1: % Use only LaTeX2e, calling the article.cls class and 12-point type.

2: 

3: %\documentclass[12pt]{article}

4: \documentclass[twocolumn,english,aps,prl,superscriptaddress]{revtex4}

5: % Users of the {thebibliography} environment or BibTeX should use the

6: % scicite.sty package, downloadable from *Science* at

7: % www.sciencmag.org/misc/con-info.shtml .  This package should properly

8: % format in-text reference calls and reference-list numbers.

9: 

10: %\usepackage{scicite}

11: 

12: % Use times if you have the font installed; otherwise, comment out the

13: % following line.

14: 

15: \usepackage{times}

16: 

17: \usepackage{graphicx}

18: \usepackage{babel}

19: \usepackage{amsmath}

20: \usepackage{amsfonts}

21: \usepackage{amssymb}

22: \usepackage{epsfig}

23: 

24: 

25:  \oddsidemargin 0.2cm \textwidth 16cm \textheight

26: 25cm \footskip 1.0cm

27: 

28: \begin{document}

29: \normalsize{\textbf{Science, Vol.313, p.499, 2006}}\\\\

30: \title{Violation of Kirchhoff's

31: Laws for a Coherent RC Circuit}

32: 

33: \author{J. Gabelli}\affiliation{Laboratoire Pierre Aigrain, D{\'e}partement

34: de Physique de l'Ecole Normale Sup\'erieure, 24 rue Lhomond, 75231

35: Paris Cedex 05, France} \author{G. F\`{e}ve}\affiliation{Laboratoire

36: Pierre Aigrain, D{\'e}partement de Physique de l'Ecole Normale

37: Sup\'erieure, 24 rue Lhomond, 75231 Paris Cedex 05,

38: France}\author{J.-M. Berroir}\affiliation{Laboratoire Pierre

39: Aigrain, D{\'e}partement de Physique de l'Ecole Normale

40: Sup\'erieure, 24 rue Lhomond, 75231 Paris Cedex 05,

41: France}\author{B. Pla\c{c}ais}\affiliation{Laboratoire Pierre Aigrain,

42: D{\'e}partement de Physique de l'Ecole Normale Sup\'erieure, 24 rue

43: Lhomond, 75231 Paris Cedex 05, France}\author{A.

44: Cavanna}\affiliation{Laboratoire de Photonique et Nanostructures,

45: UPR20 CNRS, Route de Nozay, 91460 Marcoussis Cedex,

46: France}\author{B. Etienne}\affiliation{Laboratoire de Photonique et

47: Nanostructures, UPR20 CNRS, Route de Nozay, 91460 Marcoussis Cedex,

48: France}\author{Y.Jin}\affiliation{Laboratoire de Photonique et

49: Nanostructures, UPR20 CNRS, Route de Nozay, 91460 Marcoussis Cedex,

50: France}\author{D.C. Glattli} \email{glattli@lpa.ens.fr}

51: \affiliation{Laboratoire Pierre Aigrain, D{\'e}partement de Physique

52: de l'Ecole Normale Sup\'erieure, 24 rue Lhomond, 75231 Paris Cedex

53: 05, France} \affiliation{Service de Physique de l'Etat Condens{\'e},

54: CEA Saclay, F-91191 Gif-sur-Yvette, France}

55: 

56: 

57: 

58: 

59: \begin{abstract}

60: What is the complex impedance of a fully coherent quantum

61: resistance-capacitance (RC)circuit at GHz frequencies in which a

62: resistor and a capacitor are connected in series? While Kirchhoff's

63: laws predict addition of capacitor and resistor impedances, we

64: report on observation of a different behavior. The resistance,here

65: associated with charge relaxation, differs from the usual transport

66: resistance given by the Landauer formula. In particular, for a

67: single mode conductor, the charge relaxation resistance is half the

68: resistance quantum, regardless of the transmission of the mode. The

69: new mesoscopic effect reported here is relevant for the dynamical

70: regime of all quantum devices.

71: \end{abstract}

72: 

73: \maketitle

74: For a classical circuit, Kirchhoff's laws prescribe the

75: addition

76:  of resistances in series. Its failure has been a central issue in developing our understanding

77:  of electronic transport in mesoscopic conductors. Indeed, coherent multiple electronic

78: reflections between scatterers in the conductor were

79:  found to make the conductance non-local \cite{Landauer57,Landauer70}. A new composition law of

80:  individual scatterer contribution to resistance was found that led to

81:  the solution of the problem of electron localization \cite{Anderson81} and, later,

82:  to formulation of the electronic conduction in terms of

83:  scattering of electronic waves \cite{ButtImLandauer83}. Nonadditivity of series

84:  resistances, or of parallel conductances, nonlocal effects and

85:  negative four-point resistances \cite{GaoBO} have been

86:  observed in a series of transport experiments at low temperature, where phase

87:  coherence extends over the mesoscopic scale \cite{Review1,Review2}.

88:  It is generally accepted that the conductance of a phase-coherent quantum

89:  conductor is given by the Landauer formula and its generalization to multi-lead conductors \cite{ButtLand},

90: which relate the conductance to the transmission of electronic waves

91: by the conductance quantum $e^2/h$. But, how far is this description

92: robust at finite frequency where conductance combines with

93: nondissipative circuit elements such as capacitors or inductors ?

94: Are there significant departures from the dc result ? The question

95: is important, as recent advances in quantum information highlight

96: the need for fast manipulation of quantum systems, in particular

97: quantum conductors. High frequency quantum transport has been

98: theoretically addressed, showing that a quantum RC circuit displays

99: discrepancies with its classical counterpart

100: \cite{Buttiker93,Pretre96}.

101: \begin{figure}

102: \centerline{\includegraphics[width=6 cm,

103: keepaspectratio]{fig1.eps}}\caption{{\small The quantum capacitor

104: realized using a 2DEG (A) and its equivalent circuit (B). The

105: capacitor consists of a metallic electrode (in gold) on top of a

106: submicrometer 2DEG quantum dot (in blue) defining the second

107: electrode. The resistor is a QPC linking the dot to a wide 2DEG

108: reservoir (in blue), itself connected to a metallic contact (dark

109: gold). The QPC voltage $V_G$ controls the number of electronic modes

110: and their transmission. The radio frequency voltage $V_{ac}$, and

111: eventually a dc voltage $V_{dc}$, are applied to the

112: counter-electrode whereas the ac current, from which the complex

113: conductance is deduced, is collected at the ohmic contact. As

114: predicted by theory, the relaxation resistance $R_q$, which enters

115: the equivalent circuit for the coherent conductance, is

116: transmission-independent and equal to half the resistance quantum.

117: The capacitance is the serial combination $C_\mu$ of the quantum and

118: the geometrical capacitances ($C_q$ and $C$ respectively). $C_q$ is

119: transmission-dependent and strongly modulated by $V_{dc}$ and/or

120: $V_G$. The combination of $R_q$ and $C_q$ forms the impedance

121: $1/g_q$ of the coherent quantum conductor.}}\label{HBT1}\end{figure}

122: It was shown that a counter-intuitive modification of the series

123: resistance lead to the situation in which the resistance is no

124: longer described by the Landauer formula and does not depend on

125: transmission in a direct way \cite{Buttiker93,Pretre96}. Instead it

126: is directly related to the dwell time of electrons in the capacitor.

127: Moreover, when the resistor transmits in a single electronic mode, a

128: constant resistance was found, equal to the half-resistance quantum

129: $h/2e^2$, i.e., it was not transmission-dependent. This resistance,

130: modified by the presence of the coherent capacitor, was termed a

131: "charge-relaxation resistance" to distinguish it from the usual dc

132: resistance, which is sandwiched between macroscopic reservoirs and

133: described by the Landauer formula. The quantum charge-relaxation

134: resistance, as well as its generalization in nonequilibrium systems,

135: is an important concept that can be applied to quantum information.

136: For example, it enters into the problem of quantum-limited detection

137: of charge qubits \cite{Pilgram} \cite{Clerk}, in the study of

138: high-frequency-charge quantum noise

139: \cite{ChargeNoise1,ChargeNoise2,ChargeNoise3}, or in the study of

140: dephasing of an electronic quantum interferometer \cite{Seelig}. In

141: molecular electronics, the charge relaxation resistance is also

142: relevant to the THz frequency response of systems such as carbon

143: nanotubes\cite{Burke}.

144: 

145: We report on the observation and quantitative measurement of the

146: quantum charge-relaxation resistance in a coherent RC circuit

147: realized in a two-dimensional electron gas (2DEG) (see Fig.1A). The

148: capacitor is made of a macroscopic metallic electrode on top of a

149: 2DEG submicrometer dot defining the second electrode. The resistor

150: is a quantum point contact (QPC) connecting the dot to a wide 2DEG

151: macroscopic reservoir.  We address the coherent regime in which

152: electrons emitted from the reservoir to the dot are backscattered

153: without loss of coherence. In this regime, we have checked the

154: prediction made in refs.\cite{Buttiker93,Pretre96} that the

155: charge-relaxation resistance is not given by the Landauer formula

156: resistance but instead is constant and equals $h/2e^2$, as the QPC

157: transmission is varied. Note that we consider here a spin-polarized

158: regime and that the factor $1/2$ is not the effect of spin, but a

159: hallmark of a charge-relaxation resistance. When coherence is washed

160: out by thermal broadening, the more conventional regime pertaining

161: to dc transport is recovered. The present work differs from previous

162: capacitance measurements where, for spectroscopic purpose, the dot

163: reservoir coupling was weak and the ac transport regime was

164: incoherent \cite{Ashoori92a,Ashoori92b}. As a consequence, although

165: quantum effects in the capacitance were observable, the quantum

166: charge-relaxation resistance was not accessible in these earlier

167: experiments.

168: 

169: At zero temperature in the coherent regime and when a single mode is

170: transmitted by the QPC, the mesoscopic RC circuit is represented by

171: the equivalent circuit of Fig.1B \cite{Buttiker93,Pretre96}. The

172: geometrical capacitance $C$ is in series with the quantum admittance

173: $g_{q}(\omega)$ connecting the ac current flowing in the QPC to the

174: ac internal potential of the dot:

175: \begin{equation}\label{cond0}

176: g_{q}(\omega)=\frac{1}{\frac{h}{2e^{2}}+\frac{1}{-i\omega C_{q}}}

177: \qquad (T = 0)

178: \end{equation}

179: The nonlocal quantum impedance behaves as if it were the series

180: addition of a quantum capacitance $C_{q}$ with a constant contact

181: resistance $h/2e^2$. $C_{q}=e^{2}\frac{d\mathcal{N}}{d\varepsilon}$

182: is associated with the local density of state

183: $\frac{d\mathcal{N}}{d\varepsilon}$ of the mode propagating in the

184: dot, taken at the Fermi energy. The striking effect of phase

185: coherence is that the QPC transmission probability $D$ affects the

186: quantum capacitance (see Eq.\ref{Cq}) but not the resistance. The

187: total circuit admittance $G$ is simply :

188: \begin{equation}\label{grandcon}

189:  G=\frac{-i\omega C g_{q}(\omega)}{-i\omega C+g_{q}(\omega)}= \frac{-i\omega C_{\mu}\frac{2e^{2}}{h}}{-i\omega

190:  C_{\mu}+\frac{2e^{2}}{h}}\quad , \qquad (T = 0)

191: \end{equation}

192: where $C_{\mu}=\frac{CC_{q}}{C+C_{q}}$ is the electrochemical

193: capacitance. In the incoherent regime, both resistance and quantum

194: capacitance vary with transmission. The dot forms a second reservoir

195: and the electrochemical capacitance $C_{\mu}$ is in series with the

196: QPC resistance $R$. In particular, when the temperature is high

197: enough to smooth the capacitor density of states, the Landauer

198: formula $R=\frac{h}{e^{2}}\times\frac{1}{D}$ is recovered.

199: 

200: Several samples have been measured at low temperatures, down to 30

201: mK, which show analogous features. We present results on two samples

202: made with 2DEG defined in the same high-mobility GaAsAl/GaAs

203: heterojunction, with nominal density $n_{s}=1.7 \times 10^{15}$

204: m$^{-2}$ and mobility $\mu =260$ V$^{-1}$m$^{2}$s$^{-1}$. A finite

205: magnetic field ($B=1.3$ T) is applied, so as to work in the

206: ballistic quantum Hall regime with no spin degeneracy \cite{SOM}.

207: 

208: \begin{figure}[hhh]

209: \centerline{\includegraphics[width=9 cm,

210: keepaspectratio]{fig2.eps}}\caption{{\small Complex conductance of

211: sample E3 as function of the gate voltage $V_G$ for $T=100$ mK and

212: $\omega /2\pi=1.2$ GHz, at the opening of the first conduction

213: channel (C) and its Nyquist representation in (D). The theoretical

214: circle characteristic of the coherent regime is shown as a solid

215: line. (A and B) show the corresponding curves for the simulation of

216: sample E3 using the 1D model with $C=4$ $ fF$, $C_\mu=1$ $

217: fF$.}}\label{HBT2}\end{figure}

218: 

219: The real and imaginary parts of the admittance $Im(G)$ and $Re(G)$

220: as a function of QPC gate voltage $V_{G}$ at the opening of the

221: first conduction channel are shown in Fig.2C. On increasing $V_{G}$,

222: we can distinguish three regimes. At very negative $V_{G}\leq -0.86$

223: V, the admittance is zero. Starting from this pinched state, peaks

224: are observed in both $Im(G)$ and $Re(G)$. Following a maximum in the

225: oscillations, a third regime occurs where $Im(G)$ oscillates nearly

226: symmetrically about a plateau while the oscillation amplitude

227: decreases smoothly. Simultaneously, peaks in $Re(G)$ quickly

228: disappear to vanish in the noise.

229: 

230: Comparing these observations with the results of

231: refs.\cite{Buttiker93,Pretre96},using a simplified one-dimensional

232: (1D) model for $C_q$ with one conduction mode and a constant energy

233: level spacing in the dot $\Delta$ \cite{Gabelli06}, the simulation

234: (Fig.2A) shows a striking similarity to the experimental conductance

235: traces in Fig.2C. In this simulation, $V_{G}$ determines the

236: transmission $D$ but also controls linearly the 1D dot potential.

237: The transmission is chosen to vary with $V_G$ according to a

238: Fermi-Dirac-like dependence appropriate to describe QPC transmission

239: \cite{Butt90}. This model can be used to get a better understanding

240: of the different conductance regimes. Denoting $r$ and $t$ the

241: amplitude reflection and transmission coefficients of the QPC

242: ($r^{2}=1-D$, $t=\sqrt{D}$), we first calculate the scattering

243: amplitude of the RC circuit:

244: \begin{equation}\label{S}

245: s(\varepsilon)=r-t^{2}e^{i\varphi}\sum_{n=0}^{\infty}(re^{i\varphi})^{n}=\frac{r-e^{i\varphi}}{1-re^{i\varphi}}

246: \end{equation}

247: where $\varepsilon$ is the Fermi energy relative to the dot

248: potential and $\varphi=2\pi\varepsilon/\Delta$ is the phase

249: accumulated for a single turn in the quantum dot. The

250: zero-temperature quantum capacitance is then given by:

251: \begin{equation}\label{Cq}

252: C_q=e^{2}\frac{d\mathcal{N}}{d\varepsilon}=\frac{1}{2i\pi}s^{+}\frac{\partial

253: s}{\partial

254: \varepsilon}=\frac{e^{2}}{\Delta}\frac{1-r^{2}}{1-2r\cos2\pi\frac{\varepsilon}{\Delta}+r^{2}}

255: \end{equation}

256: Therefore, $C_q$ exhibits oscillations when the dot potential is

257: varied. When $r \rightarrow 0$,  these oscillations vanish and

258: $C_q\rightarrow e^{2}/\Delta$. As reflection increases, oscillations

259: are growing with maxima $\frac{e^{2}}{\Delta} \frac{1+r}{1-r}$ and

260: minima $\frac{e^{2}}{\Delta} \frac{1-r}{1+r}$. For strong

261: reflection, Eq \ref{Cq} gives resonant Lorentzian peaks with an

262: energy width $D\Delta/2$ given by the escape rate of the dot.

263: However, at finite temperature, the conductance in equation

264: (\ref{cond0}) has to be thermally averaged to take into account the

265: finite energy width of the electron source so that :

266: \begin{eqnarray}\label{condT}

267:      g_{q}(\omega) = \int d\varepsilon \left ( - \frac{\partial f}{\partial \varepsilon}\right ) \frac{1}{h/2e^2+1/(-i\omega

268:      C_{q})}\quad \\

269:      (T\neq 0)  \nonumber

270: \end{eqnarray}

271: where $f$ is the Fermi-Dirac distribution. Again, the nonlocal

272: admittance behaves as if it were the serial association of a

273: charge-relaxation resistance $R_q$ and a capacitance that we still

274: denote $C_q$. In the weak transmission regime ($D\rightarrow 0$),

275: when $D\Delta \ll k_{B}T $, equation (\ref{condT}) yields thermally

276: broadened capacitance peaks with

277: \begin{equation}\label{fanC}

278: C_{q} \simeq

279: \frac{e^{2}}{4k_{B}T\cosh^{2}(\delta\varepsilon/2k_{B}T)}\quad,\qquad

280: (D<<1)

281: \end{equation}

282: where $\delta \varepsilon$ denotes the energy distance to a resonant

283: dot level. Note that these capacitance peaks do not depend on the

284: dot parameters and can be used as a primary thermometer. Similar but

285: transmission-dependent peaks are predicted in the inverse resistance

286: \begin{equation}\label{fanR}

287: 1/R_{q} \simeq

288: D\frac{e^{2}}{h}\frac{\Delta}{4k_{B}T\cosh^{2}(\delta\varepsilon/2k_{B}T)},\qquad

289: (D<<1)

290: \end{equation}

291: This result is reminiscent of the thermally broadened resonant

292: tunneling conductance for transport through a quantum dot. A

293: consequence of the finite temperature is the fact that the

294: resistance is no longer constant. This thermally-induced divergence

295: of $R_q$ at low transmission restores a frequency-dependent

296: pinch-off for $R_q\gg 1/C_q\omega$, as can be seen in both model and

297: experiment in Figs.2A.C. As mentioned above, for $k_{B}T \gg

298: D\Delta$, the quantum dot looks like a reservoir and the Landauer

299: formula is recovered.

300: 

301: The coherent and the thermally broadened regimes are best

302: demonstrated in the Nyquist representation $Im(G)$ versus $Re(G)$ of

303: the experimental data in Fig.2D. This representation allows to

304: easily distinguish constant resistance from constant capacitance

305: regimes, as they correspond to circles respectively centered on the

306: real and imaginary axis. Whereas, for low transmission, the Nyquist

307: diagram strongly depends on transmission, the conductance

308: oscillations observed in Fig.2C collapse on a single curve in the

309: coherent regime. Moreover this curve is the constant $R_q=h/2e^2$

310: circle. By contrast, admittance peaks at low transmission correspond

311: to a series of lobes in the Nyquist diagram, with slopes increasing

312: with transmission in qualitative agreement with Eqs.\ref{fanC} and

313: \ref{fanR}. These lobes and the constant $R_q$ regime are well

314: reproduced by the simulations in Fig.2B. Here, the value of $C_\mu$

315: and the electronic temperature are deduced from measurement. In our

316: experimental conditions, the simulated traces are virtually free of

317: adjustable parameters as $C\geq 4C_\mu \gg C_{q}$.

318: 

319: 

320: It is important to note that in a real system, the weak transmission

321: regime is accompanied by Coulomb blockade effects that are not taken

322: into account in the above model. In the weak transmission regime and

323: $T=0$, using an elastic co-tunneling approach

324: \cite{Averin92,Glattli93}, we have checked that there is no

325: qualitative change except for the energy scale that now includes the

326: charging energy so that $\Delta$ is replaced by

327: $\Delta$+$e^2/C$=$e^2/C_{\mu}$. At large transmission, the problem

328: is nonperturbative in tunnel coupling and highly nontrivial.

329: Calculations of the thermodynamic capacitance exist

330: \cite{Matveev,Glazman98,Brouwer05}, but at present, no comprehensive

331: model is available that would include both charge-relaxation

332: resistance and quantum capacitance for finite temperature and/or

333: large transmission.

334: \begin{figure}[hhh]

335: \centerline{\includegraphics[width=3.5 in,

336: keepaspectratio]{fig3.eps}}\caption{{\small Coulomb blockade

337: oscillations in the real part of the ac conductance in the

338: low-transmission regime. The control voltage is applied to the

339: counter-electrode for Sample E3 (A) and to the QPC gate for Sample

340: E1 (B). The temperature dependence is used for absolute calibration

341: of our setup, as described in the text : the peak width, shown in (C

342: and D) as a function of temperature, is deduced from theoretical

343: fits (solid lines) using Eq.(\ref{fanR}) and taking a linear

344: dependence of energy with the control voltage. Lines in (C) and (D)

345: are fits of the experimental results using a $\sqrt{T^2+T_0^2}$ law

346: to take into account a finite residual electronic temperature

347: $T_{0}$.}}\label{HBT3}\end{figure} Calibration of our admittance

348: measurements is a crucial step toward extracting the absolute value

349: of the constant charge-relaxation resistance. As at GHz frequencies,

350: direct calibration of the whole detection chain is hardly better

351: than 3dB, we shall use here an indirect, but absolute, method, often

352: used in Coulomb blockade spectroscopy, that relies on the comparison

353: between the gate voltage width of a thermally broadened Coulomb peak

354: ($\propto k_{B}T$) and the Coulomb peak spacing ($\propto

355: e^2/C_\mu$). From this, an absolute value of $C_{\mu}$ can be

356: obtained. The real part of the admittance of Sample E3 is shown as a

357: function of the dc voltage $V_{dc}$ at the counter-electrode, for a

358: given low transmission (Fig.3A). A series of peaks with periodicity

359: $\Delta V_{dc}=370\, \mu$V are observed, with the peaks accurately

360: fitted by Eq.\ref{fanR}. Their width, proportional to the electronic

361: temperature $T_{el}$, is plotted versus the refrigerator temperature

362: $T$ (see Fig.3C). When corrected for apparent electron heating

363: arising from gaussian environmental charge noise, and if we assume

364: $T_{el}=\sqrt{T^{2}+T_{0}^{2}}$, the energy calibration of the gate

365: voltage yields $C_\mu$ and the amplitude $1/C_{\mu}\omega$ of the

366: conductance plateau in Fig.2. A similar analysis is done in Fig.3, B

367: and D, for sample E1 using $V_G$ to control the dot potential. Here

368: peaks are distorted because of a transmission-dependent background

369: and show a larger periodicity $\Delta V_{G}=2$ mV, which reflects

370: the weaker electrostatic coupling to the 2DEG.

371: 

372: \begin{figure}[hhh]

373: \centerline{\includegraphics[width=9 cm,

374: keepaspectratio]{fig4.eps}}\caption{{\small Complex impedance of

375: Sample E3 (A and B) and Sample E1 (C and D) as a function of QPC

376: voltage for $T= 30$ mK and $B=1.3$ T. The dashed lines in (B and D)

377: correspond to the values of $C_{\mu}$ deduced from calibration. The

378: horizontal solid lines in (A and C) indicate the half-quantum of

379: resistance expected for the coherent regime. Uncertainties on

380: $R_{q}$ are displayed as hatched areas.}}\label{HBT4}\end{figure}

381: 

382: 

383: Finally, after numerical inversion of the conductance data, we can

384: separate the complex impedance into the contributions of the

385: capacitance, $1/C_\mu\omega$, and the relaxation resistance $R_q$.

386: The results in Fig.4 demonstrate deviations from standard

387: Kirchhoff's laws : the charge-relaxation resistance $R_q$ remains

388: constant in the regime where the quantum capacitance exhibits strong

389: transmission-dependent oscillations; this constant value equals,

390: within experimental uncertainty, half the resistance quantum as

391: prescribed by theory \cite{Buttiker93,Pretre96}.  In the weak

392: transmission regime, the Landauer formula is recovered because of

393: thermal broadening, and $R_q$ diverges as it does in the dc regime.

394: Furthermore, additional measurements at $4K$ prove that the

395: classical behavior is indeed recovered in the whole transmission

396: range whenever $k_BT\gg e^2/C_\mu$.

397: 

398: In conclusion, we have experimentally shown that the series

399: association of a quantum capacitor and a model quantum resistor

400: leads to a violation of the dynamical Kirchhoff's law of impedance

401: addition. In the fully coherent regime, the quantum resistor is no

402: longer given by the Landauer formula but by the half-quantized

403: charge-relaxation resistance predicted in

404: Ref.\cite{Buttiker93,Pretre96}.\\

405: The LPA is the CNRS-ENS UMR8551 associated with universities Paris 6

406: and Paris 7. The research has been supported by AC-Nanoscience,

407: SESAME grants and ANR-05-NANO-028.

408: 

409: 

410: \begin{thebibliography}{9}

411: 

412: \bibitem{Landauer57} R. Landauer, \emph{IBM J. Res. Dev.} \textbf{1}, 233

413: (1957).

414: 

415: \bibitem{Landauer70} R. Landauer, \emph{Phil. Mag.} \textbf{21}, 863 (1970).

416: %Electrical resistance of disordered one-dimensional lattice%  .

417: 

418: \bibitem{Anderson81} P.W. Anderson, \emph{Phys. Rev. B} \textbf{23}, 4828 (1981).

419: %New method for scaling theory of localization. II. Multichannel theory of a "wire" and possible extension to higher dimensionality %.

420: 

421: \bibitem{ButtImLandauer83} M. B\"{u}ttiker, Y. Imry, R. Landauer, S. Pinhas, \emph{Phys. Rev. B.} \textbf{31}, 6207 (1985).

422: %Generalized many Channel Conductance Formula with Application to small Rings %

423: 

424: \bibitem{GaoBO} B. Gao, A. Komnik, R. Egger, D.C. Glattli, A. Bachtold,

425: \emph{Phys. Rev. Lett.} \textbf{92}, 216804 (2004).

426: 

427: \bibitem{Review1} see for a review, S. Datta, \emph{Electronic Transport in Mesoscopic

428: Systems}, Cambridge University Press, Cambridge, (1997).

429: 

430: \bibitem{Review2} see for a review, Y Imry, \emph{Introduction

431: to Mesoscopic Physics}, Oxford University Press, Oxford (1997).

432: 

433: \bibitem{ButtLand} M. B\"uttiker, \emph{Phys. Rev. Lett.} \textbf{57}, 17611764 (1986)

434: 

435: \bibitem{Buttiker93} M. B\"{u}ttiker, A. Pr\^{e}tre, H. Thomas, \emph{Phys. Rev. Lett.} \textbf{70}, 4114

436: (1993).

437: 

438: \bibitem{Pretre96} A. Pr\^{e}tre, H. Thomas, M. B\"{u}ttiker, \emph{Phys. Rev. B.}

439: \textbf{54}, 8130 (1996).

440: 

441: \bibitem{Pilgram} S. Pilgram, M. B\"uttiker, \emph{Phys. Rev. Lett.} \textbf{89}, 200401

442: (2002).

443: 

444: \bibitem{Clerk} A.A. Clerk, S.M. Girvin, A.D. Stone, \emph{Phys. Rev. B.}

445: \textbf{67}, 165324  (2003).

446: %Quantum-limited measurement and information in mesoscopic detectors%

447: 

448: \bibitem{ChargeNoise1} M. B\"uttiker, H. Thomas, A. Pr\^{e}tre, \emph{Phys. Lett. A} \textbf{180},

449: 364 (1993).

450: %mesoscopic capacitors%

451: 

452: \bibitem{ChargeNoise2} Ya. M. Blanter, M. B\"uttiker, \emph{Phys. Rep.} \textbf{336}, 1

453: (2000).

454: %Shot Noise in Mesoscopic Conductors%

455: 

456: \bibitem{ChargeNoise3} F.W.J. Hekking, J.P. Pekola, \emph{Phys. Rev. Lett.} \textbf{96}, 056603

457: (2006).

458: 

459: \bibitem{Seelig} G. Seelig, S. Pilgram, A. N. Jordan, M. B\"uttiker, \emph{Phys. Rev. B.} \textbf{68}, 161310

460: (2003).

461: 

462: \bibitem{Burke} P. J. Burke, \emph{IEEE T Nanotechnol} \textbf{2} (1),

463: 55 (2003).

464: %"An RF Circuit Model for Carbon Nanotubes,"%

465: 

466: \bibitem{Ashoori92a} R. C. Ashoori \emph{et al.},  \emph{Phys. Rev. Lett.} \textbf{68},

467: 3088 (1992).

468: %single electron capacitance spectroscopy of discrete quantum levels%

469: 

470: \bibitem{Ashoori92b} R. C. Ashoori \emph{et al.},

471: \emph{Phys. Rev. Lett.} \textbf{71}, 613 (1992).

472: %N-electron ground state energies of a quantum dot in magnetic field%

473: 

474: \bibitem{SOM} Materials and methods are available as supporting material on

475: Science Online.

476: 

477: \bibitem{Gabelli06} J. Gabelli, thesis,

478: Universit\'e Pierre et Marie Curie, Paris, 2006;

479: %{\emph Evidence for quantum coherence in dynamical electronic transport}%.

480: On line access at (http://tel.ccsd.cnrs.fr/tel-00011619).

481: 

482: \bibitem{Butt90} M. B\"{u}ttiker,

483: \emph{Phys. Rev. B} \textbf{41}, 7906 (1990).

484: % Quantized transmission of a saddle-point constriction %.

485: 

486: \bibitem{Averin92} D.V. Averin, Y. Nazarov, \emph{Single charge tunneling}, chap.6, NATO-ASI series B, Vol.\textbf{294}, Plenum, New York, 1992.

487: 

488: \bibitem{Glattli93} D.C. Glattli,

489: \emph{Physica B} \textbf{189}, 88 (1993).

490: % Coulomb blockade and off-resonance tunneling in small electronic systems %.

491: 

492: \bibitem{Matveev} K.A. Matveev, \emph{Phys. Rev. B.}\textbf{51} , 1743

493: (1995).

494: 

495: \bibitem{Glazman98} L.I. Glazman, I.L. Aleiner, \emph{Phys. Rev. B.} \textbf{57},

496: 9608 (1998).

497: 

498: \bibitem{Brouwer05} P.W. Brouwer, A. Lamacraft, K. Flensberg

499: \emph{Phys. Rev. B.} \textbf{72}, 075316 (2005), and references

500: therein.

501: %Nonequilibrium theory of Coulomb blockade in open quantum dots%

502: 

503: %\bibitem{Chaotic}  P. W. Brouwer and M. B\(uttiker, Europhys. Lett., 37 (7), pp. 441-446 (1997)

504: 

505: \end{thebibliography}

506: 

507: \end{document}

508: