1: \documentclass[12pt]{article}
2: \usepackage{graphicx}
3: \usepackage{amsfonts}
4:
5:
6: \begin{document}
7:
8:
9:
10: \begin{titlepage}
11:
12:
13:
14: \hfill{Preprint {\bf SB/F/05-336}} \hrule \vskip 2.5cm
15:
16: \centerline{\bf Zeno and Anti Zeno effect for a two level system
17: in a squeezed bath }
18: \vskip 1.5cm
19: \centerline{D.F. Mundarain and J. Stephany} \vskip 4mm
20:
21: \centerline{{\it Departamento de F\'{\i}sica, Universidad Sim\'on
22: Bol\'{\i}var,}} \centerline{\it Apartado 89000, Caracas 1080A,
23: Venezuela.}
24:
25: \begin{abstract}
26: We discuss the appearance of Zeno (QZE) or anti-Zeno (QAE) effect
27: in an exponentially decaying system. We consider the quantum
28: dynamics of a continuously monitored two level system interacting
29: with a squeezed bath. We find that the behavior of the system
30: depends critically on the way in which the squeezed bath is
31: prepared. For specific choices of the squeezing phase the system
32: shows Zeno or anti-Zeno effect in conditions for which it would
33: decay exponentially if no measurements were done. This result
34: allows for a clear interpretation in terms of the equivalent spin
35: system interacting with a fictitious magnetic field.
36:
37: \end{abstract}
38:
39:
40: \vskip 2cm \hrule
41: \bigskip
42: \centerline{\bf UNIVERSIDAD SIM{\'O}N BOL{\'I}VAR} \vfill
43: \end{titlepage}
44:
45:
46: \section {Introduction}
47: The suppression or modification of the rate of quantum transitions
48: in a system, due to successive measurements is known as the
49: quantum Zeno effect (QZE) \cite{sud,chiu,peres}. This term has
50: been applied both to the elimination of the induced transitions as
51: in the case of Rabi oscillations on a two level system, or to the
52: reduction of the decay rate on an unstable system. The first
53: situation was experimentally achieved in 1990 \cite{ita} and the
54: second one in 2001 \cite{fis}.
55:
56: An interesting issue in relation with the QZE is wether it appears
57: or not in exponentially decaying systems. In their article of 1977
58: Chiu, Sudarshan and Misra \cite{chiu} show that in general, an
59: unstable system has three decaying regimens. For short time
60: intervals, $t \leq T_1$ or very large ones $t\geq T_2$, with
61: $T_1$ and $T_2$ some time scales, the system depart from the
62: exponentially decaying behavior shown for $T_1 \leq t \leq T_2$.
63: They also predict that frequent measurements led to the QZE if
64: the time interval between successive measurements is shorter than
65: $T_1$. In the experiment of Ref. \cite{fis} for example, for short
66: times, the decay rate of the system is remarkably slower than
67: exponential. This could lead to think that QZE only occurs when
68: the time between measurements is short enough to exploit the
69: departure from the exponential decay law. Nevertheless in a recent
70: article Koshino and Shimizu \cite{kos1} predicted the appearance
71: of QZE even for systems with an exponential decay law in the case
72: when the detector has a finite window of sensibility. For this
73: case they analyze explicitly the interaction between the quantum
74: system and the detector and interpreted the changes induced by the
75: interaction as the effect of the measurement. They refer
76: \cite{kos2} to this analysis as the dynamical formalism as
77: opposed to the conventional formalism where the measurements are
78: taken as projections consistent with the quantum collapse
79: postulate of von Neumann.
80:
81: For a closed system the theoretical description of the measurement
82: in terms of the projection postulate predicts a complete Zeno
83: effect, that is the freezing of the quantum system in the initial
84: state. For such system with a hamiltonian ${\it H}$, the
85: evolution is determined by the Schr{\"o}dinger equation
86: \begin{equation}
87: \frac{\partial}{\partial t} |\psi(t) \rangle = \frac{1}{i \hbar}
88: {\it H} |\psi(t) \rangle
89: \end{equation}
90:
91: If the observable ${\it A}$ to be measured has eigenvalues $a_m$
92: and supposing that at $t=0$ the system is in the eigenstate
93: $|a_n\rangle$, the probability of obtaining the result $a_n$ for a
94: short time interval $\Delta t\geq 0$ is given by,
95: \begin{equation}
96: P_n(\Delta t) = \left( 1 - \frac{\Delta t^2}{\hbar^2} \Delta_n^2
97: {\it H}\right)
98: \end{equation}
99: where
100: \begin{equation}
101: \Delta_n^2 {\it H} = \langle a_n |{\it H}^2|a_n\rangle -\langle
102: a_n |{\it H}|a_n\rangle^2
103: \end{equation}
104: If one considers $S$ successive projective measurements
105: separated by the same interval $\Delta t$ the probability of
106: obtaining in each case the same result $a_n$ is:
107: \begin{equation}
108: P_n(S,\Delta t) = \left( 1 - \frac{\Delta t^2}{\hbar^2} \Delta_n^2
109: {\it H}\right)^S
110: \end{equation}
111: In the limit of very frequent measurements \cite{Braginsky}, that
112: is when $ S >> 1 , \ \ \ \ S \Delta t \rightarrow t$ the
113: probability of measuring $a_n$ every time is
114: \begin{equation}
115: P_n^{(cm)} (t) = \lim_{S \rightarrow \infty} P_n(S, t/ S) =
116: \lim_{S \rightarrow \infty}\left( 1 - \frac{t^2}{S^2 \hbar^2}
117: \Delta_n^2 {\it H}\right)^S = 1
118: \end{equation}
119: which corresponds to a complete Zeno effect.
120:
121: For open systems in contact with the environment some limitations
122: affect the appearance of the QZE even if projective measurements
123: are being done. For time intervals which are greater than the
124: correlation time of the bath, the evolution may be described in
125: terms of the density matrix by a master equation of the Liouville
126: type,
127: \begin{equation}
128: \label{Liuv} \frac{\partial\rho}{\partial t} = {\it L } \left\{\rho\right\} .
129: \end{equation}
130: with ${\it L } \left\{\rho\right\}$ some appropriate operator
131: depending on $\rho$. Then, for a short time interval $\Delta t$,
132: the density operator is given in terms of its initial value by
133: \begin{equation}
134: \rho( \Delta t ) = \rho(0) + {\it L}\left\{ \rho(0)\right\} \Delta t
135: \end{equation}
136: If the initial state is $\rho(0) = |a_n\rangle \langle a_n| $ the
137: probability of measuring $a_n$ in $S$ consecutive measurements
138: separated by time intervals $\Delta t$ is,
139: \begin{equation}
140: P_n(S,\Delta t) =( 1 + \langle a_n|{\it L}\left\{
141: \rho(0)\right\}|a_n\rangle \Delta t)^S\simeq (\exp \left\{\langle
142: a_n|{\it L}\left\{ \rho(0)\right\}|a_n\rangle\ \Delta t\right\})^S
143: \end{equation}
144: In the limit $ S >> 1 , \ \ \ \ S \Delta t \rightarrow t$, of
145: very frequent measurements one obtains,
146: \begin{equation}\label{ec13}
147: P_n^{(cm)}(t)= \exp \left\{\langle a_n|{\it L}\left\{ \rho(0)\right\}|a_n\rangle\ \
148: t\right\}
149: \end{equation}
150: Here the freezing of the initial condition for continuous
151: measurements is achieved only if
152: \begin{displaymath}
153: \langle a_n|{\it L}\left\{\rho(0)\right\}|a_n\rangle = 0
154: \end{displaymath}
155: This illustrates the fact that in general, both the the
156: intrinsical properties of the system and the characteristics of
157: the measurement affect the possibility of displaying the quantum
158: Zeno effect.
159:
160: A related issue that we have to consider comes from the
161: observation that for an unstable quantum system the probability
162: of obtaining a specific result in a measurement may increase,
163: decrease or even oscillate in time as the result of its
164: undisturbed evolution. Decay rates may also be affected by
165: measurements done at particular instants of time, an effect which
166: has in principle nothing to do with the QZE. This suggests that
167: the interaction of the system with a non trivial electromagnetic
168: bath may modify the decay rates even for an exponentially
169: decaying system. In this paper we show that such mechanism can be
170: actually used to induce QZE or QAE in a two level system. For this
171: system interacting with a squeezed bath QZE or QAE may appear when
172: measuring the fictitious spin along a specific direction depending
173: on the relative phase of the squeezing and the chosen direction.
174: This may be interpreted as an effect of the orientation induced on
175: the fictitious spin by the fictitious magnetic field defined by
176: the quadratic fluctuations of the true electromagnetic field.
177:
178: \section{The two level system in a squeezed bath}
179: In the rotating wave approximation the hamiltonian which better
180: describes the atom-field interaction has the following structure,
181: \cite{gar,scully}:
182: \begin{equation}
183: {\it H} = \sum_{\bf k} \hbar \nu_k a_{\bf k}^{\dagger} a_{\bf k} +
184: \frac{1}{2} \hbar \omega \sigma_z+ \hbar \sum_{\bf k} g_{\bf k}
185: \left( \sigma_{+} a_{\bf k}+a_{\bf k}^{\dagger} \sigma_{-} \right)
186: \label{hamil}
187: \end{equation}
188: where $g_{\bf k}$ are the atom-field couplings constants, $ a_{\bf
189: k}$ and $a_{\bf k}^{\dagger}$ are the creation and annihilation
190: operators of the multimodal field and $\sigma_+$ and $\sigma_-$
191: are the ladder operators
192: \begin{equation}
193: \sigma_{+} =\left(
194: \begin{array}{cc}
195: 0&1\\
196: 0&0
197: \end{array}
198: \right)\qquad \sigma_{-} =\left(
199: \begin{array}{cc}
200: 0&0\\
201: 1&0
202: \end{array}
203: \right)\ \ ,
204: \end{equation}
205: with $\sigma_x$, $\sigma_y$ and $\sigma_z$ are the Pauli matrices,
206: \begin{equation}
207: \sigma_x =\left(
208: \begin{array}{cc}
209: 0&1\\
210: 1&0
211: \end{array}
212: \right)\qquad \sigma_y =\left(
213: \begin{array}{cc}
214: 0&-i\\
215: i&0
216: \end{array}
217: \right) \qquad \sigma_z =\left(
218: \begin{array}{cc}
219: 1&0\\
220: 0&-1
221: \end{array}
222: \right)\ \ .
223: \end{equation}
224: If the field is prepared in a broadband squeezed vacuum state
225: characterized by $\xi=r e^{i\phi}$ it was demonstrated that,
226: \cite{gar,scully}:
227: \begin{displaymath}
228: \langle a_{\bf k}\rangle =\langle a_{\bf k}^{\dagger}\rangle=0
229: \end{displaymath}
230: \begin{displaymath}
231: \langle a_{\bf k}^{\dagger} a_{\bf k'}\rangle \delta_{\bf k k'} =
232: N \delta_{\bf k k'}
233: \end{displaymath}
234: \begin{displaymath}
235: \langle a_{\bf k} a_{\bf k'}^{\dagger}\rangle= \delta_ {\bf k k'}=
236: (N+1) \delta_{\bf k k'}
237: \end{displaymath}
238: \begin{displaymath}
239: \langle a_{\bf k} a_{\bf k'} \rangle= - e^{i \phi} \cosh ( r)
240: \sinh(r) \delta_{\bf k', 2 k_0 - k} = e^{i \phi} M \delta_{\bf k',
241: 2 k_0 - k}
242: \end{displaymath}
243: \begin{equation}\label{ec4}
244: \langle a_{\bf k}^{\dagger} a_{\bf k'}^{\dagger} \rangle=- e^{-i
245: \phi} \cosh ( r) \sinh(r) \delta_{\bf k', 2 k_0 - k} = e^{-i \phi}
246: M \delta_{\bf k', 2 k_0 - k}\ \ ,
247: \end{equation}
248:
249: \noindent where $N= \sinh^2 ( r)$, $M=\sqrt{N(N+1)}$. Here ${\bf
250: k_0}$ is the wave number associated to the resonant frequency of
251: the squeezing device. In the interaction picture the master
252: equation for this system takes the form of Eq. (\ref{Liuv}) with,
253: \begin{eqnarray}\label{em1}
254: L\{\rho\} =&\frac{1}{2}\gamma \left( N+1\right) \left( 2 \sigma_{-}
255: { \rho} \sigma_{+} - \sigma_{+} \sigma_{-} { \rho} - {\rho}
256: \sigma_{+} \sigma_{-}
257: \right)\nonumber\\
258: &\frac{1}{2} \gamma N \left( 2 \sigma_{+} { \rho} \sigma_{-} -
259: \sigma_{-} \sigma_{+} {\rho} - { \rho} \sigma_{-} \sigma_{+} \right)\nonumber\\
260: &- \gamma M e^{ i \phi} \sigma_{+} { \rho} \sigma_{+} -\gamma M
261: e^{ -i \phi} \sigma_{-} { \rho} \sigma_{-}\ \ .
262: \end{eqnarray}
263: Here $\gamma$ is the decay constant of the system in the vacuum.
264: This equation may be rewritten using Bloch's representation for
265: the two level system density matrix in the form,
266: \begin{equation}\label{dm1}
267: \rho = \frac{1}{2} \left( 1 + \rho_{x} \sigma_{x}+\rho_{y}
268: \sigma_{y} + \rho_{z} \sigma_{z} \right)\ \ .
269: \end{equation}
270:
271: Using Eqs. (\ref{em1},\ref{dm1}), the master equation (\ref{Liuv})
272: takes the form,
273: \begin{eqnarray}
274: \frac{\partial \rho}{\partial t} =& -\frac{1}{2} \gamma \left(
275: N+1\right) \left( (1+\rho_z) \sigma_{z} +\frac{1}{2}
276: \rho_x \sigma_{x} +\frac{1}{2} {\rho_y} \sigma_{y} \right)\nonumber\\
277: & +\frac{1}{2} \gamma N \left( (1-\rho_z) \sigma_{z} -\frac{1}{2}
278: \rho_x \sigma_{x} -\frac{1}{2} {\rho_y} \sigma_{y} \right)\nonumber\\
279: & -\frac{1}{2} \gamma M \rho_{x} ( \cos(\phi)\sigma_{x} -\sin
280: (\phi) \sigma_{y}) \nonumber\\
281: &+\frac{1}{2} \gamma M \rho_{y} ( \sin(\phi)\sigma_{x} +\cos
282: (\phi) \sigma_{y})
283: \end{eqnarray}
284: This is equivalent to the following differential equations for
285: $(\rho_x, \rho_y, \rho_z)$:
286: \begin{displaymath}
287: \dot{\rho_x} = -\gamma \left(N+1/2+M \cos (\phi)\right) \rho_x +
288: \gamma M \sin(\phi) \rho_y
289: \end{displaymath}
290: \begin{displaymath}
291: \dot{\rho_y} = -\gamma \left(N+1/2-M \cos (\phi)\right) \rho_y +
292: \gamma M \sin(\phi) \rho_x
293: \end{displaymath}
294: \begin{equation}
295: \dot{\rho_z} = -\gamma \left(2N+1/2 \right) \rho_z -\gamma
296: \end{equation}
297: The solutions of these equations are given by,
298:
299: \begin{eqnarray}\label{ec1222}
300: \rho_x(t) =& \left( \rho_x (0) \sin^2(\phi/2) +\rho_y(0)
301: \sin(\phi/2) \cos( \phi/2) \right) e^{ -\gamma (N+1/2-M)\, t}
302: \nonumber\\
303: &+\left( \rho_x (0) \cos^2(\phi/2) -\rho_y(0) \sin(\phi/2) \cos(
304: \phi/2) \right) e^{ -\gamma (N+1/2+M)\, t}
305: \end{eqnarray}
306: \begin{eqnarray}
307: \rho_y(t) =& \left( \rho_y (0) \cos^2(\phi/2) +\rho_x(0)
308: \sin(\phi/2) \cos( \phi/2) \right) e^{ -\gamma (N+1/2-M)\, t}
309: \nonumber\\
310: &+\left( \rho_y (0) \sin^2(\phi/2) -\rho_x(0) \sin(\phi/2) \cos(
311: \phi/2) \right) e^{ -\gamma (N+1/2+M)\, t}
312: \end{eqnarray}
313:
314: \begin{equation}
315: \rho_z(t) = \rho_z(0) e^{-\gamma (2 N+1) t} + \frac{1}{2 N +1}
316: \left( e^{-\gamma (2 N+1) t}-1\right)
317: \end{equation}
318: From these expressions one can read the dependence of the decay
319: rates of the system on the phase $\phi$ of the squeezing. In
320: particular, for $\phi=0$, $\phi=\pi$ or for the critical angles $
321: \phi_z = 2 \arctan(-\rho_y(0) / \rho_x(0))$ or $\phi_{AZ}= 2
322: \arctan (\rho_x(0)/ \rho_y(0) )$, the system has a purely
323: exponential behavior with the decay rates presented in Table \ref{Table1}.
324: \vskip.3cm
325: \begin{table}[t]
326: \begin{center}
327: \mbox{\begin{tabular}{|r|r|r|} \hline
328: \ & $\rho_x (t)/\rho_x(0)$ & $\rho_y (t)/\rho_y(0)$ \\
329: \hline
330: $\phi=0\quad$ & $e^{-\gamma (N+1/2+M) t }$& $e^{-\gamma (N+1/2-M) t }$ \\
331: $\phi=\pi \quad$ & $e^{-\gamma (N+1/2-M) t}$ & $e^{-\gamma (N+1/2+M) t }$ \\
332: $\phi=\arctan \left(-\frac{\rho_y(0)}{\rho_x(0)}\right)$ &
333: $e^{-\gamma (N+1/2+M) t}$ & $e^{-\gamma (N+1/2+M)t}$ \\
334: $\phi=\arctan \left(\frac{\rho_x(0)}{\rho_y(0)}\right)$ &
335: $e^{-\gamma (N+1/2-M) t}$ & $e^{-\gamma (N+1/2-M)t}$ \\ \hline
336: \end{tabular}}
337: \end {center}
338: \caption{Decay rates for critical angles}
339: \label{Table1}
340: \end{table}
341: In Fig.(\ref{fig1}) we show the dependence of $\rho_x(t) = \langle
342: \sigma_x \rangle$ with the phase as given by Eq. (\ref{ec1222}).
343: In particular the exponential decay for the preferred values of
344: the phase may be observed.
345:
346: \section{The origin of the critical angles} Before discussing the effect of the
347: measurements in the evolution of the two level system let us first
348: explore the properties of the fictitious magnetic field associated
349: to the squeezed state in order to justify the decay rates for
350: the two critical angles appearing in Table 1.
351:
352: Consider the atomic part of the Hamiltonian (\ref{hamil}). In
353: terms of the Pauli matrices it takes the form,
354: \begin{equation}
355: {\it H}_{Atomic} = \frac{1}{2} \hbar \omega \sigma_z+ \frac{1}{2}
356: \hbar \sigma_x \sum_{\bf k} g_{\bf k} \left( a_{\bf k}+a_{\bf
357: k}^{\dagger}\right) + \frac{1}{2} \hbar \sigma_y \sum_{\bf k} i
358: g_{\bf k} \left( a_{\bf k}- a_{\bf k}^{\dagger}\right)\ .
359: \end{equation}
360: This can be rewritten in the form
361: \begin{equation}
362: {\it H}_{Atomic} = - \gamma_0 \, {\bf B \cdot S}\ \ .
363: \end{equation}
364: where $\gamma_0$ is an arbitrary constant with dimensions of
365: charge divided by mass, ${\bf S}$ is the fictitious spin
366: associated to the two level system and ${\bf B}$ is the quantum
367: fictitious magnetic field with components,
368: \begin{equation}
369: B_x = -\frac{1}{\gamma_0} \sum_{\bf k} g_{\bf k} \left( a_{\bf
370: k}+a_{\bf k}^{\dagger}\right)
371: \end{equation}
372: \begin{equation}
373: B_y = - \frac{1}{\gamma_0} \sum_{\bf k} i g_{\bf k} \left( a_{\bf
374: k}- a_{\bf k}^{\dagger}\right)
375: \end{equation}
376: \begin{equation}
377: B_z = - \frac{\omega}{\gamma_0}\ \ .
378: \end{equation}
379: Clearly $<B_x>=0$ and $<B_y>=0$. For the quadratic fluctuations
380: the result is,
381: \begin{eqnarray}
382: <B_x^2>&=&\frac{\Gamma}{4}\left(\sinh^2(r)+\cosh^2(r)-2\cos(\theta)\sinh(r)\cosh(r)\right)\\
383: &=&\Gamma \left( N +\frac{1}{2}- M \cos(\phi) \right)\nonumber\\
384: <B_y^2>&=&\frac{\Gamma}{4}\left(\sinh^2(r)+\cosh^2(r)+2\cos(\theta)\sinh(r)\cosh(r)\right)\\
385: &=& \Gamma \left( N +\frac{1}{2}+ M \cos(\phi) \right)\nonumber
386: \end{eqnarray}
387: where
388: \begin{displaymath}
389: \Gamma = \frac{1}{2 \gamma_0^2} \sum_{\bf k}| g_{\bf k}|^2 .
390: \end{displaymath}
391: Here $\Gamma=\sum_k |g_k|^2$ is taken to be finite, which means
392: that only a finite subset of the modes in the bath is coupled
393: effectively to the system. For $<B^2>$ we have,
394: \begin{eqnarray}
395: <B^2>=\frac{\Gamma}{2}\left(\sinh^2(r)+\cosh^2(r)\right) =
396: \frac{\Gamma}{2}\left( 2 N +1 \right) ,
397: \end{eqnarray}
398: which does not depend on $\phi$.
399:
400: These fluctuations may be represented in phase space as an ellipse
401: whose axis are rotated by an angle $\phi/2$. As is illustrated
402: in Fig. (\ref{fig17}), the semi-axis have magnitudes $\Gamma
403: \left( N +1/2+ M \right)$ and $\Gamma \left( N +1/2- M \right)$.
404:
405: Comparing with the results of the previous section, we can observe,
406: that for zero phase the decay rate for $\rho_x = \langle
407: \sigma_x\rangle$ is proportional to the fluctuations of the
408: fictitious magnetic field component $B_y$ and the decay rate for
409: $\rho_y = \langle \sigma_y\rangle$ is proportional to the
410: fluctuations of $B_x$. Also, the decay rate for $\rho_z$ is
411: proportional to $<B^2>$. In general, for other values of the
412: phase, the component $(\rho_x, \rho_y)$ of Bloch's vector
413: orthogonal to the major semi-axis of the phase space ellipse used
414: to represent the magnetic field fluctuations, has a decay rate
415: proportional to $\Gamma \left( N +1/2+ M \right)$ and the
416: component orthogonal to the minor semi-axis has a decay rate
417: proportional to $\Gamma \left( N +1/2- M \right)$. The phase
418: $\phi_Z=2 \arctan \left(-\rho_y(0)/ \rho_x(0)\right)$ defines a
419: critical value which corresponds to the case when initially the
420: Bloch vector is orthogonal to the major semi-axis. For this value
421: $\rho_x$ and $ \rho_y$ decay with the maximum rate. The
422: complementary case occurs for $\phi_{AZ}= 2 \arctan ( \rho_x(0)/
423: \rho_y(0))$ in which case $\rho_x$ and $ \rho_y$ decay with the
424: minimum allowed value of the decay rate.
425:
426: The fact that the decay rates for $\rho_x$ and $ \rho_y$ coincides
427: in both cases is a consequence of the coupled dynamics of these
428: two variables. But if one measures $\sigma_x$, the dynamics
429: disentangles and one would expect that the decay rate for $
430: \rho_x$ results proportional to the fluctuations of $B_y$ only
431: , as in the cases $\phi=0$ or $\phi=\pi$ when there is no
432: coupling at all. Then, we expect the Zeno effect to occur for
433: $\phi_Z$ and the anti-Zeno effect to occur for $\phi_{AZ}$ due to
434: the factor $\cos (\phi)$ that appears in the fluctuation of
435: $B_y$.
436: \section{Zeno and anti-Zeno effect}
437:
438: Let us now consider explicitly the effect of repeated
439: measurements of the observable $\sigma_x$ in the evolution of
440: the system prepared in a state defined by the initial values of
441: $(\rho_x(0),\rho_y(0), \rho_z(0))$. We suppose that the time
442: interval between measurements is very short, but still much
443: greater than the correlation time of the squeezed bath
444: \cite{GarPC1987}. Then we may describe the evolution of the system
445: by means of a master equation of the form (\ref{Liuv}). In our
446: analysis we take in fact the correlation time of the squeezed
447: bath to be zero which corresponds to broadband squeezing. For
448: considerations on the finite bandwidth effects se Ref.
449: \cite{GarPC1987,ParG1988,TanE1998}. On the experimental side
450: squeezing with a bandwidth of up to 1GHz has been reported
451: \cite{Cro1988,MacY1988,HirKI2005}.
452:
453: The probability that in a very large succession of measurements,
454: the result obtained in all of them is the eigenvalue $+1$
455: associated to the eigenstate $|+\rangle_x$ is given by,
456: \begin{equation}
457: \label{ec217} P_{+}^{(cm)}(t) = \frac{(1+\rho_x(0))}{2} \exp
458: \left\{ _x \langle+|{\it L}\left\{ \rho_1 \right\}|+\rangle_x t
459: \right\}
460: \end{equation}
461: where $\rho_1$ is the collapsed density matrix after the
462: measurements and is given by
463: \begin{equation}
464: \rho_1=|+\rangle_{x\, x} \langle +| \ \ \ , \ \ \ |+\rangle_x =
465: \frac{1}{\sqrt{2}}\left( |+\rangle + |-\rangle \right) .
466: \end{equation}
467: The probability of Eq. (\ref{ec217}) is obtained by multiplying
468: the probability corresponding to the first measurement and the
469: probability obtained in Eq. (\ref{ec13}) which is valid for the
470: following measurements. If the system is initially in the state
471: $|+\rangle_x$ then $\rho_x(0) =1$, $\rho_y(0) =0$ and $
472: \rho_z(0)=0$, in which case Eq. (\ref{ec217}) is a particular case
473: of Eq. (\ref{ec13}).
474:
475: One can show that for the squeezed bath,
476: \begin{equation}
477: _x \langle+|{\it L}\left\{ \rho_1\right\}|
478: +\rangle_x =- \frac{\gamma}{2}\left( N +\frac{1}{2} + M \cos (\phi)
479: \right).
480: \end{equation}
481: In this case Eq. (\ref{ec217}) reduces to,
482: \begin{equation}\label{ec218}
483: P_{+}^{(cm)}(t,\phi) =\frac{(1+ \rho_x(0))}{2} \exp \left\{ -
484: \frac{\gamma}{2}\left( N +\frac{1}{2} + M \cos (\phi) \right) t
485: \right\}
486: \end{equation}
487: We should compare this expression with the probability of
488: measuring the eigenvalue $+ 1$ by performing an unique measurement
489: at time $ t$
490: \begin{equation}\label{ec33}
491: P_+(t,\phi) =\frac{(1+\rho_x(t))}{2} .
492: \end{equation}
493:
494: In Fig.(\ref{fig2}) we show that the probability $P_+ (t,\phi=0)$
495: decays exponentially to the value $1/2$, that is, for $t
496: \rightarrow \infty$ we have the same probability to measure any of
497: the two eigenvalues. On the other hand the probability $P_+^{(cm)}
498: (t,\phi=0)$, i.e the probability to obtain the same value $+1$ in
499: all the measurements, decays exponentially to zero. In the same
500: figure we can see that the probability $P_+^{(cm)} (t,\phi=0)$ is
501: smaller than the probability $P_+ (t,\phi=0)$ for all $t$. Note
502: that, the probability of obtaining the result $+1$ in the last
503: measurement independently of the previous results is of course
504: greater than the probability of obtaining the value $+1$ in all
505: the measurements. Furthermore, if the evolution of the observable
506: $\sigma_x$ is not affected by the measurements, the probability to
507: obtain $+1$ in the last measurement independently of the results
508: of the previous measurements is equal to the probability to obtain
509: $+1$ at time $t$ if no other measurement has been done previously.
510: The result shown in Fig.(\ref{fig2}) suggest that in fact for
511: $\phi=0$ the evolution of the observable $\sigma_x$ is not
512: affected by the measurements.
513:
514: Changing the phase it is possible to obtain a completely
515: different result. In Fig (\ref{fig3}) we show that there exists
516: a time interval for which $P_+^{(cm)} (t,\phi_Z)$ is greater than
517: the $P_+ (t,\phi_Z)$. The natural explanation for this, comes from
518: the fact that in this case the measurements do modify the dynamics
519: of the observable.
520:
521: To study quantitatively this effect it is necessary to work out
522: the changes in the master equation related to the continuous
523: monitoring. If we have the system described by $\rho$ and perform
524: measurements of $\sigma_x$ the new density matrix is given by,
525: \begin{equation}
526: \rho^{\prime} = P \rho P + (1-P)\rho (1-P)
527: \end{equation}
528: where $P=|+\rangle_{xx}\langle+|$ is the projector to the
529: eigenvector of $\sigma_x$ with eigenvalue $+ 1$ and $(1-P)$ is the
530: projector to the eigenvector of $\sigma_x$ with eigenvalue $- 1$.
531:
532: Between consecutive measurements the free evolution is determined
533: by the free master equation. By considering the free master
534: equation and the collapse in the same expression, it is shown
535: that after the first measurement the master equation with
536: continuous measurements takes the form,
537: \begin{equation}
538: \frac{\partial \rho}{\partial t} = P {\it L}(\rho)P +(1- P) {\it
539: L}(\rho)(1-P)
540: \end{equation}
541:
542: Let us now focus in the mean value of the measured observable
543: \mbox{$\langle \sigma_x\rangle = \rho_x(t)$}. The corresponding
544: probabilities may be computed using Eq. (\ref{ec33}). In terms of
545: Bloch's vector, the master equation with continuous measurements
546: for the two level system in the squeezed bath is given by the
547: following equations,
548: \begin{eqnarray}
549: \dot{\rho_x}& = & -\gamma( N+1/2+M \cos (\phi) )\rho_x+ \gamma M \sin (\phi) \rho_y\nonumber\\
550: \dot{\rho_y}& = &0\nonumber\\
551: \dot{\rho_z}& = &0\label{ec39}
552: \end{eqnarray}
553: Since after the first measurement the values of $ \rho_y$ and
554: $\rho_z$ collapse to zero, the solutions for this system are given
555: by ,
556: \begin{equation}\label{ec332}
557: \rho_x(t) = \rho_x(0) \exp ( -\gamma( N+1/2+M \cos (\phi) )t)
558: \end{equation}
559:
560: \begin{equation}
561: \rho_y(t) = 0 \qquad \rho_z(t) =0 .
562: \end{equation}
563: As we can see from Eq. (\ref{ec332}) in presence of very frequent
564: measurements the decay rate of $\rho_x$ is proportional to the
565: quantum fluctuation of $B_y$.
566:
567: In Fig. (\ref{fig4}), it can be shown the evolution of
568: $<\sigma_x>$ for $\phi=0$ with measurements and without
569: measurements. We observe that the evolution is not affected by
570: the measurements. This agrees with the usual assumption that for
571: an unstable system with exponential decay Zeno effect is not
572: observable. In Fig. (\ref{fig5}) for $\phi=\phi_Z$, one can
573: appreciate the reduction of the decay rate when comparing with
574: the not disturbed case. For the phase $\phi=\phi_{AZ}$ the rate of
575: decaying grows and we have Anti-Zeno effect.
576:
577: \section{Indirect measurements}
578:
579: When indirect measurements are being done, the master equation
580: with continuous monitoring of $\sigma_x$ takes the form
581: \cite{Braginsky},
582: \begin{equation}
583: \frac{\partial \rho}{\partial t} = {\it L}(\rho) - \frac{1}{T_0} [
584: \, \sigma_x\, ,\, [\sigma_x,\rho]\,]
585: \end{equation}
586: where $T_0$ is the coupling constant between the measuring
587: apparatus and the system. Writing this equation in terms of
588: Bloch`s vector for the two level system in a squeezed bath we
589: have,
590: \begin{displaymath}
591: \dot{\rho_x} = -\gamma \left(N+1/2+M \cos (\phi)\right) \rho_x +
592: \gamma M \sin(\phi) \rho_y
593: \end{displaymath}
594: \begin{displaymath}
595: \dot{\rho_y} = -\gamma \left(N+1/2-M \cos (\phi)\right) \rho_y +
596: \gamma M \sin(\phi) \rho_x -\frac{4}{T_0} \rho_y
597: \end{displaymath}
598: \begin{equation}\label{ec43}
599: \dot{\rho_z} = -\gamma \left(2N+1/2 \right) \rho_z
600: -\gamma-\frac{4}{T_0} \rho_z
601: \end{equation}
602: The limit $T_0 \rightarrow \infty$ corresponds to no measurement
603: being done. For $T_0 \rightarrow 0$ equations (\ref{ec43})
604: transform into equations (\ref{ec39}). Then, for these kind of
605: measurements one obtains similar effects that those observed in
606: the previous section for the projective measurements.
607: \section{Conclusion}
608: We have presented an explicit example of a system where the
609: appearance of Zeno (or anti-Zeno) effect may be induced in a
610: regime for which it would decay exponentially if no measurements
611: were done. Working with a two level system in squeezed
612: electromagnetic bath, we found that these effects may be induced
613: by choosing adequately the phase of the squeezing of the bath.
614: This result is interpreted as the natural result of the
615: interaction of the equivalent spin system with the fluctuating
616: fictitious magnetic field.
617:
618:
619:
620: \section{Acknowledgments} This work was supported
621: by Did-Usb grant Gid-30 and by Fonacit grant G-2001000712.
622:
623:
624: \begin{thebibliography}{99}
625:
626: \bibitem{sud} B. Misra and E. C. Sudarshan, {\it J.Math.Phys. } 18,
627: 756 (1977).
628: \bibitem{chiu}
629: C. B. Chiu, E. C. Sudarshan and B. Misra, {\it Phys. Rev} D 16,
630: 520 (1977).
631: \bibitem{peres}A. Peres, {\it Am.J.Phys} 48, 931 (1980).
632: \bibitem{ita}
633: W. M. Itano, D. J. Heinzen, J. J. Bollinger and D. J. Wineland,
634: {\it Phys. Rev.} A 41, 2295 (1990).
635: \bibitem{fis}
636: M. C. Fischer, B. Gutierrez-Medina and M. G. Raizen, {\it Phys.
637: Rev. Lett} 87, 040402 (2001).
638: \bibitem{kos1}
639: K. Koshino and A. Shimizu, {\it Phys. Rev. Lett.} 92, 30401
640: (2004).
641: \bibitem{kos2}
642: K. Koshino, {\it Phys. Rev.} A 71, 034104 (2005).
643: \bibitem{Braginsky}V. B. Braginsky and F.Y. Khalili, {\it Quantum Measurement}, Cambridge University Press, 1992.
644: \bibitem{gar}
645: C. W. Gardiner, {\it Phys. Rev. Lett.} 56, 1917 (1986)
646: \bibitem{scully}
647: M. O. Scully and M. Suhail Zubairy, {\it Quantum Optics}.
648: Cambridge University Press (1997).
649: \bibitem{GarPC1987}C. W. Gardiner, A. S. Parkins and M. J. Collett, {\it J.Opt. Soc. Am} B4, 1683 (1987).
650: \bibitem{ParG1988}A. S. Parkins and C. W. Gardiner, {\it Phys Rev} A37, 3867 (1988).
651: \bibitem{TanE1998} R.Tanas and T.El-Shahat, {\it Acta.Phys.Slov. } 48, 301 (1998).
652: \bibitem{Cro1988}D. D. Crouch, {\it Phys Rev} A38, 508 (1988).
653: \bibitem{MacY1988}S. Machida and Y. Yamamoto, {\it Phys. Rev. Lett} 60, 792 (1988).
654: \bibitem{HirKI2005}T. Hirano, K. Kotani, T. Ishibashi, S. Okude and T. Kuwamoto,
655: {\it Opt. Lett.} 30, 1722 (2005).
656:
657:
658:
659: \end{thebibliography}
660: \newpage
661: \begin{figure}[ht]
662: \includegraphics[scale=0.7]{figzenon1.ps}
663: \caption{$\rho_x(t,\phi)$, $\rho_x(0)=0.2$ ,
664: $\rho_y(0)=-\sqrt{1-\rho_x(0)^2}$, $\rho_z(0)=0$, $N=1$, $\gamma =1$}\label{fig1}
665: \end{figure}
666:
667: \newpage
668: \begin{figure}[ht]
669: \centerline{\includegraphics[scale=0.7]{fig17.eps}}
670: \caption{Fluctuations of the fictitious magnetic field.}
671: \label{fig17}
672: \end{figure}
673:
674: \newpage
675: \begin{figure}[ht]
676: \includegraphics[scale=0.6,angle=-90]{grafico2.ps}
677: \caption{Solid line: $P_+ (t)$. Dashed line: $P_+^{(cm)}
678: (t,\phi=0)$. $\rho_x(0)=0.5$, $\rho_y(0)=-\sqrt{1-\rho_x(0)^2}$, $\rho_z(0)=0$,
679: $N=1$, $\gamma=1$ }\label{fig2}
680: \end{figure}
681:
682: \newpage
683: \begin{figure}[ht]
684: \includegraphics[scale=0.6,angle=-90]{grafico3.ps}
685: \caption{Solid line: $P_+ (t)$. Dashed line: $P_+^{(cm)}
686: (t,\phi_Z)$. $\rho_x(0) =0.5$, $\rho_y(0)=-\sqrt{1-\rho_x(0)^2}$, $\rho_z(0)=0$,
687: $N=1$, $\gamma=1$
688: }\label{fig3}
689: \end{figure}
690:
691:
692: \newpage
693: \begin{figure}[ht]
694: \includegraphics[scale=0.6,angle=-90]{grafico4.ps}
695: \caption{Solid line: $\rho_x(t)$ undisturbed. Dashed line:
696: $\rho_x(t)$ with measurements. $\rho_x(0)=0.5$,
697: $\rho_y(0)=-\sqrt{1-\rho_x(0)^2}$, $\rho_z(0)=0$, $\phi=0 $, $N=1$, $\gamma=1$
698: }\label{fig4}
699: \end{figure}
700:
701: \newpage
702: \begin{figure}[ht]
703: \includegraphics[scale=0.6,angle=-90]{grafico5.ps}
704: \caption{Solid line: $\rho_x(t)$ undisturbed. Dashed line:
705: $\rho_x(t)$ with measurements. $\rho_x(0) =0.5$,
706: $\rho_y(0)=-\sqrt{1-\rho_x(0)^2}$, $\rho_z(0)=0$, $\phi=\phi_Z$, $N=1$,
707: $\gamma=1$ }\label{fig5}
708: \end{figure}
709:
710: \newpage
711: \begin{figure}[ht]
712: \includegraphics[scale=0.6,angle=-90]{grafico6.ps}
713: \caption{Solid line: $\rho_x(t)$ undisturbed. Dashed line:
714: $\rho_x(t)$ with measurements. $\rho_x(0)=0.5$,
715: $\rho_y(0)=-\sqrt{1-\rho_x(0)^2}$, $\rho_z(0)=0$, $\phi=\phi_{AZ}$, $N=1$,
716: $\gamma=1$ }\label{fig6}
717: \end{figure}
718: \end{document}
719: