1: \documentclass[prb,aps,preprint]{revtex4}
2: %\documentclass[aps,prb,twocolumn]{revtex4}
3: %\documentclass[aps]{revtex4}
4:
5: \usepackage{graphicx}
6:
7: \begin{document}
8:
9: \title{Dynamics of a nanowire superlattice in an ac electric field}
10:
11: \author{Aizhen Zhang}
12: \author{L. C. Lew Yan Voon}
13: \email{lok.lewyanvoon@wright.edu}
14: \affiliation{Department of Physics, Wright State University, 3640 Colonel Glenn Highway, Dayton, Ohio 45435, USA}
15:
16: \author{M. Willatzen}
17: \affiliation{Mads Clausen Institute, University of Southern Denmark, Grundtvigs All\'{e} 150, DK-6400 S\o nderborg, Denmark}
18:
19: \date{\today}
20:
21: \begin{abstract}
22: With a one-band envelope function theory, we investigate the dynamics of a finite nanowire superlattice driven by an ac electric field by solving numerically the time-dependent Schr\"{o}dinger equation. We find that for an ac electric field resonant with two energy levels located in two different minibands, the coherent dynamics in nanowire superlattices is much more complex as compared to the standard two-level description. Depending on the energy levels involved in the transitions, the coherent oscillations exhibit different patterns.
23: A signature of barrier-well inversion phenomenon in nanowire superlattices
24: is also obtained.
25: \end{abstract}
26:
27: \pacs{73.21.La, 02.60.Cb}
28:
29: \maketitle
30:
31: \section{Introduction}
32:
33: Semiconductor quantum dots (QDs), also called ``artificial atoms,'' are
34: being considered for a variety of technological applications ranging from
35: semiconductor electronics to biological applications including optical
36: devices, quantum computing and biosensors. The study of coherent phenomena
37: is fundamental to a wide range of these applications.
38: For example, a large number of experimental and theoretical investigations of Rabi oscillations (ROs), the
39: analog of atom-light coherent nonlinear interactions, have been
40: performed in the quantum-dot two-level systems in single QDs \cite
41: {Stie,Kama,Htoon,Zrenner,Borri,Beso,Mull,Wang,Li,Ulloa} and double QDs. \cite
42: {Haya,Petta}
43:
44:
45: If we deal with an array of QDs, the discrete levels broaden into energy bands.
46: The question then arises whether coherent oscillations can be observed in such structures.
47: The formation of minibands with discrete states for a system of fifteen coupled
48: QDs in a matrix has been demonstrated.~\cite{Kouwenhoven90}
49: Quantum transport in a model one-dimensional quantum dot array under a dc bias has also been studied
50: theoretically.~\cite{Shangguan01} Another different type of QD array, so-called nanowire superlattices
51: (NWSLs),~\cite{Wu,Gudiksen,Bjork,Solanki02,Wu04} has recently been fabricated using various approaches.
52: A nanowire superlattice consists of a series of interlaced nanodots of two
53: different materials. In NWSLs, the electronic transport along the wire axis
54: is made possible by the tunneling between adjacent dots, while the
55: uniqueness of each quantum dot and its zero-dimensional (0D) characteristics
56: are maintained by the energy difference of the conduction or valence bands
57: between the different materials. The band offset not only provides some amount
58: of quantum confinement, but also creates a periodic potential for carriers
59: moving along the wire axis. This new structure offers unique features and
60: suggests a diversity of possible applications, including nanolasers,
61: nanobarcodes, one-dimensional (1D) waveguides, resonant tunneling diodes,
62: and thermoelectrics.
63:
64: The successful experimental developments of NWSLs have received increasing
65: theoretical attention. Recently considerable work has been devoted to NWSLs.
66: \cite{Lew,Galeriu,Lassen,Lin,Citrin,Madureira,Mizuno}
67: Lew Yan Voon and Willatzen \cite{Lew} have first studied the electron states and optical
68: properties in NWSLs. The thermoelectric properties of NWSLs have been
69: reported by Lin and Dresselhaus.\cite{Lin} Citrin \cite{Citrin} has
70: investigated the magnetic Bloch oscillations in nanowire superlattice rings.
71: Madureira{\sl \ et al.} \cite{Madureira} have proposed that dynamic
72: localization should be observable in NWSLs. A detailed study of acoustic
73: phonons in NWSLs has been given by Mizuno. \cite{Mizuno}
74:
75: With the development of free-electron lasers that can be continuously tuned
76: in the terahertz (THz) range, the dynamics of charged particles in
77: semiconductor nanostructures has been a subject of intensive research.
78: Motivated by the experimental progress and the importance to many applications
79: of NWSLs,
80: in this work, we study the dynamics of a NWSL driven by an ac electric field as
81: a realistic system of a finite chain of coupled QDs. Diez {\sl et al. }\cite
82: {Diez} have investigated the dynamics of a finite semiconductor
83: quantum well superlattice (SL) in an ac electric field and found that Rabi oscillations
84: between minibands are clearly identified under resonant conditions. However,
85: our system is different from SLs in a number of aspects. The coupling of
86: the superlattice longitudinal confinement to nanowire radial confinement in
87: NWSL leads to an additional position-dependent potential for the electron.
88: Furthermore, for a nanowire superlattice, using a
89: position-dependent effective mass
90: leads to qualitatively new physics.~\cite{Lew}
91: Given these unique features,
92: we expect that the dynamics of NWSLs driven by an ac electric field
93: would reveal more interesting results.
94:
95: The paper is organized as follows. In Sec. II we introduce the model and
96: present the theory. In Sec. III, we present the numerical results. A summary
97: is given in Section IV.
98:
99: \section{Theory}
100:
101: In this section we present the theoretical model and approach. The model is
102: essentially the same as the one used in Ref. \onlinecite{Lew}. We briefly describe
103: the model for completeness. The finite NWSL is modeled as an ideal cylinder surrounded by vacuum with
104: sharp modulations in the longitudinal (or $z$ ) direction (Fig.~1). The
105: electronic structure is obtained by solving the BenDaniel-Duke equation
106: \begin{equation}
107: -\frac{\hbar ^2}2\nabla \cdot \left[ \frac 1{m({\bf r})}\nabla \psi ({\bf r}%
108: )\right] +V({\bf r)}\psi ({\bf r})=E\psi ({\bf r}),
109: \end{equation}
110: where $V({\bf r)}$ is the potential experienced by the electron, $m({\bf r})$
111: is the effective mass in each layer, $\psi ({\bf r})$ is the wave function,
112: and $E$ is the energy. This model is appropriate for conduction states and
113: for large band-gap materials when nonparabolicity can be neglected. This
114: partial differential equation is separable in cylindrical circular
115: coordinates, leading to the following set of ordinary differential equations,
116: \begin{equation}
117: \frac{d^2\Phi (\phi )}{d\phi ^2}+l^2\Phi (\phi )=0,
118: \end{equation}
119: \begin{equation}
120: r\frac d{dr}(r\frac{dJ}{dr})+\left( q^2r^2-l^2\right) J(r)=0,
121: \end{equation}
122: \begin{equation}
123: -\frac{\hbar ^2}2\frac d{dz}\left( \frac 1{m\left( z\right) }\frac{dZ\left(
124: z\right) }{dz}\right) +\left[ V\left( z\right) +\frac{\hbar ^2q^2}{2m\left(
125: z\right) }\right] Z\left( z\right) =EZ\left( z\right) ,
126: \end{equation}
127: where $\psi ({\bf r,}\phi ,z)$ $=$ $J(r)\Phi (\phi )Z\left( z\right) .$
128:
129: The solution to the angular Eq. (2) is $\Phi (\phi )=e^{il\phi }$ with $l$
130: an integer. The solution to the radial Eq. (3) is a Bessel function of the
131: first kind, with the wavenumber $q$ determined by the boundary condition $%
132: J_l\left( qR\right) =0,$ where $R$ is the radius of the NWSL. Because the
133: zeros of the Bessel functions are themselves independent of the structure,
134: we only have to solve the longitudinal Eq. (4). Equation (4) describes an
135: electron moving in a one-dimensional effective potential $V^{eff}(z)=V\left(
136: z\right) +\frac{\hbar ^2q^2}{2m\left( z\right) }$. For typical
137: semiconductors, the well mass $m_W$ is smaller than the barrier mass $m_B.$
138: Thus the effective potential can be zero or negative if
139: \begin{equation}
140: R^2\leq \frac{\hbar ^2\alpha ^2\left[ m_B(z)-m_W(z)\right] }{2V_B\left(
141: z\right) m_W(z)m_B(z)}=R_c^2,
142: \end{equation}
143: where $V_B\left( z\right) $ is the real barrier height and $\alpha $ is a
144: zero of the Bessel function. Hence, below a critical radius $R_c,$ the
145: barrier layer acts as the well layer and vice versa. The barrier-well
146: inversion induced by quantum confinement, a unique phenomenon in NWSLs,
147: was predicted in Ref. \onlinecite{Lew}.
148:
149: In order to study the dynamics of the NWSL driven by an ac electric field,
150: we must determine the band structure. At flatband, the band structure of Eq.
151: (4) is computed by using the finite-element method.
152: The eigenstate $j$ of band $i$ with
153: eigenenergy $E_i^{(j)}$ is denoted as $Z_i^{(j)}(z).$ Under an ac electric
154: field, the envelope function for the electron wave packet satisfies the
155: following equation
156: \begin{equation}
157: i\hbar \frac d{dt}\Psi (z,t)=\left\{ -\frac{\hbar ^2}2\frac d{dz}\left(
158: \frac 1{m\left( z\right) }\frac d{dz}\right) +\left[ V\left( z\right) +\frac{%
159: \hbar ^2q^2}{2m\left( z\right) }\right] -eFz\sin (\omega _{ac}t)\right\}
160: \Psi (z,t),
161: \end{equation}
162: where $F$ and $\omega _{ac}$ are the strength and the frequency of the ac
163: field. Given the form of the initial wave function $\Psi (z,0),$ the
164: time-dependent Schr\"{o}dinger Eq. (6) is solved using the finite-difference
165: method and a fourth order Runge-Kutta integration.
166: The mesh spacing is 0.1 nm and the time step is of the order $10^{-6}$ ps,
167: which ensure the accuracy of the numerical calculation. For the sake of
168: simplicity, we have selected $\Psi (z,0)=Z_i^{(j)}\left( z\right) $ as the
169: initial wave packet. The probability of finding the electron in the state $%
170: Z_k^{(l)}\left( z\right) $ is given by
171: \begin{equation}
172: P_{ik}^{(jl)}=\left| \int_{-\infty }^{+\infty }dzZ_k^{(l)}\left( z\right)
173: \Psi (z,t)\right| ^2.
174: \end{equation}
175:
176: Note that, for the case of a two-level system with levels $a$ and $b$ driven
177: by a strong resonant field, the driving field couples levels $a$ and $b$ and
178: induces oscillations of the population in the two levels,
179: the so-called Rabi oscillations. The frequency $\Omega _{R}$ of the ROs is proportional to
180: the electric field amplitude and the dipole transition matrix element, which
181: is given by
182: \begin{equation}
183: \Omega _R=eF_{Thz}\left\langle \psi _a\right| x\left| \psi _b\right\rangle
184: /\hbar =eF_{Thz}x_{ab}/\hbar ,
185: \end{equation}
186: where $F_{Thz}$ is the amplitude of the
187: terahertz electric field, $\left| \psi _a\right\rangle $ and $\left| \psi
188: _b\right\rangle $ are respectively the unperturbed eigenfunctions of levels $%
189: a$ and $b$, and $ex_{ab}$ is the dipole transition matrix element with $e$ the electric charge. If the
190: particle is initially in level $a$, the probability of finding the electron
191: in level $b$ oscillates between $0$ and $1$, i.e, complete resonance occurs.
192:
193: \section{Results}
194:
195: We consider a GaAs/GaAlAs NWSL structure with $10$ periods of $5$ nm GaAlAs
196: and $9$ periods of $10$ nm\ GaAs terminated by a
197: cylindrical 100 nm GaAs layer on each side (Fig.~1).
198: Such a structure can be well described by
199: Eq. (1). In the calculation we use parameters $\alpha =2.405$ (the first
200: root of zeroth order Bessel function), $V_z=230$ meV, $m_B(z)=0.0919m_0,$ and $%
201: m_W(z)$ $=0.067m_0$, where $m_0$ is the free-electron mass. The resulting
202: critical radius $R_c=2$ nm for such a structure. We first choose $R=5$ nm,
203: which is larger than the critical radius. We begin by considering the case
204: of resonant excitation with $F=50$ kV/cm. Figure 2(a) shows $P_{12}^{(55)}(t)
205: $ at the resonant frequency $\omega _{ac}=\left( E_2^{(5)}-E_1^{(5)}\right)
206: /\hbar =127$ THz, i.e., we are monitoring the transition between the
207: central state $j=5$ in the first miniband to the central state $l=5$ in the
208: second miniband. As can be seen, the oscillations reveal the occurrence of
209: coherent oscillations between minibands.
210: This is qualitatively similar to the picture reported for one-dimensional superlattices.~\cite{Diez}
211: Figure 2(b) displays $P_{12}^{(22)}(t)$ at the
212: resonant frequency $\omega _{ac}=\left( E_2^{(2)}-E_1^{(2)}\right) /\hbar
213: =121$ THz, i.e, the transitions between the second state $j=2$ in the first
214: miniband to the second state $l=2$ in the second miniband. There are two
215: main differences in the oscillations in comparison to Fig. 2(a). Firstly,
216: the oscillation peak values are smaller. Secondly, the oscillations have a
217: decreasing amplitude tendency. This is due to the higher probability of the
218: electron tunnelling to the outer GaAs layers when the energy levels are close to
219: the boundary of the actual structure,
220: because in the calculation we have considered an overall structure
221: much larger than the actual NWSL region. In Fig. 2(c)
222: we present $P_{12}^{(56)}(t)$ at the resonant frequency $\omega _{ac}=\left(
223: E_2^{(6)}-E_1^{(5)}\right) /\hbar =131$ THz, i.e., the probability of finding
224: the electron, initially in the central state $j=5$ in the first miniband, in
225: the state $l=6$ in the second miniband at time $t$. We find that coherent oscillations between
226: minibands can also be clearly identified except that the peak values
227: are smaller than in Fig. 2(a).
228: By performing the fast Fourier transform of $%
229: P_{12}^{(55)}(t)$, $P_{12}^{(22)}(t),$ and $P_{12}^{(56)}(t),$ we find that
230: the oscillation frequencies are $15.1$ THz$,$ $19.1$ THz, and $16.0$ THz,
231: respectively.
232:
233: The different oscillation patterns and oscillation frequencies shown in Fig.~2
234: can be understood as follows. The considered nanowire superlattice
235: structure presents Bloch minibands with nine states in each miniband. These
236: states are closely spaced with separations less than 1 meV. When a resonant
237: ac electric field is applied to two energy levels respectively located in
238: the first miniband and the second miniband, a number of energy levels are
239: involved in the transitions. As a result, the probability of finding the
240: electron in the energy level in the second miniband is much less than 1
241: (as shown in Fig. 2). To further check the validity of the two-level description,
242: we calculate the oscillation
243: frequencies corresponding to the three cases shown in Fig. 2
244: using the analytic Eq. (8) of a driven two-level system. The resulting
245: frequencies are $19$ THz, $4.9\times 10^{-2}$ THz, and
246: $2.1\times 10^{-2}$ THz, respectively, which,
247: except for the first one, are very different from the values obtained by
248: performing the fast Fourier transform
249: of $P_{12}^{(55)}(t)$, $P_{12}^{(22)}(t),$ and $%
250: P_{12}^{(56)}(t)$. Thus it is clear that the two-state description is not
251: generally valid for a NWSL. In the three cases shown in Fig. 2, different energy
252: levels and therefore different eigenfunctions are involved in the
253: transitions, leading to different oscillation patterns and oscillation
254: frequencies. Although the occurrence
255: of ROs has been identified in Ref. \onlinecite{Diez},
256: the important features of the
257: sensitivity of ROs to the
258: energy levels involved in the transitions and the inapplicability of two-level
259: theory were not pointed out.
260: Furthermore, because the electron can be driven up to the third miniband, we
261: find that the total probability of finding the electron in the second
262: miniband cannot reach 1 maximally.
263:
264: We now turn to the case of nonresonant excitation.
265: Figures 3(a)--3(c) show the transition probability $P_{12}^{(55)}(t),$ $P_{12}^{(22)}(t),$
266: and $P_{12}^{(56)}(t)$ with $\omega _{ac}=100$ THz and $F=50$ kV/cm
267: respectively. In contrast to the resonant situation, one can see that the
268: oscillation amplitudes are significantly decreased in all three cases when
269: the ac driving field is out of resonance.
270:
271: Next we discuss the barrier-well inversion induced by quantum confinement
272: predicted in Ref. \onlinecite{Lew}. When $R< R_c,$ the barrier becomes the
273: well and vice versa. For the GaAs/GaAlAs NWSL structure discussed here, the
274: barrier width increases if the barrier-well inversion occurs, resulting in a
275: narrower band width and a larger band gap. The energy levels in the
276: minibands become more closely spaced and the wavefunctions change
277: dramatically. To compare with the case of $R> R_c$, in Fig. 4 we show
278: the calculated $P_{12}^{(55)}(t)$ with $R=1$ nm (less than critical radius).
279: Other parameters are the same as used in Fig. 2(a). As expected, the
280: oscillation pattern is totally different from Fig. 2(a) due to different
281: band structures. Both the amplitudes and the frequency are found to be
282: greatly decreased. Therefore the barrier-well inversion phenomenon is verified
283: by the transport properties presented here.
284:
285: \section{Summary}
286:
287: In this paper, within a one-band envelope function theory
288: and by means of
289: numerically solving the time-dependent Schr\"{o}dinger equation, we have
290: presented an analysis of the dynamics of a finite NWSL driven by an ac
291: electric field.
292: We have found that for the case of resonant ac electric field, the coherent dynamics
293: is much more complex as compared to the standard two-level description.
294: The oscillations are extremely sensitive to
295: the energy levels involved in the transitions. As a result, the oscillation
296: patterns are very different when the applied ac electric field is resonant
297: with two different energy levels respectively located in the first and second minibands.
298: Coherent oscillations between minibands can be identified when the probability
299: of the tunneling of the electron to the surrounding region is negligible. On
300: the other hand, the oscillations can also exhibit a decreasing amplitude
301: tendency due to the greater probability of the electron tunneling to the surrounding region. In
302: comparison to the resonant case, the oscillation amplitudes are
303: significantly decreased when the ac electric field is out of resonance. We
304: have also verified the existence of barrier-well inversion phenomena due to
305: the coupling of the superlattice longitudinal confinement to nanowire radial
306: confinement in NWSLs. This indicates that the radius of NWSL should be kept
307: in mind for investigating a variety of phenomena.
308:
309: Nanowire superlattices do have some advantages
310: over other nanostructures: (i) scattering events are
311: highly suppressed because of the spatial confinement of carriers along all
312: three directions, leading to a long decoherence time, (ii) the wire
313: structure lacks the transverse excitations present in quantum-well superlattice
314: structures, eliminating this source of damping, and (iii) this novel
315: structure has no wetting layer (WL) compared with
316: previously-studied single and double QDs.
317: These make NWSLs potentially more favorable for observations of coherent phenomena.
318: However, due to the structural complexity and the materials diversity in
319: these nanostructures, for practical applications and device optimization it
320: is essential to develop appropriate models to understand the behavior and to
321: predict properties of interest in these novel structures. Work along these
322: lines is currently in progress.
323:
324: \begin{acknowledgments}
325: The work was supported by an NSF CAREER award (NSF Grant No. 0454849),
326: and by a Research Challenge grant from Wright State University and the Ohio
327: Board of Regents.
328: \end{acknowledgments}
329:
330:
331: %\bibliography{WZcyl}
332: %\bibliographystyle{apsrev}
333:
334: \newpage
335:
336: \begin{thebibliography}{30}
337: \bibitem{Stie} T. H. Stievater, X. Li, D. G. Steel, D. Gammon, D. S. Katzer,
338: D. Park, C. Piermarocchi, and L. J. Sham, Phys. Rev. Lett. {\bf 87}, 133603 (2001).
339:
340: \bibitem{Kama} H. Kamada, H. Gotoh, J. Temmyo, T. Takagahara, and H. Ando,
341: Phys. Rev. Lett. {\bf 87}, 246401 (2001).
342:
343: \bibitem{Htoon} H. Htoon, T. Takagahara, D. Kulik, O. Baklenov, A. L.
344: Holmes, Jr., and C. K. Shih, Phys. Rev. Lett. {\bf 88}, 087401 (2002).
345:
346: \bibitem{Zrenner} A. Zrenner, E. Eham, S. Stufler, F. Findeis, M. Bichler,
347: and G. Abstreiter, Nature (London) {\bf 418}, 612 (2002).
348:
349: \bibitem{Borri} P. Borri, W. Langbein, S. Schneider, U. Woggon, R. L.
350: Sellin, D. Ouyang, and D. Bimberg, Phys. Rev. B {\bf 66}, 081306(R) (2002).
351:
352: \bibitem{Beso} L. Besombes, J.J. Baumberg, and J. Motohisa, Phys. Rev.
353: Lett. {\bf 90}, 257402 (2003).
354:
355: \bibitem{Mull} A. Muller, Q.Q. Wang, P. Biancci, C.K. Shih, and Q. K. Cue,
356: Appl. Phys. Lett. {\bf 84}, 981 (2004).
357:
358: \bibitem{Wang} Q. Q. Wang, A. Muller, P. Bianucci, E. Rossi, Q. K. Xue, T.
359: Takagahara, C. Piermarocchi, A. H. MacDonald, and C. K. Shih, Phys. Rev. B {\bf 72}, 035306 (2005).
360:
361: \bibitem{Li} X. Li, Y. Wu, D. Steel, D. Gammon, T.H. Stievater, D.S.
362: Katzer, D. Park, C. Piermarocchi, and L.J. Sham, Science {\bf 301}, 809 (2003).
363:
364: \bibitem{Ulloa} J. M. Villas-B\^{o}as, Sergio E. Ulloa, and A. O. Govorov, Phys. Rev. Lett. {\bf 94}, 057404 (2005).
365:
366: \bibitem{Haya} T. Hayashi, T. Fujisawa, H. D. Cheong, Y. H. Jeong, and Y.
367: Hirayama, Phys. Rev. Lett. {\bf 91}, 226804 (2003).
368:
369: \bibitem{Petta} J. R. Petta, A. C. Johnson, C. M. Marcus, M. P. Hanson, and
370: A.C. Gossard, Phys. Rev. Lett. {\bf 93}, 186802 (2004).
371:
372: \bibitem{Kouwenhoven90}
373: L. P. Kouwenhoven, F. W. J. Hekking, B. J. van Wees, C. J. P. M. Harmans,
374: C. E. Timmering, and C. T. Foxon,
375: Phys. Rev. Lett. {\bf 65}, 361 (1990).
376:
377: \bibitem{Shangguan01}
378: W. Z. Shangguan, T. C. Au Yeung, Y. B. Yu, and C. H. Kam,
379: Phys. Rev. B {\bf 63}, 235323 (2001).
380:
381: \bibitem{Wu} Y. Wu, R. Fan, and P. Yang, Nano Lett. {\bf 2}, 83 (2002).
382:
383: \bibitem{Gudiksen} M. S. Gudiksen, L. J. Lauhon, J. Wang, D. C. Smith, and
384: C. M. Lieber, Nature (London) {\bf 415}, 617 (2002).
385:
386: \bibitem{Bjork} M. T. Bj\"{o}rk, B. J. Ohlsson, T. Sass, A. I. Persson, C.
387: Thelander, M. H. Magnusson, K. Deppert, L. R. Wallenberg, and L. Samuelson,
388: Nano Lett. {\bf 2}, 87 (2002).
389:
390: \bibitem{Solanki02}
391: R. Solanki, J. Huo, J. L. Freeouf, and B. Miner,
392: Appl. Phys. Lett. {\bf 81}, 3864 (2002).
393:
394: \bibitem{Wu04}
395: Z. H. Wu, M. Sun, X. Y. Mei, and H. E. Ruda,
396: Appl. Phys. Lett. {\bf 85}, 657 (2004).
397:
398:
399: \bibitem{Lew} L. C. Lew Yan Voon and M. Willatzen,
400: J. Appl. Phys. {\bf 93}, 9997 (2003).
401:
402: \bibitem{Galeriu} C. Galeriu, L. C. Lew Yan Voon, R. Melnik, M. Willatzen,
403: Comput. Phys. Comm. {\bf 157}, 147 (2004).
404:
405: \bibitem{Lassen} L. C. Lew Yan Voon, B. Lassen, R. Melnik, and M.
406: Willatzen, J. Appl. Phys. {\bf 96}, 4660 (2004).
407:
408: \bibitem{Lin} Yu-Ming Lin and M. S. Dresselhaus, Phys. Rev. B {\bf 68}, 075304 (2003).
409:
410: \bibitem{Citrin} D. S. Citrin, Phys. Rev. Lett. {\bf 92}, 196803 (2004).
411:
412: \bibitem{Madureira} Justino R. Madureira, Peter A. Schulz, Marcelo Z.
413: Maialle, Phys. Rev. B {\bf 70}, 033309 (2004).
414:
415: \bibitem{Mizuno} S. Mizuno, Phys. Rev. B {\bf 71}, 085303 (2005).
416:
417: \bibitem{Diez} Enrique Diez, Rafael G\'{o}mez-Alcala, Francisco
418: Dom\'{\i}nguez-Adame, Angel S\'{a}nchez, and Gennady P. Berman,
419: Phys. Rev. B {\bf 58}, 1146 (1998).
420:
421: \end{thebibliography}
422:
423: \newpage
424:
425: \begin{figure}
426: \includegraphics[width=10cm]{fig1.eps}
427: \caption{\label{fig:fig1}
428: Nanowire superlattice with radius $R$, well width $L_W$, barrier
429: width $L_B,$ and potential energy $V$.
430: The wire is capped on both sides by a thick cylindrical layer of GaAs.
431: }
432: \end{figure}
433:
434: \begin{figure}
435: \includegraphics[width=10cm]{fig2.eps}
436: \caption{\label{fig:fig2}
437: (a) Temporal evolution of $P_{12}^{(55)}(t)$ with $\omega
438: _{ac}=\left( E_2^{(5)}-E_1^{(5)}\right) /\hbar =127$ THz. (b) Temporal
439: evolution of $P_{12}^{(22)}(t)$ with $\omega _{ac}=\left(
440: E_2^{(2)}-E_1^{(2)}\right) /\hbar =121$ THz. (c) Temporal evolution of $%
441: P_{12}^{(56)}(t)$ with $\omega _{ac}=\left( E_2^{(6)}-E_1^{(5)}\right)
442: /\hbar =131$ THz. In all cases, $F=50$ kV/cm and $R=5$ nm.
443: }
444: \end{figure}
445:
446: \begin{figure}
447: \includegraphics[width=10cm]{fig3.eps}
448: \caption{\label{fig:fig3}
449: Temporal evolution of (a) $P_{12}^{(55)}(t),$ (b) $P_{12}^{(22)}(t),$
450: and (c) $P_{12}^{(56)}(t)$ with $F=50$ kV/cm, $\omega _{ac}=100$ THz, and $%
451: R=5$ nm.
452: }
453: \end{figure}
454:
455: \begin{figure}
456: \includegraphics[width=10cm]{fig4.eps}
457: \caption{\label{fig:fig4}
458: Temporal evolution of $P_{12}^{(55)}(t)$ with $\omega _{ac}=\left(
459: E_2^{(5)}-E_1^{(5)}\right) /\hbar =127$ THz, $F=50$ kV/cm, and $R=1$ nm.
460: }
461: \end{figure}
462:
463: \end{document}
464: