Atomic structure and stoichiometry of In(Ga)As/GaAs quantum dots grown on anexact-oriented GaP/Si(001) substrateC. S. Schulze, X. Huang, C. Prohl, V. Füllert, S. Rybank, S. J. Maddox, S. D. March, S. R. Bank, M. L. Lee, andA. Lenz Citation: Applied Physics Letters 108, 143101 (2016); doi: 10.1063/1.4945598 View online: http://dx.doi.org/10.1063/1.4945598 View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/108/14?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Nanotemplate-directed InGaAs/GaAs single quantum dots: Toward addressable single photon emitter arrays J. Vac. Sci. Technol. B 32, 02C106 (2014); 10.1116/1.4863680 Spatial structure of In0.25Ga0.75As/GaAs/GaP quantum dots on the atomic scale Appl. Phys. Lett. 102, 123102 (2013); 10.1063/1.4798520 Long wavelength (>1.55 μm) room temperature emission and anomalous structural properties of InAs/GaAsquantum dots obtained by conversion of In nanocrystals Appl. Phys. Lett. 102, 073103 (2013); 10.1063/1.4792700 Single quantum dot emission at telecom wavelengths from metamorphic InAs/InGaAs nanostructures grown onGaAs substrates Appl. Phys. Lett. 98, 173112 (2011); 10.1063/1.3584132 InGaAs/GaAs quantum dots on (111)B GaAs substrates J. Appl. Phys. 84, 2624 (1998); 10.1063/1.368373

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 128.62.27.20 On: Tue, 05 Apr 2016

16:43:22

http://scitation.aip.org/content/aip/journal/apl?ver=pdfcovhttp://oasc12039.247realmedia.com/RealMedia/ads/click_lx.ads/www.aip.org/pt/adcenter/pdfcover_test/L-37/2051104815/x01/AIP-PT/APL_ArticleDL_032316/AIP-APL_Photonics_Launch_1640x440_general_PDF_ad.jpg/434f71374e315a556e61414141774c75?xhttp://scitation.aip.org/search?value1=C.+S.+Schulze&option1=authorhttp://scitation.aip.org/search?value1=X.+Huang&option1=authorhttp://scitation.aip.org/search?value1=C.+Prohl&option1=authorhttp://scitation.aip.org/search?value1=V.+F�llert&option1=authorhttp://scitation.aip.org/search?value1=S.+Rybank&option1=authorhttp://scitation.aip.org/search?value1=S.+J.+Maddox&option1=authorhttp://scitation.aip.org/search?value1=S.+D.+March&option1=authorhttp://scitation.aip.org/search?value1=S.+R.+Bank&option1=authorhttp://scitation.aip.org/search?value1=M.+L.+Lee&option1=authorhttp://scitation.aip.org/search?value1=A.+Lenz&option1=authorhttp://scitation.aip.org/content/aip/journal/apl?ver=pdfcovhttp://dx.doi.org/10.1063/1.4945598http://scitation.aip.org/content/aip/journal/apl/108/14?ver=pdfcovhttp://scitation.aip.org/content/aip?ver=pdfcovhttp://scitation.aip.org/content/avs/journal/jvstb/32/2/10.1116/1.4863680?ver=pdfcovhttp://scitation.aip.org/content/aip/journal/apl/102/12/10.1063/1.4798520?ver=pdfcovhttp://scitation.aip.org/content/aip/journal/apl/102/7/10.1063/1.4792700?ver=pdfcovhttp://scitation.aip.org/content/aip/journal/apl/102/7/10.1063/1.4792700?ver=pdfcovhttp://scitation.aip.org/content/aip/journal/apl/98/17/10.1063/1.3584132?ver=pdfcovhttp://scitation.aip.org/content/aip/journal/apl/98/17/10.1063/1.3584132?ver=pdfcovhttp://scitation.aip.org/content/aip/journal/jap/84/5/10.1063/1.368373?ver=pdfcov

Atomic structure and stoichiometry of In(Ga)As/GaAs quantum dotsgrown on an exact-oriented GaP/Si(001) substrate

C. S. Schulze,1 X. Huang,2 C. Prohl,1 V. F€ullert,1 S. Rybank,1 S. J. Maddox,3 S. D. March,3

S. R. Bank,3 M. L. Lee,2 and A. Lenz1,2,a)1Institut f€ur Festk€orperphysik, Technische Universit€at Berlin, D-10623 Berlin, Germany2Department of Electrical Engineering, Yale University, New Haven, Connecticut 06520, USA3Microelectronics Research Center and ECE Department, The University of Texas at Austin, Austin,Texas 78758, USA

(Received 16 October 2015; accepted 25 March 2016; published online 4 April 2016)

The atomic structure and stoichiometry of InAs/InGaAs quantum-dot-in-a-well structures grown

on exactly oriented GaP/Si(001) are revealed by cross-sectional scanning tunneling microscopy.

An averaged lateral size of 20 nm, heights up to 8 nm, and an In concentration of up to 100% are

determined, being quite similar compared with the well-known quantum dots grown on GaAs sub-

strates. Photoluminescence spectra taken from nanostructures of side-by-side grown samples on

GaP/Si(001) and GaAs(001) show slightly blue shifted ground-state emission wavelength for

growth on GaP/Si(001) with an even higher peak intensity compared with those on GaAs(001).

This demonstrates the high potential of GaP/Si(001) templates for integration of III-V optoelec-

tronic components into silicon-based technology. VC 2016 AIP Publishing LLC.[http://dx.doi.org/10.1063/1.4945598]

The incorporation of III-V laser structures on Si(001)

substrates is highly desirable for Si photonics and on-chip

interconnects.1–4 However, despite considerable interest,

direct growth of III-V semiconductors on Si is plagued by

high dislocation densities of 105–108 cm�2, which act as

non-radiative recombination centers that greatly limit device

performance.5 Often, III-V laser structures are grown on

slightly off-cut Si(001) substrates5–7 in order to eliminate

anti-phase domains, but for an integration into complemen-

tary metal-oxide-semiconductor (CMOS) technology, the

growth on exactly oriented Si(001) substrates8–12 is highly

preferable.

In order to optimize the device quality, a fundamental

understanding of the atomic structure and the electronic

properties is of major importance. However, compared with

growth on III-V substrates, relatively little is known about

the atomic structure and stoichiometry of III-V nanostruc-

tures for laser devices grown on silicon substrates.13,14

Cross-sectional scanning tunneling microscopy (XSTM)

is a powerful method to investigate the structural details of

III-V nanostructures, e.g., size, shape, and stoichiometry,

grown on different III-V substrates.15–17 A clean cross-

sectional III/V{110} surface for XSTM investigations can

easily be prepared by cleaving the III-V(001) substrate in an

ultrahigh-vacuum (UHV) chamber. In case of III-V layers

grown on a Si(001) substrate, the preparation of III-V/

Si{110} surfaces is not as trivial and frequently results in

{111} oriented terraces on the silicon substrate, also deterio-

rating the quality of the III-V/{110} cleavage surface. Thus,

only a few XSTM studies on in-situ cleaved Si{110} surfa-ces are yet reported.18

Here, we present XSTM studies on the atomic structure

and stoichiometry of InAs/InGaAs quantum-dot-in-a-well

(DWELL) structures grown in a GaAs matrix on an exact-

oriented GaP/Si(001) template and compare the results with

those of corresponding DWELL structures on bulk

GaAs(001).19

The DWELL structures were grown on a high-quality

epitaxial GaP/Si(001) template, with an offcut angle below

0.5�, provided by NAsPIII/V GmbH.20 It consists of a 46 nm

thick GaP layer grown by metalorganic vapor-phase epitaxy

(MOVPE) on an n-doped Si buffer and on a p-doped Si(001)substrate.

The molecular beam epitaxy (MBE) growth of the arse-

nide structures was performed simultaneously on the GaP/

Si(001) template and on bulk GaAs(001) positioned side-by-

side. After oxide desorption at 650 �C for 30 min under As2overpressure, the growth started with 2.3 lm n-doped GaAsat 630 �C followed by undoped layers of 150 nm GaAs and10 nm AlAs [Fig. 1(a)]. The DWELL structures are then

sandwiched between 80 nm GaAs grown with a temperature

ramp from 630 �C to 525 �C and vice versa and cappedfinally with 10 nm AlAs and 10 nm GaAs. Each of the three

DWELL structures is grown at 525 �C with the followingsequence: 20 nm GaAs, 2 nm In0.15Ga0.85As, 3.1 monolayers

(ML) InAs, 6 nm In0.15Ga0.85As, and 20 nm GaAs with

growth rates of 0.58 lm/h, 0.69 lm/h, 0.1 ML/s, 0.69 lm/h,and 0.58 lm/h, respectively.

The XSTM experiment was performed using a home-built

setup with an RHK Technology SPM 1000 control unit. The

800 lm thick sample was manually thinned down to about100 lm by using a 30 lm diamond slurry and cleaved in-situin UHV at a base pressure below 1� 10–8 Pa, providing aclean (110) cleavage surface. Electrochemically etched tung-

sten wires were used as tips, which were further cleaned in-situby electron bombardment. Photoluminescence (PL) spectros-

copy were measured at room temperature using a frequency-

doubled Nd:YAG pump laser (532 nm), a nitrogen-purgeda)Electronic mail: alenz@physik.tu-berlin.de

0003-6951/2016/108(14)/143101/4/$30.00 VC 2016 AIP Publishing LLC108, 143101-1

APPLIED PHYSICS LETTERS 108, 143101 (2016)

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 128.62.27.20 On: Tue, 05 Apr 2016

16:43:22

http://dx.doi.org/10.1063/1.4945598http://dx.doi.org/10.1063/1.4945598http://dx.doi.org/10.1063/1.4945598mailto:alenz@physik.tu-berlin.dehttp://crossmark.crossref.org/dialog/?doi=10.1063/1.4945598&domain=pdf&date_stamp=2016-04-04

0.5 lm grating spectrometer, a liquid-nitrogen-cooled InSb de-tector, and a lock-in amplifier.

Figure 1(a) shows a sketch of the entire sample structure,

while an XSTM overview image of the GaAs/GaP/Si inter-

face is presented in Fig. 1(b). From the latter, one can easily

distinguish between the III-V layers and the silicon substrate

due to the different densities and orientations of surface steps.

On the silicon surface, many steps—oriented along the h112idirections—can be seen, while the GaP and GaAs surfaces are

almost free of steps. The GaAs/GaP interface shows a high

crystal quality despite of the lattice mismatch. The three dif-

ferent materials can also be distinguished by analyzing the

different electronic properties by taking current-voltage (I-V)curves in the scanning tunneling spectroscopy (STS) mode

[Fig. 1(d)]. As expected, the I-V curve for Si shows an almostmetallic behavior,18 while for GaAs and GaP, clear plateaus

with zero slope can be seen. Furthermore, the larger extension

of the plateau in case of GaP compared with GaAs is in agree-

ment with its larger bulk band gap. An additional structural

investigation of the GaP/Si and GaAs/GaP interfaces as well

as a more detailed STS study will be presented elsewhere.

Figure 1(c) shows overlapping XSTM overview images

starting from the undoped GaAs and AlAs layers towards the

first two DWELL layers. The AlAs layer can be identified as

a dark stripe due to the larger band gap of AlAs compared

with GaAs, while the DWELL layers appear as bright stripes.

The highest contrast within these stripes is related to the quan-

tum dots (QDs) located within the In0.15Ga0.85As layers. The

bright appearance of the DWELL layers as compared with the

surrounding GaAs matrix results from a combination of the

electronic contrast and of the compressive strain due to the

lattice mismatch of about 7% between InAs and GaAs. The

latter results in an elastic outward relaxation after sample

cleavage, that can be probed with the STM tip and is highest

at the DWELL centers.21 It should be noted that the third

DWELL layer was located directly at a pronounced surface

step and could thus not be imaged.

The PL analysis presented in Fig. 2 compares the optical

properties of the side-by-side grown samples on GaP/Si(001)

and GaAs(001) substrates. For each sample, PL spectra were

taken at different positions in order to study the sample ho-

mogeneity. The DWELL ground-state emission for growth

on GaP/Si varies between 1270 and 1275 nm, while for

growth on GaAs, it varies between 1305 and 1310 nm. Thus,

the spectra for growth on GaP/Si(001) are slightly blue

shifted with respect to the growth on GaAs substrate. Such a

blue shift was already observed5,22 and attributed to the re-

sidual compressive strain in the GaAs buffer layer on Si sub-

strate, resulting in slightly smaller QDs.22 Higher PL peak

intensities for growth on GaP/Si substrates compared with

growth on GaAs substrates were observed, not only for this

specific sample but also for several samples with varying

growth parameters, all showing good PL efficiency. A more

detailed analysis on the spectral shape, e.g., as the additional

shoulders, will be given elsewhere.

Figure 3 shows XSTM images of DWELL 1 and

DWELL 2 with several slightly differently appearing QDs.

The roughly truncated-pyramidal shape of the QDs is

marked by dashed lines in Figs. 3(b) and 3(c) for guiding the

eye. It should be noted here that in filled-states images the

image contrast induced by strain relaxation dominates,15

resulting at first view in a rounder appearance of the QDs.

However, a detailed inspection of the contrast reveals a

shape very close to a truncated pyramid.

FIG. 1. (a) Schematic of the sample structure grown on GaP/Si(001) sub-

strates. Growth on GaAs(001) substrates starts at the dashed line. (b) XSTM

overview image of the GaAs/GaP/Si interface region, taken at a sample volt-

age of VS¼þ2.5 V and a tunneling current of IT¼ 40 pA, and (c) assemblyof XSTM overview images of the InAs/InGaAs DWELL-structure, taken at

VS between �3.0 V and �3.5 V and IT¼ 20 pA. (d) I-V curves of Si(squares), GaAs (triangles), and GaP (circles) acquired at a stepped (�110)-cleavage surface (not shown here) with the tip position stabilized at

Vstab¼�1.4 V and Istab¼ 70 pA.

FIG. 2. Room temperature PL spectra for DWELL structures grown on GaP/

Si(001) (solid, green) and on bulk GaAs(001) (dashed, blue), taken with an

excitation power density of 2.4 kW cm�2. The ground-state emission peaksare labeled A and B, peak C is due to the pump laser.

143101-2 Schulze et al. Appl. Phys. Lett. 108, 143101 (2016)

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 128.62.27.20 On: Tue, 05 Apr 2016

16:43:22

Within the evaluated scan area extending across

�0.9 lm, a total number of about 30 QDs are analyzed. QDswith heights up to 28 ML (�8 nm) and base lengths between15 and 25 nm are observed. This wide range of apparent base

lengths is connected to the randomly distributed cleavage

positions so that the actual average base length is closer to

25 nm, leading to a density of about 6(62)� 1010 cm�2.These structural findings are quite similar to the averaged

values for DWELL structures grown on a GaAs(001) sub-

strate with emission wavelengths slightly below 1.3 lm.19

In Fig. 3(c), the bright feature (labeled with A) with

monoatomic step height in the center of the QD is most

likely a cleavage-induced defect, as it was already observed

in XSTM investigations of the highly strained diluted-nitride

system.23 Such cleavage-induced defects could appear as

bright objects or also as dark depressions, e.g., as the feature

labeled B in Fig. 3(c). It should be noted that features labeled

C are no defects but can be attributed to a slight contamina-

tion after cleavage.

In previous XSTM studies of MBE grown DWELL

structures and of MOVPE grown In0.80Ga0.20As QDs cov-

ered by an In0.10Ga0.90As layer, also the formation of the so-

called nanovoids during growth was reported, especially in

case of larger QD sizes.19 Such nanovoids, characterized by

a material hole, are also found within both DWELL layers,

by XSTM (not shown here), as well as by transmission elec-

tron microscopy.24 Interestingly, 17 6 6% and 14 6 3%nanovoids, and about 65 6 15% and 20 6 5% cleavage-induced defects are imaged at DWELL 1 and DWELL 2,

respectively. Since both layers are grown using identical

growth parameters, a similar occurrence would be expected

based on DWELL growth alone. Thus, it is most likely that

defects are also related to dislocations running from the GaP/

Si substrate or the GaAs/GaP interface towards the sample

surface and being partially stopped by the strained DWELL

layers. It should be noted that the incorporation of strained

layers (or even DWELL layers) below the active region as

the so-called dislocation filter layers is already used in order

to improve the device quality of laser structures grown on

silicon substrates.3,6

In Fig. 4(a), a close-view XSTM image of an atomically

resolved QD from DWELL 2 is presented. The roughly

truncated-pyramidal QD shape is highlighted by dashed

lines, as also illustrated in the sketch in the inset image. The

shape of its In-rich center, characterized by straight bounda-

ries as marked by dotted lines, is similar to a reversed trun-

cated cone.

Figure 4(b) shows the variation of the local lattice param-

eter as a function of the position along growth direction.21 By

comparing the resulting graphs with simulations of the strain

relaxation of two-dimensional quantum wells, the local In

concentration x is obtained.25 The local stoichiometry is deter-mined for the QD shown in Fig. 4(a), separately for the center

and its sides, as shown in the XSTM inset image [Fig. 4(b)],

and compared with the In0.15Ga0.85As layer between the QDs,

labeled as wetting layer (WL). The In0.15Ga0.85As layer

between the QDs has a maximum In concentration of about

15% at its base, slightly decreasing along growth direction. A

much higher In concentration is found in the QD. For the QD

sides, a maximum In concentration of about 70% is observed,

while the highest In concentration of up to 100% is located in

the upper QD center. A similar In distribution was already

observed for In0.5Ga0.5As/GaAs QDs.26 It should be noted

that the reason for the data points exceeding the value for

FIG. 3. XSTM filled-states images of (a), (b) DWELL 2, and (c) DWELL 1,

showing QDs with roughly a truncated-pyramidal shape, as indicated by the

dashed lines for guiding the eye. In (c), QDs with cleavage-induced defects

(labeled A and B) are observed. A few rounder bright spots (exemplarily

marked, labeled C) are attributed to a slight surface contamination after

cleavage. The XSTM images are taken at (a), (b) VS¼�2.7 V, and (c)VS¼�3.0 V, and at IT¼ 20 pA.

FIG. 4. (a) XSTM filled-states image of a QD at DWELL 2, taken at

VS¼�2.8 V and IT¼ 20 pA. The DWELL shape is marked by dashed linesand the In-rich region by dotted lines, further illustrated by the inset sketch.

(b) Determination of the local lattice parameter (left axis) as a function of

the position along growth direction and the corresponding In concentration

(right axis) of the QD center (circles), the QD sides (squares), and the WL

(triangles). The graphs from QD center and QD side are determined within

the areas marked by the boxes in the XSTM inset image. The uncertainty of

the obtained In concentrations amounts to about 10%.

143101-3 Schulze et al. Appl. Phys. Lett. 108, 143101 (2016)

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 128.62.27.20 On: Tue, 05 Apr 2016

16:43:22

pure InAs is related only to the local bending of the cleavage

surface due to the high compressive strain in this system.21

The region of the In-rich center and the total QD height

derived from the image contrast is marked in Fig. 4(b), being

in nice agreement with the variation of the local lattice

parameter.

In conclusion, we presented XSTM and PL data of III-V

nanostructures grown on exactly oriented GaP/Si(001) sub-

strates. The preparation of a (110) cleavage plane through

the GaAs/GaP/Si layer structure allows the study of the

atomic structure and stoichiometry of In(Ga)As/GaAs QDs,

revealing lateral sizes close to 25 nm, heights up to 8 nm, and

an In concentration of up to 100%. In addition, a compara-

tive PL study of side-by-side grown DWELL structures on

GaP/Si(001) and GaAs(001) samples exhibits a slight blue

shift to a ground-state emission wavelength above 1.27 lmtogether with an increase in the PL peak intensity for growth

on GaP/Si(001). These properties demonstrate the high

potential for integration of such III-V nanostructures into

silicon-based technology.

The authors thank the Deutsche Forschungsgemeinschaft

for financial support of Project No. LE3317/1-1, U. W. Pohl

for fruitful discussions, M. D€ahne for providing the XSTMsetup, and A. Beyer and R. Straubinger for characterization of

the sample orientation. Work at UT-Austin was supported by

a Multidisciplinary University Research Initiative from the

Air Force Office of Scientific Research (AFOSR MURI

Award No. FA9550-12-1-0488).

1W. I. Wang, Appl. Phys. Lett. 44, 1149 (1984).2M. Razeghi, M. Defour, F. Omnes, Ph. Maurel, J. Chazelas, and F.

Brillouet, Appl. Phys. Lett. 53, 725 (1988).3Z. Mi, J. Yang, P. Bhattacharya, G. Qin, and Z. Ma, Proc. IEEE 97, 1239(2009).

4D. Liang and J. E. Bowers, Nat. Photonics 4, 511 (2010).5A. Y. Liu, S. Srinivasan, J. Norman, A. C. Gossard, and J. E. Bowers,

Photonics Res. 3, B1 (2015).6A. Lee, H. Liu, and A. Seeds, Semicond. Sci. Technol. 28, 015027 (2013).

7B. Corbett, C. Bower, A. Fecioru, M. Mooney, M. Gubbins, and J. Justice,

Semicond. Sci. Technol. 28, 094001 (2013).8S. Neumann, A. Bakin, P. Velling, W. Prost, H.-H. Wehmann, A.

Schlachetzki, and F.-J. Tegude, J. Cryst. Growth 248, 380 (2003).9S. Liebich, M. Zimprich, A. Beyer, C. Lange, D. J. Franzbach, S.

Chatterjee, N. Hossain, S. J. Sweeney, K. Volz, B. Kunert, and W. Stolz,

Appl. Phys. Lett. 99, 071109 (2011).10L. Desplanque, S. El Kazzi, C. Coinon, S. Ziegler, B. Kunert, A. Beyer, K.

Volz, W. Stolz, Y. Wang, P. Ruterana, and X. Wallart, Appl. Phys. Lett.

101, 142111 (2012).11Y. Song and M. L. Lee, Appl. Phys. Lett. 103, 141906 (2013).12X. Huang, Y. Song, T. Masuda, D. Jung, and M. Lee, Electron. Lett. 50,

1226 (2014).13K. K. Linder, J. Phillips, O. Qasaimeh, X. F. Liu, S. Krishna, P.

Bhattacharya, and J. C. Jiang, Appl. Phys. Lett. 74, 1355 (1999).14J. Shen, Y. Song, M. L. Lee, and J. J. Cha, Nanotechnology 25, 465702

(2014).15A. Lenz, R. Timm, H. Eisele, Ch. Hennig, S. K. Becker, R. L. Sellin, U.

W. Pohl, D. Bimberg, and M. D€ahne, Appl. Phys. Lett. 81, 5150 (2002).16A. Lenz, E. Tourni�e, J. Schuppang, M. D€ahne, and H. Eisele, Appl. Phys.

Lett. 102, 102105 (2013).17C. Prohl, A. Lenz, D. Roy, J. Schuppang, G. Stracke, A. Strittmatter, U.

W. Pohl, D. Bimberg, H. Eisele, and M. D€ahne, Appl. Phys. Lett. 102,123102 (2013).

18M. A. Lutz, R. M. Feenstra, and J. O. Chu, Surf. Sci. 328, 215 (1995); Ph.Ebert, S. Schaafhausen, A. Lenz, A. Sabitova, L. Ivanova, M. D€ahne, Y.-L. Hong, S. Gwo, and H. Eisele, Appl. Phys. Lett. 98, 062103 (2011).

19A. Lenz, H. Eisele, R. Timm, S. K. Becker, R. L. Sellin, U. W. Pohl, D.

Bimberg, and M. D€ahne, Appl. Phys. Lett. 85, 3848 (2004); A. Lenz, H.Eisele, R. Timm, L. Ivanova, H.-Y. Liu, M. Hopkinson, U. W. Pohl, and

M. D€ahne, Phys. E 40, 1988 (2008); A. Lenz, H. Eisele, R. Timm, L.Ivanova, R. L. Sellin, H.-Y. Liu, M. Hopkinson, U. W. Pohl, D. Bimberg,

and M. D€ahne, Phys. Status Solidi B 246, 717 (2009).20B. Kunert, S. Zinnkann, K. Volz, and W. Stolz, J. Cryst. Growth 310, 4776

(2008); K. Volz, A. Beyer, W. Witte, J. Ohlmann, I. N�emeth, B. Kunert,and W. Stolz, ibid. 315, 37 (2011).

21O. Flebbe, H. Eisele, T. Kalka, F. Heinrichsdorff, A. Krost, D. Bimberg,

and M. D€ahne-Prietsch, J. Vac. Sci. Technol. B 17, 1639 (1999).22D. Lacombe, A. Ponchet, J.-M. G�erard, and O. Cabrol, Appl. Phys. Lett.

70, 2398 (1997).23H. A. McKay, R. M. Feenstra, T. Schmidtling, U. W. Pohl, and J. F. Geisz,

J. Vac. Sci. Technol. B 19, 1644 (2001).24A. Beyer, private communication (2016).25H. Eisele, A. Lenz, R. Heitz, R. Timm, M. D€ahne, Y. Temko, T. Suzuki,

and K. Jacobi, J. Appl. Phys. 104, 124301 (2008).26N. Liu, J. Tersoff, O. Baklenov, A. L. Holmes, Jr., and C. K. Shih, Phys.

Rev. Lett. 84, 334 (2000).

143101-4 Schulze et al. Appl. Phys. Lett. 108, 143101 (2016)

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 128.62.27.20 On: Tue, 05 Apr 2016

16:43:22

http://dx.doi.org/10.1063/1.94673http://dx.doi.org/10.1063/1.99815http://dx.doi.org/10.1109/JPROC.2009.2014780http://dx.doi.org/10.1038/nphoton.2010.167http://dx.doi.org/10.1364/PRJ.3.0000B1http://dx.doi.org/10.1088/0268-1242/28/1/015027http://dx.doi.org/10.1088/0268-1242/28/9/094001http://dx.doi.org/10.1016/S0022-0248(02)01852-3http://dx.doi.org/10.1063/1.3624927http://dx.doi.org/10.1063/1.4758292http://dx.doi.org/10.1063/1.4824029http://dx.doi.org/10.1049/el.2014.2077http://dx.doi.org/10.1063/1.123548http://dx.doi.org/10.1088/0957-4484/25/46/465702http://dx.doi.org/10.1063/1.1533109http://dx.doi.org/10.1063/1.4795020http://dx.doi.org/10.1063/1.4795020http://dx.doi.org/10.1063/1.4798520http://dx.doi.org/10.1016/0039-6028(95)00038-0http://dx.doi.org/10.1063/1.3553022http://dx.doi.org/10.1063/1.1808884http://dx.doi.org/10.1016/j.physe.2007.09.041http://dx.doi.org/10.1002/pssb.200880587http://dx.doi.org/10.1016/j.jcrysgro.2008.07.097http://dx.doi.org/10.1016/j.jcrysgro.2010.10.036http://dx.doi.org/10.1116/1.590803http://dx.doi.org/10.1063/1.118863http://dx.doi.org/10.1116/1.1379967http://dx.doi.org/10.1063/1.3042216http://dx.doi.org/10.1103/PhysRevLett.84.334http://dx.doi.org/10.1103/PhysRevLett.84.334