1: % March 23 2004 TB after 3rd round
2: % Jan 23, 2004 from Ramon
3: % Jan 22, 2004 version from Ramon
4: % word `RABI oscillations, RABI resonance -> oscillations, resonance'. This is not a RABI oscillation,
5: % but a QM coherent oscillation! RABI oscillationsare the ones between atom and field modes.
6: % new version after referee's comments. Oct 8
7: % final version July 02, 2003
8: % comments from Buettiker
9: % new suggestions from Leo, remarks from Till
10: % June 12 : last modif. before sending to Leo: PACS
11: % June 10 : built in Ramon's latest changes from ab_June10.tex (hopefully forgot nothing);
12: % June 10 : changed := into \equiv; defined unit vector; changed cpation 3; added references: Makhlin et al, Girvin paper, Loss paper, recent noise+phonon mode paper,
13: % June 6
14: % June 2
15: \documentclass[amssymb,prl,twocolumn,showpacs]{revtex4}
16: \usepackage{amsmath,amsthm,amsfonts}
17: %\usepackage{epsf}
18: \usepackage{amssymb}
19: \usepackage{graphicx}
20: \renewcommand{\topfraction}{1}
21: \renewcommand{\textfraction}{0}
22: \begin{document}
23: \title{
24: Shot Noise Spectrum of Open Dissipative Quantum Two-level Systems}
25: \author{Ram\'on Aguado $^1$ and Tobias Brandes $^2$}
26: %\affiliation{TMR network}
27: \affiliation{1-Departamento de Teor\'{\i}a de la Materia Condensada,
28: Instituto de Ciencia de Materiales de Madrid,
29: CSIC, Cantoblanco 28049, Madrid, Spain}
30: \affiliation{2-Department of Physics, %University of Manchester Institute of
31: %Science and Technology (
32: UMIST, P.O. Box 88, Manchester M60 1QD, United Kingdom}
33: \begin{abstract}
34: We study the current noise spectrum of {qubits} under transport conditions
35: in a dissipative bosonic environment. We combine
36: (non-)Markovian master equations with correlation functions in Laplace-space to
37: derive a noise formula for both weak and strong coupling to the bath.
38: The coherence-induced reduction of noise
39: is diminished by weak dissipation and/or a large {level separation (bias)}.
40: For weak dissipation, we demonstrate that the dephasing and relaxation rates of
41: the two-level system can be extracted from noise. In the
42: strong dissipation regime, the localisation-delocalisation transition
43: becomes visible in the low-frequency noise.
44: \end{abstract}
45: \date{\today{ }}
46: \pacs{
47: 72.70.+m %Noise processes and phenomena
48: 73.23.Hk %Coulomb blockade; single-electron tunneling
49: 73.63.Kv %Quantum dots
50: }
51: \maketitle
52: The way a quantum two-level system (qubit)
53: loses coherence due to the coupling with a noisy environment has
54: been the subject of intense research for many years
55: %%
56: \cite{Legetal87,Weiss}.
57: This fundamental problem has received a great deal of
58: attention due to recent advances in solid state devices in which
59: quantum two-level systems (TLS) have been realized using different
60: degrees of freedom (charge, spin, flux) \cite{qubitexperiment}.
61: Interest in
62: current {\em noise} \cite{BB00}, in particular in presence of dephasing and dissipation \cite{SU92},
63: has risen owing to the possibility of extracting
64: valuable information not available in conventional dc transport
65: experiments.
66:
67:
68:
69: % CHANGED
70: In this Letter, we demonstrate that current noise
71: %and current fluctuations
72: in coupled quantum dots or Cooper pair boxes reveal
73: the complete dissipative, internal dynamics of qubits
74: % CHANGED two-level systems (`qubits')
75: coupled to external electron reservoirs.
76: % CHANGED
77: We develop a formalism that allows us to make quantitative
78: predictions for the frequency ($\omega$) dependent
79: charge and current noise for {\em arbitrary} dissipative environments.
80: We find a
81: reduction of noise by coherent oscillations, weakened by
82: increasing the bias or weak dissipation.
83: The latter suppresses shot noise at $\omega=0$ and large bias
84: due to spontaneous boson emission.
85: Importantly, the dephasing and relaxation rates of the TLS can be extracted from noise.
86: Our formulation
87: includes non-Markovian memory effects \cite{LD03} and
88: the strong coupling limit, where we observe
89: a re-establishing of the full shot noise
90: due to the formation of polarons as new quasi-particles.
91: %%%% OPTIONAL
92: %This could be relevant
93: %for systems where boson confinement \cite{Blietal00,DBK02,Paretal00} plays a role.
94: %%%% OPTIONAL END
95:
96: In the following, we assume that the TLS is
97: defined in a double
98: quantum dot (DQD) device \cite{Fujetal98,Hayetal03}. We point out, however, that our method can also
99: be applied to charge qubits realized in a Cooper pair (CP) box \cite{MSS01,CGNS02,Choi03}, see below.
100: %i.e, a
101: %superconducting single electron transistor (SSET) where the
102: %charging energy is much larger than the Josephson energy,
103: %$E_C \gg E_J$, see below.
104: DQD´s in the regime of strong Coulomb
105: blockade can be tuned into a regime that is governed by a (pseudo)
106: spin--boson (SB) model (dissipative two--level system
107: \cite{Legetal87}), coupled to reservoirs\cite{BK99} ${\cal
108: H}={\cal H}_{SB}+{\cal H}_{res}+{\cal H}_{T}$. Here, ${\cal
109: H}_{SB}$ describes one additional `transport' electron which
110: tunnels between a left (L) and a right (R) dot with energy
111: difference $\varepsilon$ and inter--dot coupling $T_c$, and is
112: coupled to a dissipative bosonic bath (${\cal H}_B= \sum_{\bf
113: Q}\omega_{Q} a^{\dagger}_{\bf Q} a_{\bf Q}$),
114: \begin{equation}\label{modelhamiltonian}
115: {\cal H}_{SB}\!=\! \Big[\frac{\varepsilon}{2}
116: +\sum_{\bf Q} \frac{g_{Q}}{2} \left(a_{-\bf Q} + a^{\dagger}_{\bf
117: Q}\right)\Big]
118: \hat{\sigma}_z
119: \!+\! T_c
120: \hat{\sigma}_x
121: \!+\!{\cal H}_B.
122: \end{equation}
123: The effective Hilbert space of the closed system consists of two
124: states $|L \rangle=|N_L+1,N_R \rangle$ and $|R \rangle=|N_L,N_R+1
125: \rangle$, such that the system is defined by a `pseudospin' \cite{BK99}
126: $\hat{\sigma}_z \equiv |L \rangle \langle L|-|R \rangle \langle
127: R|\equiv \hat{n}_L-\hat{n}_R$ and $\hat{\sigma}_x\equiv |L \rangle \langle
128: R|+|R \rangle \langle L|\equiv \hat{p}+\hat{p}^{\dagger}$. The effects
129: of the
130: %bosonic CHANGED
131: bath can be encapsulated in the spectral density
132: $J({\omega})\equiv\sum_{\bf Q} |g_{Q}|^2\delta(\omega-\omega_Q)$,
133: where $\omega_Q$ are the frequencies of the bosons and the $g_Q$
134: denote interaction constants.
135: %Here, we are only interested
136: %in the simplest form of boson coupling
137: %to $\sigma_z$
138: %(cf. \cite{KS02}).
139: { When showing
140: %particular CHANGED
141: results we will be using
142: $J(\omega)={2\alpha}{\omega}
143: \left[1-{\omega_d}/{\omega}\sin\left({\omega}/{\omega_d}\right)\right]
144: e^{-\omega/\omega_c}$ for piezoacoustic phonons in lateral DQD´s
145: with $\omega_d$ depending on the
146: %qubit CHANGED
147: geometry \cite{BK99}, or a generic Ohmic bath $(\omega_d\to 0)$}:
148: $J(\omega)={2\alpha}\omega e^{-\omega/\omega_c}$. The
149: dimensionless parameter $\alpha$ reflects the strength of
150: dissipation and $\omega_c$ is a high energy cutoff \cite{Weiss}.
151: The coupling to
152: %external CHANGED
153: reservoirs ${\cal H}_{res}=\sum_{k_\alpha}\epsilon_{k_\alpha}
154: c_{k_\alpha}^{\dagger}c_{k_\alpha}$
155: is described by ${\cal H}_T=\sum_{k_\alpha{}}
156: (V_k^\alpha{} c_{k_\alpha{}}^{\dagger}s_\alpha{}+H.c.)$, where
157: $\hat{s}_\alpha=|0 \rangle
158: \langle \alpha|$ ($\alpha$=L,R) and
159: the extra
160: state $|0 \rangle=|N_L,N_R \rangle$ describes an `empty' DQD, such that
161: %the completeness relation is now CHANGED
162: $1=\hat{n}_0+\hat{n}_L+\hat{n}_R$.
163:
164: The full model described by ${\cal H}$
165: %offers the possibility CHANGED
166: allows to study
167: non--equilibrium properties, such as the inelastic stationary
168: current or current noise, through an open dissipative TLS.
169: We describe its dynamics by a reduced, with respect to reservoirs,
170: statistical operator $\rho(t)$.
171: Introducing the vectors ${\bf A}\equiv
172: (\hat{n}_L,\hat{n}_R,\hat{p},\hat{p}^{\dagger})^T$, ${\bf
173: \Gamma}=(\Gamma_L,0,0,0)^T\equiv\Gamma_L{\bf e}_1$
174: %(${\bf e}_i$ are unit vectors)
175: and a matrix memory kernel ${M}$,
176: the equations of motion (EOM) of the expectation values \cite{BV02}
177: (with $\langle \hat{O} \rangle\equiv \sum_{i=0,L,R}{\rm Tr_{bath}} \langle i |\hat{O}\rho(t)|i\rangle$)
178: %can be written in matrix form as
179: read in matrix form
180: %[$\langle..\rangle\equiv {\rm Tr_{dot}}{\rm Tr_{bath}} ..\rho(t)]$,
181: \begin{equation}\label{expectation}
182: \langle {\bf A}(t)\rangle \!=\! \langle {\bf A}(0)\rangle
183: + \int_{0}^{t}dt' \left\{{M}(t-t')\langle {\bf A}(t')\rangle + {\bf \Gamma} \right\}.
184: \end{equation}
185: Eq.~(\ref{expectation}) can be %%formally
186: solved in Laplace space as
187: $\langle\hat{\bf A}(z)\rangle= [z-z\hat{M}(z)]^{-1}(\langle{\bf
188: A}(0)\rangle+{\bf \Gamma}/z)$ and serves as a starting point for the
189: analysis of stationary ($1/z$ coefficient in Laurent series for $z\rightarrow
190: 0$) and non-stationary quantities. The memory kernel has a
191: block structure
192: \begin{equation}\label{block}
193: z\hat{M}(z)=\left[
194: \begin{matrix}
195: -\hat{G} & \hat{T}_c\\ \hat{D}_{z}& \hat{\Sigma}_{z}
196: \end{matrix}\right], \quad
197: \hat{G}\equiv
198: \left(
199: \begin{matrix}
200: \Gamma_L & \Gamma_L\\0&\Gamma_R
201: \end{matrix}\right),
202: \end{equation}
203: where $\hat{T}_c \equiv -iT_c(1-\sigma_x)$,
204: and the coupling to the reservoirs within Born and Markov (BM) approximation with respect to ${\cal H}_{T}$
205: \cite{BK99,SN96} is given by
206: $\Gamma_{\alpha}=2\pi\sum_{k_{\alpha}}|V_k^{\alpha}|^2\delta(\epsilon-\epsilon_{k_{\alpha}})$
207: (we assume Fermi distributions for the reservoirs $f_L=1$ and
208: $f_R=0$; large voltage regime).
209: The blocks $\hat{D}_{z}$ and $\hat{\Sigma}_{z}$
210: are determined by the EOM for the coherences (off--diagonal elements)
211: $\langle\hat{p}\rangle=\langle\hat{p}^{\dagger }\rangle^*$
212: and contain the complete information on dephasing of the system.
213: In general, no exact solution is available
214: %%% OPTIONAL
215: %\cite{footnote1}
216: %%% OPTIONAL
217: but we will present approximate results now:
218: for weak coupling
219: to the bosons, one can use perturbation theory (PER)
220: in $\alpha$
221: in the correct basis of the hybridised states of the TLS.
222: In BM approximation, the resulting expressions are:
223: %independent of $z$,
224: \begin{equation}\label{PERb}
225: \hat{D}^{\rm PER}=\hat{T}_c+
226: \left(
227: \begin{matrix}
228: \gamma_+ & - \gamma_-\\ \gamma_+ & -\gamma_-
229: \end{matrix}\right),\quad
230: \hat{\Sigma}^{\rm PER}=
231: \left(
232: \begin{matrix}
233: E&0\\0&E^*
234: \end{matrix}\right),
235: \end{equation}
236: %%%%%%%% CHANGED
237: where $E=i\varepsilon-\gamma_p-\frac{\Gamma_R}{2}$,
238: $\gamma_p\equiv 2\pi \frac{T_c^2}{\Delta^2} J(\Delta) \coth
239: \left(\beta \Delta /2\right)$ and
240: $\gamma_{\pm}\equiv-\frac{\varepsilon T_c}{\Delta^2} \frac{\pi}{2}
241: J(\Delta) \coth \left(\beta \Delta /2\right)\mp
242: \frac{T_c}{\Delta} \frac{\pi}{2}J(\Delta)$
243: completely determine dephasing and relaxation in the system.
244: % CHANGE END
245: Here, $\Delta\equiv\sqrt{\varepsilon^2+4T_c^2}$ is the
246: hybridization splitting and $\beta=1/k_BT$.
247:
248: On the other hand,
249: for strong electron-boson coupling, one has to start from a
250: polaron--transformed frame (strong coupling, POL), leading to an
251: integral equation \cite{BK99} which involves the boson correlation
252: function $C(t)\equiv\exp({-\int_0^{\infty}d\omega
253: \frac{J(\omega)}{\omega^2} \left[ \left(1- \cos \omega t\right)
254: \coth \left(\frac{\beta \omega}{2}\right) + i \sin \omega t
255: \right]})$. Introducing
256: $C^{[*]}_\varepsilon(z)\equiv\int_{0}^{\infty}dte^{-zt}e^{[-]i\varepsilon
257: t} C^{[*]}(t)$, the resulting matrices in $z$-space are
258: \begin{equation}\label{POLb}
259: \hat{D}_{z}^{\rm POL}=iT_c
260: \left(\begin{matrix}
261: -1 & {\hat{C}^*_{-\varepsilon}}/{\hat{C}_\varepsilon}\\
262: 1 & -{\hat{C}_{-\varepsilon}}/{\hat{C}^*_\varepsilon}
263: \end{matrix}\right),
264: %\quad
265: \hat{\Sigma}_{z}^{\rm POL}=
266: \left(
267: \begin{matrix}
268: \tilde{E}&0\\0&\tilde{E}^*
269: \end{matrix}\right),
270: \end{equation}
271: with $\tilde{E}^{[*]}\equiv z-1/C_\varepsilon^{[*]}(z)-\Gamma_R/2$.
272: In contrast to the PER solution, where
273: $M(\tau)=M=z\hat{M}(z)$ is time-independent,
274: $M^{\rm POL}(\tau)$ is time-dependent and
275: $z\hat{M}(z)$ depends on $z$ in the POL approach \cite{footnote2}.
276: We note
277: that $\mbox{Re}[C_\varepsilon (z)]|_{z=\pm i\omega}=\pi
278: P(\varepsilon\mp\omega)$
279: where $P(\varepsilon)$ is the probability for inelastic tunneling
280: with energy transfer $\varepsilon$ \cite{Weiss}.
281: %%% OPTIONAL
282: %In particular, for Ohmic dissipation ($T=0$) one has
283: %$C_\varepsilon(0)=-i/\omega_c({-\varepsilon}/{\omega_c})^{2\alpha-1}
284: %e^{-\varepsilon/\omega_c} \Gamma(1-2\alpha,{-\varepsilon}/{\omega_c})$, such that
285: %$P(\varepsilon)
286: %={(\varepsilon/\omega_c)^{2\alpha-1}} e^{-\varepsilon/\omega_c}
287: %\theta(\varepsilon)/{(\omega_c\Gamma(2\alpha))}$, where $\Gamma(x)$ and $\Gamma(x,y)$ are the
288: %gamma and incomplete gamma functions respectively.
289: % OPTIONAL END
290:
291: As mentioned above, our model describes
292: a CP box as well, the transport through the DQD
293: being analogous to the Josephson Quasiparticle Cycle (JQP) of
294: the superconducting single electron transistor (SSET) with $E_C\gg E_J$,
295: such that only two charge states,
296: $|2\rangle$ (one excess CP
297: in the SSET) and $|0\rangle$ (no extra CP), are allowed. Two
298: consecutive quasiparticle events (with rates $\Gamma_{2}$ and
299: $\Gamma_{1}$) couple
300: $|2\rangle$ and $|0\rangle$ with another state $|1\rangle$ through
301: the cycle $|2\rangle\rightarrow |1\rangle\rightarrow
302: |0\rangle\Leftrightarrow |2\rangle$.
303: Interdot tunneling is analogous to coherent tunneling of a
304: CP through one of the junctions, and tunneling to and
305: from the DQD
306: %left dot and out of the right dot
307: is analogous to the two
308: quasiparticle events through the probe junction in the SSET \cite{Choi03}.
309: %The analogy is thus established by replacing $|L\rangle
310: %\rightarrow|0\rangle$, $|R\rangle \rightarrow|2\rangle$,
311: %$|0\rangle \rightarrow|1\rangle$, $2T_c\rightarrow E_J$,
312: %$\Gamma_{R(L)}\rightarrow \Gamma_{2(1)}$ \cite{Choi03}.
313:
314:
315:
316:
317: %In a quantum
318: %conductor out of equilibrium, electronic current noise originates
319: %from the dynamical fluctuations of the current away from its
320: %average $\Delta\hat{I}(t)\equiv \hat{I}(t)-\langle
321: %\hat{I}(t)\rangle$.\cite{BB00}
322: Current noise, which is described by the power spectral density $\mathcal
323: {S}_{I}(\omega)\equiv 2\int_{-\infty}^{\infty} d\tau
324: e^{i\omega\tau} \mathcal {S}_{I}(\tau)=\int_{-\infty}^{\infty}
325: d\tau e^{i\omega\tau}\langle
326: \{\Delta\hat{I}(\tau),\Delta\hat{I}(0)\} \rangle$, with $ \Delta\hat{I}(t)\equiv \hat{I}(t)-\langle
327: \hat{I}(t)\rangle$, is
328: a sensitive tool to study correlations between carriers \cite{BB00}.
329: %For uncorrelated carriers with charge $q$, $\mathcal{S}_I(0)=2qI$
330: %({\it full} shot noise or Poissonian
331: %noise).
332: The Fano factor ($\gamma\equiv\frac{\mathcal{S}_I(0)}{2qI}$)
333: quantifies deviations from the Poissonian noise,
334: $\mathcal{S}_I(0)=2qI$, which characterizes uncorrelated carriers
335: with charge $q$. Importantly, $S_I(\omega)$ has to be calculated
336: from the autocorrelations of the {\it total} current $I(t)$, i.e.
337: particle plus displacement current \cite{BB00}. Using current
338: conservation together with the Ramo-Shockley theorem, $I(t)=a
339: I_L(t) + b I_R(t)$ ($a$ and $b$, with $a+b=1$, depend on each
340: junction capacitance \cite{BB00}), one can express $\mathcal
341: {S}_{I}(\omega)$ in terms of the spectra of particle currents and
342: the charge noise spectrum $S_{Q}(\omega)$ \cite{Mozetal02},
343: %a relation for
344: %$\mathcal {S}_{I}(\omega)$ in terms of the spectra of particle
345: %currents and the charge noise spectrum:
346: \begin{equation}
347: \label{fullnoise}
348: S_I(\omega)=a S_{I_L}(\omega) + b
349: S_{I_R}(\omega)-a b\omega^2S_{Q}(\omega).
350: \end{equation}
351: %The factors $a= C_R/(C_L+C_R)$ and $b= C_L/(C_L+C_R)$ depend on
352: %the capacitance of each junction.\cite{BB00}
353: Note that in symmetric configurations, $a\approx b$, the charge
354: noise reduces the contribution from particle currents to the noise
355: spectrum. For $a\approx 1$ or $b\approx 1$ the main contribution
356: to noise comes from particle currents. At zero frequency
357: $S_I(0)=S_{I_L}(0)=S_{I_R}(0)$.
358: %The {\bf charge noise spectrum}
359: $S_{Q}(\omega)$ is defined as
360: \begin{eqnarray}
361: \label{charge}
362: S_Q(\omega) &\equiv& \lim_{t\to \infty}
363: \int_{-\infty}^{\infty} d\tau e^{i\omega\tau}\langle
364: \{\hat{Q}(t),
365: \hat{Q}(t+\tau)\}\rangle \nonumber\\
366: &=& 2{\mbox Re} \left\{ \hat{f}(z=i\omega) +
367: \hat{f}(z=-i\omega)\right\},
368: \end{eqnarray}
369: where $\hat{Q}=\hat{n}_L+\hat{n}_R$ and $\hat{f}(z)$ is the
370: Laplace transform of
371: \begin{eqnarray}
372: \label{fequation}
373: f(\tau) &=& \sum_{i,j=L,R}\langle \hat{n}_i(t)\hat{n}_j(t+\tau)\rangle
374: %\langle \hat{n}_L(t)\hat{n}_L(t+\tau)\rangle
375: %+\langle \hat{n}_L(t)\hat{n}_R(t+\tau)\rangle \nonumber\\
376: %&+& \langle \hat{n}_R(t)\hat{n}_L(t+\tau)\rangle + \langle
377: %\hat{n}_R(t)\hat{n}_R(t+\tau)\rangle.
378: \end{eqnarray}
379: %$f(\tau)$
380: and can be evaluated with the help of the charge correlation
381: functions ${\bf C}_\alpha(\tau)\equiv \langle \hat{n}_{\alpha}(t) {\bf
382: A}(t+\tau) \rangle$, as $f(\tau)=({\bf e}_1+{\bf
383: e}_2)[{\bf C}_L(\tau)+{\bf C}_R(\tau)]$. The EOM for ${\bf
384: C}_\alpha(\tau)$ can be obtained from the quantum regression theorem \cite{Carmichael}
385: whose solution is again expressed with the help of the resolvent $[z-z\hat{M}(z)]^{-1}$, cf. Eq.
386: (\ref{block}).
387: \par
388: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
389: \begin{figure}[t]
390: \includegraphics[width=0.8\columnwidth]{fig1.eps}
391: %\caption[]{a) Fano factor vs. bias $\varepsilon$ for
392: %different dissipative couplings $\alpha$.
393: %Parameters $T_c=3$, $\Gamma_L=\Gamma_R=0.15$, $\omega_c=500$,
394: %$\omega_d=10$, $T=2$ (in $\mu eV$) correspond to
395: %typical experimental values~\cite{Fujetal98} in double quantum dots.
396: %Lines: acoustic phonons, circles: generic ohmic environment $\omega_d=0$ (see text).
397: %b) Frequency dependent current noise
398: %(no dissipation, $T=0$, $\Gamma_L=\Gamma_R=0.01$).
399: %Inset: charge noise contribution.
400: %}
401: \caption[]{a) Fano factor vs. bias $\varepsilon$ for
402: different dissipative couplings $\alpha$.
403: Parameters $T_c=3$, $\Gamma=0.15$, $\omega_c=500$,
404: $\omega_d=10$, $T=2$ (in $\mu eV$) correspond to
405: typical experimental values~\cite{Fujetal98} in double quantum dots.
406: Lines: acoustic phonons, circles: generic ohmic environment $\omega_d=0$ (see text).
407: b) Frequency dependent current noise ($\alpha=0$, $T=0$, $\Gamma=0.01$).
408: Inset: (Top) Contribution to noise from particle currents $S_{I_R}(\omega)/2eI$. (Bottom) Charge noise contribution $\omega^2S_{Q}(\omega)/8eI$. $a=b=1/2$.}
409: \end{figure}
410: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
411: To calculate the contribution of particle currents to noise, we need to relate the
412: reduced dynamics of the qubit described by Eqs. (2-3) to reservoir
413: operators.
414: For $S_{I_R}(\omega)$, %this can be done by
415: we introduce
416: %degree of freedom CHANGED
417: %a counting variable $n$ (
418: the number $n$ of electrons that have
419: tunneled through the right barrier \cite{MSS01,EG02,RK03})
420: %which allows CHANGED
421: which defines generalized expectation values as
422: $O^{(n)}\equiv\sum_{i=0,L,R}{\rm Tr_{bath}} \langle n,i |\hat{O}\rho(t)|n,i\rangle$
423: (such that $\langle \hat{O} \rangle=\sum_n O^{(n)}$)
424: and write
425: \begin{eqnarray}\label{generalized}
426: \dot{n}_0^{(n)}&=&-\Gamma_L {n}_0^{(n)} + \Gamma_R {n}_R^{(n-1)}\nonumber\\
427: \dot{n}_{L/R}^{(n)}&=& \pm\Gamma_{L/R} {n}_0^{(n)} \pm iT_c \left( p^{(n)}-
428: [p^{(n)}]^{\dagger}\right)
429: \end{eqnarray}
430: and correspondingly for
431: %the coherences CHANGED
432: $p^{(n)}$ and $[p^{(n)}]^{\dagger}$
433: and the left barrier.
434: Eqs.(\ref{generalized}) allow one to calculate the particle current
435: and the noise spectrum from
436: $P_n(t)=n_0^{(n)}(t)+n_L^{(n)}(t)+n_R^{(n)}(t)$ which gives the
437: total probability of finding $n$ electrons in the collector by time
438: $t$. In particular, $I_R(t)=e\sum_n n\dot{P}_n(t)$
439: and $S_{I_R}$ can be calculated from \cite{Mac48}
440: %\cite{Mac48,RK03}
441: %the MacDonald´s
442: %formula \cite{Mac48,RK03}
443: \begin{eqnarray}
444: S_{I_R}(\omega)=2\omega e^2\int_0^{\infty} dt \sin(\omega t) \frac{d}{dt}
445: \left[ \langle n^2(t) \rangle - (t\langle I\rangle)^2 \right],
446: \end{eqnarray}
447: %where $\langle I\rangle$ is the stationary current and
448: where $\frac{d}{dt}\langle n^2(t)\rangle=\sum_n
449: n^2\dot{P}_n(t)=\Gamma_R\sum_{n=0}^{\infty} n {n}_R^{(n)} (t) +
450: \Gamma_R \sum_{n=0}^{\infty} {n}_R^{(n)} (t)$.
451: Solving Eqs.(\ref{generalized}) with the initial condition
452: $n_R^{(n)} (0)=\delta_{n,0}n_R(0)$, where $n_R(0)$ is the
453: stationary solution of Eqs. (2) \cite{RK03}, we get
454: \begin{eqnarray}\label{SIR}
455: S_{I_R}(\omega) &=& 2eI\left\{1+\Gamma_R\left[\hat{n}_R(-i\omega)
456: + \hat{n}_R(i\omega)\right]\right\},
457: \end{eqnarray}
458: with
459: $z\hat{n}_R(z)=\Gamma_Lg_+(z)/N(z)$, where
460: $N(z)\equiv[z+\Gamma_R+g_-(z)](z+\Gamma_L)$
461: $+(z+\Gamma_R+\Gamma_L)g_+(z)$ and
462: \begin{eqnarray}\label{gpm}
463: % g_{\pm}(z)=\pm iT_c\mbox{\rm Tr} \left(\left[z-\Sigma_z\right]^{-1}\hat{D}_z H_{\pm}\right),
464: g_{+[-]}(z)=\pm iT_c({\bf e}_1-{\bf e}_2)
465: \left[z-\hat{\Sigma}_z\right]^{-1}\hat{D}_z {\bf e}_{1[2]}.
466: %g^{\rm PER}_{\pm}(z)&\equiv&
467: % 2T_c\frac{T_c(\gamma_p+ \Gamma_R/2+z)-\varepsilon\delta_{\pm}}{(\gamma_p+ \Gamma_R/2+z)^2+\varepsilon
468: % ^2}\nonumber\\
469: % g^{\rm POL}_{+[-]}(z) &\equiv& T_c^2 \left[
470: %\frac{C^{[*]}_{[-]\varepsilon}(z)}{1+\frac{\Gamma_R}{2}C_\varepsilon(z)}+
471: %(C \leftrightarrow C^*)
472: %\right]
473: \end{eqnarray}
474: Eqs. (\ref{SIR},\ref{gpm}) demonstrate
475: the dependence of the current noise on the dephasing via the two-by-two blocks $\hat{D}_z$ and $\hat{\Sigma}_z$,
476: cf. Eq. (\ref{block},\ref{PERb},\ref{POLb}).
477: %where $H=\left( \begin{matrix} 1&-1\\0&0\end{matrix}\right)$.
478: %%
479: %The explicit expression are easily obtained as
480: Explicitly,
481: \begin{eqnarray}\label{gpm1}
482: %%$
483: g^{\rm PER}_{\pm}(z)&\equiv&
484: 2T_c\frac{T_c(\gamma_p+ \Gamma_R/2+z)-\varepsilon\gamma_{\pm}}{(\gamma_p+ \Gamma_R/2+z)^2+\varepsilon
485: ^2}\nonumber\\
486: %$ and
487: %%$
488: g^{\rm POL}_{+[-]}(z) &\equiv& T_c^2 \left[
489: \frac{C^{[*]}_{[-]\varepsilon}(z)}{1+\frac{\Gamma_R}{2}C_\varepsilon(z)}+
490: (C \leftrightarrow C^*)
491: \right].
492: %g^{\rm POL}_{+[-]}(z) &\equiv& T_c^2
493: %C^{[*]}_{[-]\varepsilon}(z)\left[1+({\Gamma_R}/{2})C_\varepsilon(z)\right]^{-1}
494: %+(C \leftrightarrow C^*).
495: \end{eqnarray}
496: %%%%
497: A similar derivation yields $S_{I_L}(\omega)=S_{I_R}(\omega)$.
498: The explicit expressions
499: Eq.(\ref{SIR},\ref{gpm},\ref{gpm1}), together with the inverse of a 4 by 4 matrix for the charge noise
500: Eqs.~(\ref{charge}), yield our key quantity $S_I(\omega)$, Eq.~(\ref{fullnoise}).
501:
502:
503:
504: {\it Zero frequency (Shot Noise)}.--
505: In the zero frequency limit
506: $z\rightarrow 0$,
507: one obtains
508: \begin{eqnarray}
509: \label{shot}
510: S_I(0) = 2eI \left(1+2\Gamma_R \frac{d}{dz}\left[z
511: \hat{n}_R(z)\right]_{z=0}\right).
512: \end{eqnarray}
513: Eqs.~(\ref{shot})
514: allows to
515: investigate the shot noise of open dissipative TLS's
516: for {\it arbitrary environments}. In contrast to
517: non-interacting mesoscopic conductors, the noise cannot be
518: written in the Khlus-Lesovik form $S_I(0)=
519: 2 e^2 \int \frac{dE}{2\pi} t(E) [1-t(E)]$ with an effective
520: transmission coefficient $t(E)$.
521: Without bath,
522: %one has
523: %$g_{\pm}(z) = T_c^2
524: %\frac{2z+\Gamma_R}{\left(z+\frac{\Gamma_R}{2}\right)^2+\varepsilon^2}$
525: %such that
526: %\begin{eqnarray}
527: %& & \frac{d}{dz}\left[z \hat{n}_R(z)\right]_{z=0}=
528: %-\frac{4T_c^2\Gamma_L}{\Gamma_R}\times\nonumber\\& &
529: %\frac{4\varepsilon^2(\Gamma_R-\Gamma_L)+3\Gamma_L\Gamma_R^2+\Gamma_R^3+8\Gamma_RT_c^2}
530: %{\left[\Gamma_L\Gamma_R^2 + 4\Gamma_L\varepsilon^2
531: %+4T_c^2(\Gamma_R+2\Gamma_L)\right]^2},
532: %\end{eqnarray}
533: %and
534: we recover the results of Ref.~[\onlinecite{EG02}] (shot noise
535: of DQD's) and Ref.~[\onlinecite{Choi03}] (shot noise
536: of the CP box). For $\alpha=0$ and
537: $\Gamma\equiv\Gamma_L=\Gamma_R$ (Fig. 1a, solid line), the
538: smallest Fano factor is reached for $\varepsilon =0$ where quantum
539: coherence strongly suppresses noise. The maximum suppression
540: ($\gamma=1/5$) is reached for $\Gamma =2\sqrt{2}T_c$. For large
541: $\varepsilon>0$ ($\varepsilon <0$) the charge becomes localized in
542: the right (left) level, $S_I(0)$ is dominated by only one Poisson
543: process, namely the noise of the right (left) barrier, and
544: $\gamma\rightarrow 1$.
545: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
546: \begin{figure}[t]
547: \begin{center}
548: %\includegraphics[width=0.75\columnwidth]{spin-spin1.eps}
549: \includegraphics[width=0.7\columnwidth]{fig2.eps}
550: %\includegraphics[width=0.5\columnwidth]{polaron.eps}
551: \end{center}
552: \caption[]{Effect of ohmic dissipation on current noise near resonance
553: ($\varepsilon=10$, $\Gamma=0.01$,
554: and $\alpha=0.005, 0.01, 0.02$
555: corresponding to $\gamma_p\approx 4.74\Gamma, 9.47\Gamma, 18.95\Gamma$).
556: %$S_I(\omega)$ can be directly related with the
557: %internal dissipative dynamics of the two-level system. This is shown in
558: %the right inset where
559: Inset: (Right)
560: pseudospin correlation function $S_z(\omega)$. %(see text) of the qubit is plotted.
561: Arrows indicate %the calculated
562: relaxation rate
563: $(\Gamma+\gamma_p)/\Delta\approx 0.005$ for $\alpha=0.005$.
564: (Left)
565: low frequencies region near shot noise limit $\omega=0$. %The half-width at low frequencies (relaxation)
566: %is {\it twice} that around $\Delta$ (dephasing), see main text.
567: }
568: %{$\frac{S(\omega)}{2I}$ vs.
569: %$\omega$ for different values of $g$ at fixed $\varepsilon =10$
570: %and Ohmic environment. Left inset: low frequencies region. Right
571: %inset: Spin correlation function $S_z(\omega)$ vs. $\omega$.}
572: \end{figure}
573: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
574: For $\alpha\neq 0$, this mechanism is strongly
575: affected by the possibility of exchanging energy quanta with the
576: bath.
577: %%%%%%% CHANGED
578: The effect of a bosonic bath on noise in a mesoscopic scatterer was first discussed in
579: \cite{US93}. For the TLS discussed here, spontaneous
580: emission (for $\varepsilon >0$) occurs even at very low temperatures \cite{Fujetal98,BK99}
581: and the noise is
582: reduced \cite{SU92} well below the Poisson limit (Fig. 1a).
583: %%%%%%% CHANGE END
584: The maximum suppression is now reached when the elastic and inelastic rates coincide, i.e.,
585: $\gamma_p=\Gamma_R$, as we have checked numerically.
586: For large
587: couplings, spontaneous emission leads to a very asymmetric Fano factor that goes
588: from $\gamma\approx 1$ to $\gamma\approx 0.5$ as $\varepsilon$
589: changes sign (not shown here).
590:
591: {\it Finite frequencies}.--
592: For finite $\omega$, % ($a=b=1/2$),
593: the numerical results for
594: $\alpha=0$ are shown in Fig. 1b where we plot $S_I(\omega)$ for different
595: values of $\varepsilon$. The background noise is half the Poisson
596: value as one expects for a symmetric structure. $\gamma$ deviates
597: from this value around $\omega=0$
598: where the noise has a peak
599: and $\omega=\Delta$ where the
600: noise is suppressed. The dip in the Fano factor directly reflects the
601: resonance of the subtracted charge noise $S_{Q}(\omega)$ around $\Delta$ (inset Fig. 1b), cf. Eq.
602: (\ref{fullnoise}).
603: %where the noise is enhanced
604: %(peak in the Fano factor, Fig. 1b) and $\omega=\Delta$ where the
605: %noise is suppressed (dip in the Fano factor, Fig. 1c).
606: %%As we just
607: %%mentioned,
608: An increase of $\varepsilon$ localizes the qubit and,
609: thus, the zero-frequency noise reaches $\gamma\rightarrow 1$.
610: %This behavior is accompanied by strong localization of charge and thus
611: %leads to a reduction of coherence in the two-level system.
612: Moreover, the dip in the high frequency noise at $\omega=\Delta$
613: (Fig. 1b) is progressively destroyed (reduction of quantum
614: coherence) as $\varepsilon$ increases which is consistent with the
615: previous argument.
616: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
617: \begin{figure}[t]
618: \begin{center}
619: %\includegraphics[width=0.45\columnwidth]{spin-spin1.eps}
620: \includegraphics[width=0.7\columnwidth]{fig3.eps}
621: \end{center}
622: \caption[]{Low frequency current noise in qubit with strong ohmic dissipation
623: ($T=0$, $\varepsilon=10$, $T_c=3$, $\Gamma_L=\Gamma_R=0.01$).
624: Inset: Shot noise for $\omega=0$.
625: }
626: \end{figure}
627: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
628: A similar reduction of the dip at $\omega=\Delta$ occurs at
629: fixed $\varepsilon$ and $\Gamma$ with increasing dissipation
630: (Fig 2) in the weak coupling (PER) regime. This behavior demonstrates
631: that $S_I(\omega)$ reveals the complete internal dissipative dynamics
632: of the TLS.
633: The above argument can be further substantiated by
634: plotting also the symmetrized pseudospin
635: correlation function $S_z(\omega)=1/2\int_{-\infty}^{\infty} d\omega
636: e^{i\omega\tau}\langle\{\hat{\sigma}_z(\tau),\hat{\sigma}_z\}\rangle$ (Fig. 2, right inset)
637: which is commonly used to investigate the dynamics of the SB problem \cite{Weiss}.
638: Both functions reflect in the same fashion how the coherent dynamics of the system progressively
639: gets damped by the bosonic bath.
640:
641: In particular, the {\em dephasing rate} can be extracted from the half-width of
642: $S_I(\omega)$ around
643: $\omega=\Delta$. For an Ohmic environment, $\gamma_d^b=\gamma_p/2+2\pi\alpha(\frac{\varepsilon}{\Delta})^2 k_BT$, such that
644: the total dephasing rate is
645: $\gamma_d(T=0)=\gamma_d^b+\Gamma/2=(\gamma_p+\Gamma)/2$ (Fig. 2, arrows denote full-width, i.e.
646: $2\gamma_d\approx \gamma_p$ as $\alpha$ increases).
647: Close to $\omega=0$, the peak in $S_I(\omega)$ for $\alpha=0$
648: changes into a dip around
649: $\omega=0$ reflecting incoherent relaxation dynamics for $\alpha\neq 0$.
650: The half-width is now given by the {\em relaxation rate} such that
651: the full-width of $S_I(\omega)$ around $\omega=0$ is
652: {\it twice} that of the high frequency noise (Fig 2, left inset).
653:
654: The results for the strong coupling (POL) regime are presented in Fig 3.
655: Near $\omega=0$,
656: POL and PER yield nearly identical results for the noise $S_I(\omega)$
657: at very small $\alpha$ (not shown here).
658: The cross-over to Poissonian noise near $\omega=0$ with increasing $\alpha$ indicates
659: the formation of localized polarons. The delocalisation-localisation transition \cite{Legetal87,Weiss}
660: of the spin-boson model at $\alpha=1$ is reflected in a
661: change of the analyticity of $C_\varepsilon$ and the shot noise near zero bias
662: (Fig 3, inset).
663: Similar physics has been found recently in
664: the suppression of the persistent current $I(|\varepsilon|)\propto
665: \mbox{\rm Im}C_{-|\varepsilon|}$ through
666: a strongly dissipative quantum ring containing a quantum dot with bias $\varepsilon$
667: \cite{CB01}. Although POL becomes less reliable for $\alpha<1$ and smaller bias,
668: the non-symmetry in $\varepsilon$ of the shot noise and the inelastic current $\propto
669: \mbox{\rm Re} C_\varepsilon$ reflects the
670: `open' topology of our TLS in the non-linear transport regime.
671:
672: To conclude, our results demonstrate that frequency-dependent current noise
673: provides detailed information about the internal, dissipative dynamics
674: of open quantum two-level system such as
675: double quantum dots or Cooper pair boxes. The
676: weak coupling regime should be close to current experiments \cite{Leonoise}
677: in these systems, where
678: we expect our predictions to be tested in the near future.
679:
680: We thank Markus B\"uttiker, Leo P. Kouwenhoven and Till Vorrath for useful discussions.
681: This work was supported by
682: EPSRC GR/R44690, DFG BR 1528 and by the
683: MCYT of Spain through the "Ram\'on y Cajal" program and grant MAT2002-02465 (R. A.).
684:
685:
686: %A first step towards these challenging experiments has been
687: %demonstrated recently in Ref. [\onlinecite{Leonoise}].
688:
689:
690:
691:
692: %\bibliographystyle{apsrev}
693: %\bibliography{../base/tob,../base/book,../base/ep,../base/omega,../base/ee,../base/trans,../base/opt,../base/tobias,../base/wire,../base/loc,../base/dot,../base/qbit,../base/hall,../base/noise,../base/phonon}
694:
695: %\end{document}
696:
697:
698: \begin{thebibliography}{20}
699: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
700: \expandafter\ifx\csname bibnamefont\endcsname\relax
701: \def\bibnamefont#1{#1}\fi
702: \expandafter\ifx\csname bibfnamefont\endcsname\relax
703: \def\bibfnamefont#1{#1}\fi
704: \expandafter\ifx\csname citenamefont\endcsname\relax
705: \def\citenamefont#1{#1}\fi
706: \expandafter\ifx\csname url\endcsname\relax
707: \def\url#1{\texttt{#1}}\fi
708: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
709: \providecommand{\bibinfo}[2]{#2}
710: \providecommand{\eprint}[2][]{\url{#2}}
711:
712: \bibitem[{\citenamefont{{A. J. Leggett, S. Chakravarty, A. T. Dorsey, M. P. A.
713: Fisher, A. Garg, and W. Zwerger}}(1987)}]{Legetal87}
714: \bibinfo{author}{\bibnamefont{{A. J. Leggett, {\it et al}
715: }}}, \bibinfo{journal}{Rev. Mod.
716: Phys.} \textbf{\bibinfo{volume}{59}}(\bibinfo{number}{1}),
717: \bibinfo{pages}{1} (\bibinfo{year}{1987}).
718:
719: \bibitem[{\citenamefont{Weiss}(1993)}]{Weiss}
720: \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Weiss}},
721: \emph{\bibinfo{title}{Quantum Dissipative Systems}},
722: (\bibinfo{publisher}{World Scientific}, \bibinfo{address}{Singapore},
723: \bibinfo{year}{1993}).
724:
725: \bibitem[{\citenamefont{{Y. Nakamura {\em et al}, Nature, {\bf 398} (1999),
726: 786; D. Vion {\em et al}, Science {\bf 296}, 886 (2002); I. Chiorescu {\em et
727: al}, {\em ibid} {\bf 299}, 1869 (2003); Yu. A. Pashkin {\em et al}, Nature {\bf
728: 421}, 823}}(2003)}]{qubitexperiment}
729: \bibinfo{author}{\bibnamefont{{Y. Nakamura {\em et al}, Nature, {\bf 398}
730: (1999), 786; D. Vion {\em et al}, Science {\bf 296}, 886 (2002); I. Chiorescu
731: {\em et al}, {\em ibid.} {\bf 299}, 1869 (2003); Yu. A. Pashkin {\em et al},
732: Nature {\bf 421}, 823}}} (\bibinfo{year}{2003}).
733:
734: \bibitem[{\citenamefont{{Y. M. Blanter and M. B\"uttiker}}(2000)}]{BB00}
735: \bibinfo{author}{\bibnamefont{{Y. M. Blanter and M. B\"uttiker}}},
736: \bibinfo{journal}{Phys. Rep.} \textbf{\bibinfo{volume}{336}},
737: \bibinfo{pages}{1} (\bibinfo{year}{2000}).
738:
739: \bibitem[{\citenamefont{{A. Shimizu and M. Ueda}}(1992)}]{SU92}
740: \bibinfo{author}{\bibnamefont{{A. Shimizu and M. Ueda}}},
741: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{69}},
742: \bibinfo{pages}{1403} (\bibinfo{year}{1992});
743: Phys. Rev. Lett. {\bf 86}, 3694 (2001).
744:
745:
746: \bibitem[{\citenamefont{{D.~Loss, D.~P.~DiVincenzo,
747: cond-mat/0304118}}(2003)}]{LD03}
748: \bibinfo{author}{\bibnamefont{{D.~Loss, D.~P.~DiVincenzo, cond-mat/0304118}}}
749: (\bibinfo{year}{2003}).
750:
751: % OPTIONAL START
752: %\bibitem[{\citenamefont{{R. H. Blick, F. G. Monzon, W. Wegscheider, M. Bichler,
753: % F. Stern, and M. L. Roukes}}(2000)}]{Blietal00}
754: %\bibinfo{author}{\bibnamefont{{R. H. Blick {\em et al}
755: %}}}, \bibinfo{journal}{Phys. Rev. B}
756: % \textbf{\bibinfo{volume}{62}}(\bibinfo{number}{24}), \bibinfo{pages}{17103}
757: % (\bibinfo{year}{2000}).
758:
759: %\bibitem[{\citenamefont{{ S. Debald, T. Brandes, B. Kramer}}(2002)}]{DBK02}
760: %\bibinfo{author}{\bibnamefont{{ S. Debald, T. Brandes, B. Kramer}}},
761: % \bibinfo{journal}{Phys. Rev. B (Rapid Comm.)} \textbf{\bibinfo{volume}{66}},
762: % \bibinfo{pages}{041301(R)} (\bibinfo{year}{2002}).
763:
764: %\bibitem[{\citenamefont{{H. Park, J. Park, A. K. L. Lim, E. H. Anderson, A. P.
765: % Alivisatos, and P. L. McEuen}}(2000)}]{Paretal00}
766: %\bibinfo{author}{\bibnamefont{{H. Park {\em et al}
767: %}}}, \bibinfo{journal}{Nature}
768: % \textbf{\bibinfo{volume}{407}}, \bibinfo{pages}{57} (\bibinfo{year}{2000}).
769: % OPTIONAL END
770:
771: \bibitem{Hayetal03}
772: T. Hayashi {\em et al.}, Phys. Rev. Lett., {\bf 91}, 226804 (2003).
773:
774:
775: \bibitem[{\citenamefont{{T.~Fujisawa, T.~H.~Oosterkamp, W.~G.~van~der~Wiel,
776: B.~W.~Broer, R.~Aguado, S.~Tarucha, and
777: L.~P.~Kouwenhoven}}(1998)}]{Fujetal98}
778: \bibinfo{author}{\bibnamefont{{T.~Fujisawa {\em et al}
779: }}}, \bibinfo{journal}{Science}
780: \textbf{\bibinfo{volume}{282}}, \bibinfo{pages}{932} (\bibinfo{year}{1998}).
781:
782: \bibitem[{\citenamefont{{Yu. Makhlin, G. Sch\"on, and A.
783: Shnirman}}(2001)}]{MSS01}
784: \bibinfo{author}{\bibnamefont{{Yu. Makhlin {\em et al} }}},
785: \bibinfo{journal}{Rev. Mod. Phys.} \textbf{\bibinfo{volume}{73}},
786: \bibinfo{pages}{357} (\bibinfo{year}{2001}).
787:
788: \bibitem[{\citenamefont{{A. A. Clerk, S. M. Girvin, A. K. Nguyen, and A. D.
789: Stone}}(2002)}]{CGNS02}
790: \bibinfo{author}{\bibnamefont{{ A. A. Clerk {\em et al}
791: }}}, \bibinfo{journal}{Phys. Rev. Lett.}
792: \textbf{\bibinfo{volume}{89}}, \bibinfo{pages}{176804}
793: (\bibinfo{year}{2002}).
794:
795:
796: \bibitem[{\citenamefont{Choi et~al.}(2003)\citenamefont{Choi, Plastina, and
797: Fazio}}]{Choi03}
798: \bibinfo{author}{\bibfnamefont{{M.-S. Choi {\em et al}}}},
799: % \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Plastina}}, \bibnamefont{and}
800: % \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Fazio}},
801: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{67}},
802: \bibinfo{pages}{045105} (\bibinfo{year}{2003}).
803:
804: \bibitem[{\citenamefont{Brandes and Kramer}(1999)}]{BK99}
805: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Brandes}} \bibnamefont{and}
806: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Kramer}},
807: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{83}},
808: \bibinfo{pages}{3021} (\bibinfo{year}{1999}).
809:
810: \bibitem{BV02}
811: T. Brandes and T. Vorrath, Phys. Rev. B {\bf 66}, 075341 (2002).
812:
813: %% OPTIONAL
814: %\bibitem{footnote1}
815: %This is even the case for $g_{Q} \propto \delta_{Q,Q_0}$ (one mode Rabi Hamiltonian), cf.
816: %T. Brandes, N. Lambert, Phys. Rev. B {\bf 67}, 125323 (2003).
817: %% OPTIONAL END
818:
819:
820: \bibitem{footnote2}
821: For the SB problem,
822: POL is equivalent to the double-path integral non-interacting
823: blip approximation \cite{Legetal87,Weiss} and works well for $\varepsilon=0$
824: and $\varepsilon\gg T_c$.
825:
826:
827: \bibitem[{\citenamefont{{T. H. Stoof and Yu. V. Nazarov}}(1996)}]{SN96}
828: \bibinfo{author}{\bibnamefont{{T. H. Stoof and Yu. V. Nazarov}}},
829: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{53}},
830: \bibinfo{pages}{1050} (\bibinfo{year}{1996}).
831:
832:
833: \bibitem{Mozetal02}
834: D. Mozyrsky {\em et al.}, Phys. Rev. B {\bf 66}, 161313 (2002).
835:
836:
837: \bibitem[{\citenamefont{{See, {\bf e.g.}, H. J.
838: Carmichael}}(1993)}]{Carmichael}
839: \bibinfo{author}{\bibnamefont{{See, {\bf e.g.}, H. J. Carmichael}}},
840: \emph{\bibinfo{title}{An Open System Approach to Quantum Optics}}, %vol.
841: % \bibinfo{volume}{m 18} of \emph{\bibinfo{series}{Lecture Notes in Physics}}
842: (\bibinfo{publisher}{Springer-Verlag}, \bibinfo{address}{Berlin}
843: %\bibinfo{address}{Berlin, Heidelberg},
844: \bibinfo{year}{1993}).
845:
846: \bibitem[{\citenamefont{Elattari and Gurvitz}(2002)}]{EG02}
847: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Elattari}} \bibnamefont{and}
848: \bibinfo{author}{\bibfnamefont{S.~A.} \bibnamefont{Gurvitz}},
849: \bibinfo{journal}{Physics Letters A} \textbf{\bibinfo{volume}{292}},
850: \bibinfo{pages}{289} (\bibinfo{year}{2002}).
851:
852:
853: \bibitem[{\citenamefont{{Ruskov and Korotkov}}(2003)}]{RK03}
854: \bibinfo{author}{\bibnamefont{{R. Ruskov and A. N. Korotkov}}},
855: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{67}},
856: \bibinfo{pages}{075303} (\bibinfo{year}{2003}).
857:
858: \bibitem[{\citenamefont{{D. K. C. MacDonald}}(1948)}]{Mac48}
859: \bibinfo{author}{\bibnamefont{{D. K. C. MacDonald}}}, \bibinfo{journal}{Rep.
860: Progr. Phys.} \textbf{\bibinfo{volume}{12}}, \bibinfo{pages}{56}
861: (\bibinfo{year}{1948}).
862:
863:
864: \bibitem{US93}
865: M. Ueda and A. Shimizu, J. Phys. Soc. Jap. {\bf 62}, 2994 (1993).
866:
867:
868: \bibitem[{\citenamefont{{P. Cedraschi and M. B\"uttiker}}(2001)}]{CB01}
869: \bibinfo{author}{\bibnamefont{{P. Cedraschi and M. B\"uttiker}}},
870: \bibinfo{journal}{Annals of Physics} \textbf{\bibinfo{volume}{289}},
871: \bibinfo{pages}{1} (\bibinfo{year}{2001}).
872:
873:
874: %\bibitem{Leonoise}
875: %A first step towards these challenging high frequency noise experiments has been
876: %demonstrated recently. Leo Kouwenhoven, private communication.
877: \bibitem[{\citenamefont{{R.~Deblock, E. Onac, L. Gurevich and L.~P.~Kouwenhoven}}(1998)}]{Leonoise}
878: \bibinfo{author}{\bibnamefont{{R.~Deblock {\em et al}
879: }}}, \bibinfo{journal}{Science}
880: \textbf{\bibinfo{volume}{301}}, \bibinfo{pages}{203} (\bibinfo{year}{2003}).
881:
882:
883: \end{thebibliography}
884:
885:
886: \end{document}
887: