1: \documentclass[pra,showpacs,showkeys,amsfonts,amsmath,twocolumn]{revtex4}
2: \usepackage{bm}
3: %\usepackage[cp1250]{inputenc}
4: \usepackage{graphicx}
5: \RequirePackage{mathptm}
6: \newtheorem{thm}{Theorem}
7: \newtheorem{cor}{Corollary}
8: %\newtheorem{lem}[thm]{Lemma}
9: %\newtheorem{defn}[thm]{Definition}
10: \newtheorem{rem}{Remark}
11: \numberwithin{equation}{section}
12: \newcommand{\si}[1]{\sigma_{#1}}
13: \newcommand{\sa}[2]{\sigma_{#1}^{#2}}
14: \newcommand{\ip}[2]{\langle \,{#1},\,{#2}\,\rangle}
15: \newcommand{\W}[4]{\begin{cases}
16: #1 ,\\
17: #3 ,
18: \end{cases}}
19: \newcommand{\ro}{\rho}
20: \newcommand{\ras}{\rho_{\mr{as}}}
21: \newcommand{\la}{\lambda}
22: \newcommand{\La}{\Lambda}
23: \newcommand{\al}{\alpha}
24: \newcommand{\be}{\beta}
25: \newcommand{\el}{\ell}
26: \newcommand{\g}{\gamma}
27: \newcommand{\ga}[1]{\gamma_{#1}}
28: \newcommand{\Ga}[1]{\Gamma_{#1}}
29: \newcommand{\Om}[1]{\Omega_{#1}}
30: \newcommand{\re}{\mathrm{Re}\,}
31: \newcommand{\I}{\openone}
32: \newcommand{\conj}[1]{\overline{#1}}
33: \newcommand{\ket}[1]{|{#1}\rangle}
34: \newcommand{\bra}[1]{\langle {#1} |}
35: \newcommand{\cH}{{\mathcal H}}
36: \newcommand{\Z}{\mathbb Z}
37: \newcommand{\C}{\mathbb C}
38: \newcommand{\R}{\mathbb R}
39: \newcommand{\M}{\mathbb M}
40: \newcommand{\PP}{\mathbb P}
41: \newcommand{\Wu}{\mathfrak W}
42: \newcommand{\fA}{\mathfrak A}
43: \newcommand{\fE}{\mathcal E}
44: \newcommand{\tr}{\mathrm{tr}\,}
45: \newcommand{\ptr}[1]{\mathrm{tr}_{#1}}
46: \newcommand{\tl}[1]{\boldsymbol #1}
47: \newcommand{\mr}[1]{\mathrm{#1}}
48: \newcommand{\e}[1]{{\boldsymbol e}_{#1}}
49: %\newcommand{\begin{equation}}{\begin{equation}}
50: %\newcommand{\end{equation}}{\end{equation}}
51: \newcommand{\DS}{\displaystyle}
52: \begin{document}
53: \title{Dynamical creation of entanglement versus disentanglement in a system
54: of three - level atoms with vacuum - induced coherences}
55: \author{{\L}ukasz Derkacz}
56: \affiliation{Institute of Theoretical Physics\\ University of
57: Wroc{\l}aw\\
58: Plac Maxa Borna 9, 50-204 Wroc{\l}aw, Poland}
59: \author{Lech Jak{\'o}bczyk
60: \footnote{E-mail addres: ljak@ift.uni.wroc.pl} }
61: \affiliation{Institute of Theoretical Physics\\ University of
62: Wroc{\l}aw\\
63: Plac Maxa Borna 9, 50-204 Wroc{\l}aw, Poland}
64: \begin{abstract}
65: The dynamics of entanglement between three - level atoms coupled to
66: the common vacuum is investigated. We show that the collective
67: effects such as collective damping, dipole - dipole interaction and
68: the cross coupling between orthogonal dipoles, play a crucial role
69: in the process of creation of entanglement. In particular, the
70: additional cross coupling enhances the production of entanglement.
71: For the specific initial states we find that the effect of delayed
72: sudden birth of entanglement, recently invented by Ficek and Tana\'s
73: [Phys. Rev. A \textbf{77}, 054301(2008)] in the case of two - level
74: atoms, can also be observed in the system. When the initial state is
75: entangled, the process of spontaneous emission causes destruction of
76: correlations and its disentanglement. We show that the robustness of
77: initial entanglement against the noise can be changed by local
78: operations performed on the state.
79: \end{abstract}
80: \pacs{03.67.Mn, 03.65.Yz, 42.50.-p} \keywords{three - level
81: atoms, vacuum - induced coherence , entanglement production}
82: \maketitle
83: \section{Introduction}
84: In a system of coupled multi - level atoms having closely lying
85: energy states and interacting with the vacuum, quantum interference
86: between different radiative transitions can occur, resulting in
87: coherences in a system which are known as \textit{vacuum - induced
88: coherences}. For example, when the distance between atoms is
89: comparable to the wavelength of the emitted radiation and transition
90: dipole moments involving the decay processes are parallel, the
91: coupling between the atoms via common vacuum gives rise to the
92: collective effects such as collective damping and dipole - dipole
93: interaction. Such effects are well known \cite{Ag, FS}, particularly
94: in the case of two - level atoms. In the system of three - level
95: atoms, radiative coupling can produce a new interference effect in
96: the spontaneous emission. This effect manifests by the cross
97: coupling between radiative transitions with \textit{orthogonal}
98: dipole moments \cite{AP} and is strongly dependent on the relative
99: orientation of the atoms \cite{EK,SE}. All such collective
100: properties of the system influence the quantum dynamics, which can
101: significantly differ from a corresponding single atom dynamics.
102: There were many studies on the effect of quantum interference on
103: various physical processes including: resonance fluorescence
104: \cite{HP}, quantum jumps \cite{Z}, the presence of ultranarrow
105: spectral lines \cite{ZS} or amplification without population
106: inversion \cite{Ha}.
107: \par
108: In our research we consider entanglement properties of a pair of
109: three - level atoms in the \textsf{V} configuration with vacuum
110: induced coherences. We study a dynamical creation of entanglement
111: due to the collective effects which are present in the system, as
112: well as the process of degradation of correlations, resulting in
113: disentanglement of initially entangled pairs of atoms. Both
114: processes crucially depend on the interatomic distance compared to
115: the wavelength of the emitted radiation. For large separation we
116: expect that the collective properties of two atoms are negligible
117: and dissipation causes disentanglement. On the contrary, for small
118: distance the collective effects are so strong that they can
119: partially overcome decoherence. As a result, the system can decay to
120: a stationary state which can be entangled, even if the initial state
121: was separable \cite{DJ1}.
122: \par
123: In the present paper, we study the case of distance comparable to
124: the radiation wavelength $\lambda$. Although the dynamics brings all
125: initial states into the asymptotic state in which both atoms are in
126: their ground states, still there can be some transient entanglement
127: between the atoms. In particular we show that the dynamical creation
128: of entanglement is possible in a system where only one atom is in
129: the excited state. Moreover the production of entanglement is
130: enhanced, when the cross coupling between orthogonal dipoles is
131: present. In the more accessible initial state when the both atoms
132: are excited, and if the cross coupling is absent, the interesting
133: phenomenon of delayed sudden birth of entanglement \cite{FT} can be
134: observed: unentangled atoms become entangled after some finite time,
135: despite of the fact that the correlation between the atoms existed
136: earlier. On the other hand, cross coupling causes that the
137: entanglement starts to build up immediately. We consider also the
138: process of disentanglement of initially entangled states in the
139: presence of vacuum induced coherences. Analogously to the case of
140: two - level atoms (see e.g. \cite{FTpr}), there are specific
141: entangled states of our system which decay much slower then the
142: other states. In the limit of small separation, those states
143: decouple from the environment and therefore are stable. They are
144: called (generalized) antisymmetric Dicke states \cite{BGZ} and play
145: the crucial role in characterizing disentanglement properties of
146: given initial state. In particular, the class of maximally entangled
147: states of two - qutrits i.e. (generalized) Bell states \cite{BHN}
148: can be divided into two subsets. The first set contains those states
149: which have no populations in antisymmetric Dicke states, and they
150: decay rapidly. The remaining states have equal populations in stable
151: Dicke states and decay much slower. Since all Bell states are
152: locally equivalent, local operations performed on the states may
153: change the robustness of entanglement against the noise.
154: \section{Model dynamics}
155: We start with a short description of the model studied by Agarwal
156: and Patnaik \cite{AP}. Consider two identical three - level atoms (
157: $A$ and $B$) in the $V$ configuration. The atoms have two near -
158: degenerate excited states $\ket{1_{\al}},\; \ket{2_{\al}}$ ($\al
159: =A,B$) and ground states $\ket{3_{\al}}$. Assume that the atoms
160: interact with the common vacuum and that transition dipole moments
161: of atom $A$ are parallel to the transition dipole moments of atom
162: $B$. Due to this interaction, the process of spontaneous emission
163: from two excited levels to the ground state take place in each
164: individual atom but a direct transition between excited levels is
165: not possible. Moreover, the coupling between two atoms can be
166: produced by the exchange of the photons. As it was shown by Agarwal
167: and Patnaik, in such atomic system there is also possible the
168: radiative process in which atom $A$ in the excited state
169: $\ket{1_{A}}$ loses its excitation which in turn excites atom $B$ to
170: the state $\ket{2_{B}}$. This effect manifests by the cross coupling
171: between radiation transitions with orthogonal dipole moments. The
172: evolution this atomic system can be described by the following
173: master equation \cite{AP}
174: \begin{equation}
175: \frac{d\ro}{dt}=(L^{A}+L^{B}+L^{AB})\ro\label{me}
176: \end{equation}
177: where for $\al=A,B$ we have
178: \begin{equation}
179: L^{\al}\ro=\sum\limits_{k=1}^{2}\ga{k3}\,\left(\,2\sa{3k}{\al}\ro\sa{k3}{\al}-\sa{a3}{\al}
180: \sa{3k}{\al}\ro-\ro\sa{k3}{\al}\sa{3k}{\al}\right) \label{genA}
181: \end{equation}
182: and
183: \begin{equation}
184: \begin{split}
185: L^{AB}\ro=&\hspace*{3mm}\sum\limits_{k=1}^{2}\sum\limits_{\al=A,B}\Ga{k3}\,(\,
186: 2\sa{3k}{\al}\ro\sa{k3}{\neg\al}-\sa{k3}{\neg\al}\sa{3k}{\al}\ro-\ro\sa{k3}{\neg\al}\sa{3k}{\al})\\[1mm]
187: &\hspace*{3mm}+i\,\sum\limits_{k=1}^{2}\Om{k3}\,
188: \left[\,\sa{k3}{A}\sa{3k}{B}+\sa{k3}{B}\sa{3k}{A},\ro\,\right]\\[1mm]
189: &\hspace*{3mm}+\Ga{vc}\sum\limits_{\al=A,B}(\,2\sa{31}{\al}\ro\sa{23}{\neg\al}
190: -\sa{23}{\neg\al}\sa{31}{\al}\ro-
191: \ro\sa{23}{\neg\al}\sa{31}{\al}\\[1mm]
192: &\hspace*{20mm}+2\sa{32}{\al}\ro\sa{13}{\neg\al}-\sa{13}{\neg\al}\sa{32}{\al}\ro
193: -\ro\sa{13}{\neg\al}\sa{32}{\al}\,)\\[2mm]
194: &\hspace*{3mm}+i\,\Om{vc}\sum\limits_{\al=A,B}
195: [\sa{23}{\al}\sa{31}{\neg\al}+\sa{32}{\al}\sa{13}{\neg\al},\ro\,]
196: \label{genAB}
197: \end{split}
198: \end{equation}
199: In the equations (\ref{genA}) and (\ref{genAB}), $\neg\al$ is $A$
200: for $\al=B$ and $B$ for $\al=A$, $\sa{jk}{\al}$ is the transition
201: operator from $\ket{k_{\al}}$ to $\ket{j_{\al}}$ and the
202: coefficient $\ga{j3}$ represents the single atom spontaneous - decay
203: rate from the state $\ket{j}$ ( $j=1,2$ ) to the state $\ket{3}$.
204: The coefficients $\Ga{j3}$ and $\Om{j3}$ are related to the coupling
205: between two atoms and are the collective damping and the dipole -
206: dipole interaction potential, respectively. The coherence terms
207: $\Ga{vc}$ and $\Om{vc}$ are cross coupling coefficients, which
208: couple a pair of orthogonal dipoles. This cross coupling between two
209: atoms strongly depend on the relative orientation of the atoms. To
210: see this, we put the atom A at the origin of coordinate system and
211: the position of the atom B is give by the vector $\vec{R}$ which
212: makes na angle $\phi$ with the $x$ axis and an angle $\theta$ with
213: the $z$ axis (see FIG. 1). Assume that the dipole moments
214: $\vec{d}_{13}$ and $\vec{d}_{23}$ of transitions $\ket{1_{\al}}\to
215: \ket{3_{\al}}$ and $\ket{2_{\al}}\to \ket{3_{\al}}$ are given by
216: $$
217: \vec{d}_{13}=\hat{x}d,\quad \vec{d}_{23}=\hat{y}d
218: $$
219: \begin{figure}[t]
220: \centering {\includegraphics[height=52mm]{fv1n.eps}}\caption{The
221: considered geometry of two - atomic system.}
222: \end{figure}
223: Since the states $\ket{1_{\al}}$ and $\ket{2_{\al}}$ are closely
224: lying, the transition frequencies $\omega_{13}$ and $\omega_{23}$
225: satisfy
226: $$
227: \omega_{13}\approx\omega_{23}=\omega_{0}
228: $$
229: Similarly, the spontaneous - decay rates
230: $$
231: \ga{13}\approx\ga{23}=\g
232: $$
233: As was shown in Ref.\cite{AP}, the coefficients in (\ref{genAB}) can
234: be written as
235: \begin{equation}
236: \begin{split}
237: &\Ga{13}=\frac{3\g}{2}\left(P_{i}-\sin^{2}\theta\cos^{2}\phi\,
238: Q_{i}\right)\\
239: &\Om{13}=\frac{3\g}{2}\left(P_{r}-\sin^{2}\theta\cos^{2}\phi\,
240: Q_{r}\right)\\
241: &\Ga{23}=\frac{3\g}{2}\left(P_{i}-\sin^{2}\theta\sin^{2}\phi\,Q_{i}\right)\\
242: &\Om{23}=\frac{3\g}{2}\left(P_{r}-\sin^{2}\theta\sin^{2}\phi\,Q_{r}\right)\\
243: &\Ga{vc}=-\frac{3\g}{2}\sin^{2}\theta\sin\phi\cos\phi\,Q_{i}\\
244: &\Om{vc}=-\frac{3\g}{2}\sin^{2}\theta\sin\phi\cos\phi\,Q_{r}
245: \end{split}\label{cc}
246: \end{equation}
247: where for $\xi=R\,\omega_{0}/c$
248: \begin{equation}
249: \begin{split}
250: &P_{i}=\frac{\sin\xi}{\xi}+\frac{\cos\xi}{\xi^{2}}-\frac{\sin\xi}{\xi^{3}},\quad
251: Q_{i}=\frac{\sin\xi}{\xi}+3\frac{\cos\xi}{\xi^{2}}-3\frac{\sin\xi}{\xi^{3}}\\[2mm]
252: &P_{r}=\frac{\cos\xi}{\xi}-\frac{\sin\xi}{\xi^{2}}-\frac{\cos\xi}{\xi^{3}},\quad
253: Q_{r}=\frac{\cos\xi}{\xi}-3\frac{\sin\xi}{\xi^{2}}-3\frac{\cos\xi}{\xi^{3}}
254: \end{split}\label{PQ}
255: \end{equation}
256: From the formulas (\ref{cc}) and (\ref{PQ}) it follows that the
257: coupling coefficients are small for large distance between the atoms
258: and tend to zero for $R\to \infty$. On the other hand, when $R\to
259: 0$, $\Om{13},\, \Om{23}$ and $\Om{vc}$ diverge, whereas
260: $$
261: \Ga{13},\, \Ga{23}\to \g\quad\text{and}\quad \Ga{vc}\to 0
262: $$
263: In the following we will consider two special configurations of
264: atomic system.
265: \par
266: \textit{Configuration I}: $\theta=\pi$ i.e. both atoms lie along
267: the $z$ axis and $\phi=\pi/4$. In that case
268: $$
269: \Ga{13}=\Ga{23},\quad \Om{13}=\Om{23}
270: $$
271: and the coherence terms $\Ga{vc}=\Om{vc}=0$.
272: \par
273: \textit{Configuration II}: $\theta=\pi/2$ i.e. both atoms lie on
274: the $xy$ plane and $\phi=\pi/4$. In that case
275: $$
276: \Ga{13}=\Ga{23},\quad \Om{13}=\Om{23}
277: $$
278: and the coherence terms
279: $$
280: \Ga{vc}\neq 0,\quad \Om{vc}\neq 0.
281: $$
282: \par
283: The time evolution of the initial state of two - atomic system is
284: given by the semi - group $\{ T_{t} \}_{t\geq 0}$ of completely
285: positive mappings acting on density matrices, generated by
286: $L^{A}+L^{B}+L^{AB}$. The properties of this semi - group crucially
287: depend on the distance between the two atoms and the geometry of the
288: system. It can be shown by a direct calculation, that irrespective
289: to the geometry, when the distance is large (compared to the
290: radiation wavelength $\lambda$), the semi - group $\{ T_{t}
291: \}_{t\geq 0}$ is uniquely relaxing with the asymptotic state
292: $\ket{3_{A}}\otimes\ket{3_{B}}$. Thus, for any initial state, its
293: entanglement approaches $0$ when $t\to \infty$. But still there can
294: be some transient entanglement between the atoms. In the following,
295: we study in details time evolution of some classes of initial states
296: and show how the creation of entanglement as well as the process of
297: disentanglement are sensitive to the geometry of the system.
298: \section{Negativity}
299: To describe the process of creation or destruction of entanglement
300: between the atoms, we need the effective measure of mixed - state
301: entanglement. For such a measure we take a computable measure of
302: entanglement proposed in \cite{VW}. The measure is based on the
303: trace norm of the partial transposition $\ro^{PT}$ of the state
304: $\ro$. From the Peres - Horodecki criterion of separability \cite{P,
305: HHH}, it follows that if $\ro^{PT}$ is not positive, then $\rho$ is
306: entangled and one defines the \textit{negativity} of the state
307: $\rho$ as
308: \begin{equation}
309: N(\ro)=\frac{||\ro^{PT}||-1}{2}
310: \end{equation}
311: $N(\ro)$ is equal to the absolute value of the sum of the negative
312: eigenvalues of $\ro^{PT}$ and is an entanglement monotone, but it
313: cannot detect bound entangled states \cite{H}.
314: \par
315: Although negativity of a given state is easy to compute numerically,
316: the analytical formulas for general mixed states of two qutrits can
317: be only obtained for some limited classes of states. The density
318: matrix $\ro$ which we consider to compute negativity is defined on
319: the space $\C^{3}\otimes \C^{3}$ and $\ro$ is written in the basis
320: of product states
321: \begin{equation}
322: \ket{j_{A}}\otimes\ket{k_{B}},\quad j,k=1,2,3 \label{basis}
323: \end{equation}
324: taken in the lexicographic order. In particular, for the states of
325: the form
326: \begin{equation}
327: \ro=\begin{pmatrix} 0&0&0&0&0&0&0&0&0\\
328: 0&0&0&0&0&0&0&0&0\\
329: 0&0&\ro_{33}&0&0&\ro_{36}&\ro_{37}&\ro_{38}&0\\
330: 0&0&0&0&0&0&0&0&0\\
331: 0&0&0&0&0&0&0&0&0\\
332: 0&0&\ro_{63}&0&0&\ro_{66}&\ro_{67}&\ro_{68}&0\\
333: 0&0&\ro_{73}&0&0&\ro_{76}&\ro_{77}&\ro_{78}&0\\
334: 0&0&\ro_{83}&0&0&\ro_{86}&\ro_{87}&\ro_{88}&0\\
335: 0&0&0&0&0&0&0&0&\ro_{99}
336: \end{pmatrix}\label{state13}
337: \end{equation}
338: the negativity is given by
339: \begin{equation}
340: N(\ro)=\frac{1}{2}\left[\,\sqrt{4\,(\,|\ro_{37}|^{2}+|\ro_{38}|^{2}+|\ro_{67}|^{2}+
341: |\ro_{68}|^{2}\,)+\ro_{99}^{2}}-\ro_{99}\,\right] \label{neg13}
342: \end{equation}
343: Notice that (\ref{neg13}) is equal to zero when the coherences
344: $\ro_{37},\,\ro_{38},\, \ro_{67},\,\ro_{68}$ are all equal to zero,
345: and is greater then zero when at least one of them is nonzero.
346: Similarly for the states
347: \begin{equation}
348: \ro=\begin{pmatrix} 0&0&0&0&0&0&0&0&0\\
349: 0&\ro_{22}&0&0&0&0&0&0&0\\
350: 0&0&\ro_{33}&0&0&0&\ro_{37}&0&0\\
351: 0&0&0&0&0&0&0&0&0\\
352: 0&0&0&0&0&0&0&0&0\\
353: 0&0&0&0&0&\ro_{66}&0&\ro_{68}&0\\
354: 0&0&\ro_{73}&0&0&0&\ro_{77}&0&0\\
355: 0&0&0&0&0&\ro_{86}&0&\ro_{88}&0\\
356: 0&0&0&0&0&0&0&0&\ro_{99}
357: \end{pmatrix},\label{state12}
358: \end{equation}
359: the negativity can be computed from the formula
360: \begin{equation}
361: N(\ro)=\frac{1}{2}\left[\,\sqrt{4\,(\,|\ro_{37}|^{2}+
362: |\ro_{68}|^{2}\,)+\ro_{99}^{2}}-\ro_{99}\,\right]. \label{neg12}
363: \end{equation}
364: \par
365: On the other hand, for the states
366: \begin{equation}
367: \ro=\begin{pmatrix} \ro_{11}&0&0&0&0&0&0&0&0\\
368: 0&0&0&0&0&0&0&0&0\\
369: 0&0&\ro_{33}&0&0&0&\ro_{37}&0&0\\
370: 0&0&0&0&0&0&0&0&0\\
371: 0&0&0&0&0&0&0&0&0\\
372: 0&0&0&0&0&0&0&0&0\\
373: 0&0&\ro_{73}&0&0&0&\ro_{77}&0&0\\
374: 0&0&0&0&0&0&0&0&0\\
375: 0&0&0&0&0&0&0&0&\ro_{99}
376: \end{pmatrix},\label{state11}
377: \end{equation}
378: their negativity
379: \begin{equation}
380: N(\ro)=\max\,\left(\,0,\, \widetilde{N}(\ro)\,\right)\label{neg11}
381: \end{equation}
382: where
383: \begin{equation}
384: \widetilde{N}(\ro)=\frac{1}{2}
385: \left[\sqrt{(\ro_{11}-\ro_{99})^{2}+4\,|\ro_{37}|^{2}}-\ro_{11}-\ro_{99}\right]
386: \label{neg11f}
387: \end{equation}
388: can be zero, even if the coherence $\ro_{37}$ is not zero. There is
389: a threshold for the coherence at which two atoms become entangled.
390: \section{Creation of entanglement}
391: In this section we study the process of creation of transient
392: entanglement between atoms prepared in separable initial states. We
393: fix the distance between the atoms and solve numerically the master
394: equation (\ref{me}) in two cases of configurations of the system.
395: \subsection{Initial states $\ket{1_{A}}\otimes\ket{3_{B}}$ and
396: $\ket{1_{A}}\otimes\ket{2_{B}}$}
397: \begin{figure}[h]
398: \centering
399: {\includegraphics[height=86mm,angle=270]{fv2n.eps}}\caption{The time
400: evolution of $|\ro_{37}|$ for the initial state
401: $\ket{1_{A}}\otimes\ket{3_{B}}$ when $\Ga{vc}=\Om{vc}=0$ and
402: $R/\lambda=0.2$. }
403: \end{figure}
404: When the system is prepared in the pure state $\ket{1_{A}}\otimes
405: \ket{3_{B}}$ (atom A in the excited state and atom B in the ground
406: state) and both atoms lie along the $z$ axis (\textit{Configuration
407: I}), so
408: $$
409: \Ga{12}=\Ga{23},\quad \Om{13}=\Om{23}\quad\text{and}\quad
410: \Ga{vc}=\Om{vc}=0,
411: $$
412: one can check that the density matrix at time $t$ takes the form
413: \begin{equation}
414: \ro(t)=\begin{pmatrix} 0&0&0&0&0&0&0&0&0\\
415: 0&0&0&0&0&0&0&0&0\\
416: 0&0&\ro_{33}(t)&0&0&0&\ro_{37}(t)&0&0\\
417: 0&0&0&0&0&0&0&0&0\\
418: 0&0&0&0&0&0&0&0&0\\
419: 0&0&0&0&0&0&0&0&0\\
420: 0&0&\ro_{73}(t)&0&0&0&\ro_{77}(t)&0&0\\
421: 0&0&0&0&0&0&0&0&0\\
422: 0&0&0&0&0&0&0&0&\ro_{99}(t)
423: \end{pmatrix}\label{ro13}
424: \end{equation}
425: and by (\ref{neg13})
426: \begin{equation}
427: N(t)=\frac{1}{2}\,\left[\,\sqrt{4\,|\ro_{37}(t)|^{2}+\ro_{99}(t)^{2}}-\ro_{99}(t)\,\right]
428: \label{neg13t}
429: \end{equation}
430: %\begin{figure}[h] \centering
431: %{\includegraphics[height=86mm,angle=270]{fv2n.eps}}\caption{The time
432: %evolution of $|\ro_{37}|$ for the initial state
433: %$\ket{1_{A}}\otimes\ket{3_{B}}$ when $\Ga{vc}=\Om{vc}=0$ and
434: %$R/\lambda=0.2$. }
435: %\end{figure}
436: Since the process of the photon exchange produces coherence between
437: the states $\ket{1_{A}}\otimes \ket{3_{B}}$ and $\ket{3_{A}}\otimes
438: \ket{1_{B}}$, the value of $|\ro_{37}|$ starts to grow and the
439: system becomes entangled (FIG. 2).
440: \begin{figure}[h]
441: \centering
442: {\includegraphics[height=80mm,angle=270]{fv3n.eps}}\caption{The time
443: evolution of $|\ro_{38}|$, $|\ro_{67}|$, $|\ro_{68}|$ for the
444: initial state $\ket{1_{A}}\otimes\ket{3_{B}}$ in the case when
445: $\Ga{vc}\neq 0,\, \Om{vc}\neq 0$ and $R/\lambda=0.2$. }
446: \end{figure}
447: \par
448: When both atoms lie on the $xy$ plane (\textit{Configuration II}),
449: the cross coupling coefficients $\Ga{vc}$ and $\Om{vc}$ are nonzero
450: and the dynamics of the system is changed significantly. The
451: additional coupling between orthogonal dipoles produces new
452: coherences $\ro_{36},\, \ro_{38},\, \ro_{67},\, \ro_{68}$ and
453: $\ro_{78}$, so the state at time $t$ has the form (\ref{state13}).
454: In particular, the values of $|\ro_{37}|,\, |\ro_{38}|,\,
455: |\ro_{67}|$ and $|\ro_{68}|$ become nonzero (FIG. 3), so the
456: negativity of the state can be computed from the formula
457: (\ref{neg13}).
458: \begin{figure}[h]
459: \centering
460: {\includegraphics[height=86mm,angle=270]{fv4n.eps}}\caption{The time
461: evolution of negativity of initial state
462: $\ket{1_{A}}\otimes\ket{3_{B}}$ in two cases: $\Ga{vc}\neq 0,\,
463: \Om{vc}\neq 0$ ($N_{vc}$) and $\Ga{vc}=\Om{vc}=0$ ($N$). In both
464: cases $R/\lambda =0.2$.}
465: \end{figure}
466: In FIG. 4 we plot the time evolution of negativity of initial state
467: $\ket{1_{A}}\otimes\ket{3_{B}}$ in both configurations. As we see,
468: the cross coupling between the atoms enhances the production of
469: entanglement. The same behaviour of negativity can be observed for
470: initial state $\ket{2_{A}}\otimes\ket{3_{B}}$.
471: \par
472: On the other hand, when the system is prepared in the initial state
473: $\ket{1_{A}}\otimes\ket{2_{B}}$ (both atoms in excited states) and
474: the cross coupling is absent, the entanglement production is due to
475: the coherences $\ro_{37}$ and $\ro_{68}$. As in the previous case,
476: the presence of cross coupling enhances the production of
477: entanglement, but the maximal value of negativity is much less then
478: in the case of initial state $\ket{1_{A}}\otimes \ket{3_{B}}$ (FIG.
479: 5).
480: %\begin{figure}[t]
481: %\centering {\includegraphics[height=50mm]{fv3.eps}}\caption{The time
482: %evolution of $|\ro_{37}|$ (black), $|\ro_{38}|$ (red),
483: %$|\ro_{67}|$ (green), $|\ro_{68}|$ (blue) for the initial state
484: %$\ket{1_{A}}\otimes\ket{3_{B}}$ in the case $\theta=\pi/2,\,
485: %\phi=\pi/4$ and $R/\lambda=0.2$. }
486: %\end{figure}
487: \begin{figure}[h]
488: \centering
489: {\includegraphics[height=85mm,angle=270]{fv5n.eps}}\caption{The time
490: evolution of negativity of initial state
491: $\ket{1_{A}}\otimes\ket{2_{B}}$ in two cases: $\Ga{vc}\neq 0,\,
492: \Om{vc}\neq 0$ ($N_{vc}$) and $\Ga{vc}=\Om{vc}=0$ ($N$). In both
493: cases $R/\lambda=0.2$. }
494: \end{figure}
495: \subsection{Initial state $\ket{1_{A}}\otimes\ket{1_{B}}$ and
496: delayed sudden birth of entanglement}
497: \begin{figure}[h]
498: \centering
499: {\includegraphics[height=88mm,angle=270]{fv6n.eps}}\caption{The time
500: evolution of negativity of initial state
501: $\ket{1_{A}}\otimes\ket{1_{B}}$. Here we take $\Ga{vc}=\Om{vc}=0$
502: and $R/\lambda =0.2$.}
503: \end{figure}
504: If the system is prepared in the state
505: $\ket{1_{A}}\otimes\ket{1_{B}}$ (both atoms are in the same excited
506: state), and the cross coupling is absent, the state at time $t$
507: takes the form (\ref{state11}). As in the case of initial state
508: $\ket{1_{A}}\otimes\ket{3_{B}}$, the entanglement production is due
509: to the creation of coherence $\ro_{37}$, but in the present case,
510: the nonzero coherence is only the necessary condition for
511: entanglement. As it follows from (\ref{neg11}) and (\ref{neg11f}),
512: there is a threshold for $|\ro_{37}|$ at which the negativity
513: becomes nonzero. A detailed numerical analysis shows that there is
514: no entanglement at earlier times, and suddenly at some time the
515: entanglement starts to build up (FIG. 6). This is the example of
516: phenomenon of \textit{delayed sudden birth of entanglement}, studied
517: by Ficek and Tana\'s \cite{FT} in the case of two - level atoms.
518: \begin{figure}[h]
519: \centering
520: {\includegraphics[height=85mm,angle=270]{fv7n.eps}}\caption{The time
521: evolution of negativity of initial state (\ref{superposition}) for
522: different values of $\phi$ ($\Ga{vc}=\Om{vc}=0,\, R/\lambda=0.2$). }
523: \end{figure}
524: \begin{figure}[b]
525: \centering
526: {\includegraphics[height=80mm,angle=270]{fv8n.eps}}\caption{The time
527: of birth of entanglement as a function of $\phi$.}
528: \end{figure}
529: %\begin{figure}[h]
530: %\centering {\includegraphics[height=50mm]{fv6.eps}}\caption{The time
531: %evolution of negativity of initial state
532: %$\ket{1_{A}}\otimes\ket{1_{B}}$}
533: %\end{figure}
534: To get some insight into the process of creation of entanglement in
535: this case, consider the initial state
536: \begin{equation}
537: \ket{\Psi}=\cos\phi\, \ket{1_{A}}\otimes\ket{1_{B}}+\sin\phi\,
538: \ket{1_{A}}\otimes\ket{3_{B}},\quad
539: \phi\in[0,\pi/2]\label{superposition}
540: \end{equation}
541: FIG. 7 shows that the evolution of initial state
542: (\ref{superposition}) crucially depends on the superposition angle
543: $\phi$. The smaller is the probability that the system is prepared
544: in the state $\ket{1_{A}}\otimes\ket{1_{B}}$, the earlier the atoms
545: become entangled. In FIG. 8 we plot the time of the birth of
546: entanglement as the function of the superposition angle. This time
547: is maximal for $\phi=0$ (i.e.
548: $\ket{\Psi}=\ket{1_{A}}\otimes\ket{1_{B}}$) and is equal to zero for
549: $\phi=\pi/2$.
550: \begin{figure}[h]
551: \centering
552: {\includegraphics[height=85mm,angle=270]{fv9n.eps}}\caption{The time
553: evolution of negativity of initial state
554: $\ket{1_{A}}\otimes\ket{1_{B}}$ when $\Ga{vc}\neq 0,\, \Om{vc}\neq
555: 0$ and $R/\lambda=0.2$.}
556: \end{figure}
557: %\begin{figure}[h]
558: %\centering {\includegraphics[height=50mm]{fv7.eps}}\caption{The time
559: %evolution of negativity of initial state (\ref{superposition}) for
560: %$\phi=\pi/16$ (red), $\phi=\pi/8$ (green) and $\phi=3\pi/16$ (blue)}
561: %\end{figure}
562: %\begin{figure}[h]
563: %\centering
564: %{\includegraphics[height=80mm,angle=270]{fv8.eps}}\caption{The time
565: %of birth of entanglement as a function of $\phi$}
566: %\end{figure}
567: \par
568: When the cross coupling coefficients $\Ga{vc}$ and $\Om{vc}$ are not
569: zero the numerical analysis shows that the time delayed creation of
570: entanglement does not occur. In that case, even for the initial
571: state $\ket{1_{A}}\otimes\ket{1_{B}}$ (or
572: $\ket{2_{A}}\otimes\ket{2_{B}}$), the entanglement starts to build
573: up immediately after the atoms begin to interact with the vacuum
574: (FIG. 9).
575: %\begin{figure}[h]
576: %\centering {\includegraphics[height=50mm]{fv9.eps}}\caption{The time
577: %evolution of negativity of initial state
578: %$\ket{1_{A}}\otimes\ket{1_{B}}$ when the cross coupling is present.}
579: %\end{figure}
580: \section{Disentanglement}
581: Apart from the effect of creation of entanglement, the quantum
582: evolution given by the master equation (\ref{me}) may cause also
583: destruction of correlations, resulting in disentanglement of
584: initially entangled states. In this section we study the process of
585: disentanglement for some entangled pure initial states. We start
586: with the characterization of maximally entangled states of two three
587: - level systems (two qutrits).
588: \subsection{Maximally entangled states of two qutrits and generalized Dicke states}
589: The basis $\ket{\Psi_{\alpha}},\, \alpha =1,\ldots, 9$ of the space
590: $\C^{3}\otimes \C^{3}$ consisting of maximally entangled Bell - like
591: states was constructed in \cite{BHN} (see also \cite{DJ}). The
592: states $\ket{\Psi_{\alpha}}$ can be written as follows:
593: \begin{equation}
594: \begin{split}
595: &\ket{\Psi_{1}}=\frac{1}{\sqrt{3}}\,\left(\ket{1_{A}}\otimes\ket{1_{B}}
596: +\ket{2_{A}}\otimes\ket{2_{B}}+\ket{3_{A}}\otimes\ket{3_{B}}\,\right)\\
597: &\ket{\Psi_{2}}=\frac{1}{\sqrt{3}}\,\left(\ket{1_{A}}\otimes\ket{2_{B}}+
598: \ket{2_{A}}\otimes\ket{3_{B}}+\ket{3_{A}}\otimes\ket{1_{B}}\,\right)\\
599: &\ket{\Psi_{3}}=\frac{1}{\sqrt{3}}\,\left(\ket{1_{A}}\otimes\ket{3_{B}}
600: +\ket{2_{A}}\otimes\ket{1_{B}}+\ket{3_{A}}\otimes\ket{2_{B}}\,\right)\\
601: &\ket{\Psi_{4}}=\frac{1}{\sqrt{3}}\,\left(\ket{1_{A}}\otimes\ket{1_{B}}+
602: w\ket{2_{A}}\otimes\ket{2_{B}}+\conj{w}\ket{3_{A}}\otimes\ket{3_{B}}\,\right)\\
603: &\ket{\Psi_{5}}=\frac{1}{\sqrt{3}}\,\left(\ket{1_{A}}\otimes\ket{2_{B}}+
604: w\ket{2_{A}}\otimes\ket{3_{B}}+\conj{w}\ket{3_{A}}\otimes\ket{1_{B}}\,\right)\\
605: &\ket{\Psi_{6}}=\frac{1}{\sqrt{3}}\,\left(\ket{1_{A}}\otimes\ket{3_{B}}+
606: w\ket{2_{A}}\otimes\ket{1_{B}}+\conj{w}\ket{3_{A}}\otimes\ket{2_{B}}\,\right)\\
607: &\ket{\Psi_{7}}=\frac{1}{\sqrt{3}}\,\left(\ket{1_{A}}\otimes\ket{1_{B}}+
608: \conj{w}\ket{2_{A}}\otimes\ket{2_{B}}+w\ket{3_{A}}\otimes\ket{3_{B}}\,\right)\\
609: &\ket{\Psi_{8}}=\frac{1}{\sqrt{3}}\,\left(\ket{1_{A}}\otimes\ket{2_{B}}+
610: \conj{w}\ket{2_{A}}\otimes\ket{3_{B}}+w\ket{3_{A}}\otimes\ket{1_{B}}\,\right)\\
611: &\ket{\Psi_{9}}=\frac{1}{\sqrt{3}}\,\left(\ket{1_{A}}\otimes\ket{3_{B}}+
612: \conj{w}\ket{2_{A}}\otimes\ket{1_{B}}+w\ket{3_{A}}\otimes\ket{2_{B}}\,\right)
613: \end{split}\label{bell}
614: \end{equation}
615: where
616: $$
617: w=e^{2\pi i/3}
618: $$
619: One can check that the states (\ref{bell}) have maximal negativity
620: and that they are locally equivalent.
621: \par
622: There is another class of pure entangled states of two qutrits which
623: are very important for the analysis of the dynamics of coupled three
624: - level atoms. The \textit{generalized symmetric and antisymmetric
625: Dicke states} (see e.g. \cite{BGZ}), defined by the formulas
626: \begin{equation}
627: \begin{split}
628: &\ket{s_{kl}}=\frac{1}{\sqrt{2}}\,\left(\ket{k_{A}}\otimes\ket{l_{B}}
629: +\ket{l_{A}}\otimes\ket{k_{B}}\,\right)\\
630: &\ket{a_{kl}}=\frac{1}{\sqrt{2}}\,\left(\ket{k_{A}}\otimes\ket{l_{B}}
631: -\ket{l_{A}}\otimes\ket{k_{B}}\,\right)
632: \end{split}\label{dicke}
633: \end{equation}
634: where $k,l=1,2,3;\, k<l$, are not maximally entangled (their
635: negativity is equal to $1/2$) but have a remarkable properties. As
636: it was shown in our previous paper \cite{DJ1}, in the limit of small
637: separation between the atoms, the process of photon exchange
638: produces such correlations that the dynamics is not ergodic and
639: there are nontrivial asymptotic stationary states. In that case, the
640: symmetric Dicke states $\ket{s_{kl}}$ decay to the ground state
641: $\ket{3_{A}}\otimes\ket{3_{B}}$ whereas antisymmetric states
642: $\ket{a_{13}}$ and $\ket{a_{23}}$ decouple from the environment and
643: therefore are stable. Moreover, the state $\ket{a_{12}}$ is not
644: stable, but is asymptotically nontrivial. So for the distances
645: comparable to the radiation wavelength $\lambda$, symmetric and
646: antisymmetric Dicke states will decay with significantly different
647: rates and the populations in the antisymmetric states can be used to
648: characterize disentanglement properties of given initial state.
649: \subsection{Time evolution of Dicke states}
650: Now we consider the evolution of the antisymmetric Dicke state
651: \begin{equation}
652: \ket{a_{13}}=\frac{1}{\sqrt{2}}\,\left(\ket{1_{A}}\otimes\ket{3_{B}}-
653: \ket{3_{A}}\otimes\ket{1_{B}}\,\right)\label{a13}
654: \end{equation}
655: in the case when the distance between the atoms is comparable to
656: $\lambda$. If $\Ga{vc}=\Om{vc}=0$, the state at time $t$ takes the
657: form (\ref{ro13}), so the degree of its entanglement is determined
658: by the coherence $\ro_{37}(t)$. The same is true for the symmetric
659: state
660: \begin{equation}
661: \ket{s_{13}}=\frac{1}{\sqrt{2}}\,\left(\ket{1_{A}}\otimes\ket{3_{B}}+
662: \ket{3_{A}}\otimes\ket{1_{B}}\,\right)\label{s13}
663: \end{equation}
664: which decays to the ground state even in the limit of small
665: separation. As we show numerically, time evolution of $\ro_{37}$ for
666: the symmetric state (\ref{s13}), differs significantly from that
667: for antisymmetric state (\ref{a13}) (see FIG. 10) and the latter
668: disentangle much slower that the former (FIG. 11).
669: \begin{figure}[h]
670: \centering
671: {\includegraphics[height=85mm,angle=270]{fv10n.eps}}\caption{The
672: time evolution of $|\ro_{37}|$ for the symmetric and antisymmetric
673: Dicke states in the case $\Ga{vc}=\Om{vc}=0$ and $R/\lambda=0.2
674: $.}
675: \end{figure}
676: \begin{figure}[h]
677: \centering
678: {\includegraphics[height=85mm,angle=270]{fv11n.eps}}\caption{Disentanglement
679: of Dicke states $\ket{s_{13}}$ and $\ket{a_{13}}$
680: ($\Ga{vc}=\Om{vc}=0,\, R/\lambda=0.2$).}
681: \end{figure}
682: \par
683: When the cross coupling coefficients are not zero, the dynamics of
684: Dicke states is more complicated, and at a given time $t$ they have
685: the form (\ref{state13}). Detailed analysis of the evolution of
686: negativity indicates that for the antisymmetric state
687: $\ket{a_{13}}$, additional coupling between transitions with
688: orthogonal dipole moments, slow down the process of disentanglement
689: (FIG. 12). On the other hand, this coupling does not influence rapid
690: disentanglement of the symmetric state.
691: \par
692: The antisymmetric state $\ket{a_{12}}$ is not stable in the limit of
693: small separation between the atoms, but it is asymptotically
694: nontrivial \cite{DJ1}. It can be shown that $\ket{a_{12}}$ evolves
695: to the asymptotic state $\ro$ which has the form (\ref{state12})
696: with
697: $$
698: \ro_{22}=\ro_{99}=0,\quad
699: \ro_{33}=\ro_{66}=\ro_{88}=\frac{1}{4},\quad
700: \ro_{37}=\ro_{68}=-\frac{1}{4}
701: $$
702: so by (\ref{neg12}), the asymptotic negativity of $\ket{a_{12}}$ has
703: the value $\sqrt{2}/4$. For the atom separation comparable with
704: $\lambda$, this state disentangle quicker then $\ket{a_{13}}$.
705: \begin{figure}[h]
706: \centering
707: {\includegraphics[height=85mm,angle=270]{fv12n.eps}}\caption{Disentanglement
708: of Dicke state $\ket{a_{13}}$ in two cases: $\Ga{vc},\, \Om{vc}\neq
709: 0 $ (VC) and $\Ga{vc}=\Om{vc}=0$ (NON VC). In both cases $R/\lambda
710: =0.2$}
711: \end{figure}
712: \subsection{Disentanglement of Bell states}
713: As it was stated before, in the case of small separation between the
714: atoms, the antisymmetric Dicke states $\ket{a_{12}},\, \ket{a_{23}}$
715: are stable, and the state $\ket{a_{12}}$ has a nontrivial asymptotic
716: limit. For that reason, the initial states which have the property
717: of trapping populations in $\ket{a_{13}},\, \ket{a_{23}}$ and
718: $\ket{a_{12}}$, decay to an entangled asymptotic states. In the
719: process of evolution, the initial populations in $\ket{a_{13}}$ and
720: $\ket{a_{23}}$ are conserved, whereas the population in
721: $\ket{a_{12}}$ can be transformed into $\ket{a_{13}}$ and
722: $\ket{a_{23}}$, giving the enlargement of initial populations.
723: \par
724: Consider now the Bell states (\ref{bell}). One checks that the
725: states $\ket{\Psi_{1}},\, \ket{\Psi_{4}}$ and $\ket{\Psi_{7}}$ have
726: zero populations in $\ket{a_{13}},\, \ket{a_{23}}$ and
727: $\ket{a_{12}}$, so they decay to the separable asymptotic state. In
728: fact, the limiting state in this case is the ground state
729: $\ket{3_{A}}\otimes\ket{3_{B}}$. On the other hand, the remaining
730: Bell states $\ket{\Psi_{2}},\, \ket{\Psi_{3}},\, \ket{\Psi_{5}},\,
731: \ket{\Psi_{6}},\, \ket{\Psi_{8}}$ and $\ket{\Psi_{9}}$ have equal
732: populations in antisymmetric Dicke states, so they have the same
733: asymptotic entanglement. Take for example the state
734: $\ket{\Psi_{2}}$. Since the corresponding populations are equal
735: $1/6$, by the general result of \cite{DJ1} the asymptotic state is
736: of the form
737: \begin{equation}
738: \ras=\begin{pmatrix} 0&0&\hspace*{2mm}0&0&0&\hspace*{2mm}0&\hspace*{2mm}0
739: &\hspace*{2mm}0&\hspace*{2mm}0\\[2mm]
740: 0&0&\hspace*{2mm}0&0&0&\hspace*{2mm}0&\hspace*{2mm}0&\hspace*{2mm}0&\hspace*{2mm}0\\[2mm]
741: 0&0&\hspace*{2mm}\frac{1}{8}&0&0&-\frac{1}{12}&-\frac{1}{8}
742: &\hspace*{2mm}\frac{1}{12}&\hspace*{2mm}\frac{1}{12}\\[2mm]
743: 0&0&\hspace*{2mm}0&0&0&\hspace*{2mm}0&\hspace*{2mm}0
744: &\hspace*{2mm}0&\hspace*{2mm}0\\[2mm]
745: 0&0&\hspace*{2mm}0&0&0&\hspace*{2mm}0&\hspace*{2mm}0&\hspace*{2mm}0&\hspace*{2mm}0\\[2mm]
746: 0&0&-\frac{1}{12}&0&0&\hspace*{2mm}\frac{1}{8}
747: &\hspace*{2mm}\frac{1}{12}&-\frac{1}{8}&-\frac{1}{12}\\[2mm]
748: 0&0&-\frac{1}{8}&0&0&\hspace*{2mm}\frac{1}{12}
749: &\hspace*{2mm}\frac{1}{8}&-\frac{1}{12}&-\frac{1}{12}\\[2mm]
750: 0&0&\hspace*{2mm}\frac{1}{12}&0&0&-\frac{1}{8}
751: &-\frac{1}{12}&\hspace*{2mm}\frac{1}{8}&\hspace*{2mm}\frac{1}{12}\\[2mm]
752: 0&0&\hspace*{2mm}\frac{1}{12}&0&0&-\frac{1}{12}
753: &-\frac{1}{12}&\hspace*{2mm}\frac{1}{12}&\hspace*{2mm}\frac{1}{2}
754: \end{pmatrix}
755: \end{equation}
756: Its negativity can be computed numerically and one obtains that
757: $N(\ras)\simeq 0.0968$.
758: \par
759: If the distance $R$ is comparable with the wavelength $\lambda$ and
760: the cross coupling is absent, the populations in $\ket{a_{13}}$ and
761: $\ket{a_{23}}$ are no longer conserved but increase at the beginning
762: (since the population in $\ket{a_{12}}$ decreases) and then decay
763: much slower than the populations of the remaining states (FIG. 13).
764: \begin{figure}[h]
765: \centering
766: {\includegraphics[height=85mm,angle=270]{fv13n.eps}}\caption{Time
767: evolution of population in antisymmetric state $\ket{a_{13}}$,
768: symmetric state $\ket{s_{13}}$ and antisymmetric state
769: $\ket{a_{12}}$ for the initial Bell state $\ket{\Psi_{2}}$. We take
770: here $\Ga{vc}=\Om{vc}=0$ and $R/\lambda =0.08$.}
771: \end{figure}
772: In this way, the entanglement of state $\ket{\Psi_{2}}$ is more
773: robust against the noise than the entanglement of $\ket{\Psi_{1}}$
774: (FIG. 14).
775: \begin{figure}[h]
776: \centering
777: {\includegraphics[height=85mm,angle=270]{fv14n.eps}}\caption{Disentanglement
778: of the Bell state $\ket{\Psi_{1}}$ and $\ket{\Psi_{2}}$ . We take
779: here $\Ga{vc}=\Om{vc}=0$ and $R/\lambda=0.08$.}
780: \end{figure}
781: The same is true for the other states from this class. So all Bell
782: states can be divided into two classes. The states from the first
783: class containing $\ket{\Psi_{1}},\, \ket{\Psi_{4}}$ and
784: $\ket{\Psi_{7}}$ decay rapidly, whereas remaining states decay much
785: slower. Since the Bell states are locally equivalent, local
786: operations performed on the initial state can change the robustness
787: of entanglement against the noise. On the other hand, as we show
788: numerically, the influence of cross coupling between the atoms on
789: the process of disentanglement of Bell states is negligible (FIG.
790: 15).
791: \begin{figure}[ht]
792: \centering
793: {\includegraphics[height=85mm,angle=270]{fv15n.eps}}\caption{Disentanglement
794: of the Bell state $\ket{\Psi_{2}}$ in two cases: $\Ga{vc}=\Om{vc}=0$
795: (NON VC) and $\Ga{vc},\, \Om{vc}\neq 0$(VC). In both cases
796: $R/\lambda=0.08$.}
797: \end{figure}
798: \section{Conclusions}
799: We have studied the dynamics of entanglement in the system of three
800: - level atoms in the \textsf{V} configuration, coupled to the
801: common vacuum and separated by a distance comparable to the
802: radiation wavelength. In this case only some transient entanglement
803: between the atoms can exist but the dynamical generation of such
804: entanglement is possible. It happens for example, when the cross
805: coupling between orthogonal dipoles is absent and initially only one
806: atom is excited. Additional coupling enhances the production of
807: entanglement and causes that entanglement can be produced also in
808: the case when two atoms are excited. Initial states with two atoms
809: excited lead also to the interesting phenomenon of delayed sudden
810: birth of entanglement. The process of disentanglement of initially
811: entangled states is less sensitive to cross coupling between the
812: atoms. We have shown this for the maximally entangled Bell states.
813: On the other hand, the rate of disentanglement of Bell states
814: crucially depends on populations of initial state in the
815: antisymmetric Dicke states, which are more robust against the noise
816: then the Bell states. We have demonstrated that those Bell states
817: which have no populations in the antisymmetric Dicke states rapidly
818: disentangle, whereas remaining Bell states disentangle much slower.
819: \begin{thebibliography}{99}
820: \bibitem{Ag} G. S. Agarwal, \textit{Quantum Statistical Theories of
821: Spontaneous Emission and Their Relation to Other Approaches}
822: (Springer, Berlin, 1974).
823: \bibitem{FS} Z. Ficek and S. Swain, \textit{Quantum Interference and
824: Coherence: Theory and Experiments} (Springer, Berlin, 2005).
825: \bibitem{AP} G. S. Agarwal, A. K. Patnaik, Phys. Rev. A \textbf{63},
826: 043805(2001).
827: \bibitem{EK} J. Evers, M. Kiffner, M. Macovei and Ch. H. Keitel,
828: Phys. Rev. A \textbf{73}, 023804(2006).
829: \bibitem{SE} S.I. Schmid and J. Evers, Phys. Rev. A \textbf{77},
830: 013822(2008).
831: \bibitem{HP} G.C. Hegerfeldt and M.B. Plenio, Phys. Rev. A
832: \textbf{46}, 373(1992).
833: \bibitem{Z} P. Zoller, M. Marte and D.W. Walls, Phys. Rev. A
834: \textbf{35}, 198(1987).
835: \bibitem{ZS} P. Zhou and S. Swain, Phys. Rev. Lett. \textbf{77},
836: 3995(1996).
837: \bibitem{Ha} S.E. Harris, Phys. Rev. Lett. \textbf{62}, 1033(1989).
838: \bibitem{DJ1} {\L}. Derkacz and L. Jak\'obczyk, J. Phys. A
839: \textbf{41}, 205304(2008).
840: \bibitem{FT} Z. Ficek and R. Tana\'s, Phys. Rev. A \textbf{77},
841: 054301(2008).
842: \bibitem{FTpr} Z. Ficek and R. Tana\'s, Phys. Rep. \textbf{372},
843: 369(2002).
844: \bibitem{BGZ}I.V. Bargatin, B.A. Grishanin and V.N. Zadkov, Phys.
845: Rev. A \textbf{61}, 052305(2000).
846: \bibitem{BHN} B. Baumgartner, B.C. Hiesmayr and H. Narnhofer, Phys.
847: Rev. A \textbf{74}, 032327(2006).
848: \bibitem{VW} G. Vidal and R.F. Werner, Phys. Rev. A \textbf{65},
849: 032314(2002).
850: \bibitem{P} A. Peres, Phys, Rev. Lett. \textbf{77}, 1413(1996).
851: \bibitem{HHH} M. Horodecki, P. Horodecki and R. Horodecki, Phys.
852: Lett. A \textbf{223}, 1(1996).
853: \bibitem{H} P. Horodecki, Phys. Lett. A \textbf{232}, 333(1997).
854: \bibitem{DJ} {\L}. Derkacz and L. Jak\'obczyk, Phys. Rev. A
855: \textbf{76}, 042304(2007).
856: \end{thebibliography}
857: \end{document}
858: