M. R. Wegewijs's research while affiliated with Bavarian Research Alliance and other places

What is this page?


This page lists the scientific contributions of an author, who either does not have a ResearchGate profile, or has not yet added these contributions to their profile.

It was automatically created by ResearchGate to create a record of this author's body of work. We create such pages to advance our goal of creating and maintaining the most comprehensive scientific repository possible. In doing so, we process publicly available (personal) data relating to the author as a member of the scientific community.

If you're a ResearchGate member, you can follow this page to keep up with this author's work.

If you are this author, and you don't want us to display this page anymore, please let us know.

Publications (88)


RealTimeTransport: An open-source C++ library for quantum transport simulations in the strong coupling regime
  • Preprint
  • File available

May 2024

·

3 Reads

Konstantin Nestmann

·

Martin Leijnse

·

Maarten R. Wegewijs

The description of quantum transport in the strong system-reservoir coupling regime poses a significant theoretical and computational challenge that demands specialized tools for accurate analysis. RealTimeTransport is a new open-source C++ library that enables the computation of both stationary and transient transport observables for generic quantum systems connected to metallic reservoirs. It computes the Nakajima-Zwanzig memory kernels for both dynamics and transport in real-time going beyond traditional expansions in the bare system-reservoir couplings. Currently, several methods are available: (i) A renormalized perturbation theory in leading and next-to-leading order which avoids the low-temperature breakdown that limits the traditional theory. (ii) Starting from this well-behaved reference solution a 2- and 3-loop self-consistent renormalization-group transformation of the memory kernels is implemented. This allows refined quantitative predictions even in the presence of many body resonances, such as the Kondo enhancement of cotunneling. This paper provides an overview of the theory, the architecture of RealTimeTransport and practical demonstrations of the currently implemented methods. In particular, we analyze the stationary transport through a serial double quantum dot and showcase for the $T=0$ interacting Anderson model the complete time-development of single-electron tunneling (SET), cotunneling-assisted SET (CO-SET) and inelastic cotunneling resonances throughout the entire gate-bias stability diagram. We discuss the range of applicability of the implemented methods and benchmark them against other advanced approaches.

Download
Share

FIG. 1. Transient transport spectroscopy of a proximized quantum dot: For the example shown, an initialized state at t = 0 decays to the final 1-electron |1) state which is stationary for gate voltage . The initialization starts at time −t 0 < 0 from a state which before the switch is stationary at gate voltage 0 , in the example a doubly occupied state. It is transformed to a mixture of Andreev states ρ 0 , either by a fast switch (t 0 α −1 , bottom inset) or a slow switch (t 0 α −1 , top inset). As explained in Sec. III, in both cases the effective description by an initial mixture of Andreev energy eigenstates is appropriate. For the fast switch, this requires a charge and heat current readout that is coarse grained on the timescale α −1 of Cooperpair oscillations, which is much shorter than the dissipative transport timescale (scale γ −1 p ) on which the metal probes the proximized dot.
FIG. 3. Transient charge transport spectroscopy (no superconductor, α = 0). Plotted is the transient charge current I N (0) flowing right after a gate-voltage switch 0 → in units of the tunnel rate . For low T U this results in a decay of the dot charge from a nearly integer initial stationary value N z 0 towards another nearly integer stationary value N z . Possible values of (N z 0 , N z ) are indicated. Red indicates positive, blue negative currents.
FIG. 4. Overlap factor (13) for the fast-switch initial condition (30) depending on initial and final gate voltage (red = +1, blue = −1). The horizontal line cuts are taken at positions indicated by arrows with corresponding color. We show the cases of weak (i) and strong (iii) pairing labeled as in Fig. 2 and following.
FIG. 5. Invariants and stationary observables versus gate voltage − μ and bias μ < 0 for α values of Fig. 2. Line cuts in even rows are taken at arrows indicated in odd rows with same color. In row 1 the superconductor resonance (dashed line − μ = −μ − U/2) and Andreev resonances (dashed curves where μ = E η,τ for η, τ = ±) are indicated.
FIG. 6. Stationary charge current I N (∞) in units of as function of gate and bias voltage. Parameters and conventions are the same as in Fig. 5. Note the faint Andreev resonances occurring at low bias |μ| | U for α < U .

+7

Transient transport spectroscopy of an interacting quantum dot proximized by a superconductor: Charge and heat currents after a switch

August 2023

·

40 Reads

·

1 Citation

We analyze the time evolution of a quantum dot which is proximized by a large-gap superconductor and weakly probed using the charge and heat currents into a wide-band metal electrode. We map out the full time dependence of these currents after initializing the system by a switch. We find that due to the proximity effect there are two simple yet distinct switching procedures which initialize a nonstationary mixture of the gate-voltage-dependent eigenstates of the proximized quantum dot. We find in particular that the ensuing time-dependent heat current is a sensitive two-particle probe of the interplay of strong Coulomb interaction and induced superconducting pairing. The pairing can lead to a suppression of charge and heat current decay which we analyze in detail. The analysis of the results makes crucial use of analytic formulas obtained using fermionic duality, a “dissipative symmetry” of master equations describing this class of open systems.


Solution of master equations by fermionic-duality: Time-dependent charge and heat currents through an interacting quantum dot proximized by a superconductor

May 2023

·

19 Reads

·

2 Citations

SciPost Physics

We analyze the time-dependent solution of master equations by exploiting fermionic duality, a dissipative symmetry applicable to a large class of open systems describing quantum transport. Whereas previous studies mostly exploited duality relations after partially solving the evolution equations, we here systematically exploit the invariance under the fermionic duality mapping from the very beginning when setting up these equations. Moreover, we extend the resulting simplifications -so far applied to the local state evolution- to non-local observables such as transport currents. We showcase the exploitation of fermionic duality for a quantum dot with strong interaction -covering both the repulsive and attractive case- proximized by contact with a large-gap superconductor which is weakly probed by charge and heat currents into a wide-band normal-metal electrode. We derive the complete time-dependent analytical solution of this problem involving non-equilibrium Cooper pair transport, Andreev bound states and strong interaction. Additionally exploiting detailed balance we show that even for this relatively complex problem the evolution towards the stationary state can be understood analytically in terms of the stationary state of the system itself via its relation to the stationary state of a dual system with inverted Coulomb interaction, superconducting pairing and applied voltages.


Transient transport spectroscopy of an interacting quantum dot proximized by a superconductor: Charge- and heat-currents after a switch

April 2023

·

25 Reads

We analyze the time-evolution of a quantum dot which is proximized by a large-gap superconductor and weakly probed using the charge and heat currents into a wide-band metal electrode. We map out the full time dependence of these currents after initializing the system by a switch. We find that due to the proximity effect there are two simple yet distinct switching procedures which initialize a non-stationary mixture of the gate-voltage dependent eigenstates of the proximized quantum dot. We find in particular that the ensuing time-dependent heat current is a sensitive two-particle probe of the interplay of strong Coulomb interaction and induced superconducting pairing. The pairing can lead to a suppression of charge and heat current decay which we analyze in detail. The analysis of the results makes crucial use of analytic formulas obtained using fermionic duality, a ``dissipative symmetry'' of the master equation describing this class of open systems.


Solution of master equations by fermionic-duality: Time-dependent charge and heat currents through an interacting quantum dot proximized by a superconductor

October 2022

·

19 Reads

·

1 Citation

We analyze the time-dependent solution of master equations by exploiting fermionic duality, a dissipative symmetry applicable to a large class of open systems describing quantum transport. Whereas previous studies mostly exploited duality relations after partially solving the evolution equations, we here systematically exploit the invariance under the fermionic duality mapping from the very beginning when setting up these equations. Moreover, we extend the resulting simplifications -- so far applied to the local state evolution- to non-local observables such as transport currents. We showcase the exploitation of fermionic duality for a quantum dot with strong interaction -- covering both the repulsive and attractive case -- proximized by contact with a large-gap superconductor which is weakly probed by charge and heat currents into a wide-band normal-metal electrode. We derive the complete time-dependent analytical solution of this problem involving non-equilibrium Cooper pair transport, Andreev bound states and strong interaction. Additionally exploiting detailed balance we show that even for this relatively complex problem the evolution towards the stationary state can be understood analytically in terms of the stationary state of the system itself via its relation to the stationary state of a dual system with inverted Coulomb interaction, superconducting pairing and applied voltages.


Figure 3: Transient occupation 〈n〉(t) for U = 4Γ , ε = −U/2, bias V = Γ and several temperatures. Initially the dot is empty, ρ 0 = |0〉〈0|. We note that the O(δt 2 ) contributions of Eq. (41) are negligible here, but can play a role, see Fig. 5.
Figure 5: Reentrant effect for dot occupation and correlation for parameters U = 8Γ , ε = 2.75Γ , V = Γ at T = 0. The reentrant effect for time t r = Γ −1 is realized by the initial state ρ 0 with 〈n ↑ 〉 ρ 0 = 〈n ↓ 〉 ρ 0 ≈ 0.131 and 〈n ↑ n ↓ 〉 ρ 0 ≈ 0.007.
Renormalization group for open quantum systems using environment temperature as flow parameter

April 2022

·

24 Reads

·

4 Citations

SciPost Physics

We present the T-flow renormalization group method, which computes the memory kernel for the density-operator evolution of an open quantum system by lowering the physical temperature T of its environment. This has the key advantage that it can be formulated directly in real time, making it particularly suitable for transient dynamics, while automatically accumulating the full temperature dependence of transport quantities. We solve the T-flow equations numerically for the example of the single impurity Anderson model. We benchmark in the stationary limit, readily accessible in real-time for voltages on the order of the coupling or larger using results obtained by the functional renormalization group, density-matrix renormalization group and the quantum Monte Carlo method. Here we find quantitative agreement even in the worst case of strong interactions and low temperatures, indicating the reliability of the method. For transient charge currents we find good agreement with results obtained by the 2PI Green's function approach. Furthermore, we analytically show that the short-time dynamics of both local and non-local observables follow a "universal" temperature-independent behaviour when the metallic reservoirs have a flat wide band.


Renormalization group for open quantum systems using environment temperature as flow parameter

November 2021

·

6 Reads

We present the $T$-flow renormalization group method, which computes the memory kernel for the density-operator evolution of an open quantum system by lowering the physical temperature $T$ of its environment. This has the key advantage that it can be formulated directly in real time, making it particularly suitable for transient dynamics, while automatically accumulating the full temperature dependence of transport quantities. We solve the $T$-flow equations numerically for the example of the single impurity Anderson model. In the stationary limit, readily accessible in real-time for voltages on the order of the coupling or larger, we benchmark using results obtained by the functional renormalization group, density-matrix renormalization group and the quantum Monte Carlo method. Here we find quantitative agreement even in the worst case of strong interactions and low temperatures, indicating the reliability of the method. For transient charge currents we find good agreement with results obtained by the 2PI Green's function approach. Furthermore, we analytically show that the short-time dynamics of both local and non-local observables follow a "universal" temperature-independent behavior when the metallic reservoirs have a flat wide band.


General connection between time-local and time-nonlocal perturbation expansions

October 2021

·

9 Reads

·

7 Citations

There exist two canonical approaches to describe open quantum systems by a time-evolution equation: the Nakajima-Zwanzig quantum master equation, featuring a time-nonlocal memory kernel K, and the time-convolutionless equation with a time-local generator G. These key quantities have recently been shown to be connected by an exact fixed-point relation [Phys. Rev. X 11, 021041 (2021)]. Here we show that this implies a recursive relation between their perturbative expansions, allowing a series for the kernel K to be translated directly into a corresponding series for the more complicated generator G. This leads to an elegant way of computing the generator using well-developed, standard memory-kernel techniques for strongly interacting open systems. Moreover, it allows for an unbiased comparison of time-local and time-nonlocal approaches independent of the particular technique chosen to calculate expansions of K and G (Nakajima-Zwanzig projections, real-time diagrams, etc.). We illustrate this for leading and next-to-leading-order calculations of K and G for the single impurity Anderson model using both the bare expansion in the system-environment coupling and a more advanced renormalized series. We compare the different expansions obtained, quantify the legitimacy of the generated dynamics (complete positivity) and benchmark with the exact result in the noninteracting limit.


Figure 3: Resonant level model for strong coupling and detuning −µ = Γ /2. Plotted is the modulus of the complex valued matrix element (0|ˆΠ0|ˆ 0|ˆΠ(ω)|0) in units of Γ in the complex frequency plane where |0) = |0〉〈0| denotes the unoccupied state. (a) Finite temperature T = Γ /4: distinct from the two infinite sets of equidistant poles there are four poles (marked ×) at ω = 0, ω = −iΓ and ω = ± − iΓ /2. The last two poles are not visible here but appear in other matrix elements. (b) For T → 0 two branch cuts develop from the sets of equidistant poles. At resonance ( − µ T ) these poles (branch cuts) cancel exactly leaving just a single pole at −iΓ whereas off-resonance ( − µ T, Γ ) they move to the sides where they become suppressed in amplitude. Only in these two limits four poles remain and the dynamics is a semigroup [23]. For T → ∞ the first case always applies.
Figure 5: Resonant level model: Nonperturbative semigroup [Eq. (77), dashed] and initial slip approximation [Eq. (80), dash-dotted] compared to the exact occupation dynamics (solid line). (a) For generic level positions and couplings Γ 2πT the slip-corrected dynamics coincides with the exact result well before reaching the stationary value, while the semigroup approximation converges much later. In this case the time-nonlocal QME (64) solved by selection of the poles [Eq. (83)]-automatically including semigroup plus initial slip-is advantageous. (b) Surprisingly, near isolated points in the , Γ -parameter space the initial slip correction worsens the reliable semigroup approximation (77) based on the time-local QME. The precise positions at which this failure occurs are predicted by fermionic duality: As explained after Eq. (85), they are a consequence of the constraints it imposes on the slippage superoperator. The increased error introduced by the slip correction can also be understood as a failure to account for cancellation by eigenvector poles responsible for the branch cuts of the dynamics at T = 0.
Fermionic duality: General symmetry of open systems with strong dissipation and memory

September 2021

·

37 Reads

·

8 Citations

SciPost Physics

We consider the exact time-evolution of a broad class of fermionic open quantum systems with both strong interactions and strong coupling to wide-band reservoirs. We present a nontrivial fermionic duality relation between the evolution of states (Schrödinger) and of observables (Heisenberg). We show how this highly nonintuitive relation can be understood and exploited in analytical calculations within all canonical approaches to quantum dynamics, covering Kraus measurement operators, the Choi-Jamiołkowski state, time-convolution and convolutionless quantum master equations and generalized Lindblad jump operators. We discuss the insights this offers into the divisibility and causal structure of the dynamics and the application to nonperturbative Markov approximations and their initial-slip corrections. Our results underscore that predictions for fermionic models are already fixed by fundamental principles to a much greater extent than previously thought.


The connection between time-local and time-nonlocal perturbation expansions

July 2021

·

17 Reads

There exist two canonical approaches to describe open quantum systems by a time-evolution equation: the Nakajima-Zwanzig quantum master equation, featuring a time-nonlocal memory kernel $\mathcal{K}$, and the time-convolutionless equation with a time-local generator $\mathcal{G}$. These key quantities have recently been shown to be connected by an exact fixed-point relation [Phys. Rev. X 11, 021041 (2021)]. Here we show that this implies a recursive relation between their perturbative expansions, allowing a series for the kernel $\mathcal{K}$ to be translated directly into a corresponding series for the more complicated generator $\mathcal{G}$. This leads to an elegant way of computing the generator using well-developed, standard memory-kernel techniques for strongly interacting open systems. Moreover, it allows for an unbiased comparison of time-local and time-nonlocal approaches independent of the particular technique chosen to calculate expansions of $\mathcal{K}$ and $\mathcal{G}$ (Nakajima-Zwanzig projections, real-time diagrams, etc.). We illustrate this for leading and next-to-leading order calculations of $\mathcal{K}$ and $\mathcal{G}$ for the single impurity Anderson model using both the bare expansion in the system-environment coupling and a more advanced renormalized series. We compare the different expansions obtained, quantify the legitimacy of the generated dynamics (complete positivity) and benchmark with the exact result in the non-interacting limit.


Citations (65)


... The former, in particular, allows for resummation techniques that restore the Andreev bound state picture [58][59][60], and can even be extended to capture perturbatively the effect of the quasiparticles [61]. Recent works have employed fermionic duality to investigate the dynamics in the time domain of junctions with large gap superconductors [62,63]. ...

Reference:

Non-equilibrium cotunneling in interacting Josephson junctions
Transient transport spectroscopy of an interacting quantum dot proximized by a superconductor: Charge and heat currents after a switch

... The former, in particular, allows for resummation techniques that restore the Andreev bound state picture [58][59][60], and can even be extended to capture perturbatively the effect of the quasiparticles [61]. Recent works have employed fermionic duality to investigate the dynamics in the time domain of junctions with large gap superconductors [62,63]. ...

Solution of master equations by fermionic-duality: Time-dependent charge and heat currents through an interacting quantum dot proximized by a superconductor

SciPost Physics

... We consider two distinct physical preparation procedures, denoted "fast switch" and "slow switch", respectively, and map out their full time-evolution of charge and heat currents based on master and transport equations derived earlier [58]. We make use of the analytical solution of these equations reported in a recent work [59] which did not address the specific initialization. This work fully exploited the above mentioned fermionic duality which is applicable to proximized quantum dot systems [16,18] finding that the interesting interplay of strong repulsive interaction and strong induced supercon-ducting pairing is exhibited only in the time-scale characteristic of the quasiparticle charge decay (γ −1 c ): similar to systems without a superconductor, the Andreev-state parity decay time (γ −1 p ) is insensitive to any parameter except the metal contact tunnel-rate constant Γ. ...

Solution of master equations by fermionic-duality: Time-dependent charge and heat currents through an interacting quantum dot proximized by a superconductor

... In prior work this was done by a "T -flow" from infinitely high temperature to the low temperature of interest. 29 This involved solving a differential RG equation with respect to the physical temperature T , which at low T approaches a fixed point (in the space of time-dependent memory kernels and vertex functions). ...

Renormalization group for open quantum systems using environment temperature as flow parameter

SciPost Physics

... It is instructive to analyze the "worst case" T → 0, where the bare perturbation theory suffers from a complete breakdown due to the very slow and oscillatory decay of the contraction function γ − (t) ∝ e −iη 1 µ r t /t. 44 By contrast, the limit T → 0 is unproblematic in Eq. (26) because the intermediate renormalized propagator is still exponentially decaying towards the infinte-temperature stationary state Π ∞ (t) → 1 d |1)(1|, which is a left zero eigenvector of the D + 2 creation superfermion on the right. 37 The next-to-leading order corrections are ...

General connection between time-local and time-nonlocal perturbation expansions
  • Citing Article
  • October 2021

... This is not an isolated observation but was noted in different contexts [21][22][23][24]. It was subsequently shown to be a consequence of fermionic duality [16] which is applicable to a large class of interacting dissipative nanostructures [18,25]. Transient transport spectroscopy of a proximized quantum dot: For the example shown, an initialized state at t = 0 decays to the final 1-electron |1) state which is stationary for gate voltage . ...

Fermionic duality: General symmetry of open systems with strong dissipation and memory

SciPost Physics

... Next-to-leading order renormalized perturbation theory at T = 0: Transient occupation of the excited state ⟨n ↑ ⟩(t) as function of bias voltage V = 2µ L = −2µ R and gate voltage Vg = −ϵ for an initially empty Anderson quantum dot ρ 0 = |0⟩⟨0| with strong interaction U = 30Γ and magnetic field B = 5Γ .ator.47,48 ...

How Quantum Evolution with Memory is Generated in a Time-Local Way

Physical Review X

... In these quantum-dot realizations of heat engines, electron-electron interaction can play an important role in the performance. This is true for the standard repulsive Coulomb interaction, but strong attractive interaction has also been analyzed in pumping through quantum dots [10] and for steadystate thermoelectric systems [11]. The most basic setup allowing for such a nano-electronic implementation of a cyclic thermal machine is a driven single-level quantum dot coupled to two electronic baths. ...

Thermovoltage in quantum dots with attractive interaction

Applied Physics Letters

... E. We focus on transport observables but emphasize that we have checked that every computed propagator is a completely positive map at each time t. This is a basic criterion for the physicality of an approximation ensuring it also properly evolves the system when it is entangled [30,38]. Furthermore, in App. ...

Density-operator evolution: Complete positivity and the Keldysh real-time expansion

SciPost Physics

... If neither of these parameters are small (as is often the case in experimental systems), then there is no simple theory, one requires a complicated theory of quantum non-equilibrium systems [52], such as formulated in terms of Keldysh theory. There are different ways to formulate Keldysh theory, and we present one of them below, which is called real-time diagrammatics or realtime transport theory. 1 However, formulating the problem in terms of a Keldysh theory is usually not sufficient to solve the problem in question, further approximations are usually required, and progress is difficult. ...

Five approaches to exact open-system dynamics: Complete positivity, divisibility, and time-dependent observables
  • Citing Article
  • July 2019

The Journal of Chemical Physics