1: \documentclass[12pt]{iopart}

2: 

3: \usepackage{iopams}

4: \usepackage{epsf}

5: \usepackage{graphicx}

6: \usepackage{fancyhdr}

7: 

8: \newcommand{\dtot}[2] { \frac{d {#1} } {d {#2}} }

9: \newcommand{\ndtot}[3] { \frac{ d^{#3} {#1} } {d #2 ^{#3} }}

10: \newcommand{\dpar}[2] { \frac{\partial {#1} } {\partial {#2}} }

11: \newcommand{\ndpar}[3] { \frac{\partial^{#3} {#1} } {\partial #2 ^{#3} }}

12: \newcommand{\dparh}[2] { {\partial {#1} }/ {\partial {#2}} }

13: \newcommand{\dpard}[3] { \frac{\partial^2 {#1} } {\partial {#2} \partial {#3}} }

14: \newcommand{\OL}[1] {\textrm{\emph{\large O}}\!\left({#1}\right)}

15: \newcommand{\eqref}[1] {(\ref{#1})}

16: \renewcommand{\Re}[1] {\mbox{Re}(#1)}

17: \renewcommand{\Im}[1] {\mbox{Im}(#1)}

18: \newcommand{\II}[0] {\mbox{i}}

19: \newcommand{\Complejos}[0] {\mathbb{C}}

20: \newcommand{\sign}[0] {\mbox{sign}\,}

21: \newcommand{\text}[1] {\mbox{#1}}

22: \newcommand{\arccot}[0] {\mbox{arccot}\,}

23: \newcommand{\dis}[0] {\displaystyle}

24: \newcommand{\fracdis}[2] { \frac {\dis  {#1} } {\dis {#2}} }

25: \newcommand{\abs}[1]{\left| #1 \right|}

26: 

27: \begin{document}

28: 

29: \title[A transfer matrix method for the analysis of fractal quantum

30: potentials] {A transfer matrix method for the analysis of fractal

31: quantum potentials}

32: 

33: \author{Juan A. Monsoriu$^1$\footnote[7]{To whom

34: correspondence should be addressed (jmonsori@fis.upv.es)},

35: Francisco R. Villatoro$^2$, Mar{\'\i}a J. Mar{\'\i}n$^3$, Javier

36: F. Urchuegu{\'\i}a$^1$ and Pedro Fern\'andez de C\'ordoba$^4$}

37: 

38: \address{$^1$ Departamento de F{\'\i}sica Aplicada,

39:  Universidad Polit\'ecnica de Valencia, E-46022 Valencia, Spain}

40: 

41: \address{$^2$ Departamento de Lenguajes y Ciencias de la Computaci\'on,

42:   Universidad de M\'alaga, E-29071 M\'alaga, Spain}

43: 

44: \address{$^3$ Departamento de Termodin\'amica,

45:  Universitat de Val\`{e}ncia, E-46100 Burjassot, Spain}

46: 

47: \address{$^4$ Departamento de Matem\'atica Aplicada,

48:  Universidad Polit\'ecnica de Valencia, E-46022 Valencia, Spain}

49: 

50: 

51: 

52: \begin{abstract}

53: The scattering properties of quantum particles on fractal

54: potentials at different stages of fractal growth are obtained by

55: means of the transfer matrix method. This approach can be easily

56: adopted for project assignments in introductory quantum mechanics

57: for undergraduates. The reflection coefficients for both the

58: fractal potential and the finite periodic potential are calculated

59: and compared. It is shown that the reflection coefficient for the

60: fractal has a self-similar structure associated with the fractal

61: distribution of the potential.

62: \end{abstract}

63: 

64: \pacs{03.65.Nk, 05.45.Df}

65: 

66: \submitto{\EJP}

67: 

68: \maketitle

69: 

70: \section{Introduction}

71: \label{Intro}

72: 

73: Both quantum mechanics and elementary solid state physics courses

74: illustrate the energy band structure in solids through the

75: one-dimensional Kronig-Penney model that consists of a periodic

76: configuration of square-well

77: potentials~\cite{KronigPenney,Liboff}. This problem is usually

78: solved by matching the boundary conditions of the wavefunctions at

79: the cell boundaries, thus requiring the computation of the

80: determinant of a $4\times 4$ matrix~\cite{Szmulowicz}. Recently,

81: some less tedious approaches have been proposed which usually can

82: be readily adapted to finite periodic potentials~\cite{Sprung}.

83: Among these methods, those based on the transfer matrix approach

84: which only uses $2\times 2$ matrix operations in a purely

85: algebraic way are the most appropriate ones for

86: beginners~\cite{Griffiths}. Moreover, this method allows to

87: introduce a numerical method based on a piecewise constant

88: approximation~\cite{Kalotas} for a general potential, the analysis

89: of defects on slightly aperiodic potentials, and even the

90: consideration of more complicated potentials. Among the last,

91: fractal potentials is the one which we considered here.

92: 

93: In recent years the study of fractals has attracted much attention

94: because many physical phenomena, natural structures and

95: statistical processes can be analyzed and described by using a

96: fractal approach~\cite{Mandelbrot,Ficker}. From a mathematical

97: point of view, fractals are self-similar structures obtained by

98: performing a basic operation, called {\it generator}, on a given

99: geometrical object called {\it initiator}, and repeating this

100: process on multiple levels; in each one of them, an object

101: composed of sub-units of itself is created that resembles the

102: structure of the whole object. Mathematically, this property

103: should hold in all scales. However, in the real world, there are

104: lower and upper bounds over which such self-similar behavior

105: applies. Fractals are becoming a useful tool to be able to model

106: diverse physical systems~\cite{Berry,Karman}, and have new

107: technological applications~\cite{Saavedra,Monsoriu}.

108: 

109: In non-relativistic quantum mechanics, fractals have been used to

110: generate new solutions of the Schr\"odinger equation which are

111: continuous but nowhere differentiable wave

112: functions~\cite{Wojcik}, and models for the so-called fractal

113: potentials~\cite{Albeverio}. Fractal potentials allow the analysis

114: of quasi-periodic and nearly stochastic potentials using the

115: symmetries induced by the self-similar structure of the potential.

116: Here we consider the simplest fractal, the (triadic) Cantor set,

117: as a fractal potential for quantum scattering~\cite{Makarov} and

118: tunnelling~\cite{Chuprikov1,Chuprikov2}.

119: 

120: In this paper, we present a simple transfer matrix method to

121: obtain the scattering properties of Cantor set fractal potentials,

122: which can easily be automated by computers. The present method

123: makes easier for the comparison with the finite periodic case and

124: shows how the reflection coefficient for the fractal case has a

125: self-similar structure associated with the fractal distribution of

126: the potential. Moreoever, it can be easily implemented in any

127: computer language, e.g., the \emph{Mathematica} software package,

128: accesible to undergraduate students with only a basic programming

129: experience, so that it can be adopted for project assignments in

130: computer physics courses. Starting with the implementation of the

131: transfer matrix method for a potential barrier, the extension to

132: finite periodic potential is a straightforward one; its extension

133: to Cantor set potentials can be based on a recursive

134: implementation, involving the possible improvement of student

135: programming skills. Furthermore, these projects can introduce the

136: students to the analysis of computational complexity of

137: algorithms, since the Cantor set prefractal has a large number of

138: potential barriers and its simulation requires an exponential

139: number of matrix products.

140: 

141: This paper is organized as follows: The next section describes the

142: main facts about the transfer matrix method for quantum scattering

143: implemented by piecewise constant potentials. In Section~3, the

144: reflection coefficient for the tunnelling on both the finite

145: periodic potential as well as the Cantor set pre-fractals are

146: determined and compared. Finally, the last section is devoted to

147: the conclusions.

148: 

149: 

150: \section{The transfer matrix method in quantum scattering}

151: \label{Problem}

152: 

153: \begin{figure}

154:  \begin{center}

155:   \epsfbox{fig1.eps}

156:  \end{center}

157:  \caption{\label{figuraMatrix}(a) Local scattering with the

158:  $i$-th interface of the piecewise constant potential

159:  among the values $V_{i-1}$ and~$V_{i}$. (b) Wave

160:  propagation through the constant potential $V_i$ as

161:   used to calculate the propagation matrix.}

162: \end{figure}

163: 

164: Let us consider the one-dimensional, steady-state, linear

165: Schr\"odinger equation

166: \begin{equation}

167:  \label{LSE}

168:  -\frac{\hbar^2}{2\,m}\,\ndpar{\psi(x)}{x}{2} + V(x)\,\psi(x) =

169:  E\,\psi(x),

170: \end{equation}

171: where $\psi(x)$, $m$ and $E$ are the wavefunction, mass and energy

172: of the particle, respectively. The constant $\hbar$ is Planck's

173: constant, and $V(x)$ is the quasiperiodic potential which can be

174: represented by a piecewise constant function.

175: Figure~\ref{figuraMatrix}(a) shows the quantum scattering at the

176: $i$-th interface between two successive constant values of the

177: piecewise potential, whose position, without loss of generality,

178: has been taken as $x=0$. In this figure, both $\psi^+_i$ and

179: $\psi^-_i$ are forward and backward plane wavefunctions,

180: respectively, on the region where the potential value is $V_i$,

181: and $\psi_i=\psi^+_i+\psi^-_i$. These wave functions are given by

182: 

183: \begin{equation}

184:  \psi^\pm_i = A^\pm_i\,e^{\dis  \pm\II\,k_i\,x},

185: \end{equation}

186: where $ k_i = \frac{1}{\hbar}\,\sqrt{2\,m\,(E-V_i)}$ is the local

187: particle momentum, and $A^\pm_i$ are integration constants to be

188: determined by applying the standard boundary conditions at the

189: interface. The continuity of the wavefunctions and the derivatives

190: at the boundary are given by

191: \begin{equation}

192:  \label{eq:Three}

193:  \eqalign{

194:  \psi_{i-1}(x=0)=\psi_{i}(x=0), \qquad

195:    A^+_{i-1} + A^-_{i-1} = A^+_i + A^-_i,\\

196:  \psi'_{i-1}(x=0)=\psi'_{i}(x=0), \qquad

197:    k_{i-1}\,A^+_{i-1} - k_{i-1}\,A^-_{i-1} = k_{i}\,A^+_i -

198:    k_{i}\,A^-_i,}

199: \end{equation}

200: where the prime denotes differentiation. Eq.~\eqref{eq:Three} is a

201: linear system of equations which can be written in matrix notation

202: as

203: \begin{equation}

204:  \left(\begin{array}{cc}

205:     1 &  1 \\ k_{i-1} & -k_{i-1}

206:  \end{array}\right)\,

207:  \left(\begin{array}{c}

208:     A^+_{i-1} \\ A^-_{i-1}

209:  \end{array}\right)

210:       =

211:  \left(\begin{array}{cc}

212:     1 &  1 \\ k_{i} & -k_{i}

213:  \end{array}\right)\,

214:  \left(\begin{array}{c}

215:     A^+_{i} \\ A^-_{i}

216:  \end{array}\right),

217: \end{equation}

218: and yielding

219: \begin{equation}

220:  \left(\begin{array}{c}

221:     A^+_{i-1} \\ A^-_{i-1}

222:  \end{array}\right)

223:     =

224:  D_{i-1}^{-1}\,D_{i}\,

225:  \left(\begin{array}{c}

226:     A^+_{i} \\ A^-_{i}

227:  \end{array}\right),

228:  \qquad

229:  D_i =

230:  \left(\begin{array}{cc}

231:     1 &  1 \\ k_{i} & -k_{i}

232:  \end{array}\right).

233: \end{equation}

234: Here on, the matrix $D_{i-1}^{-1}\,D_{i}$ is referred to as the

235: wave scattering matrix.

236: 

237: After crossing the $i$-th interface, the plane wave propagates

238: through the constant potential $V_i$ until it finds the next

239: interface at a distance $d_i$. Using the notation shown in

240: Figure~\ref{figuraMatrix}(b), this wavefunction is given by

241: \begin{equation}

242:  \widetilde{\psi}^\pm_i = A^\pm_i\,e^{\dis  \pm\II\,k_i\,d_i}\,e^{\dis  \pm\II\,k_i\,x}

243:  = \widetilde{A}^\pm_i\,e^{\dis  \pm\II\,k_i\,x},

244: \end{equation}

245: and a wave propagation matrix $P_i$ can be defined as

246: \begin{equation}

247:  \left(\begin{array}{c}

248:     \widetilde{A}^+_{i-1} \\ \widetilde{A}^-_{i-1}

249:  \end{array}\right)

250:     =

251:  \left(\begin{array}{cc}

252:     e^{\dis  \II\,k_i\,d_i} & 0 \\

253:     0  & e^{\dis -\II\,k_i\,d_i}

254:  \end{array}\right)\,

255:  \left(\begin{array}{c}

256:     A^+_{i} \\ A^-_{i}

257:  \end{array}\right)

258:    = P_i\,

259:  \left(\begin{array}{c}

260:     A^+_{i} \\ A^-_{i}

261:  \end{array}\right).

262: \end{equation}

263: \begin{figure}

264:  \begin{center}

265:   \epsfbox{fig2.eps}

266:  \end{center}

267:  \caption{\label{figuraPiecewisePotential}Piecewise constant potential with

268:   $N$ potential wells $V_i$ with $d_i$ as the corresponding width.

269:   $V_0$ and $V_{N+1}$ are the

270:   surrounding constant potential values extended to infinity.}

271: \end{figure}

272: Both the scattering and propagation matrices can be used to solve

273: the general problem of the scattering with a piecewise constant

274: potential with $N$ potential wells, as shown in

275: Figure~\ref{figuraPiecewisePotential}. The successive application

276: of the scattering and propagation matrices yield

277: \begin{equation}

278:  \left(\begin{array}{c}

279:     A^+_{0} \\ A^-_{0}

280:  \end{array}\right)

281:    = D_{0}^{-1}\,D_{1}\,

282:  \left(\begin{array}{c}

283:     A^+_{1} \\ A^-_{1}

284:  \end{array}\right)

285:    = D_{0}^{-1}\,D_{1}\,P_1\,D_{1}^{-1}\,D_{2}\,

286:  \left(\begin{array}{c}

287:     A^+_{2} \\ A^-_{2}

288:  \end{array}\right),

289: \end{equation}

290: and, in the most general form,

291: \begin{equation}

292:  \left(\begin{array}{c}

293:     A^+_{0} \\ A^-_{0}

294:  \end{array}\right)

295:    = M\,

296:  \left(\begin{array}{c}

297:     A^+_{N+1} \\ A^-_{N+1}

298:  \end{array}\right),

299:  \qquad

300:  M = D_{0}^{-1}\,

301:    \left(

302:     \prod_{i=1}^N

303:        D_{i}\,P_i\,D_{i}^{-1}

304:     \right)\,D_{N+1}.

305: \end{equation}

306: Both the reflection and transmission coefficients of the

307: scattering of a quantum particle, incoming from the left, with the

308: $N$-well potential is determined by the coefficients of the matrix

309: $M$,

310: \begin{equation}

311:  \left(\begin{array}{c}

312:     A^+_{0} \\ A^-_{0}

313:  \end{array}\right)

314:    = \left(\begin{array}{cc}

315:     M_{11} & M_{12} \\

316:     M_{21} & M_{22}

317:  \end{array}\right) \,

318:  \left(\begin{array}{c}

319:     A^+_{N+1} \\ 0

320:  \end{array}\right),

321: \end{equation}

322: where no backward moving particle can be found on the right side

323: of the potential, so $A^-_{N+1}=0$. The reflection and the

324: transmission coefficients~\cite{Liboff,Pedro} are given by

325: \begin{equation}

326:  R = \frac{\abs{A^-_{0}}^2}{\abs{A^+_{0}}^2} =

327:   \frac{\abs{M_{21}}^2}{\abs{M_{11}}^2},

328:  \qquad \mbox{and} \qquad

329: T = \frac{k_{N+1}\,\abs{A^+_{N+1}}^2}{k_{0}\,\abs{A^+_{0}}^2} =

330:   \frac{k_{N+1}}{k_{0}\,\abs{M_{11}}^2},

331: \end{equation}

332: respectively.

333: 

334: \begin{figure}

335:  \begin{center}

336:   \epsfbox{fig3.eps}

337:  \end{center}

338:  \caption{\label{figuraPeriodicCantor}Finite periodic (a) and Cantor set quasiperiodic (b) potentials

339:   where the white and black regions denote the potential values $0$ and $\mathcal{V}$,

340:   respectively.}

341: \end{figure}

342: 

343: \section{Presentation of results}

344: \label{Results}

345: 

346: The simplest fractal potential is the Cantor set, shown in

347: Figure~\ref{figuraPeriodicCantor}(b), which can be obtained by

348: means of an iterative construction. The first step ($S=0$) is to

349: take a segment of unit length. The next one ($S=1$) is to divide

350: the segment in three equal parts of length $1/3$ and remove the

351: central one. In general, at the stage $S$, there are $2^S$

352: segments of length $3^{-S}$ with $2^S-1$ gaps in between. Stage

353: $S+1$ is obtained by dividing each of these segments into three

354: parts of length $3^{-S-1}$ and removing the central ones. In

355: Figure~\ref{figuraPeriodicCantor}(b), only the four first stages

356: are shown for clarity. Note that the $S$-th stage Cantor set

357: pre-fractal can be interpreted as a quasiperiodic distribution of

358: segments which can be obtained by removing some segments in a

359: finite periodic distribution as shown in

360: Figure~\ref{figuraPeriodicCantor}(a). This distribution at stage

361: $p_M$ has $(3^M-1)/2+1$ potential barriers of length $3^{-M}$,

362: separated by potential wells of the same length, so the ``period"

363: of this finite structure is $\Lambda=2\,\cdot\,3^{-M}$.

364: 

365: The scattering problem for both the quasiperiodic, Cantor set,

366: pre-fractal potential, and the finite periodic potential can be

367: easily solved by means of the matrix transfer theory presented in

368: Sec. 2. It is standard to normalize both the energy and the height

369: of the potential barrier by the period $\Lambda$, introducing the

370: non-dimensional variables

371: \[

372:  \phi=\Lambda\,\frac{\sqrt{2\,m\,E}}{\hbar}, \qquad \mbox{and} \qquad

373:  \phi_\mathcal{V}=\Lambda\,\frac{\sqrt{2\,m\,\mathcal{V}}}{\hbar}.

374: \]

375: 

376: Figures~\ref{figuraRforPeriodic} and~\ref{figuraRforCantor} show

377: the reflection coefficient, $R$, for the finite periodic potential

378: and Cantor set fractal potential, respectively, around the

379: interval which contains the first band gap of the infinite

380: periodic one. Using the standard Kronig-Penney

381: model~\cite{KronigPenney2}, this band gap can be numerically

382: calculated yielding $3.2519 < \phi < 3.6222$ for the potential

383: $\phi_\mathcal{V}=2$.

384: \begin{figure}

385:  \begin{center}

386:   \epsfbox{fig4.eps}

387:  \end{center}

388:  \caption{\label{figuraRforPeriodic}Scattering reflection coefficient for the

389:   finite periodic potentials of stages $p_2$ (a), $p_3$ (b),

390:   and $p_4$ (c) as a function of the normalized energy $\phi$

391:   for the potential $\phi_\mathcal{V}=2$.}

392: \end{figure}

393: In this energy interval, a Bloch wavefunction does not propagate

394: in a infinite periodic potential and, therefore, the transmission

395: coefficient should vanish ($R=1$). Only evanescent wavefunctions

396: characterized by a complex wavevector, $k$, are solutions of the

397: Schr\"odinger equation. For this reason, when the number of

398: periods is finite, the quantum particle may pass through the

399: potential distribution by the tunnelling effect.

400: \begin{figure}

401:  \begin{center}

402:   \epsfbox{fig5.eps}

403:  \end{center}

404:  \caption{\label{figuraRforCantor}Scattering reflection coefficient for the

405:   Cantor set pre-fractal potentials of stages $S=2$ (a), $S=3$ (b),

406:   and $S=4$ (c) as a function of the normalized energy $\phi$

407:   for the potential $\phi_\mathcal{V}=2$.}

408: \end{figure}

409: Figure~\ref{figuraRforPeriodic} shows that the reflection

410: coefficient approaches unity as the number of periods in the

411: spatial interval increases, illustrating the process of appearance

412: of the band gap of the (full) periodic structure. Although, at the

413: graphical resolution of Figure~\ref{figuraRforPeriodic}(c) the

414: value $R=1$ is apparently reached, the reflection coeficient is

415: always smoller than unity in finity peridodic structures.

416: 

417: Figure~\ref{figuraRforCantor} shows the reflection coefficient for

418: the Cantor set pre-fractal potential for $S=2$ (top), $S=3$

419: (middle) and $S=4$ (bottom). It is shown that the reflection at

420: each higher stage is a modulated version of that associated with

421: the previous stage. That is, the reflection spectrum exhibits a

422: characteristic fractal profile that reproduces the self-similarity

423: of the potential distribution. In fact, any wide peak at stage $S$

424: is transformed into three narrower and taller peaks at stage

425: $S+1$. Zero reflection from these fractal quantum potential occurs

426: at specific discrete energies, while near total reflection is

427: possible at other discrete energies. Comparing

428: Figures~\ref{figuraRforPeriodic} and~\ref{figuraRforCantor}, an

429: increasing number of zeros inside the band gap is observed. These

430: zeros represent resonances due to the presence of ``defects" in

431: the quasiperiodic potential obtained by removing some segments in

432: the finite periodic sequence.

433: 

434: 

435: 

436: \section{Conclusions}

437: \label{Conclusions}

438: 

439: The transfer matrix method is becoming the standard method for the

440: calculation of the tunnelling of quantum particles on constant

441: piecewise potentials because it can be used for simple,

442: textbook-like problems and as a numerical method for computer

443: simulations. This procedure has been applied to Cantor set fractal

444: potentials, which are constant value potentials with support on a

445: Cantor set. For pre-fractals, the $S$-th stage fractal, the

446: reflection coefficient was numerically calculated and compared

447: with that of a finite periodic potential of the same period. The

448: appearance of the first band gap of the Kronig-Penney model in the

449: finite periodic potential has been illustrated. The reflection

450: coefficient for the Cantor set potential is self-similar.

451: 

452: The transfer matrix method presented in this paper can be easily

453: adopted in computer laboratories for undergraduate quantum

454: mechanics courses, providing a powerful method for developing

455: students skill on physics by means of computational tools.

456: Furthermore, fractal geometry is a highly motivating topic for the

457: students providing a great opportunity to undertake projects

458: closely related to research ones.

459: 

460: \ack{The authors are thankful to Prof. Sarira Sahu from the

461: Instituto de Ciencias Nucleares at the Universidad Aut\'onoma de

462: M\'exico, M\'exico, and Prof. Juan I. Ramos from the Universidad

463: de M\'alaga, Spain, for their valuable comments and suggestions.

464: J.A. Monsoriu and P. Fern\'andez de C\'ordoba were supported by

465: the Plan Nacional I+D+I under project TIC 2002-04527-C02-02

466: (Spain). F.R. Villatoro was supported by Project BFM2001-1902 from

467: the Direcci\'on General de Investigaci\'on, Ministerio de Ciencia

468: y Tecnolog\'{\i}a, Spain. Part of this work was done during the

469: visit of J.A. Monsoriu to the Universidad de M\'alaga with a grant

470: from the Universidad Polit\'ecnica de Valencia, under the

471: ``Programa de Incentivo a la Investigaci\'on de la UPV 2004".}

472: 

473: 

474: %\bookref{authors,year,title,editors}

475: \newcommand{\bookref}[4]{#1 #2 {\em {#3}} (#4)}

476: 

477: %\paperref{authors,year,journal,vol,pp}

478: \newcommand{\paperref}[5]{#1 #2 {\em {#3}} {\bf #4} #5}

479: 

480: \Bibliography{19}

481: 

482: 

483:  \bibitem{KronigPenney} \bookref

484:  {Kittel C}

485:  {1996}

486:  {Introduction to Solid State Physics}

487:  {Wiley, New York}

488: 

489:  \bibitem{Liboff} \bookref

490:  {Liboff R}

491:  {2003}

492:  {Introductory Quantum Mechanics}

493:  {Benjamin Cummings, Redwood City, CA}

494: 

495:  \bibitem{Szmulowicz} \paperref

496:  {Szmulowicz F}

497:  {1997}

498:  {Eur. J. Phys.}

499:  {18} {392}

500: 

501:  \bibitem{Sprung} \paperref

502:  {Sprung D W L, Sigetich J D, Wu H, and Martorell J}

503:  {2000}

504:  {Am. J. Phys.}

505:  {68} {715}

506: 

507:  \bibitem{Griffiths} \paperref

508:  {Griffiths D J and Steinke C A}

509:  {2001}

510:  {Am. J. Phys.}

511:  {69} {137}

512: 

513:  \bibitem{Kalotas} \paperref

514:  {Kalotas T M and Lee A R}

515:  {1991}

516:  {Eur. J. Phys.}

517:  {12} {275}

518: 

519:  \bibitem{Mandelbrot} \bookref

520:  {Mandelbrot B B}

521:  {1982}

522:  {The Fractal Geometry of Nature}

523:  {Freeman, San Francisco}

524: 

525:  \bibitem{Ficker} \paperref

526:  {Ficker T and Benesovsky P}

527:  {2002}

528:  {Eur. J. Phys.}

529:  {23} {403}

530: 

531:  \bibitem{Berry} \paperref

532:  {Berry M and Klein S}

533:  {1996}

534:  {J. Mod. Opt.}

535:  {43} {2139}

536: 

537:  \bibitem{Karman} \paperref

538:  {Karman G P, McDonald G S, New G H C, and Woederman J P}

539:  {1999}

540:  {Nature}

541:  {402} {138}

542: 

543:  \bibitem{Saavedra} \paperref

544:  {Saavedra G, Furlan W D, and Monsoriu J A}

545:  {2003}

546:  {Opt. Lett.}

547:  {28}  {971}

548: 

549:  \bibitem{Monsoriu} \paperref

550:  {Monsoriu J A, Furlan W D, and Saavedra G}

551:  {2004}

552:  {Opt. Express}

553:  {12}  {4227}

554: 

555:  \bibitem{Wojcik} \paperref

556:  {Wojcik D, Bialynicki-Birula I, and Zyczkowski K}

557:  {2000}

558:  {Phys. Rev. Lett.}

559:  {85}  {5022}

560: 

561:  \bibitem{Albeverio} \paperref

562:  {Albeverio S and Koshmanenko V}

563:  {2000}

564:  {Rep. Math. Phys.}

565:  {45} {307}

566: 

567:  \bibitem{Makarov} \paperref

568:  {Makarov K A}

569:  {1994}

570:  {J. Math. Phys.}

571:  {35} {1522}

572: 

573:  \bibitem{Chuprikov1} \paperref

574:  {Chuprikov N L}

575:  {2000}

576:  {J. Phys. A: Math. Gen.}

577:  {33} {4293}

578: 

579:  \bibitem{Chuprikov2} \paperref

580:  {Chuprikov N L and Zhabin D N}

581:  {2000}

582:  {J. Phys. A: Math. Gen.}

583:  {33} {4309}

584: 

585:  \bibitem{Pedro} \bookref

586:  {Fern\'andez de C\'ordoba P and Urchuegu{\'\i}a J}

587:  {2004}

588:  {Fundamentos de F{\'\i}sica Cu\'antica para Ingenier{\'\i}a}

589:  {Textbook, Servicio de Publicaciones, Universidad Polit\'ecnica de Valencia, Spain}

590: 

591:  \bibitem{KronigPenney2} \paperref

592:  {Schulkin B, Sztancsik L, and  Federici J F}

593:  {2004}

594:  {Am. J. Phys.}

595:  {72}

596:  {1051}

597: 

598: \endbib

599: 

600: 

601: \end{document}

602: