1: \documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
2: %\documentclass[showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
3: %\documentclass[preprint,tightenlines,aps,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
4: %\documentclass[preprint,showpacs,amsmath,amssymb]{revtex4}
5:
6: % Some other (several out of many) possibilities
7: %\documentclass[preprint,aps]{revtex4}
8: %\documentclass[preprint,aps,draft]{revtex4}
9: %\documentclass[prb]{revtex4}% Physical Review B
10:
11: \usepackage{graphicx}% Include figure files
12: %\usepackage{dcolumn}% Align table columns on decimal point
13: \usepackage{bm}% bold math
14: %\usepackage{dcolumn}
15: %\nofiles
16: %\tighten
17:
18: \begin{document}
19:
20: \title{Generation and control of resonance states in crossed magnetic
21: and electric fields}
22:
23: \author{K. Krajewska and J. Z. Kami\'nski}
24:
25: \affiliation{Institute of Theoretical Physics, Warsaw University,
26: Ho\.za 69, 00-681 Warszawa, Poland}
27:
28: %\date{\today}
29:
30: \begin{abstract}
31: A two-dimensional electron system interacting with an impurity and
32: placed in crossed magnetic and electric fields is under
33: investigation. Since it is assumed that an impurity center
34: interacts as an attractive $\delta$-like potential a
35: renormalization procedure for the retarded Green's function has to
36: be carried out. For the vanishing electric field we obtain a close
37: analytical expression for the Green's function and we find one
38: bound state localized between Landau levels. It is also shown by
39: numerical investigations that switching on the electric field new
40: long-living resonance states localized in the vicinity of Landau
41: levels can be generated.
42: \end{abstract}
43:
44: \pacs{03.65.Ge, 71.70.Di, 73.43.-f}
45: \maketitle
46:
47:
48: %\begin{keyword}
49: %resonances, lifetime, quantum Hall effect, Landau levels,
50: %zero-range potential, point interaction
51: %\end{keyword}
52:
53:
54: % main text
55: \section{Introduction}
56: \label{sec:Introduction}
57:
58: The quantum mechanical motion of electrons in two-dimensional
59: space, subject to crossed magnetic and electric fields and in the
60: presence of a single impurity described by a zero-range potential,
61: provides a basis for the explanation of the integer quantum Hall
62: effect \cite{P81,PJ82,JP84,PC91,CC98,GM99,HL99}. Until now in all
63: considerations the electric field has been treated as a small
64: perturbation. The purpose of the present work is to reconsider
65: these investigations and to discuss some novel phenomena which
66: appear for nonperturbative electric fields.
67:
68: In our approach an attractive impurity is modelled as a zero-range
69: potential center, as has been already exploited in
70: \cite{PC91,CC98}. It turns out that in order to define such a
71: point interaction one needs to renormalize the strength of the
72: $\delta$-function and that such a renormalized potential is
73: capable of supporting one localized state per Landau level. This
74: finding also corresponds to the result obtained by Gyger and
75: Martin \cite{GM99}. Investigating the case of a weak electric
76: field they have proven that all localized states change into
77: resonances with lifetimes varying with the electric field in a
78: Gaussian way. With reference to these papers we shall focus our
79: attention on the opposite regime, i.e., when the electric fields
80: is strong. Our aim is to answer the following questions:
81: \begin{enumerate}
82: \item
83: how many resonance states exist per one Landau level for a non-vanishing
84: electric field?
85: \item
86: what is a maximum electric field above which resonance states of a
87: given lifetime cannot exist?
88: \end{enumerate}
89: We shall see that apart from the well-known localized states lying
90: between Landau levels \cite{PC91,CC98,GM99} \textit{new resonance
91: states} are created by a strong electric field, the lifetime of
92: which can be comparable to or larger than typical times
93: encountered in solid state physics.
94:
95: Throughout this paper we use units in which $\hbar=1$.
96:
97: \section{Attractive point interaction}
98: \label{sec:math}
99:
100: The two-dimensional system we are concerned with is governed by
101: the hamiltonian which, for convenience, we decompose into two
102: parts, $H_0$ related to the physical situation for which the
103: Green's function is known and a $\delta$-like potential with the
104: strength given by the so-called bare coupling constant $\lambda$,
105: \begin{equation}
106: H=H_0+\lambda \delta^{(2)} (\bm{r}) \, . \label{e2.1}
107: \end{equation}
108: Such a model has been extensively studied in the literature
109: \cite{AGHH88,B75,FFR90,FFR91,MFSF00,W91} proving its usefulness in
110: many branches of physics. It is, however, important to realize
111: that the model is not free from difficulties. To be more precise,
112: it exhibits divergencies which have to be tempered by
113: renormalizing the bare coupling constant $\lambda$. Therefore, let
114: us address this problem and show how to remove all divergent terms
115: by considering the retarded Green's function for the hamiltonian
116: $H$.
117:
118: Our starting point is the Lippmann-Schwinger equation which, in
119: this particular case, adopts the form
120: \begin{eqnarray}
121: G(\bm{r},\bm{r}';E) =
122: G_0^{(\mathrm{reg})}(\bm{r},\bm{r}';E) \nonumber \\
123: + \int \mathrm{d}^2\bm{r}''
124: G_0^{(\mathrm{reg})}(\bm{r},\bm{r}'';E)\lambda
125: \delta^{(2)} (\bm{r}'') G(\bm{r}'',\bm{r}';E) \, ,
126: \label{e2.2}
127: \end{eqnarray}
128: where the regularized Green's function for the hamiltonian $H_0$,
129: denoted by $G_0^{(\mathrm{reg})}(\bm{r},\bm{r}';E)$, is
130: assumed to be known. Now performing the foregoing integral over
131: $\text{d}^2\bm{r}''$ and making some algebraic manipulations
132: we arrive at the retarded Green's function for our model
133: \begin{eqnarray}
134: G(\bm{r},\bm{r}';E)=
135: G_0^{(\text{reg})}(\bm{r},\bm{r}';E) \nonumber\\
136: + \frac{G_0^{(\text{reg})}(\bm{r},\bm{0};E)
137: G_0^{(\text{reg})}(\bm{0},\bm{r}';E)}
138: {\lambda^{-1}-G_0^{(\text{reg})}(\bm{0},\bm{0};E)} \, .
139: \label{e2.3}
140: \end{eqnarray}
141: It appears that $G_0^{(\text{reg})}(\bm{0},\bm{0};E)$ is
142: the only term which needs to be regularized in order to avoid
143: divergencies. Let us explain how to cope with them analyzing the
144: simplest case of a free electron with an effective mass $m^*$
145: interacting only with the $\delta$-like potential. In our approach
146: we choose a particular regularization scheme which we call the
147: proper time regularization. We expect that although the bare
148: coupling constant $\lambda$ depends on the regularization scheme
149: chosen, the observable quantities, like poles of the full Green's
150: function in the complex energy plane, cannot depend on this
151: scheme. Indeed, we shall demonstrate that our method leads to the
152: same equation for energies of localized states in the magnetic
153: field as in \cite{CC98}.
154:
155: For a free particle the retarded Green's function in the so-called
156: Fock-Schwinger proper time represantation \cite{IZ80} has the form
157: \begin{eqnarray}
158: G_0^{(\text{reg})}(\bm{r},\bm{r}';E)
159: =-\text{i}\int_{-\text{i}\sigma}^{\infty} \text{d}t
160: \text{e}^{\text{i}(E+\text{i}\varepsilon)t} \nonumber\\
161: \int \frac{\text{d}^2\bm{k}}{(2\pi)^2} \exp\biggl
162: (\text{i}\bm{k}\cdot (\bm{r} -\bm{r}')
163: -\text{i}\frac{\bm{k}^2}{2m^*}t \biggr ) \, , \label{e2.4}
164: \end{eqnarray}
165: where the singularity for $\bm{r}=\bm{r}'=\bm{0}$ is
166: avoided by modifying the integration contour over the proper time
167: $t$ in such a way that it lies just below the real axis (this
168: means that $\sigma$ is a small positive parameter). Naturally, the
169: retarded Green's function for non-vanishing $\bm{r}$ or
170: $\bm{r}'$ is recovered by putting $\sigma=0$. Such a
171: prescription we call the proper time regularization. Now, it is
172: straightforward to show that the denominator in eq. (\ref{e2.3})
173: reads
174: \begin{equation}
175: D(E)=
176: \lambda^{-1}+\frac{m^*}{2\pi}\int_{-\text{i}\sigma}^{\infty}
177: \text{d}t \text{e}^{\text{i}(E+\text{i}\varepsilon)t}
178: t^{-1} \, . \label{e2.5}
179: \end{equation}
180: It is also clear that the integral above diverges logarithmically
181: for small values of $t$ when $\sigma=0$. This divergence, however,
182: can be absorbed into the bare coupling constant $\lambda$ giving
183: the finite expression for the full Green's function even for
184: $\sigma$ equal to 0. Indeed, performing the proper time
185: integration by parts we arrive at
186: \begin{equation}
187: D(E)=\lambda_R^{-1}+\frac{m^*}{2\pi} \biggl ( \ln(\text{i})
188: -\text{i}E\int_0^{\infty} \text{d}t
189: \text{e}^{\text{i}(E+\text{i}\varepsilon)t}\ln(t/t_0) \biggr
190: ) \, , \label{e2.6}
191: \end{equation}
192: where $t_0$ is an arbitrary positive parameter of the same
193: dimensionality as $t$, whereas $\lambda_R$ is the renormalized
194: coupling constant defined by the relation
195: \begin{equation}
196: \lambda_R^{-1}=\lim_{\sigma\rightarrow 0} \Bigl (
197: \lambda^{-1}-\frac{m^*}{2\pi}\ln \frac{\sigma}{t_0} \Bigr ) \, .
198: \label{e2.7}
199: \end{equation}
200: Let us remark that in order to temper a singular part in the
201: formula above the bare coupling constant $\lambda$ has to be
202: negative. This means, in fact, that in two dimensions there cannot
203: exist a repulsive point interaction. Having this in mind we shall
204: restrict ourselves only to the case with an attractive point
205: interaction for which, after performing the remaining integral, we
206: have
207: \begin{equation}
208: D(E)=\lambda_R^{-1}+\frac{m^*}{2\pi}\ln\biggl (
209: \frac{\text{e}^{-\gamma}}{-Et_0} \biggr ) \, , \label{e2.8}
210: \end{equation}
211: where $\gamma$ is the Euler's constant. This equation shows that
212: independently of numerical values of $\lambda_R$ and $t_0$ there
213: exists $E$ such that the denominator of the full Green's function
214: vanishes, i.e., the two-dimensional zero-range potential always
215: supports a bound state. Denoting the energy of this state by $E_B$
216: we end up with the following expression for the denominator
217: $D(E)$,
218: \begin{equation}
219: D(E)=\frac{m^*}{2\pi}\ln\biggl (\frac{E_B}{E} \biggr ) \, .
220: \label{e2.9}
221: \end{equation}
222: Let us add in closing that there still remains a free parameter of
223: the theory, i.e., the energy of a bound state $E_B$. Nevertheless,
224: it has now the clear physical meaning in the sense that its value
225: can be measured, contrary to the artificial and physically unclear
226: parameters $t_0$ and $\lambda_R$.
227:
228: \section{Green's function in external fields}
229: \label{sec:Green'sfunction}
230:
231: In this section we consider the two-dimensional system composed of
232: an electron and $\delta$-like impurity placed in the external
233: magnetic and electric fields for which the hamiltonian $H_0$ can
234: be written as
235: \begin{equation}
236: H_0=\frac{1}{2m^*}\Biggl[ \Bigl(
237: -\text{i}\partial_x+\frac{e\cal{B}}{2}y\Bigr )^2+
238: \Bigl(-\text{i}\partial_y-\frac{e\cal{B}}{2}x\Bigr )^2\Biggr
239: ]-e{\cal E}x \, , \label{e3.1}
240: \end{equation}
241: where the electric $\bm{\mathcal{E}}$ and magnetic
242: $\bm{\mathcal{B}}$ fields are directed towards $x$ and $z$
243: axes, respectively, whereas the vector potential is taken in the
244: circular gauge,
245: $\bm{A}=\frac{1}{2}\bm{\mathcal{B}}\times\bm{r}$. As
246: we have already pointed out, the problem of finding energies of
247: impurity-induced resonance states consists in searching for zeros
248: of the denominator in eq. (\ref{e2.3}). Therefore, what we have to
249: know is the regularized Green's function for the hamiltonian
250: $H_0$. Applying the proper time regularization scheme discussed
251: above we arrive at the regularized retarded Green's function
252: $G_0^{(\text{reg})}$,
253: \begin{eqnarray}
254: G_0^{(\text{reg})}(\bm{r},\bm{r}';E) =
255: -\frac{m^*\omega}{4\pi}\int_{-\text{i}\sigma}^{\infty}
256: \text{d}t \frac{\text{e}^{\text{i}(E+\text{i}\varepsilon)t}}
257: {\sin\frac{\omega t}{2}} \nonumber\\
258: \times \exp\Biggl[ \frac{\text{i}m^*\omega}{4}\bigl( (x-x')^2
259: + (y-y')^2\bigr )\cot\frac{\omega t}{2} \nonumber\\
260: +\frac{\text{i}m^*\omega}{2}(xy'-x'y)
261: +\frac{\text{i}e{\cal E}t}{2}(x+x') \nonumber\\
262: + \frac{\text{i}e\cal{E}}{\omega}\Bigl( \frac{\omega
263: t}{2} \cot\frac{\omega t}{2}-1\Bigr ) \Bigl(
264: -y+y'+\frac{e{\cal E}t}{2m^*\omega}\Bigr )\Biggr ] \, ,
265: \label{e3.2}
266: \end{eqnarray}
267: and the denominator $D(E)$,
268: \begin{eqnarray}
269: D(E)=\lambda^{-1}+\frac{m^*\omega}{4\pi}\int_{
270: -\text{i}\sigma}^{\infty} \text{d}t
271: \frac{\text{e}^{\text{i}(E+\text{i}\varepsilon)t}}
272: {\sin\frac{\omega t}{2}} \nonumber\\
273: \times \exp\Biggl[
274: \frac{\text{i}t}{2m^*}\Bigl( \frac{e\cal{E}}{\omega}\Bigr
275: )^2 \Bigl( \frac{\omega t}{2}\cot\frac{\omega t}{2}-1\Bigr )\Biggl
276: ] \, , \label{e3.3}
277: \end{eqnarray}
278: where $\omega=|e|{\cal B}/m^*$ is the cyclotron frequency.
279: Comparing this formula with eq. (\ref{e2.5}) one can notice that
280: in both these expressions the integrals over $t$ diverge for small
281: arguments exactly in the same way. Hence, we gather that now it is
282: also possible to temper the divergent term by renormalizing the
283: coupling constant $\lambda$. At this point let us introduce new
284: dimensionless units. As a typical length scale for the problem we
285: take $1/\sqrt{m^{*}\omega}$, whereas the dimensionless scaled
286: electric field and energy are
287: $\tilde{\cal E}=|e|{\cal E}/\sqrt{m^* \omega^3}$ and
288: $\tilde{E}=2 E/\omega$, respectively. Additionally, choosing a new
289: variable $s=\omega t/2$ we find that
290: \begin{eqnarray}
291: D(\tilde{E})=\frac{m^*}{2\pi}\Biggl [\ln\biggl(
292: \frac{\tilde{E}_B}{\tilde{E}}\biggr )+ \int_0^{\infty}\text{d}s
293: \text{e}^{\text{i}(\tilde{E}+ \text{i}\varepsilon)s} \nonumber\\
294: \biggl(\frac{\exp\bigl( \text{i} {\tilde{\cal E}}^2 s(s\cot
295: s-1)\bigr )}{\sin s}-\frac{1}{s}\biggr )\Biggr ] \, . \label{e3.4}
296: \end{eqnarray}
297: At first glance it might seem that the integral above is divergent
298: for $s=n\pi$, where $n=1, 2,..$. It is not, however, the case.
299: Indeed, analyzing the integrand in the close vicinity of these
300: points one can prove the convergence of the foregoing integral.
301: Having this in mind we shall discuss below the special case when
302: the electric field is switched off.
303: \begin{figure}
304: \begin{center}
305: \includegraphics[width=7.5cm]{fig1.eps}
306: \end{center}
307: \caption{Impurity-induced localized states for the vanishing
308: electric field} \label{fig1}
309: \end{figure}
310: \begin{figure}
311: \begin{center}
312: \includegraphics[width=7.5cm]{fig2.eps}
313: \end{center}
314: \caption{Real part of resonance states as the function of
315: $\tilde{E}_B$ for $\text{Im}(\tilde{E})=10^{-4}$.} \label{fig2}
316: \end{figure}
317: \begin{figure}
318: \begin{center}
319: \includegraphics[width=7.5cm]{fig3.eps}
320: \end{center}
321: \caption{Real part of resonance states as the function of
322: $\tilde{\cal{E}}$ for $\text{Im}(\tilde{E})=10^{-4}$.}
323: \label{fig3}
324: \end{figure}
325: \begin{figure}
326: \begin{center}
327: \includegraphics[width=7.5cm]{fig4.eps}
328: \end{center}
329: \caption{Real part of resonance states as the function of
330: $\tilde{\cal{E}}$ for $\text{Im}(\tilde{E})=10^{-6}$.}
331: \label{fig4}
332: \end{figure}
333:
334: \section{Localized states in the external magnetic field}
335: \label{sec:localizedstates}
336:
337: The problem without an electric field has been already examined by
338: Cavalcanti and de Carvalho in \cite{CC98}. However, we have found
339: it interesting to derive their equation for energies of the
340: impurity-induced localized states using the proper time
341: regularization scheme. Our point of departure is the function
342: defined by eq. (\ref{e3.4}) with $\tilde{\cal E}=0$,
343: \begin{equation}
344: D(\tilde{E})=\frac{m^{*}}{2\pi}\Biggl[ \ln\biggl(
345: \frac{\tilde{E}_B}{\tilde{E}}\biggr ) +\int_0^{\infty}\text{d}s
346: \text{e}^{\text{i}(\tilde{E}+\text{i}\varepsilon)s}
347: \biggl(\frac{1}{\sin s}-\frac{1}{s}\biggr) \Biggr] \, .
348: \label{e3.5}
349: \end{equation}
350: In order to treat these two integrals separately we regularize
351: them first. This allows us to integrate them by parts and put
352: finally the regularization parameter $\sigma$ equal to 0, keeping
353: in mind, however, that the integration contour over $s$ lies just
354: below the real axis,
355: \begin{eqnarray}
356: D(\tilde{E})=\frac{m^{*}}{2\pi}\Biggl[ \gamma+\ln(2\tilde{E}_B)
357: -\frac{\text{i}\pi}{2} \nonumber\\
358: -\text{i}\tilde{E} \int_{0}^{\infty}\text{d}s
359: \text{e}^{\text{i}(\tilde{E}+\text{i}\varepsilon)s} \ln\tan
360: (s/2) \Biggr] \, . \label{e3.6}
361: \end{eqnarray}
362: Next, let us expand $\ln\tan(s/2)$ into series of
363: $\text{e}^{-\text{i}s}$. This is allowed because the
364: inequality $|\text{e}^{-\text{i}s}|<1$ holds for $s$
365: possessing a negative imaginary part, which is guaranteed by our
366: choice of a regularization scheme. Hence, inserting into eq.
367: (\ref{e3.6}) the series expansion
368: \begin{equation}
369: \ln\tan(s/2)=-\frac{\text{i}\pi}{2}-2\sum_{n=0}^{\infty}
370: \frac{\text{e}^{-\text{i}(2n+1)s}}{2n+1} \, , \label{e3.7}
371: \end{equation}
372: and carrying out all integrations term by term, we arrive at
373: \begin{eqnarray}
374: D(\tilde{E})=\frac{m^{*}}{2\pi}\Biggl[
375: \gamma+\ln(2\tilde{E}_B)-\text{i}\pi \nonumber\\
376: + 2\tilde{E}\sum_{n=0}^{\infty}\frac{1}{(2n+1)(2n+1-\tilde{E})}
377: \Biggr] \, . \label{e3.8}
378: \end{eqnarray}
379: Finally, relating the sum above to a digamma function
380: $\psi(z)$, we find that the eigenenergies of our system are
381: zeros of the following function
382: \begin{equation}
383: D(\tilde{E})=\frac{m^{*}}{2\pi}\Biggl[ \ln\biggl(
384: \frac{|\tilde{E}_B|}{2}\biggr)
385: -\psi\biggl(\frac{1-\tilde{E}}{2}\biggr) \Biggr] \, , \label{e3.9}
386: \end{equation}
387: which is adequate to the result derived in \cite{CC98} with the
388: help of another regularization scheme. Zeros of $D(\tilde{E})$
389: determine the dependence of $\tilde{E}$ on $\tilde{E}_B$ for those
390: localized states which are extracted from the Landau levels by the
391: zero-range interaction. It is shown in Fig. \ref{fig1} that only
392: one state is extracted per Landau level, which is understandable,
393: since in the degenerate Landau level there is only one state the
394: wavefunction of which does not vanish at the origin, i.e., the
395: cylindrically symmetric state \cite{BBCK}. The situation changes
396: if the electric field is present. It mixes this state with those
397: Landau states which explicitly depend on the polar angle. For
398: small electric fields such a mixing is also small and the effect
399: of a point interaction can be neglected. However, for strong
400: electric fields this effect becomes considerable as we are going
401: to show this below.
402: \begin{figure}
403: \begin{center}
404: \includegraphics[width=7.5cm]{fig5.eps}
405: \end{center}
406: \caption{Two resonance states labeled by $\#1$ and $\#2$ in Fig.
407: \ref{fig3}.} \label{fig5}
408: \end{figure}
409: \begin{figure}
410: \begin{center}
411: \includegraphics[width=7.5cm]{fig6.eps}
412: \end{center}
413: \caption{Real parts and lifetimes of some resonance states as the
414: function of $\tilde{\cal E}$ for $\tilde{E}_B=-3$.}
415: \label{fig6}
416: \end{figure}
417:
418: \section{Resonance states in external fields}
419: \label{sec:resonance}
420:
421: In order to determine resonance energies for a non-vanishing
422: electric field we have to find complex zeros of $D(\tilde{E})$
423: defined by eq. (\ref{e3.4}) and lying below the real axis. This
424: means that we have to solve one complex equation,
425: $D(\tilde{E})=0$, with four unknown real parameters:
426: $\text{Re}(\tilde{E})$, $\text{Im}(\tilde{E})$, $\tilde{E}_B$
427: and the scaled electric field $\tilde{\cal{E}}$. To this end
428: we shall fix one of these parameters, treat another one as a free
429: parameter, and find the remaining two parameters by solving this
430: equation, i.e., by determining poles of the full Green's function.
431: At the beginning let us fix the imaginary part of the resonance
432: energy and find all possible zeros of $D(\tilde{E})$. The results
433: are shown in Figs. \ref{fig2} and \ref{fig3}. In Fig. \ref{fig2}
434: we present the real part, $\text{Re}(\tilde{E})$, of all
435: resonance states of a given $\text{Im}(\tilde{E})=10^{-4}$ for
436: the bound state energy $\tilde{E}_B$ between $-30$ and 0, as it
437: has been already presented in Fig. \ref{fig1} for the vanishing
438: electric field (in both these figures we show only those states
439: for which $\text{Re}(\tilde{E})>1$, i.e., lying above the first
440: Landau level). We observe that, apart from the already existing
441: impurity-induced bound states for $\tilde{\cal{E}}=0$, a
442: non-vanishing electric field generates new resonance states which
443: are located in the vicinity of Landau levels of energy
444: $\hbar\omega (n+\frac{1}{2})$ for $n$ greater than 0. Moreover, as
445: it follows from our numerical investigations, the number of these
446: new resonance states is equal to $n$. This means that for a
447: non-vanishing electric field and for a given quantum system,
448: specified by the binding energy $\tilde{E}_B$, the number of
449: impurity-induced resonances per Landau level is equal to $n+1$,
450: where the quantum number $n$ labels the Landau energy. It appears
451: that for the vanishing electric field real parts of energies of
452: these new resonances approach the Landau energy, as it is shown in
453: Fig. \ref{fig3}, in which we present the real part,
454: $\text{Re}(\tilde{E})$, of all resonance states of a given
455: $\text{Im}(\tilde{E})=10^{-4}$ as the function of
456: $\tilde{\cal{E}}$. We observe that for a given imaginary part
457: of resonances there is a maximum electric field above which no
458: such resonance states exist. This is not a surprising fact. What
459: is unexpected, however, is that the maximum electric field very
460: little depends on the imaginary part of resonance energies chosen,
461: as it is shown in Fig. \ref{fig4} for
462: $\text{Im}(\tilde{E})=10^{-6}$; although the lifetime of these
463: \textit{new states} has increased hundred times the maximum
464: electric field has decreased only by a small fraction. This
465: finding could suggest that even for nonperturbative electric
466: fields there could exist long-living resonance states, as it will
467: be discussed in details in \cite{KK2}.
468:
469: The lines in Figs. \ref{fig3} and \ref{fig4} corresponding to the
470: electric-field-induced resonances are, in fact, a narrow loops. In
471: Fig. \ref{fig5} we present in an enlarged scale two such
472: resonances labelled by $\#1$ and $\#2$ in Fig. \ref{fig3}. It is
473: clearly seen that for these particular resonances paths of the
474: real part of resonance energies with a changing electric field
475: represent closed loops which start from a Landau level. For a
476: fixed electric field there are two states which correspond to two
477: different binding energies $\tilde{E}_B$ and, as could be
478: expected, the lower parts of these loops correspond to the
479: stronger attraction, i.e., to the larger values of
480: $|\tilde{E}_B|$.
481:
482: \begin{table}
483: \caption{Physical values of the
484: magnetic field $\cal{B}$, electric field $\cal{E}$ and
485: lifetime $\tau$ for some chosen values of the bound state energy
486: $E_B$ for a particular resonance.}
487: \label{tab1}
488: \begin{ruledtabular}
489: \begin{center}
490: \begin{tabular}{cccc}
491: $E_B$ [meV] & $B$ [T] & $\cal{E}$ [kV/m] & $\tau$ [ns]\\ \hline
492: 1&0.506&6.423&7.52\\
493: 2&1.013&18.167&3.76\\
494: 4&2.025&51.383&1.88\\
495: 6&3.038&94.396&1.25
496: \end{tabular}
497: \end{center}
498: \end{ruledtabular}
499: \end{table}
500:
501: In order to estimate the lifetime of one of these new resonance
502: states let us take $\tilde{\cal{E}}=0.2647$ and
503: $\tilde{E}_B=-2.2860459726451$ for which there exists the
504: resonance of energy $\tilde{E}=3.0703456182811-10^{-4}\text{i}$.
505: This resonance corresponds in Fig. \ref{fig3} to the rightmost
506: point lying near the first excited Landau level. In Table
507: \ref{tab1} we show some numerical values of the magnetic field
508: $\cal{B}$ (in teslas), electric field $\cal{E}$ (in
509: kilovolts per meter) and lifetime $\tau$ (in nanoseconds) for this
510: resonance state and for some typical values of the binding energy
511: $E_B$ of an electron with effective mass $m^*=0.067$ (in the
512: electron's mass unit) trapped by an attractive impurity in the
513: GaAs quantum well \cite{PL90}. Our evaluations show that in spite
514: of applying strong electric fields, which, in general, are
515: expected to destabilize a system, one can observe generation of
516: long-living states with lifetimes comparable to or even larger
517: than typical timescales encountered in semiconductor
518: heterostructures.
519:
520: Let us close this paper discussing the resonances extracted by an
521: attractive impurity and electric field from the third excited
522: Landau level of $n=3$. We fix now the binding energy by assuming
523: that $\tilde{E}_B=-3$ and draw the dependence of the complex
524: resonance energy $\tilde{E}$ on the electric field intensity
525: $\tilde{\cal{E}}$. As we have shown this above, the number of
526: such states is equal to $n+1$ and $n$ of them approach the $n$-th
527: Landau level for the vanishing electric field. The same conclusion
528: can be reached looking at the upper panel of Fig. \ref{fig6},
529: while in the lower panel we present lifetimes of these states. We
530: see that the state 'a', related to the old states discussed before
531: in \cite{P81,CC98}, is very stable for weak electric fields, but
532: with an increasing $\tilde{\cal{E}}$ its lifetime suddenly
533: decreases. On the contrary, the lifetimes of new
534: electric-field-induced resonances, discussed in this paper, are
535: shorter for weak fields, but with the increasing
536: $\tilde{\cal{E}}$ we observe a peculiar behavior for one of
537: them, i.e., for the resonance 'd'; instead of a continuous
538: decrease we observe that for some particular values of
539: $\tilde{\cal{E}}$ the lifetime increases, approaches its
540: maximum value, and then starts decreasing. We call this phenomenon
541: the electric-field-induced stabilization and we relate it to
542: quantum vortices, positions of which are controlled by an electric
543: field. A detailed analysis of this problem will be presented in
544: our subsequent paper \cite{KK2}.
545:
546: \acknowledgments
547: This work has been supported in part by the Polish Committee
548: for Scientific Research (Grant No. KBN 2 P03B 039 19).
549:
550: \begin{thebibliography}{99}
551: \bibitem{P81}
552: R. E. Prange, Phys. Rev. B 23 (1981) 4802.
553: \bibitem{PJ82}
554: R. E. Prange, R. Joynt, Phys. Rev. B 25 (1982) 2943.
555: \bibitem{JP84}
556: R. Joynt, R. E. Prange, Phys. Rev. B 29 (1984) 3303.
557: \bibitem{PC91}
558: F. Perez, F. A. B. Coutinho, Am. J. Phys. 59 (1991) 52.
559: \bibitem{CC98}
560: R. M. Cavalcanti, C. A. A. de Carvalho, J. Phys. A 31 (1998) 2391.
561: \bibitem{GM99}
562: S. Gyger, P. A. Martin, J. Math. Phys. 40 (1999) 3275.
563: \bibitem{HL99}
564: E. H. Hauge, J. M. J. van Leeuwen, Physica A 268 (1999) 525.
565: \bibitem{AGHH88}
566: S. Albeverio, F. Gesztesy, R. H\o egh-Krohn, H. Holden, Solvable
567: Models in Quantum Mechanics, Springer, Heidelberg, 1988.
568: \bibitem{B75}
569: I. J. Berson, J. Phys. B 8 (1975) 3078.
570: \bibitem{FFR90}
571: F. H. M. Faisal, P. Filipowicz, K. Rz\c a\.{z}ewski, Phys. Rev. A
572: (1990) 6176.
573: \bibitem{FFR91}
574: P. Filipowicz, F. H. M. Faisal, K. Rz\c a\.{z}ewski, Phys. Rev. A
575: (1991) 2210.
576: \bibitem{MFSF00}
577: N. L. Manakov, M. V. Frolov, A. F. Starace, I. I. Fabrikant, J.
578: Phys. B 33 (2000) R141.
579: \bibitem{W91}
580: K. W\'odkiewicz, Phys. Rev. A 43 (1991) 68.
581: \bibitem{IZ80}
582: C. Itzykson, J. B. Zuber, Quantum Field Theory, McGraw-Hill, New
583: York, 1980.
584: \bibitem{BBCK}
585: I. Bia\l ynicki-Birula, M. Cieplak, J. Kami\'nski, Theory of
586: Quanta, Oxford University Press, New York, 1992.
587: \bibitem{KK2}
588: K. Krajewska, J. Z. Kami\'nski, in preparation.
589: \bibitem{PL90}
590: T. Pang, S. G. Louie, Phys. Rev. Lett. 65 (1990) 1635.
591: \end{thebibliography}
592:
593: \end{document}
594: