betaFeedback

WorksPhysics and Astronomy

Thermal-energy relaxation in chiral quantum Hall edge mode

Abstract

We study the nonequilibrium state generated by a real Ohmic contact (ROC) in the charge-fractionalization regime. We then investigate its thermal-energy relaxation along a quantum Hall edge mediated by a fictitious Ohmic contact (FOC) with large capacitance and perfect coupling to the edge. For cold Ohmic contacts, the electron distribution function and current noise are fully described by their equilibrium forms and the corresponding Fermi-Dirac distribution. In contrast, in the case of hot Ohmic contacts, strong thermalization effects do not drive the outgoing state into complete equilibrium. Both the distribution function and the noise retain signatures of the initial nonequilibrium state and therefore cannot be characterized solely by the effective contact temperature, contrary to common assumptions.

Relaxation.pdf

The experimental progress in hybrid mesoscale systems based on chiral QH edge states has triggered more accurate theoretical studies of phase coherence [1, 2, 3, 4, 5], charge and heat quantization [6, 7, 8, 9, 10, 11], relaxation and equilibration [12, 13, 14, 15, 16, 17, 18], and entanglement [19, 20, 21, 22], which are fundamental questions of mesoscopic physics. Among them, the investigation of the electron distribution function and equilibration along the QH edge states, for example, the unresolved issue of heat leakage [23, 24], remains an open question. In these experiments a quantum point contact (QPC) serves as an essential element. Namely, one prepares a non-equilibrium double-step distribution function with the help of biased partition QPC. Later, the evolution of distribution function along edge states is studied by measuring the conductance of tunneling current between two edges using the quantum dot or the second QPC as a detector [25].

Further progress in experimental techniques allowed experimentalists to realize systems, where the QH edge states are perfectly coupled to Ohmic contact, a mesoscale piece of metal with zero level spacing and finite charging energy, Ec=e2/2CE_c=e^2/2C, where CC is the capacitance [26]. It is worth mentioning that the Ohmic contact has become a key element of new experimental and theoretical research [27]. The remarkable examples are the suppression of charge quantization caused by quantum fluctuations both in quantum and thermal regime [6, 7], heat Coulomb blockade effect [8, 9, 11], interaction induced recovery of the phase coherence [28, 29, 30], charge Kondo effect [31, 32], quantization of the anyonic heat flow [33], observation of the half-integer thermal Hall conductance [34, 35], the measurement of full counting statistics of electron currents with the help of quantum ammeter [36], and tunneling effects in multisite quantum Hall circuits [37, 38, 39]. Recently, an implementation of Ohmic contact with chiral edge channels allowed experimentalists to investigate relaxation and equilibration, measuring zero frequency noise for both integer and fractional filling factors [40]. This is the alternative method compared to conventional conductance measurements with respect to applied bias [25]. Subsequently, it has been used to

study quantized heat flow in the Hofstadter butterfly [41], the ballistic-to-diffusive transition in quantum Hall channels [42], dynamical Coulomb blockade under a temperature bias [43], and topological effects in thermal Hall conductance [44]. In addition to quantum Hall systems based on two-dimensional electron gas, this experimental technique has been used for other two-dimensional materials, including graphene, to investigate heat equilibration between integer and fractional quantum Hall edge modes [45] and heat transport in Moiré Chern bands of bilayer graphene [46, 47].

In this Letter, we study the energy relaxation by Ohmic contact strongly coupled to the QH edge state at filling factor ν=1\nu=1 (see Figure 1). The setup consists of a real Ohmic contact (ROC), which at zero capacitance (C→0C\to0) serves as an ideal electron fractionalizer. Two independently biased current sources, S1S_1 and S2S_2 with base temperature TinT_{\mathrm{in}} inject current return to the ROC. The sources biases, V1=−V2=VV_1=-V_2=V, are chosen such that the total accumulated charge in the ROC vanishes, namely ⟨Qc⟩=(V1+V2)C/2Rq=0\langle Q_{\mathrm{c}}\rangle=(V_1+V_2)\mathcal{C}/2R_q=0, while the charge fluctuations remain nonzero, where Rq=2πℏ/e2R_q=2\pi\hbar/e^2 is the quantum resistance. The equilibration in the ROC increases it’s temperature Tc>TinT_{\mathrm{c}}>T_{\mathrm{in}} with non-zero dissipated power. Further, the measured distribution function and noise allow the study of the relaxation that occurs after the ROC. Relaxation is modeled by a fictitious Ohmic contact (FOC) with capacitance CfC_{\mathrm{f}} and negligible level spacing. The main interest is the study of relaxation modeled by the FOC with a large capacitance. Throughout the paper, we set ∣e∣=ℏ=kB=1|e|=\hbar=k_{\mathrm{B}}=1.

Schematic of setup with ROC and FOC

Figure 1. Schematic of setup. It consists of the real Ohmic contact (left), a micrometer size piece of metal with negligible level spacing and capacitance CC. The real Ohmic contact is used to prepare the initial non-equilibrium distribution function with the help of two biased sources S1S_1 and S2S_2. Relaxation of initially prepared state is modeled by a fictitious Ohmic contact with large capacitance CfC_f, located after real Ohmic contact. The setup can be used to measure the distribution function, fout,1(ϵ)f_{\mathrm{out},1}(\epsilon), and noise, Sout,1(ω=0)S_{\mathrm{out},1}(\omega=0), right before and after passing the region modeled by fictitious Ohmic contact.

Model and formulation of the problem. We use effective field theory [48] and the bosonization technique [49] to describe the QH edge states perfectly coupled to the Ohmic contact. According to effective theory, the collective fluctuations of the charge densities ρα(x,t)\rho_\alpha(x,t) and currents jα(x,t)j_\alpha(x,t) in the QH edge channels are expressed in terms of the bosonic fields ϕα(x,t)\phi_\alpha(x,t),

ρα=(1/2π)∂xϕα,jα=−(1/2π)∂tϕα,(1)\rho_\alpha=(1/2\pi)\partial_x\phi_\alpha,\qquad j_\alpha=-(1/2\pi)\partial_t\phi_\alpha, \tag*{(1)}

where the index α\alpha denotes the incoming and outgoing channels (see Figure 1). As stated in the bosonization technique, the electrons in these channels are represented by the vertex operators ψα(x,t)∝eiϕα(x,t)\psi_\alpha(x,t)\propto e^{i\phi_\alpha(x,t)}, so that the electron distribution function in the outgoing channel right after real or fictitious Ohmic contact reads

fα(ϵ)=∫dte−iϵtKα(t),f_\alpha(\epsilon)=\int dt e^{-i\epsilon t}K_\alpha(t),
Kα(t)∝⟨e−iϕα(t)eiϕα(0)⟩.(2)K_\alpha(t)\propto\langle e^{-i\phi_\alpha(t)}e^{i\phi_\alpha(0)}\rangle. \tag*{(2)}

The prefactor in the expression for Kα(t)K_\alpha(t) can be fixed in the end of the calculations by comparing with the equilibrium correlation function for free electrons. In particular, the equilibrium two-point correlation function of the incoming states Kineq(t)=−iTin/sinh⁡[πTin(t−i0)]K_{\mathrm{in}}^{\mathrm{eq}}(t)=-iT_{\mathrm{in}}/\sinh[\pi T_{\mathrm{in}}(t-i0)] results in a Fermi-Dirac distribution function fFD(ϵ/Tin)=1/(1+eϵ/Tin)f_{\mathrm{FD}}(\epsilon/T_{\mathrm{in}})=1/(1+e^{\epsilon/T_{\mathrm{in}}}). Apart from defining the electron distribution function, one can study the non-symmetrized noise of outgoing current after ROC and FOC,

Sout(ω)=∫dteiωt⟨δjout(t)δjout(0)⟩,(3)S_{\mathrm{out}}(\omega)=\int dt e^{i\omega t}\langle\delta j_{\mathrm{out}}(t)\delta j_{\mathrm{out}}(0)\rangle, \tag*{(3)}

where δjout(t)=jout(t)−⟨jout(t)⟩\delta j_{\mathrm{out}}(t)=j_{\mathrm{out}}(t)-\langle j_{\mathrm{out}}(t)\rangle and the average is taken with respect to the incoming states. The symmetrized noise is defined as the sum of the positive and negative frequency components of the non-symmetrized noise, [Sout(ω)+Sout(−ω)]/2[S_{\mathrm{out}}(\omega)+S_{\mathrm{out}}(-\omega)]/2, which accounts for fluctuations associated with both emission and absorption processes and is an even function of frequency. At zero frequency, the non-symmetrized and symmetrized noises coincide. It is important to note that, in general, one cannot apply the relation

Sout(0)=Gq∫dϵ[1−fFD(ϵ/Ti)]fFD(ϵ/Ti),i=c,f,(4)S_{\mathrm{out}}(0)=G_q\int d\epsilon[1-f_{\mathrm{FD}}(\epsilon/T_i)]f_{\mathrm{FD}}(\epsilon/T_i),\qquad i=\mathrm{c},\mathrm{f}, \tag*{(4)}

for zero frequency noise of current right after the ROC and FOC, since Wick’s theorem is invalid for the fermions outgoing from both Ohmic contacts. It will be shown later that this behavior is related to the nonequilibrium character of the outgoing state prepared by the ROC; due to the presence of electron-electron interactions between two incoming states and strong thermalization effects inside the ROC. In Eq. (4), Gq=1/Rq=e2/2πℏG_q=1/R_q=e^2/2\pi\hbar is the quantum conductance, and TcT_{\mathrm{c}} is the temperature of the ROC and TfT_{\mathrm{f}} is the temperature of the FOC, which are obtained in Eqs. (8) and (9) below. These temperatures are different from TinT_{\mathrm{in}} due to thermalization effects.

Langevin equations and noise in the outgoing channel. We apply the Langevin equation method to evaluate the electron correlation function Kα(t)K_\alpha(t) in Eq. (2). According to this method, the Ohmic contact serves as a reservoir of neutral excitations, which are accounted for by introducing Langevin currents jc,lj_{\mathrm{c,l}} into the equation of motion for the accumulated charge at the contact. The Langevin equations can be obtained rigorously using the bosonization technique for dynamics of bosonic fields [7, 8, 29, 36, 50, 51]. In the frequency domain, the outgoing currents immediately after the ROC are expressed in terms of the incoming currents and Langevin sources

j1neq(ω)=jc,i(ω)+∑l=1,2T(ω)[jin,l(ω)−jc,l(ω)],(5)j_1^{\mathrm{neq}}(\omega)=j_{\mathrm{c},i}(\omega)+\sum_{l=1,2}\mathcal{T}(\omega)[j_{\mathrm{in},l}(\omega)-j_{\mathrm{c},l}(\omega)], \tag*{(5)}

where the subscript i=1,2i=1,2 labels the channel and T(ω)=1/(2−iωRqC)\mathcal{T}(\omega)=1/(2-i\omega R_q\mathcal{C}) is the scattering coefficient. The superscript on the left hand side denotes that the outgoing state is non-equilibrium due to thermalization effects and charging energy. The first term in the denominator of the scattering coefficient, T(ω)\mathcal{T}(\omega), counts the number of incoming and outgoing channels with respect to ROC. Similar equations can be written in the presence of additional FOC. The outgoing current after FOC has the form

jout,1(ω)=A(ω)j1neq(ω)+B(ω)jf,1(ω),(6)j_{\mathrm{out},1}(\omega)=\mathcal{A}(\omega)j_1^{\mathrm{neq}}(\omega)+\mathcal{B}(\omega)j_{\mathrm{f},1}(\omega), \tag*{(6)}

where the scattering coefficients are given by A(ω)=1/(1−iωRqCf)\mathcal{A}(\omega)=1/(1-i\omega R_qC_{\mathrm{f}}) and B(ω)=1−A(ω)\mathcal{B}(\omega)=1-\mathcal{A}(\omega) and CfC_{\mathrm{f}} is the capacitance of FOC. For the case of no FOC, Cf→0C_{\mathrm{f}}\to0, and one has jout,1=j1neqj_{\mathrm{out},1}=j_1^{\mathrm{neq}}. The two-point correlation functions of the incoming currents and Langevin sources according to the fluctuation-dissipation theorem are given by the equilibrium spectral function [52]

⟨δjk,l(ω)δjk′,l′(ω′)⟩=2πδkk′δll′δ(ω+ω′)Sk,l(ω),(7)\langle\delta j_{k,l}(\omega)\delta j_{k',l'}(\omega')\rangle=2\pi\delta_{kk'}\delta_{ll'}\delta(\omega+\omega')S_{k,l}(\omega), \tag*{(7)}

where Sk,l(ω)=Gqω/(1−e−ω/Tk,l)S_{k,l}(\omega)=G_q\omega/(1-e^{-\omega/T_{k,l}}) and k=in,c,fk=\mathrm{in},\mathrm{c},\mathrm{f} and l=1,2,3l=1,2,3. These correlation functions are used to determine the temperatures of the ROC and FOC, as well as to calculate the distribution function and noise.

Heating effects in Ohmic contacts. Before calculating the distribution function and noise, one needs to take into account the thermalization effects inside the Ohmic contacts, i.e., one needs to find the temperatures TcT_c and TfT_f in Eq. (7). To do this, we solve self-consistently the energy balance equations, i.e. conservation of incoming and outgoing heat currents, where the heat current is defined by Joule’s law, Jα=Rq⟨jα2⟩/2J_\alpha=R_q\langle j_\alpha^2\rangle/2.

Further, using Joule’s law, we obtain the equation determining the ROC temperature,

V2Rq+πTin212[2+F(πTinRqC2)]=πTc212[2+F(πTcRqC2)],\frac{V^2}{R_q}+\frac{\pi T_{\rm in}^2}{12}\left[2+\mathcal{F}\left(\frac{\pi T_{\rm in}R_qC}{2}\right)\right]=\frac{\pi T_c^2}{12}\left[2+\mathcal{F}\left(\frac{\pi T_cR_qC}{2}\right)\right],

where F(x)=(3/x2)[x+ln⁡(2x)+Ψ(1/2x)]\mathcal{F}(x)=(3/x^2)[x+\ln(2x)+\Psi(1/2x)] is a dimensionless function, and Ψ(x)=dln⁡Γ(x)/dx\Psi(x)=d\ln\Gamma(x)/dx denotes the digamma function and Γ(x)=∫0∞yx−1e−y dy\Gamma(x)=\int_0^\infty y^{x-1}e^{-y}\,dy is the gamma function. In the absence of FOC and in the small-capacitance regime, TcRqC≪1T_cR_qC\ll1, the dimensionless function approaches F(x)→−1\mathcal{F}(x)\to-1. Therefore, the ROC temperature is given by [53]

Tc2=Tin2+6V2π2.(8)T_c^2=T_{\rm in}^2+\frac{6V^2}{\pi^2}. \tag*{(8)}

This equation expresses the heating of the ROC as a result of thermalization. Incoming channels initially have the base temperature TinT_{\rm in}, while the symmetric voltage bias, V1=V2=−VV_1=V_2=-V, introduces an additional nonequilibrium energy proportional to V2V^2. Furthermore, electron–electron interactions within the ROC redistribute this excess energy among the neutral modes jc,lj_{c,l}, converting it into thermal energy and producing the temperature TcT_c, which is indeed higher than the base temperature.

For the FOC, a similar analysis in the large-capacitance regime, TinRqCf≫1T_{\rm in}R_qC_f\gg1, yields the following expression for the fictitious contact temperature [53]

Tf2=Tin2+3V2π2.(9)T_f^2=T_{\rm in}^2+\frac{3V^2}{\pi^2}. \tag*{(9)}

Compared with the ROC, the FOC reaches a lower temperature because it receives only part of the excess energy generated by the two biased incoming channels; because it is coupled to one edge channel. The heating effects discussed here are used below for the calculation of the distribution function and noise. At V=0V=0 the system is in equilibrium and at Tf=Tc=TinT_f=T_c=T_{\rm in}.

Noise in the outgoing channel. To find the outgoing noise after the FOC, we use Eq. (6) and arrive at the following equation

Sout,1(ω)=Sfeq(ω)+∣A(ω)∣2[Sin,1neq(ω)−Sfeq(ω)].(10)S_{\mathrm{out},1}(\omega)=S_f^{\mathrm{eq}}(\omega)+|A(\omega)|^2\left[S_{\mathrm{in},1}^{\mathrm{neq}}(\omega)-S_f^{\mathrm{eq}}(\omega)\right]. \tag*{(10)}

where Sfeq(ω)S_f^{\mathrm{eq}}(\omega) is an equilibrium spectral function at temperature TfT_f from Eq. (7). The non-equilibrium incoming noise, Sin,1neq(ω)S_{\mathrm{in},1}^{\mathrm{neq}}(\omega), in the case of fractionalizer (after ROC), when C→0C\to0 is given by [53]

Sin,1neq(ω)=[Sineq(ω)+Sceq(ω)]/2.(11)S_{\mathrm{in},1}^{\mathrm{neq}}(\omega)=\left[S_{\mathrm{in}}^{\mathrm{eq}}(\omega)+S_c^{\mathrm{eq}}(\omega)\right]/2. \tag*{(11)}

Furthermore, at zero frequency and ωRqCf→0\omega R_qC_f\to0 and ω/Tin,c→0\omega/T_{\rm in,c}\to0 we have ∣A(ω)∣2→1|A(\omega)|^2\to1; thus, the outgoing noise is given by the incoming non-equilibrium noise [53], namely,

Sout,1(0)=Sin,1neq(0)=Gq(Tin+Tc)/2.(12)S_{\mathrm{out},1}(0)=S_{\mathrm{in},1}^{\mathrm{neq}}(0)=G_q(T_{\rm in}+T_c)/2. \tag*{(12)}

From the above equation, one can see that in the case of cold ROC, i.e. Tc/Tin→1T_c/T_{\rm in}\to1, the noise is given by the equilibrium result Sout,1(0)=GqTinS_{\mathrm{out},1}(0)=G_qT_{\rm in}. However, in the hot ROC limit, Tc/Tin→∞T_c/T_{\rm in}\to\infty, the result deviates from the equilibrium expectation at temperature TcT_c. Instead, the zero-frequency outgoing noise is given by Sout,1(0)=GqTc/2S_{\mathrm{out},1}(0)=G_qT_c/2. Thus, Eq. (4) does not provide the correct result in this regime, since it assumes an equilibrium Fermi-Dirac distribution with temperature TcT_c, yielding Sout,1(0)=GqTcS_{\mathrm{out},1}(0)=G_qT_c. It means that the relation for noise obtained using Wick’s theorem is not applicable, since the outgoing state is highly non-equilibrium and keeps the information about ROC, where the strong electron-electron interaction between two incoming states is present. However, in the limits ωRqCf→∞\omega R_qC_f\to\infty followed by ω/Tin,c→0\omega/T_{\rm in,c}\to0, one obtains Sout,1(0)=GqTcS_{\mathrm{out},1}(0)=G_qT_c. Nevertheless, the zero-frequency noise of the nonequilibrium ”in,1” outgoing ballistic channel downstream of the ROC remains described by Eq. (12).

Non-equilibrium distribution function. As a first step, we consider the prepared non-equilibrium state after the injection source, i.e. ROC. Our interest is the regime of small capacitance C→0C\to0. In this case, the correlation function in Eq. (2) for the distribution function splits into two factors

Kin,1neq(t)=Kineq(t)⋅Kceq(t),(13)K_{\mathrm{in},1}^{\mathrm{neq}}(t)=\sqrt{K_{\mathrm{in}}^{\mathrm{eq}}(t)}\cdot\sqrt{K_c^{\mathrm{eq}}(t)}, \tag*{(13)}

where Kαeq(t)=−iπTα/sinh⁡[πTα(t−i0)]K_\alpha^{\mathrm{eq}}(t)=-i\pi T_\alpha/\sinh[\pi T_\alpha(t-i0)], with α=in,c\alpha={\rm in},c. The square roots in Eq. (13) arise from the fact that at C→0C\to0, the ROC acts as the current fractionalizer and T(ω)→1/2\mathcal{T}(\omega)\to1/2 in Eq. (5). In case of cold ROC, Tc/Tin→1T_c/T_{\rm in}\to1, expanding the integrand with respect to Tc/TinT_c/T_{\rm in} in Eq. (2), the result for the distribution function to leading order is given by the equilibrium FD function at base temperature

fin,1neq(ϵ)=fFD(ϵ/Tin).(14)f_{\mathrm{in},1}^{\mathrm{neq}}(\epsilon)=f_{\mathrm{FD}}(\epsilon/T_{\rm in}). \tag*{(14)}

The sub-leading contribution is equal to (1−Tc2/Tin2)∂fFD(ϵ)/4∂ϵ(1-T_c^2/T_{\rm in}^2)\partial f_{\mathrm{FD}}(\epsilon)/4\partial\epsilon at z→1z\to1. Qualitatively, the result of Eq. (14) is clear; the whole system is almost in equilibrium at Tc/Tin→1T_c/T_{\rm in}\to1 and the result in leading order is given by the equilibrium distribution function of FD.

In case of hot ROC, Tc/Tin→∞T_c/T_{\rm in}\to\infty, one gets the following results in low and high energy regimes [53]

fin,1neq(ϵ)=1/2−C∣ϵ∣/4Tc,∣ϵ∣/Tc≪1,f_{\mathrm{in},1}^{\mathrm{neq}}(\epsilon)=1/2-\mathcal{C}|\epsilon|/4T_c,\qquad|\epsilon|/T_c\ll1,
fin,1neq(ϵ)=e−ϵ/Tc/πϵ/Tc,ϵ/Tc≫1.(15)f_{\mathrm{in},1}^{\mathrm{neq}}(\epsilon)=e^{-\epsilon/T_c}/\sqrt{\pi\epsilon/T_c},\qquad\epsilon/T_c\gg1. \tag*{(15)}

where C=(2/π2)∫dxx/sinh⁡(x)≈1.5573\mathcal{C}=(2/\pi^2)\int dx\sqrt{x}/\sinh(x)\approx1.5573 and finneq(ϵ)→1f_{\mathrm{in}}^{\mathrm{neq}}(\epsilon)\to1 at ϵ→−∞\epsilon\to-\infty. From Eq. (15), one can see that the tail of the distribution function differs sufficiently from the equilibrium one at temperature TcT_c, which has a pure exponential tail exp⁡(−ϵ/Tc)\exp(-\epsilon/T_c). Moreover, the sub-leading term in the low energy regime is different by factor C>1C > 1 compared to the equilibrium result C=1C = 1. It is worth mentioning that in case of cold ROC the leading order contribution is given by the FD factor for both low- and high energy regimes. In contrast, the hot ROC’s case indicates the non-equilibrium nature of the prepared state more precisely in both low- and high energy regimes, since the temperature difference between incoming and outgoing states is huge, Tc/Tin→∞T_c/T_{\mathrm{in}} \to\infty. Despite the conventional assumption that the ROC completely equilibrates the incoming states, our results, Eqs. (12) and (15), show that this picture is not accurate. These findings emphasize the limitations of assuming complete equilibration at the ROC.

In order to model the relaxation at the QH edge, we consider the case of one large FOC (Cf→∞C_f \to\infty) and TinRqCf≫1T_{\mathrm{in}}R_qC_f \gg1. In this case according to the scattering problem in Eq. (6), the correlation function in Eq. (2) is given by the product of the equilibrium two-point correlation function and the non-equilibrium part [53]

Kout,1(t)=Kfeq(t)⋅exp⁡[Φ(t)],(16)K_{\mathrm{out},1}(t) = K_f^{\mathrm{eq}}(t) \cdot\exp[\Phi(t)], \tag*{(16)}

where Kfeq(t)K_f^{\mathrm{eq}}(t) is the equilibrium two-point correlation function at temperature TfT_f as in Eq. (13) and the dimensionless function is responsible for non-equilibrium effects in exponent given by the integral over frequency [53]

Φ(t)=Rq∫dωe−iωt−1ω2+η2[Sout,1(ω)−Sfeq(ω)],η→0,(17)\Phi(t) = R_q \int d\omega\frac{e^{-i\omega t}-1}{\omega^2+\eta^2}\left[S_{\mathrm{out},1}(\omega)-S_f^{\mathrm{eq}}(\omega)\right], \qquad\eta\to0, \tag*{(17)}

and, to recall, Sout,1(ω)S_{\mathrm{out},1}(\omega) is given in Eq. (10). In equilibrium exp⁡[Φ(t)]=1\exp[\Phi(t)] = 1 and Kfeq(t)=Kineq(t)K_f^{\mathrm{eq}}(t) = K_{\mathrm{in}}^{\mathrm{eq}}(t). Further, we substitute Eq. (16) in Eq. (2) and take into account that the main contribution to the integral comes from times t≲RqCft \lesssim R_qC_f, where one can replace the function ∣A(ω)∣2=1/[1+(ωRqCf)2]|A(\omega)|^2 = 1/[1+(\omega R_qC_f)^2] in the integrand of Φ(t)\Phi(t) in Eq. (17) by the delta function, πδ(ω)/RqCf\pi\delta(\omega)/R_qC_f. After some algebra, one arrives at the following result [53]

fout,1(ϵ)=fFD(ϵ/Tf)+δf(ϵ),f_{\mathrm{out},1}(\epsilon) = f_{\mathrm{FD}}(\epsilon/T_f) + \delta f(\epsilon),
δf(ϵ)=π2Cflim⁡ω→0[Sin,1neq(ω)−Sfeq(ω)]∂2∂ϵ2fFD(ϵ/Tf).(18)\delta f(\epsilon) = \frac{\pi^2}{C_f}\lim_{\omega\to0}\left[S_{\mathrm{in},1}^{\mathrm{neq}}(\omega)-S_f^{\mathrm{eq}}(\omega)\right]\frac{\partial^2}{\partial\epsilon^2}f_{\mathrm{FD}}(\epsilon/T_f). \tag*{(18)}

The pre-factor in the sub-leading term is given by zero frequency, i.e. lim⁡ω→0[Sin,1neq(ω)−Sfeq(ω)]=Gq(Tin+Tc−2Tf)/2\lim_{\omega\to0}\left[S_{\mathrm{in},1}^{\mathrm{neq}}(\omega)-S_f^{\mathrm{eq}}(\omega)\right] = G_q(T_{\mathrm{in}}+T_c-2T_f)/2. From the above equation, one can see that at Tf=Tc=TinT_f = T_c = T_{\mathrm{in}} the system is in equilibrium and the distribution function is given only by the leading order term, which indeed has the FD form at base temperature. Noise generated by QPC. The non-symmetrized zero-frequency noise can be calculated using the Landauer-Buttiker scattering approach locally at QPC [52]. The noise of the outgoing current in channel ”out, 1” after QPC is found to be given by the sum of two terms

Sout11(0)/Gq=St(0)+Sg(0).(19)S_{\mathrm{out}}^{11}(0)/G_q = S_t(0) + S_g(0). \tag*{(19)}

The first term corresponds to the transmitted noise and has the form

St(0)=D2Tin+12R2(Tin+Tc),(20)S_t(0) = \mathcal{D}^2T_{\mathrm{in}}+\frac{1}{2}\mathcal{R}^2(T_{\mathrm{in}}+T_c), \tag*{(20)}

where D\mathcal{D} is the transparency of the QPC and R=1−D\mathcal{R}=1-\mathcal{D}. The first term is the contribution of incoming channel 3 and the second term is responsible for the noise coming after the ROC in the regime of charge fractionalizer (see Figure 1). One can see that in an almost equilibrium condition, Tc→TinT_c \to T_{\mathrm{in}}, the transmitted noise is given by the equilibrium base temperature, St(0)=TinS_t(0)=T_{\mathrm{in}}. The second term in Eq. (19) describes the noise generated by QPC due to mixing of the incoming states from ROC and channel 3. It is given by the following time integral [53]

Sg(0)/2RD∝∫dtKin,1neq(t)⋅Kin,3eq(t),(21)S_g(0)/2\mathcal{R}\mathcal{D} \propto\int dt K_{\mathrm{in},1}^{\mathrm{neq}}(t)\cdot K_{\mathrm{in},3}^{\mathrm{eq}}(t), \tag*{(21)}

where the first factor is given in Eq. (13) in the fractionalizer regime and the second multiplier corresponds to free fermions at temperature TinT_{\mathrm{in}} originating from the channel in ”3”, thus Kin,3eq(t)∼xTin/2sinh⁡(πTint−i0)K_{\mathrm{in},3}^{\mathrm{eq}}(t) \sim xT_{\mathrm{in}}/2\sinh(\pi T_{\mathrm{in}}t-i0).

For cold ROC, in the leading order of Tc/Tin→1T_c/T_{\mathrm{in}}\to1, one can replace the contact’s temperature TcT_c with TinT_{\mathrm{in}} and obtain the cross-term [53]

Sg(0)/2RD=Tin.(22)S_g(0)/2\mathcal{R}\mathcal{D}=T_{\mathrm{in}}. \tag*{(22)}

The sub-leading correction to Eq. (22) is proportional to Tc−Tin≪TinT_c-T_{\mathrm{in}}\ll T_{\mathrm{in}}. In case of a hot ROC at Tc/Tin→∞T_c/T_{\mathrm{in}}\to\infty, one can replace the finite-temperature TinT_{\mathrm{in}} correlation functions in the integrand of Eq. (21) with the zero temperature case and obtain the leading order behavior as follows [53]

Sg(0)/2RD=MTc,M≈0.4739.(23)S_g(0)/2\mathcal{R}\mathcal{D}=\mathcal{M}T_c,\qquad\mathcal{M}\approx0.4739. \tag*{(23)}

The sub-leading correction to Eq. (23) is proportional to Tc(Tin2/Tc2)≪TcT_c(T_{\mathrm{in}}^2/T_c^2)\ll T_c. In the case of FOC, the transmitted noise in Eq. (19) is given by Eq. (20). The noise generated by QPC has the following integral form [53]

Sg(0)/2RD∝∫dtKin,3eq(t)⋅Kfeq(t)⋅exp⁡[Φ(t)],(24)S_g(0)/2\mathcal{R}\mathcal{D}\propto\int dt K_{\mathrm{in},3}^{\mathrm{eq}}(t)\cdot K_f^{\mathrm{eq}}(t)\cdot\exp[\Phi(t)], \tag*{(24)}

where Φ(t)\Phi(t) is given in Eq. (17). The effect of Φ(t)\Phi(t) is perturbative and the main contribution to the integral, as in the case of the distribution function above, comes from times t≲RqCft\lesssim R_qC_f. First, we consider the case of cold Ohmic contacts, namely Tc→Tf→TT_c\to T_f\to T. In this case, one can replace all temperatures in leading order with TinT_{\mathrm{in}} and get [53]

Sg(0)/2RD=Tin+δTg(0),(25)S_g(0)/2\mathcal{R}D = T_{\mathrm{in}} + \delta T_g(0), \tag*{(25)}
δTg(0)=π26TinCflim⁡ω→0[Sin,1neq(ω)−Sfeq(ω)].\delta T_g(0) = \frac{\pi^2}{6T_{\mathrm{in}}C_f}\lim_{\omega\to0}\left[S_{\mathrm{in},1}^{\mathrm{neq}}(\omega)-S_f^{\mathrm{eq}}(\omega)\right].

In this case of hot Ohmic contacts Tc∼Tf≫TinT_c \sim T_f \gg T_{\mathrm{in}} one can replace the correlation function at temperature TinT_{\mathrm{in}} in the leading order with zero temperature case and get [53]

Sg(0)/2RD=Tfln⁡2+δT~g(0),(26)S_g(0)/2\mathcal{R}D = T_f\ln2+\delta\widetilde{T}_g(0), \tag*{(26)}
δT~g(0)=π24TfCflim⁡ω→0[Sin,1neq(ω)−Sfeq(ω)].\delta\widetilde{T}_g(0) = \frac{\pi^2}{4T_fC_f}\lim_{\omega\to0}\left[S_{\mathrm{in},1}^{\mathrm{neq}}(\omega)-S_f^{\mathrm{eq}}(\omega)\right].

In summary, we have investigated thermal relaxation at a quantum Hall edge, modeled by the FOC with a large capacitance. The initial nonequilibrium state is generated by the ROC in the charge-fractionalization regime, providing an alternative mechanism to QPC partitioning. Both, electron distribution function and the electric current noise are calculated. For a cold ROC, when the system is close to equilibrium and thermalization effects are negligible, the electron distribution function and the noise at the quantum Hall edge are described by the corresponding equilibrium expressions determined by the FD distribution. However, when thermalization effects become significant, the outgoing state immediately downstream of the ROC cannot be described by an equilibrium FD distribution and by Johnson–Nyquist noise at the ROC temperature, despite the conventional assumption of complete equilibration by the ROC. Furthermore, we calculated the electron distribution function and the current noise generated by a QPC placed downstream of the ROC and FOC. We found that the subleading corrections to both the electron distribution function and the QPC-generated noise contain a universal prefactor determined by the difference between the nonequilibrium noise generated by the ROC and the noise of a free chiral channel.

The author thanks Patrick R. Behr for careful reading of the manuscript and fruitful discussions. The author acknowledges financial support from the United Arab Emirates University under the Startup Grant No. G00004974, UPAR Grant G00005479, and UAEU-AUA Grant G00005522.

AppendixRelaxation.pdf

HEAT CURRENT IN CHIRAL EDGE MODE

Using the definition of heat flux in the main text, Jα=Rq⟨j˙α2(t)⟩/2J_{\alpha}=R_q\langle\dot{j}_{\alpha}^{2}(t)\rangle/2 and subtracting the contribution of the ground state, we get the following expression for the thermal current

Jα=Rq2∫dω2π[Sαeq(ω)−Sq(ω)]=∫0∞dω2πωeω/Tα−1=πTα212,α=in,c,f.(27)J_{\alpha}=\frac{R_q}{2}\int\frac{\mathrm{d}\omega}{2\pi}\left[S_{\alpha}^{\mathrm{eq}}(\omega)-S_q(\omega)\right]=\int_{0}^{\infty}\frac{\mathrm{d}\omega}{2\pi}\frac{\omega}{e^{\omega/T_{\alpha}}-1}=\frac{\pi T_{\alpha}^{2}}{12},\qquad\alpha=\mathrm{in},\mathrm{c},\mathrm{f}. \tag*{(27)}

This is the quantum heat flux and can be as well derived using FDT in equilibrium. If the channel is additionally biased, then according to Joules law, one needs to add the term proportional to square of electric current, namely V2/2RqV^{2}/2R_q. The Eq. (S-1) will be used below to calculate the thermalization effects in real and fictitious Ohmic contacts.

NON-EQUILIBRIUM FINITE FREQUENCY NOISE FOR REAL OHMIC CONTACT

We consider the real Ohmic contact (ROC) with capacitance CC as a source of heat current here. The solution to the outgoing current in channel ”1” in the frequency representation for the case of ROC is expressed in terms of the incoming currents and the Langevin sources

j1neq(ω)=jc,1(ω)+∑l=1,2T(ω)[jin,l(ω)−jc,l(ω)],T(ω)=1/(2−iωRqC).(28)j_{1}^{\mathrm{neq}}(\omega)=j_{\mathrm{c},1}(\omega)+\sum_{l=1,2}\mathcal{T}(\omega)\left[j_{\mathrm{in},l}(\omega)-j_{\mathrm{c},l}(\omega)\right],\qquad\mathcal{T}(\omega)=1/(2-\mathrm{i}\omega R_q C). \tag*{(28)}

The first term of the scattering coefficient T(ω)\mathcal{T}(\omega) in the denominator counts the number of incoming and outgoing channels with respect to the ROC. Next, using Eq. (S-2), one can calculate the finite frequency noise of the outgoing current right after the ROC, namely S1neq(ω)=R∫dteiωt⟨δj1neq(t)δj1neq(0)⟩S_{1}^{\mathrm{neq}}(\omega)=R\int\mathrm{d}t e^{\mathrm{i}\omega t}\langle\delta j_{1}^{\mathrm{neq}}(t)\delta j_{1}^{\mathrm{neq}}(0)\rangle and δj1neq(t)=j1neq(t)−⟨j1neq(t)⟩\delta j_{1}^{\mathrm{neq}}(t)=j_{1}^{\mathrm{neq}}(t)-\langle j_{1}^{\mathrm{neq}}(t)\rangle. After some algebra, the result is given by the following expression

S1neq(ω)/Gq=2∣T(ω)∣2[ω1−e−ω/Tin+ω1−e−ω/Tc]+[1−2Re⁡T(ω)]ω1−e−ω/Tc.(29)S_{1}^{\mathrm{neq}}(\omega)/G_q=2|\mathcal{T}(\omega)|^{2}\left[\frac{\omega}{1-e^{-\omega/T_{\mathrm{in}}}}+\frac{\omega}{1-e^{-\omega/T_{\mathrm{c}}}}\right]+\left[1-2\operatorname{Re}\mathcal{T}(\omega)\right]\frac{\omega}{1-e^{-\omega/T_{\mathrm{c}}}}. \tag*{(29)}

where the ROC temperature, TcT_{\mathrm{c}} is unknown and can be calculated using self-consistent energy balance equations, i.e., the conservation of the incoming and outgoing heat currents. When real Ohmic contact functions as fractionalizer, the scattering coefficient is T(ω)→1/2\mathcal{T}(\omega)\to1/2 at C→0C\to0. Then the finite frequency noise is equal to

S1neq(ω)/Gq=12[ω1−e−ω/Tin+ω1−e−ω/Tc]=Rq2[Sineq(ω)+Sceq(ω)].(30)S_{1}^{\mathrm{neq}}(\omega)/G_q=\frac{1}{2}\left[\frac{\omega}{1-e^{-\omega/T_{\mathrm{in}}}}+\frac{\omega}{1-e^{-\omega/T_{\mathrm{c}}}}\right]=\frac{R_q}{2}\left[S_{\mathrm{in}}^{\mathrm{eq}}(\omega)+S_{\mathrm{c}}^{\mathrm{eq}}(\omega)\right]. \tag*{(30)}

Thus, two incoming equilibrium noises are split into two parts due to two outgoing chiral channels from ROC. This result is given in Eq. (11) of the main text. We are not interested in the noise after the fictitious Ohmic contact (FOC), but similarly one can show that after the FOC, i.e. Sout,1(ω)=∣A(ω)∣2[S1neq(ω)+Sfeq(ω)]+[1−2Re⁡A(ω)]Sfeq(ω)S_{\mathrm{out},1}(\omega)=|A(\omega)|^{2}[S_{1}^{\mathrm{neq}}(\omega)+S_{\mathrm{f}}^{\mathrm{eq}}(\omega)]+[1-2\operatorname{Re}A(\omega)]S_{\mathrm{f}}^{\mathrm{eq}}(\omega), where A(ω)=1/(1−iωRqCf)A(\omega)=1/(1-\mathrm{i}\omega R_q C_{\mathrm{f}}) and Sfeq(ω)S_{\mathrm{f}}^{\mathrm{eq}}(\omega) is an equilibrium spectral function at FOC’s temperature TfT_{\mathrm{f}}.

DETERMINATION OF TEMPERATURE TcT_{\mathrm{c}} OF REAL OHMIC CONTACT

To find the temperature TcT_{\mathrm{c}}, we solve the equations of the energy balance self-consistently, i.e., conservation of the incoming and outgoing heat currents, where the heat current is defined by Joule’s law,

(Rq/2)∑i=1,2[⟨δjin,i(t)δjin,i(t)⟩+⟨jin,i(t)⟩2]=(Rq/2)∑i=1,2[⟨δjineq(t)δjineq(t)⟩+⟨jineq(t)⟩2].(31)(R_q/2)\sum_{i=1,2}\left[\langle\delta j_{\mathrm{in},i}(t)\delta j_{\mathrm{in},i}(t)\rangle+\langle j_{\mathrm{in},i}(t)\rangle^{2}\right]=(R_q/2)\sum_{i=1,2}\left[\langle\delta j_{i}^{\mathrm{neq}}(t)\delta j_{i}^{\mathrm{neq}}(t)\rangle+\langle j_{i}^{\mathrm{neq}}(t)\rangle^{2}\right]. \tag*{(31)}

where jineq(t)j_i^{\mathrm{neq}}(t) outgoing non-equilibrium current emitted from ROC. Using the above equation and the scattering matrix of currents for ROC, Eq. (S-2), we arrive at the following expression for determining contact’s temperature.

V122Rq+V122Rq+2Rq2∫dω2π[S1eq(ω)−Sq(ω)]−2Rq2∫dω2π2∣T(ω)∣2[Sineq(ω)−Sq(ω)]=2Rq2∫dω2π[Sceq(ω)−Sq(ω)]−2Rq2∫dω2π2∣T(ω)∣2[Sceq(ω)−Sq(ω)].(32)\begin{aligned} \frac{V_1^2}{2R_q}+\frac{V_1^2}{2R_q}+2\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left[S_1^{\mathrm{eq}}(\omega)-S_q(\omega)\right]-2\frac{R_q}{2}\int\frac{d\omega}{2\pi}2\lvert T(\omega)\rvert^2\left[S_{\mathrm{in}}^{\mathrm{eq}}(\omega)-S_q(\omega)\right] \\ &=2\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left[S_c^{\mathrm{eq}}(\omega)-S_q(\omega)\right]-2\frac{R_q}{2}\int\frac{d\omega}{2\pi}2\lvert T(\omega)\rvert^2\left[S_c^{\mathrm{eq}}(\omega)-S_q(\omega)\right]. \tag*{(32)} \end{aligned}

The first integrals on the left hand and right hand sides are given by Eq. (S-1). The second integral with capacitance can be rewritten using special functions as follows (α=in,c\alpha={\rm in},c)

−2Rq2∫dω2π2∣T(ω)∣2[Sαeq(ω)−Sq(ω)]=−2∫dω4π2(ωRqC)2+4[ω1−e−ω/Tα−ωθ(ω)]=2∫0∞dω4π11+(ωRqC/2)2ω1−e−ω/Tα=πTα212⋅F(πTαRqC/2),(33)\begin{aligned} -2\frac{R_q}{2}\int\frac{d\omega}{2\pi}2\lvert T(\omega)\rvert^2\left[S_\alpha^{\rm eq}(\omega)-S_q(\omega)\right]=-2\int\frac{d\omega}{4\pi}\frac{2}{(\omega R_qC)^2+4}\left[\frac{\omega}{1-e^{-\omega/T_\alpha}}-\omega\theta(\omega)\right] \\ &=2\int_0^\infty\frac{d\omega}{4\pi}\frac{1}{1+(\omega R_qC/2)^2}\frac{\omega}{1-e^{-\omega/T_\alpha}}=\frac{\pi T_\alpha^2}{12}\cdot\mathcal{F}(\pi T_\alpha R_qC/2), \tag*{(33)} \end{aligned}

where the dimensionless function equals the following expression

F(x)=3x2∫0∞dz1+z22z1−e−πz/x=3x2[x+log⁡(2x)+Ψ(1/2x)],(34)\mathcal{F}(x)=\frac{3}{x^2}\int_0^\infty\frac{dz}{1+z^2}\frac{2z}{1-e^{-\pi z/x}}=\frac{3}{x^2}\left[x+\log(2x)+\Psi(1/2x)\right], \tag*{(34)}

and Ψ(x)=dlog⁡Γ(x)/dx\Psi(x)=d\log\Gamma(x)/dx is a digamma function, and Γ(x)=∫0∞yx−1e−ydy\Gamma(x)=\int_0^\infty y^{x-1}e^{-y}dy is a gamma function. Further, choosing the symmetric bias and thus setting V1=−V2=VV_1=-V_2=V, we obtain the equation to determine the temperature of ROC

V2Rq+πTin212[2+F(πTinRqC/2)]=πTc212[2+F(πTcRqC/2)].(35)\frac{V^2}{R_q}+\frac{\pi T_{\rm in}^2}{12}\left[2+\mathcal{F}(\pi T_{\rm in}R_qC/2)\right]=\frac{\pi T_c^2}{12}\left[2+\mathcal{F}(\pi T_cR_qC/2)\right]. \tag*{(35)}

We are interested when ROC functions as a fractionalizer, i.e., C→0C\to0. In this case F(x)→−1\mathcal{F}(x)\to-1 at x→0x\to0 and we get the final result Tc2=Tin2+6V2/π2T_c^2=T_{\rm in}^2+6V^2/\pi^2 from Eq. (S-9), which is presented in Eq. (8) of the main text. At V=0V=0 the system is in equilibrium and at Tc=TinT_c=T_{\rm in}.

Determination of temperature TfT_f of fictitious Ohmic contact

In order to calculate the distribution function and noise further, one needs to take into account the thermalization effects inside the fictitious Ohmic contact (FOC). For the FOC it is given by conservation law, namely the incoming heat flux is equal to the outgoing heat flux. Using the definition of heat flux and the same steps as in the previous section, we arrive at the following equation

Rq2∫dω2π(S1neq(ω)−Sq(ω))=Rq2∫dω2π∣A(ω)∣2(S1neq(ω)−Sq(ω))+Rq2∫dω2π[1−∣A(ω)∣2](Sf,1eq(ω)−Sq(ω))(36)\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left(S_1^{\rm neq}(\omega)-S_q(\omega)\right)=\frac{R_q}{2}\int\frac{d\omega}{2\pi}\lvert A(\omega)\rvert^2\left(S_1^{\rm neq}(\omega)-S_q(\omega)\right)+\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left[1-\lvert A(\omega)\rvert^2\right]\left(S_{f,1}^{\rm eq}(\omega)-S_q(\omega)\right) \tag*{(36)}

where S1neq(ω)S_1^{\rm neq}(\omega) are given by Eq. (S-3) for ROC. Next, we find the temperature TfT_f in the case of small and large capacitance CfC_f. To find it, we can use the above equation Eq. (S-10)

Rq2∫dω2π[1−∣A(ω)∣2][Sin,1neq(ω)−Sq(ω)]=Rq2∫dω2π[1−∣A(ω)∣2][Sfeq(ω)−Sq(ω)].(37)\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left[1-\lvert A(\omega)\rvert^2\right]\left[S_{\rm in,1}^{\rm neq}(\omega)-S_q(\omega)\right]=\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left[1-\lvert A(\omega)\rvert^2\right]\left[S_f^{\rm eq}(\omega)-S_q(\omega)\right]. \tag*{(37)}

Using Eq. (S-1), the right hand side can be rewritten as follows

Rq2∫dω2π[1−∣A(ω)∣2][Sfeq(ω)−Sq(ω)]=πTf212+Rq2∫dω2π(−∣A(ω)∣2)[Sfeq(ω)−Sq(ω)].(38)\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left[1-\lvert A(\omega)\rvert^2\right]\left[S_f^{\rm eq}(\omega)-S_q(\omega)\right]=\frac{\pi T_f^2}{12}+\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left(-\lvert A(\omega)\rvert^2\right)\left[S_f^{\rm eq}(\omega)-S_q(\omega)\right]. \tag*{(38)}

After introducing the dimensionless variable z=ωRqCfz=\omega R_qC_f, the second term of the above equation takes the form

Rq2∫dω2π(−∣A(ω)∣2)[Sfeq(ω)−Sq(ω)]=πTf212⋅3(πTfRqCf)2∫0∞dz1+z22z1−e−z/TfRqCf=πTf212⋅F(πTfRqCf),(39)\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left(-\lvert A(\omega)\rvert^2\right)\left[S_f^{\rm eq}(\omega)-S_q(\omega)\right]=\frac{\pi T_f^2}{12}\cdot\frac{3}{(\pi T_fR_qC_f)^2}\int_0^\infty\frac{dz}{1+z^2}\frac{2z}{1-e^{-z/T_fR_qC_f}}=\frac{\pi T_f^2}{12}\cdot\mathcal{F}(\pi T_fR_qC_f), \tag*{(39)}

where F(x)\mathcal{F}(x) is given in Eq. (S-8). Thus, we obtain the following equation to determine the temperature of FOC

Rq2∫dω2π[1−∣A(ω)∣2][S1neq(ω)−Sq(ω)]=πTf212[1+F(πTfRqCf)].(40)\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left[1-\lvert A(\omega)\rvert^2\right]\left[S_1^{\rm neq}(\omega)-S_q(\omega)\right]=\frac{\pi T_f^2}{12}\left[1+\mathcal{F}(\pi T_fR_qC_f)\right]. \tag*{(40)}

For the case of ROC as a non-equilibrium source using in the fractionalizer regime, Eq. (S-4) we get the following expression for the left hand side of Eq. (S-11)

Rq2∫dω2π[1−∣A(ω)∣2][S1neq(ω)−Sq(ω)]=12⋅πTin212⋅[1+F(πTinRqCf)]+12⋅πTc212⋅[1+F(πTcRqCf)].(41)\frac{R_q}{2}\int\frac{d\omega}{2\pi}\left[1-\lvert A(\omega)\rvert^2\right]\left[S_1^{\mathrm{neq}}(\omega)-S_q(\omega)\right]=\frac{1}{2}\cdot\frac{\pi T_{\mathrm{in}}^2}{12}\cdot\left[1+\mathcal{F}(\pi T_{\mathrm{in}}R_qC_f)\right]+\frac{1}{2}\cdot\frac{\pi T_c^2}{12}\cdot\left[1+\mathcal{F}(\pi T_cR_qC_f)\right]. \tag*{(41)}

Therefore, the final equation for the determination of the temperature of FOC takes the following form

12⋅πTin212⋅[1+F(πTinRqCf)]+12⋅πTc212⋅[1+F(πTcRqCf)]=πTf212⋅[1+F(πTfRqCf)].(42)\frac{1}{2}\cdot\frac{\pi T_{\mathrm{in}}^2}{12}\cdot\left[1+\mathcal{F}(\pi T_{\mathrm{in}}R_qC_f)\right]+\frac{1}{2}\cdot\frac{\pi T_c^2}{12}\cdot\left[1+\mathcal{F}(\pi T_cR_qC_f)\right]=\frac{\pi T_f^2}{12}\cdot\left[1+\mathcal{F}(\pi T_fR_qC_f)\right]. \tag*{(42)}

At large capacitance Cf→∞C_f\to\infty, one can use F(x)≈−3/x\mathcal{F}(x)\approx-3/x for x≫1x\gg1, thus the temperature of the FOC is given by Tf2=(Tin2+Tc2)/2T_f^2=(T_{\mathrm{in}}^2+T_c^2)/2. At small capacitance Cf→0C_f\to0, one can use F(x)≈−1+2x2/5\mathcal{F}(x)\approx-1+2x^2/5 for x≪1x\ll1 and the temperature of the FOC is given by Tf4=(Tin4+Tc4)/2T_f^4=(T_{\mathrm{in}}^4+T_c^4)/2. Our interest is the large capacitance case, which is discussed in the main text, Eq. (9).

CALCULATION OF INITIAL DISTRIBUTION FUNCTION FOR HOT ROC

According to Eqs. (2) and (9) of the main text, the initial non-equilibrium distribution function after ROC has the following form

fin,1neq(ϵ)=∫dt2πie−iϵtπTinsinh⁡(πTint−i0)πTcsinh⁡(πTct−i0).(43)f_{\mathrm{in},1}^{\mathrm{neq}}(\epsilon)=\int\frac{dt}{2\pi i}e^{-i\epsilon t}\sqrt{\frac{\pi T_{\mathrm{in}}}{\sinh(\pi T_{\mathrm{in}}t-i0)}}\sqrt{\frac{\pi T_c}{\sinh(\pi T_ct-i0)}}. \tag*{(43)}

where TcT_c is given in Eq. (S-9). Calculations for the case of ”cold” ROC, Tc/Tin→1T_c/T_{\mathrm{in}}\to1, are straightforward, and since the system is almost in equilibrium, the main contribution is given by equilibrium FD distribution function (See Eq. (14) in the main text). In the regime of hot ROC, Tc/Tin→∞T_c/T_{\mathrm{in}}\to\infty and low energies ϵ/Tc≪1\epsilon/T_c\ll1 (for definiteness, positive values of ϵ>0\epsilon>0 are considered), we introduce the dimensionless variable x=πTctx=\pi T_ct, expand the oscillatory exponent and replace the correlation function with TinT_{\mathrm{in}} by the zero temperature case, namely

fin,1neq(ϵ)=∫dx2πie−iϵπTcxsinh⁡[x−i0]Tin/Tcsinh⁡[(Tin/Tc)(x−i0)]≈∫dx2πi1+−iϵπTcxx−i0sinh⁡[x−i0].(44)f_{\mathrm{in},1}^{\mathrm{neq}}(\epsilon)=\int\frac{dx}{2\pi i}\frac{e^{-i\frac{\epsilon}{\pi T_c}x}}{\sqrt{\sinh[x-i0]}}\frac{\sqrt{T_{\mathrm{in}}/T_c}}{\sqrt{\sinh[(T_{\mathrm{in}}/T_c)(x-i0)]}}\approx\int\frac{dx}{2\pi i}\frac{1+\frac{-i\epsilon}{\pi T_c}x}{\sqrt{x-i0}\sqrt{\sinh[x-i0]}}. \tag*{(44)}

Next, keeping the leading order terms, we get the result for initial non-equilibrium distribution function at low energies

fin,1neq(ϵ)=12−Cϵ4Tc,C=4π2∫0∞x dxsinh⁡x≈1.5573>1.(45)f_{\mathrm{in},1}^{\mathrm{neq}}(\epsilon)=\frac{1}{2}-\frac{C\epsilon}{4T_c},\qquad C=\frac{4}{\pi^2}\int_0^\infty\frac{\sqrt{x}\,dx}{\sqrt{\sinh x}}\approx1.5573>1. \tag*{(45)}

This result is given in Eq. (15) of the main text. Now, for the case of high energies, ϵ/Tc≫1\epsilon/T_c\gg1 the oscillatory term is essential and cannot be expanded in Eq. (S-18). Therefore, in the leading order of Tc/Tin→∞T_c/T_{\mathrm{in}}\to\infty the form of the integral is

fin,1neq(ϵ)≈∫dx2πie−iϵπTcxsinh⁡(x−i0)x−i0.(46)f_{\mathrm{in},1}^{\mathrm{neq}}(\epsilon)\approx\int\frac{dx}{2\pi i}\frac{e^{-i\frac{\epsilon}{\pi T_c}x}}{\sqrt{\sinh(x-i0)}\sqrt{x-i0}}. \tag*{(46)}

To estimate this integral, we first shift the integration variable x=u−iπ/2x=u-i\pi/2 and get

fin,1neq(ϵ)=e−ϵ2Tc∫du2πe−iϵπTcucosh⁡u112+iu=∫dzπe−πz22∫du2πe−i(ϵπTc+z2)ucosh⁡u.(47)f_{\mathrm{in},1}^{\mathrm{neq}}(\epsilon)=e^{-\frac{\epsilon}{2T_c}}\int\frac{du}{2\pi}\frac{e^{-i\frac{\epsilon}{\pi T_c}u}}{\sqrt{\cosh u}}\frac{1}{\sqrt{\frac{1}{2}+iu}}=\int\frac{dz}{\sqrt{\pi}}e^{-\frac{\pi z^2}{2}}\int\frac{du}{2\pi}\frac{e^{-i\left(\frac{\epsilon}{\pi T_c}+z^2\right)u}}{\sqrt{\cosh u}}. \tag*{(47)}

In the above equation, the inner integration with respect to the variable uu is given by a gamma function and has the form

fin,1neq(ϵ)=1πe−ϵ2Tc∫dzπe−πz2212π∣Γ[1/4+i(ϵ/πTc+z2)/2]∣2.(48)f_{\mathrm{in},1}^{\mathrm{neq}}(\epsilon)=\frac{1}{\pi}e^{-\frac{\epsilon}{2T_c}}\int\frac{dz}{\sqrt{\pi}}e^{-\frac{\pi z^2}{2}}\frac{1}{\sqrt{2\pi}}\left\lvert\Gamma\left[1/4+i\left(\epsilon/\pi T_c+z^2\right)/2\right]\right\rvert^2. \tag*{(48)}

Using further ∣Γ(x+iy)∣2≃2πy2x−1e−πy\lvert\Gamma(x+iy)\rvert^2\simeq2\pi y^{2x-1}e^{-\pi y} at y→+∞y\to+\infty we arrive at the following result

fin,1neq(ϵ)=1πe−ϵTc∫e−πz2 dzz2+ϵ/πTc=1πe−ϵTcK0(ϵ/2Tc),(49)f_{\mathrm{in},1}^{\mathrm{neq}}(\epsilon)=\frac{1}{\pi}e^{-\frac{\epsilon}{T_c}}\int\frac{e^{-\pi z^2}\,dz}{\sqrt{z^2+\epsilon/\pi T_c}}=\frac{1}{\pi}e^{-\frac{\epsilon}{T_c}}K_0(\epsilon/2T_c), \tag*{(49)}

where K0(x)K_0(x) is the modified Bessel function of the second kind. Taking into account that K0(x)≃π/2xe−xK_0(x)\simeq\sqrt{\pi/2x}e^{-x} for x≫1x\gg1 we arrive at Eq. (15) from the main text.

Calculation of distribution function after FOC

The distribution function after the FOC can be written as follows

fout,1(ϵ)=∫dt2πe−i(ϵ−2πIin)t⋅Kfeq(t)⋅eΦ(t),(50)f_{\mathrm{out},1}(\epsilon)=\int\frac{dt}{2\pi}e^{-i(\epsilon-2\pi I_{\mathrm{in}})t}\cdot K_{\mathrm{f}}^{\mathrm{eq}}(t)\cdot e^{\Phi(t)}, \tag*{(50)}

where Iin=⟨j1neq⟩=0I_{\mathrm{in}}=\langle j_{1}^{\mathrm{neq}}\rangle=0 is an incoming non-equilibrium current after ROC and is equal to zero since the real Ohmic contact is not biased V1+V2=0V_{1}+V_{2}=0, and Kfeq(t)∝−iTf/sinh⁡(πTft−i0)K_{\mathrm{f}}^{\mathrm{eq}}(t)\propto-iT_{\mathrm{f}}/\sinh(\pi T_{\mathrm{f}}t-i0) is a two-point free-fermion correlation function. The second factor is given by

Φ(t)=Rq∫dωe−iωt−1ω2+η2S1neq(ω)−Sfeq(ω)1+(ωRqCf)2,η→+0,(51)\Phi(t)=R_{\mathrm{q}}\int d\omega\frac{e^{-i\omega t}-1}{\omega^{2}+\eta^{2}}\frac{S_{1}^{\mathrm{neq}}(\omega)-S_{\mathrm{f}}^{\mathrm{eq}}(\omega)}{1+(\omega R_{\mathrm{q}}C_{\mathrm{f}})^{2}},\qquad\eta\to+0, \tag*{(51)}

One can see that Φ∗(t)=Φ(t)=Φ(−t)\Phi^{*}(t)=\Phi(t)=\Phi(-t) and note that at Tin=Tc=TfT_{\mathrm{in}}=T_{\mathrm{c}}=T_{\mathrm{f}}, Φ(t)=0\Phi(t)=0, and we recover the equilibrium Fermi-Dirac distribution. We are interested in large capacitance FOC at Cf→∞C_{\mathrm{f}}\to\infty, thus the effect of Φ(t)\Phi(t) is perturbative and expanding the exponent in Eq. (S-24) we have

fout,1(ϵ)≈f1(ϵ)+f2(ϵ).(52)f_{\mathrm{out},1}(\epsilon)\approx f_{1}(\epsilon)+f_{2}(\epsilon). \tag*{(52)}

The first term in Eq. (S-26) is given by the equilibrium FD distribution function at temperature TfT_{\mathrm{f}}

f1(ϵ)=∫dt2πe−iϵt⋅Kfeq(t)=fFD(ϵ/Tf)=11+eϵ/Tf.(53)f_{1}(\epsilon)=\int\frac{dt}{2\pi}e^{-i\epsilon t}\cdot K_{\mathrm{f}}^{\mathrm{eq}}(t)=f_{\mathrm{FD}}(\epsilon/T_{\mathrm{f}})=\frac{1}{1+e^{\epsilon/T_{\mathrm{f}}}}. \tag*{(53)}

The second term is given by

f2(ϵ)=Rq∫dt2πe−iϵt⋅Kfeq(t)⋅Φ(t)=∫dt2πe−iϵt⋅Kfeq(t)∫dωe−iωt−1ω2+η2S1neq(ω)−Sfeq(ω)1+(ωRqCf)2.(54)f_{2}(\epsilon)=R_{\mathrm{q}}\int\frac{dt}{2\pi}e^{-i\epsilon t}\cdot K_{\mathrm{f}}^{\mathrm{eq}}(t)\cdot\Phi(t)=\int\frac{dt}{2\pi}e^{-i\epsilon t}\cdot K_{\mathrm{f}}^{\mathrm{eq}}(t)\int d\omega\frac{e^{-i\omega t}-1}{\omega^{2}+\eta^{2}}\frac{S_{1}^{\mathrm{neq}}(\omega)-S_{\mathrm{f}}^{\mathrm{eq}}(\omega)}{1+(\omega R_{\mathrm{q}}C_{\mathrm{f}})^{2}}. \tag*{(54)}

At large capacitance Cf→∞C_{\mathrm{f}}\to\infty, the main contribution to the integral comes from finite times t≲RqCft\lesssim R_{\mathrm{q}}C_{\mathrm{f}} and small frequencies ωt≪1\omega t\ll1, where the fraction 1/(1+(ωRqCf)2)1/(1+(\omega R_{\mathrm{q}}C_{\mathrm{f}})^{2}) can be replaced by the delta function (π/RqCf)δ(ω)(\pi/R_{\mathrm{q}}C_{\mathrm{f}})\delta(\omega). Therefore, we get

f2(ϵ)=−π2Cflim⁡ω→0[S1neq(ω)−Sfeq(ω)]∫dt2πe−iϵtt2Kfeq(t)=π2Cflim⁡ω→0[S1neq(ω)−Sfeq(ω)]⋅∂2∂ϵ2fFD(ϵ/Tf).(55)f_{2}(\epsilon)=-\frac{\pi^{2}}{C_{\mathrm{f}}}\lim_{\omega\to0}\left[S_{1}^{\mathrm{neq}}(\omega)-S_{\mathrm{f}}^{\mathrm{eq}}(\omega)\right]\int\frac{dt}{2\pi}e^{-i\epsilon t}t^{2}K_{\mathrm{f}}^{\mathrm{eq}}(t)=\frac{\pi^{2}}{C_{\mathrm{f}}}\lim_{\omega\to0}\left[S_{1}^{\mathrm{neq}}(\omega)-S_{\mathrm{f}}^{\mathrm{eq}}(\omega)\right]\cdot\frac{\partial^{2}}{\partial\epsilon^{2}}f_{\mathrm{FD}}(\epsilon/T_{\mathrm{f}}). \tag*{(55)}

This result is given in Eq. (18) of the main text. The prefactor can be estimated using Eq. (S-3) for ROC. That is,

lim⁡ω→0[S1neq(ω)−Sfeq(ω)]=(Gq/2)(Tin+Tc−2Tf).(56)\lim_{\omega\to0}\left[S_{1}^{\mathrm{neq}}(\omega)-S_{\mathrm{f}}^{\mathrm{eq}}(\omega)\right]=(G_{\mathrm{q}}/2)(T_{\mathrm{in}}+T_{\mathrm{c}}-2T_{\mathrm{f}}). \tag*{(56)}

The above two results are presented in Eq. (18) of the main text.

QPC-generated noise right after the ROC

Using Landauer-Buttiker scattering approach, the non-symmetrized zero frequency noise (at zero frequency the non-symmetrized and symmetrized noises coincide) in the channel 1 right after the QPC is given by

Sout11(ω=0)/Gq)=D2P33+R2P11(ω=0)+2RDP13(ω=0).(57)S_{\mathrm{out}}^{11}(\omega=0)/G_{\mathrm{q})}=D^{2}\mathcal{P}_{33}+R^{2}\mathcal{P}_{11}(\omega=0)+2R D\mathcal{P}_{13}(\omega=0). \tag*{(57)}

In the above Eq. (S-31) the first and second terms are given by the FDT theorem; therefore,

P33(ω=0)∝∫dt⟨δjin,3(t)δjin,3(0)⟩=Tin,P11(ω=0)∝∫dt⟨δj1neq(t)δj1neq(0)⟩=(Tin+Tc)/2.(58)\mathcal{P}_{33}(\omega=0)\propto\int dt\langle\delta j_{\mathrm{in},3}(t)\delta j_{\mathrm{in},3}(0)\rangle=T_{\mathrm{in}},\qquad\mathcal{P}_{11}(\omega=0)\propto\int dt\langle\delta j_{1}^{\mathrm{neq}}(t)\delta j_{1}^{\mathrm{neq}}(0)\rangle=(T_{\mathrm{in}}+T_{\mathrm{c}})/2. \tag*{(58)}

The last term of Eq. (S-31) is given by the following time integral

P13(ω=0)∝∫dt⟨(ψ1neq)†(t)ψ1neq(0)⟩⟨ψin,3(t)ψin,3†(0)⟩=∫dtKin,1neq(t)⋅Kin,3eq(t).(59)\mathcal{P}_{13}(\omega=0)\propto\int dt\langle(\psi_{1}^{\mathrm{neq}})^{\dagger}(t)\psi_{1}^{\mathrm{neq}}(0)\rangle\langle\psi_{\mathrm{in},3}(t)\psi_{\mathrm{in},3}^{\dagger}(0)\rangle=\int dt K_{\mathrm{in},1}^{\mathrm{neq}}(t)\cdot K_{\mathrm{in},3}^{\mathrm{eq}}(t). \tag*{(59)}

Here, the second multiplier corresponds to free fermions at temperature TinT_{\mathrm{in}} originating from the channel ”in, 3”, thus Kin,3eq(t)∝−iTin/2sinh⁡(πTint−i0)K^{\mathrm{eq}}_{\mathrm{in,3}}(t) \propto-iT_{\mathrm{in}}/2\sinh(\pi T_{\mathrm{in}}t-i0). The first multiplier can be presented as follows

K1neq(t)∝⟨e2πi∫−∞tj1neq(t′)eηt′ dt′e−2πi∫−∞0j1neq(t′)eηt′ dt′⟩=exp⁡[M1neq(t)].(60)K^{\mathrm{neq}}_{1}(t) \propto\left\langle e^{2\pi i\int_{-\infty}^{t}j^{\mathrm{neq}}_{1}(t')e^{\eta t'}\,dt'}e^{-2\pi i\int_{-\infty}^{0}j^{\mathrm{neq}}_{1}(t')e^{\eta t'}\,dt'}\right\rangle=\exp[M^{\mathrm{neq}}_{1}(t)]. \tag*{(60)}

After some algebra, one can rewrite it as follows

M1neq(t)=iRq∫0tdt′⟨j1neq(t′)⟩+Rq∫dωe−iωt−1ω2+η2S1neq(ω),η→+0.(61)M^{\mathrm{neq}}_{1}(t)=iR_q\int_{0}^{t}dt'\langle j^{\mathrm{neq}}_{1}(t')\rangle+R_q\int d\omega\frac{e^{-i\omega t}-1}{\omega^2+\eta^2}S^{\mathrm{neq}}_{1}(\omega),\qquad\eta\to+0. \tag*{(61)}

Here, the first term is zero because the outgoing state of the ROC is not biased; therefore, we have

M1neq(t)=Rq∫dωe−iωt−1ω2+η2Sin,1neq(ω),(62)M^{\mathrm{neq}}_{1}(t)=R_q\int d\omega\frac{e^{-i\omega t}-1}{\omega^2+\eta^2}S^{\mathrm{neq}}_{\mathrm{in,1}}(\omega), \tag*{(62)}

where S1neq(ω)S^{\mathrm{neq}}_{1}(\omega) is given in Eq. (S-4). We are interested in the limit of C→0C\to0 (fractionalizer); thus, using Eq. (S-4) one gets

M1neq(t)=∫dωe−iωt−1ω2+η2[ω/21−e−ω/Tin+ω/21−e−ω/Tc]=log⁡[−iTin2sinh⁡(πTint−i0)−iTc2sinh⁡(πTct−i0)].(63)M^{\mathrm{neq}}_{1}(t)=\int d\omega\frac{e^{-i\omega t}-1}{\omega^2+\eta^2}\left[\frac{\omega/2}{1-e^{-\omega/T_{\mathrm{in}}}}+\frac{\omega/2}{1-e^{-\omega/T_c}}\right]=\log\left[\sqrt{\frac{-iT_{\mathrm{in}}}{2\sinh(\pi T_{\mathrm{in}}t-i0)}}\sqrt{\frac{-iT_c}{2\sinh(\pi T_ct-i0)}}\right]. \tag*{(63)}

After substitution into Eq. (S-33), we finally get the following integral

P13(ω=0)=∫dt2π−iπTinsinh⁡(πTint−i0)−iπTcsinh⁡(πTct−i0)−iπTinsinh⁡(πTint−i0).(64)\mathcal{P}_{13}(\omega=0)=\int\frac{dt}{2\pi}\sqrt{\frac{-i\pi T_{\mathrm{in}}}{\sinh(\pi T_{\mathrm{in}}t-i0)}}\sqrt{\frac{-i\pi T_c}{\sinh(\pi T_ct-i0)}}\frac{-i\pi T_{\mathrm{in}}}{\sinh(\pi T_{\mathrm{in}}t-i0)}. \tag*{(64)}

For cold ROC in the leading order of Tc/Tin→1T_c/T_{\mathrm{in}}\to1, one can replace the contact’s temperature TcT_c with TinT_{\mathrm{in}}, and introducing the dimensionless variable x=πTintx=\pi T_{\mathrm{in}}t we obtain the following.

P13(ω=0)=Tin∫(−1/2)dxsinh⁡2(x−i0)=Tin.(65)\mathcal{P}_{13}(\omega=0)=T_{\mathrm{in}}\int\frac{(-1/2)dx}{\sinh^2(x-i0)}=T_{\mathrm{in}}. \tag*{(65)}

The sub-leading correction is proportional to Tc−Tin≪TinT_c-T_{\mathrm{in}}\ll T_{\mathrm{in}}. In case of hot ROC at Tc/Tin→∞T_c/T_{\mathrm{in}}\to\infty, one can replace the finite-temperature TinT_{\mathrm{in}} correlation functions in the integrand of Eq. (S-38) with zero temperature case and introducing the dimensionless variable x=πTctx=\pi T_ct we obtain the leading order behavior

P13(0)=MTc,M=∫(−1/2)dx(x−i0)3/2sinh⁡(x−i0)≈0.4739.(66)\mathcal{P}_{13}(0)=\mathcal{M}T_c,\qquad\mathcal{M}=\int\frac{(-1/2)dx}{(x-i0)^{3/2}\sqrt{\sinh(x-i0)}}\approx0.4739. \tag*{(66)}

The sub-leading correction is proportional to Tc(Tin2/Tc2)≪TcT_c(T_{\mathrm{in}}^2/T_c^2)\ll T_c.

QPC-GENERATED NOISE RIGHT AFTER THE FOC

In case of FOC, the first two terms P11(ω=0)\mathcal{P}_{11}(\omega=0) and P33(ω=0)\mathcal{P}_{33}(\omega=0) are given by Eq. (S-32). The last term of noise has the following integral form

P13(ω=0)∝∫dt2πKin,3eq(t)⋅Kfeq(t)⋅eΦ(t)=∫dt2π−iπTinsinh⁡(πTint−i0)×−iπTfsinh⁡(πTft−i0)×eΦ(t)(67)\mathcal{P}_{13}(\omega=0)\propto\int\frac{dt}{2\pi}K^{\mathrm{eq}}_{\mathrm{in,3}}(t)\cdot K^{\mathrm{eq}}_{\mathrm{f}}(t)\cdot e^{\Phi(t)}=\int\frac{dt}{2\pi}\frac{-i\pi T_{\mathrm{in}}}{\sinh(\pi T_{\mathrm{in}}t-i0)}\times\frac{-i\pi T_{\mathrm{f}}}{\sinh(\pi T_{\mathrm{f}}t-i0)}\times e^{\Phi(t)} \tag*{(67)}

where Φ(t)\Phi(t) is given in Eq. (S-25). At large frequencies of FOC the effect of Φ(t)\Phi(t) is perturbative; therefore, the leading two terms are given by P13(ω=0)=I1+I2\mathcal{P}_{13}(\omega=0)=\mathcal{I}_1+\mathcal{I}_2 and

I1=∫dt2π−iπTinsinh⁡(πTint−i0)−iπTfsinh⁡(πTft−i0),I2=∫dt2π−iπTinsinh⁡(πTint−i0)−iπTfsinh⁡(πTft−i0)×Φ(t).(68)\mathcal{I}_1=\int\frac{dt}{2\pi}\frac{-i\pi T_{\mathrm{in}}}{\sinh(\pi T_{\mathrm{in}}t-i0)}\frac{-i\pi T_{\mathrm{f}}}{\sinh(\pi T_{\mathrm{f}}t-i0)},\qquad\mathcal{I}_2=\int\frac{dt}{2\pi}\frac{-i\pi T_{\mathrm{in}}}{\sinh(\pi T_{\mathrm{in}}t-i0)}\frac{-i\pi T_{\mathrm{f}}}{\sinh(\pi T_{\mathrm{f}}t-i0)}\times\Phi(t). \tag*{(68)}

First, we consider the case of cold Ohmic contacts, namely Tc∼Tf→TT_c\sim T_{\mathrm{f}}\to T. In this case, we can replace TfT_{\mathrm{f}} in the first term I1\mathcal{I}_1 with TinT_{\mathrm{in}} and introducing the dimensionless variable of integration x=πTinx=\pi T_{\mathrm{in}} we obtain the following.

I1=Tin∫(−1/2)dxsinh⁡2(x−i0)=Tin.(69)\mathcal{I}_1=T_{\mathrm{in}}\int\frac{(-1/2)dx}{\sinh^2(x-i0)}=T_{\mathrm{in}}. \tag*{(69)}

The sub-leading correction is proportional to Tf−Tin≪TinT_{\mathrm{f}}-T_{\mathrm{in}}\ll T_{\mathrm{in}}. The second integral can be estimated as in Sec. F and after some algebra we obtain the following result,

I2=1TinCflim⁡ω→0[S1neq(ω)−Sfeq(ω)]∫x2dx2sinh⁡2(x−i0)=π212TinRqCf(Tc+Tin−2Tf).(70)\mathcal{I}_2=\frac{1}{T_{\mathrm{in}}C_{\mathrm{f}}}\lim_{\omega\to0}\left[S_1^{\mathrm{neq}}(\omega)-S_{\mathrm{f}}^{\mathrm{eq}}(\omega)\right]\int\frac{x^2dx}{2\sinh^2(x-i0)}=\frac{\pi^2}{12T_{\mathrm{in}}R_{\mathrm{q}}C_{\mathrm{f}}}(T_{\mathrm{c}}+T_{\mathrm{in}}-2T_{\mathrm{f}}). \tag*{(70)}

In this case of hot Ohmic contacts Tc∼Tf≫TinT_{\mathrm{c}}\sim T_{\mathrm{f}}\gg T_{\mathrm{in}} one can replace the correlation function at temperature TinT_{\mathrm{in}} in the leading order with zero temperature case and introducing the dimensionless variable of integration, x=πtTfx=\pi tT_{\mathrm{f}}, we get

I1=Tf∫(−1/2)dy(y−i0)sinh⁡(y−i0)=Tfln⁡2,(71)\mathcal{I}_1=T_{\mathrm{f}}\int\frac{(-1/2)dy}{(y-i0)\sinh(y-i0)}=T_{\mathrm{f}}\ln2, \tag*{(71)}

and the sub-leading correction is proportional to Tf(Tin2/Tf2)≪TfT_{\mathrm{f}}(T_{\mathrm{in}}^2/T_{\mathrm{f}}^2)\ll T_{\mathrm{f}}. The second integral is given by

I2=1TfCflim⁡ω→0[S1neq(ω)−Sfeq(ω)]∫x2dx2(x−i0)sinh⁡(x−i0)=π28TinRqCf(Tc+Tin−2Tf).(72)\mathcal{I}_2=\frac{1}{T_{\mathrm{f}}C_{\mathrm{f}}}\lim_{\omega\to0}\left[S_1^{\mathrm{neq}}(\omega)-S_{\mathrm{f}}^{\mathrm{eq}}(\omega)\right]\int\frac{x^2dx}{2(x-i0)\sinh(x-i0)}=\frac{\pi^2}{8T_{\mathrm{in}}R_{\mathrm{q}}C_{\mathrm{f}}}(T_{\mathrm{c}}+T_{\mathrm{in}}-2T_{\mathrm{f}}). \tag*{(72)}

References

  1. [1]S. Tewari, P. Rouleau, C. Grenier, F. Portier, A. Cavanna, U. Gennser, D. Mailly, and P. Roche, Phys. Rev. B 93, 035420 (2016).
  2. [2]R. Bisognin, A. Marguerite, B. Roussel, M. Kumar, C. Cabart, C. Chapdelaine, A. Mohammad-Djafari, J.-M. Berroir, E. Bocquillon, B. Plaçais, A. Cavanna, U. Gennser, Y. Jin, P. Degiovanni, and G. Fève, Nature Communications 10, 3379 (2019).
  3. [3]H. Duprez, E. Sivre, A. Anthore, A. Aassime, A. Cavanna, A. Ouerghi, U. Gennser, and F. Pierre, Phys. Rev. X 9, 021030 (2019).
  4. [4]R. Bhattacharyya, M. Banerjee, M. Heiblum, D. Mahalu, and V. Umansky, Phys. Rev. Lett. 122, 246801 (2019).
  5. [5]J. Nakamura, S. Liang, G. C. Gardner, and M. J. Manfra, Nature Physics 16, 931 (2020).
  6. [6]S. Jezouin, Z. Iftikhar, A. Anthore, F. D. Parmentier, U. Gennser, A. Cavanna, A. Ouerghi, I. P. Levkivskyi, E. Idrisov, E. V. Sukhorukov, L. I. Glazman, and F. Pierre, Nature 536, 58 (2016).
  7. [7]E. G. Idrisov, I. P. Levkivskyi, and E. V. Sukhorukov, Phys. Rev. B 96, 155408 (2017).
  8. [8]A. O. Slobodeniuk, I. P. Levkivskyi, and E. V. Sukhorukov, Phys. Rev. B 88, 165307 (2013).
  9. [9]E. Sivre, A. Anthore, F. D. Parmentier, A. Cavanna, U. Gennser, A. Ouerghi, Y. Jin, and F. Pierre, Nature Physics 14, 145 (2018).
  10. [10]C. Lin, M. Hashisaka, T. Akiho, K. Muraki, and T. Fujisawa, Nature Communications 12, 131 (2021).
  11. [11]E. Sivre, H. Duprez, A. Anthore, A. Aassime, F. D. Parmentier, A. Cavanna, A. Ouerghi, U. Gennser, and F. Pierre, Nature Communications 10, 5638 (2019).
  12. [12]C. Altimiras, H. le Sueur, U. Gennser, A. Cavanna, D. Mailly, and F. Pierre, Nature Physics 6, 34 (2010).
  13. [13]D. L. Kovrizhin and J. T. Chalker, Phys. Rev. Lett. 109, 106403 (2012).
  14. [14]I. P. Levkivskyi and E. V. Sukhorukov, Phys. Rev. B 85, 075309 (2012).
  15. [15]A. O. Slobodeniuk, E. G. Idrisov, and E. V. Sukhorukov, Phys. Rev. B 93, 035421 (2016).
  16. [16]K. Itoh, R. Nakazawa, T. Ota, M. Hashisaka, K. Muraki, and T. Fujisawa, Phys. Rev. Lett. 120, 197701 (2018).
  17. [17]C. Lin, R. Eguchi, M. Hashisaka, T. Akiho, K. Muraki, and T. Fujisawa, Phys. Rev. B 99, 195304 (2019).
  18. [18]R. H. Rodriguez, F. D. Parmentier, D. Ferraro, P. Roulleau, U. Gennser, A. Cavanna, M. Sassetti, F. Portier, D. Mailly, and P. Roche, Nature Communications 11, 2426 (2020).
  19. [19]C. Grenier, R. Hervé, G. Fève, and P. Degiovanni, Modern Physics Letters B 25, 1053 (2011).
  20. [20]T. Jullien, P. Roulleau, B. Roche, A. Cavanna, Y. Jin, and D. C. Glattli, Nature 514, 603 (2014).
  21. [21]C. Bäuerle, D. C. Glattli, T. Meunier, F. Portier, P. Roche, P. Roulleau, S. Takada, and X. Waintal, Reports on Progress in Physics 81, 056503 (2018).
  22. [22]J. D. Fletcher, N. Johnson, E. Locane, P. See, J. P. Griffiths, I. Farrer, D. A. Ritchie, P. W. Brouwer, V. Kashcheyevs, and M. Kataoka, Nature Communications 10, 5298 (2019).
  23. [23]C. Altimiras, H. le Sueur, U. Gennser, A. Cavanna, D. Mailly, and F. Pierre, Phys. Rev. Lett. 105, 226804 (2010).
  24. [24]H. le Sueur, C. Altimiras, U. Gennser, A. Cavanna, D. Mailly, and F. Pierre, Phys. Rev. Lett. 105, 056803 (2010).
  25. [25]I. P. Levkivskyi, Mesoscopic Quantum Hall Effect, 1st ed. (Springer-Verlag Berlin Heidelberg, 2012).DOI
  26. [26]S. Jezouin, F. D. Parmentier, A. Anthore, U. Gennser, A. Cavanna, Y. Jin, and F. Pierre, Science 342, 601 (2013).
  27. [27]Z. Iftikhar, Charge Quantization and Kondo Quantum Criticality in Few-Channel Mesoscopic Circuits, 1st ed. (Springer International Publishing, 2018).
  28. [28]A. A. Clerk, P. W. Brouwer, and V. Ambegaokar, Phys. Rev. Lett. 87, 186801 (2001).
  29. [29]E. G. Idrisov, I. P. Levkivskyi, and E. V. Sukhorukov, Phys. Rev. Lett. 121, 026802 (2018).
  30. [30]H. Duprez, E. Sivre, A. Anthore, A. Aassime, A. Cavanna, U. Gennser, and F. Pierre, Science 366, 1243 (2019).
  31. [31]Z. Iftikhar, S. Jezouin, A. Anthore, U. Gennser, F. D. Parmentier, A. Cavanna, and F. Pierre, Nature 526, 233 (2015).
  32. [32]Z. Iftikhar, A. Anthore, A. K. Mitchell, F. D. Parmentier, U. Gennser, A. Ouerghi, A. Cavanna, C. Mora, P. Simon, and F. Pierre, Science 360, 1315 (2018).
  33. [33]M. Banerjee, M. Heiblum, A. Rosenblatt, Y. Oreg, D. E. Feldman, A. Stern, and V. Umansky, Nature 545, 75 (2017).
  34. [34]M. Banerjee, M. Heiblum, V. Umansky, D. E. Feldman, Y. Oreg, and A. Stern, Nature 559, 205 (2018).
  35. [35]D. F. Mross, Y. Oreg, A. Stern, G. Margalit, and M. Heiblum, Phys. Rev. Lett. 121, 026801 (2018).
  36. [36]E. G. Idrisov, I. P. Levkivskyi, and E. V. Sukhorukov, Phys. Rev. B 101, 254526 (2020).
  37. [37]W. Pouse, L. Peeters, C. L. Hsueh, U. Gennser, A. Cavanna, M. A. Kastner, A. K. Mitchell, and D. Goldhaber-Gordon, Nature Physics 19, 492 (2023).
  38. [38]D. B. Karki, Phys. Rev. B 113, 155411 (2026).
  39. [39]D. B. Karki, Phys. Rev. B 111, 035426 (2025).
  40. [40]A. Rosenblatt, S. Konyzheva, F. Lafont, N. Schiller, J. Park, K. Snizhko, M. Heiblum, Y. Oreg, and V. Umansky, Phys. Rev. Lett. 125, 256803 (2020).
  41. [41]A. Zhang, G. Aissani, Q. Dong, Y. Jin, K. Watanabe, T. Taniguchi, C. Altimiras, P. Roche, J.-M. Berroir, F. Baudin, G. Fève, C. Ménard, O. Maillet, and F. D. Parmentier, Phys. Rev. Lett. 137, 146603 (2026).
  42. [42]A. Zhang, K. Watanabe, T. Taniguchi, P. Roche, C. Altimiras, F. D. Parmentier, and O. Maillet, Phys. Rev. Res. 7, L042037 (2025).
  43. [43]H. Duprez, F. Pierre, E. Sivre, A. Aassime, F. D. Parmentier, A. Cavanna, A. Ouerghi, U. Gennser, I. Safi, C. Mora, and A. Anthore, Phys. Rev. Res. 3, 023122 (2021).
  44. [44]A. K. Paul, P. Tiwari, R. Melcer, V. Umansky, and M. Heiblum, Phys. Rev. Lett. 133, 076601 (2024).
  45. [45]G. Le Breton, R. Delagrange, Y. Hong, M. Garg, K. Watanabe, T. Taniguchi, R. Ribeiro-Palau, P. Roulleau, P. Roche, and F. D. Parmentier, Phys. Rev. Lett. 129, 116803 (2022).
  46. [46]R. Kumar, S. K. Srivastav, U. Roy, U. Singhal, K. Watanabe, T. Taniguchi, V. Singh, P. Roulleau, and A. Das, Nature Physics 20, 1941 (2024).
  47. [47]S. Samai, D. Sarkar, A. Halder, T. Taniguchi, K. Watanabe, S. Mukerjee, S. K. Srivastav, and A. Das, (2026), arXiv:2607.19205 [cond-mat.mes-hall].arxiv.org/abs/2607.19205
  48. [48]X. G. Wen, Phys. Rev. B 41, 12838 (1990).
  49. [49]T. Giamarchi, Quantum Physics in One Dimension (Claverdon Press Oxford, 2004).
  50. [50]E. G. Idrisov, I. P. Levkivskyi, and E. V. Sukhorukov, Phys. Rev. B 106, L121405 (2022).
  51. [51]E. G. Idrisov, I. P. Levkivskyi, and E. V. Sukhorukov, Phys. Rev. Res. 7, 013042 (2025).
  52. [52]E. M. Lifshitz and L. P. Pitaevskii, Statistical Physics, Part 2: (Landau and Lifshitz Course of Theoretical Physics, Vol.9) (Butterworth-Heinemann Oxford, 1980).
  53. [53]See Supplemental Material at ... for calculation of contacts temperature, electron distribution function and noise of electric current.

Paper details

vlnc:W_69njxyy7
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY-SA 4.0
Usage

views · 3

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents