of 33 ACS Paragon Plus Environment Nano Letters 1 2 3 4 5 6 7 8 9

36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57 ... with atomically sharp interfaces. Valentina Zannier. 1. , F...
0 downloads 0 Views 3MB Size
Subscriber access provided by University of Florida | Smathers Libraries

Communication

Nanoparticle stability in axial InAs-InP nanowire heterostructures with atomically sharp interfaces Valentina Zannier, Francesca Rossi, Vladimir G. Dubrovskii, Daniele Ercolani, Sergio Battiato, and Lucia Sorba Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.7b03742 • Publication Date (Web): 29 Nov 2017 Downloaded from http://pubs.acs.org on December 1, 2017

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 free 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 accessible to all readers and 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.

Nano Letters 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 33 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

289x229mm (150 x 150 DPI)

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 33

Nanoparticle stability in axial InAs-InP nanowire heterostructures with atomically sharp interfaces Valentina Zannier1, Francesca Rossi2, Vladimir G. Dubrovskii3, 4, Daniele Ercolani1, Sergio Battiato1, and Lucia Sorba1

1

NEST, Istituto Nanoscienze – CNR and Scuola Normale Superiore, Piazza San Silvestro 12,

56127 Pisa, Italy 2

IMEM – CNR, Parco Area delle Scienze 37/A, 43124 Parma, Italy

3

ITMO University, Kronverkskiy Pr. 49, 197101 St. Petersburg, Russia

4

Ioffe Institute RAS, Politekhnicheskaya 26, 194021 St. Petersburg, Russia

ABSTRACT The possibility to expand the range of material combinations in defect-free heterostructures is one of the main motivations for the great interest in semiconductor nanowires. However, most axial nanowire heterostructures suffer from interface compositional gradients and kink formation, as a consequence of nanoparticle-nanowire interactions during the metal-assisted growth. Understanding such interactions and how they affect the growth mode is fundamental to achieve a full control over the morphology and the properties of nanowire heterostructures for device applications. Here we demonstrate that the sole parameter affecting the growth mode (straight or kinked) of InP segments on InAs nanowire stems by the Au-assisted method is the nanoparticle composition. Indeed, straight InAs-InP nanowire heterostructures are obtained only when the In/Au ratio in the nanoparticles is low, typically smaller than 1.5. For higher In content, 1   

ACS Paragon Plus Environment

Page 3 of 33 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

the InP segments tend to kink. Tailoring the In/Au ratio by the precursor fluxes at a fixed growth temperature enables us to obtain straight and radius-uniform InAs-InP nanowire heterostructures (single and double) with atomically sharp interfaces. We present a model that is capable of describing all the experimentally observed phenomena: straight growth versus kinking, the stationary nanoparticle compositions in pure InAs and InAs-InP nanowires, the crystal phase trends, and the interfacial abruptness. Taking into account different nanowire/nanoparticle interfacial configurations (forming wetting or non-wetting monolayers in vertical or tapered geometry), our generalized model provides the conditions of nanoparticle stability and abrupt heterointerfaces for a rich variety of growth scenarios. Therefore, our results provide a powerful tool for obtaining high quality InAs-InP nanowire heterostructures with well-controlled properties and can be extended to other material combinations based on the group V interchange.

KEYWORDS: III-V nanowires, axial heterostructures, VLS growth, nanoparticle stability, modeling.

Semiconductor nanowires (NWs) enable dislocation-free integration of lattice-mismatched materials due to a more efficient relaxation of elastic strain than in the corresponding epitaxial films.1-3 Consequently, NW heterostructures have become relevant for a wide range of device applications. The successful growth of axial NW heterostructures has been reported since 19964 and several types of devices based on such structures have already been demonstrated.5-9 The Au-assisted vapor-liquid-solid (VLS) growth10 is the common approach to obtain high quality III-V NWs. However, when NW heterostructures are concerned, the VLS mechanism is not 2   

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 33

straightforward and far from being fully understood and controlled. The two main issues in Auassisted axial NW heterostructures are their interfacial abruptness and kinking. The difficulty to form sharp interfaces is attributed to the solubility of the growth species in the Au seed nanoparticle (NP) which remain when one vapor flux is switched to the other. This so-called reservoir effect has been investigated in detail experimentally and theoretically,11-13 and many efforts have been made to obtain sharp interfaces essential for device applications.13-15 The problem of kinking in Au-assisted III-V NW heterostructures, that is, a spontaneous change of the growth direction at the liquid-solid interface, has also been reported for many material systems. Usually, kinking is attributed to different interfacial energies of the Au NP with the two materials and to the tendency of the NP to preserve the more stable interface when changing the growing material.16 In III-V NW heterostructures based on the group III interchange, the NWs typically grow straight only in one of the two growth sequences, whereas the other sequence often produces kinked geometries.17,18 Furthermore, the interfaces are usually broad because of the high solubility of the group III metals in Au. Axial NW heterostructures based on the group V interchange are usually easier to obtain and sharper interfaces are expected thanks to the lower solubility of the group V species in liquid Au. For instance, atomically sharp interfaces have been reported for InAs-InP19 and GaAs-GaP20 NW superlattice structures. However, kinking and changes in growth direction have also been observed in these systems.21 In particular, for the InAs-InP system, a good yield of straight NWs has been obtained only within a narrow range of NW diameter and density, and only for very thin (≤ 10 nm) InP insertions.5, 21, 22 Therefore, a full control over the NW growth mode, which should allow for the realization of InAs-InP NW heterostructures with the desired morphology, is still challenging. Understanding the mechanisms behind the kink formation is fundamental to achieve this goal.

3   

ACS Paragon Plus Environment

Page 5 of 33 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

Here, we present a detailed study of InAs-InP NW heterostructure morphology as a function of precursor fluxes, crystal structure of the NWs and chemical composition of the catalyst NP. Our results clearly demonstrate that the sole parameter which affects the kink formation is the NP composition (In/Au ratio). We interpret this mechanism in terms of the NP stability on top of the NW as it undergoes changes in chemical composition and shape. Understanding the NP stability under varying fluxes requires a significant modification of the VLS growth theory in terms of surface energetics influencing possible configurations of the growth interface (changing the NP contact angle, or the NW radius, or the NP sliding down). We find that increasing the In/Au ratio first leads to a corresponding increase of the NP contact angle, then the NW top radius, and further to the development of a truncated wetted edge at the growth interface. When the contact angle exceeds a certain critical value, vertical VLS growth becomes unstable and the InP segment kinks. These new findings provide the understanding of the mechanisms behind the kink formation in InAs-InP NW heterostructures and the way to suppress kinking. More generally, our results can be applied to many other material systems based on group V interchange, giving a new method to control the growth of NW heterostructures with the desired morphology and sharp heterointerfaces. The NWs of the present study were grown by chemical beam epitaxy (CBE) on InAs(111)B substrates by the VLS mechanism.23 We used Au NPs as seeds for the NW growth and trimethylindium (TMIn), tert-butylarsine (TBAs) and tert-butylphosphine (TBP) as metalorganic precursors. We first grew InAs NW stems using constant TMIn and TBAs fluxes at the growth temperature of 430 ±10oC, which is close to the optimal growth temperature to obtain InAs-InP nanowire heterostructures of good quality19. Then, in order to obtain InAs-InP axial NW heterostructures, the InP growth was carried out on the InAs stems at the same temperature, 4   

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 33

using different combinations of TMIn and TBP fluxes. All the samples were cooled down in ultra-high-vacuum (UHV) environment and characterized by means of scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Details of growth protocols, SEM and TEM analysis are presented in the Methods section. The effect of precursor fluxes on the morphology of the NW heterostructures is summarized in Figure 1, which shows the InAs-InP NW heterostructures obtained using Au colloids of 40 nm in diameter, as a function of the TMIn and TBP fluxes employed for the InP segment growth. We find that the InP growth mode and the resulting heterostructure morphology strongly depend on the precursor fluxes. When we use high TMIn flux (0.6 Torr) and low TBP flux (1.5 Torr), referred to as the HL fluxes hereinafter, the InP segment grows kinked, yielding NWs with a bad morphology [see Fig. 1(a, b)]. On the other hand, when we use high TMIn flux (0.6 Torr) and high TBP flux (2.2 Torr), (HH fluxes hereinafter) straight NW heterostructures are obtained [Fig. 1 (c, d)]. Straight InP segments are also obtained when we use a low TMIn flux (0.1 Torr), for any TBP flux used, that is a combination of the LL and LH fluxes [Figs. 1 (e, f) and (g, h)]. The interface between the two materials is well visible (thanks to an increase of diameter) in all samples, except for the one grown with the LH fluxes [Fig. 1 (g, h)]. We believe that the larger diameter of some InP segments arises from the vapor-solid (VS) lateral InP growth which is different from sample to sample due to the different crystal structure and fluxes employed, rather than from a change of the nanoparticle diameter. As will be demonstrated in the following, straight NW heterostructures obtained using different precursor fluxes have the NPs with similar diameter on their tips, but only some of them show a pronounced increase of diameter at the InAs-InP interface and/or for the whole InP segment length. The InP lateral (shell) growth,

5   

ACS Paragon Plus Environment

Page 7 of 33 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

which occurs concomitantly with the axial growth, depends on the NW crystal structure, as will be shown later based on the TEM analysis.

Figure 1. Morphology and NP composition of the InAs-InP axial NW heterostructures obtained with Au colloids of 40 nm in diameter as a function of the precursor fluxes. (a, c, e, g): 45o-tilted SEM images of the as-grown samples and (b, d, f, h): corresponding STEM images of the tip of one representative NW for each sample. The chemical composition of the NPs (the In/Au ratio), measured by EDX, is given in each panel. TMIn and TBP fluxes indicated in the figure are those employed for the top InP segment: 0.6 and 1.5 Torr (HL) in panels (a) and (b), 0.6 and 2.2 Torr (HH) in panels (c) and (d), 0.1 and 1.5 Torr (LL) in panels (e) and (f), 0.1 and 2.2 Torr (LH) in panels (g) and (h). The InAs stems were grown using the same TMIn and TBAs fluxes for all the samples.

6   

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 33

In order to determine the factors influencing these different NW morphologies, we analyzed the NW heterostructures by means of TEM. Panels (b, d, f, h) of Figure 1 show the scanning TEM (STEM) images of one representative NW from each sample presented in panels (a, c, e, g). We measured also the chemical composition of the NPs on their tips by energy dispersive xray spectroscopy (EDX). For all the NWs, the detected amount of P and As in the NPs never exceeds 1 at% and 2 at% respectively, so that the NP composition is well characterized by the In/Au ratio. In order to verify that the composition of the NPs measured ex-situ by EDX at room temperature (RT) is representative of the actual NP composition during growth, we analyzed the InAs-InP NW heterostructures cooled in UHV (without any flux) and the NWs grown under the same conditions but cooled with As flux. In this way, we were able to quantify the amount of In left in the NPs and the amount of In possibly consumed from the NPs during cooling (by forming an InAs neck under the NP). By using this procedure we were able to deduce the real NP composition during growth.24 The details of these experiments are reported in Supporting Information S1. The obtained results indicate that the cooling process in the UHV environment does not alter the NP chemical composition. As a consequence, the chemical analysis performed ex situ at RT by EDX gives the actual NP composition during growth. The kinked NWs obtained with the HL fluxes have the NPs with 64 at% of In, corresponding to an In/Au ratio of 1.8 [Fig. 1(a, b)]. On the other hand, straight NW heterostructures grown using HH, LL or LH fluxes [Figs. 1 (c, d), (e, f) and (g, h), respectively] have less In in the NPs, between 54 and 58 at%, corresponding to the In/Au ratios of 1.2 to 1.4. The NPs on the tips of InAs NW stems grown prior to the InP segments (not shown) have 35 at% of In, corresponding to a much lower In/Au ratio of 0.5. We also grew the InAs NW stems using a lower TMIn flux 7   

ACS Paragon Plus Environment

Page 9 of 33 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

(0.1 Torr instead of 0.6 Torr), and found the same In/Au ratio of 0.5 in the NPs in this case. Therefore, the chemical composition of the Au-In NPs depends on both precursor fluxes for InP segment growth, whereas it does not depend on the TMIn flux for InAs NWs in this parameter window. Furthermore, we found a systematic increase of the In/Au ratio as we switch from As to P even if the TMIn flux remains the same, as already reported by other researchers.22, 25 The observed increase of the In/Au ratio in the NPs when changing InAs to InP can be understood using the theory of ternary III-V NWs and NW-based heterostructures. Two approaches have previously been used to describe the compositions of ternary III-V NWs based either on the binary nucleation theory with macroscopic nuclei26-28 or the irreversible growth concept in which meeting of any two atoms of group III and V at the liquid-solid growth interface immediately produces stable III-V pair in the solid state.29-31 The first approach is reduced to the second one at high supersaturations where the critical size of classical nucleation theory tends to only one III-V pair.32,33 For ternary systems without miscibility gaps, such as InAsP,27 both approaches yield similar results for the compositional diagrams (that is, the liquid versus solid composition at a given temperature).32 Therefore, we have developed a generalized model for the InAsP composition based on a combination of the two approaches. The model and the details of calculations are described in detail in Supporting Information S2. We find that, while both As and P concentrations in liquid are very low and have negligible effect on the NP size, the In concentration may be changed dramatically upon the group V flux commutation. In particular, the NPs become more In-rich during the growth of InP due to (i) a twice lower crystallization rate of InP, which leaves much more In in the NP relative to InAs and (ii) a higher sensitivity of the NP composition to the In flux due to a lower group V to III influx imbalance for InP. The latter also explains how the NP composition can be changed by tuning the precursor 8   

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 33

fluxes when growing the InP segment. This is extremely important because the In/Au ratio during InP growth strongly affects the resulting heterostructure morphology (straight or kinked NWs). The possibility to change the NW growth direction by changing the NP composition has already been reported for pure InP NWs.24 We have shown here that the NP composition plays a fundamental role in driving the growth mode of the InP segment also on top of InAs stems: straight InAs-InP NW heterostructures obtained using different combinations of precursor fluxes have a similar NP composition (In/Au ratio = 1.2-1.4) despite the very different morphology (with or without lateral growth, with a more or less uniform shell thickness along the InP segment). Conversely, kinked NW heterostructures obtained under the HL fluxes have a significantly different NP composition, with a higher amount of In (the In/Au ratio = 1.8). For several material systems it has been shown that kinked or straight morphologies in axial NW heterostructures depend on the respective crystal structures at the two sides of the heterointerface.21,

34

Therefore, we consider now the crystal structures of straight NW

heterostructures obtained with different material flux combinations. Panels (a) and (b) of Figure 2 show the high resolution TEM (HRTEM) images with the corresponding selected area electron diffraction (SAED) patterns at the InAs/InP interface of the two NWs grown using HH fluxes [Fig. 1 (c, d)] and LH fluxes [Fig. 1 (g, h)]. While the InAs stem exhibits an almost defect-free wurtzite (WZ) crystal structure in both NWs, the InP segments show a very different crystal structure and quality. The InP segment grown with the HH fluxes [Fig. 2 (a)] has a zinc blende (ZB) crystal structure with many twins, whereas the one grown with the LH fluxes [Fig. 2 (b)] has a WZ structure with few stacking faults. Straight InP segments obtained with the LL fluxes (an intermediate value of the III/V flux ratios) have a mixed WZ/ZB structure (not shown). In 9   

ACS Paragon Plus Environment

Page 11 of 33 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

most of Au-assisted III-V NWs, the III/V flux ratio and the growth rate (which is often proportional to the group III flux) strongly affect the NW crystal structure.35,36 Also in our case, the crystal structure of the InP segment is affected by the precursor fluxes employed. Low group III and high group V fluxes favor the growth of WZ InP segments with a very good crystal quality. As the III/V ratio is increased, we obtain a mixed WZ/ZB crystal and, for the highest III/V ratio employed, a pure ZB crystal.

Figure 2. HRTEM images of the InAs/InP interface region of the two NWs grown using HH (a) and LH (b) fluxes, respectively. The insets are the correspondent false-color SAED patterns taken along [2, -1, 1,0] zone axis, with red and blue spots representing InAs and InP, respectively.

Another difference between these straight InAs-InP NW heterostructures lies in the lateral InP (shell) growth. We observe a significant shell growth only for ZB and mixed WZ/ZB InP segments [Figs. 1 (a-f) and Fig. 2(a)], while the shell growth is negligible around the WZ InP segment [Fig. 1(g, h) and Fig. 2(b)]. In III-V radial (core/shell) NW heterostructures it has been reported that the shell growth proceeds at different growth rates on the sidewalls of NW cores 10   

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 33

having different crystal structures37-39 and generally ZB NWs have an enhanced radial growth compared to WZ ones.40 This consideration may also be applied to explain also our results. It is clear at this point that InP segments grow straight on top of the InAs stems despite the differences in precursor fluxes, growth rate, morphology and crystal structure. The only common parameter for these straight InAs-InP NW heterostructures is the NP chemical composition, namely, a low In/Au ratio of 1.2 to 1.4. On the other hand, kinked InP segments obtained under the HL fluxes have a similar crystal structure (not shown) to the one of straight InP segments obtained under the HH fluxes (mainly ZB), but a very different chemical composition of the catalyst NP. This allows us to conclude that the precursor fluxes affect both the chemical composition of the catalyst NP and the crystal structure of the InP segment; however, the InP growth mode (straight or kinked) and the resulting NW morphology depend only on the NP composition. We observe the same trend also for NWs with different diameters, grown either from Au colloids of 20 nm in diameter or by Au film dewetting, which results in a broad distribution over the NW diameter.23 We always obtain kinked NW heterostructures when growing InP segments using the HL fluxes, and systematically detect a higher amount of In in their NPs compared to straight NWs obtained using the LH fluxes [see Supporting Information S3]. Therefore, also the NP diameter does not have any significant effect on the InP growth mode. We have thus demonstrated that the sole parameter controlling stable vertical growth or kinking of InP segments on InAs stems is the In/Au ratio in the NP, regardless of what combination of precursor fluxes has led to that NP composition [see Fig. 1]. In VLS growth, the NP stability is the key parameter to control the NW growth mode.41-46 In the specific case of

11   

ACS Paragon Plus Environment

Page 13 of 33 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

axial NW heterostructures, the flux commutation is often accompanied by changes in the NP composition. This in turn leads to variations in the NP volume, contact angle and surface energies of different interfaces which may alter the stability of vertical VLS growth, triggering phenomena like the NP unpinning, downward growth or kinking as already reported for many systems.16,21,47-50 Understanding the phenomena behind the stability of vertical VLS growth versus kinking is fundamental for controlling the growth mode and achieving the NW heterostructures with the desired morphology.21, 24, 50 Similarly to Ref. [50], we speculate that the growth mode of InP on InAs stems is related to the contact angle of the NP which is known to be crucial for maintaining the NP on the NW top.46, 50-52 The data shown in Fig. 1 demonstrate that straight growth of InP on InAs stems systematically requires low In/Au ratios. The trend was opposite for InAs-GaAs NW heterostructures50, but in that case the NP base radius was constant and the contact angle proportional to the group III/Au ratio. For InAs-InP NW heterostructures the scenario is more complex, as it will be shown in the following. In order to quantify the dependence of the InP growth mode from the NP composition in terms of the contact angle, we analyzed the NPs at the early stages of InP growth using the two opposite flux combinations (LH and HL). We measured the NP volume by integrating NP thickness values from TEM images [see Supporting Information S4 for the details], their chemical composition by EDX and the NW/NP base radius ( R ) from the HRTEM images like those shown in the insets of Figure 3. Based on these data, we calculated the NP contact angle (

 ) by using the equation for the volume of a spherical cap [see Supporting Information S4] as a function of R and  . The obtained results are summarized in Table 1.

12   

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 33

Table 1: The In/Au ratios and contact angles of NPs on top of InAs-InP NW heterostructures grown using the LH and HL fluxes for different durations InP growth time (seconds)

TMIn and TBP fluxes

NP composition (In/Au ratio)

 (deg)

Percentage of straight NWs

0

-

0.5 ± 0.1

113 ± 3

100%

3600

LH

1.3 ± 0.1

104 ± 3

100%

30

HL

1.5 ± 0.1

113 ± 3

100%

60

HL

1.6 ± 0.1

112 ± 1

100%

120

HL

1.8 ± 0.3

115 ± 4

40±5%

We found that the vertical growth of pure InAs NWs is always stable and proceeds without any tapering even at a high  of 113  3o. Straight growth of InP segments on InAs stems without strong tapering (as in the LH sample) requires NPs with much smaller contact angles, typically in the range of   104  3o, even if the In/Au ratio is higher. Bigger NPs (as in the HL samples) yield radial extension at the NW top [see Table S3 in the Supporting Information] and finally the InP segment starts to kink at  = 115  4o as in the longest HL sample [see Table 1 and panels (a-e) of Figure 3]. In order to understand this complex picture, we use the generalized model which takes into account different NW/NP interfacial geometries depending on the NP contact angle. In particular, we consider the surface energies of instantaneously forming truncated NW

13   

ACS Paragon Plus Environment

Page 15 of 33 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

monolayers in different configurations with respect to vertical non-wetted monolayer in the standard VLS growth mode51

Gn w  Gw 

 V   0V  ( SL   LV cos  ) tan  cos 

 L   0V   SL tan    LV sin  cos 

(1)

(2)

Here, the G values are relative to the surface energy of the standard VLS growth mode (forming vertical non-wetted monolayer). G n w are the surface energies of forming non-wetted inward or outward tapered monolayers with symmetrical facets making the angles   to the substrate normal. Positive angle corresponds to inward tapered narrowing facet. Gw is the surface energy of forming wetted monolayer with narrowing facets. Equations (1) and (2) contain the effective surface energies of vertical facets in contact with vapor (  0V ), horizontal top plane in contact with liquid (  SL ), tapered facets in contact with vapor (  V ) and with liquid (  L ), and liquid NP in contact with vapor (  LV ). In order to describe possible changes in the growth direction, we extend the analysis of the VLS energetics by introducing the surface energy of forming a kink:

1 Gkink  (Gw  Gn w ) 2

(3)

assuming that the InP segment kinks when one half of the facets are inward tapered and wetted and the other half is outward tapered and non-wetted. In this case, the NW will most probably change the growth direction toward the side of wetted facet.

14   

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 16 of 33

The liquid-vapor surface energy in our case should be given by that of liquid In as the lowest surface energy metal50 and hence  LV   In  0.459 J/m2 (Ref. [53]). The surface energy of ZB sidewalls of InP NWs is  0V   InP = 0.641 J/m2 (Refs. 54-56). Unfortunately, other surface energies are unknown. However, the  V / cos   0V values must be positive for both narrowing and widening facets, because the initial sidewall planes are stable against faceting.51, 52 This

approach

generalizes

the

Nebolsin-Shchetinin

model46

which

assumes

that

 V / cos   0V  0 for any facet, and yields a rich variety of VLS growth scenarios, some of which are discussed in Ref. [52]. Most importantly for our analysis, when  V / cos   0V  0 , there are two stable contact angles of the NP corresponding to vertical growth of the NW. The NP changes its volume by decreasing or increasing the top NW radius for small and large contact angles, respectively, or by changing the contact angle in between. Increasing the contact angle above a certain value first leads to the formation of truncated wetted facets and then to the NW kinking. To illustrate this effect, we take  V / cos   0V  0.035 and 0.01 J/m2 for narrowing and widening non-wetted facets, respectively, with   19.5o as for (111)A and (111)B facets of ZB III−V NWs. The values of  SL  0.13 J/m2 and  L  0.243 J/m2 are adjusted to fit the observed stability of straight growth at   104o and kinking for    c  115o. Figure 3 shows the minimum surface energies versus the contact angle obtained from Eqs. (1) to (3) with these model parameters. It is seen that narrowing facets tend to form at small contact angles below point 1 in the figure. The region between points 1 and 2, with the contact angles between ~94o and ~111o, corresponds to stable vertical facets without outward tapering, as observed in our LH sample. Within this range, the NPs just change their contact angles without changing the base radius. In the range of contact angles between 111o and 115o, outward tapered facets yield radial 15   

ACS Paragon Plus Environment

Page 17 of 33 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

extension of the NW tops, as observed in HL samples before kinking. Finally, the InP segments kink for higher contact angles. The observed crystal phase switching from pure WZ for InAs NWs to mainly ZB in InP segments at higher In/Au ratios can also be attributed to their different contact angles. Indeed, the Glas condition for the WZ phase formation57 for vertical facets and planar growth interface is better fulfilled for smaller contact angles. In the Tersoff picture,58, 59 the WZ phase formation in III-V NWs is suppressed by truncated wetted facets which tend to occur at larger contact angles.

Figure 3. Minimum surface energy landscape versus the contact angle for the model parameters of InP: vertical facets between points 1 and 2, widening between points 2 and 3 and kinking after point 3. Inserts show the representative TEM images of vertical InP segments on InAs stems for small contact angles in the LH sample (a), widening InP segments for intermediate contact angles (b, c, d: the HL samples with the InP growth times of 30, 60 and 120 s, respectively), and kinked InP segments after passing the critical contact angle of about 115o: the longest InP segment obtained with the HL fluxes (e).

16   

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 18 of 33

The controlled growth of InAs-InP NW heterostructures with straight geometry through the fine tuning of the NP composition and contact angle is of particular interest when more complex nanostructures, like superlattices or quantum dots are desired. Indeed, an InAs quantum dot can be realized within a single InAs NW by growing two InP barriers with a slice of InAs in between. The growth of such structures has always been challenging because of kink formation after a certain InP barrier thickness (usually very small, < 10 nm).21 However, by adopting a growth protocol in which the InP barriers are forced to grow with the NPs having a low In/Au ratio and small contact angle (within the vertical facet growth window), this problem can be overcome and straight thick barriers can be realized. This is demonstrated in Figure 4 by showing the comparison between two InAs quantum dot samples in which the InP barriers have been grown under different precursor fluxes. In this particular case, we used the HL fluxes for the sample depicted in panels (a) and (b) and the LH fluxes for the sample depicted in panels (c) and (d).

Figure 4. Comparison between the InAs quantum dot samples in which the InP barriers are grown using the HL (a, b) and LH (c, d) fluxes. Panels (a, c) show the 45° tilted SEM images of as-grown samples and (b, d) the STEM-HAADF micrographs of the NW portion with two InP barriers. 17   

ACS Paragon Plus Environment

Page 19 of 33 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

As clearly visible in Fig. 4, only a small fraction of the NWs (⁓30%) has a straight morphology in the first sample and the thickness of the barriers is only around 10 nm. Conversely, the barrier thickness is 30 nm in average in the second sample and the yield of straight NWs is extremely high (more than 80%) compared to what was reported so far.21 We now investigate the interfacial abruptness in the InP/InAs/InP NW heterostructures. For this purpose, we measured the High Angle Annular Dark Field (HAADF) intensity line profile and modeled the variation of P concentration in the NP (y) and in the NW (x) across the two heterojunctions, with the results shown in Fig. 5.

The details of modeling are given in

Supporting Information S5. In brief, due to a low As and P solubility in the NPs, we find that the P content in the NP, y , is well described by the universal equations

   0  ys  , y  y  ys tanh 1  (  1 ) /  *  * 

(4)

which apply for the InAs to InP and InP to InAs transition, respectively. Here,  is the distance along the NW growth axis,  0 ( 1 ) is the distance at which the As (P) flux is changed to the P (As) one, and y s is the stationary P composition in liquid. The InP fraction x in the solid InAs1-xPx is obtained as described in the Supporting Information S5. The sole parameter of the model:

*  ys

 S f (  ) R  In v In (c P  c As ) .  P vP 3 L

(5)

determines the interfacial abruptness. Here,  S is the elementary volume per III-V pair in solid,

 L is the elementary volume per atom in liquid, f (  ) is the geometrical function of the contact

18   

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 20 of 33

angle,  In v In /  P v P is the effective In to P flux ratio and c P  c As is the total concentration of the group V atoms (P and As) dissolved in the NP. Our  * , compared to the results of Refs. [28, 60] for heterostructures based on the group III interchange, contains a very small factor c P  c As on the order of 0.01, showing why heterostructures based on the group V interchange must be extremely sharp. Indeed, if we take a plausible set of parameters: y s =0.9,   105o, R  30 nm,

 In vIn /(  P vP ) = 0.1, the  * value equals only 0.3 nm (even for a high c P  c As of 0.05), that is, on the order of one monolayers, while it would be on the order of 30 monolayers for GaAs/AlxGa1-xAs NWs of the same radius.60 According to Eq. (4), the direct InAs to InP switch should be sharper than the reverse one, because the relaxation of the composition with the distance is exponential in the first and only inverse power law in the second case. However we could not appreciate this difference in the experimental profile obtained by the HAADF intensity, due to the experimental error. Atomic resolution in the HAADF data (through High Resolution TEM) is likely necessary to capture the details of such sharp interfaces. The suppression of the reservoir effect in axial NW heterostructures based on the interchange of highly volatile As and P species is intuitively clear and has been discussed elsewhere (see, for example. Ref. [61] and references therein). However, the quantification of the interfacial abruptness in this case, such as given by our simple Eq. (5), has not been achieved before to our knowledge. Moreover, we note that the developed approach is not critically sensitive to the epitaxy technique used to grow the NW heterostructures and in particular to the material inputs. Indeed, the arrival rates of the group III and V atoms enter Eq. (5) only in the ratio vIn / vP , while the total concentration of the group V atoms in the liquid ( cP  cAs ) is much smaller than unity in any case, regardless of the growth technique. 19   

ACS Paragon Plus Environment

Page 21 of 33 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 5. Interfacial abruptness: (a) STEM-HAADF image of the InP/InAs/InP NW heterostructure used for composition profiling. (b) P content across the heterostructure depicted in (a): experimental profile obtained from HAADF intensity (columns) and theoretical curves (lines) obtained from Eqs. (4) with

 *  0.3 nm.

In conclusion, we have developed a new general procedure for the controlled realization of straight InAs-InP axial NW heterostructures with atomically sharp heterointerfaces. Straight growth of InP segments on InAs stems systematically requires low In/Au ratios in the NPs and small contact angles to ensure the stability of vertical facets with respect to kinking. This is realized here by tuning the precursor fluxes at a fixed growth temperature, but in principle can be achieved also by varying the growth temperature, if the temperature window is appropriately chosen to ensure the relevant conditions for VLS growth of both materials forming the heterostructure.

Low group III and high group V fluxes yield straight WZ InAs-InP 20 

 

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 33

heterostructures with the best crystal quality. The interfacial abruptness of InP/InAs/InP NW quantum dots is greatly improved compared to the case of the group III interchange due to a low group V content in the NPs. We presented a theoretical model which allows for quantitative description of the NW composition versus material fluxes, interfacial abruptness and NW morphology. We strongly believe that the obtained results provide a general understanding of the growth of axial NW heterostructures based on group V interchange and can be used for the fine tuning of NW heterostructures in other material systems.

METHODS The NWs were grown in a Riber Compact-21 CBE system. Details of the growth chamber setup are given in Ref. [23]. We used InAs(111)B substrates and Au nanoparticles obtained either by drop casting of monodispersed colloidal solution or by thermal dewetting of a thin (0.5 nm) Au film deposited on the substrate at room temperature in a thermal evaporator. In the case of Au film, an annealing step (20 min at 470 ± 10°C under As flux of 1 Torr) was performed prior to the growth in order to create the nanoparticles. The InAs NW stems were grown for 30 minutes at the growth temperature of 430 ± 10°C, using precursor fluxes of 0.6 Torr and 1.5 Torr for TMIn and TBAs, respectively. To produce the axial InAs-InP NW heterostructures, the InAs NW growth was followed by the InP deposition at the same temperature using different combinations of TMIn and TBP fluxes. It should be specified that a direct As-to-P switch was performed when the TMIn flux was kept constant for the two materials, while a growth interruption between InAs and InP (keeping the sample under the As flux for 4 min at the growth temperature) was needed when the TMIn flux was changed in order to adjust and stabilize the 21   

ACS Paragon Plus Environment

Page 23 of 33 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

flux before starting the growth of InP segments. All the samples were cooled down in UHV environment, after simultaneously switching off both group III and group V precursor fluxes. Morphological characterization of the NWs was performed using a Zeiss field-emission SEM operated at 5 kV. Structural characterization by TEM was carried out using a JEOL JEM-2200FS microscope, equipped with an in-column  filter, operated at 200 kV. For TEM observation, the NWs were transferred to carbon-coated Cu or Ni grids by gentle rubbing. Imaging was carried out either in high resolution TEM mode after zero-loss energy filtering, or in STEM mode using an HAADF detector for Z-contrast imaging. EDX measurements were performed in STEM spot mode, enabling drift correction by probe tracking, with a spot size of 1.5 nm. The results (expressed in atomic percent units, with a statistical dispersion of about 2%) were averaged over 10-12 measured NWs for each sample.

AUTHOR INFORMATION Corresponding Author

* E-mail: [email protected] Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

ACKNOWLEDGMENTS

22   

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 33

We gratefully acknowledge the bilateral project CNR/RFBR. VGD thanks the Ministry of Education and Science of the Russian Federation for financial support received under grant 14.587.21.0040 (project ID RFMEFI58717X0040).

Supporting Information available:

S1: Cooling experiments and EDX measurements S2: Quantification of the stationary NP composition for InAs and InAs-InP NWs S3: Morphology of InAs-InP NW heterostructures obtained with 20 nm colloids and Au film dewetting S4: Contact angle determination from the NP volume S5: Modeling of interfacial abruptness

REFERENCES

1. Lieber, C. M. Nano Lett. 2002, 2 (2), 81–82. 2. Larsson, M. W.; Wagner, J. B.; Wallin, M.; Håkansson, P.; Fröberg, L. E.; Samuelson, L.; Wallenberg, L. R. Nanotechnology 2007, 18 (1), 15504. 3. De la Mata, M.; Magén, C.; Caroff, P.; Arbiol, J. Nano Lett. 2014, 14 (11), 6614–6620. 4. Hiruma, K.; Murakoshi, H.; Yazawa, M.; Katsuyama, T. J. Cryst. Growth 1996, 163, 226– 231. 5. Björk, M. T.; Thelander, C.; Hansen, A. E.; Jensen, L. E.; Larsson, M. W.; Wellenberg, L. R.; Samuelson, L.; Wallenberg, L. R.; Samuelson, L. Nano Lett. 2004, 4 (9), 1621–1625.

23   

ACS Paragon Plus Environment

Page 25 of 33 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

6. Nilsson, H. A.; Duty, T.; Abay, S.; Wilson, C.; Wagner, J. B.; Thelander, C.; Delsing, P.; Samuelson, L. Nanowire. Nano Lett. 2008, 8, 872–875. 7. Rossella, F.; Bertoni, A.; Ercolani, D.; Rontani, M.; Sorba, L.; Beltram, F.; Roddaro, S. Nat. Nanotechnol. 2014, 9, 997–1001.

8. Lind, E.; Persson, A. I.; Samuelson, L.; Wernersson, L. E. Nano Lett. 2006, 6, 1842–1846. 9. Borgstrom, M. T.; Zwiller, V.; Muller, E.; Imamoglu, A. Nano Lett. 2005, 5, 1439–1443. 10. Wagner, R. S.; Ellis, W. C. Appl. Phys. Lett. 1964, 4 (5), 89–90. 11. Clark, T. E.; Nimmatoori, P.; Lew, K.; Pan, L.; Redwing, J. M.; Dickey, E. C. Nano Lett. 2008, 8 (4), 1246–1252.

12. Li, N.; Tan, T. Y.; Gösele, U. Appl. Phys. A Mater. Sci. Process. 2008, 90 (4), 591–596. 13. Dick, K.; Bolinsson, J.; Borg, B. M.; Johansson, J. Nano Lett. 2012, 12, 3200–3206. 14. Wen, C.-Y.; Reuter, M. C.; Bruley, J.; Tersoff, J.; Kodambaka, S.; Stach, E. A.; Ross, F. M. Science 2009, 326 (5957), 1247–1250.

15. Perea, D. E.; Li, N.; Dickerson, R. M.; Misra, A.; Picraux, S. T. Nano Lett. 2011, 11 (8), 3117–3122. 16. Paladugu, M.; Zou, J.; Guo, Y. N.; Auchterlonie, G. J.; Joyce, H. J.; Gao, Q.; Tan, H. H.; Jagadish, C.; Kim, Y. Small 2007, 3 (11), 1873–1877. 17. Paladugu, M.; Zou, J.; Guo, Y. N.; Zhang, X.; Kim, Y.; Joyce, H. J.; Gao, Q.; Tan, H. H.; Jagadish, C. Appl. Phys. Lett. 2008, 93 (10), 91–94. 18. Messing, M. E.; Wong-leung, J.; Joyce, H. J.; Zanolli, Z.; Gao, Q.; Tan, H. H.; Reine, L.; Johansson, J.; Jagadish, C. Nano Lett. 2011, 11, 3899–3905. 19. Björk, M. T.; Ohlsson, B. J.; Thelander, C.; Persson, A. I.; Deppert, K.; Wallenberg, L. R.; Samuelson, L. Appl. Phys. Lett. 2002, 81, 4458–4460.

24   

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 26 of 33

20. Gudiksen, M. S.; Lauhon, L. J.; Wang, J.; Smith, D. C.; Lieber, C. M. Nature 2002, 415 (6872), 617–620. 21. Svensson, S. F.; Jeppesen, S.; Thelander, C.; Samuelson, L.; Linke, H.; Dick, K. a. Nanotechnology 2013, 24 (34), 345601.

22. Haapamaki, C. M.; Lapierre, R. R. Nanotechnology 2011, 22 (33), 335602. 23. Gomes, U. P.; Ercolani, D.; Zannier, V.; Beltram, F.; Sorba, L. Semicond. Sci. Technol. 2015, 30 (11), 115012.

24. Wang, J.; Plissard, S. R.; Verheijen, M. A.; Feiner, L. F.; Cavalli, A.; Bakkers, E. P. A. M. Nano Lett. 2013, 13 (8), 3802–3806.

25. Froberg, L. E.; Wacaser, B. A.; Wagner, J. B.; Ohlsson, B. J.; Deppert, K.; Samuelson, L. Nano Lett. 2008, 8 (11), 3815–3818.

26. Ghasemi, M.; Johansson, J. Cryst. Growth Des. 2017, 17, 1630–1635. 27. Dubrovskii, V. G.; Koryakin, A. A.; Sibirev, N. V. Materials and Design 2017, 132, 400– 408. 28. Glas, F. Cryst. Growth Des. 2017, doi: 10.1021/acs.cgd.7b00732. 29. Ameruddin, A. S.; Caroff, P.; Tan, H. H.; Jagadish, C.; Dubrovskii, V. G. Nanoscale 2015, 7, 16266−16272. 30. Dubrovskii, V. G. Cryst. Growth Des. 2015, 15, 5738–5743. 31. Dubrovskii, V. G.; Sibirev, N. V. Cryst. Growth Des. 2016, 16, 2019–2023. 32. Dubrovskii, V. G. J. Phys. D: Appl. Phys. 2017, 50, 453001. 33. Dubrovskii, V. G. J. Chem. Phys. 2009, 131, 164514. 34. Messing, M. E.; Wong-leung, J.; Joyce, H. J.; Zanolli, Z.; Gao, Q.; Tan, H. H.; Reine, L.; Johansson, J.; Jagadish, C. Nano Lett. 2011, 11, 3899–3905. 35. Dick, K. a; Caroff, P.; Bolinsson, J.; Messing, M. E.; Johansson, J.; Deppert, K.; Wallenberg, L. R.; Samuelson, L. Semicond. Sci. Technol. 2010, 25 (2), 024009. 36. Dubrovskii, V. G. J. Chem. Phys. 2015, 142 (20). 37. Ghalamestani, S. G.; Heurlin, M.; Wernersson, L.-E.; Lehmann, S.; Dick, K. A. Nanotechnology 2012, 23 (28), 285601. 25   

ACS Paragon Plus Environment

Page 27 of 33 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

38. Rieger, T.; Scha, T.; Gru, D.; Lepsa, M. I. Cryst. Growth Des. 2014, 14, 1167–1174. 39. Namazi, L.; Nilsson, M.; Lehmann, S.; Thelander, C.; Dick, K. A. Nanoscale, 2015,7, 10472-10481. 40. Ercolani, D.; Gemmi, M.; Nasi, L.; Rossi, F.; Pea, M.; Li, A.; Salviati, G.; Beltram, F.; Sorba, L. Nanotechnology 2012, 23 (11), 115606. 41. Roper, S. M.; Davis, S. H.; Norris, S. A.; Golovin, A. A.; Voorhees, P. W.; Weiss, M. J. Appl. Phys. 2007, 102 (3).

42. Roper, S. M.; Anderson, A. M.; Davis, S. H.; Voorhees, P. W. J. Appl. Phys. 2010, 107 (11). 43. Ross, F. M. Rep.Prog. Phys. 2010, 73, 114501 44. Schwarz, K. W.; Tersoff, J. Nano Lett. 2011, 11 (2), 316–320. 45. Schwarz, K. W.; Tersoff, J.; Kodambaka, S.; Ross, F. M. Phys. Rev. Lett. 2014, 113 (5), 1–5. 46. Nebol’sin, V. A.; Shchetinin, A. A. Inorg. Mater. 2003, 39, 899−903. 47. Hocevar, M.; Immink, G.; Verheijen, M.; Akopian, N.; Zwiller, V.; Kouwenhoven, L.; Bakkers, E. Nat. Commun. 2012, 3, 1266. 48. Madras, P.; Dailey, E.; Drucker, J. Nano Lett. 2009, 9 (11), 3826–3830. 49. Zannier, V.; Grillo, V.; Martelli, F.; Plaisier, J. R.; Lausi, A.; Rubini, S. Nanoscale 2014, 6 (14), 8392–8399. 50. Zannier, V.; Ercolani, D.; Gomes, U. P.; David, J.; Gemmi, M.; Dubrovskii, V. G.; Sorba, L. Nano Lett. 2016, 16 (11), 7183–7190.

51. Dubrovskii, V. G. Cryst. Growth Des. 2017, 17, 2544–2548. 52. Kim, W.; Dubrovskii, V. G.; Vukajlovic-Plestina, J.; Tütüncüoglu, G.; Francaviglia, L.; Güniat, L.; Potts, H.; Friedl, M.; Leran, J. P.; Fontcuberta i Morral, A. Nano Lett. 2017, submitted. 53. Jasper, J. J. J. Phys. Chem. Ref. Data 1972, I (4), 841−1009. 54. Hilner, E.; Håkanson, U.; Froberg, L. E.; Karlsson, M.; Kratzer, P.; Lundgren, E.; Samuelson, L.; Mikkelsen, A. Nano Lett. 2008, 8, 3978−3982. 55. Joyce, H. J.; Wong-Leung, J.; Gao, Q.; Tan, H. H.; Jagadish, C. Nano Lett. 2010, 10, 908−915. 56. Rosini, M.; Magri, R. ACS Nano 2010, 4, 6021−6031. 57. Glas, F.; Harmand, J. C.; Patriarche, G. Phys. Rev. Lett. 2007, 99, 146101.

26   

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 28 of 33

58. Wen, C.-Y.; Tersoff, J.; Hillerich, K.; Reuter, M. C.; Park, J. H.; Kodambaka, S.; Stach, E. A.; Ross, F. M. Phys. Rev. Lett. 2011, 107, 025503. 59. Jacobsson, D.; Panciera, F.; Tersoff, J.; Reuter, M. C.; Lehmann, S.; Hofmann, S.; Dick, K. A.; Ross, F. M. Nature 2016, 531, 317. 60. Priante, G.; Glas, F.; Patriarche, G.; Pantzas, K.; Oehler, F.; Harmand, J. C. Nano Lett. 2016, 16, 1917–1924.

61. Priante, G.; G. Patriarche, G.; Oehler, F.; Glas, F.; Harmand, J. C. Nano Lett. 2015, 15, 6036–6041.

Graphical table of contents:

The growth mode of InP on InAs nanowire stems depends on the stability of vertical VLS growth, which is strongly affected by the chemical composition and contact angle of the nanoparticle on the nanowire top. Stable vertical growth of InP occurs within a well-defined range of contact angles (94 – 111°). Below (above) this range narrowing (widening) of the nanowire radius occurs. The InP segments kink above the critical contact angle of 115°.

27   

ACS Paragon Plus Environment

Page 29 of 33 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

241x198mm (150 x 150 DPI)

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

749x369mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 30 of 33

Page 31 of 33 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

286x232mm (150 x 150 DPI)

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

474x160mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 32 of 33

Page 33 of 33 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

145x162mm (150 x 150 DPI)

ACS Paragon Plus Environment