1: %%% Intrinsic decoherence paper %%%%%
2:
3:
4: \documentclass[aps,epsfig,rotate,showpacs]{revtex4} % DON'T CHANGE
5: %\documentstyle[aps,pra,epsfig,twocolumn]{revtex} % DON'T CHANGE
6: \usepackage{epsfig}
7: \usepackage{graphicx}
8: \usepackage{amsmath}
9:
10: %
11: %
12: \newcommand{\MF}{{\large{\manual META}\-{\manual FONT}}}
13: \newcommand{\manual}{rm} % Substitute rm (Roman) font.
14: \newcommand\bs{\char '134 } % add backslash char to \tt font
15:
16: %
17: %
18: \begin{document} % INITIALIZE - DONT CHANGE
19: %
20: %
21: \title{
22: Intrinsic Decoherence Dynamics in Smooth Hamiltonian Systems: Quantum-classical
23: Correspondence}
24: \author{Jiangbin Gong and Paul Brumer}
25: \affiliation{Chemical Physics Theory Group,\\Department of Chemistry,\\
26: University of Toronto\\ Toronto, Canada M5S 3H6}
27: %
28: \date{\today}
29:
30:
31:
32: \begin{abstract} % DON'T CHANGE THIS LINE
33: A direct classical analog of
34: the quantum dynamics of intrinsic decoherence in Hamiltonian systems,
35: characterized by the time dependence of the linear entropy of the reduced density operator,
36: is introduced.
37: The similarities and differences between
38: the classical and quantum decoherence dynamics of an initial quantum state
39: are exposed using both analytical and computational results.
40: In particular, the classicality of early-time intrinsic decoherence dynamics
41: is explored analytically using a second-order perturbative treatment, and an interesting
42: connection between
43: decoherence rates and the stability nature
44: of classical trajectories is revealed in a simple approximate
45: classical theory of intrinsic decoherence dynamics.
46: The results offer new insights into decoherence, dynamics of quantum
47: entanglement,
48: and quantum chaos.
49: \end{abstract}
50:
51: \pacs{03.65.Yz, 03.65.Ud, 05.45.Mt}
52: \maketitle
53:
54:
55: \section{Introduction}
56: Quantum dynamics induces unitary transformations in Hilbert space, but most often
57: it is only the dynamics projected onto a Hilbert {\it subspace}
58: that is of interest. In general this reduced dynamics is nonunitary and
59: therefore displays decoherence \cite{zurekreview}.
60: For example, if a system of interest is coupled to a bath, then
61: averaging over the bath degrees of freedom introduces decoherence in the system dynamics.
62: Likewise, in an isolated system, the reduced dynamics of
63: a subsystem of this isolated system can display decoherence.
64: We have termed decoherence in the latter case
65: ``{\it intrinsic decoherence}'' since it does not involve an external bath \cite{batista}.
66:
67: Understanding decoherence is of crucial importance
68: to a variety of modern fields such as quantum information
69: processing \cite{qcbook} and quantum control of atomic and molecular
70: processes \cite{brumerreview,ricebook,brumerbook}. Our interest here is in
71: the quantum-classical correspondence (QCC) between classical and
72: quantum descriptions of the {\it dynamics} of decoherence.
73: Specifically, we consider an initial quantum state subjected to either quantum
74: or classical dynamics and compare the time evolution of the decoherence
75: in both cases. We note that
76: the formal theory of correspondence between quantum dynamics and
77: classical Liouville dynamics \cite{wilkie} suggests that
78: classical Liouville dynamics projected
79: onto a subspace should also display decoherence.
80: That is, as in the quantum
81: case, the classical Liouville dynamics considered in the entire phase space is unitary and
82: the classical Liouville dynamics projected onto a subspace
83: is nonunitary. We therefore expect that
84: the reduced classical Liouville dynamics propagated
85: classically will show decoherence dynamics that is, at least qualitatively,
86: parallel to that seen in the reduced quantum dynamics insofar as the loss of phase
87: information, entropy production, etc.
88: In the case of
89: bath-induced decoherence we recently showed analytically that (a)
90: one can indeed introduce a direct classical analog of quantum
91: decoherence, and (b) examining the dynamics of decoherence
92: classically gives new insights into both the dynamics of
93: decoherence described quantum mechanically and into the conditions
94: for QCC in decoherence dynamics \cite{gongprl}.
95:
96: Here we extend these considerations to intrinsic decoherence, both analytically and computationally.
97: Specifically, in this paper we study QCC in the dynamics
98: of intrinsic decoherence in smooth Hamiltonian systems, with an emphasis on the usefulness
99: of classical dynamics in describing intrinsic decoherence.
100: In particular, the classicality of early-time intrinsic decoherence dynamics
101: is studied using a second-order perturbative treatment, and the interesting
102: connection between
103: decoherence rates at later times and the stability
104: properties of classical trajectories is revealed by considering
105: a simple approximate
106: classical theory of intrinsic decoherence dynamics.
107: The analytic and computational results shed new light on
108: decoherence, dynamics of quantum entanglement,
109: and quantum chaos. This study is also of interest to semiclassical
110: decoherence studies \cite{bmiller}, e.g., semiclassical descriptions of
111: intrinsic decoherence dynamics
112: in large molecular systems \cite{batista}.
113:
114: This paper is organized as follows. In Sec. \ref{perturbation-section}
115: we introduce a second-order perturbation theory in an effort to understand
116: QCC in early-time intrinsic decoherence dynamics.
117: For simplicity we focus upon two degree-of-freedom systems, but the extension to larger systems
118: is straightforward.
119: Computational results of two sample cases in coupled-oscillator model systems,
120: which strongly support
121: the physical picture afforded by the perturbative treatment, are presented in
122: Sec. \ref{sample}.
123: Then, a classical theory
124: of intrinsic decoherence dynamics for initially localized states is derived in Sec.
125: \ref{classical-theory}. In the same section,
126: detailed computational studies using this
127: simple theory are carried out for the
128: quartic oscillator model and one of its variants.
129: Discussions and a summary comprise Sec. \ref{summary}.
130:
131:
132: \section{Early-time Intrinsic Decoherence Dynamics}
133: \label{perturbation-section}
134: Consider a conservative
135: system composed of two subsystems, with the total Hamiltonian given by
136: \begin{eqnarray}
137: H(Q,P,q,p)=\frac{P^{2}}{2}+\frac{p^{2}}{2}+V_{1}(Q)+V_{2}(q)+V_{12}(Q,q),
138: \end{eqnarray}
139: where $(Q,P)$ and ($q,p$) are dimensionless
140: phase space conjugate variables,
141: $V_{i}$ is the potential of the $i$-th subsystem, and
142: $V_{12}(Q,q)$ describes arbitrary coupling between the two subsystems.
143: As the system evolves
144: the total system wavefunction $|\psi(t)\rangle$
145: becomes inseparable due to quantum entanglement,
146: even if it is initially separable in $Q$ and $q$.
147: As a result, measuring a subsystem would collapse the system wavefunction
148: and therefore affect the properties
149: of the other subsystem. Similarly,
150: ignoring a subsystem decoheres the other one.
151: The degree of intrinsic decoherence,
152: which is induced by, and is a manifestation of, quantum entanglement between the two subsystems,
153: can be measured by
154: a well-known quantity: the quantum linear entropy \cite{linear-entropy}
155: \begin{eqnarray}
156: S_{{\rm q}}=
157: 1-Tr_{1}(\hat{\tilde{\rho}}^{2}),
158: \end{eqnarray}
159: where $Tr_{i}$ denotes a trace over the $i$-th subsystem, and
160: $\hat{\tilde{\rho}}\equiv Tr_{2} \left(|\psi(t)\rangle\langle\psi(t)|\right)$
161: is the reduced density operator
162: for the first subsystem.
163: An increase in $S_{{\rm q}}$ suggests
164: an increase of $1/(1-S_{{\rm q}})$, which gives
165: the number of orthogonal quantum states that are incoherently populated if the second subsystem
166: is ignored. Below we choose $q, p$ as the ``bath" variables and $P,Q$ as the system variables.
167:
168: Let $\rho_{c}(Q,P,q,p,t)$ denote the phase space
169: distribution function evolved classically, and $\rho_{W}(Q,P,q,p,t)$ denote the
170: quantum (Wigner) phase space
171: distribution function.
172: Their time evolution equations are given by
173: \begin{eqnarray}
174: \frac{\partial \rho_{c}}{\partial t}& = &\{H,\rho_{c}\}, \\
175: \frac{\partial \rho_{W}}{\partial t}&=& \{H,\rho_{W}\}_{M},
176: \end{eqnarray}
177: where $\{\cdot\}$ denotes classical Poisson bracket and
178: $\{\cdot\}_{M}$ denotes quantum Moyal bracket \cite{moyal}.
179: We define classical and quantum reduced distribution functions as
180: \begin{eqnarray} \tilde{\rho}_{c}(Q,P,t) &\equiv &
181: \int \rho_{c}(Q,P,q,p,t)\ dq \ dp, \\
182: \tilde{\rho}_{W}(Q,P,t) & \equiv & \int \rho_{W}(Q,P,q,p,t)\ dq\ dp.
183: \end{eqnarray}
184: Since \begin{eqnarray}
185: S_{{\rm q}}(t)=1-2\pi\hbar\int\tilde{\rho}_{W}^{2}(Q,P,t)\ dQ\ dP,
186: \end{eqnarray}where $\hbar$ is the effective
187: Planck constant,
188: we can define a
189: classical analog [denoted $S_{c}(t)$] to $S_{{\rm q}}(t)$
190: by replacing $\tilde{\rho}_{W}$ with
191: $\tilde{\rho}_{c}$. That is,
192: \begin{eqnarray}
193: S_{c}(t)\equiv 1-2\pi\hbar\int\tilde{\rho}_{c}^{2}(Q,P,t)\ dQ\ dP.
194: \end{eqnarray}
195: The main focus here is to
196: compare $S_{c}(t)$
197: with $S_{{\rm q}}(t)$, i.e., the classical
198: vs. quantum evolution of the intrinsic decoherence dynamics, as measured by the
199: classical vs. quantum entropy.
200:
201: Perturbative treatments have proved to very useful in understanding
202: decoherence dynamics \cite{gongprl,kim,duan,bacon}. Here, to analytically examine
203: classical vs. quantum intrinsic decoherence dynamics at early times,
204: we apply the perturbative approach developed
205: in our previous work \cite{gongprl} to the case of intrinsic decoherence dynamics.
206: Specifically, consider a second-order
207: perturbative expansion with respect to the time variable $t$ for both $S_{{\rm q}}$ and $S_{c}$, i.e.,
208: \begin{eqnarray}
209: S_{c}(t)& = & S_{c}(0)+\frac{t}{\tau_{c,1}}+\frac{t^{2}}{\tau_{c,2}^{2}}+\cdots, \nonumber \\
210: S_{{\rm q}}(t)& = & S_{{\rm q}}(0)+\frac{t}{\tau_{{\rm q},1}}+\frac{t^{2}}{\tau_{{\rm q},2}^{2}}+\cdots.
211: \end{eqnarray}
212: Then, from the classical and quantum dynamics of the entire system
213: one obtains
214: \begin{eqnarray}
215: \frac{1}{\tau_{c,1}}=-4\pi\hbar\int \tilde{\rho}_{c}(Q,P,0)
216: \int \{H, \rho_{c}(Q,P,q,p,0)\}\ dq\ dp \ dQ\ dP,
217: \end{eqnarray}
218: \begin{eqnarray}
219: \frac{1}{\tau_{{\rm q},1}}=-4\pi\hbar\int \tilde{\rho}_{W}(Q,P,0)
220: \int \{H, \rho_{W}(Q,P,q,p,0)\}_{M}\ dq\ dp \ dQ \ dP,
221: \end{eqnarray}
222: \begin{eqnarray}
223: \frac{1}{\tau_{c,2}^{2}}& = &-2\pi\hbar\int \tilde{\rho}_{c}(Q,P,0)
224: \int \left\{H, \left\{H,
225: \rho_{c}(Q,P,q,p,0)\right\}\right\}\ dq\ dp \ dQ \ dP
226: \nonumber \\
227: & & -2\pi\hbar\int \left[ \int \left\{H,
228: \rho_{c}(Q,P,q,p,0)\right\}\ dq\ dp\right]^{2} \ dQ\ dP,
229: \label{t2c-a}
230: \end{eqnarray}
231: and
232: \begin{eqnarray}
233: \frac{1}{\tau_{{\rm q},2}^{2}}& = &-2\pi\hbar\int \tilde{\rho}_{W}(Q,P,0)
234: \int \left\{H, \left\{H,
235: \rho_{W}(Q,P,q,p,0)\right\}_{M}\right\}_{M}\ dq \ dp \ dQ \ dP
236: \nonumber \\
237: & & -2\pi\hbar\int \left[ \int \left\{H,
238: \rho_{W}(Q,P,q,p,0)\right\}_{M}\ dq\ dp \right]^{2}\ dQ\ dP.
239: \label{t2q-a}
240: \end{eqnarray}
241: Further, using the definitions of the classical Poisson and quantum Moyal brackets and
242: assuming that initial classical and quantum distribution functions are identical and
243: separable, i.e.,
244: \begin{eqnarray}
245: \rho_{c}(Q,P,q,p,0)=\rho_{W}(Q,P,q,p,0)=\tilde{\rho}_{1}^{0}(Q,P)\tilde{\rho}_{2}^{0}(q,p),
246: \end{eqnarray}
247: we have
248: \begin{equation}
249: \frac{1}{\tau_{c,1}}=\frac{1}{\tau_{{\rm q},1}}=0,
250: \label{firstorder}
251: \end{equation}
252: \begin{equation}
253: \frac{1}{\tau_{c,2}^{2}}=2\pi\hbar \int
254: \left[\frac{\partial\tilde{\rho}_{1}^{0}(Q,P)}{\partial
255: P}\right]^{2}C(0,0)\ dQ \ dP,
256: \label{cla-2nd}
257: \end{equation}
258: and
259: \begin{eqnarray}
260: \frac{1}{\tau_{{\rm q},2}^{2}}&=&\frac{1}{\tau_{c,2}^{2}}
261: +2\pi\hbar\int \sum_{l_{1}\ne l_{2}\ge
262: 0}\frac{[\hbar/(2i)]^{(2l_{1}+2l_{2})}}{(2l_{1}+1)!(2l_{2}+1)!} \nonumber \\
263: &&\times
264: \frac{\partial^{(2l_{1}+1)}\tilde{\rho}_{1}^{0}(Q,P)}{\partial P^{(2l_{1}+1)}}
265: \frac{\partial^{(2l_{2}+1)}\tilde{\rho}_{1}^{0}(Q,P)}{\partial P^{(2l_{2}+1)}}C(l_{1},l_{2})\ dQ\ dP,
266: \label{quan-2nd}
267: \end{eqnarray}
268: where $C(l_{1},l_{2})$ is a correlation function given by
269: \begin{eqnarray}
270: C(l_{1},l_{2})&\equiv&
271: \left\langle \frac{\partial^{(2l_{1}+1)} V(Q,q)}{\partial Q^{(2l_{1}+1)}}
272: \frac{\partial^{(2l_{2}+1)} V(Q,q)}{\partial Q^{(2l_{2}+1)}}
273: \right\rangle_{\tilde{\rho}_{2}^{0}}\nonumber \\
274: &-&
275: \left\langle\frac{\partial^{(2l_{1}+1)} V(Q,q)}{\partial Q^{(2l_{1}+1)}}\right\rangle_{\tilde{\rho}_{2}^{0}}
276: \left\langle\frac{\partial^{(2l_{2}+1)} V(Q,q)}{\partial
277: Q^{(2l_{2}+1)}}\right\rangle_{\tilde{\rho}^{0}_{2}}.
278: \end{eqnarray}
279: Here $\langle\cdot\rangle_{\tilde{\rho}^{0}_{2}}$ denotes the ensemble average over
280: the zero-time ``bath distribution function" $\tilde{\rho}_{2}^{0}(q,p)$.
281: It is worth emphasizing that in our derivations we have used the same
282: initial state for the classical and quantum dynamics.
283:
284: Equation (\ref{firstorder}) shows that
285: the zero first-order linear entropy increase rate, i.e.,
286: $1/\tau_{{\rm q},1}$=0,
287: has a strict classical analog. Further, Eq. (\ref{cla-2nd})
288: indicates that
289: classical Liouville dynamics also predicts a second-order entropy production
290: rate $1/\tau_{c,2}^{2}$
291: that is the analog of the second-order quantum decoherence rate $1/\tau_{{\rm q},2}^{2}$.
292: Thus, we can identify two categories of early-time intrinsic decoherence dynamics:
293: {\it classical} if $\tau_{c,2}\approx \tau_{{\rm q},2}$,
294: and {\it nonclassical} if $\tau_{{\rm q},2}$ appreciably differs from $\tau_{c,2}$.
295:
296: To simplify Eqs. (\ref{cla-2nd}) and (\ref{quan-2nd}) we introduce the Fourier transform [denoted $
297: F(Q_{1},Q_{2})$]
298: of $\tilde{\rho}_{1}^{0}(Q,P)$, i.e.,
299: \begin{eqnarray}
300: F(Q_{1},Q_{2})&\equiv&
301: \int \tilde{\rho}_{1}^{0}(\overline{Q}
302: ,P)\exp\left[\frac{i\Delta QP}{\hbar}\right]\ dP,
303: \end{eqnarray}
304: where $\overline{Q}\equiv (Q_{1}+Q_{2})/2$ and $\Delta Q=Q_{1}-Q_{2}$.
305: We then obtain
306: \begin{equation}
307: \frac{1}{\tau_{c,2}^{2}}=\frac{1}{\hbar^{2}} \int
308: |F(Q_{1},Q_{2})|^{2} \Delta Q^{2} C(0,0)\ dQ_{1}\ dQ_{2},
309: \label{cla-2nd-n}
310: \end{equation}
311: and
312: \begin{eqnarray}
313: \frac{1}{\tau_{{\rm q},2}^{2}}=\frac{1}{\tau_{c,2}^{2}}
314: +\frac{1}{\hbar^{2}}\int |F(Q_{1},Q_{2})|^{2}\sum_{l_{1}\ne l_{2}\ge
315: 0}\frac{\Delta Q^{(2l_{1}+2l_{2}+2)}}{(2l_{1}+1)!(2l_{2}+1)!}\frac{1}{2^{(2l_{1}+2l_{2})}}
316: C(l_{1},l_{2})\ dQ_{1}\ dQ_{2}.
317: \label{quan-2nd-n}
318: \end{eqnarray}
319: Equations (\ref{cla-2nd-n}) and (\ref{quan-2nd-n}) are general results.
320: For the special case of $V_{12}(Q,q)=f(Q)g(q)$,
321: Eqs. (\ref{cla-2nd-n}) and
322: (\ref{quan-2nd-n}) can be rewritten in a simple and more enlightening form:
323: \begin{eqnarray}
324: \frac{1}{\tau_{c,2}^{2}}=
325: \frac{\langle g^{2}(q)\rangle_{\tilde{\rho}_{2}^{0}} -
326: \langle g(q)\rangle^{2}_{\tilde{\rho}_{2}^{0}}}{\hbar^{2}}\int|F(Q_{1},Q_{2})|^{2}
327: \Delta Q^{2}\left[\frac{d f(\overline{Q})}{dQ}\right]^{2}\ dQ_{1}\ dQ_{2},
328: \label{cfg}
329: \end{eqnarray}
330: and
331: \begin{eqnarray}
332: \frac{1}{\tau_{{\rm q},2}^{2}}=
333: \frac{
334: \langle g^{2}(q) \rangle_{\tilde{\rho}_{2}^{0}} -
335: \langle g(q)\rangle^{2}_{\tilde{\rho}_{2}^{0}}}{\hbar^{2}}\int
336: |F(Q_{1},Q_{2})|^{2}
337: \Delta Q^{2} \left[\frac{\Delta f(\overline{Q})}{\Delta Q}\right]^{2}\ dQ_{1}\ dQ_{2},
338: \label{qfg}
339: \end{eqnarray}
340: where $\Delta f(\overline{Q})/\Delta Q$ is the finite-difference function
341: \begin{eqnarray}
342: \frac{\Delta f(\overline{Q})}{\Delta Q}&\equiv &
343: \frac{f(\overline{Q}+\Delta Q/2)-f(\overline{Q}-\Delta Q/2)}{\Delta Q}
344: \nonumber \\
345: &=&\frac{f(Q_{1})-f(Q_{2})}{Q_{1}-Q_{2}}.
346: \end{eqnarray}
347: As a result: (1) If $f(Q)$ depends only
348: linearly or quadratically upon the coupling coordinate $Q$, a common approximation, then
349: $(1/\tau_{{\rm q},2}^{2}-1/\tau_{c,2}^{2})=0$ for any initial state. That is,
350: in this case there exists
351: perfect QCC in early-time dynamics of intrinsic decoherence, regardless of $\hbar$,
352: and irrespective of the potentials $V_{1}(Q)$ and $V_{2}(q)$.
353: (2) Even in the case of highly nonlinear $f(Q)$, as long as
354: $F(Q_{1},Q_{2})$ decays fast enough
355: with $|Q_{1}-Q_{2}|$
356: such
357: that $\Delta f/\Delta Q \approx df/dQ$, QCC would still be excellent.
358: The smaller the $\hbar$, the more rigorous
359: is this requirement. (3) If $\Delta f/\Delta Q$ differs significantly from
360: $ df/dQ$ over the $Q$-coordinate scale of the initial state,
361: quantum entropy production can be totally unrelated to classical entropy production.
362: Such cases of poor QCC are of fundamental interest, but are not the focus of this paper.
363:
364: The second-order perturbative treatment is most reliable for early-time
365: dynamics and for relatively weak decoherence.
366: The above results are particularly
367: significant for studies on the control of intrinsic decoherence,
368: where early-time
369: dynamics of weak decoherence is important.
370: In these circumstances
371: it is useful to understand the extent to which (quantum) intrinsic
372: decoherence is equivalent to classical entropy production, i.e. to
373: increasing $S_c(t)$. In particular, if there exists good
374: correspondence between classical and quantum decoherence dynamics,
375: then the essence of decoherence control is equivalent to the
376: suppression of classical entropy production, and various classical
377: tools may be considered to achieve decoherence control. If not,
378: then fully quantum tools are required.
379:
380: The above perturbation
381: results clearly demonstrate that quantum dynamics of intrinsic decoherence
382: has a direct analog in classical Liouville dynamics.
383: %Hence, despite the fact that there is no classical analog of quantum
384: %entanglement
385: %between two individual subsystems, it is still possible
386: %to simulate
387: %the dynamics of entanglement-induced intrinsic
388: %decoherence by
389: %classical correlations between classical subensembles.
390: This rather intriguing result motivates us to
391: computationally examine QCC in the dynamics of intrinsic decoherence
392: over all time scales.
393:
394: \section{Computational Results: Two Sample Cases}
395: \label{sample}
396:
397: To computationally examine QCC in the dynamics of
398: intrinsic decoherence, we consider
399: coupled-oscillator model systems with smooth Hamiltonians.
400: In all the model systems studied below, we choose
401: \begin{eqnarray}
402: V_{1}(Q)+V_{2}(q)
403: =\frac{\beta}{4}(Q^{4}+q^{4}),
404: \end{eqnarray}
405: where $\beta=0.01$.
406: Since $V_{1}(Q)$ and $V_{2}(q)$ have no simple harmonic terms, any
407: observed agreement between
408: classical and quantum behavior cannot be attributed to the similarity between
409: classical and quantum harmonic oscillator dynamics.
410: If the coupling potential $V_{12}(Q,q)$ is quadratic in both $Q$ and $q$, i.e.,
411: $V_{12}(Q,q)=\alpha Q^{2}q^{2}/2$, then
412: the resultant coupled-oscillator system is the well-known
413: quartic oscillator model \cite{eckhardt,gongpre,channel}.
414: Because this model is well-studied and can display
415: strongly chaotic (e.g.,
416: $\alpha=1.0$, $\beta=0.01$) or integrable (e.g.,
417: $\alpha=0.03$, $\beta=0.01$) dynamics, it is used in Sec.
418: \ref{classical-theory}
419: as an ideal model to
420: study QCC in intrinsic decoherence dynamics for
421: both integrable and chaotic cases.
422:
423: Our perturbation theory approach predicts good classical-quantum agreement
424: at short times for some potentials and initial conditions and poor
425: agreement for others. We examine both these cases computationally.
426:
427: It suffices to consider one case of poor agreement, since poor QCC at early times
428: invariably translates to similar behavior at later times.
429: Consider then $V_{12}(Q,q)$ to
430: be some highly nonlinear potential.
431: Computations of the quantum dynamics and thus the time dependence of $S_{{\rm q}}(t)$
432: are straightforward \cite{gongpre}. $S_{c}(t)$ is computed directly
433: using
434: Monte-Carlo simulations
435: with an importance sampling
436: technique (where the Monte-Carlo simulations are based upon Eq. (\ref{seq1}) below).
437: From the analytical results above we see that $V_{12}(Q,q)$
438: and the scale of the initial state play decisive roles in QCC in early-time
439: intrinsic decoherence dynamics. In particular, we expect poor
440: QCC if $V_{12}(Q,q)=f(Q)g(q)$ differs significantly from a linear or quadratic function of $Q$ such that
441: $\Delta f/\Delta Q$ differs significantly from
442: $(df/dQ)$ over the $Q$-coordinate scale (i.e., the support) of the initial state.
443: To confirm this computationally
444: we consider $f(Q)g(q)=\sin^{2}(10Q)q^{2}$, with the initial distribution functions of the two subsystems
445: given by
446: \begin{eqnarray}
447: \tilde{\rho}_{1}^{0}(Q,P)&=&\frac{1}{\pi\hbar}
448: \exp\left[-\frac{(Q-Q_{0})^{2}}{2\sigma_{Q}^{2}}-\frac{(P-P_{0})^{2}}{2\sigma_{P}^{2}}\right],
449: \nonumber
450: \\
451: \tilde{\rho}_{2}^{0}(q,p)&=&\frac{1}{\pi\hbar}\exp\left[-\frac{(q-q_{0})^{2}}{2\sigma_{q}^{2}}-
452: \frac{(p-p_{0})^{2}}{2\sigma_{p}^{2}}
453: \right].
454: \label{initialstate}
455: \end{eqnarray}
456: Here the dimensionless effective Planck constant is chosen to be
457: $\hbar=0.005$ throughout, except for one case in Sec.
458: \ref{summary},
459: and $\sigma_{Q}/25=25\cdot \sigma_{P}=\sqrt{\hbar/2}$, $\sigma_{q}=\sigma_{p}=\sqrt{\hbar/2}$,
460: $Q_{0}=0.5$, $P_{0}=0.5$, $q_{0}=0$, with $H(Q_{0}, P_{0}, q_{0}, p_{0})=0.24$.
461: Note that $\tilde{\rho}_{1}^{0}(Q,P)$ is strongly squeezed in $P$ and that this initial distribution
462: function is considerably delocalized in $Q$.
463: Further, since $|f(Q)|=|\sin^{2}(10Q)|\leq 1.0$,
464: $|\Delta Q|\cdot |df/dQ|$ can be much larger
465: than $|\Delta f(Q)|$. Thus, for this case
466: the perturbation result predicts that at early times
467: there can be substantial classical entropy production with insignificant
468: quantum decoherence. As shown in Fig. \ref{nonlinear1},
469: this is nicely confirmed by the numerical results of
470: $S_{{\rm q}}(t)$ and $S_{c}(t)$.
471: In particular, Fig. \ref{nonlinear1} shows that, at $t=1.0$
472: $S_{c}(t)$ (discrete points) is $\sim$ 0.9 while $S_{{\rm q}}(t)$ (solid line) is still less
473: than 0.2.
474: Evidently, QCC in this case
475: is indeed very poor from the very beginning.
476:
477: \begin{figure}[ht]
478: \begin{center}
479: \epsfig{file=bs-cfaster.ps,width=9.0cm}
480: \end{center}
481: \caption{A comparison between $S_{{\rm q}}(t)$ (dashed line) and $S_{c}(t)$ (discrete
482: circular points) in the first
483: sample case. The coupling potential is highly nonlinear such that at early times
484: classical entropy production is much faster than quantum entropy production.
485: See the text for details.
486: All variables
487: are in dimensionless units.}
488: \label{nonlinear1}
489: \end{figure}
490:
491:
492: There remains then the important question of the quantitative degree of QCC in circumstances where our
493: perturbative analysis predicts good short-time QCC. In particular,
494: it is important to investigate whether or not good QCC predicted perturbatively
495: remains for a considerable amount of time.
496: If so, then the perturbative treatment provides a useful
497: guide to our understanding of QCC in intrinsic decoherence dynamics.
498: If not, then our perturbative results make sense only for
499: extremely weak decoherence. Dramatically, our computational studies
500: strongly support our analytical perturbation results, even in the presence of
501: significant decoherence. For example,
502: consider the case, where the parameters for the initial state are
503: the same as in
504: the previous case (therefore the initial state is also much delocalized),
505: but the coupling potential is given by $V_{12}(Q,q)=Q^{2}\sin^{2}(q)$.
506: This coupling potential is highly nonlinear in $q$ but still quadratic in $Q$.
507: In accord with the second-order perturbation
508: results, such a coupling potential should still give rise to
509: good early-time QCC in the intrinsic decoherence dynamics of the first subsystem.
510: This is confirmed by the quantitative comparison between $S_{{\rm q}}(t)$ and $S_{c}(t)$
511: shown in Fig. \ref{nonlinear2}. More importantly,
512: Fig. \ref{nonlinear2} shows that outstanding
513: QCC remains even when both $S_{{\rm q}}(t)$ and $S_{c}(t)$ have increased
514: to close to their saturation value of unity.
515: Numerous other computational results (not shown) are consistent with the two cases shown here.
516:
517: \begin{figure}[ht]
518: \begin{center}
519: \epsfig{file=bs-q.ps,width=9.0cm}
520: \end{center}
521: \caption{ A comparison between $S_{{\rm q}}(t)$ (dashed line) and $S_{c}(t)$ (discrete
522: circular points) in the second
523: sample case. The coupling potential is highly nonlinear in terms of the position of the second
524: subsystem,
525: but is quadratic in terms of the position of the first subsystem, resulting
526: in excellent quantum-classical correspondence in intrinsic decoherence
527: dynamics even though the initial distribution function
528: of the first subsystem is considerably delocalized.
529: See the text for details. All variables
530: are in dimensionless units.}
531: \label{nonlinear2}
532: \end{figure}
533:
534:
535: These results
536: show the usefulness of the second-order
537: perturbation theory
538: in understanding QCC in intrinsic decoherence dynamics emanating from squeezed initial states.
539: Also of interest is intrinsic decoherence dynamics associated with
540: sufficiently localized initial states, which, in accord with the previous perturbation results,
541: should display excellent early-time QCC for any coupling potential
542: $V_{12}(Q,q)$.
543: We now
544: computationally examine QCC at later times for localized states as initial conditions
545: and explain the results in terms of a simple classical theory of intrinsic decoherence dynamics.
546:
547: \section{Localized Initial States}
548: \label{classical-theory}
549:
550: \subsection{Simple Classical Approach}
551:
552: Below we show that $S_{{\rm q}}(t)$ and $S_{c}(t)$ are often in
553: excellent agreement, over large time scales, for initially localized
554: states, significantly extending the perturbation theory result.
555: In doing so we compare the quantum $S_{{\rm q}}(t)$ with full
556: classical mechanics as well as with a simple classical theory derived
557: in this section. The latter provides further insight into the origins
558: of increasing $S_{c}(t)$.
559:
560: To derive the simplified classical result we first use Liouville's theorem to
561: reexpress $S_{c}(t)$ as
562: \begin{eqnarray}
563: S_{c}(t)&=& 1-2\pi\hbar\int
564: \int \rho_{c}(Q(t),P(t),q(t),p(t),t) \nonumber \\
565: & & \times
566: \rho_{c}(Q(t),P(t),q',p',t)\ dq'\ dp'\ dQ(t)\ dP(t)\ dq(t)\ dp(t) \nonumber \\
567: &=& 1-2\pi\hbar\int \int \rho_{c}(Q,P,q,p,0)
568: \rho_{c}(Q(t),P(t),q',p',t)\ dq'\ dp' \ dQ\ dP\ dq\ dp \nonumber \\
569: &=& 1-2\pi\hbar\int \int \rho_{c}(Q,P,q,p,0)
570: \rho_{c}(Q'',P'',q'',p'',0)\ dq'\ dp'\ dQ\ dP\ dq\ dp,
571: \label{seq1}
572: \end{eqnarray}
573: where $(Q(t),P(t),q(t),p(t))$ is the phase space location of the trajectory
574: emanating from $(Q,P,q,p)$ at $t=0$,
575: and $(Q'',P'',q'',p'')$ is the phase space location of the trajectory at time zero
576: if the classical trajectory is propagated backwards from $(Q(t),P(t),q',p')$.
577: Because the initial state $\rho_{c}(Q,P,q,p,0)$
578: is assumed highly localized in phase
579: space,
580: $(Q'',P'',q'',p'')$ must be very close to $(Q,P,q,p)$ in order for
581: the term $\rho_{c}(Q,P,q,p,0)
582: \rho_{c}(Q'',P'',q'',p'',0)$ in Eq. (\ref{seq1}) to be appreciable and thus
583: to contribute to $S_{c}(t)$.
584: Hence a convenient approximation can be made:
585: we assume that, at time $t$,
586: only those backward trajectories near $(Q(t),P(t),q(t),p(t))$ need be taken into account.
587: This means that we treat
588: $Q''-Q$, $P''-P$, $q''-q$, $p''-p$,
589: $\delta q' \equiv [q'-q(t)]$, and $\delta p'\equiv [p'-p(t)]$ as sufficiently small
590: such that
591: \begin{eqnarray}
592: Q''&\approx&Q+M_{13}(t)\delta q' + M_{14}(t)\delta p', \nonumber \\
593: P''&\approx&P+M_{23}(t)\delta q' + M_{24}(t)\delta p', \nonumber \\
594: q''&\approx &q+M_{33}(t)\delta q' + M_{34}(t)\delta p', \nonumber\\
595: p''&\approx &p+ M_{43}(t)\delta q' + M_{44}(t)\delta p',
596: \label{seq2}
597: \end{eqnarray}
598: where $M_{ij}$ ($i,j=1,2,3,4$) is
599: the stability matrix associated with the backward trajectories emanating from
600: $(Q(t),P(t),q(t),p(t))$:
601: \begin{eqnarray}
602: M_{ij}=\frac{\partial (Q, P, q, p)}{\partial (Q(t), P(t), q(t), p(t))}.
603: \label{seq3}
604: \end{eqnarray}
605: Although this approximation should be less reliable
606: for chaotic systems, we demonstrate below that, it is, nonetheless,
607: computationally
608: useful in both integrable and chaotic cases.
609:
610: For simplicity we consider below a specific example
611: in which $\tilde{\rho}_{1}^{0}(Q,P)$ and $\tilde{\rho}^{0}_{2}(q,p)$ are
612: symmetric Gaussian states given by
613: Eq. (\ref{initialstate}) with $\sigma_{Q}=\sigma_{P}=$$\sigma_{q}=\sigma_{p}\equiv \sigma=\sqrt{\hbar/2}$.
614: Other distributions can be considered in an analogous fashion.
615: Substituting Eqs. (\ref{initialstate}) and (\ref{seq2}) into
616: Eq. (\ref{seq1}) and evaluating the integrals, we obtain
617: \begin{equation}
618: S_{c}(t)=1-\frac{1}{2}\left\langle \frac{\exp\left[
619: \frac{U^{2}X^{2}+V^{2}Y^{2}-2UVZ}{2\sigma^{2}(X^{2}Y^{2}-Z^{2})}\right]}{\sqrt{
620: X^{2}Y^{2}-Z^{2}}} \right\rangle_{\rho_{c}'},
621: \label{seq4}
622: \end{equation}
623: where
624: \begin{eqnarray}
625: X&=&M_{13}^{2}+M_{23}^{2}+M_{33}^{2}+M_{43}^{2}, \nonumber \\
626: Y&=& M_{14}^{2}+M_{24}^{2}+M_{34}^{2}+M_{44}^{2}, \nonumber \\
627: Z&=& M_{13}M_{14}+M_{23}M_{24}+M_{33}M_{34}+M_{43}M_{44}, \nonumber \\
628: U&=& (Q-Q_{0})M_{14}+
629: (P- P_{0})M_{24}+
630: (q- q_{0})M_{34}+
631: (p-p_{0})M_{44}, \nonumber \\
632: V&=& (Q- Q_{0})M_{13}+
633: (P- P_{0})M_{23}+
634: (q-q_{0})M_{33}+
635: (p- p_{0})M_{43},
636: \label{seq5}
637: \end{eqnarray}
638: and
639: \begin{eqnarray}
640: \rho_{c}'(Q,P,q,p)=4\pi^{2}\hbar^{2}[\tilde{\rho}_{1}^{0}(Q,P)]^{2}[\tilde{\rho}_{2}^{0}(q,p)]^{2}.
641: \end{eqnarray}
642: Equations (\ref{seq4}) and (\ref{seq5}) indicate that the classical dynamics of
643: intrinsic decoherence
644: is closely related to the classical stability matrix elements averaged over a rescaled
645: initial distribution
646: function.
647: This interesting connection
648: provides insight into
649: a variety of interesting aspects of quantum intrinsic decoherence
650: dynamics \cite{furuya,miller,arul,fujisaki,prosen}.
651: For example,
652: for chaotic dynamics in which classical trajectories are highly unstable
653: and therefore in which
654: $|M_{ij}|$ increases rapidly,
655: $S_{c}(t)$ should increase much faster than for the case of integrable
656: dynamics. This observation can,
657: with the assumption that there is fairly good QCC in intrinsic decoherence
658: dynamics, directly explain
659: previous results on quantum signatures of classical chaos
660: in the dynamics of quantum entanglement
661: \cite{furuya}. Further,
662: because Eqs. (\ref{seq4}) and (\ref{seq5}) are expressed in terms of classical
663: stability matrices,
664: characteristics of the time
665: dependence of $S_{c}(t)$ and therefore of $S_{{\rm q}}(t)$ can be easily related to
666: the time and space fluctuations in the instability of classical trajectories.
667:
668: \subsection{Computational Results: Localized Initial States}
669:
670: \begin{figure}[ht]
671: \begin{center}
672: \epsfig{file=Inte.ps,width=9.0cm}
673: \end{center}
674: \caption{ A comparison between
675: $S_{{\rm q}}(t)$ (dashed line) and the approximate $S_{c}(t)$ (solid line) calculated
676: from Eq. (\ref{seq4})
677: for the quartic oscillator model in the case of integrable dynamics ($\alpha=0.03,
678: \beta=0.01$). The initial state is given by Eq. (\ref{initialstate}), with
679: $\sigma_{P}=\sigma_{Q}=\sigma_{p}=\sigma_{q}=\sqrt{\hbar/2}$,
680: $\hbar=0.005$,
681: $Q_{0}=0.4$, $P_{0}=0.5$, $q_{0}=0.6$,
682: and $H(Q_{0}, P_{0}, q_{0}, p_{0})=0.24$.
683: Full classical results based upon Eq. (\ref{seq1}) are represented by discrete
684: circular points.
685: All variables
686: are in dimensionless units.
687: }
688: \label{Fig-inte}
689: \end{figure}
690:
691: \begin{figure}[ht]
692: \begin{center}
693: \epsfig{file=chaos.ps,width=9.0cm}
694: \end{center}
695: \caption{Same as in Fig. \ref{Fig-inte} except for strongly chaotic dynamics ($\alpha=1.0, \beta=0.01$).}
696: \label{Fig-chaotic}
697: \end{figure}
698:
699:
700: Consider then QCC over large time scales for localized initial states
701: for both integrable and chaotic cases. To do so
702: we examine
703: the quartic oscillator model as well as results where
704: the coupling potential is replaced by
705: the nonlinear potential $V_{12}(Q,q)=0.5Q^{2}q^{2}+Q^{4}q^{2}$.
706: The initial states are chosen to be localized initial states, and both
707: the full classical dynamics and the
708: approximate time dependence of $S_{c}(t)$
709: in Eq. (\ref{seq4}) are compared to the quantum result.
710: Specifically, we realize the ensemble average in
711: Eq. (\ref{seq4}) by Monte-Carlo simulations, using
712: only $2\times 10^{4}$ sampling
713: classical trajectories from which the stability matrix elements $M_{ij}$ are evaluated.
714: The initial Gaussian states are chosen symmetric
715: with $\sigma=\sqrt{\hbar/2}=0.05$ and are sufficiently localized so that
716: Eq. (\ref{seq2}) should be a valid approximation.
717: Full classical results for $S_{c}(t)$ (represented again by discrete circular points), which
718: are much more demanding computationally,
719: are also provided below.
720:
721:
722: Figures \ref{Fig-inte} and \ref{Fig-chaotic}
723: compare results for $S_{c}(t)$ obtained from
724: Eq. (\ref{seq4}) and from exact classical results
725: with $S_{{\rm q}}(t)$ (dashed line)
726: for integrable and chaotic
727: dynamics in the quartic oscillator model, respectively.
728: A number of observations are in order.
729: First, it is clear that in both cases the approximate $S_{c}(t)$
730: are in excellent agreement with the full classical results, confirming
731: the utility of the simple model [Eq. (\ref{seq4})].
732: Second, both $S_{c}(t)$ and $S_{{\rm q}}(t)$ are seen,
733: in the chaotic case, to relax faster towards $1.0$
734: than they do in the integrable case.
735: Third, the oscillation amplitudes of $S_{{\rm q}}(t)$
736: in the chaotic case are much smaller than that in the integrable case.
737: Hence, the fast relaxation and small-amplitude oscillations
738: of $S_{{\rm q}}(t)$ shown in Fig. \ref{Fig-chaotic}
739: may be regarded as fingerprints of the underlying classical chaos.
740: Finally, and most importantly, the entire time dependence, including
741: oscillations in Fig. \ref{Fig-inte}
742: and Fig. \ref{Fig-chaotic} of $S_{{\rm q}}(t)$
743: are beautifully captured by both the exact and the approximate $S_{c}(t)$.
744: It should also be noted that
745: the QCC time scale shown in Fig. \ref{Fig-chaotic}
746: is appreciably longer than is the QCC break time $t_{b}\sim 5.0$ for the same $\hbar$,
747: obtained
748: by quantitatively comparing the structure of the
749: classical and quantum distribution functions
750: \cite{gongthesis}.
751: This can be understood by the fact
752: that $S_{c}(t)$ [or $S_{{\rm q}}(t)$]
753: describes the reduced distribution functions $\tilde{\rho}_{c}(Q,P,t)$ [or $\tilde{\rho}_{W}(Q,P,t)$]
754: which is insensitive to the fine structure of $\rho_{c}(Q,P,q,p,t)$
755: [or $\rho_{W}(Q,P,q,p,t)$].
756:
757:
758: Calculations for many other initial states confirm
759: that the QCC results shown in
760: Figs. \ref{Fig-inte} and \ref{Fig-chaotic} are typical,
761: indicating that (a) QCC is essentially exact over large time scales and (b)
762: the simple classical
763: theory of intrinsic decoherence dynamics introduced above
764: provides a useful approximation to
765: the exact results.
766:
767: \begin{figure}[ht]
768: \begin{center}
769: \epsfig{file=channel.ps,width=9.0cm}
770: \end{center}
771: \caption{Same as in Fig. \ref{Fig-inte} except for
772: strongly chaotic dynamics ($\alpha=1.0
773: , \beta=0.01$), and for a special initial Gaussian state, with
774: $Q_{0}=P_{0}=0$, $q_{0}=0.6$ and $H(Q_{0}, P_{0}, q_{0}, p_{0})=0.24$.}
775: \label{Fig-chan}
776: \end{figure}
777:
778:
779: Figure \ref{Fig-chan} shows one case, however,
780: where the approximate $S_{c}(t)$ and exact classical or quantum
781: results differ quantitatively. Here
782: the system is still the quartic oscillator model with $\alpha=1.0$, $\beta=0.01$, but
783: with an initial state of special type. In particular, both the initial average position $Q_{0}$
784: and the initial average momentum $P_{0}$ are set to zero. Initial states of this type
785: are called channel states \cite{channel}, and
786: effectively give rise to very weak coupling
787: between the two subsystems over a considerably large time scale.
788: Indeed, since at short times $df/d\overline{Q} \approx \Delta
789: f(\overline{Q})/\Delta Q\approx 0$ for $\overline{Q}\approx Q_{0}=0$,
790: one obtains from
791: Eqs. (\ref{cfg}) and (\ref{qfg}) that
792: $1/\tau_{c,2}^{2}=1/\tau_{2,{\rm q}}^{2}\approx 0$.
793: Therefore the early-time intrinsic decoherence rate should be small, as
794: seen in Fig. \ref{Fig-chan}, although the underlying classical dynamics is strongly chaotic.
795: For this reason one expects that
796: the dynamical behavior of $S_{c}(t)$ and $S_{{\rm q}}(t)$ should
797: differ from previous cases.
798: As shown in Fig. \ref{Fig-chan}, in this case both
799: $S_{{\rm q}}(t)$ and $S_{c}(t)$ increase in a step-wise fashion, distinctly
800: different from that in Figs. \ref{Fig-inte}
801: and \ref{Fig-chaotic}. Agreement between them remains excellent.
802: However, the approximate
803: $S_{c}(t)$ misses some of the important structure.
804:
805: \begin{figure}[ht]
806: \begin{center}
807: \epsfig{file=chaos-nonlinear.ps,width=9.0cm}
808: \end{center}
809: \caption{Same as in Fig. \ref{Fig-inte} except for
810: a modified
811: quartic oscillator model in which
812: $V_{12}(Q,q)=0.5Q^{2}q^{2}
813: +Q^{4}q^{2}$.
814: }
815: \label{Fig-nonlinear}
816: \end{figure}
817:
818:
819: In Fig. \ref{Fig-nonlinear} we show the QCC result for a simple variant of the quartic oscillator model,
820: i.e., $V_{12}(Q,q)=0.5 Q^{2}q^{2}+Q^{4}q^{2}$, a coupling potential that is
821: neither linear nor quadratic.
822: As seen in Fig. \ref{Fig-nonlinear},
823: even with such nonlinear coupling
824: $S_{c}(t)$ and $S_{{\rm q}}(t)$ are in excellent agreement
825: over large time scales. This emphasizes the fact
826: that the good QCC results
827: observed in the quartic oscillator model are not due to the fact
828: that the coupling potential
829: therein is quadratic.
830: Hence,
831: we conclude that for initially localized states,
832: our simple classical theory of
833: intrinsic decoherence dynamics [see Eq. (\ref{seq4})] is generally useful
834: in describing intrinsic decoherence dynamics in smooth Hamiltonian systems.
835:
836:
837:
838: \section{discussion and summary}
839: \label{summary}
840:
841: Quantum entanglement between individual subsystems
842: has no classical analog.
843: Nevertheless, as shown in this work, the quantum dynamics of quantum entanglement,
844: as manifest in the quantum dynamics of intrinsic
845: decoherence, does have a classical analog in
846: classical Liouville dynamics describing classical correlations between
847: classical subensembles.
848: Hence it is useful to isolate the conditions under which
849: there is good QCC in intrinsic decoherence dynamics. This is done analytically, for early-time dynamics
850: and for weak decoherence,
851: by a second-order
852: perturbative theory. Interestingly, as demonstrated by our computational studies
853: in Sec. \ref{sample} and Sec. \ref{classical-theory}, the physical picture of QCC
854: afforded by the perturbative treatment can be still very useful even when the time scale under
855: investigation is relatively long and the degree of intrinsic decoherence is significant.
856: In particular,
857: under the circumstances where there is good early-time QCC,
858: classical Liouville dynamics can provide a simple means of understanding
859: different aspects of intrinsic decoherence dynamics, for relatively large time scales and
860: for both integrable and chaotic dynamics.
861: Further,
862: we have derived an approximate but very simple
863: classical theory of linear entropy production of
864: intrinsic decoherence dynamics associated with localized initial states,
865: and shown that the rate of entropy production is closely
866: related to the stability properties of classical trajectories.
867:
868: Clearly, the linear entropy is just one of many possible representation-independent
869: measures of intrinsic decoherence,
870: and $S_{{\rm q}}(t)\approx S_{c}(t)$ does not mean that
871: the quantum dynamics is equivalent to the corresponding classical Liouville dynamics.
872: For example, if the saturation value of the linear entropy
873: in the long time limit is of particular interest, then the measures
874: $1/[1-S_{c}(t)]$ and $1/[1-S_{{\rm q}}(t)]$ (which gives the number of orthogonal states that are incoherently
875: populated) should be more useful in describing QCC.
876: Indeed,
877: our results in Figs. \ref{Fig-chaotic} and \ref{Fig-nonlinear} suggest that
878: as time increases one has $1/[1-S_{c}(t)]>>1/[1-S_{{\rm q}}(t)]$.
879: This is consistent with our previous observation \cite{gongprl} that,
880: decoherence can dramatically improve QCC, but
881: even strong decoherence does not necessarily suffice to ensure that quantum entropy
882: production is the same as classical entropy production.
883:
884: \begin{figure}[ht]
885: \begin{center}
886: \epsfig{file=bigger-hbar.ps,width=9.0cm}
887: \end{center}
888: \caption{A comparison between $S_{{\rm q}}(t)$ (dashed line) and $S_{c}(t)$ (discrete
889: circular points)
890: for strongly chaotic dynamics
891: of the quartic oscillaor model ($\alpha=1.0
892: , \beta=0.01$) and for $\hbar=0.05$.
893: The initial state is given by Eq. (\ref{initialstate}), with
894: $\sigma_{P}=\sigma_{Q}=\sigma_{p}=\sigma_{q}=\sqrt{\hbar/2}$,
895: $Q_{0}=0.4$, $P_{0}=0.5$, $q_{0}=0.6$,
896: and $H(Q_{0}, P_{0}, q_{0}, p_{0})=0.24$.
897: All variables
898: are in dimensionless units.
899: }
900: \label{bighbar}
901: \end{figure}
902:
903:
904: It should also be pointed out that
905: the model quantum systems studied in this paper are still far from
906: the semiclassical regime.
907: This is indicated, in the chaotic case of the quartic oscillator
908: model for example, by the fact that the QCC break time
909: is relatively short compared to the time scale that we examined. Correspondence
910: will worsen quantitatively with increasing $\hbar$, although,
911: as discussed above, $\hbar$
912: is far from the only factor influencing the quality of the QCC.
913: However, we note that for localized initial states the qualitative features of the time dependence of
914: classical and quantum linear entropy may remain similar to one another
915: with much larger effective Planck
916: constants. For example, Fig. \ref{bighbar} displays fairly
917: good QCC between $S_{{\rm q}}(t)$ and $S_{c}(t)$,
918: in the chaotic case of the quartic oscillator model, with $\hbar=0.05$ and
919: with an initial
920: symmetric Gaussian state.
921:
922:
923: A number of interesting extensions of this work are under
924: consideration. First, it seems straightforward but necessary to consider cases in which the
925: coupling potential depends upon both position and
926: momentum. Second, we
927: propose to further investigate the role of the dynamics of
928: the subsystems in addition to that of the coupling potential (e.g.,
929: the dynamics in coupled Morse oscillator systems). Third,
930: it is interesting to study QCC
931: in intrinsic decoherence dynamics in terms of the decay of off-diagonal
932: density matrix elements. Such studies are ongoing, with preliminary studies \cite{han}
933: indicating that comparing the time dependence of off-diagonal
934: density matrix elements to its direct classical analog \cite{gongprl} will
935: provide deeper insights into QCC in
936: intrinsic decoherence dynamics.
937:
938: To summarize, we have shown that classical dynamics can be very useful in describing
939: intrinsic decoherence dynamics in smooth Hamiltonian systems. In particular, we have identified
940: conditions under which excellent
941: quantum-classical correspondence
942: in the early-time dynamics of intrinsic decoherence is
943: possible via a second-order perturbative treatment,
944: have presented a simple classical theory
945: of intrinsic decoherence dynamics emanating from localized initial states, and
946: have provided supporting computational results.
947: The hope is that by extending this study to high-dimensional Hamiltonian systems,
948: we may use purely classical approaches to describe (at least qualitatively)
949: the dynamics of quantum entanglement or
950: intrinsic decoherence in polyatomic molecular systems.
951:
952: {\bf Acknowledgments:} This work was supported by the U.S. Office of
953: Naval Research
954: and by the Natural Sciences and Engineering
955: Research Council of Canada.
956:
957: \begin{thebibliography}{100}
958: \bibitem{zurekreview} W.H. Zurek, preprint quant-ph/0105127.
959: \bibitem{batista} V.S. Batista and P. Brumer, \prl{\bf 89}, 143201 (2002).
960: \bibitem{qcbook} M.A. Nielsen and I.L. Chuang, {\it Quantum Computation
961: and Quantum Information} (Cambridge University Press, Cambridge, 2000).
962: \bibitem{brumerreview}
963: M. Shapiro and P. Brumer, Adv. Atom. Mol. and Opt.
964: Phys., {\bf 42}, 287 (2000).
965: \bibitem{ricebook}
966: S.A. Rice and M. Zhao, {\it Optical Control of Molecular Dynamics} (John
967: Wiley, New York, 2000).
968: \bibitem{brumerbook}M. Shapiro and P. Brumer, {\it Principles of
969: the Quantum Control of Molecular Processes} (John Wiley, New York, 2003).
970: \bibitem{wilkie} J. Wilkie and P. Brumer, Phys. Rev. A {\bf 55}, 27 (1997);
971: Phys. Rev. A {\bf 55}, 43 (1997).
972: \bibitem{gongprl} J. Gong and P. Brumer, Phys. Rev. Lett. {\bf 90},
973: 050402 (2003);
974: J. Gong and P. Brumer, quant-ph/0212106.
975: \bibitem{bmiller}
976: H. Wang, M. Thoss, K.L. Sorge, R.X. Gim\'{e}nez, and
977: W.H. Miller, \jcp{\bf 114}, 2562 (2001);
978: F. Grossmann, \jcp{\bf 103}, 3696 (1995);
979: A.M.O. de Almeida, preprint quant-ph/0208094.
980: \bibitem{linear-entropy}
981: P.C. Lichtner and J.J. Griffin, \prl{\bf 37}, 1521 (1976);
982: W.H. Zurek, S. Habib and J.P. Paz, \prl{\bf 70}, 1187 (1993);
983: X-P. Jiang and P. Brumer, Chem. Phys. Lett. {\bf 208}, 179 (1993);
984: a. Isar, A. Sandulescu, and W. Scheid,
985: \pre{\bf 60}, 6371 (1999);
986: G. Manfredi and M. R. Feix, \pre{\bf 62}, 4665 (2000);
987: J.N. Bandyopadhyay and A. Lakshminarayan, \prl{\bf 89}, 060402 (2002).
988: \bibitem{moyal} J.E. Moyal, Proc. Cambridge Phil. Soc. {\bf 45}, 99 (1949).
989: \bibitem{kim} J.I. Kim, M.C. Nemes, A.F.R. Toledo Piza, and H.E. Borges, \prl{\bf
990: 77}, 207
991: (1996).
992: \bibitem{duan} L.M. Duan and G.C. Guo, \pra{\bf 56},
993: 4466 (1997).
994: \bibitem{bacon}D. Bacon, D.A. Lidar, and K.B. Whaley, \pra{\bf 60},
995: 1944 (1999).
996: \bibitem{eckhardt} B. Eckhardt, G. Hose, and E. Pollak, \pra{\bf 39}, 3776
997: (1989).
998: \bibitem{gongpre} J. Gong and P. Brumer, \pre{\bf 60}, 1643 (1999).
999: \bibitem{channel}
1000: M.S. Santhanam, V.B. Sheorey, and A. Lakshminarayan, \pre{\bf 57}, 345 (1998).
1001: \bibitem{furuya}K. Furuya, M.C. Nemes, and G.Q. Pellegrino, \prl{\bf 80}, 5524 (1998).
1002: \bibitem{miller}P.A. Miller and S. Sarkar, \pre{\bf 60}, 1542 (1999).
1003: \bibitem{arul} A. Lakshminarayan, \pre{\bf 64}, 036207 (2001).
1004: \bibitem{fujisaki} A. Tanaka, H. Fujisaki, and T. Miyadera, \pre{\bf 66}, 045201 (2002).
1005: \bibitem{prosen}M. Znidaric and T. Prosen, J. Phys. A {\bf 36}, 2463 (2003).
1006: %\bibitem{eckhardt} B. Eckhardt, G. Hose, and E. Pollak, \pra{\bf 39}, 3776 (1989).
1007: %\bibitem{gongpre} J. Gong and P. Brumer, \pre{\bf 60}, 1643 (1999).
1008: %\bibitem{channel}
1009: %M.S. Santhanam, V.B. Sheorey, and A. Lakshminarayan, \pre{\bf 57}, 345 (1998).
1010: %\bibitem{kim} J.I. Kim, M.C. Nemes, A.F.R. Toledo Piza, and H.E. Borges, \prl{\bf 77}, 207
1011: %(1996).
1012: \bibitem{gongthesis} J. Gong, Ph.D thesis (unpublished), University of Toronto, 2001.
1013: \bibitem{han} H. Han, J. Gong, and P. Brumer, to be published.
1014: \end{thebibliography}
1015:
1016:
1017:
1018: \end{document}
1019:
1020: