1: \documentclass[prb,,twocolumn,
2: %draft,preprint,
3: showpacs,preprintnumbers,amsmath,amssymb,superscriptaddress,affilletter]{revtex4}
4:
5: %\documentclass[twocolumn,showpacs
6: %,preprint,draft,prl, nobibnotes,secnumroman%
7: %,tightenlines%
8: %,amssymb, amsmath, aps, prb]{revtex4}
9:
10: \usepackage{color}
11: \usepackage{graphicx}
12: %\usepackage{srcltx}
13: % Include figure files
14: %\usepackage{dcolumn}% Align table columns on decimal point
15: %\usepackage{bm}% bold math
16: %\nofiles
17: \usepackage[T1]{fontenc} %polish fonts
18: \usepackage[cp1250]{inputenc} %polish fonts
19: \newcommand{\ea}{{\em et al.}}
20: \newcommand{\be}{\begin{equation}}
21: \newcommand{\ee}{\end{equation}}
22: \newcommand{\CdMnTe}{Cd$_{1-x}$Mn$_x$Te}
23: \newcommand{\CdMgTe}{Cd$_{1-y}$Mg$_y$Te}
24:
25:
26: \begin{document}
27: \title{Mapping of quantum well eigenstates with semimagnetic probes}% Force line breaks with \\
28:
29: \author{Ł.~Kłopotowski} \affiliation{Institute of Physics, Polish Academy of Sciences,
30: al. Lotników 32/46 02-668 Warsaw, Poland}
31:
32: \author{A.~Gruszczyńska} \affiliation{Institute of Physics, Polish Academy of Sciences,
33: al. Lotników 32/46 02-668 Warsaw, Poland}
34:
35: \author{E.~Janik} \affiliation{Institute of Physics, Polish Academy of Sciences,
36: al. Lotników 32/46 02-668 Warsaw, Poland}
37:
38: \author{M.~Wiater} \affiliation{Institute of Physics, Polish Academy of Sciences,
39: al. Lotników 32/46 02-668 Warsaw, Poland}
40:
41: \author{P.~Kossacki} \affiliation{Institute of Experimental Physics, Warsaw University,
42: ul. Hoża 69 00-681 Warsaw, Poland}
43:
44: \author{G.~Karczewski} \affiliation{Institute of Physics, Polish Academy of Sciences,
45: al. Lotników 32/46 02-668 Warsaw, Poland}
46:
47: \author{T.~Wojtowicz} \affiliation{Institute of Physics, Polish Academy of Sciences,
48: al. Lotników 32/46 02-668 Warsaw, Poland}
49:
50:
51:
52: \date{\today}% It is always \today, today,
53: % but any date may be explicitly specified
54: \begin{abstract}
55:
56: We present results of transmission measurements on CdTe quantum
57: wells with thin semimagnetic \CdMnTe\ probe layers embedded in
58: various positions along the growth axis. The presence of the
59: probes allow us to map the probability density functions by two
60: independent methods: analyzing the exciton energy position and the
61: exciton Zeeman splitting. We apply both approaches to map the
62: first three quantum well eigenstates and we find that both of them
63: yield equally accurate results.
64:
65: \end{abstract}
66: \pacs{68.65.Fg 78.66.Hf 71.35.Cc 71.70.Gm}%
67:
68: %\keywords{Suggested keywords}%Use showkeys class option if keyword
69: %display desired
70: \maketitle
71:
72:
73: %\textcolor[rgb]{1.00,0.00,0.00}{xxx I guess an Introduction is
74: %missing}
75:
76:
77: \section{Introduction}
78:
79: The information about a quantum state is given by its energy $E$
80: and its wave-function $\psi$. Although it is easy to measure the
81: former, the experimental access to the latter is more difficult.
82: In semiconductors, since the advent of epitaxial growth methods,
83: band-gap engineering and tailoring of the eigenstates have become
84: possible and retrieving the information inscribed in the
85: wave-functions became necessary. The access to $\psi$ is gained
86: through the probability density (PD) functions $|\psi|^2$. One way
87: of probing of the PDs in quantum wells (QWs) is to introduce a
88: highly localized potential perturbation with precisely controlled
89: position along the growth axis. The perturbation shifts the
90: eigenenergies of the system allowing to obtain the PD at the
91: location of the probe in either an optical \cite{mar89} or a
92: transport \cite{sal97} experiment. Another way is to introduce a
93: layer containing magnetic ions, which in an external magnetic
94: field give rise to a Zeeman effect larger than in the case of an
95: unperturbed QW \cite{pre00,yan00,lee99}. This latter method
96: allowed mapping of PDs in single \cite{pre00} or coupled multiple
97: QWs \cite{lee99}. In the first case, only the ground state PD was
98: accessed and in the second, only relative, integrated PD values
99: were obtained. In this work we apply both approaches to extract
100: the PDs: we introduce magnetic probes and measure both the Zeeman
101: effect and the shift of excitonic transitions in interband
102: absorption. In this way, we map PD functions of the ground and the
103: first two excited states of a CdTe QW sandwiched between \CdMgTe\
104: barriers. As a PD probe, we use a layer of \CdMnTe, where a part
105: of the Cd cations are substituted with Mn$^{2+}$ ions.
106:
107: Incorporation of magnetic ions into a semiconductor matrix
108: gives a new class of materials usually referred to as diluted
109: magnetic semiconductors (DMS) \cite{fur88}. Most commonly, the
110: substituting atoms are transition metal ions with
111: partially filled $d$-shells (Mn$^{2+}$ ions have a half filled
112: $d$-shell), which gives rise to a localized magnetic moment.
113: Exchange interaction between localized spins of the $d$-shell
114: electrons and band carriers leads to Zeeman effects enhanced by
115: up to three orders of magnitude. To write the electronic
116: wave-function in a DMS, we assume that the electrons adjust
117: quasi-instantaneously to the arrangement of localized spins. In
118: this adiabatic approximation the electronic wave-function reads
119: \cite{mau85}:
120: \be
121: \Psi(\vec{r};\vec{S_1},\ldots,\vec{S_N})=\Psi(\vec{r};{\vec{S_i}})=\psi(\vec{r};{\vec{S_i}})
122: \Phi(\{\vec{S_i}\})
123: \ee
124: where $\{\vec{S_i}\}$ denotes the set of all quantum numbers
125: describing the system of magnetic ions. The $s,p-d$ exchange
126: interaction is described by the Hamiltonian:
127:
128: \be
129: H_{sp-d} = \sum_{\vec{R_i}} J^{sp-d} (\vec{r}-\vec{R_i}) \vec{S_i}
130: \vec{\sigma}
131: \ee
132: where $\vec{r}$ and $\vec{R_i}$ are the spatial and $\vec{\sigma}$
133: and $ \vec{S_i}$ are the spin coordinates of a band electron and a
134: localized ion, respectively. As a consequence of the localized
135: character of the $d$-shell electrons, the exchange constant is
136: usually approximated by a collision term:
137: $J^{sp-d}(\vec{r}-\vec{R_i}) = J^{sp-d}
138: \delta(\vec{r}-\vec{R_i})$.
139:
140: The $s$-$d$ exchange interaction leads therefore to a conduction
141: band splitting given by:
142: \be
143: \label{delta1}
144: \Delta E_c=\Sigma_i \langle \Phi | S_i | \Phi
145: \rangle N_0 \alpha |\phi_c(X_i, Y_i)|^2 |\varphi_c(Z_i)|^2
146: \ee
147: where $(X_i,Y_i,Z_i)$, are the coordinates of a $i$th Mn ion,
148: $N_0$ is the number of cation sites per unit volume, and
149: $\alpha$ is the $s$-$d$ exchange integral. In the above, we
150: factorized the electron wave-function into components dependent
151: on the in-plane and perpendicular coordinates: $\psi(\vec{r})=
152: \phi(x,y)\varphi(z)$. Such a
153: procedure is not always justified when interband absorption is
154: involved as the electron-hole Coulomb interaction mixes these
155: degrees of freedom. However, we checked that in our case this
156: mixing is negligible and in mapping experiments it leads to
157: errors smaller than those resulting from compositional
158: fluctuations and temperature instability.
159:
160: If the function $\varphi(z)$ does not change substantially
161: along the thickness of the probe layer, we can substitute the
162: summation over $Z_i$ with a value of $\varphi(z)$ at the Mn
163: layer location $Z_{Mn}$. Moreover, assuming uniform
164: distribution of Mn ions in the QW plane we can average
165: $\phi_c(x,y)$ over in-plane ion coordinates and obtain the
166: electron Zeeman splitting proportional to the layer
167: magnetization $M_L$ \cite{pre03}:
168:
169: \be
170: \label{delta2}
171: \begin{array}{ll}
172: \Delta E_c & = N_0 \alpha |\varphi_c(Z_{Mn})|^2
173: \overline{|\phi_c(X_i, Y_i)|^2} \Sigma_i \langle \Phi | S_i | \Phi
174: \rangle = \\ & = N_0 \alpha |\varphi_c(Z_{Mn})|^2 \cdot M_L
175: \end{array}
176: \ee
177: where for \CdMnTe\ the conduction band exchange constant
178: is\cite{gaj79}: $N_0 \alpha=0.22$ eV.
179:
180: It is thus seen from Eq. \ref{delta2} that electron Zeeman
181: splitting is proportional to the PD of finding an electron in the
182: Mn layer. However, in an interband absorption experiment we
183: measure the excitonic Zeeman splitting, which is a sum of electron
184: and hole splittings:
185:
186: \be
187: \label{mapZeeman}
188: \Delta E_Z = \left( N_0 \alpha \cdot |\varphi_c(Z_{Mn})|^2 - N_0
189: \beta \cdot |\varphi_v(Z_{Mn})|^2 \right) \cdot M_L
190: \ee
191: where $N_0 \beta = -0.88$ eV is the valence band exchange constant
192: \cite{gaj79}.
193:
194: It can be seen from the above that measuring excitonic Zeeman
195: splitting for a series of samples, where the Mn ions are located
196: at various positions $Z_{Mn}$, allows to map a PD function {\em
197: weighted} with $sp$-$d$ exchange integrals $N_0 \alpha$ and $N_0
198: \beta$ contrary to the usual assumption \cite{lee99,pre00} that
199: the heavy hole PD is mapped.
200:
201: Magnetic dopants not only give rise to magnetooptical effects, but
202: also introduce a local potential. In the first order perturbation
203: theory, a potential of the form: $V(\delta(z-Z_{Mn}))$ located at
204: the position $Z_{Mn}$ shifts the electron energy $E^{0}$ of the
205: eigenstate $\varphi_c(z)$ by:
206:
207: \be
208: E^{\prime}_c-E^{0}_c = V |\varphi_c(Z_{Mn})|^2
209: \ee
210: where $V$ is the perturbing potential, given by the chemical shift
211: of the respective bands. Therefore, for electrons and holes the
212: shift is proportional to conduction and valence band offset,
213: respectively. As a result, the shift of the excitonic transition
214: reads:
215:
216: \be
217: \label{mapShift}
218: \Delta E_S = E^{\prime}_X-E^{0}_X = \varrho_c V
219: |\varphi_c(Z_{Mn})|^2 + \varrho_v V |\varphi_v(Z_{Mn})|^2
220: \ee
221: where $\varrho_c$ and $\varrho_v$ are conduction and valence band
222: offsets, respectively and $E^{0}_X$ is the energy of the
223: unperturbed state. Therefore, measuring the shift of the exciton
224: energy, we can map a PD function weighted with band offsets. In
225: the following, we took a valence band offset $\varrho_v=0.4$
226: \cite{kut97,siv99} and assumed a linear dependence of the chemical
227: shift on the Mn composition\cite{fur88}: $V=1592 {\mbox { meV}}
228: \cdot x_{Mn}$.
229:
230: \section{Samples and Experiment}
231:
232: Designing samples for mapping experiments using semimagnetic
233: probes one has to bear in mind that the profiles of CdTe/\CdMnTe\
234: interfaces are broadened along the growth axis due to complete
235: exchange of Cd and Mn ions during growth and thus absence of
236: segregation processes \cite{gri96}. Consequently, although we aim
237: to obtain thin probe layers, the probe ions are always distributed
238: among a couple of adjacent monolayers with the composition profile
239: peaked at $Z_{Mn}$. Moreover, we have to take into account the
240: antiferromagnetic coupling between Mn ions, which decreases
241: substantially the magnetization and as a result also the
242: splitting, as seen in Eq. \ref{mapZeeman}. Therefore, the
243: composition of the probe layers has to be low enough to assure a
244: small number of nearest neighbor Mn pairs to avoid the
245: antiferromagnetic coupling.
246:
247: The samples were grown on (001) oriented GaAs substrates by
248: molecular beam epitaxy. Substrate temperature was
249: 230$^{\circ}$C, which assures a high sample quality and
250: relatively low interface broadening\cite{gri96}. 3.5 $\mu$m
251: \CdMgTe\ buffer was deposited before the growth of the QWs to
252: relax the strain resulting from the lattice mismatch between
253: the QW structure and the substrate. Next, five CdTe QWs, 117
254: \AA , wide were grown, separated by 300 \AA\ \CdMgTe\ barrier
255: layers. Magnesium composition $y$ was chosen as high as 33\% in
256: order to assure that more than one confined state is present in
257: the QW. Growth of each of the QWs was interrupted for the
258: deposition of a single probe consisting of 2 monolayers of
259: \CdMnTe\ with intentional Mn molar fraction of 12\%.
260: For schematics of the probing
261: heterostructure, see Fig. \ref{spectra}b. Four samples with
262: different positions of the probe layer along the QW axis were
263: grown. Additionally, a reference sample with no probe was
264: prepared.
265:
266: %\begin{figure}[h]
267: %\begin{center}
268: %\includegraphics[width=.5\textwidth]{sample.eps}
269: %\end{center}
270: %\caption{}\label{calc}
271: %\end{figure}
272:
273: To obtain exciton transition energies and Zeeman splitting we
274: measured transmission as a function of magnetic field. To
275: optically access the QWs, we first had to remove the
276: nontransparent GaAs substrate, which was done by mechanical
277: polishing and wet etching in hydrogen peroxide. The thick
278: transparent buffer produced Fabry-P\'erot oscillations, which
279: obscured the absorption spectrum and thus most of it was removed
280: by chemical etching in a 0.6\% solution of bromine in methanol.
281: The sample was immersed in superfluid liquid helium at a
282: temperature of 1.8 K. Magnetic field up to 7 T was applied in
283: Faraday configuration. A halogen lamp was used as a white-light
284: source and the transmission signal was detected by a liquid
285: nitrogen cooled CCD camera and a monochromator. In order to
286: analyze transitions in two circular polarizations, a quarter
287: wave-plate and a linear polarizer were placed in the way of the
288: transmitted beam.
289:
290:
291: \section{Results and Discussion}
292:
293: Optical density spectra were evaluated according to Beer-Lambert
294: law as $A=-\log (I/I_0)$, where $I$ and $I_0$ are transmitted and
295: incident beam intensities, respectively. In Fig. \ref{spectra}, we
296: show the spectra obtained from a sample with the probe layer
297: located at the center of the QW. Transitions corresponding to
298: three heavy hole excitons, labelled ($n_{el},n_{hh}$) with
299: $n=1,2,3$ numbering electron and heavy hole states, can be
300: resolved. Only diagonal transitions, i.e. those satisfying
301: $n_{el}=n_{hh}$, are observed. The oscillator strength of
302: parity-allowed nondiagonal transitions is very low, since in a
303: deep rectangular QW the eigenstates are nearly orthogonal. The
304: feature below the (2,2) transition is related to the light hole
305: (1,1) exciton. The increase of the absorption above 2.1 eV is due
306: to \CdMgTe\ barrier excitons. Immediately from Fig. 1 the effect
307: of the Mn probe layer on the exciton Zeeman splitting can be
308: deduced: the odd-number excitons exhibit a giant Zeeman effect
309: since there is a non-vanishing PD of finding carriers in the
310: center of the QW (see Fig. 1). On the other hand, the Zeeman
311: splitting of the (2,2) state is limited to the direct interaction
312: between carriers and the magnetic field and so the splitting is
313: smaller then the transition line-width.
314:
315: \begin{figure}[h]
316: \begin{center}
317: \includegraphics[angle=0,width=.45\textwidth]{fig1.1.eps}
318: \end{center}
319: \caption{Left: Optical density spectra for various magnetic fields
320: obtained for a sample with the probe layer at the center of the
321: quantum well. Right: Schematic of this sample shown together with
322: the first three electron PD functions. }\label{spectra}
323: \end{figure}
324:
325: The identification of the excitonic transitions in Fig. 1a is
326: based on results of effective mass approximation calculations
327: of electron and hole energy levels in the QW as a function of
328: the magnetic field. In the calculations, we took into account the diffusion of the probe interfaces during growth. The broadened probe shape was modeled
329: by a gaussian function \cite{pre00}. The band-edge Zeeman
330: splitting of the probe was described by a modified Brillouin
331: function \cite{fur88} with effective parameters $S_0$ and $T_0$
332: reflecting the antiferromagnetic coupling between the Mn ions
333: adjusted to take into account the nonuniform number of nearest
334: neighbors. Valence band states were calculated using a 4x4 Luttinger
335: Hamiltonian \cite{lut56}. The lattice mismatch between the barrier
336: and the QW layer introduced a strain, which was taken into account
337: in the framework of the Bir-Pikus theory by adding a complete deformation potential Hamiltonian \cite{bir74}. We neglected all the
338: excitonic effects and used the exciton binding energy as a free
339: parameter. In Fig. \ref{calc}, we present the experimental and
340: calculated transition energies for the same sample as in Fig.
341: 1, i.e. with the Mn probe layer in the center of the QW. Heavy
342: hole exciton binding energies resulting from the presented fits
343: were found to be between 14 and 19 meV remaining in good
344: agreement with calculations in framework of an analytical model
345: by Mathieu \ea \cite{mat92}, which yields for (1,1) exciton a
346: binding energy of 14 meV. A very good agreement between the
347: measured and calculated transition energies points out that the
348: model includes the most important features of the system.
349:
350: \begin{figure}[h]
351: \begin{center}
352: \includegraphics[angle=0,width=.50\textwidth]{calc.eps}
353: \end{center}
354: \caption{Points: Exciton transitions measured for a sample with
355: the probe layer at the center of the quantum well. Lines: The
356: results of effective mass calculations allowing the identification
357: of the transitions. Full (empty) points correspond to transitions
358: seen in $\sigma^+$ ($\sigma^-$) polarization.}\label{calc}
359: \end{figure}
360:
361: Using the above procedure for the reference sample with a flat
362: (i.e. without the probe layer) QW, we fitted the excitonic
363: transitions and calculated electron and heavy hole PD functions
364: $|\varphi_{c}(z)|^2$ and $|\varphi_{v}(z)|^2$. In Fig.
365: \ref{mapping}a, we plot these functions weighted with exchange
366: constants as derived in Eq. \ref{mapZeeman} for the first three QW
367: eigenstates. On the same graph, we present the the exciton Zeeman
368: splitting $\Delta E_Z$ measured as a function of the position of
369: the center of probe layer $Z_{Mn}$. A very good correlation
370: between the Zeeman splitting and the PD value at $Z_{Mn}$ is
371: obtained confirming that the Zeeman splitting provides a good
372: estimation of the PDs. In Fig. \ref{mapping}b, we plot the flat-QW
373: PDs weighted with band offsets as derived in Eq. \ref{mapShift}.
374: On the same graph, we plot the exciton energy shifts $\Delta E_S$
375: given by the difference between transition energies for the
376: samples with probe layers and the reference sample with a flat QW.
377: The shifts for the (1,1) and (2,2) excitons are measured and for
378: the remaining (3,3) is calculated since this transition could not
379: be resolved in the reference sample. Again, a good correlation
380: between the measured shift and the PD at the probe location is
381: obtained.
382:
383:
384: \begin{figure*}[bt!]
385: \begin{center}
386: \includegraphics[angle=0,width=.95\textwidth]{fig3.eps}
387: \end{center}
388: \caption{a) The PD functions of three lowest quantum well
389: eigenstates weighted with exchange constants (lines) and
390: corresponding Zeeman splittings (points, right scale) plotted as a
391: function of the probe layer location. b) The PD functions of three
392: lowest quantum well eigenstates weighted with conduction and
393: valence band offsets (lines) and corresponding exciton shifts
394: (points, right scale) plotted as a function of the probe layer
395: location. c) and d) Correlations between weighted PD functions and
396: Zeeman splitting and exciton shifts, respectively.}
397: \label{mapping}
398: \end{figure*}
399:
400: The accuracies of our mapping procedures are summarized in
401: Figs. \ref{mapping}c and d, which show how the measured Zeeman
402: splittings and excitonic shifts are correlated with values of
403: PD functions weighted with exchange integrals and band offsets,
404: respectively. In both cases, we obtain correlation coefficient
405: values $R\approx 0.9$, proving a high accuracy of the approach.
406: Inaccuracies are caused mainly by the fact that the probe layer
407: has a nonzero thickness, contrary to the assumption taken to
408: derive Eqs. \ref{mapZeeman} and \ref{mapShift}. Indeed, two
409: monolayers deposited during growth are further broadened by the
410: intermixing of the interface profile. To fit the excitonic
411: transition dependencies on the magnetic field (see Fig.
412: \ref{calc}), we had to introduce a gaussian broadened profiles
413: with halfwidths between 2.5 and 4 monolayers depending on the
414: sample. Our effective mass approximation calculations show that
415: this nonzero thickness of the probe layers leads to a
416: noticeable modification of the PD functions. In the PD
417: functions compared with measured quantities in Figs.
418: \ref{mapping} also not included are the excitonic effects. The
419: electron-hole Coulomb interaction modifies importantly the
420: shape of PD functions with respect to a noninteracting case. A
421: method to obtain Coulomb-correlated $\varphi_c$ and $\varphi_v$
422: is based on a Hartree approach. In this calculation the
423: electron wave function is self-consistently calculated by
424: solving a one dimensional Schr\"{o}dinger equation with
425: an effective potential resulting from the hole wave function
426: and vice-versa \cite{kyr00}. Since in CdTe the hole is about 5
427: times heavier than the electron, the Coulomb-correlated
428: $\varphi_c$ tends to unperturbed $\varphi_v$. For this reason,
429: previous works on Zeeman mapping successfully compared the
430: results to only heavy hole PD functions \cite{lee99,pre00}.
431:
432: Another issue not included in the above considerations is the
433: variation of the exchange constants with confinement. As
434: pointed out by Mackh \ea \cite{mac96} and Merkulov \ea
435: \cite{mer99}, an admixture of higher
436: \boldmath$k$\unboldmath-vector states leads to a significant
437: decrease of the exchange parameters. Possible reasons include
438: turning on the kinetic exchange between the conduction
439: electrons and localized Mn ions\cite{mer99} and a hopping
440: interference in the valence band\cite{mac96}. In our case, both
441: effects would have highest impact on the (3,3) state -- the one
442: with highest admixture of the nonzero-\boldmath$k$\unboldmath\
443: states\cite{mer99}. We compared the PDs weighted with exchange
444: integrals that were decreased according to the
445: higher-\boldmath$k$\unboldmath\ states admixture and found a
446: worse agreement with measured Zeeman splittings. We therefore
447: conclude that the reduction of the exchange integrals is less
448: important than the effects related to a nonzero thickness of
449: the probe layers.
450:
451: In order to obtain a higher mapping accuracy one should design
452: thinner (e.g. submonolayer) mapping probes that introduce a
453: smaller perturbation potential. In that case, the observed Zeeman
454: splitting and the shift of the excitonic transition will be
455: smaller, but one will gain a weaker modification of the mapped PD
456: functions. Another approach is to completely bury the semimagnetic
457: probes in a CdMgTe QW exploiting the same band offsets for MgTe
458: and MnTe with respect to CdTe \cite{kut97,kuh93}. In this method,
459: the Mn composition of the probes and the Mg composition of the QW
460: bottom are chosen to assure a flat QW potential so at zero
461: magnetic field no perturbation is introduced. Applying a small
462: magnetic field should make mapping feasible with only a minor
463: modification of the PD functions.
464:
465: In summary, we have compared two independent methods for mapping
466: of the quantum well eigenstates with semimagnetic probes. One is
467: based on the analysis of the position of exciton transition which
468: in the first order of perturbation theory is proportional to a
469: value of a wave function weighted with band offsets. The second
470: approach exploits the fact that for samples with thin semimagnetic
471: probes the exciton Zeeman splitting is proportional to a value of
472: a wave function weighted with exchange constants. We find a good
473: agreement between the calculated wave functions and measured
474: excitonic positions and Zeeman splittings and we conclude that
475: both methods are equally well suited for mapping purposes.
476:
477:
478:
479: \begin{thebibliography}{19}
480: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
481: \expandafter\ifx\csname bibnamefont\endcsname\relax
482: \def\bibnamefont#1{#1}\fi
483: \expandafter\ifx\csname bibfnamefont\endcsname\relax
484: \def\bibfnamefont#1{#1}\fi
485: \expandafter\ifx\csname citenamefont\endcsname\relax
486: \def\citenamefont#1{#1}\fi
487: \expandafter\ifx\csname url\endcsname\relax
488: \def\url#1{\texttt{#1}}\fi
489: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
490: \providecommand{\bibinfo}[2]{#2}
491: \providecommand{\eprint}[2][]{\url{#2}}
492:
493: \bibitem[{\citenamefont{Marzin and G\'erard}(1989)}]{mar89}
494: \bibinfo{author}{\bibfnamefont{J.-Y.} \bibnamefont{Marzin}} \bibnamefont{and}
495: \bibinfo{author}{\bibfnamefont{J.-M.} \bibnamefont{G\'erard}},
496: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{62}},
497: \bibinfo{pages}{2172} (\bibinfo{year}{1989}).
498:
499: \bibitem[{\citenamefont{Salis et~al.}(1997)\citenamefont{Salis, Graf, Ensslin,
500: Campman, Maranowski, and Gossard}}]{sal97}
501: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Salis}},
502: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Graf}},
503: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Ensslin}},
504: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Campman}},
505: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Maranowski}},
506: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{A.~C.}
507: \bibnamefont{Gossard}}, \bibinfo{journal}{Phys. Rev. Lett.}
508: \textbf{\bibinfo{volume}{79}}, \bibinfo{pages}{5106} (\bibinfo{year}{1997}).
509:
510: \bibitem[{\citenamefont{Prechtl et~al.}(2000)\citenamefont{Prechtl, Heiss,
511: Bonnani, Jantsch, Maćkowski, Janik, and Karczewski}}]{pre00}
512: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Prechtl}},
513: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Heiss}},
514: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Bonnani}},
515: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Jantsch}},
516: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Maćkowski}},
517: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Janik}}, \bibnamefont{and}
518: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Karczewski}},
519: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{61}},
520: \bibinfo{pages}{15617} (\bibinfo{year}{2000}).
521:
522: \bibitem[{\citenamefont{Yang et~al.}(2000)\citenamefont{Yang, Furdyna, and
523: Luo}}]{yan00}
524: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Yang}},
525: \bibinfo{author}{\bibfnamefont{J.~K.} \bibnamefont{Furdyna}},
526: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Luo}},
527: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{62}},
528: \bibinfo{pages}{4226} (\bibinfo{year}{2000}).
529:
530: \bibitem[{\citenamefont{Lee et~al.}(1999)\citenamefont{Lee, Dobrowolska,
531: Furdyna, and Ram-Mohan}}]{lee99}
532: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Lee}},
533: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Dobrowolska}},
534: \bibinfo{author}{\bibfnamefont{J.~K.} \bibnamefont{Furdyna}},
535: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{L.~R.}
536: \bibnamefont{Ram-Mohan}}, \bibinfo{journal}{Phys. Rev. {\bf B}}
537: \textbf{\bibinfo{volume}{59}}, \bibinfo{pages}{10302} (\bibinfo{year}{1999}).
538:
539: \bibitem[{\citenamefont{Furdyna}(1988)}]{fur88}
540: \bibinfo{author}{\bibfnamefont{J.~K.} \bibnamefont{Furdyna}},
541: \bibinfo{journal}{J. Appl. Phys.} \textbf{\bibinfo{volume}{64}},
542: \bibinfo{pages}{R29} (\bibinfo{year}{1988}).
543:
544: \bibitem[{\citenamefont{Mauger and Mills}(1985)}]{mau85}
545: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Mauger}} \bibnamefont{and}
546: \bibinfo{author}{\bibfnamefont{D.~L.} \bibnamefont{Mills}},
547: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{31}},
548: \bibinfo{pages}{8024} (\bibinfo{year}{1985}).
549:
550: \bibitem[{\citenamefont{Prechtl et~al.}(2003)\citenamefont{Prechtl, Heiss,
551: Bonnani, Jantsch, Maćkowski, and Janik}}]{pre03}
552: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Prechtl}},
553: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Heiss}},
554: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Bonnani}},
555: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Jantsch}},
556: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Maćkowski}},
557: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Janik}},
558: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{68}},
559: \bibinfo{pages}{165313} (\bibinfo{year}{2003}).
560:
561: \bibitem[{\citenamefont{Gaj et~al.}(1979)\citenamefont{Gaj, Planel, and
562: Fishman}}]{gaj79}
563: \bibinfo{author}{\bibfnamefont{J.~A.} \bibnamefont{Gaj}},
564: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Planel}}, \bibnamefont{and}
565: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Fishman}},
566: \bibinfo{journal}{Solid State Comm.} \textbf{\bibinfo{volume}{29}},
567: \bibinfo{pages}{435} (\bibinfo{year}{1979}).
568:
569: \bibitem[{\citenamefont{Kutrowski et~al.}(1997)\citenamefont{Kutrowski,
570: Wojtowicz, Cywinski, Karczewski, Janik, Dynowska, Kossut, Kossacki,
571: Fiederling, Pfeuffer-Jeschke et~al.}}]{kut97}
572: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Kutrowski}},
573: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Wojtowicz}},
574: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Cywinski}},
575: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Karczewski}},
576: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Janik}},
577: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Dynowska}},
578: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Kossut}},
579: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Kossacki}},
580: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Fiederling}},
581: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Pfeuffer-Jeschke}},
582: \bibnamefont{et~al.}, \bibinfo{journal}{Acta Phys. Pol. {\bf A}}
583: \textbf{\bibinfo{volume}{92}}, \bibinfo{pages}{887} (\bibinfo{year}{1997}).
584:
585: \bibitem[{\citenamefont{Siviniant et~al.}(1999)\citenamefont{Siviniant,
586: Kyrychenko, Semenov, Coquillat, Scalbert, and Lascaray}}]{siv99}
587: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Siviniant}},
588: \bibinfo{author}{\bibfnamefont{F.~V.} \bibnamefont{Kyrychenko}},
589: \bibinfo{author}{\bibfnamefont{Y.~G.} \bibnamefont{Semenov}},
590: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Coquillat}},
591: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Scalbert}}, \bibnamefont{and}
592: \bibinfo{author}{\bibfnamefont{J.~P.} \bibnamefont{Lascaray}},
593: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{59}},
594: \bibinfo{pages}{10276} (\bibinfo{year}{1999}).
595:
596: \bibitem[{\citenamefont{Grieshaber et~al.}(1996)\citenamefont{Grieshaber,
597: Haury, Cibert, d'Aubign\'e, Wasiela, and Gaj}}]{gri96}
598: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Grieshaber}},
599: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Haury}},
600: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Cibert}},
601: \bibinfo{author}{\bibfnamefont{Y.~M.} \bibnamefont{d'Aubign\'e}},
602: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Wasiela}}, \bibnamefont{and}
603: \bibinfo{author}{\bibfnamefont{J.~A.} \bibnamefont{Gaj}},
604: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{53}},
605: \bibinfo{pages}{4891} (\bibinfo{year}{1996}).
606:
607: \bibitem[{\citenamefont{Luttinger}(1956)}]{lut56}
608: \bibinfo{author}{\bibfnamefont{J.~M.} \bibnamefont{Luttinger}},
609: \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{102}},
610: \bibinfo{pages}{1030} (\bibinfo{year}{1956}).
611:
612: \bibitem[{\citenamefont{Bir and Pikus}(1974)}]{bir74}
613: \bibinfo{author}{\bibfnamefont{G.~L.} \bibnamefont{Bir}} \bibnamefont{and}
614: \bibinfo{author}{\bibfnamefont{G.~E.} \bibnamefont{Pikus}},
615: \emph{\bibinfo{title}{Symmetry and Strain-induced Effects in Semiconductors}}
616: (\bibinfo{publisher}{Wiley, New York}, \bibinfo{year}{1974}).
617:
618: \bibitem[{\citenamefont{Mathieu et~al.}(1992)\citenamefont{Mathieu, Lefebvre,
619: and Christol}}]{mat92}
620: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Mathieu}},
621: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Lefebvre}}, \bibnamefont{and}
622: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Christol}},
623: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{46}},
624: \bibinfo{pages}{4092} (\bibinfo{year}{1992}).
625:
626: \bibitem[{\citenamefont{Kyrychenko et~al.}(2000)\citenamefont{Kyrychenko,
627: Ryabchenko, and Semenov}}]{kyr00}
628: \bibinfo{author}{\bibfnamefont{F.~V.} \bibnamefont{Kyrychenko}},
629: \bibinfo{author}{\bibfnamefont{S.~M.} \bibnamefont{Ryabchenko}},
630: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{Y.~G.}
631: \bibnamefont{Semenov}}, \bibinfo{journal}{Physica E}
632: \textbf{\bibinfo{volume}{8}}, \bibinfo{pages}{275} (\bibinfo{year}{2000}).
633:
634: \bibitem[{\citenamefont{Mackh et~al.}(1996)\citenamefont{Mackh, Ossau, Waag,
635: and Landwehr}}]{mac96}
636: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Mackh}},
637: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Ossau}},
638: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Waag}}, \bibnamefont{and}
639: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Landwehr}},
640: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{54}},
641: \bibinfo{pages}{5227} (\bibinfo{year}{1996}).
642:
643: \bibitem[{\citenamefont{Merkulov et~al.}(1999)\citenamefont{Merkulov, Yakovlev,
644: Keller, Ossau, Geurts, Waag, Landwehr, Karczewski, Wojtowicz, and
645: Kossut}}]{mer99}
646: \bibinfo{author}{\bibfnamefont{I.~A.} \bibnamefont{Merkulov}},
647: \bibinfo{author}{\bibfnamefont{D.~R.} \bibnamefont{Yakovlev}},
648: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Keller}},
649: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Ossau}},
650: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Geurts}},
651: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Waag}},
652: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Landwehr}},
653: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Karczewski}},
654: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Wojtowicz}},
655: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Kossut}},
656: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{83}},
657: \bibinfo{pages}{1431} (\bibinfo{year}{1999}).
658:
659: \bibitem[{\citenamefont{Kuhn-Heinrich et~al.}(1993)\citenamefont{Kuhn-Heinrich,
660: Ossau, Heinke, Fischer, Litz, Waag, and Landwehr}}]{kuh93}
661: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Kuhn-Heinrich}},
662: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Ossau}},
663: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Heinke}},
664: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Fischer}},
665: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Litz}},
666: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Waag}}, \bibnamefont{and}
667: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Landwehr}},
668: \bibinfo{journal}{App. Phys. Lett.} \textbf{\bibinfo{volume}{63}},
669: \bibinfo{pages}{2932} (\bibinfo{year}{1993}).
670:
671: \end{thebibliography}
672:
673:
674:
675: \end{document}