AlGaAs Core-Shell Nanowires - ACS Publications

For example, V-shaped grooves in GaAs substrates, typically composed of ..... The closest corresponding Miller indices are {1 3 4} and {19 1 0}. The t...
0 downloads 0 Views 780KB Size
Subscriber access provided by Kaohsiung Medical University

Communication

Connecting Composition-Driven Faceting with Facet-Driven Composition Modulation in GaAs-AlGaAs Core-Shell Nanowires Nari Jeon, Daniel Ruhstorfer, Markus Döblinger, Sonja Matich, Bernhard Loitsch, Gregor Koblmueller, and Lincoln J. Lauhon Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b02104 • Publication Date (Web): 11 Jul 2018 Downloaded from http://pubs.acs.org on July 12, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Connecting Composition-Driven Faceting with Facet-Driven Composition Modulation in GaAsAlGaAs Core-Shell Nanowires Nari Jeon1,†, Daniel Ruhstorfer2, Markus Döblinger3, Sonja Matich2, Bernhard Loitsch2, Gregor Koblmüller,,*2, Lincoln Lauhon1,* 1

Department of Materials Science and Engineering, Northwestern University, Evanston, Illinois

60208, United States 2

Walter Schottky Institut, Physik Department, and Center for Nanotechnology and

Nanomaterials, Technische Universität München, Garching, 85748, Germany 3

Department of Chemistry, Ludwig-Maximilians-Universität München, Munich, 81377,

Germany Keywords: nanowire, III-V, heterostructure, alloy composition, non-planar geometry

Ternary III-V alloys of tunable bandgap are a foundation for engineering advanced optoelectronic devices based on quantum-confined structures including quantum wells, nanowires, and dots. In this context, core-shell nanowires provide useful geometric degrees of freedom in heterostructure design, but alloy segregation is frequently observed in epitaxial shells even in the absence of interface strain. High-resolution scanning transmission electron 1 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

microscopy and laser-assisted atom probe tomography were used to investigate the driving forces of segregation in non-planar GaAs-AlGaAs core-shell nanowires. Growth temperature dependent studies of Al-rich regions growing on radial {112} nanofacets suggest that facet-dependent bonding preferences drive the enrichment, rather than kinetically limited diffusion. Observations of the distinct interface faceting when pure AlAs is grown on GaAs confirm the preferential bonding of Al on {112} facets over {110} facets, explaining the decomposition behavior. Furthermore, three dimensional composition profiles generated by atom probe tomography reveal the presence of Al-rich nanorings perpendicular to the growth direction; correlated electron microscopy shows that short zincblende insertions in a nanowire segment with predominantly wurtzite structure are enriched in Al, demonstrating that crystal phase engineering can be used to modulate composition. The findings suggest strategies to limit alloy decomposition and promote new geometries of quantum confined structures.

Group III-V heterostructure nanowires show great promise for compact electronic and optoelectronic

devices

lasers,1-3

including

photodetectors,10-12 and transistors13,

14

solar

cells,4-6

light-emitting

diodes,7-9

. Optimization of device performance requires precise

control over injection, confinement, transport and collection of charge carriers, each of which are influenced by energy band discontinuities and built-in potentials at heterointerfaces. In core-shell nanowires, it is often advantageous to employ shells of ternary III-V alloys to tune interfacial strain, band offsets, and carrier confinement. For example, shells of higher bandgap can passivate surface states and confine charge carriers to the core1, 9, 10, 13 or quantum wells formed within multilayer shells.15,

16

However, the use of ternary alloys creates the possibility of

2 ACS Paragon Plus Environment

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

composition fluctuations and associated perturbations of the potential landscape, which may localize charge carriers and undermine efforts to controllably tune the bandgap and emission wavelength in light emitting devices. Indeed, the ternary shells in many III-V core-shell nanowires display evidence of segregation including Ga rich nanostructures acting as quantum dots,17-20 which may be useful as single photon sources, and alloy enriched bands along radial directions on {112} nanofacets.18,

20, 21

Greater understanding of the driving forces of

compositional segregation on the non-planar growth surfaces is necessary to develop effective bandgap tuning strategies in ternaries and multicomponent systems, and to exploit these quantum dot and quantum wire features in quantum electronics and photonics. Epitaxy of III-V ternary alloys on patterned planar substrates has been explored as a route to ordered arrays of quantum heterostructures. For example, V-shaped grooves in GaAs substrates, typically composed of {100} or {111} main facets and high-index side facets, induce the formation of ordered arrays of quantum wires and quantum dots on the grooves;22, 23 the gradient of the chemical potential established by multiple facets generates preferential diffusion of one species into desired locations.24, 25 Similarly, vicinal substrates with high step densities have been used to form arrays of quantum dots and quantum wires.26-28 However, the same driving forces that are exploited for nanostructure formation can make it difficult to grow quantum heterostructures on nanowire sidewalls in a well-controlled fashion.29, 30 For example, multiple groups have observed segregation of group III species into bands along the radial directions as shown schematically in Figure 1(a),17-21,

31-33

and alloy fluctuations on {110}

growth facets.18, 19, 31 Given the negligible misfit strain at GaAs-AlGaAs interfaces, GaAs-AlGaAs core-shell nanowires are a good model system in which to study the correlation of composition and surface 3 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

faceting in the absence of strain driven segregation, which would otherwise complicate the analysis. Here we explore how the distinct binding preferences of Ga and Al influence the evolution of the GaAs, AlGaAs, and AlAs growth interfaces, and we relate these preferences to non-uniformities in the ternary alloy shell. In addition, we find that crystal phase modulations in the GaAs core seed axial composition modulations in the ternary shell in the form of nanodisks, which have not been previously reported. These findings provide new insights into how composition influences the growth interface in heterostructures, and how crystal phase and faceting can be used to engineer new geometries of semiconductor heterostructures. Molecular beam epitaxy was used to grow GaAs nanowires by a vapor-liquid-solid mechanism mediated by Ga droplets on Si (111) substrates.34 The crystal phase is predominantly zincblende with stacking faults, but regions of mixed wurtzite and zincblende phases are also observed as described in the Supporting Information (Supporting Figures S1a and S2). AlGaAs shells were subsequently grown at temperatures ranging from 370 °C to 560 °C under high V/III ratio as previously reported.31 Both zincblende and wurtzite regions of GaAs-AlGaAs nanowires were analyzed in this work. For comparison, a sample was also grown consisting of an AlAs-GaAs multishell structure with variable AlAs/GaAs layer thicknesses. The zincblende region of the AlAs-GaAs nanowire was analyzed as discussed below. Atom probe tomography (APT), scanning transmission electron microscopy (STEM) and TEM were used to analyze the composition and structure of the nanowires. Cross-sectional lamellas (~100-nm thin) were prepared using a focused ion beam for Z-contrast imaging using high-angle annual dark field (HAADF)-STEM. Imaging by HAADF-STEM, TEM and energy dispersive x-ray (EDX) spectroscopy in STEM mode were also conducted in plan view on nanowires dispersed on carbon film coated copper grids. A FEI Titan Themis microscope at 300 kV was used for both 4 ACS Paragon Plus Environment

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

TEM and STEM. APT samples were prepared by methods reported previously (Supporting Figure S1b),19 and laser assisted APT was performed under the following conditions: 0.02 – 0.05 pJ laser power, 500 kHz laser pulse repetition rate, 0.2 – 0.5% detection rate. In the work described in Figures 1–3 below, cross-sectional STEM was used to determine the shape of growth interfaces in core-shell and core-multishell structures, and APT was used primarily to analyze compositional segregation effects. Figure 1(b) shows an Al mole fraction map of a predominantly zincblende region of a GaAs-AlGaAs core-shell nanowire generated by APT analysis. Only a portion of the nanowire cross-section is visible because the field-of-view of APT is limited, so a schematic of the three-fold symmetric cross-section is shown in Figure 1(a) for clarity. For a nanowire specimen with interesting features towards their exterior, such as in the shell, a “tilted” specimen orientation may be preferable so that the features of interest are observed closer to the center of the detector. Figure 1(c) presents one-dimensional composition profiles extracted from the AlGaAs shell reconstruction along an azimuthal pathway depicted in Figure 1(b). Additional Al mole fraction maps from which Fig. 1c was derived are provided in the Supporting Information Figure S3. The compositional variations observed in Figures 1(b) and 1(c) derive from the underlying faceting of the growth surface, which we now describe. GaAs nanowires grown by MBE18 and MOCVD35, 36 commonly exhibit a orientation and a cross-section consisting of {110} type facets bridged by short {112} corner facets (Fig. 1a). While the cross-sections appear approximately hexagonal, the growth axis is actually threefold symmetric due to the zincblende crystal symmetry, as manifest in the varying polarity of the {112} facets and the slightly different widths of {112}A and {112}B facets.21 For the MBE growth conditions used here, the deposition rate on the nanowire sidewalls is low, and the Ga diffusion length sufficiently long, such that the cross-section is close to the Wulff shape17, 18, 37 5 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

dominated by thermodynamically preferred {110} facets.38 In general, the Ga diffusion length is in the order of ~1 µm whereas the Al diffusion length is known to be much smaller than the Ga diffusion length by orders of magnitude on GaAs surfaces.39, 40 When Al is introduced to grow a ternary alloy shell of low aluminum mole fraction (30% here), the nanowire shape is preserved, but the shell material grown on {112} facets is enriched in Al (Fig. 1b). The preferential segregation of Al to {112} facets slightly depletes adjacent regions on {110} facets (Fig. 1c). Similar alloy segregation has been widely reported for different III-V alloy core-shell nanowires including GaAs-AlGaAs,13, 18-21 GaAs-AlInP,32 GaAs-GaAsP41, and InGaAs-InAlAs.33 GaAs and Al0.3Ga0.7As are nearly lattice matched, so this behavior is not driven by strain. Also, there is no miscibility gap in the phase diagram of bulk AlGaAs.26 One known driver of ternary alloy segregation at nanofacets is the change in the chemical potential due to curvature (capillarity).42-44 On nonplanar growth interfaces including nanowires, the chemical potential of nanofacets is decreased or increased compared to the neighboring “wide” facets depending on whether the region is concave or convex, respectively. The increased chemical potential associated with the positive curvature of {112} nanofacets induces the impinging adatoms on the nanofacets to diffuse away toward the neighboring facets. The diffusion toward and incorporation on {110} facets result in the increase of the width of {112} facets, which in turn decreases the chemical potential of the nanofacets. In this way, the growth rate of {112} nanofacets is balanced with that of {110} facets. For a ternary such as AlGaAs, if one species (Ga) diffuses faster than the other species (Al), the slow-diffusing element may be over-represented on the energetically unfavorable nanofacets, leading to alloy segregation. At typical growth temperatures and V/III ratios of AlGaAs in MBE, the diffusion length of Ga is

6 ACS Paragon Plus Environment

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

much longer than that of Al,45 consistent with the reduced Ga mole fraction in regions of higher curvature (vicinity of {112} nanofacets). We grew AlGaAs shells on GaAs nanowires at successively higher temperatures in order to increase Al diffusivity and minimize Al segregation. According to Koshiba et al,40, the activation energy for Ga diffusion is smaller than that for Al diffusion on (001) GaAs surface, so one might expect the Al diffusion length to increase more rapidly than Ga diffusion length with temperature. However, the AlGaAs shells show very similar profiles, and the magnitudes of enhancement and depletion actually increase (Fig. 1c). At all the tested temperatures, the adjacent {112} facets are not symmetric, consistent with our prior work.21,

41

These profiles

cannot be explained by an angular variation in sticking coefficient, as one would not expect variation along {110} facets, and sticking coefficients of Ga and Al on (Al)GaAs are known to be nearly unity at temperatures up ~650 °C.46 Instead, the observation suggests that the equilibrium composition of the surface varies with position, with Al bonding preferred over Ga bonding in {112} facet regions, and a slight preference for Ga in the transition region between {110} and {112} facets. Generically, these observations imply that ternary alloy shells will develop some degree of inhomogeneity due to different incorporation preference of alloying element on different facets. Indeed, in GaAs-GaAsP core-shell nanowires containing P rich bands along directions, P content was higher at {112}A nanofacets compared to {112}B nanofacets due to the distinct bonding environments,41 pointing to the importance of the local surface structure as a major driver of segregation. We note that for wurtzite nanowires, the {112} type facets ({101-1} in hexagonal notation) are equivalent, and no asymmetry in the degree of Al segregation is expected.

7 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

The growth of pure GaAs and AlAs on GaAs nanowires highlights differences in bonding preferences via changes in the shape of the growth interface. Because there is very little Al and Ga intermixing at the growth temperature, we can assume that the shape of the AlAs surface during growth can be deduced from the shape of the AlAsGaAs interface measured after growth. Figures 2a–c show that AlAs grows selectively on {112} facets; AlAs layers are obviously thicker along directions (Fig. 2a, green arrow) than along directions (Fig. 2a, red arrow), as shown schematically in Fig. 2b and summarized quantitatively in Fig. 2c. The thinner AlAs regions adjacent to the {112} facets (Figure 2a, blue arrow) are reminiscent of the dips in Al mole fraction near the {112} facets in GaAs-AlGaAs nanowires (Fig. 1c). Considering that Al0.3Ga0.7As growth preserves the GaAs growth interface shape, we observe that the “removal” of the Ga flux leads to a concave growth front between {110} and {112} facet regions. In a kinetic Wulff analysis of a convex (concave) crystal growth interface, the fastestgrowing facets shrink (expand) and the slowest growing facets expand (shrink) until the shrinkage is balanced by capillary forces. Furthermore, the convex (concave) surface remains convex (concave). Apparently, the AlAs layers as shown in Fig. 2a are not thick enough yet to generate the steady-state interface shape. Instead, the concave growth fronts close to {112} facets represent a transition stage to accommodate growth rate differences between {110} and {112}.21 To enable the AlAs growth front to evolve the kinetic Wulff shape and more clearly reveal differences in bonding preferences, we then realized core-multishell nanowires with alternating GaAs/AlAs layers of increasing thickness in the shell (Fig. 3a). For all layer thicknesses studied, the growth interface of pure GaAs is convex and dominated by wide {110} facets bridged by {112} nanofacets. We conclude that the Ga adatom diffusion length is large enough to recover 8 ACS Paragon Plus Environment

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

and maintain the kinetically favored shape. This observation is consistent with a recent model analysis of shell growth in which it was shown that the facet lengths are not self-limiting, but the relative length of {110} vs. {112} facets is stable up to a diameter related to the diffusion length.47 The shape of the AlAs growth interfaces is strikingly distinct, exhibiting both concave and convex regions (Fig. 3b). AlAs again grows preferentially on corners consisting of GaAs {112} nanofacets, creating concave regions between the nominally {110} facets and the {112} corner regions. In the initial stages of growth revealed by the innermost AlAs layer (labeled “1” in Fig. 3a), preferential incorporation of Al atoms on/nearby the {112} nanofacets may induce the facets to become atomically rough, with the atomically rough regions extending from {112} corners to the neighboring {110} facets. When thicker AlAs layers are grown, the growth interface appears to evolve two distinct facets with orientations that deviate from {110} growth fronts (“2” and “3” in Fig. 3a). The change in the length ratio from the first to second layer is consistent with expectations that, for a concave growth front, the faster growing facet widens and the slower growing facet shrinks until the capillary force equilibrates the two growth rates.48 This may be taken as an indication that the growth fronts of the two layers are approaching the kinetic Wulff shape, established by angular variations in preferential incorporation. Indeed, the AlAs  GaAs interfaces denoted as “2” and “3” in Fig. 3a are self-similar, with the long and short facets are misoriented from {110} by ~5° and ~14°, respectively, and the length ratio of two neighboring facets is also the same (Supporting Information S4). The closest corresponding Miller indices are {1 3 4} and {19 1 0}. The three-fold symmetry due to the variations in polarity of zincblende {112} facets also emerges in layers 2 and 3, with the 14° misoriented planes flanking the {112}B nanofacets. The outermost AlAs layer, denoted by “4” in Fig. 3a, does not display the same interface shape as the 9 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

inner layers. Along with the two vicinal facets identified in layers 2 and 3, the layer 4 growth front contains a small {110} region connecting the two vicinal facets. This region likely develops once the diffusion length of Al adatoms is less than the distance between facets, preventing attainment of a fixed interface shape. In contrast, the Ga adatom diffusion length remains long enough to permit recovery of the energetically favorable {110} facets. The differences in bonding preferences of Al and Ga drive segregation in the ternary alloy shell and the distinct growth front evolution in the binary layer sequences. Both the different strengths of Ga-As and Al-As bonds,49 as well as the arrangement of surface atoms and states of dangling bonds on the {110}, {112}A, and {112}B facets and bridging regions, influence the surface diffusion lengths and the incorporation rates.39,

50

Hence, a complete model of the different

incorporation efficiencies of Al atoms among {110}, {112}A, and {112}B facets must consider surface reconstructions. The surface reconstructions are sensitive to the coverage of As atoms on the given surface, which in turn depends on temperature and III/V ratio.51 We also note that the discussion above assumes that the imaging direction is parallel to the nanowire sidewall, but it is important to consider alternative scenarios. For example, {112}A nanofacets of high surface energy have been reported to dissociate into {111}A, {113}A, {001} and {112}A facets along the sidewall of GaAs nanowires grown by MOCVD with Au catalysts.35 Similarly, the two misoriented facets of AlAs layers in the intermediate stage may have high surface energies, and may at least partially dissociate into other low index facets not parallel to the nanowire axis. The HAADF-STEM analysis, which was conducted parallel to the growth direction, does not confirm nor rule out this possibility. We also note that rotational twins within the cross-sectioned volume change the symmetry of transmission images along the zone axis (Supporting Information S5). While cross-sectional STEM studies revealed how Al bonding preferences influenced the 10 ACS Paragon Plus Environment

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

evolution of nanowire structure, three-dimensional atom probe analysis revealed additional Al segregation driven by structural features that were subsequently identified in plan view STEM studies (Figure 4). Specifically, perturbations of the GaAs core crystal structure, including stacking faults and phase changes, induce axial Al segregation in the AlGaAs shell in both zincblende and wurtzite regions. Figures. 4a and b show additional 3D representations of the nanowire grown at 490 °C in a region containing both zincblende and wurtzite material (see Supporting Information S1 and S6). The GaAs core is represented by a purple isosurface of 90% Ga mole fraction, and the Al distribution in the Al0.3Ga0.7As shell is visualized using a green isosurface of 42% Al mole fraction. The green Al isosurface reveals the presence of Al-rich nanorings (disk-shaped with toroidal topology), which were found in all the examined nanowires regardless of shell growth temperature. The two nanorings shown in Fig. 4b occur in a region where the asymmetric segregation to {112}A and {112}B facets is interrupted, suggesting that the nanorings occur within a wurtzite segment of hexagonal symmetry about the growth axis. Indeed, plan view HAADF-STEM analysis (Fig. 4c) confirms the existence of zincblende insertions of a few nanometers within larger wurtzite segments. TEM and STEM-EDX analysis (Fig. 4d, e) determined that the zincblende segments are enriched in Al relative to wurtzite segments for all shell growth temperatures tested. Furthermore, the density of Al-rich nanorings detected by APT is in agreement with the density of Al rich ZB segments observed by (S)TEM (see also Supporting Information S6). Based on the similarities between the features observed by APT and (S)TEM, we conclude that the Al rich regions identified by APT are induced by the crystal structure transfer from distinct zincblende segments of the core.

11 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

Prior reports of the influence of crystal phase on composition include the observation of In segregation in cubic phase inclusions52 and in regions where m- and r- planes meet53 in InGaN quantum wells in InGaN/GaN multi quantum well nanowires. Here, the zincblende insertions arise from deviations from wurtzite stacking. Stacking faults introduce deviations from a strictly planar {110} side facets, changing the surface termination and creating new types of bonding sites.54, 55 We propose that the new bonding sites selectively incorporate Al, while noting that differences in strain between wurtzite and zincblende regions could also play a role in other III-V systems. Based on the comparison in Supporting Information S6, stacking faults also create binding sites at which Al is preferred over Ga. Preferential incorporation of Al atoms at step edges has been previously observed in epitaxy of AlGaAs on {110} and {100} vicinal surfaces.56 A kinetic Monte Carlo simulation suggests that strong and anisotropic Al-Al interactions govern Al segregation along the step edges.57, 58 The studies presented here do not provide the necessary resolution to confirm the association of Al with new binding sites. However, the correlated STEM analysis clearly shows that the composition, and therefore bandgap, of a ternary shell is modulated by the crystal phase of the core, suggesting new opportunities for engineering energy barriers and wells with a disk-like shape and toroidal topology. In summary, Al segregation in GaAs-AlGaAs core-shell nanowires is driven by the preferential incorporation of Al atoms in sites on and adjacent to {112} nanofacets. The diffusion lengths and bonding preferences of alloying elements on different facets control the resulting nanostructure. Some degree of composition modulation is likely unavoidable when the alloy is grown on corrugated surfaces composed of certain facet combinations. As such, digital alloys in the form of short-period superlattices59 are an attractive approach to engineering uniform potentials. We also note that the AlGaAs shell growth interface will be perturbed at an Al group III mole 12 ACS Paragon Plus Environment

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

fraction between the 30% and 100% values studied here. Directed evolution of the growth shape interface and/or selective segregation to nanofacets could be exploited to define quantum wires on the surfaces of nanowires. Furthermore, crystal phase engineering of the nanowire core can provide a template for the growth of quantum barriers and/or wells both in the radial direction via deliberate composition modulation and periodically along the nanowire growth axis through spontaneous segregation.

Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI:. Al mole fraction maps of GaAs-AlGaAs core-shell nanowires, X-TEM images of GaAs-AlAs core-multishell nanowires (PDF) AUTHOR INFORMATION Corresponding Authors LJL: [email protected]; GK: [email protected] Present Addresses †Materials Science Division, Argonne National Laboratory, Argonne, Illinois, 60439, USA Author Contributions Atom probe tomography was carried out at Northwestern University. Nanowire growth and electron microscopy was carried out at TU Munich. The manuscript was written with the

13 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

contributions of all authors. All authors have given approval to the final version of the manuscript. ACKNOWLEDGMENT NJ and LL gratefully acknowledge support of the National Science Foundation via DMR1611341 and DMR-1308654. Atom-probe tomography was performed at the Northwestern University Center for Atom-Probe Tomography (NUCAPT). The LEAP tomograph at NUCAPT was purchased and upgraded with grants from the NSF-MRI (DMR-0420532) and ONR-DURIP (N00014-0400798, N00014-0610539, N00014-0910781, N00014-1712870) programs. NUCAPT received support from the MRSEC program (NSF DMR-1720139) at the Materials Research Center, the SHyNE Resource (NSF ECCS-1542205), and the Initiative for Sustainability and Energy (ISEN) at Northwestern University. GK acknowledges support of the excellence program Nanosystems Initiative Munich (NIM) funded by the German Research Foundation (DFG), the DFG-Grant KO-4005/6-1, and the International Graduate School for Science and Engineering (TUM-IGSSE). LJL and NJ acknowledge Peter Voorhees for useful discussions.

14 ACS Paragon Plus Environment

Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

REFERENCES 1.

Li, K. H.; Liu, X.; Wang, Q.; Zhao, S.; Mi, Z. Nat. Nanotechnol. 2015, 10, 140-144.

2.

Eaton, S. W.; Fu, A.; Wong, A. B.; Ning, C.-Z.; Yang, P. Nat. Rev. Mater. 2016, 1, 16028.

3.

Koblmüller, G.; Mayer, B.; Stettner, T.; Abstreiter, G.; Finley, J. J. Semicond. Sci. Tech. 2017,

32, 053001. 4.

Yao, M.; Cong, S.; Arab, S.; Huang, N.; Povinelli, M. L.; Cronin, S. B.; Dapkus, P. D.; Zhou, C.

Nano Lett. 2015, 15, 7217-7224. 5.

Yao, M.; Huang, N.; Cong, S.; Chi, C.-Y.; Seyedi, M. A.; Lin, Y.-T.; Cao, Y.; Povinelli, M. L.;

Dapkus, P. D.; Zhou, C. Nano Lett. 2014, 14, 3293-3303. 6.

Otnes, G.; Borgström, M. T. Nano Today 2017, 12, 31-45.

7.

Zhao, S.; Connie, A. T.; Dastjerdi, M. H. T.; Kong, X. H.; Wang, Q.; Djavid, M.; Sadaf, S.; Liu,

X. D.; Shih, I.; Guo, H.; Mi, Z. Sic. Rep. 2015, 5, 8332. 8.

Koester, R.; Sager, D.; Quitsch, W.-A.; Pfingsten, O.; Poloczek, A.; Blumenthal, S.; Keller, G.;

Prost, W.; Bacher, G.; Tegude, F.-J. Nano Lett. 2015, 15, 2318-2323. 9.

Sadaf, S. M.; Zhao, S.; Wu, Y.; Ra, Y. H.; Liu, X.; Vanka, S.; Mi, Z. Nano Lett. 2017, 17, 1212-

1218. 10.

Erhard, N.; Zenger, S.; Morkötter, S.; Rudolph, D.; Weiss, M.; Krenner, H. J.; Karl, H.;

Abstreiter, G.; Finley, J. J.; Koblmüller, G.; Holleitner, A. W. Nano Lett. 2015, 15, 6869-6874. 11.

LaPierre, R. R.; Robson, M.; Azizur-Rahman, K. M.; Kuyanov, P. J. Phys. D: Appl. Phys. 2017,

50, 123001.

15 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

12.

Page 16 of 25

Dai, X.; Tchernycheva, M.; Soci, C., Chapter Three - Compound Semiconductor Nanowire

Photodetectors. In Semiconduct. Semimet.; Dayeh, S. A.; Fontcuberta i Morral, A.; Jagadish, C., Eds; Elsevier: 2016; Vol. 94, pp 75-107. 13.

Morkötter, S.; Jeon, N.; Rudolph, D.; Loitsch, B.; Spirkoska, D.; Hoffmann, E.; Döblinger, M.;

Matich, S.; Finley, J. J.; Lauhon, L. J.; Abstreiter, G.; Koblmüller, G. Nano Lett. 2015, 15, 3295-3302. 14.

Tomioka, K.; Yoshimura, M.; Fukui, T. Nature 2012, 488, 189-192.

15.

Stettner, T.; Zimmermann, P.; Loitsch, B.; Döblinger, M.; Regler, A.; Mayer, B.; Winnerl, J.;

Matich, S.; Riedl, H.; Kaniber, M.; Abstreiter, G.; Koblmüller, G.; Finley, J. J. Appl. Phys. Lett. 2016, 108, 011108. 16.

Badada, B. H.; Shi, T.; Jackson, H. E.; Smith, L. M.; Zheng, C.; Etheridge, J.; Gao, Q.; Tan, H.

H.; Jagadish, C. Nano Lett. 2015, 15, 7847-7852. 17.

Heiss, M.; Fontana, Y.; Gustafsson, A.; Wust, G.; Magen, C.; O'Regan, D. D.; Luo, J. W.;

Ketterer, B.; Conesa-Boj, S.; Kuhlmann, A. V.; Houel, J.; Russo-Averchi, E.; Morante, J. R.; Cantoni, M.; Marzari, N.; Arbiol, J.; Zunger, A.; Warburton, R. J.; Fontcuberta i Morral, A. Nat. Mater. 2013, 12, 439444. 18.

Rudolph, D.; Funk, S.; Döblinger, M.; Morkötter, S.; Hertenberger, S.; Schweickert, L.; Becker,

J.; Matich, S.; Bichler, M.; Spirkoska, D.; Zardo, I.; Finley, J. J.; Abstreiter, G.; Koblmüller, G. Nano Lett. 2013, 13, 1522-1527. 19.

Jeon, N.; Loitsch, B.; Morkoetter, S.; Abstreiter, G.; Finley, J.; Krenner, H. J.; Koblmueller, G.;

Lauhon, L. J. ACS Nano 2015, 9, 8335-8343. 20.

Mancini, L.; Fontana, Y.; Conesa-Boj, S.; Blum, I.; Vurpillot, F.; Francaviglia, L.; Russo-

Averchi, E.; Heiss, M.; Arbiol, J.; Morral, A. F. i.; Rigutti, L. Appl. Phys. Lett. 2014, 105, 243106. 16 ACS Paragon Plus Environment

Page 17 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

21.

Zheng, C.; Wong-Leung, J.; Gao, Q.; Tan, H. H.; Jagadish, C.; Etheridge, J. Nano Lett. 2013, 13,

3742-3748. 22.

Tsukamoto, S.; Nagamune, Y.; Nishioka, M.; Arakawa, Y. J. Appl. Phys. 1992, 71, 533-535.

23.

Dwir, B.; Leifer, K.; Kapon, E. Phys. Rev. B 2003, 67, 075302.

24.

Biasiol, G.; Kapon, E. Phys. Rev. Lett. 1998, 81, 2962-2965.

25.

Biasiol, G.; Leifer, K.; Kapon, E. Phys. Rev. B 2000, 61, 7223-7226.

26.

Petroff, P. M.; Miller, R. C.; Gossard, A. C.; Wiegmann, W. Appl. Phys. Lett. 1984, 44, 217.

27.

Etienne, B.; Lelarge, F.; Wang, Z. Z.; Laruelle, F. Appl. Surf. Sci. 1997, 113-114, 66-72.

28.

Heyn, C.; Stemmann, A.; Schramm, A.; Hansen, W. J. Cryst . Growth 2009, 311, 1825-1827.

29.

Yan, X.; Zhang, X.; Ren, X.; Lv, X.; Li, J.; Wang, Q.; Cai, S.; Huang, Y. Nano Lett. 2012, 12,

1851-1856. 30.

Mata, M. d. l.; Zhou, X.; Furtmayr, F.; Teubert, J.; Gradecak, S.; Eickhoff, M.; Fontcuberta i

Morral, A.; Arbiol, J. J. Mater. Chem. C 2013, 1, 4300-4312. 31.

Loitsch, B.; Jeon, N.; Döblinger, M.; Winnerl, J.; Parzinger, E.; Matich, S.; Wurstbauer, U.;

Riedl, H.; Abstreiter, G.; Finley, J. J.; Lauhon, L. J.; Koblmüller, G. Appl. Phys. Lett. 2016, 109, 093105. 32.

Sköld, N.; Wagner, J. B.; Karlsson, G.; Hernán, T.; Seifert, W.; Pistol, M.-E.; Samuelson, L.

Nano Lett. 2006, 6, 2743-2747. 33.

Treu, J.; Stettner, T.; Watzinger, M.; Morkötter, S.; Döblinger, M.; Matich, S.; Saller, K.; Bichler,

M.; Abstreiter, G.; Finley, J. J.; Stangl, J.; Koblmüller, G. Nano Lett. 2015, 15, 3533-3540.

17 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

34.

Page 18 of 25

Rudolph, D.; Hertenberger, S.; Bolte, S.; Paosangthong, W.; Spirkoska, D. e.; Döblinger, M.;

Bichler, M.; Finley, J. J.; Abstreiter, G.; Koblmüller, G. Nano Lett. 2011, 11, 3848-3854. 35.

Jiang, N.; Wong-Leung, J.; Joyce, H. J.; Gao, Q.; Tan, H. H.; Jagadish, C. Nano Lett. 2014, 14,

5865-5872. 36.

Fickenscher, M.; Shi, T.; Jackson, H. E.; Smith, L. M.; Yarrison-Rice, J. M.; Zheng, C.; Miller,

P.; Etheridge, J.; Wong, B. M.; Gao, Q.; Deshpande, S.; Tan, H. H.; Jagadish, C. Nano Lett. 2013, 13, 1016-1022. 37.

Peter, K.; Henrik, I. J.; Erik, J.; Morten Hannibal, M.; Claus, B. S.; Anna Fontcuberta i, M.;

Martin, A.; Jesper, N.; Frank, G. J. Phys. D: Appl. Phys. 2013, 46, 313001. 38.

Messmer, C.; Bilello, J. C. J. Appl. Phys. 1981, 52, 4623-4629.

39.

López, M.; Nomura, Y. J. Cryst . Growth 1995, 150, Part 1, 68-72.

40.

Koshiba, S.; Nakamura, Y.; Tsuchiya, M.; Noge, H.; Kano, H.; Nagamune, Y.; Noda, T.; Sakaki,

H. J. Appl. Phys. 1994, 76, 4138-4144. 41.

Zhang, Y.; Sanchez, A. M.; Wu, J.; Aagesen, M.; Holm, J. V.; Beanland, R.; Ward, T.; Liu, H.

Nano Lett. 2015, 15, 3128-3133. 42.

Biasiol, G.; Gustafsson, A.; Leifer, K.; Kapon, E. Phys. Rev. B 2002, 65, 205306.

43.

Sato, T.; Tamai, I.; Hasegawa, H. J. Vac. Sci. Technol. B 2005, 23, 1706-1713.

44.

Steinke, L.; Cantwell, P.; Zakharov, D.; Stach, E.; Zaluzec, N. J.; Fontcuberta i Morral, A.;

Bichler, M.; Abstreiter, G.; Grayson, M. Appl. Phys. Lett. 2008, 93, 193117.

18 ACS Paragon Plus Environment

Page 19 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

45.

Nishinaga, T.; Nishioka, K.; Harada, J.; Sasaki, A.; Takei, H., Advances in the Understanding of

Crystal Growth Mechanisms. Elsevier Science: 2012. 46.

Fischer, R.; Klem, J.; Drummond, T. J.; Thorne, R. E.; Kopp, W.; Morkoç, H.; Cho, A. Y. J.

Appl. Phys. 1983, 54, 2508-2510. 47.

Zhang, Q.; Aqua, J.-N.; Voorhees, P. W.; Davis, S. H. J. Mech. Phys. Solids 2016, 91, 73-93.

48.

Sun, Q.; Yerino, C. D.; Leung, B.; Han, J.; Coltrin, M. E.; Zhang, Z.; Lagally, M. G. J. Appl.

Phys. 2011, 110, 053517. 49.

Landgren, G.; Ludeke, R. Solid State Commun. 1981, 37, 127-131.

50.

Hjort, M.; Kratzer, P.; Lehmann, S.; Patel, S. J.; Dick, K. A.; Palmstrøm, C. J.; Timm, R.;

Mikkelsen, A. Nano Lett. 2017, 17, 3634-3640. 51.

LaBella, V. P.; Krause, M. R.; Ding, Z.; Thibado, P. M. Surf. Sci. Rep. 2005, 60, 1-53.

52.

Poliani, E.; Wagner, M. R.; Reparaz, J. S.; Mandl, M.; Strassburg, M.; Kong, X.; Trampert, A.;

Sotomayor Torres, C. M.; Hoffmann, A.; Maultzsch, J. Nano Lett. 2013, 13, 3205-3212. 53.

Ra, Y.-H.; Navamathavan, R.; Kang, S.; Lee, C.-R. J. Mater. Chem. C 2014, 2, 2692-2701.

54.

Caroff, P.; Dick, K. A.; Johansson, J.; Messing, M. E.; Deppert, K.; Samuelson, L. Nat.

Nanotechnol. 2008, 4, 50. 55.

Pankoke, V.; Kratzer, P.; Sakong, S. Phys. Rev. B 2011, 84, 075455.

56.

Tsuchiya, M.; Petroff, P. M.; Coldren, L. A. Appl. Phys. Lett. 1989, 54, 1690-1692.

57.

Lu, Y. T.; Petroff, P.; Metiu, H. Appl. Phys. Lett. 1990, 57, 2683-2685.

19 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

58.

Metiu, H.; Lu, Y.-T.; Zhang, Z. Science 1992, 255, 1088-1092.

59.

Irber, D. M.; Seidl, J.; Carrad, D. J.; Becker, J.; Jeon, N.; Loitsch, B.; Winnerl, J.; Matich, S.;

Page 20 of 25

Döblinger, M.; Tang, Y.; Morkötter, S.; Abstreiter, G.; Finley, J. J.; Grayson, M.; Lauhon, L. J.; Koblmüller, G. Nano Lett. 2017, 17, 4886-4893.

20 ACS Paragon Plus Environment

Page 21 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 1. (a) End view schematic of Al segregation (green regions) in AlGaAs shell (light blue region) on zincblende GaAs nanowire core. (b) Atom probe tomography generated Al mole fraction map of Al0.3Ga0.7As shell grown at 560 °C. Scale bar is 10 nm. (c) Al composition profile along the azimuthal direction at 560 °C, 490 °C, 420 °C, and 370 °C from top to bottom. Black dashed lines in (b) show the region profiled.

21 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 25

Figure 2. (a) Cross-sectional HAADF-STEM image of GaAs-AlAs multilayers on zincblende GaAs nanowire. (b) AlAs layer thicknesses along paths indicated by the red, blue, and green arrows shown in (b). AlAs layers are thickest on {112} corners and thinnest adjacent to these corners. (c) End view schematic of {112} corners of GaAs nanowire core (purple) with AlAs shell (green) capped by GaAs shell (purple). The Al segregation observed in AlGaAs shells manifests as thicker regions on {112} corners for AlAs shells. Pure GaAs growth restores the original interface shape.

22 ACS Paragon Plus Environment

Page 23 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 3. (a) HAADF-STEM image of GaAs-AlAs multilayers of increasing thickness. (b) Schematic showing evolution of the AlAs growth interface.

23 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 25

Figure 4. (a) APT reconstruction (62 x 62 x 20 nm3) of GaAs-AlGaAs core-shell nanowire with shell grown at 490 °C. Purple and green isosurfaces were generated at 90% Ga and 43% Al mole fractions, respectively. The purple surface renders the GaAs core, and the green surfaces reveal Al-rich nanorings. (b) Longer APT reconstruction (62 x 62 x 80 nm3) of the same nanowire shown in (a). The ZB (WZ) segments are assigned based on the presence (absence) of strong Al enrichment on {112} corners. (c) STEM image of nanowire with the shell deposited at 620 °C showing inclusion of a short ZB section. (d) HAADF-STEM image of the AlGaAs shell region (grown at 370 °C) containing a short zincblende (ZB) insertion in wurtzite (WZ) segment indicated by blue (ZB) and red (ZB) lines, respectively. (e) EDX line scan along the black dotted arrow in (d) showing increased Al concentration in the ZB region. We note that the reported concentrations are not quantitative for the AlGaAs shell due to the varying projection through the 5 nm GaAs capping layer.

24 ACS Paragon Plus Environment

Page 25 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

TOC Figure

25 ACS Paragon Plus Environment