1: \documentclass[pra,aps,twocolumn]{revtex4}
2: % \documentclass[12pt]{iopart}
3: \usepackage{amsmath}
4: \usepackage{epsf}
5:
6: \newcommand\fra[2]{{\textstyle{\frac{#1}{#2}}}}
7: \def \mbf#1{{\mbox{\boldmath$#1$}}} % bold greek
8: \renewcommand{\thesection}{\arabic{section}}
9: \def \new{$\clubsuit$~}
10: % \usepackage[notcite]{showkeys}
11:
12: \begin{document}
13:
14: \title{Two-mode optical state truncation and generation of
15: maximally entangled states in pumped nonlinear couplers}
16:
17: \author{Adam Miranowicz and Wies\l{}aw Leo\'nski}
18: \affiliation{Nonlinear Optics Division, Institute of Physics, Adam
19: Mickiewicz University, 61-614 Pozna\'n, Poland}
20:
21: \date{\today}
22:
23: \begin{abstract}
24: Schemes for optical-state truncation of two cavity modes are
25: analyzed. The systems, referred to as the nonlinear quantum
26: scissors devices, comprise two coupled nonlinear oscillators (Kerr
27: nonlinear coupler) with one or two of them pumped by external
28: classical fields. It is shown that the quantum evolution of the
29: pumped couplers can be closed in a two-qubit Hilbert space spanned
30: by vacuum and single-photon states only. Thus, the pumped couplers
31: can behave as a two-qubit system. Analysis of time evolution of
32: the quantum entanglement shows that Bell states can be generated.
33: A possible implementation of the couplers is suggested in a pumped
34: double-ring cavity with resonantly enhanced Kerr nonlinearities in
35: an electromagnetically-induced transparency scheme. The fragility
36: of the generated states and their entanglement due to the standard
37: dissipation and phase damping are discussed by numerically solving
38: two types of master equations.
39: \end{abstract}
40:
41: \maketitle
42:
43: \pagenumbering{arabic}
44:
45: \section{Introduction}
46:
47: Methods for preparation and manipulation of nonclassical states of
48: light have become an important research area in quantum optics
49: \cite{state}, especially in relation to possible optical
50: implementations of quantum computers and systems for quantum
51: communication and quantum cryptography \cite{Nie00}. Among various
52: schemes for optical-qubit generation, the so-called {\em quantum
53: scissors} device of Pegg {\em et al.} \cite{Peg98} produces a
54: superposition of vacuum and single-photon states,
55: $c_0|0\rangle+c_1|1\rangle$, by optical-state truncation of an
56: input single-mode coherent light. The Pegg {\em et al.} quantum
57: scissors device was studied in numerous papers (see, e.g.,
58: \cite{Vil99,Kon00,Par00,Ozd01,Villas01,Mir04,Mir05}), and tested
59: experimentally by Babichev {\em et al.} \cite{Bab02} and Resch
60: {\em et al.} \cite{Res02}. This simple scheme and its
61: generalizations for truncation of an input optical state to a
62: superposition of $d$ Fock states (the so-called qudits)
63: \cite{Villas01,Mir05} are based on linear optical elements, and
64: thus referred to as the {\em linear quantum scissors} devices.
65: Optical-state `truncation' can also be achieved in systems
66: comprising nonlinear elements (e.g., Kerr media)
67: \cite{Leo97,Dar00}, and thus will be referred to as the {\em
68: nonlinear quantum scissors} devices. The above-mentioned schemes
69: are restricted to the single-mode optical truncation. Here, by
70: generalizing our former scheme \cite{Leo04}, we present a
71: realization of nonlinear quantum scissors for optical-state
72: truncation of two cavity modes by means of a pumped nonlinear
73: coupler.
74:
75: Two-mode nonlinear couplers have become, shortly after their
76: introduction by Jensen \cite{Jen82} and Maier \cite{Mai82}, one of
77: the important topics of photonics due to their wide potential
78: applications and relative simplicity (see, e.g., reviews
79: \cite{Sny91,Per00}). Among various types of the nonlinear optical
80: couplers, those based on Kerr effect have attracted especial
81: interest both in classical \cite{Jen82,Mai82,Sny91,Gry01} and
82: quantum \cite{Che96,Hor89,Kor96,Fiu99,Ibr00,Ari00,San03,ElOrany05}
83: regimes. The Kerr nonlinear couplers can exhibit variations of
84: self-trapping, self-modulation and self-switching effects. In
85: quantum regime, they can also generate sub-Poissonian and squeezed
86: light \cite{Hor89,Kor96,Fiu99,Ibr00,Ari00}. Possibilities of
87: entanglement generation were also studied in nonlinear couplers
88: operating by means of the Kerr effect \cite{San03,ElOrany05} or
89: degenerate parametric down-conversion \cite{Her03}.
90:
91: Here, we analyze Kerr nonlinear couplers, which can be modelled by
92: systems composed of two quantum nonlinear oscillators linearly
93: coupled to each other and placed inside a double-ring cavity. We
94: discuss two schemes based on the coupler with an external
95: excitation of a single mode and the coupler with two modes pumped.
96: We show that the states generated in the excited nonlinear
97: couplers under suitable conditions can be limited to a
98: superposition of only vacuum and single-photon states,
99: $c_{00}|00\rangle+c_{01}|01\rangle+c_{10}|10\rangle+c_{11}|11\rangle$.
100: We compare the possibilities of generation of maximally entangled
101: states by the couplers excited in single and two modes. We also
102: discuss effects of dissipation on the fidelity of truncation and
103: suggest a method to achieve strong Kerr interactions at low
104: intensities in our system.
105:
106: %----------------------------------------------------------
107: \begin{figure} % figure 1
108: \epsfxsize=6cm\centerline{\epsfbox{fig01.eps}} \caption{A
109: realization of the two-mode nonlinear quantum scissors device via
110: a pumped nonlinear coupler in a double-ring cavity. Symbols are
111: explained in the text.}
112: \end{figure}
113:
114: \section{Coupler pumped in a single mode}
115:
116: We consider a system, referred to as the pumped Kerr nonlinear
117: coupler, which consists of two nonlinear oscillators, with Kerr
118: nonlinearities $\chi_a$ and $\chi_b$, linearly coupled to each
119: other and additionally linearly coupled to an external classical
120: field. In this section, we assume that the field is coupled to one
121: of the oscillators only. Thus, the system can be described by the
122: following Hamiltonian \cite{Leo04} ($\hbar=1$):
123: \begin{eqnarray}
124: \hat{H}&=&\hat{H}_{0}+\hat{H}_1,
125: \nonumber \\
126: \hat{H}_0&=&\omega_a\hat{a}^\dagger\hat{a}+\omega_b\hat{b}^\dagger\hat{b},
127: \nonumber \\
128: \hat{H}_1&=&\hat{H}^{(a)}_{\rm nonl} + \hat{H}^{(b)}_{\rm
129: nonl}+\hat{H}_{\rm int}+\hat{H}^{(a)}_{\rm ext}, \label{N01}
130: \end{eqnarray}
131: where
132: \begin{eqnarray}
133: \hat{H}^{(a)}_{\rm nonl}+\hat{H}^{(b)}_{\rm nonl} &=&
134: \frac{\chi_a}{2}(\hat{a}^\dagger )^2\hat{a}^2 +
135: \frac{\chi_b}{2}(\hat{b}^\dagger )^2\hat{b}^2, \label{N02}
136: \end{eqnarray}
137: \begin{eqnarray}
138: \hat{H}_{\rm int}&=&\epsilon\hat{a}^\dagger\hat{b} +\epsilon^*
139: \hat{a}\hat{b}^\dagger,
140: \label{N03} \\
141: \hat{H}^{(a)}_{\rm
142: ext}&=&\alpha\hat{a}^\dagger+\alpha^*\hat{a}, \label{N04}
143: \end{eqnarray}
144: and $\hat{a}$ ($\hat{b}$) is the bosonic annihilation operator
145: corresponding to the mode $a$ ($b$) of frequency $\omega_a$
146: ($\omega_b$). Hamiltonian (\ref{N01}) for $\hat{H}^{(a)}_{\rm
147: ext}=0$ describes the standard nonlinear coupler
148: \cite{Che96,Hor89,Kor96,Fiu99,Ibr00,Ari00,San03,ElOrany05}
149: composed of two Kerr nonlinear oscillators linearly coupled to
150: each other, where the parameter $\epsilon$ is the strength of this
151: coupling. However, our model differs from the standard by
152: inclusion of the extra term $\hat{H}^{(a)}_{\rm ext}$, which
153: describes linear coupling between the driving single-mode
154: classical field (say, with frequency $\omega_{\rm ext}^{(a)}$) and
155: the cavity mode $a$. The parameter $\alpha$ describes the strength
156: of this coupling and is proportional to the classical field
157: amplitude. A possible physical realization of the model is
158: presented in figure 1, where an external pump, described by
159: $\hat{H}^{(b)}_{\rm ext}$, is off in the present analysis. Kerr
160: media, marked by $\hat{H}^{(a)}_{\rm nonl}$ and
161: $\hat{H}^{(b)}_{\rm nonl}$, are linearly coupled as described by
162: grey central region corresponding to $\hat{H}_{\rm int}$.
163:
164: The evolution of our system in the interaction picture can be
165: described by the Schr\"odinger equation
166: \begin{equation}
167: i\frac{d }{d t}|\psi (t)\rangle =\hat{H}_{1}|\psi (t)\rangle,
168: \label{N05}
169: \end{equation}
170: where
171: \begin{equation}
172: |\psi (t)\rangle =\sum_{m,n=0}^\infty c_{mn}(t)|mn\rangle .
173: \label{N06}
174: \end{equation}
175: Thus, the complex probability amplitudes $c_{mn}(t)$ satisfy the
176: set of equations of motion:
177: \begin{eqnarray}
178: && i\frac{d }{d t}c_{mn} = \left[ \chi _{a}m(m-1)+\chi
179: _{b}n(n-1)\right]c_{mn} \notag \\ && + \epsilon c_{m-1,n+1}
180: \sqrt{m(n+1)} + \epsilon ^{\ast } c_{m+1,n-1} \sqrt{(m+1)n} \notag
181: \\ && + \alpha c_{m-1,n} \sqrt{m} + \alpha ^{\ast } c_{m+1,n}
182: \sqrt{m+1}. \label{N07}
183: \end{eqnarray}
184: A superficial analysis of (\ref{N07}) could lead to a conclusion
185: that the evolution of the system pumped by classical external
186: field cannot be restricted to two lowest photon-numer states, but
187: will also include states with a greater number of photons.
188: However, by generalizing the method of single-mode nonlinear
189: quantum scissors proposed in \cite{Leo97} (for a review see
190: \cite{MLI01,LM01}), we have observed in \cite{Leo04} that
191: evolution can be restricted to the only four states: $|00\rangle$,
192: $|10\rangle$, $|01\rangle$ and $|11\rangle$ as a result of
193: degeneracy of Hamiltonians $\hat{H}^{(a)}_{\rm nonl}$ and
194: $\hat{H}^{(b)}_{\rm nonl}$. By assuming that the couplings
195: $|\alpha|$, and $|\epsilon|$ are much smaller than the Kerr
196: nonlinearities $\chi_a$ and $\chi_b$, we can interpret the
197: evolution between the four states as resonant transitions, while
198: the negligible evolution to other states as out of resonance,
199: analogously to the single-mode case \cite{LM01}. This phenomenon
200: can be shown explicitly as follows: under the assumption of $\chi
201: _{a},\chi _{b}\gg \max (|\epsilon |,|\alpha |)$ and short
202: evolution times, equation (\ref{N07}) for $m,n\neq 0,1$ can be
203: approximated by
204: \begin{equation}
205: i\frac{d }{d t}c_{mn}\approx \left[ \chi _{a}m(m-1)+\chi
206: _{b}n(n-1)\right] c_{mn} \label{N08}
207: \end{equation}
208: which has the simple solution
209: \begin{equation}
210: c_{mn}(t)\approx \exp \{-i\left[ \chi _{a}m(m-1)+\chi
211: _{b}n(n-1)\right] t\}c_{mn}(0). \label{N09}
212: \end{equation}
213: By setting the initial condition $c_{mn}(0)=0$ for $m,n\neq 0,1$,
214: one gets $c_{mn}(t)\approx 0$. By contrast, for $m,n\in \{0, 1\}$,
215: the terms proportional to $\chi_a$ and $\chi_b$ are vanishing due
216: to degeneracy of the Kerr Hamiltonian and so the remaining terms
217: proportional to $\epsilon$ and $\alpha$ are significant. Thus, the
218: ideally `truncated' two-mode state generated in the system has the
219: following simple form
220: \begin{equation}
221: |\psi (t)\rangle_{\rm cut} =c_{00}(t)| 0 0\rangle +c_{01}(t)| 0
222: 1\rangle +c_{10}(t)| 1 0\rangle+c_{11}(t)| 1 1\rangle, \label{N10}
223: \end{equation}
224: where the evolution of $c_{mn}$, precisely given by (\ref{N07}),
225: can approximately be described by the following equations
226: \begin{eqnarray}
227: i\frac{d}{dt}c_{00}&=&\alpha^*c_{10},\quad
228: i{\frac{d}{dt}}c_{01}\;=\;\epsilon^*c_{10}+\alpha^*c_{11},\nonumber\\
229: i\frac{d}{dt}c_{11}&=&\alpha\, c_{01},\quad\;
230: i\frac{d}{dt}c_{10}\;=\;\epsilon\, c_{01}+\alpha\,
231: c_{00}.\label{N11}
232: \end{eqnarray}
233: Hereafter, in equations for the probability amplitudes under the
234: discussed assumptions, sign `=' should be understood as `$\approx
235: $'. Although approximate equations (\ref{N11}) are independent of
236: $\chi_a$, our derivation clearly shows that the Kerr nonlinearity
237: plays a crucial role in the truncation process. By assuming that
238: both oscillators are initially in vacuum states, $|\psi
239: (t=0)\rangle =| 0 0\rangle$, and parameters $\alpha$ and
240: $\epsilon$ are real, we find the following solutions of
241: (\ref{N11}) for the time-dependent probability amplitudes:
242: \begin{eqnarray}
243: c_{00}&=&\frac{1}{2\gamma}\left[(\gamma-\epsilon)\cos\tau
244: _{1}+(\gamma+\epsilon)\cos\tau _{2}\right],
245: \notag \\
246: c_{01}&=&\frac{\alpha}{\gamma}\left(\cos\tau _{1}-\cos\tau
247: _{2}\right),
248: \notag \\
249: c_{10}&=&-\frac{i(\gamma+\epsilon)\Omega_2}{4 \alpha
250: \gamma}(\sin\tau _{1}+\sin\tau _{2}),
251: \notag \\
252: c_{11}&=&-\frac{i}{2\gamma}\left(\Omega_2\sin\tau
253: _{1}-\Omega_1\sin\tau _{2}\right), \label{N12}
254: \end{eqnarray}
255: where
256: $\Omega_{j}=\sqrt{2[2\alpha^2+\epsilon^2+(-1)^{j-1}\epsilon\gamma]}$,
257: $\gamma=\sqrt{4\alpha ^{2}+\epsilon ^{2}}$, and $\tau _{j}=\Omega
258: _{j}t/2$ for $j=1,2$. Note that the solution for $c_{10}$ can be
259: written in a more symmetric form since the properties hold:
260: $(\gamma+\epsilon)\Omega_2=(\gamma-\epsilon)\Omega_1=2
261: \sqrt{\alpha^2(\gamma^2-\epsilon^2)}$. In a special case of equal
262: couplings $\alpha$ and $\epsilon$, (\ref{N12}) simplifies to our
263: former solution \cite{Leo04}:
264: \begin{eqnarray}
265: c_{00} &=&\cos (\sqrt{5}\tau)\cos (\tau)+\frac{1}{\sqrt{5}}\sin
266: (\sqrt{5}\tau)\sin (\tau), \notag
267: \\
268: c_{01} &=&-\frac{2}{\sqrt{5}}\sin (\sqrt{5}\tau)\sin (\tau), \notag \\
269: c_{10} &=&-i\frac{2}{\sqrt{5}}\sin (\sqrt{5}\tau)\cos (\tau), \label{N13} \\
270: c_{11} &=&-i\cos (\sqrt{5}\tau)\sin (\tau)+\frac{i}{\sqrt{5}}\sin
271: (\sqrt{5}\tau)\cos (\tau), \notag
272: \end{eqnarray}
273: where $\tau=\alpha t /2$. To estimate the quality of the
274: optical-state truncation (up to single-photon states) of the
275: generated light, we apply fidelity as a measure of discrepancy
276: between the ideally truncated two-qubit state $\hat{\rho}_{\rm
277: cut}=|\psi (t)\rangle_{\rm cut}\,_{\rm cut}\langle \psi (t)|$,
278: given by (\ref{N10}), and the actually generated output state
279: $\hat{\rho}(t)=|\psi(t)\rangle\langle\psi(t)|$ calculated
280: numerically from
281: \begin{eqnarray}
282: |\psi(t)\rangle=\exp(-i\hat{H}t)|00\rangle
283: \label{N14}
284: \end{eqnarray}
285: for a large (practically infinite-dimensional) two-mode Hilbert
286: space. Specifically, in our numerical analysis, we have chosen the
287: dimension equal to $20$ for each subspace associated with single
288: mode of the field. The fidelity, also referred to as the Uhlmann's
289: transition probability for mixed states, is defined by (see, e.g.,
290: \cite{Ved02}):
291: %----------------------------------------------------------
292: \begin{figure} % figure 2
293: \epsfxsize=8cm\centerline{\epsfbox{fig02a.eps}}
294: \epsfxsize=8cm\centerline{\epsfbox{fig02b.eps}}%
295: \caption{Fidelity between the actually generated state
296: $|\psi(t)\rangle$ and the ideally truncated state $|\psi_{\rm cut
297: } (t)\rangle$ if the coupling strengths are (a) $\epsilon=\alpha$
298: and (b) $\epsilon=\alpha/10$ for the system pumped in one mode
299: with $\beta=0$ (dashed curves), and that pumped in two modes with:
300: $\beta=\alpha$ (solid), $\beta=-\alpha$ (dotted), $\beta=i\alpha$
301: (dot-dashed curves). In figures 2--7, we assume that the
302: nonlinearity coefficients are $\chi_a=\chi_b=10^8$ rad/sec,
303: $\alpha=\chi_a/200$, and the coupler is initially in two-mode
304: vacuum state $|00\rangle$.}
305: \end{figure}
306: % %----------------------------------------------------------
307: \begin{figure} % figure 3
308: \centerline{ \epsfxsize=4.4cm\epsfbox{fig03a.eps}
309: \epsfxsize=4.4cm\epsfbox{fig03b.eps}} %
310:
311: \centerline{ \epsfxsize=4.4cm\epsfbox{fig03c.eps}
312: \epsfxsize=4.4cm\epsfbox{fig03d.eps}} %
313: \caption{Evolution of the entropy of entanglement $E$ of the
314: generated states $|\psi(t)\rangle$ (dots) and the desired
315: truncated states $|\psi_{\rm cut }(t)\rangle$ (solid curves) by
316: the coupler pumped in a single mode with (a) $\beta=0$ and in
317: two modes with: (b) $\beta=\alpha$, (c) $\beta=-\alpha$, (d)
318: $\beta=i\alpha$, where $\alpha=\epsilon$.}
319: \end{figure}
320: %----------------------------------------------------------
321: \begin{figure} % figure 4
322: \centerline{\epsfxsize=4.4cm\epsfbox{fig04a.eps}
323: \epsfxsize=4.4cm\epsfbox{fig04b.eps}} %
324: \caption{Evolution of the entropy of entanglement $E$ for the
325: coupler pumped in (a) single mode ($\beta=0$) and (b) two
326: modes ($\beta=\alpha$) using line styles and parameters same as
327: in figure 3 except $\epsilon=\alpha/10$. On this figure scale,
328: evolutions of $E$ for $\beta=-\alpha$ can be described by the same
329: curve as in figure (b), while the entropy of entanglement for
330: $\beta=i\alpha$ is negligible as less than $5\times 10^{-4}$.}
331: \end{figure}
332: \begin{eqnarray}
333: F(\hat{\rho},\hat{\rho}_{\rm cut}) &=& \left\{ {\rm
334: Tr}[(\sqrt{\hat{\rho}_{\rm cut}}\hat{\rho} \sqrt{\hat{\rho}_{\rm
335: cut}})^{1/2}]\right\}^2.
336: \label{N15}
337: \end{eqnarray}
338: By assuming that one of the states is pure (say $| \psi
339: \rangle_{\rm cut}$), then (\ref{N15}) simplifies to $F=\,_{\rm
340: cut}\langle \psi | \hat{\rho} | \psi \rangle_{\rm cut}$. Instead
341: fidelity, the Bures distance $D_B(\hat{\rho}|| \hat{\rho}_{\rm
342: cut}) = 2-2 \sqrt{F(\hat{\rho},\hat{\rho}_{\rm cut})}$ is often
343: applied as a measure of discrepancy between the states. Note that
344: the Bures distance satisfies the usual metric properties including
345: symmetry $D_B(\hat{\rho}||\hat{\rho}_{\rm
346: cut})=D_B(\hat{\rho}_{\rm cut}||\hat{\rho})$, contrary to the
347: quantum Kullback-Leibler `distance' (quantum relative entropy)
348: often used in quantum-information context \cite{Ved02}. General
349: expression (\ref{N15}) will be applied in our description of the
350: effect of damping on the optical-state truncation in section 4. In
351: this section, we are focused on pure-state truncation, for which
352: equation (\ref{N15}) simplifies to the familiar expression
353: $F=||\langle \psi (t)|\psi (t)\rangle_{\rm cut} ||^2$, where the
354: ideally truncated state $|\psi (t)\rangle_{\rm cut}$ is given by
355: (\ref{N10}) and the actually generated state $|\psi (t)\rangle$ is
356: calculated numerically from (\ref{N14}). The fidelity for perfect
357: truncation is equal to one. Figure 2 clearly indicates that the
358: fidelity of truncation using the pumped coupler is close to one
359: for short times and the coupling strengths much smaller than the
360: nonlinearity parameters ($|\alpha|,|\epsilon|\ll \chi_{a,b}$). The
361: numerical results shown in figure 2 confirm the validity of our
362: analytical approach at least for short evolution times. Thus, we
363: can refer to the system as a kind of nonlinear (as operating by
364: means of Kerr nonlinearity) quantum scissors device.
365:
366: Solutions for probability amplitudes of the truncated states
367: enable a simple calculation of quantum entanglement, which is one
368: of the most fundamental resources of quantum information theory
369: \cite{Nie00}. It is well known that the entanglement of a
370: bipartite pure state, described by a density matrix $\hat \rho=
371: |\psi\rangle\langle\psi|$, can be described by the von Neumann
372: entropy of either the reduced density matrix $\hat \rho_a={\rm
373: Tr}_b \hat \rho$ or $\hat \rho_b={\rm Tr}_a \hat \rho$ or,
374: equivalently, by the Shannon entropy of the squared Schmidt
375: coefficients $p_i$ \cite{Ved02}:
376: \begin{eqnarray}
377: E(\hat \rho)&=&-{\rm Tr}(\rho_a\log_2\rho_a)=-{\rm
378: Tr}(\rho_b\log_2\rho_b)
379: \notag\\
380: &=&-\sum_{i=1}^N p _i\log_2 p_i \equiv h(p_1,...,p_{N-1}).
381: \label{N16}
382: \end{eqnarray}
383: For bipartite pure states this measure is often referred to as the
384: {\em entropy of entanglement}. In a special case of two qubits in
385: a pure state, the entropy of entanglement $E$ ranges from zero for
386: a separable state to 1 ebit for a maximally entangled state, and
387: it is simply given in terms of the binary entropy $h(p)= -p\log
388: _{2}p-(1-p)\log _{2}(1-p)$. In fact, for a general two-qubit pure
389: state, given by (\ref{N10}) with arbitrary amplitudes $c_{mn}$
390: ($m,n=0,1$), the entropy of entanglement given by (\ref{N16}) can
391: simply be calculated as
392: \begin{equation}
393: E(t) \equiv E(|\psi(t)\rangle_{\rm cut})
394: = {\cal E}\left(2|c_{00}(t)c_{11}(t)-c_{01}(t)c_{10}(t)|\right),
395: \label{N17}
396: \end{equation}
397: where
398: \begin{eqnarray}
399: {\cal E}(x) \equiv h\left(\frac{1}{2}(
400: 1+\sqrt{1-x^{2}})\right) \label{N18}
401: \end{eqnarray}
402: and $h$ is the binary entropy. If the probability amplitudes
403: $c_{mn}(t)$ evolve according to (\ref{N12}) then the evolution of
404: the entropy of entanglement is given by
405: \begin{eqnarray}
406: E(t) = {\cal E}\Big( \sum_{j=1,2}\frac{\Omega_j}
407: {\gamma^{2}}\big\{ \epsilon\cos(\fra12 \Omega_{3-j}t)
408: -[\epsilon+(-1)^j\gamma]
409: \nonumber \\
410: \times \cos(\fra12 \Omega_{j}t)\big \}
411: \sin(\fra12 \Omega_{3-j}t)\Big). \label{N19}
412: \end{eqnarray}
413: Solution (\ref{N19}) is further simplified by assuming real
414: $\alpha=\epsilon$ then the amplitudes $c_{mn}$ are given by
415: (\ref{N13}). Thus, we obtain
416: \begin{eqnarray}
417: E (t)= {\cal E}\Big(\frac15 \{ [4+\cos (\sqrt{5}\alpha t)] \sin
418: (\alpha t)
419: \nonumber \\
420: -\sqrt{5}\cos (\alpha t)\sin (\sqrt{5}\alpha t)\}\Big). \label{N20}
421: \end{eqnarray}
422: Figures 3(a) and 4(a) show the plots of the entropy of
423: entanglement for the case discussed here, i.e., for the
424: single-mode pumped coupler with the coupling parameters $|\alpha|$
425: and $|\epsilon|$ much smaller than the nonlinearities
426: $\chi_a=\chi_b$. We see in figure 3(a) that the rapid oscillations
427: in time (with a period $T_1=\pi/|\alpha|$) are modulated by
428: oscillations of low frequency (with period $T_2=8\pi/|\alpha|$).
429: As a consequence, their maxima are of various values but some of
430: them approach 1 ebit corresponding to the formation of Bell
431: states. To show this explicitly, we represent the generated state
432: in the basis $|\psi\rangle=\sum_{j=1}^4 b_j|B_j\rangle$ spanned by
433: the Bell-like states:
434: %----------------------------------------------------------
435: \begin{figure} % figure 5
436: \centerline{\epsfxsize=4.4cm\epsfbox{fig05a.eps}
437: \epsfxsize=4.4cm\epsfbox{fig05b.eps}} %
438: \caption{Probabilities $|b_i|^2$ for finding the coupler pumped in
439: a single mode with $\epsilon=\alpha$ in the Bell states:
440: $|B_{1}\rangle$ and $|B_{3}\rangle$ (solid curves) as well as
441: $|B_{2}\rangle$ and $|B_{4}\rangle$ (dashed curves).}
442: \end{figure}
443: %----------------------------------------------------------
444: \begin{figure} % figure 6
445: \centerline{\epsfxsize=4.4cm\epsfbox{fig06a.eps}
446: \epsfxsize=4.4cm\epsfbox{fig06b.eps}} %
447: \caption{Probabilities for finding the single-mode pumped
448: coupler with $\epsilon=\alpha/10$ in the Bell states. Same symbols
449: as in figure 5.}
450: \end{figure}
451: %----------------------------------------------------------
452: \begin{figure} % figure 7
453: \centerline{\epsfxsize=4.4cm\epsfbox{fig07a.eps}
454: \epsfxsize=4.4cm\epsfbox{fig07b.eps}} %
455: \caption{Probabilities for finding the coupler pumped in two
456: modes with $\alpha=\beta$ and $\epsilon=\alpha/10$ in the Bell
457: states. Same symbols as in figure 5.}
458: \end{figure}
459: \begin{eqnarray}
460: |B_1\rangle&=&\frac{|11\rangle+i|00\rangle}{\sqrt{2}},\quad
461: |B_2\rangle\,=\,\frac{|00\rangle+i|11\rangle}{\sqrt{2}},\nonumber\\
462: |B_3\rangle&=&\frac{|01\rangle-i|10\rangle}{\sqrt{2}},\quad
463: |B_4\rangle\,=\,\frac{|10\rangle-i|01\rangle}{\sqrt{2}},
464: \label{N21}
465: \end{eqnarray}
466: which differ from the standard Bell states only by the phase
467: factor $i$. Clearly, $\sum_j |B_j\rangle\langle B_j|=1$ and
468: $E(|B_j\rangle)=1$ for $i=1,...,4$, so states (\ref{N21}) will
469: shortly be referred to as the Bell states. Figures 5 and 6 show
470: the probabilities for the generation of the Bell states as a
471: function of time for the single-mode driven couplers, when the
472: initial state is the two-mode vacuum state. It is seen that states
473: $|B_1\rangle$ and $|B_2\rangle$, being superpositions of
474: $|00\rangle$ and $|11\rangle$, are generated with a high accuracy.
475: In detail, the maxima of (\ref{N20}) occur approximately at times
476: $t(m,n) = \frac12 [(2m-1)T_1+(2n-1)T_2]$ if $\alpha=\epsilon$.
477: Since the frequencies of oscillations in (\ref{N20}) are
478: incommensurate, by waiting long enough (assuming no dissipation),
479: we can achieve 1 ebit with high precision. For example, in the
480: first four periods of $T_2$, one could observe generation of
481: entangled states approaching Bell state $|B_1\rangle$ with
482: $E[t(1,1)]=0.994$ and $E[t(1,2)]=0.997$ ebits, and approaching
483: $|B_2\rangle$ with $E[t(2,3)]=0.992$ and $E[t(2,4)]=0.998$. Note
484: that only the first period is shown in figures 3(a) and 5(a), for
485: which the highest entanglement of 0.994 ebits occurs approximately
486: at the time $t(1,1)=4.5 \times 10^{-6}$. Thus, we can effectively
487: treat our system as a source of maximally entangled states. On the
488: other hand, the Bell states $|B_3\rangle$ and $|B_4\rangle$, being
489: superpositions of $|10\rangle$ and $|01\rangle$, are not generated
490: if the initial state is vacuum and $\epsilon=\alpha$. This
491: conclusion can be drawn by observing that the probabilities for
492: $|B_3\rangle$ and $|B_4\rangle$ can reach only $0.8$ in figure
493: 5(b). However, by relaxing the condition of equal couplings
494: $\epsilon$ and $\alpha$, as shown in figure 6(b), Bell states
495: $|B_3\rangle$ and $|B_4\rangle$ can be generated from vacuum with
496: high precision in the dissipation-free system.
497:
498: Hitherto, we have assumed that both cavity modes were initially in
499: vacuum states. Now, we analyze a more general evolution when the
500: cavity modes are initially not only in vacuum but also in
501: single-photon Fock states, i.e., $|\psi (0)\rangle \equiv|\psi
502: ^{(kl)}(0)\rangle =|kl\rangle$, where $k,l=0,1$. Thus, by assuming
503: as usual that $|\alpha|= |\epsilon|\ll\chi_a,\chi_b$, the
504: evolutions of the initial states are found to be
505: \begin{eqnarray}
506: |\psi ^{(01)}(\tau )\rangle_{\rm cut} &=&c_{01}|00\rangle
507: +\widetilde{c}_{00}|01\rangle +
508: \widetilde{c}_{11}|10\rangle +c_{10}|11\rangle , \notag \\
509: |\psi ^{(10)}(\tau )\rangle_{\rm cut} &=&c_{10}|00\rangle
510: +\widetilde{c}_{11}|01\rangle +
511: \widetilde{c}_{00}|10\rangle +c_{01}|11\rangle , \notag \\
512: |\psi ^{(11)}(\tau )\rangle_{\rm cut} &=&c_{11}|00\rangle
513: +c_{10}|01\rangle +c_{01}|10\rangle +c_{00}|11\rangle \notag \\
514: \label{N22}
515: \end{eqnarray}
516: in terms of the time-dependent amplitudes $c_{mn}(\tau )$ given by
517: (\ref{N13}) and
518: \begin{eqnarray}
519: \widetilde{c}_{00}(\tau ) =&\cos (\sqrt{5}\tau )\cos (\tau
520: )-\frac{1}{\sqrt{5}}
521: \sin (\sqrt{5}\tau )\sin (\tau ), \notag \\
522: \widetilde{c}_{11}(\tau ) =&-i\cos (\sqrt{5}\tau )\sin (\tau
523: )-\frac{i}{\sqrt{5}} \sin (\sqrt{5}\tau )\cos (\tau ), \label{N23}
524: \end{eqnarray}
525: where, as usual, $\tau=\alpha t/2$. Please note that
526: $\widetilde{c}_{jj}(\tau)$ and $c_{jj}(\tau)$ differ in sign. We
527: find that the generalized expression for the entropies of
528: entanglement for the initial states $|kl\rangle$ with $k,l=0,1$
529: reads as
530: \begin{eqnarray}
531: E^{(kl)}(t) = {\cal E}\Big(\frac15[4 +\cos (\sqrt{5}\alpha t)]
532: \sin (\alpha t)\hspace{1.5cm}
533: \nonumber \\
534: -(-1)^{k-l}\sqrt{5}\cos (\alpha t)\sin (\sqrt{5}\alpha t)\Big),
535: \label{N24}
536: \end{eqnarray}
537: which implies that
538: \begin{eqnarray}
539: E^{(00)}(t)&=& E^{(11)}(t),
540: \notag \\
541: E^{(01)}(t)&=& E^{(10)}(t). \label{N25}
542: \end{eqnarray}
543: It is worth noting that our system for the initial states
544: $|01\rangle$ or $|10\rangle$ quasi-periodically evolves into the
545: Bell states $|B_3\rangle$ and $|B_4\rangle$ with high precision
546: but does not evolve into $|B_1\rangle$ or $|B_2\rangle$ assuming
547: $\alpha=\epsilon$. This is by contrast to the evolutions of the
548: initial states $|00\rangle$ (see figure 5) or $|11\rangle$ also
549: for $\epsilon=\alpha$. For brevity, we will not present any graphs
550: of the evolutions of the initial states $|01\rangle$,
551: $|10\rangle$, and $|11\rangle$, which would correspond to figures
552: 2--5 plotted for the initial vacua.
553:
554: The above solutions for the initial single-photon states are
555: included for completeness of our mathematical approach to show
556: that, in principle, all Bell states can be generated in our system
557: even for $\epsilon=\alpha$. But it should be stressed that the
558: system with the initial Fock states is much more experimentally
559: challenging than that assuming initially the vacuum states only.
560: Despite of experimental difficulty, our system enables generation
561: of the one-photon Fock states from vacuum assuming no coupling
562: $\epsilon$ between the modes, which corresponds to having two
563: independent pumped cavities with nonlinear Kerr media. A
564: possibility of producing single-photon states in such systems was
565: demonstrated in \cite{Leo97} (for a review see \cite{LM01} and
566: references therein).
567:
568: %------------------------------------------------------------------
569: \section{Coupler pumped in two modes}
570:
571: This section is devoted to the most general scheme presented in
572: figure 1, namely that involving two external excitations. We
573: assume here that both modes of the coupler are excited by external
574: fields, whereas for the case discussed previously we assumed that
575: only one of the modes was coupled to the external field. The
576: Hamiltonian describing such a system is of the form
577: \begin{equation}
578: \hat{H}_1=\hat{H}^{(a)}_{\rm nonl}+\hat{H}^{(b)}_{\rm
579: nonl}+\hat{H}_{\rm int} +\hat{H}^{(a)}_{\rm
580: ext}+\hat{H}^{(b)}_{\rm ext}, \label{N26}
581: \end{equation}
582: which is the same as that defined by (\ref{N01})--(\ref{N04}),
583: except for the extra term given by
584: \begin{eqnarray}
585: \hat{H}^{(b)}_{\rm ext}&=&\beta\hat{b}^\dagger+\beta^*\hat{b}
586: \label{N27}
587: \end{eqnarray}
588: corresponding to the coupling of the cavity mode $b$ with an
589: external driving single-mode classical field (say, with frequency
590: $\omega_{\rm ext}^{(b)}$), where the parameter $\beta$ describes
591: the strength of this interaction being proportional to the
592: classical field amplitude. The evolution of our system, described
593: by Hamiltonian (\ref{N26}), can be given by the Schr\"odinger
594: equation from which we find the following set of equations for the
595: amplitudes $c_{mn}(t)$ of the wavefunction (\ref{N06}) in the
596: interaction picture:
597: \begin{eqnarray}
598: && i\frac{d }{d t}c_{mn} = \left[ \chi _{a}m(m-1)+\chi
599: _{b}n(n-1)\right]c_{mn} \notag \\ && + \epsilon c_{m-1,n+1}
600: \sqrt{m(n+1)} +
601: \epsilon ^{\ast } c_{m+1,n-1} \sqrt{(m+1)n} \notag \\
602: && + \alpha c_{m-1,n} \sqrt{m} + \alpha ^{\ast } c_{m+1,n}
603: \sqrt{m+1}\notag \\ && +\beta c_{m,n-1} \sqrt{n} +\beta ^{\ast }
604: c_{m,n+1} \sqrt{n+1}. \label{N28}
605: \end{eqnarray}
606: Analogously to the analysis in the former section, we assume short
607: evolution times as well as the couplings $|\alpha|$, $|\beta|$,
608: and $|\epsilon|$ to be much smaller than the Kerr nonlinearities
609: $\chi_a$ and $\chi_b$. Then, equation (\ref{N28}) for $m,n\neq
610: 0,1$ can be approximated by (\ref{N08}) with the solution
611: (\ref{N09}), which vanishes for the initial condition
612: $c_{mn}(0)=0$. Thus, under the above assumptions, the infinite set
613: of equations (\ref{N28}) reduces to the following four equations:
614: \begin{eqnarray}
615: i\frac{dc_{00}}{dt}&=&\alpha^* c_{10}+\beta^*c_{01},\nonumber\\
616: i\frac{dc_{01}}{dt}&=&\epsilon^* c_{10}+\alpha^* c_{11}+\beta c_{00},\nonumber\\
617: i\frac{dc_{10}}{dt}&=&\epsilon c_{01}+\alpha c_{00}+\beta^*c_{11},\nonumber\\
618: i\frac{dc_{11}}{dt}&=&\alpha c_{01}+\beta c_{10}. \label{N29}
619: \end{eqnarray}
620: Assuming that at the time $t=0$, both oscillator modes are in the
621: vacuum states, i.e., $c_{00}=1$ and $c_{01}=c_{10}=c_{11}=0$, we
622: can obtain analytical solutions of (\ref{N29}). To solve
623: (\ref{N29}) we need to find zeros of a fourth-order polynomial
624: and, hence, the solutions in their general form are rather
625: complicated and unreadable. However, if we assume that all
626: coupling constants are real and the couplings with two external
627: fields are of the same strength ($\alpha=\beta$) then the
628: solutions become much simpler and easier to interpret. Thus, under
629: these assumptions, the solutions are found to be:
630: \begin{eqnarray}
631: c_{00} &=&\frac{1}{2}\left\{1+\left[\cos (\frac{\lambda
632: t}{2})+i\frac{\epsilon}{\lambda} \sin (\frac{\lambda
633: t}{2})\right]e^{-i\epsilon t/2}\right\}, \notag \\
634: c_{01} &=&c_{10}=-i2\frac{\alpha }{\lambda }\sin (\frac{\lambda
635: t}{2})\,e^{-i\epsilon t/2}, \notag \\
636: c_{11} &=&c_{00}-1, \label{N30}
637: \end{eqnarray}
638: where we have introduced the effective coupling constant
639: $\lambda=\sqrt{16\alpha^2+\epsilon^2}$. As in the case discussed
640: in the previous section, the system's dynamics is closed within
641: the finite set of $n$-photon states. Figure 2 shows that for the
642: assumed parameters and evolution times shorter than $5\times
643: 10^{-6}$s, the fidelity between the ideally truncated states and
644: the actually generated states by means of the coupler pumped in
645: two modes deviates from 1 by the values less than $0.03$ in figure
646: 2(a) or even $<6 \times 10^{-4}$ in figure 2(b). This again
647: confirms the validity of our analysis and justifies referring to
648: this system as a kind of the quantum scissors device. For the
649: parameters assumed in figure 2, truncation with higher fidelity is
650: usually observed for the coupler pumped in a single mode rather
651: than in two modes. In the latter case, the truncation fidelity
652: depends on the relative phase between the pumping-field couplings
653: $\alpha$ and $\beta$. Note that state $|\psi_{\rm cut}\rangle$ for
654: $\beta=\pm\alpha$, contrary to $\beta=i\alpha$, can be directly
655: calculated from solution (\ref{N12}). For brevity, we have not
656: presented here an analogous solution for $\beta=i\alpha$, but we
657: have used it for plotting the corresponding curves in figures 2
658: and 3(d). It is seen, by comparing figures 2(a) and 2(b), that by
659: decreasing $\epsilon$ in comparison to $\alpha$, the fidelity of
660: truncation can be improved for $\beta\neq 0$. Thus, pumping the
661: coupler in two modes can lead sometimes to truncation better than
662: that for the system driven in single mode as presented in figure
663: 2(b) by dot-dashed curve corresponding to $\beta=i\alpha$ and
664: $\epsilon=\alpha/10$.
665:
666: The evolution of the pure-state entanglement, given by
667: (\ref{N16}), generated in the coupler pumped in two modes can be
668: calculated from (\ref{N17}) with the probability amplitudes given
669: by (\ref{N30}). Thus, we obtain
670: \begin{equation}
671: E(t) = {\cal E}\Big(\frac1{2}\big|1-\frac{e^{-i\epsilon
672: t}}{\lambda^{2}}\big[16 \alpha^2 +\epsilon^2 \cos(\lambda t) +i
673: \epsilon \lambda \sin(\lambda t)\big]\big|\Big).\label{N31}
674: \end{equation}
675: Figures 3(b-d) and 4(b) show the evolution of the entropy of
676: entanglement measured in ebits as a function of time for the
677: system pumped in two modes in comparison to the results for the
678: single-mode driven coupler shown in figures 3(a) and 4(a). It is
679: seen that the first maximum in figures 3(b-d) is the highest,
680: contrary to the case shown in figures 3(a) for the single-mode
681: pumping. Nevertheless, the most important fact is that the value
682: of the entropy of entanglement $E$ can approach unity to a high
683: precision. So, as in the case of the single-mode pumping, the
684: two-mode driven system effectively generates Bell states. To find
685: which Bell states are generated, we can also transform the
686: resulting wavefunction into the Bell basis. Thus, figure 7 depicts
687: probabilities for the four Bell states. As expected from the form
688: of our analytical solutions for $c_{mn}$ ($m,n=0,1$), the
689: entanglement occurs for the states $|00\rangle$ and $|11\rangle$,
690: and leads to the generation of the states $|B_1\rangle$ and
691: $|B_2\rangle$ (with some unimportant global phase factor).
692: Clearly, the highest peaks of the entropy $E(t)$ and those of the
693: probabilities $|b_{1,2}(t)|^2$ of the Bell state generation occur
694: at the same evolution times, as seen by comparing figures: 3(a)
695: with 5(a), 4(a) with 6(a), and 4(b) with 7. Analysis of the
696: evolutions of the probabilities $|b_{1,2}(t)|^2$ and the entropy
697: of entanglement $E(t)$ for relatively longer times reveals that
698: the oscillations are modulated and some long-time oscillations
699: occur in the system, which can be interpreted as a result of
700: quantum beats. It is seen from (\ref{N31}) that two various
701: frequencies appear in our solution and one of them is considerably
702: greater than the other, e.g., the effective coupling constant
703: $\lambda$ is $\sqrt{17}$ times greater than the internal coupling
704: constant of the coupler $\epsilon=\alpha$. We are not presenting
705: here dissipation-free evolutions exhibiting modulated oscillations
706: at times longer than those in figure 4. It would be meaningless
707: since the entanglement is lost at such evolution times due to
708: dissipation, which inevitably occurs in real physical
709: implementations of the coupler as will be discussed in the next
710: section.
711:
712: Although our analysis, including all figures, is focused on
713: evolution of the initial vacuum states, we present shortly some
714: results for other states too. We find that the evolution of the
715: initial Fock state $|\psi ^{(kl)}(0)\rangle =|kl\rangle$ with
716: $k,l=0,1$ is of the form (\ref{N22}) but with the probability
717: amplitudes $c_{mn}(\tau )$ given by (\ref{N30}) and
718: $\widetilde{c}_{mn}(\tau )$ equal to
719: \begin{eqnarray}
720: \widetilde{c}_{00}(\tau ) &=& \frac12 \Big\{e^{i\epsilon
721: t}+e^{-i\epsilon t/2} \Big[ \cos (\frac{\lambda t}{2})
722: -i\frac{\epsilon}{\lambda}
723: \sin (\frac{\lambda t}{2}) \Big] \Big\}, \notag \\
724: \widetilde{c}_{11}(\tau ) &=& \widetilde{c}_{00}(\tau
725: )-e^{i\epsilon t}. \label{N32}
726: \end{eqnarray}
727: With the help of these formulae, we can calculate the entropies of
728: entanglement explicitly as
729: \begin{eqnarray}
730: E^{(kl)}(t) ={\cal E}\Big( \frac1{2}
731: \Big|1-\lambda^{-2}\exp[-i(2|k-l|+1)\epsilon t] \hspace{0.5cm}
732: \nonumber \\
733: +\big[16 \alpha^2 +\epsilon^2 \cos(\lambda t)(-1)^{k-l}i \epsilon
734: \lambda \sin(\lambda t)\big]\Big|\Big) \label{N33}
735: \end{eqnarray}
736: implying the same properties as those given by (\ref{N25}) for the
737: single-mode excited system. We point out that both single- and
738: two-mode pumped couplers for $\alpha=\epsilon$ and the initial
739: states $|01\rangle$ or $|10\rangle$ evolve into the Bell states
740: $|B_3\rangle$ and $|B_4\rangle$ but do not evolve into
741: $|B_1\rangle$ or $|B_2\rangle$. This is contrary to the evolutions
742: of the initial vacuum states (or $|11\rangle$) as, for example,
743: presented in figures 5 and 7.
744:
745: \section{Dissipation}
746:
747: In a more realistic description both fields $a$ and $b$ lose their
748: photons from the cavities. According to the standard techniques in
749: theoretical quantum optics, dissipation of our system can be
750: modelled by its coupling to reservoirs (heat baths) as described
751: by the interaction Hamiltonian
752: \begin{eqnarray}
753: \hat H &=& \hat H_1 + \hat{H}_{\rm loss}, \label{N34}
754: \\
755: \hat{H}_{\rm loss} &=& \hat \Gamma_a \hat f(\hat a,\hat{a}^{\dag
756: })+ \hat \Gamma_b \hat f(\hat b,\hat{b}^{\dag })+{\rm h.c.},
757: \label{N35}
758: \end{eqnarray}
759: where
760: \begin{eqnarray}
761: \hat \Gamma_{a} = \sum_{j=0}^\infty g_{j}^{(a)}
762: \hat{c}_{j}^{(a)},\quad \hat \Gamma_{b} = \sum_{j=0}^\infty
763: g_{j}^{(b)} \hat{c}_{j}^{(b)} \label{N36}
764: \end{eqnarray}
765: are the reservoir operators; $\hat{c}_{j}^{(a,b)}$ are the boson
766: annihilation operators of the reservoir oscillators coupled with
767: mode $a$ or $b$, respectively; $g_{j}^{(a,b)}$ are the coupling
768: constants of the interaction with the reservoirs; $\hat H_1$ is
769: given either by (\ref{N01}) or (\ref{N26}) dependent on the system
770: analyzed. We assume two kinds of functions $\hat f(\hat
771: a,\hat{a}^{\dag })$ and $\hat f(\hat b,\hat{b}^{\dag })$ to
772: describe standard damping and dephasing.
773:
774: The standard description of a damped system is obtained for
775: (\ref{N35}) with $\hat f(\hat a,\hat{a}^{\dag })=\hat a^\dag$ and
776: $\hat f(\hat b,\hat{b}^{\dag })=\hat b^\dag$, or explicitly
777: \begin{eqnarray}
778: \hat{H}_{\rm loss} &=& \hat \Gamma_a\hat{a}^{\dag }+ \hat
779: \Gamma_a^{\dag }\hat{a}+ \hat \Gamma_b \hat{b}^{\dag }+\hat
780: \Gamma_b^{\dag } \hat{b}, \label{N37}
781: \end{eqnarray}
782: corresponding to energy transfer between the system and
783: reservoirs. It should be stressed that the process described by
784: (\ref{N37}) leads to combined effect of amplitude and phase
785: damping. The evolution in the Markov approximation of the reduced
786: density operator $\hat{\rho}$ of the two cavity modes after
787: tracing out over the reservoirs can be described in the
788: interaction picture by the following master equation (see, e.g.,
789: \cite{Gardiner})
790: \begin{eqnarray}
791: \frac{d \hat{\rho}}{d t }&=& -i[\hat{H}_1,\hat{\rho}]+ \hat {\cal
792: L}_{\rm loss}\hat \rho, \label{N38}
793: \end{eqnarray}
794: where the Liouvillian
795: \begin{eqnarray}
796: \hat {\cal L}_{\rm loss}\hat \rho = \frac{\gamma_a}{2}
797: ([\hat a\hat \rho,\hat a^\dagger ]+[\hat a,\hat \rho\hat a^\dagger ])
798: +\frac{\gamma_b}{2}
799: ([\hat b\hat \rho,\hat b^\dagger ]
800: \nonumber \\
801: \quad +[\hat b,\hat \rho\hat b^\dagger ])
802: +\gamma_a \bar n_a [[\hat a,\hat \rho],\hat a^\dagger]
803: +\gamma_b \bar n_b [[\hat b,\hat \rho],\hat b^\dagger]
804: \label{N39}
805: \end{eqnarray}
806: is the usual loss term corresponding to $\hat{H}_{\rm loss}$,
807: given by (\ref{N37}); $\gamma_k$ is the damping rate of the $k$th
808: ($k=a,b$) (ring) cavity, and $\bar n_k=[\exp(\hbar\omega_k/k_B
809: T)-1]^{-1}$ is the mean number of thermal photons at the reservoir
810: temperature $T$. We will analyze both `noisy'reservoirs (at $T>0$
811: implying $\bar n_a,\bar n_b> 0$) and `quiet' reservoirs (at
812: $T\approx 0$, so $\bar n_a=\bar n_b\approx 0$). The latter
813: assumption implies that diffusion of fluctuations from the
814: reservoirs into the system modes is negligible. But still this
815: simplified master equation describes the loss of photons from the
816: system modes to the reservoirs.
817:
818: One can raise some doubts \cite{Alicki} against using Liouvillian
819: (\ref{N39}) in a description of lossy anharmonic oscillator models
820: given by (\ref{N02}). Also the approximation of $T=0$ in
821: (\ref{N39}) is problematic. Nevertheless, master equation
822: (\ref{N38}) with Liouvillian (\ref{N39}) and Hamiltonian $\hat
823: H_1$ set to $\hat{H}^{(a)}_{\rm nonl}$ was used in a number of
824: works both for $T>0$ (see, e.g., \cite{Dan89,Per90,Chat91,Tanas})
825: but also for $T=0$ (see, e.g.,
826: \cite{Mil86,Per88,Mil89,Dan89,Gardiner,Tanas}). Moreover, the same
827: master equation, given by (\ref{N38}) for $\hat H_1$ set to
828: (\ref{N02}), for coupled anharmonic oscillators was applied in,
829: e.g., \cite{Chat91,Per94}. Liouvillian (\ref{N39}) for $T=0$ was
830: also used in (\cite{Gardiner} p. 210) to describe a model
831: essentially similar to ours comprising a nonlinear system, in
832: which two quantized field modes in a cavity interact with a
833: classical pump field. The standard Liouvillian for $T=0$ was also
834: used in \cite{Mil91} to describe a system of Kerr nonlinearity,
835: given by $\hat{H}^{(a)}_{\rm nonl}$, and a parametric amplifier
836: driven by a pulsed classical field. Nevertheless, it should be
837: noted that a realistic master equation for Kerr medium
838: \cite{Dun99,Haus93} is more complicated.
839:
840: Phase damping (also referred to as dephasing) can be described by
841: (\ref{N35}) assuming $\hat f(\hat a,\hat{a}^{\dag })=\hat{a}^{\dag
842: } \hat a$ and $\hat f(\hat b,\hat{b}^{\dag })=\hat{b}^{\dag }\hat
843: b$, which gives the following loss Hamiltonian \cite{Gardiner}:
844: \begin{eqnarray}
845: \hat{H}_{\rm loss} &=& ( \hat \Gamma_a
846: +\hat \Gamma_a^\dag )\hat{a}^{\dag } \hat{a} + ( \hat \Gamma_b
847: +\hat \Gamma_b^\dag )\hat{b}^{\dag } \hat{b} . \label{N40}
848: \end{eqnarray}
849: This interaction can be interpreted as a scattering process, where
850: the number of photons remains unchanged contrary to the
851: interaction described by (\ref{N37}). Phase damping is essential
852: in a fully quantum picture of dissipation of our system. As a
853: simple generalization of the Gardiner-Zoller master equation for a
854: single harmonic oscillator (\cite{Gardiner}, equation (6.1.15)),
855: we describe phase damping of our two-mode nonlinear system by
856: master equation (\ref{N38}) for the Liouvillian
857: \begin{eqnarray}
858: \hat {\cal L}_{\rm loss}\hat \rho = \frac{\gamma_a}{2} (2 \bar
859: n_a+1) [2\hat a^\dag \hat a \hat \rho \hat a^\dag \hat a - (\hat
860: a^\dag
861: \hat a)^2 \hat \rho-\hat \rho (\hat a^\dag \hat a)^2] \notag \\
862: \quad + \frac{\gamma_b}{2} (2 \bar n_b+1) [2\hat b^\dag \hat b
863: \hat \rho \hat b^\dag \hat b - (\hat b^\dag \hat b)^2 \hat
864: \rho-\hat \rho (\hat b^\dag \hat b)^2], \label{N41}
865: \end{eqnarray}
866: which is significantly different from (\ref{N39}).
867:
868: To analyze dissipative evolution, governed by master equations
869: (\ref{N38}) for Liouvillians (\ref{N39}) and (\ref{N41}), we
870: apply standard numerical procedures for solving ordinary
871: differential equations with constant coefficients as an
872: exponential series.
873: %----------------------------------------------------------
874: \begin{figure} % figure 8
875: \epsfxsize=7cm\centerline{\epsfbox{fig08a.eps}}
876: \epsfxsize=7cm\centerline{\epsfbox{fig08b.eps}} \caption{Effect of
877: the standard damping, described by (\ref{N38}) and (\ref{N39}) for
878: quiet reservoirs, on (a) the entanglement of formation and (b)
879: fidelity $F(\hat{\rho},\hat{\rho}_{\rm cut})$, of the states
880: generated in the single-mode pumped coupler from the truncated
881: two-qubit mixed states for $\chi_a=\chi_b=10^8$rad/sec,
882: $\alpha=\chi_a/20$, $\epsilon=\alpha/2$, and the damping constants
883: $\gamma_a=\gamma_b$ equal to 0 (solid), $\chi_a/500$ (dashed), and
884: $\chi_a/200$ (dot-dashed curves). Large dots in figure (a)
885: correspond to the exact solution.}
886: \end{figure}
887: %----------------------------------------------------------
888: \begin{figure} % figure 9
889: \epsfxsize=7cm\centerline{\epsfbox{fig09a.eps}} %
890: \epsfxsize=7cm\centerline{\epsfbox{fig09b.eps}} %
891: \caption{Same as in figure 8 but for noisy reservoirs with mean
892: number $\bar n_a=\bar n_b=0.1$ of thermal photons.}
893: \end{figure}
894: %----------------------------------------------------------
895: \begin{figure} % figure 10
896: \epsfxsize=7cm\centerline{\epsfbox{fig10a.eps}}
897: \epsfxsize=7cm\centerline{\epsfbox{fig10b.eps}} \caption{Effect of
898: the phase damping described by (\ref{N38}) and (\ref{N41})
899: assuming quiet reservoirs for the same parameters as in figure~8.}
900: \end{figure}
901: %----------------------------------------------------------
902: \begin{figure} % figure 11
903: \epsfxsize=7cm\centerline{\epsfbox{fig11a.eps}} %
904: \epsfxsize=7cm\centerline{\epsfbox{fig11b.eps}} %
905: \caption{Same as in figure 10 but for noisy reservoirs with $\bar
906: n_a=\bar n_b=1$, which is 10 times more than in figure 9.}
907: \end{figure}
908:
909: The entropy of entanglement $E$, given by (\ref{N16}), is valid
910: for qudits of arbitrary dimension in a pure state, but it fails to
911: determine entanglement of a system in a mixed state. Thus, for a
912: two-qubit mixed state $\hat \rho$, we have to apply more general
913: measure, e.g., the Wootters measure of entanglement of formation
914: given by \cite{Woo98}:
915: \begin{equation}
916: E_F(\hat \rho)={\cal E}(C(\hat \rho)), \label{N42}
917: \end{equation}
918: where ${\cal E}$ is given by (\ref{N18}) with the argument being
919: the concurrence $C$ defined as
920: \begin{equation}
921: C(\hat \rho) =\max \{2\max_{i}\lambda _{i}-\sum_{i=1}^{4}\lambda
922: _{i},0\}, \label{N43}
923: \end{equation}
924: and $\lambda _{i}$ are the square roots of the eigenvalues of
925: $\hat{\rho}(\hat{\sigma}^{(a)}_{y}\otimes
926: \hat{\sigma}^{(b)}_{y})\hat{\rho} ^{\ast }(\hat{\sigma}^{(a)}_{y}
927: \otimes \hat{\sigma}^{(b)}_{y})$, while $\hat{\sigma}^{(k)} _{y}$
928: is the Pauli spin matrix of the $k$th qubit ($k=a,b$). It is
929: well-known that the entanglement of formation goes into the
930: entropy of entanglement for any two-qubit pure states. Examples of
931: evolution of the entanglement of formation and fidelity for the
932: dissipative systems are shown in figures 8--11 both for quiet and
933: noisy reservoirs. By analyzing the figures, the most important
934: observation is that the generated two-qubit entangled states are
935: very fragile to the leakage of photons from the cavities in the
936: analyzed couplers. Our system is more fragile to losses described
937: by the standard master equation, given by (\ref{N38}) and
938: (\ref{N39}), rather than the losses due to only phase damping as
939: described by the Gardiner-Zoller master equation, given by
940: (\ref{N38}) and (\ref{N41}). Inclusion of reservoir noise, at
941: least for the assumed low mean numbers $\bar n\equiv \bar n_a=\bar
942: n_b$ of thermal photons, does not cause a dramatic deterioration
943: of fidelity and entanglement in comparison to the losses caused by
944: coupling the system to the zero-temperature reservoirs. Note that
945: we have chosen $\bar n=1$ in the dephasing model shown in figure
946: 11 and much smaller value of $\bar n$ in the standard dissipation
947: model in figure 9. The reason is that for the latter dissipation
948: model, the number of photons in the system can be increased by
949: absorbing thermal photons from the reservoirs. If this absorption
950: of thermal photons exceeded the loss of the system photons to the
951: reservoirs then the generated fields could not be approximated as
952: two-qubit states and the Wootters function $E_F$, given by
953: (\ref{N42}), would fail to be a good entanglement measure. For the
954: parameters chosen in figure 9, the generated states are well
955: approximated by two-qubit density matrices and thus its
956: entanglement of formation is well described by (\ref{N42}). By
957: contrast, the number of system photons is not affected by the
958: reservoirs in the dephasing model; thus (\ref{N42}) can be used
959: for any number $\bar n$ of thermal photons.
960:
961: Finally, it should be noted that fragility of our system to
962: dissipation seems to be a serious drawback from an experimental
963: point of view. Nevertheless, a method which enables a significant
964: improvement of the entanglement robustness of the generated
965: states, has recently been suggested for a similar system
966: \cite{Leo05}.
967: %----------------------------------------------------------
968: \begin{figure} % figure 12
969: \hspace*{1cm}\epsfxsize=9cm\centerline{\epsfbox{fig12.eps}}
970: \caption{Level structure of atoms in the Schmidt-Imamo\v{g}lu
971: system \cite{Sch96,Ima97} exhibiting a resonantly enhanced Kerr
972: nonlinearity in mode $a$. An analogous figure can be drawn for
973: mode $b$.}
974: \end{figure}
975:
976: \section{Discussion and conclusions}
977:
978: One of the crucial conditions for the successful truncation and
979: generation of the Bell states in our scheme concerns the couplings
980: $|\alpha|$, $|\beta|$ and $|\epsilon|$ to be much smaller than the
981: Kerr nonlinearities $\chi_a$ and $\chi_b$. This implies that the
982: Kerr interaction should be strong at very low light intensities.
983: Thus, it is desirable to discuss an implementation, in which such
984: stringent conditions can experimentally be satisfied. A possible
985: realization can be based on the effect of the
986: electromagnetically-induced transparency (EIT) or atomic dark
987: resonances as proposed by Schmidt and Imamo\v{g}lu
988: \cite{Sch96,Ima97} (see also \cite{Gra98}) and observed
989: experimentally \cite{Hau99,Kan03}. The Schmidt-Imamo\v{g}lu EIT
990: scheme can be realized in a low density system of four-level
991: atoms, which level structure is shown in figure 12, exhibiting
992: giant resonantly-enhanced Kerr nonlinearity at very low
993: intensities. The atoms are placed in cavity $a$ (and analogously
994: in cavity $b$) tuned to frequency $\omega_a$ of the mode $a$
995: resonant with the transition $|1\rangle \leftrightarrow|3\rangle$
996: and detuned by $\Delta\omega_a$ of the transition
997: $|2\rangle\leftrightarrow|4\rangle$. The EIT effect is created by
998: a classical pumping field of frequency $\omega_{\rm ext}^{(a)}$
999: resonant with the transition $|2\rangle \leftrightarrow
1000: |3\rangle$. By assuming $|g_{13}|^2 n_{\rm atom}/\Omega_{a}^{2}<1$
1001: (see \cite{Gra98}), all the atomic levels can adiabatically be
1002: eliminated, which results in the following formula for the Kerr
1003: nonlinearity \cite{Ima97}:
1004: \begin{equation}
1005: 2\chi_{a} \sim \frac{3\hbar \omega_{a}^2}{2\epsilon_0 V_a}{\rm Re}
1006: (\chi_a^{(3)}) = \frac{3|g^{(a)}_{13}|^2|g^{(a)}_{24}|^2}{
1007: \Omega^2_{a} \Delta\omega_a} n_{\rm atom},\label{N44}
1008: \end{equation}
1009: where $\chi_a^{(3)}$ is third-order nonlinear susceptibility,
1010: $g_{ij}=\mu_{ij}\sqrt{\omega_i/(2\hbar \epsilon_0 V_a)}$ are the
1011: coupling coefficients, $\Omega_{a}$ is the Rabi frequency of the
1012: classical driving (and coupling) field, $\mu_{ij}$ is the electric
1013: dipole matrix element between the states $|i\rangle$ and
1014: $|j\rangle$, $n_{\rm atom}$ is the total number of atoms contained
1015: in the cavity of volume $V_{a}$, and $\epsilon_0$ is the
1016: permittivity of free space. By replacing subscript $a$ by $b$ in
1017: (\ref{N44}), an analogous expression for $\chi_b$ can be obtained.
1018: By putting the stringent limit on the required cavity parameters
1019: \cite{Gra98}, Imamo\v{g}lu {\em et al.} estimated $\chi_{a}\sim
1020: 10^8$ rad/sec \cite{Ima97}. Note that some quantum information
1021: applications of these giant Kerr nonlinearities have already been
1022: studied \cite{Dua00,Vit00,Ott03,Baj04,Mir04}. More details about
1023: application of the Schmidt-Imamo\v{g}lu scheme of
1024: resonantly-enhanced Kerr nonlinearity in the analysis of
1025: dissipation effects on entanglement generation are given by one of
1026: us in \cite{Mir04a}.
1027:
1028: It is worth stressing the differences between the present paper
1029: and our former works. (i) In \cite{Leo04,Leo05}, only the
1030: single-mode pumped systems were studied. Here, we analyze also
1031: two-mode pumped systems, which is basically a different model.
1032: (ii) The model described in \cite{Leo05} is crucially different
1033: from the one used here as was based on the two-mode {\em
1034: nonlinear} interaction term $\hat H_{\rm int}=\epsilon \hat
1035: a^{2\dagger}\hat b^2+{\rm h.c.}$. In the present work, as well as
1036: in \cite{Leo04}, we apply the two-mode {\em linear} interaction
1037: Hamiltonian $\hat H_{\rm int}=\epsilon \hat a^{\dagger}\hat b+{\rm
1038: h.c.}$, which is the same (by neglecting the pump terms $\hat
1039: H^{(a)}_{\rm ext}$ and $\hat H^{(b)}_{\rm ext}$) as that used in
1040: \cite{Che96,Hor89,Kor96,Fiu99,Ibr00,Ari00,San03,ElOrany05}. The
1041: other novelties of the present work in comparison to
1042: \cite{Leo04,Leo05} can be summarized as follows: (1) for
1043: single-mode pumped system, we found a new generalized solution
1044: (\ref{N12}), which, in a special case of equal couplings $\alpha$
1045: and $\epsilon$, simplifies to our former solution obtained in
1046: \cite{Leo04}. (2) Analytical approximate solutions were found here
1047: for various initial Fock states $|n\rangle$. In
1048: \cite{Leo04,Leo05}, solutions were given for initial vacuum states
1049: only. (3) Here, we describe a possible realization of the model
1050: based on the effect of the electromagnetically-induced
1051: transparency. (4) In present numerical analysis based on the EIT
1052: scheme, in comparison to \cite{Leo04,Leo05}, more realistic
1053: parameters were chosen for the coupling constants, Kerr
1054: nonlinearities, and damping constants. (5) The deterioration of
1055: the fidelity of the generated Bell states due to the standard
1056: dissipation and phase damping was analyzed here within two types
1057: of master equations both for quiet and noisy reservoirs. By
1058: contrast, analysis of losses in \cite{Leo04} was limited to the
1059: standard master equation and for the quiet reservoir only. No
1060: effect of dissipation was studied in \cite{Leo05}. (6) In the
1061: present manuscript, entanglement of formation was calculated
1062: analytically for dissipation-free systems and calculated
1063: numerically for dissipative systems. No analytical formulae were
1064: given in \cite{Leo04}, while the entanglement of formation was not
1065: at all studied in \cite{Leo05}. (7) The quality of truncation was
1066: described here, but it was not studied in \cite{Leo04,Leo05}.
1067: Discrepancy between the really generated states and the exact
1068: truncated states was measured here by the fidelity.
1069:
1070: In conclusion, we have described a realization of the generalized
1071: two-mode optical-state truncation of two coherent modes via a
1072: nonlinear process. Our system is a two-mode generalization of the
1073: single-mode nonlinear quantum scissors device described in
1074: \cite{Leo97}. We have described an implementation of the Kerr
1075: nonlinear couplers, where resonantly enhanced nonlinearities can
1076: be achieved in the Schmidt-Imamo\v{g}lu EIT scheme. We have
1077: compared Kerr nonlinear couplers linearly excited in one or two
1078: modes by external classical fields. We have shown under assumption
1079: of the coupling strengths to be much smaller than the nonlinearity
1080: parameters $\chi_a$ and $\chi_b$ that the optical states generated
1081: by the couplers are the two-qubit truncated states spanned by
1082: vacuum and single-photon states. Although our approximate
1083: solutions of the Schr\"odinger equation are independent of
1084: $\chi$-constants, the Kerr nonlinearity plays a crucial role in
1085: the physics as we have derived from the complete $\chi$-dependent
1086: Hamiltonian. In fact, the Kerr interaction is the mechanism in our
1087: model, which enables truncation of the generated state at some
1088: energy level. By contrast, the system without the Kerr
1089: nonlinearities and pumped by an external field would gain more and
1090: more energy. To confirm our predictions, we have compared `exact'
1091: (accurate up to double-precision) direct numerical solutions of
1092: the Schr\"odinger equation and compared with our approximate
1093: analytical solutions. The discrepancies between the exact and
1094: approximate solutions are relatively small as shown by fidelities
1095: in figures 2--4. We have demonstrated that our system initially in
1096: vacuum state or single-photon Fock states $|mn\rangle$ ($m,n=0,1$)
1097: evolves into Bell states. We have discussed the fragility of the
1098: entanglement of formation of the generated states due to the
1099: standard dissipation and dephasing in the two distinct master
1100: equation approaches.
1101:
1102: \begin{acknowledgments}
1103: We thank Professor Ryszard Tana\'s, Professor Ryszard Horodecki
1104: and Professor Robert Alicki for discussions. This work was
1105: supported by the Polish State Committee for Scientific Research
1106: under grant No. 1 P03B 064 28.
1107: \end{acknowledgments}
1108:
1109:
1110: \begin{thebibliography}{9}
1111:
1112: \bibitem{state}
1113: Special issue on {\em Quantum State Preparation and Measurement},
1114: 1997 {\em J. Mod. Opt.} {\bf 44} No. 11/12
1115:
1116: \bibitem{Nie00}
1117: Nielsen M A and Chuang I L 2000 {\em Quantum Computation and
1118: Quantum Information} (Cambridge: University Press)
1119:
1120: \bibitem{Peg98}
1121: Pegg D T, Phillips L S and Barnett S M 1998 {\em Phys. Rev. Lett.}
1122: {\bf 81} 1604
1123:
1124: \item[] Barnett S M and Pegg D T 1999 {\em Phys. Rev.
1125: A} {\bf 60} 4965
1126:
1127: \bibitem{Vil99}
1128: Villas-B\^oas C J, de Almeida N G, and Moussa M H Y 1999 {\em
1129: Phys. Rev. A} {\bf 60} 2759
1130:
1131: \bibitem{Kon00}
1132: Koniorczyk M, Kurucz Z, Gabris A, and Janszky J 2000 {\em Phys.
1133: Rev. A} {\bf 62} 013802
1134:
1135: \bibitem{Par00}
1136: Paris M G A 2000 {\em Phys. Rev. A} {\bf 62} 033813
1137:
1138: \bibitem{Ozd01}
1139: \"Ozdemir \c{S} K, Miranowicz A, Koashi M, and Imoto N 2001 {\em
1140: Phys. Rev. A} {\bf 64} 063818
1141:
1142: \item[]
1143: --- 2002 {\em Phys Rev A} {\bf 66} 053809
1144:
1145: \item[]
1146: --- 2002 {\em J. Mod. Opt.} {\bf 49} 977
1147:
1148: \bibitem{Villas01}
1149: Villas-B\^oas C J, Guimar\v{a}es Y, Moussa M H Y, and Baseia B
1150: 2001 {\em Phys. Rev. A} {\bf 63} 055801
1151:
1152: \bibitem{Mir04}
1153: Miranowicz A and Leo\'nski W 2004 {\em J. Opt. B: Quantum
1154: Semiclass. Opt.} {\bf 6} S43
1155:
1156: \bibitem{Mir05}
1157: Miranowicz A 2005 {\em J. Opt. B: Quantum Semiclass. Opt.} {\bf 7}
1158: 142
1159:
1160: \bibitem{Bab02}
1161: Babichev S A, Ries J, and Lvovsky A I 2003 {\em Europhys. Lett.}
1162: {\bf 64} 1
1163:
1164: \bibitem{Res02}
1165: Resch K J, Lundeen J S, and Steinberg A M 2002 {\em Phys. Rev.
1166: Lett.} {\bf 88} 113601
1167:
1168: \bibitem{Leo97}
1169: Leo\'nski W and Tana\'s R 1994 {\em Phys. Rev. A} {\bf 49} R20
1170:
1171: \item[] Leo\'nski W 1997 {\em Phys. Rev. A} {\bf 55} 3874
1172:
1173: \bibitem{Dar00}
1174: D'Ariano G M, Maccone L, Paris M G A, and Sacchi M F 2000 {\em
1175: Phys. Rev. A} {\bf 61} 053817
1176:
1177: \bibitem{Leo04}
1178: Leo\'nski W and Miranowicz A 2004 {\em J. Opt. B: Quantum
1179: Semiclass. Opt.} {\bf 6} S37
1180:
1181: \bibitem{Jen82}
1182: Jensen S M 1982 {\em IEEE J. Quantum Elect.} {\bf QE-18} 1580
1183:
1184: \bibitem{Mai82}
1185: Maier A M 1982 {\em Kvant. Elektron. (Moscow)} {\bf 9} 2996
1186:
1187: \bibitem{Sny91}
1188: Snyder A W, Mitchell D J, Poladian L, Rowland D R, and Chen Y 1991
1189: {\em J. Opt. Soc. Am. B} {\bf 8} 2102
1190:
1191: \bibitem{Gry01}
1192: Grygiel K and Szlachetka P 2001 {\em J. Opt. B: Quant. Semiclass.
1193: Opt.} {\bf 3} 104
1194:
1195: \bibitem{Per00}
1196: Pe\v{r}ina J Jr and Pe\v{r}ina J 2000 {\em Progress in Optics},
1197: ed. E Wolf (Amsterdam: Elsevier) {\bf 41} 361
1198:
1199: \bibitem{Che96}
1200: Chefles A and Barnett S~M 1996 {\em J. Mod. Opt.} {\bf 43} 709
1201:
1202: \bibitem{Hor89}
1203: Horak R, Sibilia C, Bertolotti M, and Pe\v{r}ina J 1989 {\em J.
1204: Opt. Soc. B} {\bf 6} 199
1205:
1206: \bibitem{Kor96}
1207: Korolkova N and Pe\v{r}ina J 1997 {\em Opt. Commun.} {\bf 136} 135
1208: % \item[] --- 1997 {\em J. Mod. Opt.} {\bf 4} 1525
1209:
1210: \bibitem{Fiu99}
1211: Fiur\'a\v{s}ek J, K\v{r}epelka J, and Pe\v{r}ina J 1999 {\em Opt.
1212: Commun.} {\bf 167} 115
1213:
1214: \bibitem{Ibr00} Ibrahim A -B M A, Umarov B A, Wahiddin M R B
1215: 2000 {\em Phys. Rev.} A {\bf 61} 043804
1216:
1217: \bibitem{Ari00}
1218: Ariunbold G and Perina J 2000 {\em Opt. Commun.} {\bf 176} 149
1219: % J. Mod. Opt. {\bf 48} 1005 (2001).
1220:
1221: \bibitem{San03}
1222: Sanz L, Angelo R M, and Furuya K 2003 {\em J. Phys. A: Math. Gen.}
1223: {\bf 36}, 9737
1224:
1225: \bibitem{ElOrany05}
1226: El-Orany F A A, Sebawe Abdalla M and Pe\v{r}ina J 2005 {\em Eur.
1227: Phys. J. D} {\bf 33} 453
1228:
1229: \bibitem{Her03}
1230: Herec J, Fiur\'a\v{s}ek J and Mi\v{s}ta L Jr 2003 {\em J. Opt. B:
1231: Quant. Semiclass. Opt.} {\bf 5} 419
1232:
1233: \bibitem{MLI01}
1234: Miranowicz A, Leo\'nski W, and Imoto N 2001 {\em Adv. Chem. Phys.}
1235: (New York: Wiley) {\bf 119(I)} 155
1236:
1237: \bibitem{LM01}
1238: Leo\'nski W and Miranowicz A 2001 {\em Adv. Chem. Phys.} (New
1239: York: Wiley) {\bf 119(I)} 195
1240:
1241: \bibitem{Ved02}
1242: Vedral V 2002 {\em Rev. Mod. Phys.} {\bf 74} 197
1243:
1244: \bibitem{Gardiner}
1245: Gardiner C W and Zoller P 2000 {\em Quantum Noise} (Berlin:
1246: Springer)
1247:
1248: \bibitem{Alicki}
1249: Alicki R, private communication.
1250:
1251: \bibitem{Dan89}
1252: Daniel D J and Milburn G J 1989 {\em Phys. Rev. A} {\bf 39} 4628
1253:
1254: \bibitem{Per90}
1255: Pe\v{r}inov\'a V and Luk\v{s} A 1990 {\em Phys. Rev. A} {\bf 41}
1256: 414
1257:
1258: \bibitem{Chat91}
1259: Chaturvedi S and Srinivasan V 1991 {\em Phys. Rev. A} {\bf 43}
1260: 4054
1261:
1262: \bibitem{Tanas}
1263: Tana\'s R 2003 {\em Theory of Non-Classical States of Light}, eds
1264: V Dodonov and V I Man'ko (London: Taylor \& Francis) p 267
1265:
1266: \bibitem{Mil86}
1267: Milburn G J and Holmes C A 1986 {\em Phys. Rev. Lett.} {\bf 56}
1268: 2237
1269:
1270: \bibitem{Per88}
1271: Pe\v{r}inov\'a V and Luk\v{s} A 1988 {\em J. Mod. Opt.} {\bf 35}
1272: 1513
1273:
1274: \bibitem{Mil89}
1275: Milburn G J, Mecozzi A, and Tombesi P 1989 {\em J. Mod. Opt.} {\bf
1276: 36} 1607
1277:
1278: \bibitem{Per94}
1279: Pe\v{r}inov\'a V and Luk\v{s} A 1994 {\em Progress in Optics} vol
1280: {\bf 33}, ed E Wolf (Amsterdam: North-Holland) p 129
1281:
1282: \bibitem{Mil91}
1283: Milburn G J and Holmes C A 1991 {\em Phys. Rev. A} {\bf 44} 4704
1284:
1285: \bibitem{Dun99}
1286: A. M. Dunlop, W. J. Firth, and E. M. Wright 1999 {\em Optics
1287: Express} {\bf 2} 204
1288:
1289: \bibitem{Haus93} Haus H A, Moores J D, and Nelson
1290: L E 1993 {\em Opt. Lett.} {\bf 18} 51
1291: % Effect of third-order dispersion on passive mode locking.
1292:
1293: \bibitem{Woo98}
1294: Wootters W K 1998 {\em Phys. Rev. Lett.} {\bf 80} 2245
1295:
1296: \bibitem{Leo05}
1297: Leo\'nski W and Kowalewska-Kud\l{}aszyk A 2005 {\em Acta Phys.
1298: Hung.} A {\bf 23} 55
1299:
1300: \bibitem{Sch96}
1301: Schmidt H and Imamo\v{g}lu A 1996 {\em Opt. Lett.} {\bf 21} 1936
1302:
1303: \bibitem{Ima97}
1304: Imamo\v{g}lu A, Schmidt H, Woods G, and Deutsch M 1997 {\em Phys.
1305: Rev. Lett.} {\bf 79} 1467
1306:
1307: \item[]
1308: --- 1998 {\em Phys. Rev. Lett.} {\bf 81} 2836
1309:
1310: \bibitem{Gra98} Grangier P, Walls D F, and Gheri K M 1998
1311: {\em Phys. Rev. Lett.} {\bf 81} 2833
1312:
1313: \bibitem{Hau99}
1314: Hau L V, Harris S E, Dutton Z, and Behroozi C H 1999 {\em Nature
1315: (London)} {\bf 397} 594
1316:
1317: \bibitem{Kan03}
1318: Kang H and Zhu Y 2003 {\em Phys. Rev. Lett.} {\bf 91} 093601
1319:
1320: \bibitem{Dua00}
1321: Duan L -M, Giedke G, Cirac J I, and Zoller P 2000 {\em Phys. Rev.
1322: Lett.} {\bf 84} 4002
1323:
1324: \bibitem{Vit00}
1325: Vitali D, Fortunato M, and Tombesi P 2000 {\em Phys. Rev. Lett.}
1326: {\bf 85} 445
1327:
1328: \bibitem{Ott03}
1329: Ottaviani C, Vitali D, Artoni M, Cataliotti F, and Tombesi P 2003
1330: {\em Phys. Rev. Lett.} {\bf 90} 197902
1331:
1332: \bibitem{Baj04}
1333: Bajer J, Miranowicz A and Andrzejewski M 2004 {\em J. Opt. B:
1334: Quantum Semiclass. Opt.} {\bf 6} 387
1335:
1336: \bibitem{Mir04a}
1337: Miranowicz A 2004 {\em J. Phys. A: Math. Gen.} {\bf 37} 7909
1338:
1339: \end{thebibliography}
1340:
1341: \end{document}
1342: