Novel nanostructure architectures

ARTICLE IN PRESS

Physica E 25 (2004) 280–287 www.elsevier.com/locate/physe

Novel nanostructure architectures O.G. Schmidta,*, A. Rastellia, G.S. Kara, R. Songmuanga, S. Kiravittayaa, c M. Stoffela, U. Denkera, S. Stuflerb, A. Zrennerb, D. Grutzmacher . , B.-Y. Nguyend, P. Wennekerse Max-Planck-Institut fur Heisenbergstrasse 1, D-70569 Stuttgart, Germany . Festkorperforschung, . Universitat . Paderborn, Experimentalphysik, Warburgerstr. 100, D-33098 Paderborn, Germany c Paul-Scherrer-Institut, CH-5232, Switzerland d Advanced Products R&D Laboratory, Motorola incorporated, 3501 Ed Bluestein Boulevard, Austin, TX 78721, USA e Digital DNA Laboratories—EMEA, Motorola incorporated, Am Borsigturm 130, D-13507, Berlin, Germany a

b

Available online 28 July 2004

Abstract We combine self-assembled island growth with in situ and ex situ selective etching techniques to create advanced semiconductor nanostructure architectures. Such architectures include lateral InAs/GaAs quantum dot molecules and unstrained inverted GaAs/AlGaAs quantum dots (QDs) with photoluminescence (PL) peak linewidths as narrow as 8.9 meV. Micro-PL of single GaAs/AlGaAs QDs reveals well-behaved and sharp exciton emission lines in the red spectral range, which renders these novel nanostructures interesting candidates for future coherent optical investigations. Furthermore, we overgrow self-assembled Ge islands with a thin Si cap layer and use standard ebeam lithography and reactive ion etching to define mesa structures on the surface. The SiGe island core is then selectively etched away, which leaves ultra-thin free-standing Si bridges as remainders on the surface. Such thin Si bridges present an alternative route to produce SiGe-free Si-on-nothing (SON) for advanced complementary metal oxide semiconductor (CMOS) technology. r 2004 Elsevier B.V. All rights reserved. PACS: 78.67.Hc; 81.07.Ta; 81.65.Cf; 81.16.Dn; 81.15.Hi; 73.21.La; 68.65.Hb Keywords: Hierarchical self-assembly; Molecular beam epitaxy; In situ etching; MOSFET

1. Introduction The long-lasting miniaturisation of functional units in semiconductor technology requires both fundamentally new approaches in semiconductor physics as well as rigorous optimisation concepts *Corresponding author. E-mail address: [email protected] (O.G. Schmidt).

in existing complementary metal oxide semiconductor (CMOS) technology. Fundamentally new concepts are currently explored in the rapidly evolving field of quantum information technology [1], while the more traditional CMOS technology moves towards highly optimised field effect transistors that rely on concepts such as strained Si [2–4], Si-on-insulator (SOI) [5], or Si-on-nothing (SON) [6,7]. In this work, we show that a

1386-9477/$ - see front matter r 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.physe.2004.06.027

ARTICLE IN PRESS O.G. Schmidt et al. / Physica E 25 (2004) 280–287

281

combination of self-assembled quantum dot (QD) growth with selective etching techniques creates interesting new semiconductor architectures that are worth considering for quantum coherent experiments as well as for advanced CMOS technology.

2. Quantum dot growth and in situ etching The great success of molecular beam epitaxy (MBE) relies on the ability to create thin pseudomorphic layers and heterostructure interfaces with monolayer precision and close to perfect structural integrity. The reverse process, i.e. the removal of single monolayers under ultra-high vacuum conditions, has also been reported applying atomic-layer precise in situ etching (ALPISE) to III–V surfaces [8,9]. The combination of MBE and ALPISE with AsBr3 has led to an additional degree of design freedom to modify semiconductor nanostructures on the atomic scale [10,11]. In particular, novel self-assembled QD configurations such as lateral QD molecules [12–14] and unstrained red-light emitting GaAs/AlGaAs QDs [15] were realised, which are interesting quantum structures in the modern field of quantum information technology. Fig. 1 schematically illustrates the multi-step (hierarchical) self-assembly processes that can be obtained by combining MBE and ALPISE. The process starts with an initial layer of selfassembled QDs (Step 1) covered by a 5 to 20 nm thick GaAs cap layer (Step 2). The GaAs surface is then exposed to the AsBr3 etching gas (Step 3), which in the beginning preferably etches away those areas of the GaAs, which are disturbed by the strain fields of the buried islands. This etching step leads to the occurrence of nanometer sized holes on the GaAs surface above the buried QDs [11–16]. For deeper etching the In(Ga)As QDs are completely removed, since InGaAs experiences a higher etching rate than GaAs [17]. The nanoholes have depths up to 10 nm and widths of 50–80 nm. This surface decorated with nanoholes serves as a new template for further self-assembly processes. On the left-hand side of Fig. 1, the GaAs nanoholes are overgrown with InAs, which forms

Fig. 1. Different paths to novel quantum dot configurations combining MBE and atomic-layer precise in situ etching (ALPISE). Steps 1–3: GaAs capped InAs QDs, etched in situ with AsBr3. The result are nanometer sized holes on the surface. Step A4: InAs QD molecule formation around the nanoholes. A5: Overgrowth and in situ etching yielding two closely spaced nanoholes. Step B4: AlGaAs overgrowth, converting the GaAs into AlGaAs nanoholes. Step B5: Fill-up of the AlGaAs nanoholes with GaAs QDs.

laterally closely spaced QDs at the edges of the nanoholes (step A4). These structures are called lateral quantum dot molecules (QDM) [12–14], although electronic coupling of charge carriers still needs to be demonstrated. By carefully tuning the growth conditions of step A4, it is possible to create three, four, five, and six closely spaced QDs around one nanohole [12,13]. It is obvious that steps 2 and 3 can be applied again to the lateral QDMs, thus creating laterally closely spaced nanoholes (step A5), which then might serve as a template to fabricate for example two closely spaced QDMs, i.e. a chain of four quantum dots. We have demonstrated that the optical quality of ensembles of lateral QDMs overgrown with GaAs do not suffer from the etching process [12–14], thus proving that ALPISE is an MBE-compatible process. An alternative assembly path is illustrated on the right-hand side of Fig. 1. In this case the

ARTICLE IN PRESS O.G. Schmidt et al. / Physica E 25 (2004) 280–287

(a)

(b)

(c)

(d) 100 nm

[110] 6

80

(e)

Hole width (nm)

nanoholes are overgrown with AlGaAs (step B4). For low growth temperatures the Al atom surface diffusion length is sufficiently small to preserve the nanohole structures during growth. The AlGaAs nanoholes are then overgrown with a few nanometers GaAs. The sample is annealed for a short period at higher temperatures, so that the nanoholes become completely filled up with GaAs (Step B5). After overgrowth with AlGaAs, this multistep (hierarchical) self-assembly yields inverted unstrained GaAs/AlGaAs QDs [15]. In contrast to random GaAs/AlGaAs quantum well thickness fluctuations [18–21]—which have also been exploited as QDs to demonstrate coherent optical manipulation [19–21]—the GaAs QDs depicted in Fig. 1 possess a well-known position and geometry, because the position and geometry of the nanoholes are well-defined by the underlying InAs QD layer. In addition, due to the absence of strain we can avoid strain-driven material intermixing, which people commonly observe during overgrowth of, for example, InAs/GaAs QDs [22,23] or Ge/Si islands [24–27]. It is therefore important to appreciate that the electronic spectrum of these novel GaAs QDs can be accurately predicted by k:p theory [15]. Fig. 2(a)–(d) shows atomic force microscopy (AFM) images at different stages of the hierarchical self-assembly processes described in Fig. 1. In Fig. 2(a) the morphology of a single layer of 1.8 ML InAs QDs grown at 0.01 ML/s at 500 C is shown. The QDs have a density of 3  109 cm 2, an average height of 12 nm and lateral size of 35 nm. After overgrowth with 10 nm GaAs, 5 nm deep etching and 2.5 ML InAs regrowth (steps 2 to A4 in Fig. 1), an array of lateral quantum dot bimolecules has formed on the surface as shown in Fig. 2(b). The AlGaAs nanohole arrays in Fig. 2(c) and (d) are obtained by applying GaAs growth/ ALPISE/AlGaAs (steps 2 to B4) to the single QDs in Fig. 2(a) and the lateral QDMs in Fig. 2(b), respectively. The geometry of the AlGaAs nanoholes precisely defines the size and shape and therefore the electronic properties of the GaAs/AlGaAs QDs. A thorough understanding of the hole geometry as a function of growth parameters is therefore desirable. In Fig. 2(e) and (f) we plot the hole depth and

Hole depth (nm)

282

5 4 3 2

70 60

Wx 50

0

5

10

15

20

(f)

Wy 0

5

10

15

20

AlGaAs barrier thickness D (nm)

Fig. 2. AFM scans showing (a) single QD layer, (b) lateral quantum dot bi-molecules, (c) single AlGaAs nanoholes and (d) double AlGaAs nanoholes. (e), (f) Nanohole geometry as a function of AlGaAs thickness D.

width as a function of AlGaAs overgrowth thickness D (step B4 in Fig. 1). With increasing D (0 to 20 nm) the hole depth decreases from 5.3 nm to only 1.5 nm. At the same time the nanoholes slightly elongate into the [1 1 0] direction (68–78 nm), but become strongly constricted (75–48 nm) in the [1–10] direction. The anisotropic change of the nanohole shape is attributed to different surface diffusivities of the Al and Ga atoms [28]. The photoluminescence (PL) spectra of an ensemble of GaAs/AlGaAs QDs with D=7 nm is given in Fig. 3 as a function of excitation

ARTICLE IN PRESS O.G. Schmidt et al. / Physica E 25 (2004) 280–287

E

Peak pos. (eV)

D = 7 nm T=8K φspot ≈ 180 µm 0

E1

1.80 1.75

QW

1.70 1.65

QDs

1.60

Intesity (arb. Units)

E2

5

10

15

20

D (nm)

E3 E4

145 W/cm2 34 W/cm2

13 W/cm2 QDs

1.55

1.60

1.65

Iexc=1.5 W/cm2

1.70 1.75 Energy (eV)

QW

1.80

1.85

Fig. 3. Excitation power dependent photoluminescence spectra of newly developed GaAs/AlGaAs quantum dots. E1 to E4 denote recombination of 1st to 4th excited states of the QDs. Inset shows transition energies of the GaAs QDs and the QW as a function of lower AlGaAs barrier thickness D.

power density (1.5–145 W/cm2). As expected for zero-dimensional quantum structures, the spectra reveal several well-resolved high-energy peaks with increasing excitation power, which are caused by the recombination of excited states in the QDs. For 145 W/cm2 four excited states can be identified and a pronounced signal from the thin GaAs QW (that exists between the QDs (see B5 in Fig. 1)) is visible. At the same time, the ground state shows a small red-shift, which originates from bi- and multi-exciton recombination occurring at high excitation densities. The inset of Fig. 3 shows the range of transition energies that we can achieve by changing the lower AlGaAs barrier thickness D. The black diamonds represent our experimental data, whereas the white diamonds are calculated . values solving the monodimensional Schrodinger equation with the Ben Daniel–Duke Hamiltonian for electrons and heavy holes in the G valley. For D=5 nm the ground state emission occurs at 1.61 eV. For D=20 nm the energy transition

283

blue-shifts by 100 meV to 1.71 eV. Simultaneously, a small red shift of the QW transition is recorded. The blue-shift of the QD transition is easy to understand, if we keep in mind that with increasing D the nanoholes become shallower (see Fig. 2(e)). This experimental trend is well-reproduced by our simple calculations. A technique to increase the nanohole depth— and therefore to increase the localisation potential for charge carriers in GaAs/AlGaAs QDs—is depicted in Fig. 4(a). Multiple layers of InAs/ GaAs QDs are stacked and capped with GaAs. The strain fields of the vertically aligned QDs superimpose [24,25] and cause the formation of deeper nanoholes during the ALPISE process. The corresponding AFM scans of the nanoholes are given in Fig. 4(b) for a single QD layer (left side) and a two-fold stacked QD layer with a spacer layer of 10 nm (right side). The histograms in Fig. 4(c) show that the average hole depth is 5.4 nm in the case of the single QD layer and 9.0 nm for the double QD layer. A collection of PL spectra of GaAs/AlGaAs QD ensembles is given in Fig. 4(d). The spectra are labelled with (D,d), denoting the thickness of the lower AlGaAs barrier and the thickness of the GaAs QD layer, respectively. An S in front of the brackets indicates that the holes were etched into a buried two-fold stack of InAs QDs. The top-most spectrum shows that PL line widths as narrow as 8.9 meV can be obtained, and a summary of the line widths as a function of hole depths is given in the inset of Fig. 4(d). The line widths narrow with increasing hole depths, although the relative hole depth fluctuation increases (see Fig. 4(c)). This observation is confirmed by our calculations given as white triangles in the inset of Fig. 4(d). The reason for the narrow line width is mainly caused by a shallow slope of the ‘‘energy–hole depth’’ dispersion relation for deep holes, i.e. the confinement energies change only slightly if the hole depth is altered. We carried out optical spectroscopy on a single GaAs/AlGaAs QD, by focusing the laser on the sample through a low-temperature microscope with high numerical aperture. For this purpose a low-density array of nanoholes was created, using about 1.7 ML InAs QDs grown at 500 C as the

ARTICLE IN PRESS O.G. Schmidt et al. / Physica E 25 (2004) 280–287

10 nm

After in-situ etching FWHM (meV)

284

8K

10 nm GaAs

In(Ga)As QDs

8.9 meV

In(Ga)As WL

GaAs

S(5,2)

45

15

(a)

3

Intensity (arb. units)

(b)

40 30

5.4 nm

20

±6%

6

h (nm)

9

QW

12 meV (7,2) 19 meV (10,2) 21 meV

9.0 nm

20

(7,1)

±9%

QDs

10

24 meV

QW

(15,2)

0 4

(c)

12 meV S(10,2)

30

10 0

Exp. Calc.

30

6

8 10 12

4

6

1.60

8 10 12

Hole depth h (nm)

(d)

1.65

1.70

1.75

1.80

Energy (eV)

Fig. 4. Method to obtain deeper nanoholes. (a) In situ etching of closely stacked double-dots yields deeper nanoholes due to larger strain fields. (b) AFM scans of holes obtained from single and double QD layer etching. Histograms in (c) reveal that the nanoholes from the double QD layer etching are deeper. (d) Various PL spectra from GaAs/AlGaAs QDs (see text for further explanations). A peak linewidth as narrow as 8.9 meV is achieved. The inset shows linewidth versus hole depth.

initial template. An AFM scan in Fig. 5(a) illustrates the low density of the nanoholes as well as the typical size of the focused laser spot. The corresponding m-PL spectra for two different excitation densities are given in Fig. 5(b). At 100 W/cm2 two sharp lines, separated by 2.8 meV, are clearly resolved. We attribute these lines to the ground state exciton (X) and biexciton (XX) recombination in a single GaAs/AlGaAs QD [29]. At higher excitation density multiple lines arise at the low energy side, which possibly originate from multiple exciton recombination [30]. At the high-energy side another peak appears at 1.637 eV (labelled E1), which we appoint to a transition involving excited states of the QD. Our measurements show that GaAs/AlGaAs QDs with well-defined geometry can be created by a combination of MBE and atomic-layer precise in situ etching (ALPISE). The GaAs QDs emit light in the optimum working range (red spectral range) of Si-photodetectors and Ti-sapphire lasers and therefore seem to be predestined for future coherent optical experiments and devices [19– 21,31–33].

3. Island growth and ex situ etching Recently, it has been shown that ex situ selective chemical etching constitutes a powerful tool to gain deep insight into the formation process as well as the compositional state [34–41] of selfassembled SiGe islands in single and multiple layers. Here, we refine this technique to create ultra-thin free-standing Si bridges on a Si (0 0 1) substrate surface. In the first experiment we have grown a single layer of SiGe dome islands on a Si (0 0 1) substrate surface at a high growth temperature (840 C) and capped the islands with 30 nm Si at low temperature (350 C), which is known to preserve the shape of the domes [42–44]. The sample is then removed from the MBE chamber, cleaved and inserted into 1HF:2H2O2:3CH3COOH for 2 min, which is known to selectively etch SiGe alloys over Si [45]. The result after the etching procedure is shown in Fig. 6. At the cleaving edge the etchant has gained access and therefore removed the SiGe core of the capped islands. As a result only the Si cap layer is left over and a tiny Si cavity is created at the sample edge.

ARTICLE IN PRESS O.G. Schmidt et al. / Physica E 25 (2004) 280–287

285

Typical laser spot size

500 nm

Intensity (arb. Units)

(a)

D = 7 nm T=4K φspot ≈ 1 µm

103 W/cm2 X XX

1.61 (b)

E0

E1

102 W/cm2

1.62 1.63 Energy (eV)

1.64

Fig. 5. (a) Low-density AlGaAs nanoholes and typical area of focussed laser spot. (b) Single QD spectra at low and high excitation power density. Narrow lines from exciton (X), biexciton (XX) and excited state (E1) recombination are visible.

A similar approach is outlined in Fig 7(a) in order to fabricate free-standing Si bridges. First, a single layer of low-temperature Si-capped Ge islands is grown on a Si (0 0 1) substrate. Using electron beam lithography and reactive ion etching, 105 nm wide and 65 nm deep mesa structures are processed into the surface. At those positions, where the mesa crosses the middle of an island, the flanks of the capped islands are truncated and the SiGe core becomes accessible to an etchant from two sides. The sample is treated by the selective etchant, which removes the SiGe core and leaves over a thin free-standing Si bridge. The result of this process is given in Fig. 7(b)–(d) showing a variety of different Si bridges. The bridge in Fig. 7(d) spans over a distance of 315 nm and has a height and thickness of 45 and 31 nm, respectively. Since the size of self-assembled SiGe

Fig. 6. Selective etching of Si capped SiGe islands. At the cleaved edge a small Si cavity is created.

islands is scalable over a large range [46], the structural parameters of the Si-bridge are also tuneable. As a next step the island sites need to be controlled [47–49] in order to create bridges at predefined positions. Finally, we note that our free-standing Si bridges put forward an alternative route to create Si-on-nothing (SON) [6,7]. SON has gained attention in advanced CMOS technology, because it attenuates short channel effects in ultra-small MOSFET devices.

Acknowledgements The continuous interest and support of K.v. Klitzing and the technical support of W. Winter and C. Muller . are greatfully acknowledged. The authors thank the Bundesministerium fur . Bildung und Forschung for financial support (contract number: 03N8711).

ARTICLE IN PRESS 286

O.G. Schmidt et al. / Physica E 25 (2004) 280–287

Fig. 7. Fabrication of ultra-thin free-standing Si bridges. (a) Schematic illustration of the required processing steps: Mesa definition on surface with Si-capped SiGe islands and subsequent selective etching of the SiGe core results in free-standing Si bridges. (b–d) Different views of the fabricated Si bridges.

References [1] D. Bouwmeester, A. Ekert, A. Zeilinger (Eds.), The Physics of Quantum Information, Springer, Berlin, 2001. [2] J.L. Hoyt et al., IEEE IEDM Technical Digest, December 2002, p. 23. [3] S.H. Olsen, A.G. O’Neill, L.S. Driscoll, K.S.K. Kwa, S. Chattopadhyay, A.M. Waite, Y.T. Tang, A.G.R. Evans, D.J. Norris, A.G. Cullis, D.J. Paul, D.J. Robbins, IEEE Trans. Electron Devices 50 (2003) 1961. [4] D.L. Harame, S.J. Koester, G. Freeman, P. Cottrel, K. Rim, G. Dehlinger, D. Ahlgren, J.S. Dunn, D. Greenberg, A. Joseph, F. Anderson, J.S. Rieh, S.A.S.T. Onge, D. Coolbaugh, V. Ramachandran, J.D. Cressler, S. Subbanna, Appl. Surf. Sci. 224 (2004) 9. [5] Z.Y. Cheng, A.J. Pitera, M.L. Lee, J.W. Jung, J.L. Hoyt, D.A. Antoniadis, E.A. Fitzgerald, IEEE Electron Devices Lett. 25 (2004) 147.

[6] M. Jurczak, T. Skotnicki, M. Paoli, B. Tormen, J. Martins, J.L. Regolini, D. Dutartre, P. Ribot, D. Lenoble, R. Pantel, S. Monfray, IEEE Trans. Electron Devices 47 (2000) 2179. [7] K. Pretet, S. Monfray, S. Cristoloveanu, T. Skotnicki, IEEE Trans. Electron Devices 51 (2000) 240. [8] W.T. Tsang, T.H. Chiu, R.M. Capre, Appl. Phys. Lett. 63 (1993) 3502. [9] T. Kaneko, P. Smilauer, B.A. Joyce, T. Kawamura, D.D. Vvedensky, Phys. Rev. Lett. 74 (1995) 3289. [10] H. Schuler, N.Y. Jin-Phillipp, F. Phillipp, K. Eberl, Semicond. Sci. Technol. 13 (1998) 1341. [11] H. Schuler, K. Eberl, Microelectron. J. 30 (1999) 341. [12] O.G. Schmidt, Ch. Deneke, S. Kiravittaya, R. Songmuang, H. Heidemeyer, Y. Nakamura, R. Zapf-Gottwick, C. Muller, . N.Y. Jin-Phillipp, IEEE J. Sel. Top. Quant. Electron. 8 (2002) 1025.

ARTICLE IN PRESS O.G. Schmidt et al. / Physica E 25 (2004) 280–287 [13] R. Songmunag, S. Kiravittaya, O.G. Schmidt, Appl. Phys. Lett. 82 (2003) 2892. [14] S. Kiravittaya, R. Songmuang, S. Panyakeow, O.G. Schmidt, J. Cryst. Growth 251 (2003) 258. [15] A. Rastelli, S. Stufler, A. Schliwa, R. Songmuang, C. Manzano, G. Costantini, K. Kern, A. Zrenner, D. Bimberg, O.G. Schmidt, Phys. Rev. Lett. 92 (2004) 166104. [16] T. Yang, S. Kohmoto, H. Nakamura, K. Asakawa, Physica E 13 (2002) 1151. [17] H. Schuler, T. Kaneko, M. Lipinski, K. Eberl, Semicond. Sci. Technol. 15 (2000) 169. [18] K. Matsuda, T. Saiki, S. Nomura, M. Mihara, Y. Aoyagi, S. Nair, T. Takagahara, Phys. Rev. Lett. 91 (2003) 177401. [19] T.H. Stievater, X. Li, D.G. Steel, D. Gammon, D.S. Katzer, D. Park, C. Piermarocchi, L.J. Sham, Phys. Rev. Lett. 87 (2001) 133603. [20] J. Hours, S. Varoutsis, M. Gallart, J. Bloch, I. RobertPhilip, A. Cavanna, I. Abram, F. Laruelle, J.M. Gerard, Appl. Phys. Lett. 82 (2003) 2206. [21] X. Li, Y. Wu, D. Steel, D. Gammon, T.H. Stievater, D.S. Katzer, D. Park, C. Piermarocchi, L.J. Sham, Science 301 (2003) 809. [22] M.O. Lipinski, H. Schuler, O.G. Schmidt, K. Eberl, N.Y. Jin-Phillipp, Appl. Phys. Lett. 77 (2000) 1789. [23] P.B. Joyce, T.J. Krzyzewski, P.H. Steans, G.R. Bell, J.H. Neave, T.S. Jones, Surf. Sci. 492 (2001) 345. [24] O.G. Schmidt, K. Eberl, Phys. Rev. B 61 (2000) 13721. [25] O.G. Schmidt, K. Eberl, Y. Rau, Phys. Rev. B 62 (2000) 16715. [26] A. Rastelli, M. Kummer, H. von K.anel, Phys. Rev. Lett. 87 (2001) 256101. [27] A. Hesse, J. Stangl, V. Holy, T. Roch, G. Bauer, O.G. Schmidt, U. Denker, B. Struth, Phys. Rev. B 66 (2002) 085321. [28] K. Shiraishi, Appl. Phys. Lett. 60 (1992) 1363. . [29] K. Brunner, G. Abstreiter, G. Bohm, G. Tr.ankle, G. Weimann, Phys. Rev. Lett. 73 (1992) 1138. [30] F. Findeis, A. Zrenner, G. Bohm, G. Abstreiter, Solid State Commun. 114 (2000) 227. [31] P. Michler, A. Kiraz, C. Becher, W.V. Schoenfeld, P.M. Petroff, L.D. Zhang, E. Hu, A. Imamoglu, Science 290 (2000) 2282.

287

[32] A. Zrenner, E. Beham, S. Stufler, F. Findeis, M. Bichler, G. Abstreiter, Nature 418 (2002) 612. [33] Z.L. Yuan, B.E. Kardynal, R.M. Stevenson, A.J. Shields, C.J. Lobo, K. Cooper, N.S. Beattie, D.A. Ritchie, M. Pepper, Science 295 (2002) 102. [34] U. Denker, O.G. Schmidt, N.Y. Jin-Phillipp, K. Eberl, Appl. Phys. Lett. 78 (2001) 3723. [35] O.G. Schmidt, U. Denker, S. Christiansen, F. Ernst, Appl. Phys. Lett. 81 (2002) 2614. [36] U. Denker, M. Stoffel, O.G. Schmidt, Phys. Rev. Lett. 90 (2003) 196102. [37] A. Malachias, S. Kycia, G. Medeiros-Ribeiro, R. Magalhaes-Paniago, T.I. Kamins, R.S. Williams, Phys. Rev. Lett. 91 (2003) 176101. [38] U. Denker, H. Sigg, O.G. Schmidt, Mater. Sci. Eng. B 101 (2003) 89. [39] U. Denker, M. Stoffel, O.G. Schmidt, Appl. Phys. Lett. 83 (2003) 1432. [40] U. Denker, H. Sigg, O.G. Schmidt, Appl. Surf. Sci. 224 (2004) 127. [41] T.U. Schulli, . M. Stoffel, A. Hesse, R.T. Lechner, E. Wintersberger, M. Sztucki, T.H. Metzger, O.G. Schmidt, G. Bauer, Phys. Rev. B, unpublished. [42] O.G. Schmidt, U. Denker, M. Dashiell, N.Y. Jin-Phillipp, F. Ernst, Science and Technology of Silicon Materials— 2001: FSRC Book of Abstracts, Bulletin of the Stefan University, Vol. 13, No. 113, August 2001, p. 59, ISBN: 1-889545–40–6. [43] O.G. Schmidt, U. Denker, M. Dashiell, N.Y. Jin-Phillipp, K. Eberl, R. Schreiner, H. Gr.abeldinger, H. Schweizer, S. Christiansen, F. Ernst, Mater. Sci. Eng. B 89 (2002) 101. [44] A. Rastelli, E. Muller, . H. von K.anel, Appl. Phys. Lett. 80 (2002) 1438. [45] T.K. Carns, M.O. Tanner, K.L. Wang, J. Electrochem. Soc. 142 (1995) 1260. [46] O.G. Schmidt, C. Lange, K. Eberl, Appl. Phys. Lett. 75 (1999) 1905. [47] O.G. Schmidt, N.Y. Jin-Phillipp, C. Lange, U. Denker, K. Eberl, R. Schreiner, H. Gr.abeldinger, H. Schweizer, Appl. Phys. Lett. 77 (2000) 4139. [48] O.G. Schmidt, K. Eberl, IEEE Electron Devices Lett. 48 (2001) 1175. [49] Z. Zhong, G. Bauer, Appl. Phys. Lett. 84 (2004) 1922.