Mesoscopic Transport

Mesoscopic transport studies charge, spin, heat, and quasiparticle flow through devices small enough that discreteness, phase coherence, Coulomb blockade, or quantum statistics matter. From the open-system point of view, the central idea is simple: a small quantum system is coupled to large reservoirs, and the observable current is generated by exchange events across the system-reservoir boundaries.

This page is an application guide. It does not replace a full condensed-matter treatment of scattering theory, Keldysh Green functions, or device fabrication. It focuses on the open-system language: reservoirs and leads, tunneling rates, master equations, jump processes, shot noise, and counting statistics. Quantum Coherence in Conductors owns orbital phase memory, diffusive dephasing, and interference-based extraction. Quantum Dots owns confinement and microscopic addition spectra; Coulomb Blockade owns the orthodox capacitance network, event free energies, and diamond interpretation. The generic bulk dictionary of conductivity, diffusion, viscosity, and Green–Kubo limits belongs to Transport Coefficients Preview.

Leads as Reservoirs

A common microscopic model splits the Hamiltonian into

$$H = H_S + \sum_\alpha H_\alpha + H_T,$$

where $H_S$ describes the central device, $H_\alpha$ are leads, and $H_T$ tunnels particles between them. For noninteracting electronic leads,

$$H_\alpha = \sum_k \epsilon_{k\alpha} c_{k\alpha}^\dagger c_{k\alpha}.$$

Each lead is usually assumed to be in a grand-canonical state,

$$\rho_\alpha \propto \exp \left[ - \beta_\alpha \left( H_\alpha-\mu_\alpha N_\alpha \right) \right],$$

with temperature $T_\alpha$ and chemical potential $\mu_\alpha$. The Fermi function is

$$f_\alpha(E) = \frac{1} {e^{\beta_\alpha(E-\mu_\alpha)}+1}.$$

The Markovian lead approximation assumes that the leads remain large, rapidly relaxing reservoirs even as particles enter and leave the small device.

Single-Level Quantum Dot

The simplest sequential-tunneling model is a spinless single-level quantum dot with energy $\epsilon$. Strong Coulomb blockade allows only two states:

$$\lvert 0\rangle, \qquad \lvert 1\rangle=d^\dagger\lvert 0\rangle.$$

Lead $\alpha$ fills the dot with rate

$$\Gamma_\alpha^+ = \Gamma_\alpha f_\alpha(\epsilon),$$

and empties it with rate

$$\Gamma_\alpha^- = \Gamma_\alpha \left[ 1-f_\alpha(\epsilon) \right].$$

The probabilities obey the rate equation

$$\dot p_1 = \sum_\alpha \Gamma_\alpha^+ p_0 - \sum_\alpha \Gamma_\alpha^- p_1, \qquad p_0=1-p_1.$$

This is a classical master equation for populations, but it has a quantum origin: the rates come from tunneling between a discrete quantum state and fermionic reservoirs. For the general population-equation structure, see Pauli Rate Equations.

Detailed Balance with Chemical Potential

For one equilibrium lead,

$$\frac{\Gamma_\alpha^+} {\Gamma_\alpha^-} = \frac{f_\alpha(\epsilon)} {1-f_\alpha(\epsilon)} = e^{-\beta_\alpha(\epsilon-\mu_\alpha)}.$$

This is local detailed balance. Adding one electron to the dot changes the reservoir particle number and energy, so the thermodynamic exponent contains $\epsilon-\mu_\alpha$, not only $\epsilon$.

When all leads have the same temperature and chemical potential, the dot relaxes to the grand-canonical occupation

$$p_1^{\mathrm{eq}} = f(\epsilon).$$

Transport appears when different reservoirs impose different chemical potentials or temperatures. Chemical Potential distinguishes the thermodynamic, electrochemical, and voltage conventions behind those reservoir parameters.

Current

With the sign convention that $I_\alpha$ is charge current from lead $\alpha$ into the dot,

$$I_\alpha = e \left( \Gamma_\alpha^+ p_0 - \Gamma_\alpha^- p_1 \right).$$

At steady state, define

$$\Gamma^+ = \sum_\alpha\Gamma_\alpha^+, \qquad \Gamma^- = \sum_\alpha\Gamma_\alpha^-.$$

Then

$$p_1^{\mathrm{ss}} = \frac{\Gamma^+} {\Gamma^+ + \Gamma^-}, \qquad p_0^{\mathrm{ss}} = \frac{\Gamma^-} {\Gamma^+ + \Gamma^-}.$$

For two leads, the steady current from the left lead is

$$I_L^{\mathrm{ss}} = e\, \frac{ \Gamma_L^+\Gamma_R^- - \Gamma_L^-\Gamma_R^+ }{ \Gamma^+ + \Gamma^- }.$$

The numerator is a competition between left-to-right and right-to-left cycles. At equilibrium the two terms cancel by detailed balance.

High-Bias Limit

In the ideal high-bias sequential-tunneling limit,

$$f_L(\epsilon)\simeq1, \qquad f_R(\epsilon)\simeq0.$$

Then the dot is filled from the left and emptied to the right:

$$\Gamma_L^+=\Gamma_L, \qquad \Gamma_R^-=\Gamma_R,$$

with the reverse rates negligible. The steady current is

$$I = e \frac{\Gamma_L\Gamma_R} {\Gamma_L+\Gamma_R}.$$

The smaller rate is the bottleneck. If $\Gamma_L\gg\Gamma_R$, the dot is almost always occupied and the right barrier controls the current. If $\Gamma_R\gg\Gamma_L$, the dot is almost always empty and the left barrier controls the current.

Counting Statistics

Counting statistics keeps track of the number of transferred particles, not just the mean current. For the single-level dot, let $\chi$ count electrons entering the right lead. In the basis $(p_0,p_1)^T$, a counting-field rate matrix is

$$W(\chi) = \begin{pmatrix} -\Gamma^+ & \Gamma_L^-+\Gamma_R^-e^{i\chi} \\ \Gamma_L^+ + \Gamma_R^+e^{-i\chi} & -\Gamma^- \end{pmatrix}.$$

The moment-generating function has the long-time form

$$Z(\chi,t) \sim e^{\lambda_0(\chi)t},$$

where $\lambda_0(\chi)$ is the eigenvalue of $W(\chi)$ that approaches zero at $\chi=0$. Current cumulants are obtained by differentiating $\lambda_0(\chi)$ with respect to $i\chi$.

This is the transport version of a quantum-jump unraveling: jumps across a selected barrier are counted, and the statistics of those jumps carries information beyond the average current.

Shot Noise and Fano Factor

Current noise reveals correlations between transfer events. A common low-frequency definition is

$$S = 2eI F,$$

where $F$ is the Fano factor. Independent Poisson transfer events have $F=1$. Sequential tunneling through a single dot in the high-bias limit gives

$$F = \frac{ \Gamma_L^2+\Gamma_R^2 }{ (\Gamma_L+\Gamma_R)^2 }.$$

For symmetric barriers, $F=1/2$: Coulomb blockade anticorrelates tunneling events because the dot must empty before it can be filled again. Super-Poissonian noise can occur when transport switches between slow and fast configurations, when avalanches occur, or when internal dynamics bunches transfer events.

In coherent scattering language, a quantum point contact with transmission eigenvalues $T_n$ has conductance proportional to $\sum_n T_n$ and zero-temperature shot noise proportional to

$$\sum_n T_n(1-T_n).$$

Fully open and fully closed channels do not produce partition noise. Partially transmitting channels do.

Relation to Master Equations

Mesoscopic transport uses several levels of description:

Regime	Common method	Open-system interpretation
coherent noninteracting conductor	scattering matrix or Landauer-Büttiker theory	reservoirs set incoming occupations
weak tunneling and Coulomb blockade	rate equation or quantum master equation	jumps between charge states
coherent dot dynamics with weak leads	Redfield or Lindblad-type equation	lead-induced transitions and dephasing
strong coupling or structured contacts	nonequilibrium Green functions or reaction-coordinate methods	reservoir memory and level broadening matter
monitored transport	counting statistics or quantum trajectories	individual transfers are records

No one description dominates every regime. A rate equation can be excellent in sequential tunneling and wrong in coherent cotunneling. A Lindblad equation can preserve positivity while using rates that are not appropriate for a biased conductor. A scattering approach can compute coherent conductance while hiding internal relaxation and measurement backaction.

Measurement and Backaction

Mesoscopic devices can also measure other quantum systems. A quantum point contact or single-electron transistor coupled to a charge qubit turns qubit state information into a current signal. The same coupling causes dephasing because different qubit states imprint different phases, transmission probabilities, or tunneling rates on the detector.

This is the electronic analogue of optical or circuit-QED readout:

information leaves through a monitored current channel;
unobserved information causes dephasing;
the measurement record can be modeled stochastically.

The correct trajectory description depends on detector bandwidth and whether individual tunneling events are resolved. When events are resolved, jump trajectories are natural. When only a noisy current is measured, a diffusive approximation may be better.

Thermodynamics of Transport

Each transferred particle carries energy and charge. For a transition involving lead $\alpha$, the entropy contribution contains

$$\beta_\alpha \left( \Delta E-\mu_\alpha\Delta N \right).$$

This is why transport fluctuation theorems involve affinities built from temperature and chemical-potential differences. For isothermal charge transport, the bias contribution is controlled by

$$\beta eV,$$

where $eV=\mu_L-\mu_R$ for a two-terminal conductor. See Fluctuation Theorems and Entropy Production.

Common Mistakes

• Treating leads as ordinary finite baths while still assuming fixed chemical potentials.
• Forgetting the chemical-potential term in detailed balance.
• Using a population rate equation when coherences or level hybridization are essential.
• Assuming a Lindblad form automatically gives the right transport thermodynamics.
• Mixing sign conventions for current into the dot and current into a lead.
• Calling all current noise shot noise; equilibrium Johnson-Nyquist noise has a different origin.
• Ignoring spin degeneracy or Coulomb blockade restrictions in rate counting.
• Using a Markovian rate model when the contact has structured density of states or strong hybridization.

Exercises

Local Detailed Balance

Show that

$$\frac{\Gamma_\alpha^+} {\Gamma_\alpha^-} = e^{-\beta_\alpha(\epsilon-\mu_\alpha)}$$

for the single-level dot rates $\Gamma_\alpha^+=\Gamma_\alpha f_\alpha(\epsilon)$ and $\Gamma_\alpha^-=\Gamma_\alpha[1-f_\alpha(\epsilon)]$.

Solution

Using

$$f_\alpha(\epsilon) = \frac{1} {e^{\beta_\alpha(\epsilon-\mu_\alpha)}+1},$$

we have

$$1-f_\alpha(\epsilon) = \frac{ e^{\beta_\alpha(\epsilon-\mu_\alpha)} }{ e^{\beta_\alpha(\epsilon-\mu_\alpha)}+1 }.$$

Therefore

$$\frac{f_\alpha(\epsilon)} {1-f_\alpha(\epsilon)} = e^{-\beta_\alpha(\epsilon-\mu_\alpha)}.$$

The factors $\Gamma_\alpha$ cancel.

Steady Current

Derive the steady current for the two-lead single-level dot.

Solution

The steady occupation is

$$p_1^{\mathrm{ss}} = \frac{\Gamma^+} {\Gamma^+ + \Gamma^-}, \qquad p_0^{\mathrm{ss}} = \frac{\Gamma^-} {\Gamma^+ + \Gamma^-}.$$

The left current into the dot is

$$I_L = e \left( \Gamma_L^+p_0 - \Gamma_L^-p_1 \right).$$

Substituting the steady probabilities gives

$$I_L^{\mathrm{ss}} = e \frac{ \Gamma_L^+\Gamma^- - \Gamma_L^-\Gamma^+ }{ \Gamma^+ + \Gamma^- }.$$

Using $\Gamma^\pm=\Gamma_L^\pm+\Gamma_R^\pm$, the same-lead products cancel:

$$I_L^{\mathrm{ss}} = e \frac{ \Gamma_L^+\Gamma_R^- - \Gamma_L^-\Gamma_R^+ }{ \Gamma^+ + \Gamma^- }.$$

High-Bias Occupation

In the high-bias limit $\Gamma_L^+=\Gamma_L$ and $\Gamma_R^-=\Gamma_R$, with reverse rates zero. Find $p_1^{\mathrm{ss}}$ and $I$.

Solution

Here

$$\Gamma^+=\Gamma_L, \qquad \Gamma^-=\Gamma_R.$$

Thus

$$p_1^{\mathrm{ss}} = \frac{\Gamma_L} {\Gamma_L+\Gamma_R}.$$

The right-going current is the rate at which the occupied dot empties into the right lead:

$$I = e\Gamma_R p_1^{\mathrm{ss}} = e \frac{\Gamma_L\Gamma_R} {\Gamma_L+\Gamma_R}.$$

Counting-Field Signs

In the counting matrix $W(\chi)$ above, why does the right-emptying rate carry $e^{i\chi}$ while the right-filling rate carries $e^{-i\chi}$?

Solution

The convention was to count electrons entering the right lead. When the dot empties to the right, one electron is added to the counted lead, so the jump is weighted by $e^{i\chi}$. When an electron enters the dot from the right lead, the counted lead loses one electron, so the jump is weighted by $e^{-i\chi}$. Reversing the current convention would reverse these signs.

Cross-Links

• Quantum Coherence in Conductors
• Universal Conductance Fluctuations
• Quantum Dots
• Coulomb Blockade
• Single-Electron Devices
• Quantum Point Contacts
• Conductance Quantization
• Edge and Surface States
• Transport Coefficients Preview
• Grand-Canonical Ensemble
• Fermi–Dirac Statistics
• System–Bath Hamiltonians
• Detailed Balance
• Thermal Master Equations
• Fluctuation Theorems
• Entropy Production
• Noise Spectra
• Quantum Jump Trajectories
• Redfield Equation
• Reaction-Coordinate Mapping
• Approximation Checklist

References

• S. Datta, Electronic Transport in Mesoscopic Systems, Cambridge University Press (1995).
• Y. V. Nazarov and Y. M. Blanter, Quantum Transport: Introduction to Nanoscience, Cambridge University Press (2009).
• Ya. M. Blanter and M. Buttiker, “Shot noise in mesoscopic conductors,” Physics Reports 336, 1-166 (2000).
• L. S. Levitov and G. B. Lesovik, “Charge distribution in quantum shot noise,” JETP Letters 58, 230-235 (1993).
• S. A. Gurvitz and Ya. S. Prager, “Microscopic derivation of rate equations for quantum transport,” Physical Review B 53, 15932-15943 (1996).
• D. A. Bagrets and Yu. V. Nazarov, “Full counting statistics of charge transfer in Coulomb blockade systems,” Physical Review B 67, 085316 (2003).
• M. Esposito, U. Harbola, and S. Mukamel, “Nonequilibrium fluctuations, fluctuation theorems, and counting statistics in quantum systems,” Reviews of Modern Physics 81, 1665-1702 (2009).