0804.2669/lz.tex
1: %2345678901234567890123456789012345678901234567890123456789012345678901234567890
2: %        1         2         3         4         5         6         7         
3: \documentclass[pre,onecolumn,floatfix,preprintnumbers,amsmath,
4: amssymb]{revtex4}
5: 
6: \bibliographystyle{apsrev}
7: \usepackage{graphicx}
8: \usepackage{amsmath}
9: \usepackage{natbib}
10: 
11: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
12: \begin{document}
13: \title{Landau-Zener transitions in a semiconductor quantum dot}
14: \author{G.\ E.\ Murgida}
15: \author{D.\ A.\ Wisniacki}
16: \author{P.\ I.\ Tamborenea}
17: \affiliation{Departamento de F\'{\i}sica ``J.\ J.\ Giambiagi",
18: Universidad de Buenos Aires, Ciudad Universitaria, Pab.\ I,
19: C1428EHA Buenos Aires, Argentina}
20: \date{\today}
21: 
22: \begin{abstract}
23: We study the transitions between neighboring energy levels in a quasi-one-dimensional
24: semiconductor quantum dot with two interacting electrons in it, when it is subject to
25: a linearly time-dependent electric field.
26: We analyze the applicability of simple two-level Landau-Zener model to describe the evolution of the probability amplitudes in this realistic system.
27: We show that the Landau-Zener model works very well when it is viewed in the adibatic
28: basis, but it is not as robust in the diabatic basis. 
29: 
\end{abstract}
30: 
31: \maketitle
32: 
33: \section{Introduction}
34: 
35: The Landau-Zener (LZ) model describes in a simplified way the ubiquitous situation
36: of avoided crossings of energy levels in quantum mechanics \cite{zen}.
37: This happens in numerous areas of physics such as quantum optics, atomic
38: physics, nuclear physics, etc.
39: In spite of its simplified nature, the LZ model often captures the essential
40: features of avoided level crossings in realistic systems and the list of 
41: its applications continues to grow.
42: 
43: Recently, we have proposed a generic quantum control method based on the navigation 
44: of the energy spectrum.
45: The navigation of the spectrum is done varying a control parameter diabatically
46: and adiabatically.
47: The possibility of traveling through a complex spectrum depends crucially on
48: the nature of the energy level crossings.
49: Our method requires that the system behave locally (at avoided crossings) like
50: a LZ model, in the sense that complete diabatic and adiabatic transitions be 
51: possible.
52: So, the first step in the application of this method to a realistic system must
53: be a careful examination of the validity of this condition.
54: Note that in a realistic system the interaction between levels is often intricate
55: and the possibility that the LZ model worked has been discussed \cite{bul-dod-kus,
56: san-ver-wis}. 
57: Recently, we successfully applied this control strategy to a quantum dot system 
58: and the isomerization of a LiCN molecule.
59: In the present paper we analyse in detail the issue of the applicability of the
60: LZ model to the avoided crossings of the two-electron quantum-dot system studied
61: in Refs.\ \cite{mur-wis-tam,wis-mur-tam}.
62: 
63: Quantum dots are prime candidates to study the ideas and proposals of quantum control,
64: given their flexible and tunable properties.
65: In this paper, we continue the study of a quasi-one-dimensional double-dot system
66: with two interacting electrons.
67: Because of the one-dimensionality, this system is well suited to investigate
68: new methods of quantum control, and at the same time it incorporates the
69: important aspect of the interparticle interaction treated exactly.
70: We remark that the presence of interactions between particles is crucial in the
71: new science of quantum information processing.
72: 
73: The article is organized as follows: 
74: In order to make the work as self-contained as possible, in the next section  
75: we review the well-known Landau-Zener Model.
76: In Section \ref{sec:system} we describe our system and in the following section
77: we present the results concerning the applicability of the LZ model.
78: We conclude with some final remarks. 
79: 
80: 
81: 
82: \section{Landau-Zener Model}
83: \label{sec:LZ_model}
84: 
85: The LZ model \cite{zen} attempts to describe the universal situation of two levels 
86: interacting at an avoided crossing when a parameter $\lambda$ in the Hamiltonian 
87: is varied. The model consists of a two-level system described by
88: a parameter-dependent Hamiltonian, which expressed in the diabatic basis reads
89: 
90: %
91: \begin{equation}
92:  H=\left[\begin{array}{cc}
93:         \varepsilon_1(\lambda)  & \delta \\
94:         \delta                  & \varepsilon_2(\lambda)
95:      \end{array}\right],
96: \label{eq:hamiltonian}
97: \end{equation}
98: where  $\delta$ is a constant while $\varepsilon_1$ and $\varepsilon_2$ are linear 
99: functions of $\lambda$: 
100: $\varepsilon_1 = \overline{\varepsilon} + \alpha_1 (\lambda - \overline{\lambda})$, 
101: $\varepsilon_2 = \overline{\varepsilon} + \alpha_2 (\lambda - \overline{\lambda})$.
102: The center of the avoided crossing is located at $\overline{\lambda}$ and 
103: $\overline{\varepsilon}$ [see Fig.\ \ref{fig:avoided}].
104: % 
105: The diabatic basis, $|1\rangle$ and $|2\rangle$, are parameter-independent 
106: eigenstates of the Hamiltonian Eq.\ (\ref{eq:hamiltonian}) with $\delta = 0$. 
107: 
108: 
109: The eigenenergies  $E_1(\lambda)$ and $E_2(\lambda)$ of the Hamiltonian
110: (\ref{eq:hamiltonian}) are two hyperbolas (the adiabatic curves) as shown in 
111: Fig.\ \ref{fig:avoided}.
112: The eigenstates associated to those energies are the so-called adiabatic
113: states, which we denote $|\phi_1(\lambda)\rangle$ and 
114: $|\phi_2(\lambda)\rangle$.
115: %
116: The asymptotes to the energy hyperbolas are the diabatic straight 
117: lines $\varepsilon_1(\lambda)$ and $\varepsilon_2(\lambda)$. 
118: %
119: The shortest distance between the hyperbolas is $2 \delta$. 
120: 
121: In his seminal paper, Zener considered the parameter $\lambda$ as a linear
122: function of time and obtained the asymptotic probabilities of transitions
123: between the diabatic states in this time-dependent problem.
124: Assuming that the state $|1 \rangle $ is the initial state 
125: (at $t \rightarrow -\infty$) and $\lambda(t)= \beta \, t$, 
126: and calling $|\psi(t)\rangle$ the evolving wave function, the asymptotic 
127: probability to end up in the other diabatic state is  
128: %
129: \begin{eqnarray}
130: P_2(t \rightarrow \infty) &=& 
131:                  |\langle2|\psi(t \rightarrow \infty)\rangle|^2 = 
132:                  1 - \exp{\left[\frac{-2\pi \delta^2}
133:                             {\hbar |\dot{\varepsilon}_1-\dot{\varepsilon}_2|}
134:                  \right]} \nonumber \\
135:     &=& 1 - \exp{\left[\frac{-2\pi \delta^2}
136:                             {\hbar \beta |\alpha_1-\alpha_2|}
137:                  \right]}.
138: \label{eq:asymp_prob}             
139: \end{eqnarray}
140: %
141: 
142: 
143: 
144: \begin{figure}
145: \begin{center}
146: \includegraphics[height=7cm,angle=-90]{fig1.ps}
147: \caption{Schematic plot of an avoided crossing.
148: The solid lines represent the adiabatic levels $E_1(\lambda)$ and $E_2(\lambda)$. 
149: The asymptotes to the energy hyperbolas are the diabatic straight 
150: lines $\varepsilon_1(\lambda)$ and $\varepsilon_2(\lambda)$ (dashed lines).
151: The center of the avoided crossing at $\overline{\lambda}$ and 
152: $\overline{\varepsilon}$ is inidicated with dotted lines.}
153: \label{fig:avoided}
154: \end{center}
155: \end{figure}
156: 
157: The two limiting cases in terms of $\beta$ (the rate of change of the parameter 
158: $\lambda$) are: \\
159: %
160: i) Slow transition: 
161: $\beta \ll \frac{2 \pi \delta^2}{\hbar |\alpha_1-\alpha_2|}$. 
162: In this case the system follows the adiabatic curve going from the initial diabatic 
163: state to the other one. \\
164: %
165: ii) Rapid transition:
166: $\beta \gg \frac{2 \pi \delta^2}{\hbar |\alpha_1-\alpha_2|}$.
167: The evolution takes place on the diabatic curve and the system remains in the initial 
168: diabatic state.\\
169: %
170: These limiting cases play a central role in our control method \cite{mur-wis-tam}, 
171: as they give us quantitative criteria to choose either the diabatic
172: or adiabatic paths in traversing an avoided crossing \cite{wis-mur-tam}.
173: 
174: More recently, the complete time dependence of the occupation probabilities in both 
175: the diabatic ($|\langle 1 | \psi(t)\rangle|^2$, 
176: $|\langle 2 | \psi(t)\rangle|^2$)
177: and adiabatic basis  
178: ($|\langle \phi_1(\lambda) | \psi(t)\rangle|^2, 
179: |\langle \phi_2(\lambda) | \psi(t)\rangle|^2$)
180: sets has been obtained \cite{vit,dam-zur}.
181: 
182: 
183: 
184: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
185: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
186: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
187: 
\section{The system: Quasi-one-dimensional doble quantum dot with two interacting 
188: electrons}
189: \label{sec:system}
190: 
191: Let us consider a quasi-one-dimensional double quantum dot with two interacting 
192: electrons in the presence of a uniform longitudinal electric field.
193: This system is interesting for two reasons. 
194: %
195: First, this type of system is experimentally realizable nowadays, and second,
196: the nonperturbative interparticle interaction is taken into account.
197: 
198: We have chosen a semiconductor system with realistic dimensions: it is an elongated
199: quantum dot $100 \, \mbox{nm}$ long and $50 \, \mbox{\AA}$ wide. 
200: Due to the small thickness of the structure, the energies of the transverse modes 
201: are widely spaced and it is enough to consider only the lowest transverse state.
202: Therefore, an effective Hamiltonian that depends only on the longitudinal coordinate
203: $z$ describes the dynamics of the system \cite{tam-met}
204: %
205: \begin{eqnarray}
206: \label{hamiltonian}
207: H&\equiv&-\frac{\hbar^2}{2m}(\frac{\partial^2}{\partial z_1^2} + 
208: \frac{\partial^2}{\partial z_2^2}) + V(z_1) + V(z_2) \nonumber
209:  \\
210: &&+ V_C(|z_1-z_2|) - e(z_1 + z_2)E(t) \, ,
211: \end{eqnarray}
212: %
213: where $m$ is the electron effective mass in the semiconductor material, 
214: $V_C$ is the Coulomb interaction between the electrons, 
215: $V$ is the confining potential, and 
216: $E(t)$ is a time-dependent external electric field.
217: There is no restriction regarding the choice of the confining potential in the 
218: $z$-direction, but we have selected the double-well configuration shown 
219: in  Fig.\ \ref{fig:system}. 
220: Double-well potentials are interesting due to the interplay between tunneling
221: and localization.
222: Moreover, in our two-electron system these important phenomena can be related
223: to the Coulomb interaction.
224: In all of the time evolutions that we will analyze, we assume that the wave function
225: is initially a singlet (antisymmetric spin wave function). 
226: Since the Hamiltonian is spin independent, the spin wave function remains a singlet
227: and the orbital part of the wave function is symmetric at all times.
228: 
229: \begin{figure}
230: \begin{center}
231: \includegraphics[height=7cm,angle=-90]{fig2.ps}
232: \caption{Confining double well potential in the longitudinal direction of the coupled quantum dot structure. The external electric field is  $E=0$ (solid lines) and $E=12 \, \mbox{kV/cm}$ (dashed lines).}%
233: \label{fig:system}
234: \end{center}
235: \end{figure}
236: 
237: We have used the time-dependent electric field $E(t)$ as the control parameter
238: \cite{mur-wis-tam,wis-mur-tam}. 
239: The first step is to understand the behavior of our system when the electric 
240: field is taken as a constant.
241: So, we have computed numerically the eigenergies and eigenfunctions of the system as a function of a constant electric field. 
242: We numerically diagonalize the Hamiltonian of Eq.\ (\ref{hamiltonian}) expanded  
243: in the basis set of Slater determinants constructed with the 12 bound single-particle
244: states of the double-well potential.
245: The two-particle basis set has then 12*(12+1)/2=78 states.
246: 
247: \begin{figure}
248:   \begin{center}
249:     \includegraphics[height=12cm,angle=-90]{fig3.ps}
250:     \caption{The energy spectrum of the two interacting electrons confined in
251:     a quasi-one-dimensional double-well semiconductor nanostructure as a function of 
252:     an external uniform electric field. It is plotted the first 31 energy levels.
253:     See text for details.}%
254:   \end{center}
255:   \label{fig:spectrum}
256: \end{figure}
257: 
258: The energy spectrum as a function of the external electric field is shown in 
259: Fig.\ \ref{fig:spectrum}. 
260: We can see that the spectrum is composed by fairly straight lines which never cross
261: each other, resulting in frequent avoided crossings.
262: For the low-lying states included in our spectrum, and far from the avoided crossings,
263: the adiabatic states have clear localization properties connected with their 
264: slope (see Fig.\ 2 in Ref.\ \cite{mur-wis-tam}): \\
265: %
266: i) in the eigenstates corresponding to negative slope both electrons are in the left 
267: well,\\ 
268: %
269: ii) the states associated to the positive slope are localized on the right well, and \\
270: iii) the states with neutral slope have one electron in each well.\\
271: %
272: At avoided crossings, the eigenstates mix their localization characteristics
273: reaching the maximal degree of mixing at the center of the avoided crossing.
274: 
275: 
276: 
277: 
278: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
279: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
280: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
281: 
282: \section{Analysis of the applicability of the Landau-Zener model}
283: \label{sec:analysis-applicability}
284: 
285: In our previous works we introduced a method of quantum control via traveling
286: in the energy spectrum of a quantum system \cite{mur-wis-tam,wis-mur-tam}.
287: The building blocks of this method are, on the one hand, the adiabatic evolutions
288: far from avoided crossings, and on the other, the slow  and fast 
289: evolutions employed at avoided crossings in order to shift in a controlled way
290: from one adiabatic path to a neighboring one.
291: If the system behaves locally like a LZ this possibility will be guaranteed.
292: For this reason, in this section we will analyze the range of validity or 
293: applicability of the 
294: Landau-Zener model to describe the transitions at the avoided crossings of 
295: our system.
296: 
297: We begin by studying the avoided crossing labelled ``A" in Fig.\ \ref{fig:spectrum},
298: between the ground state and the first excited state near the value of the 
299: electric field $E = 5 \,\mbox{kV/cm}$.
300: %
301: Initially the system is in the ground state with no electric field 
302: (state labelled ``a" in Fig.\ \ref{fig:spectrum})  
303: and we study the probability to remain in the ground state when the electric field 
304: is increased linearly with time at different velocities.
305: %
306: This corresponds to the adiabatic probability 
307: $|\langle \phi_1(E) | \psi(t)\rangle|^2$
308: introduced at the end of Section \ref{sec:LZ_model}.
309: These probabilities are shown in Fig.\ \ref{fig:avoided_1_adiab}.
310: %
311: We remark that in order to compare the results for different velocities, we
312: plot in Fig.\ \ref{fig:avoided_1_adiab} the probabilities as functions of the electric field rather than as functions of time.
313: %
314: In Fig.\ \ref{fig:avoided_1_adiab} we present the adiabatic probability for the 
315: following velocities,
316: $\dot{E} = 0.07, 0.27, 0.53, 1.07, 4.27 \, \mbox{kV/cm ps}$.
317: 
318: We now compute the adiabatic probabilities in the LZ model.
319: The first step is to fit the parameters $\delta$, the location of the avoided crossing,
320: $\alpha_1$, and $\alpha_2$ \cite{footnote} 
321: of the two-level Hamiltonian of Eq.\ (\ref{eq:hamiltonian}) to the avoided crossing 
322: under study.
323: As initial state we take the adiabatic LZ state for the value of $\lambda$ that 
324: corresponds to state ``a" in Fig.\ \ref{fig:spectrum}.   
325: We compute the adiabatic probabilities in the LZ model for the previous set of rates
326: of change of the electric field. 
327: These results are plotted with solid lines in Fig.\ \ref{fig:avoided_1_adiab}.
328: We can see that our system is well described by the LZ model for the whole
329: range of velocities considered.
330: Moreover, we see that the asymptotic probability obtained by Zener (horizontal
331: dashed lines in Fig.\ \ref{fig:avoided_1_adiab}) gives accurate results far from the 
332: avoided crossing.
333: Note that far from the avoided crossing the diabatic and adiabatic states in the
334: LZ model are essentially the same.
335: 
336: \begin{figure}
337:   \begin{center}
338:     \includegraphics[height=10cm,angle=-90]{fig4.ps}
339:   \end{center}
340:     \caption{Adiabatic transition probabilities 
341:              $|\langle \phi_1(E) | \psi(t)\rangle|^2$
342:              at the lowest avoided crossing (labelled ``A" in 
343:              Fig.\ \ref{fig:spectrum}) for various rates of change of 
344:              the control parameter.
345:               The velocities are: 
346:              $\dot{E} = 0.07 (\triangle), 0.27 (\Box), 0.53 (\times),     
347:              1.07 (\circ), 4.27 \,(*)\, \mbox{kV/cm ps}$.
348:              The initial state in the exact evolution is the one labelled ``a" 
349:              in Fig.\ \ref {fig:spectrum}.
350:              The solid lines give the adiabatic probabilities in the LZ model
351:              and the dotted lines are the asymptotic LZ probabilities
352:              given in Eq.\ (\ref{eq:asymp_prob}).}
353:  
354:   \label{fig:avoided_1_adiab}
355: \end{figure}
356: 
357: 
358: We have done similar analyses for other avoided crossings of the energy spectrum.
359: Namely, we start with the states labelled ``b", ``c", and ``d" in 
360: Fig.\ \ref{fig:spectrum}, and we study the transition probabilities in the 
361: adjacent avoided crossings ``B", ``C", and ``D", respectively.
362: In Fig.\ \ref{fig:avoided_2_adiab} we show the results, which verify the previous conclusion, in the sense that in the adiabatic basis the two-level LZ model fits 
363: very well the exact results.
364: It is worth noting here that in our previous work of 
365: Ref.\ \cite{mur-wis-tam,wis-mur-tam}
366: we travel in the spectrum (that is, we attempt to go from a given adiabatic
367: state to another one). 
368: In this sense, the above given results are the most relevant ones to judge
369: the applicability of our control method. 
370: 
371: \begin{figure}
372:   \begin{center}
    \includegraphics[height=14cm,angle=-90]{fig5.ps}
373:    \end{center}
374:     \caption{Adiabatic transition probabilities 
375:              at various avoided crossings (labeled "B", "C" and "D" in 
376:              Fig.\ \ref{fig:spectrum}).
377:              The initial states in these evolutions are labelled in Fig.\ 
378:              \ref {fig:spectrum} as 
379:              "b" [panel (a)] , "c" [panel (b)], and "d" [panel (c)] . 
380:              The velocities are: 
381:              (a) $\dot{E} = 0.0015 (\triangle), 0.0077 (\Box), 0.015 (\times),     
382:              0.07 (\circ), 4.27 \,(*)\, \mbox{kV/cm ps}$, 
383:              (b) $\dot{E} = 0.07 \,(\triangle), 0.53 \,(\Box), 1.27 \,(\times),     
384:              4.27 \,(\circ), 40.0 \,(*)\, \mbox{kV/cm ps}$, and 
385:              (c) $\dot{E} = 0.003 \,(\triangle), 0.07 \,(\Box), 0.27 \,(\times),     
386:              0.53 \,(\circ), 4.27 \, (*)\, \mbox{kV/cm ps}$,}
387:   
388:   \label{fig:avoided_2_adiab}
389: \end{figure}
390: 
391: Since the LZ model is defined on the basis of diabatic states, it is perhaps
392: more natural to perform the former analysis on that basis set.
393: However, the question arises of what the diabatic states are in our realistic
394: system.
395: Indeed, in a multilevel system like ours the two states involved in the
396: avoided crossing become mixed with other states and therefore acquire a 
397: dependence on the control parameter (which is not allowed for in the 
398: usual LZ model).
399: It is thus an interesting question to ask whether it is possible to find a 
400: ``fixed" basis set which could play the role of the diabatic basis in 
401: the LZ model. 
402: For example, we now calculate the probability  
403: $|\langle \phi_1(E_0) | \psi(t)\rangle|^2$, where $|\phi_1(E_0)\rangle$ is the 
404: the initial state in the dynamic passage of an avoided crossing.
405: That is, we are considering $|\phi_1(E_0)\rangle$ as being one of the diabatic 
406: basis states. 
407: We now do this for the lowest crossing taking $E_0=0$, and compare with 
408: the results of using the two-level LZ model in Fig.\ \ref{fig:avoided_1_diab}.
409: One can clearly see that the agreement between the two calculations is 
410: not very good.
411: This can be understood with the help of the inset of 
412: Fig.\ \ref{fig:avoided_1_diab}, which shows
413: the overlaps $|\langle \phi_1(E=0) | \phi_1(E) \rangle|^2$ and
414: $|\langle \phi_1(E=0) | \phi_2(E) \rangle|^2$ as functions of
415: the electric field $E$.
416: It is clear from the inset, especially from  
417: $|\langle \phi_1(E=0) | \phi_2(E) \rangle|^2$ (dashed line), 
418: that the hypothesis of a parameter-independent diabatic state is not satisfied
419: (that the overlap is not equal to one far from the avoided crossing), 
420: and therefore the LZ model tends to fail.
421: %
422: However, in other avoided crossings we have observed that it is possible to find 
423: good diabatic states (which are fairly parameter-independent around the avoided crossing).
424: For example, we repeated the previous analysis for the avoided crossing labelled
425: ``B" in Fig.\ \ref{fig:spectrum}, choosing 
426: $|\phi_{16}(E=1.4 \, \mbox{kV/cm})\rangle$ as one of the diabatic states.
427: In Fig.\ \ref{fig:avoided_2_diab}, we plot the probability 
428: $|\langle \phi_{16}(E=1.4 \, \mbox{kV/cm}) | \psi(t)\rangle|^2$, which shows
429: a better agreement than the one in Fig.\  \ref{fig:avoided_1_diab}.
430: We remark that, as can be seen in the inset of Fig.\ \ref{fig:avoided_2_diab}, the state 
431: $|\phi_{16}(E=1.4 \, \mbox{kV/cm})\rangle$ is a good choice of diabatic state,
432: since the overlap 
433: $|\langle \phi_{16}(E=1.4 \, \mbox{kV/cm}) | \phi_{17}(E) \rangle |^2$
434: is close to one at the right of the crossing and close to zero to the left
435: (see dashed line).
436: The behavior seen in the inset is exactly what one obtains in the LZ model
437: for the overlaps between the diabatic and adiabatic bases.
438: 
439: 
440: \begin{figure}
441:   \begin{center}
442:     \includegraphics[height=10cm,angle=-90]{fig6.ps}
443:     \end{center}
444:     \caption{Diabatic transition probabilities 
445:              $|\langle \phi_1(E_0=0) | \psi(t)\rangle|^2$ (dotted lines)
446:              at the lowest avoided crossing, labelled ``A" in Fig.\ \ref{fig:spectrum}. 
447:              The rates of change of the control parameter, the electric field $E$, are the
448:              same as Fig.\ \ref{fig:avoided_1_adiab}.
449:              In solid lines, the corresponding probabilities in the LZ model. 
450:              Inset: overlaps $|\langle \phi_1(E=0) | \phi_1(E) \rangle|^2$ (solid line)
451:              and $|\langle \phi_1(E=0) | \phi_2(E) \rangle|^2$ (dashed line).}
452:   \label{fig:avoided_1_diab}
453: \end{figure}
454: 
455: 
456: \begin{figure}
457:   \begin{center}
458:     \includegraphics[height=10cm,angle=-90]{fig7.ps}
459:     \end{center}
460:     \caption{Diabatic transition probabilities 
461:              $|\langle \phi_{16}(E=1.4 \, \mbox{kV/cm)} | \psi(t)\rangle|^2$ (dotted lines)
462:              at the avoided crossing, labelled ``B" in Fig.\ \ref{fig:spectrum}. 
463:              The rates of change of the control parameter, the electric field $E$, are the
464:              same as in Fig.\  \ref{fig:avoided_2_adiab}(a).
465:              In solid lines, the corresponding probabilities in the LZ model. 
466:              Inset: overlaps 
467:              $|\langle \phi_{16}(E=1.4 \, \mbox{kV/cm}) | \phi_{16}(E) \rangle |^2$ 
468:              (solid line)
469:              and $|\langle \phi_{16}(E=1.4 \, \mbox{kV/cm}) | \phi_{17}(E) \rangle |^2$
470:              (dashed line).}
471:   \label{fig:avoided_2_diab}
472: \end{figure}
473: 
474: 
475: 
476: \section{Final Remarks}
477: 
478: We have studied the applicability of the LZ model in a realistic system:
479: a quasi-one-dimensional double quantum dot with two interacting electrons.
480: We showed that the LZ model works very well when it is viewed in the adiabatic
481: basis.
482: This result is the cornerstone for the quantum controlability using the
483: method of control introduced in \cite{mur-wis-tam}.
484: 
485: However, when seen in the diabatic basis the results are not so robust
486: as in the case of the adiabatic basis.
487: This is due to the fact that, for multilevel systems, a proper diabatic basis 
488: does not exist.
489: Rather, the pair of interacting levels at an avoided crossing become mixed with 
490: other states and acquire a dependence with the control parameter even far from
491: the avoided crossings.
492: 
493: 
494: \section*{Acknowledgement(s)}
495: The authors acknowledge support from CONICET (PIP-6137, PIP-5851) and
496: UBACyT (X248, X179).
497: D.A.W.\ and P.I.T.\ are researchers of CONICET.
498: 
499: 
500: 
501: \begin{thebibliography}{99}
502: 
503: \bibitem{zen}
504: C.\ Zener, Proc.\ R.\ Soc.\ London, Ser.\ A {\bf 137}, 696 (1932).
505: 
506: \bibitem{bul-dod-kus}
507: A.\ Bulgac, G.\ Do Dang, and D.\ Kusnezov,
508: Ann.\ Phys.\ (Leipzig) {\bf 242}, 1 (1995);
509: Phys.\ Rep.\ {\bf 264}, 67 (1996).
510: 
511: \bibitem{san-ver-wis}
512: M.\ J.\ S\'anchez, E.\ Vergini, and D.\ A.\ Wisniacki,
513: Phys.\ Rev.\ E {\bf 54}, 4812 (1996).
514: 
515: \bibitem{mur-wis-tam}
516: G.\ E.\ Murgida, D.\ A.\ Wisniacki, and P.\ I.\ Tamborenea,
517: Phys.\ Rev.\ Lett.\ {\bf 99}, 036806 (2007).
518: 
519: \bibitem{wis-mur-tam}
520: D.\ A.\ Wisniacki, G.\ E.\ Murgida, and P.\ I.\ Tamborenea,
521: AIP Proc.\ {\bf 963}, 840 (2007).
522: 
523: \bibitem{vit}
524: N.\ V.\ Vitanov, Phys.\ Rev.\ A {\bf 59}, 988 (1999).
525: 
526: \bibitem{dam-zur}
527: B.\ Damski and W.\ H.\ Zurek, 
528: Phys.\ Rev.\ A {\bf 73}, 063405 (2006).
529: 
530: \bibitem{tam-met}
531: P.\ I.\ Tamborenea and H.\ Metiu,
532: Phys.\ Rev.\ Lett.\ {\bf 83}, 3912 (1999).
533: 
534: \bibitem{footnote}  Actually only the difference $\alpha_1 - \alpha_2$ enters in the
535: evolution of the two-level LZ system.  The location of the avoided crossing is irrelevant
536: in terms of the probabilities, but is is necessary to compare the results of the 
537: two calculations.
538: 
539: \end{thebibliography}
540: 
541: \label{lastpage}
542: 
543: \end{document}
544: 
545: 
546: