}=0$. Otherwise, these two peaks are split,
as displayed in \Fig{fig2}(a).
(ib) In contrast, for $\phi=(2m+1)\pi$,
only AT processes occur
between the QD and wire.
This results in the intrinsic Rabi oscillation
through the channels of
$\varepsilon_{\D}$ (electron) and $-\varepsilon_{\M}$ (hole)
in the even parity subspace (\{$|11\ra$ and $|00\ra$\}),
characterized by the Rabi frequency $\Delta_+$ [ see \Eq{Delta}].
%
Here, $\Delta_-=|\varepsilon_{\D}-\varepsilon_{\M}|$
is not the Rabi frequency but indicates the energy difference
between the states of $|10\ra$ and $|01\ra$.
%
Unlike case (ia), the two conductance peaks
are positioned at $E_1=E_3=\Delta_+/2$ and $E_2=E_4=-\Delta_+/2$,
symmetrically situated around zero-bias voltage
for $\varepsilon_{\D}=\varepsilon_{\M}$.
%
In other scenarios, these two peaks are also split, as indicated in \Fig{fig2} (e).
%%%%
Notably, the models in cases (ia) and (ib) can be interpreted as two
coupled QDs (QD-QD) \cite{Zen17127302}.
Specifically, the QD with energy $\varepsilon_{\D}$ is coupled to
another QD with energies $\varepsilon_{\M}$ and $-\varepsilon_{\M}$
for case(ia) and case(ib), respectively.
As a result, the characteristics of the differential conductance
exhibit similarities, and they are even identical for $\varepsilon_{\M}=0$.
%%%
(ii) Of particular interest is the case where $\phi= (2m+1)\pi/2$, in which
the coherent coupling between the QD and wire involves both
electron NT and AT processes with equal weights.
Rabi oscillations occur in both the odd and even parity subspaces, as indicated
in cases (ia) and (ib), respectively.
%
The ZBP appears at $V=E_2 = E_3 = 0$ in the differential conductance,
as displayed in \Fig{fig2} (c) and (g).
The height of ZBP is nearly twice that of the other two peaks due to the overlap of
two peaks at zero-bias voltage.
%
This characteristic is analogous to that a QD coupled with
a single Majorana mode \cite{Zha25023006}.
Indeed, using the definition
$\hat\gamma_{12}=\hat\gamma_1\pm\hat\gamma_2$ \cite{Fle11090503},
the Hamiltonian given in \Eq{HS1} can be rewritten as
\begin{align}\label{HS2}
H_{\s}=\varepsilon_{\D}\hat d^\dg \hat d
\!+\! \lambda(\hat d^\dg\hat\gamma_{12}+{\rm H.c.}),
\end{align}
where we assume $\varepsilon_{\M}=0$ for simplicity.
(iii) For other values of phase,
both the electron NT and AT processes are also involved,
but with different weights.
The ZBP splits into two peaks at $E_2=-E_3$, symmetrically positioned around zero-bias voltage.
The resulting differential conductance exhibits four peaks, as illustrated in \Fig{fig2} (b), (d), (f) and (h).
%
This feature resembles the scenario where the QD is coupled with the quasi-MZMs
as demonstrated in Ref.\,\onlinecite{Zha25023006}.
From the above analysis, we find that the characteristic of $G(V)$ changing with the phase
is analogous to the variation of $G(V)$ with the topological quality factor $q$, as
studied in Ref.\,\onlinecite{Zha25023006}. Here,
we focus on the periodic behavior of the conductance.
%
Under the symmetrical bias voltage condition ($\mu_{\tL}=-\mu_{\tR}=V/2$),
the electron channel ($\varepsilon_{\M}$) and hole channel (-$\varepsilon_{\M}$) exhibit
symmetry around $\varepsilon_{\D}=0$.
As a result, the differential conductance features in both cases (ia) and (ib) are
identical when $\varepsilon_{\D}\varepsilon_{\M}=0$ (either $\varepsilon_{\D}=0$ or $\varepsilon_{\M}=0$),
leading to a $\pi$-periodic $G(V)$.
%
In contrast, when $\varepsilon_{\D}\varepsilon_{\M}\neq0$ (both $\varepsilon_{\D}\neq0$
and $\varepsilon_{\M}\neq0$)
the symmetry between the two channels is broken, resulting in a
$2\pi$-periodic differential conductance.
%
Thus, we conclude that whether
the period of the differential conductance,
is $\pi$ or $2\pi$, is determined by $\varepsilon_{\D}\varepsilon_{\M}=0$
or $\varepsilon_{\D}\varepsilon_{\M}\neq0$.
This result extends the earlier qualitative
picture \cite{Zen17127302,Mat22155409,Bah25045920} to a quantitative one.
%%%
We would like to note that the width of the peaks arising from the measurements is determined by
the temperature in the considered weak-coupling regime, $\Gamma\lesssim k_{\B}T$.
Whereas the height of the peaks is affected by both the coupling strength
and temperature.
This behavior can be directly understood from the analytical expression for differential conductance,
as derived through a second-order perturbation quantum master equation (QME) approach \cite{Xu22064130},
%
consistent with the methodology presented in
Ref.\,\cite{Zha25023006}.
%
Taking $\varepsilon_{\D}=\varepsilon_{\M}=0$ as an example,
we have
\begin{align}\label{G00}
G(V)&=\frac{\Gamma \beta}{2^6}
\Big({\rm sech}^2[\beta(V/4-E_1)]+{\rm sech}^2[(V/4+E_1)]
\nl&
\!+\!{\rm sech}^2[\beta(V/4-E_2)]\!+\!{\rm sech}^2[\beta(V/4+E_2)]\Big).
\end{align}
Here $E_1$ and $E_2$ given by \Eq{peaks} are reduced to
\bsube\label{peak0}
\begin{align}
E_1=-E_4=2\lambda\big(\cos\frac{\phi}{2}+\sin\frac{\phi}{2}\big),
\\
E_2=-E_3=2\lambda\big(\cos\frac{\phi}{2}-\sin\frac{\phi}{2}\big).
\end{align}
\esube
%
In terms of \Eq{G00}, one can easily get the corresponding half-width of the peaks as
$\Gamma_d=2\ln(1+\sqrt{2})k_{\B}T\approx1.763 k_{\B}T$.
This explains why the ZBP exists at $\phi= (2m+1)\pi/2$
even for $\varepsilon_{\D}\varepsilon_{\M}\neq0$,
as long as $|\Delta_+-\Delta_-|<2\Gamma_d$.
%
Furthermore, due to the temperature-dependent width, the ZBP
cannot serve as a reliable hallmark for the existence
of MBSs in SC wires, as demonstrated in Ref.\,\onlinecite{Zha25023006}.
%
Notably, the ZBP is quite robust against the energy
fluctuations of the electron level in the dot ($\varepsilon_{\D}$)
for ideal MZMs ($\varepsilon_{\M}=0$).
%
%
In practical experiments situations, it is
preferable to tune the energy level of the dot to $\varepsilon_{\D}\rightarrow0$,
as the coupling energy ($\varepsilon_{\M}$) of Majorana modes is
determined by the length of the wire.
%%
Note that the comparison of numerical results of the differential conductance
based on QME approach given by \Eq{G00} and DEOM method
is displayed in \Fig{appfig1} of \App{appdeom}.
%
The QME method captures the essential features of
of the differential conductance, at least qualitatively well when
$\Gamma\lesssim k_{\B}T$.
The accuracy of the QME results improves as
the coupling strength decreases relative to the temperature.
%%%%%%%%
\begin{figure*}
\centering
\includegraphics[width=0.70\textwidth,clip=true]{fig3.pdf}
%\centerline{\includegraphics*[width=1.0\columnwidth,angle=0]{fig2.eps}}
\caption{ Physical quantities as functions of the phase $\phi$.
(a) Resonance energy positions $E_j$ calculated using \Eq{G00}.
(b) Shot noise (solid line) and steady-state current (dashed line)
under different bias voltages.
(c) Fano factor $F=S/\bar I$.
(d) Peak characteristics of the shot noise (solid line) and steady-state current (dashed line)
at low bias voltages.
\label{fig3}}
\end{figure*}
\subsection{Shot noise}
\label{thshot}
%%%
As well-known, the current noise spectrum
contains the information much beyond the average current or conductance \cite{Her927061,Bla001,Imr02,Naz03}.
%
The zero--frequency noise, referred to as shot noise, primarily describes the
statistical behavior of the steady state.
It has been utilized to measure the effective carrier charge, which can be
either fraction \cite{Pic97162,Rez99238,Bid09236802} or integer
\cite{Koz003398,Lef03067002}.
%
The finite-frequency noise spectrum
contains the rich correlated dynamics information
\cite{Ent07193308,Eng04136602,Jin11053704,Jin13025044,Rot09075307,Jin20235144}.
%
For the system studied here, the noise spectrum would exhibit
coherent Rabi dip signals at $\omega\approx\pm\Delta_{\rm p}$.
Similar to the differential conductance, the characteristics of the noise spectrum
are periodic with the phase, with one Rabi dip at $\Delta_-$ for case(ia) and at $\Delta_+$ for case(ib).
In case (ii), Rabi dips degeneracy (RDD) occur at $\Delta_+=\Delta_-$,
while, in case (iii),
the RDD splits into two.
This Rabi dip feature is quite similar to that demonstrated in Ref.\,\onlinecite{Zha25023006},
where the noise spectrum sensitive to the topological quality factor has been studied.
%
Here, we do not illustrate the finite-frequency noise spectrum, but instead focus
on the characteristics of the zero-frequency shot noise.
Actually, the periodic behavior of the differential conductance
demonstrated above is also applicable
for the current noise characteristics.
%
Since we focus on the underlying physical mechanism tuned by the flux,
we use $\varepsilon_{\D}=\varepsilon_{\M}=0$
as an illustrative example in the subsequent discussions on the shot noise.
In this example, the characteristics of the cases (ia) and (ib) are identical.
Consequently, it is unnecessary to distinguish between them,
as both will be treated the same as case(i).
The corresponding peak positions ($E_j$) describing the resonance transitions in the system
are given by \Eq{peak0}.
%
For the convenience of discussions, we plot
$E_j$
which change periodically with
the phase ($\phi$), as shown in \Fig{fig3} (a).
%
Specifically, two, three, and four peaks emerge for case (i) with $\phi=m\pi$,
case (ii) with $\phi=(2m+1)\pi/2$ , and the general case (iii), respectively.
The detailed characteristics of these conductance peaks have been demonstrated above.
%
Here, we denote $V_{\rm c}$ as a critical value of the bias voltage,
at which
the Fermi surface ($\mu_{\tL,\tR}=\pm\frac{V}{2}$) is in resonance
with all the energy transitions for case (i) ($\phi = n\pi$),
as indicated by the vertical dotted-line in \Fig{fig3} (a).
For the parameters considered here, the critical bias voltage
is $|V_{\rm c}|=4$.
%%%%
Figure \ref{fig3} (b) displays that
the shot noise ($S$) follows the average current ($\bar I$) and
is periodic with the phase $\phi$, exhibiting a periodicity
of $\pi$.
%
However, their detailed characteristics depending on the phase
is different.
This is because the generated current is only related to the opening of
the transport channels, whereas
the shot noise is influenced by both the transport channels and
charge statistical of the quasiparticles involved in the transport processes.
%
We elucidate their characteristics below.
%
%%%
For $|V|\geq V_{\rm c}$, which includes the two resonance transitions at $E_2$ and $E_3$,
the steady-state current and its shot noise exhibit a bowl-shaped profile.
As the voltage increases, the bottom of the bowl gradually rises
and eventually flattens.
%
This behavior is due to
the occurrence of two additional
resonance transitions at $E_{1,4}$ for case (iii) as the bias voltage increases,
contributing to the edge features of the bowl in \Fig{fig3} (b).
%
When the voltage reaches around $|V| = 6$, the other two resonances at $E_{1,4}$ for case (ii) also
become relevant. As the voltage continues to increase,
a high-voltage scenario ($|V/2|\gg |E_j|$) is reached (for instance, $V=8$) where
involves all the resonance transitions,
%
the resulting steady current $\bar I$ reaches a saturated
value regardless of the phase.
%
%%%
However, the shot noise displays the dips at $\phi=n\pi$ for case (i)
in the large bias voltage regime ($|V/2|\gg |E_j|$).
%
The underlying reason is that case (i) involves
only one tunneling processes, either NT or AT, where
the Pauli principle comes into play during the transport process.
%
This lead to a sub-Poissonian distribution with the Fano factor $F=0.5$,
as displayed in \Fig{fig3} (c).
%
In contrast, in both cases (ii) and (iii),
the shot noise display a Poissonian distribution ($F=1$),
indicating that the quasiparticles in the transport are bosonic
due to the involvement of both NT and AT processes.
%
It is noteworthy that
when the two resonance transitions at $E_{1}$ and $E_{4}$ occur,
the value of shot noise is always slightly greater than the current value,
as shown at the edges of the bowl in \Fig{fig3}(b). This results in a super-Poissonian distribution as
illustrated in \Fig{fig3} (c), indicating the occurrence of bounching behavior at the resonances.
This behavior is enhanced with the decrease of the bias voltage.
%
When $|V|< V_{\rm c}=4$, the dual resonances at $E_{1}$ and $E_4$ are fully quenched,
resulting in an inversion of the profiles of the steady current and its shot noise,
forming an anti-bowl configuration.
%
For case (i) with $\phi=m\pi$ in which all transport channels are outside the
bias window, both the steady current and the shot noise approach
zero at low temperatures.
In this scenario, the shot noise displays Poissonian
distribution with $F=1$ for finite temperature, which originates solely
from the thermal fluctuations.
%
As a function of the phase,
the edge features of the bowl correspond to the emergence of the
remaining resonance transitions at $E_{2}$ and $E_3$ for case (iii).
Similarly, at these resonance transitions, the value of the shot noise is
slightly greater than the current value, resulting in a Fano factor $F>1$.
%
These two resonance transitions at $E_{2}$ and $E_3$ are gradually quenched
as the bias voltage decreases. Simultaneously,
the current and shot noise drop to zero outside the bowl's rim,
and the bottom of the bowl progressively narrows.
%
At $V=0$ with no transport channels open,
the steady-state current is zero regardless of the phase.
%
Counterintuitively, the shot noise exhibits
a peak at $\phi=(2m+1)\pi/2$ for the balance between NT and AT. This peak arises due to
the occurrence of
the two resonance transitions at $V=E_2=E_3=0$,
indicating that a giant Fano factor occurs for case (ii) at zero bias voltage.
%
To clearly observe this phenomenon, we specially plot \Fig{fig3} (d)
for very low bias voltage regime, i.e., $V\leq1$.
We find that as long as $V>0$, the current increases rapidly, while the shot noise grows slowly.
When the bias reaches $V=1\gg k_BT=0.1$, indicating the onset of far-from-equilibrium situation,
the values of the current and shot noise become nearly consistent,
exhibiting a Poisson distribution.
%
We further plot the Fano factor as a function of the bias voltage for case (ii)
in the inset of \Fig{fig3} (d). It shows that the Fano factor
decreases rapidly from a giant value to $F=1$ with the increase in the bias voltage.
%
This Giant value of Fano factor at small voltages
may be used as an additional signature for MBSs.
%
However, in actual measurements, the peak at $\phi=(2m+1)\pi/2$ of
the shot noise still has a certain broadening.
This broadening is primarily determined by the temperature for
the considered weak-coupling regime.
%%%%%%%%%%%%%%%
It should be noted that
within the QME approach,
a compact analytic expression analogous to \Eq{G00} for
the differential conductance cannot be derived for the shot noise.
This stems from the significantly greater complexity involved in evaluating shot
noise compared to
the steady-state current.
%
Specifically, the steady-current is computed using
\Eq{Acurr}, which depends only on the steady state solution for
$\rho^{(1)\rm st}_{j}$. In contrast, the shot noise is evaluated via
\Eq{curr-curr}, whose solution requires ont only the steady state values from \Eq{initial_final},
but also the time-dependent evolution of $ {\rho}^{(1)}_j(t;\hat I_{\alpha'})$.
%
Moreover, an accurate discussion of the Fano factor ($S(0)/I$)
demands high numerical precision such as from DEOM theory rather than
merely qualitative accuracy as obtained by QME approach.
\section{Summary}
\label{thsum}
In summary, we have investigated the quantum transport characteristics of a hybrid ring system
comprising a QD coupled to MBSs in a topological
superconducting nanowire, with a magnetic flux threading the loop.
Using the numerically accurate dissipaton equations-of-motion (DEOM) method,
we identified distinct magnetic-flux-dependent transport signatures.
%%%%
The differential conductance exhibits periodic behavior with respect to the magnetic flux phase,
with $\pi$-periodicity when $\varepsilon_{\D}\varepsilon_{\M}=0$ and $2\pi$
otherwise.
This modulation arises from the
interplay between normal tunneling (NT) and anomalous tunneling (AT),
controlled by the magnetic flux.
%
A prominent zero-bias peak (ZBP) appears at
$\phi=(2m+1)\pi/2$ under the condition $\varepsilon_{\D}\varepsilon_{\M}=0$.
This is the typically MBS-contributed feature, coming from
the balance between the NT and AT processes.
This ZBP remains robust against fluctuation the QD energy level when $\varepsilon_{\M}=0$,
although its width is thermally broadened, limiting its utility as
a definitive signature of Majorana states.
%%%%
Similarly,
the shot noise demonstrates phase-dependent periodic behavior.
Beyond this feature,
the shot noise
provides insights into the statistical behavior of the transport processes.
%
At high bias voltages ($|V|\geq V_{\rm c}$),
the shot noise exhibiting a bowl-shaped profile displays three voltage-dependent statistical regimes:
(1) sub-Poissonian (Fano factor $F = 0.5$)
for the presence of only one tunneling process (either NT or AT),
(2) Poissonian ($F = 1$), for the coexistence NT and AT processes,
and (3) super-Poissonian ($F > 1$) at resonance edges, indicating quasiparticle bounching effects.
%%%
At low bias ($|V|<V_{\rm c}$), the quenching of resonances inverts the profiles
into an anti-bowl shape. Notably, a giant Fano factor emerges at zero bias ($V = 0$)
under the resonant condition ($E_2=E_3=0$) for
the balance point.
This anomaly rapidly diminishes with increasing voltage,
converging to Poissonian statistics ($F = 1$)
for $V\gg k_BT$.
%
The giant Fano factor at low bias may serve as a supplementary signature of MBS,
though thermal effects challenge its experimental resolution.
%
Our results provide a detailed and quantitative
characterization of the interplay between MBSs, magnetic flux,
and non-equilibrium transport in the weak-coupling regime.
%
\acknowledgments
We acknowledge helpful discussions with Professor YiJing Yan and Professor Xin-Qi Li.
\appendix
\section{
DEOM theory }
\label{appdeom}
\begin{center}
{\bf A.1 The DEOM construction}
\end{center}
In this Appendix, we briefly introduce the DEOM method, for the details, refer to
Ref.\,\onlinecite{Yan14054105,Jin15234108,Yan16110306}.
We then provide the numerical comparison between exact DEOM and QME approaches
for weak system-reservoir coupling \cite{Xu22064130}.
The system-reservoir coupling
is generally given by
\be\label{Hsb1}
H_{\SB}\!=\!\sum_{\alpha u }\left(t_{\alpha u k} \hat a^{\dg}_{u} c_{\alpha k}
+ {\rm H.c.} \right)\!\!=\!\sum_{\alpha u \sigma}\hat a^{\bar\sigma}_{u}
\wti F^{\sigma}_{\alpha u},
\ee
which accounts for multi-modes in the system (labeled by $u$) compared to the single-mode considered in \Eq{Hcoup}.
%
Here $ \hat F^-_{\alpha u}
=\sum_k t_{\alpha u k} c_{\alpha k}=(\hat F^+_{\alpha u})^\dg$
and
$\wti F^{\sigma}_{\alpha u} \equiv \bar\sigma \hat F^{\sigma}_{\alpha u}$,
%\ee
with $\sigma =+,-$ (and $\bar\sigma$ is its opposite sign) %$\bar\sigma = -\sigma$
for identifying the creation/annihilation operators.
For the dissipaton description of bath interaction, \cite{Jin15234108,Yan14054105}
the bath correlation function is decomposed as
\be\label{FF_corr}
c^{(\sigma)}_{\alpha uv }(t-\tau)\!\equiv\!\big\la \hat F^{\sigma}_{\alpha u}(t)
\hat F^{\bar\sigma}_{\alpha v}(0)\big\ra_{\B}
\!=\! \sum_{m=1}^{M} \eta^{\sigma}_{\alpha uv m}e^{-\gamma^{\sigma}_{\alpha m}t},
\ee
%
using the optimal Pad\'{e} spectrum decomposition
of the Fermi function \cite{Hu10101106,Hu11244106}.
Here $\hat F^{\sigma}_{\alpha u}(t)\equiv
e^{i H_{\B}t}\hat F^{\sigma}_{\alpha u}e^{-i H_{\B}t}$
and $\la \cdots \ra_{\B}$ stands for the statical average over
the electrodes reservoirs in
local thermal equilibrium state.
%
%%
Correspondingly, the bath operators are decomposed into quasi--particle (dissipatons),
\be\label{wtiF_f}
\wti F^{\sigma}_{\alpha u} \equiv -\sigma \hat F^{\sigma}_{\alpha u}
\equiv \sum_{m=1}^{M} \hat f^{\sigma}_{\alpha u m},
\ee
%
which satisfy
\be\label{ff_corr}
\big\la\hat f^{\sigma}_{\alpha u m}(t)\hat f^{\sigma'}_{\alpha' v m'}(0)\big\ra_{\B}
=-\delta_{\sigma\bar\sigma'}\delta_{\alpha\alpha'}\delta_{mm'}\,
\eta^{\sigma}_{\alpha u v m}\, e^{-\gamma^{\sigma}_{\alpha m} t}.
\ee
%%%
Using composite indices
$j\equiv(\sigma\alpha u m)$ and $\bar j\equiv(\bar\sigma\alpha u m)$,
so that $f_j\equiv f^{\sigma}_{\alpha u m}$, etc.,
%
dynamical variables in DEOM are the reduced dissipaton density
operators (DDOs), which
are defined as,
\be\label{DDO_def}
\rho^{(n)}_{\bf j}(t)\equiv \rho^{(n)}_{j_1\cdots j_n}(t)\equiv
{\rm tr}_{\B}\Big[\big(\hat f_{j_n}\cdots\hat f_{j_1}\big)^{\circ}
\rho_{\rm tot}(t)\Big]\, ,
\ee
where the product of dissipatons inside the circled parentheses, $(\,\cdots)^{\circ}$,
is \emph{irreducible}.
%
Exchanging any two fermionic dissipatons
introduces a minus sign, e.g.,
%%
$\big(\hat f_{j}\hat f_{j'}\big)^{\circ}=-\big(\hat f_{j'}\hat f_{j}\big)^{\circ}$.
%
Moreover, the irreducible notation is related to the
generalized Wick's theorem, such as \cite{Yan14054105,Jin15234108,Yan16110306},
\begin{align}\label{Wick}
&\quad\, \text{tr}_{\B}\left[\big(\hat f_{j_n}\!\cdots\!\hat f_{j_1}\big)^{\circ}
\hat f_j\rho_{\rm tot}(t)\right] \nl
&=\rho^{(n+1)}_{j{\bf j}}
+ \sum_{r=1}^n (-)^{r-1} \La\hat f_{j_r}\hat f_j\Ra_{\B}
\rho^{(n-1)}_{{\bf j}^{-}_r} .
\end{align}
Followed
by using the above dissipaton algebra,
one can perform the time derivative on
the reduced system density operator $\rho(t)\equiv{\rm tr}_{\B}[\rho_{\rm tot }(t)]$
with
$\dot{\rho}_{\rm tot}(t)=-i[H_{\tS}+H_{\B}+H_{\SB},{\rho}_{\rm tot}(t)]$.
Consequently, the DEOM is obtained as
\cite{Yan14054105,Jin15234108,Yan16110306}
%
\begin{align}\label{DEOM}
\dot\rho^{(n)}_{\bf j}(t)&=-\bigg(i{\cal L}_{\tS}
+\sum_{r=1}^n \gamma_{j_r}\bigg)\rho^{(n)}_{\bf j}(t)
-i\sum_{j} {\cal A}_{\bar j}\rho^{(n+1)}_{{\bf j}j}(t) \nl
&\quad
-i \sum_{r=1}^n (-)^{n-r}{\cal C}_{j_r}\rho^{(n-1)}_{{\bf j}^-_r}(t),
\end{align}
which is the HEOM formalism \cite{Jin08234703}. Here,
${\cal L}_{\tS}\,(\cdot)=[H_{\tS},\,(\cdot)]$
and the superoperators, ${\cal A}_{\bar j}\equiv {\cal A}^{\bar\sigma}_{\alpha u m} = {\cal A}^{\bar\sigma}_{u}$
and ${\cal C}_{j}\equiv {\cal C}^{\sigma}_{\alpha u m}$
are defined via
%
\be\label{calAC}
\begin{split}
{\cal A}^{\sigma}_{u} \Opm &\equiv
a^{\sigma}_{ u}\Opm \pm \Opm \hat a^{\sigma}_{u}
\equiv \big[\hat a^{\sigma}_{u},\Opm\big]_\pm \, ,
\\
{\cal C}^{\sigma}_{\alpha u m} \Opm &\equiv
\sum_{v} \big(\eta^{\sigma}_{\alpha uv m}\hat a^{\sigma}_{v}\Opm
%%%
\mp \eta^{\bar \sigma\,{\ast}}_{\alpha uv m}\Opm \hat a^{\sigma}_{m}\big),
\end{split}
\ee
%
where, $\Opm$ is an arbitrary operator,
with even ($+$) or odd ($-$) fermionic parity,
such as $\rho^{(2k)}$ or $\rho^{(2k+1)}$, respectively.
%
The reduced system density operator is just
$\rho(t) \equiv {\rm tr}_{\B}\rho_{\rm tot}(t)= \rho^{(0)}(t)$.
%
All $\big\{\rho^{(n\geq1)}_{\bf j}\big\}$
are physically well--defined DDOs, as depicted in \Eq{DDO_def},
capturing entangled system--bath dynamics.
%
It is important to emphasize that
the DEOM method offers a unified framework for studying both noninteracting
and interacting open quantum systems, as characterized by
the system Hamiltonian $H_{\tS}$ in the first term of \Eq{DEOM}.
%
For noninteracting systems, as considered here,
DEOM evaluations converge exactly
at tier level $L=2$.
Indeed, we have previously derived an exact equation for the
single-particle density matrix,
$\varrho_{uv}={\rm tr}_{\s}[a^\dg_v a_u\rho(t)]$,
by truncating the hierarchy at the second tier($L=2$),
and obtained an exact
current expression \cite{Jin08234703}
that agrees with the results from Landauer–Büttiker scattering theory
\cite{Dat95} and nonequilibrium Green function method \cite{Hau08}.
%
For interacting systems,
DEOM exhibits rapid and uniform convergence to
the exact outcomes, as the truncated tier level $L$ (where $\rho^{(n>L)}_{\bf j}=0$)
increases \cite{Zhe121129,Li12266403,%
Zhe13086601,Hou15104112,Ye16608}. The minimal tier $L$ required for
convergence depends on the specific configurations
of both the system and the bath.
%
Furthermore, the DEOM, \Eq{DEOM} forms a set of linear
differential equations, enabling the use of standard quantum dynamics pictures in
the DEOM-space descriptions of open quantum systems \cite{Yan16110306}.
%%%
This foundational structure supports consistent
DEOM-based evaluations of various physical quantities, including transport current
and its correlation function, as summarized below.
%
\begin{center}
{\bf A.2 Transport current and its correlation function}
\end{center}
The transient current though the $\alpha$-lead is calculated via
$I_{\alpha}(t) = {\rm Tr}[\hat I_{\alpha}\rho_{\rm tot}(t)]$. Substituting the current operator
\be
\hat I_{\alpha}=-\dot{\hat N}_{\alpha}
=i[\hat N_{\alpha},H_{\rm tot}]
=-i\sum_u \big(\hat a^{+}_u \hat F^-_{\alpha u}
-\hat F^{+}_{\alpha u}\hat a^-_u \big),
\ee
one directly obtains
\be\label{Acurr}
I_{\alpha}(t) = -i\! \sum_{j_{\alpha}\in j}
{\rm tr}_{\tS}\!\big[\ti a_{\bar j}\rho^{(1)}_{j}(t)\big],
\ee
where $\ti a_{\bar j}\equiv \ti a^{\bar \sigma}_{\alpha u k}
=\bar \sigma\hat a^{\bar \sigma}_{u}$
and $j_{\alpha}\equiv \{ \sigma u k\}\in j\equiv\{\sigma\alpha u k\}$.
%
Clearly, the transport current is directly related to the first-tier DDO.
%
Similarly,the lead-specific current correlation function
can be written as
%
\begin{align}\label{curr-curr}
% C_{\alpha\alpha'}(t) =
\La\hat I_{\alpha}(t)\hat I_{\alpha'}(0)\Ra
&=
{\rm Tr}\big[\hat I_{\alpha}\rho_{\rm tot}(t;\hat I_{\alpha'})\big]
\nl&=- i\!\sum_{j_{\alpha}\in j}
{\rm tr}_{\tS} \big[\ti a_{\bar j}
{\rho}^{(1)}_j(t;\hat I_{\alpha'})\big],
\end{align}
which also depends on the first-tier DDOs. The key difference
lies in the initial condition: $\rho_{\rm tot}(0;\alpha')=\hat I_{\alpha'}\rho^{\rm st}_{\rm tot}$,
where $\rho^{\rm st}_{\rm tot}$ is the steady state of the whole system.
%
Using dissipaton-algebra, the initial
DDOs are given by
\begin{align}\label{initial_final}
\rho^{(n)}_{\bm j}(0; \alpha')
&\!=\! -i\!\!\!\sum_{j'_{\alpha'}\in j'}{\ti a}_{\bar j'} \rho^{(n+1);{\rm st}}_{{\bm j}j'}
- i\!\sum_{r=1}^n(-)^{n-r} \wti C_{j_r}\rho^{(n-1);{\rm st}}_{{\bm j}^{-}_r},
\end{align}
Further details can be found in Refs. \cite{Yan14054105,Jin15234108,Yan16110306},
along with numerical applications \cite{Jin20235144,Mao21014104}.
\begin{center}
{\bf A.3 Weak system-reservoir coupling}
\end{center}
\begin{figure}
\includegraphics[width=1.0\columnwidth,angle=0]{appfig1.pdf}
\caption{
Differential conductance $G$ as a function of bias voltage
for $\phi=0.5\pi$ and
$\varepsilon_{\D}=\varepsilon_{\M}=0$.
The evaluations are based on
QME approach (black solid-line) given by the analytical expression
\Eq{G00} and
exact DEOM theory (red dash-line)
with
(a) $\Gamma= k_{\B}T=0.1$,
(b) $\Gamma=0.1$ and $k_{\B}T=0.2$,
(c) $\Gamma=0.05$ and $k_{\B}T=0.1$, and
(d) $\Gamma=0.05$ and $k_{\B}T=0.2$. }
\label{appfig1}
\end{figure}
For the case of weak QD-reservoir coupling[c.f.\Eq{Hsb1}],
the hierarchy is truncated at the lowest tier, i.e.,
\be
\rho^{1}_j(t)
=\rho^{\sigma}_{\alpha u m}(t)\equiv
{\rm tr}_{\B}\big[\hat f^{\sigma}_{\alpha u m}\rho_{\rm tot}(t)\big].
\ee
The DEOM in \Eq{DEOM} then reduces to
\bsube\label{TL-EOM}
\begin{align}
&\dot\rho(t) = -i{\cal L}_{\tS}\rho(t)
- i\sum_{\alpha u\sigma m}
\big[ \hat a^{\bar\sigma}_u,\rho^{\sigma}_{\alpha u m}(t)\big],
\label{rho0t}
\\
&\dot\rho^{\sigma}_{\alpha u m}(t)
=\!-i({\cal L}_{\tS}\!+\!\gamma^\sigma_{\alpha m})\rho^{\sigma}_{\alpha u m}(t)
-i{\cal C}^{(\sigma)}_{\alpha u m} \rho(t).
\label{rho1t}
\end{align}
\esube
After straightforward algebra, we obtain
\be\label{rho1}
\sum_m\rho^{\sigma}_{\alpha u m}(t)
= -i\!\int_{t_0}^t\!\!{\mathrm d}\tau\, e^{-i{\cal L}_{\tS} (t-\tau)}
{\cal C}^{(\sigma)}_{\alpha u}(t-\tau) \rho(\tau),
\ee
where
\be\label{Ct}
{\cal C}^{(\sigma)}_{\alpha u }(t) \bullet
\equiv \sum_v\big[c^{(\sigma)}_{\alpha uv}(t)\hat a^\sigma_v \bullet
- c^{(\bar\sigma)\ast}_{\alpha uv}(t)\bullet \hat a^{\sigma}_v \big],
\ee
with $c^{(\sigma)}_{\alpha uv }(t) $
being the bath correlation function defined in [\Eq{FF_corr}].
%
This leads directly to the well-known
second-order perturbation quantum master equation (QME) with memory effects
\cite{Jin11053704,Yan05187,Mak01357,Jin16083038,Xu22064130},
%%%%%%%%%
\be\label{QME}
\dot\rho(t)\! =\! -i{\cal L}_{\s}\rho(t)\!-\!\sum_{\alpha \sigma u }
\!\int_{t_0}^t\!\!{\mathrm d}\tau \big[\hat a^{\bar\sigma u},
%\Pi_0(t,\tau)
{\cal C}^{(\sigma)}_{\alpha u }({\cal L}_{\s},t-\tau) \rho(\tau) \big],
\ee
where $
{\cal C}^{(\sigma)}_{\alpha u }({\cal L}_{\s} ,t)
\equiv \sum_v e^{-i{\cal L}_{\s} t} {\cal C}^{(\sigma)}_{\alpha u v}(t)$.
%%%%
%
Accordingly, the current expression given in \Eq{Acurr} simplifies to
\cite{Jin11053704,Shi16095002,Xu22064130}
\be\label{curr-exp}
I_{\alpha}(t)
=-\!\sum_{\sigma u} \!\!\int_{t_0}^t\!\! {\mathrm d}\tau\, {\rm tr}_{\rm s}
[ \sigma \hat a^{\sigma u}{\cal C}^{(\bar\sigma)}_{\alpha u}({\cal L}_{\s} ,t-\tau) \rho(\tau) ].
\ee
The stationary current is obtained by taking $t\rightarrow\infty$ and $\rho(\tau)\rightarrow\rho^{\rm st}$.
%
For the current correlation function, the initial condition in \Eq{curr-curr} must be accounted for.
%
The corresponding expression for the current correlation and its spectrum
can be derived similarly, and the details can be found in
Ref.\,\onlinecite{Xu22064130}.
%%%
Figure \ref{appfig1} shows the differential conductance $G$
computed using
the QME method (black solid-line) and the exact DEOM theory (red dashed-line).
%
Clearly, the QME method captures well
all the basic features of $G$ at least qualitatively for $\Gamma\lesssim k_{\B}T$.
The smaller the coupling strength is compared to the temperature,
the more accurate the QME result becomes.
\begin{thebibliography}{10}
\bibitem{Kit01131}
A.~Y. Kitaev,
\newblock Physics-Uspekhi {\bf 44}, 131 (2001).
\bibitem{Kit032}
A.~Kitaev,
\newblock Annals of Physics {\bf 303}, 2 (2003).
\bibitem{Nay081083}
C.~Nayak, S.~H. Simon, A.~Stern, M.~Freedman, and S.~Das~Sarma,
\newblock Rev. Mod. Phys. {\bf 80}, 1083 (2008).
\bibitem{Ali10125318}
J.~Alicea,
\newblock Phys. Rev. B {\bf 81}, 125318 (2010).
\bibitem{Lei11210502}
M.~Leijnse and K.~Flensberg,
\newblock Phys. Rev. Lett. {\bf 107}, 210502 (2011).
\bibitem{Aas16031016}
D.~Aasen et~al.,
\newblock Phys. Rev. X {\bf 6}, 031016 (2016).
\bibitem{Lut10077001}
R.~M. Lutchyn, J.~D. Sau, and S.~Das~Sarma,
\newblock Phys. Rev. Lett. {\bf 105}, 077001 (2010).
\bibitem{Ore10177002}
Y.~Oreg, G.~Refael, and F.~von Oppen,
\newblock Phys. Rev. Lett. {\bf 105}, 177002 (2010).
\bibitem{Fle10180516}
K.~Flensberg,
\newblock Phys. Rev. B {\bf 82}, 180516 (2010).
\bibitem{Sar16035143}
S.~Das~Sarma, A.~Nag, and J.~D. Sau,
\newblock Phys. Rev. B {\bf 94}, 035143 (2016).
\bibitem{Moo18165302}
C.~Moore, T.~D. Stanescu, and S.~Tewari,
\newblock Phys. Rev. B {\bf 97}, 165302 (2018).
\bibitem{Law09237001}
K.~T. Law, P.~A. Lee, and T.~K. Ng,
\newblock Phys. Rev. Lett. {\bf 103}, 237001 (2009).
\bibitem{Dan20036801}
J.~Danon et~al.,
\newblock Phys. Rev. Lett. {\bf 124}, 036801 (2020).
\bibitem{Men20036802}
G.~C. M\'enard et~al.,
\newblock Phys. Rev. Lett. {\bf 124}, 036802 (2020).
\bibitem{Cao12115311}
Y.~Cao, P.~Wang, G.~Xiong, M.~Gong, and X.-Q. Li,
\newblock Phys. Rev. B {\bf 86}, 115311 (2012).
\bibitem{Zoc13036802}
B.~Zocher and B.~Rosenow,
\newblock Phys. Rev. Lett. {\bf 111}, 036802 (2013).
\bibitem{Fen21123032}
W.~Feng, L.~Qin, and X.-Q. Li,
\newblock New Journal of Physics {\bf 23}, 123032 (2021).
\bibitem{Liu15081405}
D.~E. Liu, M.~Cheng, and R.~M. Lutchyn,
\newblock Phys. Rev. B {\bf 91}, 081405 (2015).
\bibitem{Smi17063020}
S.~Smirnov,
\newblock New Journal of Physics {\bf 19}, 063020 (2017).
\bibitem{Fen22035148}
G.-H. Feng and H.-H. Zhang,
\newblock Phys. Rev. B {\bf 105}, 035148 (2022).
\bibitem{Smi22205430}
S.~Smirnov,
\newblock Phys. Rev. B {\bf 105}, 205430 (2022).
\bibitem{Smi19165427}
S.~Smirnov,
\newblock Phys. Rev. B {\bf 99}, 165427 (2019).
\bibitem{Cao23121407}
Z.~Cao et~al.,
\newblock Phys. Rev. B {\bf 108}, L121407 (2023).
\bibitem{Zha25023006}
H.~Zhao, J.~Wang, H.~Mao, and J.~Jin,
\newblock New Journal of Physics {\bf 27}, 023006 (2025).
\bibitem{Fle11090503}
K.~Flensberg,
\newblock Phys. Rev. Lett. {\bf 106}, 090503 (2011).
\bibitem{Liu11201308}
D.~E. Liu and H.~U. Baranger,
\newblock Phys. Rev. B {\bf 84}, 201308 (2011).
\bibitem{Ver14165314}
E.~Vernek, P.~H. Penteado, A.~C. Seridonio, and J.~C. Egues,
\newblock Phys. Rev. B {\bf 89}, 165314 (2014).
\bibitem{Den161557}
M.~Deng et~al.,
\newblock Science {\bf 354}, 1557 (2016).
\bibitem{Ram18045301}
J.~P. Ramos-Andrade, P.~A. Orellana, and S.~E. Ulloa,
\newblock Journal of Physics: Condensed Matter {\bf 30}, 045301 (2018).
\bibitem{Wan181}
X.-Q. Wang, S.-F. Zhang, C.~Jiang, G.-Y. Yi, and W.-J. Gong,
\newblock Physica E: Low-dimensional Systems and Nanostructures {\bf 104}, 1
(2018).
\bibitem{Zen17127302}
Q.-B. Zeng, S.~Chen, L.~You, and R.~L\"{u}ml,
\newblock Frontiers of Physics {\bf 12}, 127302 (2017).
\bibitem{Cal201900409}
A.~M. Calle, M.~Pacheco, P.~A. Orellana, and J.~A. Otálora,
\newblock Annalen der Physik {\bf 532}, 1900409 (2020).
\bibitem{Smi21205406}
S.~Smirnov,
\newblock Phys. Rev. B {\bf 104}, 205406 (2021).
\bibitem{Smi24195410}
S.~Smirnov,
\newblock Phys. Rev. B {\bf 109}, 195410 (2024).
\bibitem{Gon22122}
T.~Gong, L.~Zhang, X.~Dai, and W.-J. Gong,
\newblock Eur. Phys. J. Plus {\bf 137}, 122 (2022).
\bibitem{Mat22155409}
L.~M\'ath\'e, D.~Sticlet, and L.~P. Z\^arbo,
\newblock Phys. Rev. B {\bf 105}, 155409 (2022).
\bibitem{Mat231616}
L.~Máthé et~al.,
\newblock Nanomaterials {\bf 13}, 1616 (2023).
\bibitem{Bah25045920}
M.~Ait~Bahadou, A.~El~Allati, and K.~El~Anouz,
\newblock Physica Scripta {\bf 100}, 045920 (2025).
\bibitem{Yan14054105}
Y.~J. Yan,
\newblock J. Chem. Phys. {\bf 140}, 054105 (2014).
\bibitem{Jin15234108}
J.~S. Jin, S.~K. Wang, X.~Zheng, and Y.~J. Yan,
\newblock J. Chem. Phys. {\bf 142}, 234108 (2015).
\bibitem{Yan16110306}
Y.~J. Yan, J.~S. Jin, R.~X. Xu, and X.~Zheng,
\newblock Frontiers Phys. {\bf 11}, 110306 (2016).
\bibitem{Tan89101}
Y.~Tanimura and R.~Kubo,
\newblock J. Phys. Soc. Jpn. {\bf 58}, 101 (1989).
\bibitem{Tan906676}
Y.~Tanimura,
\newblock Phys. Rev. A {\bf 41}, 6676 (1990).
\bibitem{Tan06082001}
Y.~Tanimura,
\newblock J. Phys. Soc. Jpn. {\bf 75}, 082001 (2006).
\bibitem{Tan20020901}
Y.~Tanimura,
\newblock J. Chem. Phys {\bf 153}, 020901 (2020).
\bibitem{Xu05041103}
R.~X. Xu, P.~Cui, X.~Q. Li, Y.~Mo, and Y.~J. Yan,
\newblock J. Chem. Phys. {\bf 122}, 041103 (2005).
\bibitem{Jin08234703}
J.~S. Jin, X.~Zheng, and Y.~J. Yan,
\newblock J. Chem. Phys. {\bf 128}, 234703 (2008).
\bibitem{Fey63118}
R.~P. Feynman and F.~L. \mbox{Vernon, Jr.},
\newblock Ann. Phys. {\bf 24}, 118 (1963).
\bibitem{Wei08}
U.~Weiss,
\newblock {\em Quantum Dissipative Systems},
\newblock World Scientific, Singapore, 2008,
\newblock 3rd ed. Series in Modern Condensed Matter Physics, Vol.\ 13.
\bibitem{Kle09}
H.~Kleinert,
\newblock {\em Path Integrals in Quantum Mechanics, Statistics, Polymer
Physics, and Financial Markets},
\newblock World Scientific, Singapore, 5th edition, 2009.
\bibitem{Zhe121129}
X.~Zheng et~al.,
\newblock Prog. Chem. {\bf 24}, 1129 (2012).
\bibitem{Li12266403}
Z.~H. Li et~al.,
\newblock Phys. Rev. Lett. {\bf 109}, 266403 (2012).
\bibitem{Ye16608}
L.~Z. Ye et~al.,
\newblock WIREs Comp. Mol. Sci. {\bf 6}, 608 (2016).
\bibitem{Dat95}
S.~Datta,
\newblock {\em Electronic Transport in Mesoscopic Systems},
\newblock Oxford University Press, New York, 1995.
\bibitem{Hau08}
H.~Haug and A.-P. Jauho,
\newblock {\em Quantum Kinetics in Transport and Optics of Semiconductors},
\newblock Springer-Verlag, Berlin, 2nd, substantially revised edition, 2008,
\newblock Springer Series in Solid-State Sciences 123.
\bibitem{Zhe08184112}
X.~Zheng, J.~S. Jin, and Y.~J. Yan,
\newblock J. Chem. Phys. {\bf 129}, 184112 (2008).
\bibitem{Zhe08093016}
X.~Zheng, J.~S. Jin, and Y.~J. Yan,
\newblock New J. Phys. {\bf 10}, 093016 (2008).
\bibitem{Zhe09164708}
X.~Zheng, J.~S. Jin, S.~Welack, M.~Luo, and Y.~J. Yan,
\newblock J. Chem. Phys. {\bf 130}, 164708 (2009).
\bibitem{Zhe13086601}
X.~Zheng, Y.~J. Yan, and M.~Di~Ventra,
\newblock Phys. Rev. Lett. {\bf 111}, 086601 (2013).
\bibitem{Ye14165116}
L.~Z. Ye et~al.,
\newblock Phys. Rev. B {\bf 90}, 165116 (2014).
\bibitem{Wan13035129}
S.~K. Wang, X.~Zheng, J.~S. Jin, and Y.~J. Yan,
\newblock Phys. Rev. B {\bf 88}, 035129 (2013).
\bibitem{Zha18780}
H.~D. Zhang, R.~X. Xu, X.~Zheng, and Y.~J. Yan,
\newblock Mol. Phys. {\bf 116}, 780 (2018),
\newblock Special Issue, ``Molecular Physics in China''.
\bibitem{Wan20041102}
Y.~Wang, R.~X. Xu, and Y.~J. Yan,
\newblock J. Chem. Phys. {\bf 152}, 041102 (2020).
\bibitem{Zha16237}
H.~D. Zhang, Q.~Qiao, R.~X. Xu, and Y.~J. Yan,
\newblock Chem. Phys. {\bf 481}, 237 (2016).
\bibitem{Zha16204109}
H.~D. Zhang, Q.~Qiao, R.~X. Xu, and Y.~J. Yan,
\newblock J. Chem. Phys. {\bf 145}, 204109 (2016).
\bibitem{Jin20235144}
J.~S. Jin,
\newblock Phys. Rev. B {\bf 101}, 235144 (2020).
\bibitem{Eng04136602}
H.-A. Engel and D.~Loss,
\newblock Phys. Rev. Lett. {\bf 93}, 136602 (2004).
\bibitem{Rot09075307}
E.~A. Rothstein, O.~Entin-Wohlman, and A.~Aharony,
\newblock Phys. Rev. B {\bf 79}, 075307 (2009).
\bibitem{Yan14115411}
P.-Y. Yang, C.-Y. Lin, and W.-M. Zhang,
\newblock Phys. Rev. B {\bf 89}, 115411 (2014).
\bibitem{Bas10166801}
J.~Basset, H.~Bouchiat, and R.~Deblock,
\newblock Phys. Rev. Lett. {\bf 105}, 166801 (2010).
\bibitem{Bas12046802}
J.~Basset et~al.,
\newblock Phys. Rev. Lett. {\bf 108}, 046802 (2012).
\bibitem{Del18041412}
R.~Delagrange, J.~Basset, H.~Bouchiat, and R.~Deblock,
\newblock Phys. Rev. B {\bf 97}, 041412 (2018).
\bibitem{Bla001}
Y.~M. Blanter and M.~B\"{u}ttiker,
\newblock Phys. Rep. {\bf 336}, 1 (2000).
\bibitem{Cla17201109}
D.~J. Clarke,
\newblock Phys. Rev. B {\bf 96}, 201109 (2017).
\bibitem{Xu22064130}
Y.~Xu, J.~Jin, S.~Wang, and Y.~Yan,
\newblock Phys. Rev. E {\bf 106}, 064130 (2022).
\bibitem{Her927061}
S.~Hershfield,
\newblock Phys. Rev. B {\bf 46}, 7061 (1992).
\bibitem{Imr02}
I.~Imry,
\newblock {\em Introduction to Mesoscopic Physics},
\newblock Oxford university press, 2002.
\bibitem{Naz03}
{\em Quantum Noise in Mesoscopic Physics},
\newblock Kluwer, Dordrecht, 2003,
\newblock edited by Y. V. Nazarov.
\bibitem{Pic97162}
R.~de~Picciotto et~al.,
\newblock Nature {\bf 389}, 162 (1997).
\bibitem{Rez99238}
M.~Reznikov, R.~d. Picciotto, T.~G. Griffiths, M.~Heiblum, and V.~Umansky,
\newblock Nature {\bf 399}, 238 (1999).
\bibitem{Bid09236802}
A.~Bid, N.~Ofek, M.~Heiblum, V.~Umansky, and D.~Mahalu,
\newblock Phys. Rev. Lett. {\bf 103}, 236802 (2009).
\bibitem{Koz003398}
A.~A. Kozhevnikov, R.~J. Schoelkopf, and D.~E. Prober,
\newblock Phys. Rev. Lett. {\bf 84}, 3398 (2000).
\bibitem{Lef03067002}
F.~Lefloch, C.~Hoffmann, M.~Sanquer, and D.~Quirion,
\newblock Phys. Rev. Lett. {\bf 90}, 067002 (2003).
\bibitem{Ent07193308}
O.~Entin-Wohlman, Y.~Imry, S.~A. Gurvitz, and A.~Aharony,
\newblock Phys. Rev. B {\bf 75}, 193308 (2007).
\bibitem{Jin11053704}
J.~S. Jin, X.~Q. Li, M.~Luo, and Y.~J. Yan,
\newblock J. Appl. Phys. {\bf 109}, 053704 (2011).
\bibitem{Jin13025044}
J.~S. Jin, M.~Marthaler, P.-Q. Jin, D.~Golubev, and G.~Sch\"on,
\newblock New J. Phys. {\bf 15}, 025044 (2013).
\bibitem{Hu10101106}
J.~Hu, R.~X. Xu, and Y.~J. Yan,
\newblock J. Chem. Phys. {\bf 133}, 101106 (2010).
\bibitem{Hu11244106}
J.~Hu, M.~Luo, F.~Jiang, R.~X. Xu, and Y.~J. Yan,
\newblock J. Chem. Phys. {\bf 134}, 244106 (2011).
\bibitem{Hou15104112}
D.~Hou et~al.,
\newblock J. Chem. Phys. {\bf 142}, 104112 (2015).
\bibitem{Mao21014104}
H.~Mao, J.~Jin, S.~Wang, and Y.~Yan,
\newblock The Journal of Chemical Physics {\bf 155}, 014104 (2021).
\bibitem{Yan05187}
Y.~J. Yan and R.~X. Xu,
\newblock Annu. Rev. Phys. Chem. {\bf 56}, 187 (2005).
\bibitem{Mak01357}
Y.~Makhlin, G.~{Sch\"{o}n}, and A.~Shnirman,
\newblock Rev. Mod. Phys. {\bf 73}, 357 (2001).
\bibitem{Jin16083038}
J.~S. Jin, C.~Karlewski, and M.~Marthaler,
\newblock New J. Phys. {\bf 18}, 083038 (2016).
\bibitem{Shi16095002}
P.~Shi, M.~Hu, Y.~Ying, and J.~Jin,
\newblock AIP ADVANCES {\bf 6}, 095002 (2016).
\end{thebibliography}
\end{document}
%
%
%%%%%%%%%%%%%%%%%%%%%