Catalyst Composition Tuning: The Key for the ... - ACS Publications

Subscriber access provided by RYERSON UNIV

Communication

Catalyst composition tuning: the key for the growth of straight axial nanowire heterostructures with group III interchange Valentina Zannier, Daniele Ercolani, Umesh Prasad Gomes, Jeremy David, Mauro Gemmi, V. G. Dubrovskii, and Lucia Sorba Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.6b03524 • Publication Date (Web): 19 Oct 2016 Downloaded from http://pubs.acs.org on October 20, 2016

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 26

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

Catalyst composition tuning: the key for the growth of straight axial nanowire heterostructures with group III interchange Valentina Zannier* #, Daniele Ercolani #, Umesh Prasad Gomes #, Jérémy David §, Mauro Gemmi §, Vladimir G. Dubrovskii †& and Lucia Sorba #.

#

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

56127 Pisa, Italy §

Center for Nanotechnology Innovation @NEST, Istituto Italiano di Tecnologia, Piazza S.

Silvestro 12, 56127 Pisa, Italy †

St. Petersburg Academic University, Khlopina 8/3, 194021, St. Petersburg, Russia

&

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

ABSTRACT Au-catalyzed III-V nanowire heterostructures based on the group III interchange usually grow straight only in one of the two growth sequences, while the other sequence produces kinked geometries, and thus the realization of double heterostructures remains challenging. Here, we investigate the growth of Au‐assisted InAs‐GaAs and GaAs-InAs axial nanowire heterostructures. A detailed study of the heterostructure morphology as a function of growth parameters and chemical composition of the catalyst nanoparticle is performed by means of

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 26

scanning electron microscopy, transmission electron microscopy, and energy dispersive x‐ray analysis. Our results clearly demonstrate that the nanoparticle composition, rather than other growth parameters, as postulated so far, controls the growth mode and the resulting nanowire morphology. While GaAs easily grows straight on InAs, straight growth of InAs on GaAs is achieved only if the nanoparticle composition is properly tuned. We find that straight InAs segments on GaAs require high group III/Au ratios in the nanoparticle (greater than 0.8), otherwise the droplet wets the sidewalls and the nanowire kinks. We discuss the observed behavior within a theoretical model which relates the nanoparticle stability to group III/Au ratio. Based on this finding, we demonstrate the growth of straight nanowire heterostructures for both sequences. The proposed strategy can be extended to other III‐V nanowire heterostructures based on the group III-interchange, allowing for straight morphology regardless of the growth sequence, and ultimately for designing nanowire heterostructures with the required properties for different applications.

KEYWORDS: nanowires, axial heterostructures, VLS growth, catalyst composition

Semiconductor nanowires (NWs) are considered promising for both fundamental studies and device applications. One of the greatest advantages of NWs is their ability to relax elastic stress in two dimensions. This feature enables defect-free growth in lattice-mismatched material systems and yields structures that are not achievable in thin films or Stranski-Krastanow quantum dots. Several types of devices based on axial NW heterostructures have already been demonstrated, including resonant tunneling diodes,1 single-electron transistors,2,3 field-effect


Page 3 of 26

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

transistors,4 and optically bright NW quantum dots for single photon emitters.5 Although epitaxial growth of axial NW heterostructures was started in 1996,6 their formation process is not straightforward and far from being fully understood. The main issue in the vapor-liquid-solid (VLS) growth of axial NW heterostructures is obtaining straight junctions with sharp interfaces between the two materials. Considering Au-assisted III-V NWs, straight and abrupt axial heterointerfaces are easier produced based on the group V interchange.7,8 A low solubility of most group V elements in Au-group III melts allows for minimizing the reservoir effect and the surface energy changes at the flow commutation.9,10 Much higher solubility of group III elements in liquid Au increases the reservoir effect and broadens the interfaces.10 Furthermore, straight NW heterostructures are often obtained only for one of the two interfaces and hence the growth of a double heterostructure remains very challenging.9,11 The non-equivalence of the two interfaces has previously been explained within a thermodynamic model that involves the surface energies between the pure Au particle and each material and the interface energy between the two materials.9 In the specific case of InAs-GaAs NW heterostructures, straight axial growth has been reported for GaAs on top of InAs (hereafter InAs-GaAs NWs), whereas the other growth sequence (InAs on top of GaAs, GaAs-InAs NWs) usually yields kinked morphologies.9,12,13 It has also been observed that the GaAs to InAs transition is much sharper than the reversed one. This result is consistent with thermodynamic model of Ref. [9], where the calculated Au/GaAs surface energy is lower than for Au/InAs. Therefore, the Au particle prefers to remain in contact with GaAs on surface energetic grounds and the top InAs segment tends to kink. These considerations apply to the equilibrium case and pure materials. In non-equilibrium situation that in fact occurs in any VLS growth, the deposition conditions are known to affect the


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 26

resulting growth mode, morphology and even the crystal structure of III-V NWs, including ternary systems. For example, more recent results of Ref. [14] demonstrate that straight InAs segments on top of GaAs NWs can be achieved by combining high temperatures and high III/V flow ratios in Au-catalyzed metal organic chemical vapor deposition (MOCVD). Such conditions favor the wurtzite crystal structure of InAs segments and then a significant fraction (40-60%) of the GaAs-InAs NW heterostructures grow straight. If the growth parameters are tuned to promote zincblende InAs segments, no straight GaAs-InAs NW growth is observed. A refined model of Ref. [14] takes into account the different surface energies for the two crystal structures and supports the experimental results. However, this model also refers to a system at equilibrium and considers the catalyst consisting of pure Au rather than an Au-III-V alloy. Both Ga and In are expected to strongly modify the catalyst composition and shape that determine the resulting NW growth mode and morphology. Consequently, here we report on the chemical beam epitaxy (CBE) growth of both GaAs-InAs and InAs-GaAs axial NW heterostructures and identify the conditions for obtaining straight double heterostructures in this material system.

RESULTS We started our study from the growth of GaAs on top of InAs NWs. Since the optimal temperature window for the GaAs NW growth is higher than that for InAs NWs, we had to develop a multi-step protocol in order to grow InAs stems and simultaneously increase the temperature for the subsequent GaAs segment growth. We adjusted the growth parameters (substrate temperature and NW density) to promote the axial growth and reduce the radial growth of GaAs on the InAs stems. Details of the growth protocol optimization can be found in the supporting information S1. Figure 1 shows the InAs-GaAs NW heterostructures with the


Page 5 of 26

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

optimized morphology. Panel (a) is a 45°-tilted scanning electron microscopy (SEM) image of the as-grown sample, while panels (b) and (c) present the results of transmission electron microscopy (TEM) analysis of individual NWs transferred onto the TEM grid. As visible in Figure 1 (b), in which we show a scanning-TEM (STEM) image and the energy dispersive x-ray (EDX) elemental line profile of a single NW, the InAs/GaAs interface (indicated by the yellow arrow) is quite sharp. There is a short (less than 100 nm) GaAs shell around the InAs wire, but only in the proximity of the heterointerface (see the green arrow, which indicates the shell end). This shell is formed by lateral overgrowth around the core and can be explained by the model of Ref. [15], suggesting a sphere of “activated” Ga around the NP which is available for growing GaAs. From the length of the GaAs shell, the diffusion length of the “activated” Ga can be estimated at approximately 100 nm. Panel (c) of Figure 1 shows the TEM image of the upper part of another nanowire, along with the selected area electron diffraction (SAED) pattern of the two segments. Both InAs and GaAs parts have the wurtzite structure. The InAs stem is almost free from structural defects, while the GaAs top contains many stacking faults that are randomly distributed along its length.


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 26

Figure 1. (a) 45o-tilted SEM image of dense InAs-GaAs NWs. We have coloured one single NW in order to highlight the InAs portion (green), the GaAs one (blue) and the Au nanoparticle (yellow). (b) EDX line profile (top) and STEM image (bottom) of an individual NW. The yellow arrow points at the InAs/GaAs interface; the green arrow indicates the end of the GaAs shell grown around the InAs core. (c) TEM image with the SAED patterns of the InAs and GaAs parts, taken along [2, -1, -1, 0] zone axis.

As regards the nanoparticle (NP) composition after growth, it is known that the affinity of In for Au is much higher than that of Ga.16-19 As previously reported,10 Au/In alloy particles (with very little or no measurable Ga) are found even after growth of long (300-1000 nm) GaAs segments on top of InAs stems. For NWs shown in Figure 1, the GaAs segment length after 5 minutes of growth is around 140 nm and the NPs on their tips contain 62% of In and 38% of Au, while the Ga concentration is below our detection limit ( ̴ 4%). In order to investigate the compositional changes in the NPs after the GaAs deposition for different durations, we grew a series of samples following the protocol shown in Figure 2 (a). The growth of InAs (5 min at 410 °C plus 10 min temperature ramp up to 460 °C) was followed by GaAs (5 min temperature ramp up to 530 °C plus the GaAs deposition at this temperature for different growth times tGaAs). Figure 2 (b) shows the morphology of InAs-GaAs heterostructures obtained after tGaAs of 5, 15 and 55 min.


Page 7 of 26

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 2. (a) Growth protocol for InAs-GaAs NWs with different GaAs growth time tGaAs. (b) SEM images of representative InAs-GaAs NWs grown for the different tGaAs reported in the panel. (c) Plot of the nanoparticle composition (In/Au and Ga/Au ratios) versus tGaAs.

The NPs on the tips of pure InAs NWs (tGaAs=0) are 67/33 In/Au alloys. According to the Au-In phase diagram20 and previous study of the Au-assisted InAs NWs,21 such alloy NPs should be liquid at the growth temperature, suggesting the vapor-liquid-solid (VLS) growth mechanism. As we start growing the GaAs segment, the amount of In decreases and the amount of Ga slightly increases. Since adding Ga to Au/In alloy lowers its melting temperature,22 the NPs remain liquid and the VLS growth mechanism pertains when we switch from one material to the other. Figure 2 (c) shows that the amount of In quickly decreases from 67% to 40% after tGaAs=15 min, corresponding to the GaAs segment length around 450 nm. Further increase of


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 26

tGaAs does not change much the amount of In, which remains at around 30% even after 95 min of growth (resulting in ~ 1700 nm long GaAs segment). The Ga percentage is below the detection limit before 15 min of growth and reaches 12-16% only after very long times (55-95 min, corresponding to GaAs segments longer than 1000 nm). So, the overall group III/Au ratio of the NPs decreases with tGaAs. We then performed a 10 min long growth of InAs (while ramping down the temperature, T = -70°C) on top of InAs-GaAs NWs with different GaAs lengths, as shown in Figure 3 (a). For comparison with these double InAs-GaAs-InAs NW heterostructures, we also grew single GaAsInAs NW heterostructures following the standard procedure shown in Figure 3 (b). The results were quite surprising. It turned out that the InAs part can grow straight only on top of the shortest GaAs segments in InAs-GaAs NWs (tGaAs = 5 min, the GaAs length about 140 nm). Figure 3 (c) shows the SEM image of as-grown sample, where 75% of all NWs are straight InAsGaAs-InAs heterostructures. Figure 3 (d) shows the EDX line profiles for As, In, Ga and Au elements along a single NW, where the two interfaces are easily identified. The GaAs/InAs interface is sharper than the InAs/GaAs one, as evidenced by the slope of the single-element lines and the length of the In/Ga mixed regions (around 100 for InAs/GaAs and less than 50 nm for GaAs/InAs ). This result confirms again a higher affinity of In for Au with respect to Ga.10,1619


Page 9 of 26

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,b) Growth protocols for (a) InAs-GaAs-InAs NWs and (b) GaAs-InAs NWs. (c) SEM image of InAs NWs grown on the InAs-GaAs stems with short GaAs segments (tGaAs = 5 min). (d) EDX line profile of a straight InAs-GaAs-InAs NW heterostructure (e) SEM image of InAs NWs grown on the InAs-GaAs stems with longer GaAs segments (tGaAs = 55min). (f) SEM image of GaAs-InAs NWs. In panels (c), (e) and (f), one representative NW is colored with InAs shown in green, GaAs in blue and the nanoparticle in yellow.

For longer GaAs segments, the InAs parts tend to grow downward and produce kinked geometries. Only 15% of the double NW heterostructures obtained with tGaAs = 55 min are straight, as visible in Figure 3 (e). Similar behavior was found in the limiting case of single GaAs-InAs heterostructure (Figure 3 (f)), and independent of the InAs growth temperature (lower, higher or equal to that of GaAs). As mentioned above, the downward growth of InAs on top of GaAs NWs was previously observed in Au-catalyzed MOCVD11-13 and interpreted on surface energetic grounds for pure materials. Our results suggest a more complex picture which


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 26

depends critically on the NP composition at the flow commutation. Indeed, the main difference between double InAs-GaAs-InAs heterostructures with short and long GaAs segments is the amount of group III element stored in the NP on top of the GaAs segment. According to Figure 2 (c), at tGaAs = 5 min the NP is an In/Au alloy with a high In percentage (In/Au=1.63) and without any Ga. At tGaAs = 55 min, the alloy is ternary In/Ga/Au but the total group III to Au ratio is only 0.61. The InAs parts grow straight on short GaAs segments from NPs with high group III percentage and kink on longer GaAs segments with lower group III percentage. Therefore, higher group III content in the NP seems to favor straight heterostructure geometry. Such behavior is further confirmed by growth interruption tests described in the supporting information S2, demonstrating that a depletion of the group III species from the NPs before InAs growth causes kinking. Since a high percentage of In in the NPs before the InAs growth may also be important for the straight geometry of InAs segments on GaAs, we have investigated the possibility to seed the GaAs NWs by Au/In (or Au/In/Ga) catalysts. To realize this, we performed the In pre-deposition (using a TMIn flux of 0.3 Torr) on Au-GaAs(111)B substrates immediately after annealing. We tested different In pre-deposition times and found an optimum of 10 seconds. This is enough to form the Au/In alloys but at the same time allows us to avoid any nucleation of pure In droplets on the surface. Figure 4 (a) shows the SEM image of the surface after this treatment, along with the typical EDX spectrum acquired from a single NP. Unfortunately, in this case it is not possible to decouple the Ga signals originating from the NP and GaAs substrate and hence to precisely determine the composition of the initial NP. The presence of In signal in the EDX spectra confirms, however, the In alloying with Au.


Page 11 of 26

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 4. (a) SEM image of a GaAs(111)B substrate covered with 0.1 nm thick Au layer after the annealing and 10 s of In pre-deposition. EDX spectrum is taken from the NP marked with green cross. (b) 45°-tilted SEM image of GaAs NWs grown for 60 min. (c) TEM image of a single GaAs NW and the corresponding SAED pattern along [2,-1,-1,0] zone axis.

After the In pre-deposition, GaAs NWs were grown following the standard protocol. The resulting NWs shown in Figure 4 (b) are very similar in morphology and crystal structure to the standard GaAs NWs grown without any In. They are uniformly oriented, slightly tapered, show the wurtzite crystal structure with many randomly distributed stacking faults (see Figure 4 (c)).


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 26

However, the chemical composition of the catalysts NPs on their tips is different. After 60 min of GaAs growth, we found that the NPs are 45/55 Ga/Au alloys (with no In detected), compared to 33/67 Ga/Au composition in the standard case with no In pre-deposition. This suggests that the presence of In in the Au NPs before the GaAs deposition increases the solubility of Ga in Au in a later growth stage. Quite interestingly, the growth of InAs segments on top of GaAs NWs obtained with the In pre-deposition produces straight GaAs-InAs NW heterostructures with a high yield (85%), regardless of the InAs growth temperature. Figure 5 shows the NWs obtained after 10 min of InAs growth with the temperature ramp down (T=-50°C). The InAs and GaAs segments are clearly visible in the SEM image of Figure 5 (a) and in the EDX profile of Fig. 5 (b). The TEM image in Figure 5 (c) demonstrates high quality wurtzite structure of the InAs segment, while the wurtzite GaAs part contains many stacking faults, as we saw earlier in the reversed InAs-GaAs NW heterostructures. This is not surprising because the crystal structure of III-V NWs depends on the growth parameters23 that are kept the same for both InAs and GaAs in the two growth sequences.


Page 13 of 26

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. Straight growth of InAs segments on top of GaAs NWs obtained with the In pre-deposition: (a) 45°-tilted SEM image of as-grown NWs. In the colored NW, the GaAs portion is shown in blue and the InAs portion in green. (b) EDX line profile of the elements along the growth axis of the NW displayed below. (c) TEM image and the corresponding SAED pattern (along [2,-1,-1,0] zone axis) of a NW portion close to the GaAs-InAs heterointerface.

DISCUSSION We now compare the growth modes of InAs segments on GaAs NWs obtained with and without the In pre-deposition with our previous results on the InAs-GaAs-InAs double heterostructures. Ga-rich Ga/Au alloy NPs on top of GaAs NWs obtained with the Inpredeposition (Ga/Au ratio = 0.82) yield straight GaAs-InAs heterostructures. Straight InAs growth is also observed on top of InAs-GaAs NWs with a high initial In content in the In/Au NPs (In/Au ratio = 1.63). Conversely, InAs grows downward on the Au-seeded GaAs NWs with a lower Ga content in their Ga/Au NPs (Ga/Au ratio = 0.49), and on top of InAs-GaAs NW heterostructures with a lower group III content in their NPs [(In+Ga)/Au ratio ≤ 0.79]. A more detailed experimental data on the InAs growth modes on GaAs versus the initial NP composition are given in Table 1. There is no problem growing straight GaAs segments on top of InAs, where


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 26

the initial In content in the In/Au NPs is always very high (average In/Au ratio = 2). Therefore, straight Au-catalyzed VLS growth of InGaAs NW heterostructures systematically requires high group III/Au ratios in the initial NPs (with a minimum of 0.82 according to the data of Table 1), and actually regardless of the In to Ga composition inside them. This property is always fulfilled for GaAs on InAs but requires special care for InAs on GaAs. Similar effects have been observed in other systems: prefilling of the Au NP with In prior to growth has been demonstrated very effective to control the growth direction of InP NWs and prevent kinking.24 Likewise, an excess of Ga in the Au particle has resulted essential to grow straight GaP on top of Si NWs.25

Table 1. Summary of the experimental data on the InAs growth modes on GaAs at different conditions yielding different NP compositions

GaAs growth time (min)

Straight InAs growth

Initial NP composition Ga/Au

In/Au

Group III/Au

InAs-GaAs-InAs double heterostructures 0

0

2.0

2.0

Yes

5

0

1.63

1.63

Yes

15

0.09

0.70

0.79

No

35

0.15

0.54

0.69

No

55

0.19

0.42

0.61

No

95

0.28

0.47

0.75

No

GaAs-InAs single heterostructures with In pre-deposition 60

0.82

0

0.82

Yes

GaAs-InAs single heterostructures without In pre-deposition


Page 15 of 26

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

60

0.49

0

0.49

No

To understand why high group III content in the initial NPs promotes straight NW growth, we use the Nebol’sin-Shchetinin model for the NP stability on the NW top,26 which is a simplified version of a more general condition considered later in Refs. [27,28]. The model states that stable VLS growth occurs when

 SV  sin   cos  .  LV

(1)

This inequality connects the solid-vapor surface energy of the NW sidewalls  SV and the liquidvapor surface energy of the NP  LV with the contact angle of the droplet  , and is independent of the NW radius. Condition given by Eq. (1) applies for straight NW sidewalls and planar growth interface, which is indeed the case for wurtzite III-V NWs29 without a truncated edge30. When the inequality given by Eq. (1) is inverted, the NP slides downward and produces kinked geometries. Very importantly, in this model the NP stability is not determined by the solid-vapor or liquid-vapor surface energy alone, but rather by their ratio, and does not depend on the solidliquid surface energy9. Neglecting

the

As

concentration,

the

droplet

volume

is

given

by

( N Au  N 3 )  (R3 / 3) f (  ) , with f (  )  (1  cos  )( 2  cos  ) /[1  cos  ) sin  ] . Here, the left hand side is the volume of the liquid NP occupied by N Au atoms of Au and N 3  N In  N Ga atoms of the group III elements ( N In atoms of In and N Ga atoms of Ga), with  as the elementary volume in the liquid phase. The right hand side is the geometrical volume of a spherical cap having the base radius R and contact angle  . Clearly, the group III/Au ratio


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 26

equals N 3 / N Au . In our simple model, we consider only axial growth via the VLS mechanism, without subsequent vapor-solid lateral overgrowth on the sidewalls which occurs after the growth direction is decided. We assume that the NP size is pre-determined by its initial group III content (at N Au  const ) and does not change much after switching the group III fluxes to grow a heterostructure. In this case, the NW radius can be treated as time-independent and then any two 0 group III/Au ratios N 3 / N Au and N3 / N Au are related through the corresponding contact angles of

the NPs  and  0 as N3  N 0  f ( )  1  3  1 . N Au  N Au  f (  0 )

(2)

0 Measuring the contact angle  0 for a NP with a known N3 / N Au gives the dependence of the

group III/Au ratio on the contact angle within the entire range of compositions and with no free parameters. From Eq. (2), we can then relate the quantity ( sin   cos  ) to the group III/Au 0 ratio. This dependence is shown by line in Fig. 6 (a) for the measured  0  118o at N 3 / N Au 

1.63 [Fig. 6 (b)]. The curve shows how ( sin   cos  ) increases with the group III/Au ratio, because the contact angle of the NP is larger for higher group III content. Now, the NP stability on the NW top is entirely determined by the surface energy ratio

 SV /  LV (see Eq. (1)). The liquid-vapor energy  LV generally depends on the NP composition and is often taken as a linear interpolation between the surface energies of pure liquids,31 assuming a spatially homogeneous alloy. In our case, it depends on both N In and N Ga , which vary in opposite directions. However, In is the lowest energy liquid metal 32 and, if present in the NP, should accumulate at the surface to minimize the surface energy. Our data confirm a much higher In affinity for Au compared to Ga, which is why its initial concentration in the NPs even


Page 17 of 26

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

before growing InAs on InAs-GaAs NWs is systematically larger than Ga. Therefore, we additionally can assume that switching on the In flux to grow InAs on GaAs quickly replaces Ga to In inside the NP. Both considerations lead to the assumption of a composition-independent

 LV   In for pure In liquid.

Figure 6. (a) Graph of ( sin   cos  ) versus the group III/Au ratio obtained from eqs. (1) and (2) for

0  118o at N 3 / N Au

= 1.63, as shown in (b). Horizontal line separates the domains of straight growth

of InAs on GaAs and kinking. Symbols represent the data given in Table 1 and fall exactly within the correct growth domains. Open circle in (a) shows the point corresponding to the measured contact angle and group III/Au ratio of InAs-GaAs (tGaAs = 5 min) NWs as the one depicted in (b).

Based on the data of Ref. [32], corrected for the average growth temperature of InAs33, the  In value equals 0.459 J/m2. If we take the published value for the lowest energy vertical sidewalls of pure wurtzite InAs NWs,  SV   InAs  0.576 J/m2 (Refs. [34,35]), the horizontal line


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 26

in Figure 6 (a) which represents the  InAs /  In ratio, yields exactly the experimentally observed domains for straight InAs growth at high group III/Au ratios and kinking at low group III/Au ratios. This picture should not be significantly different for InGaAs or even pure GaAs NWs. Indeed, the solid-vapor surface energy of GaAs is only slightly larger (  GaAs  0.693 J/m2 according to Ref. [36]) and can be compensated by simultaneously increasing  LV due to the surface accumulation of some Ga when growing GaAs. Finally, the observed increase of Ga concentration in the NP with In pre-deposition can be understood from the same stationary considerations. Alloying In with Au increases the initial group III content in the droplet. In steady state VLS growth with a fixed droplet volume, In will be replaced by Ga at a fixed N3 in the course of growth and hence the final Ga content will be larger than without In pre-deposition.

CONCLUSIONS We have shown that the optimization of the growth protocols allows for fabrication of straight GaAs-InAs and InAs-GaAs-InAs heterostructures by Au-assisted CBE. The well-known problem of growing InAs on top of GaAs is solved by choosing the conditions that yield high group III content in the Au/In or Au/In/Ga NPs. This can be achieved by either growing short GaAs segments in double heterostructures or pre-depositing In before growth of GaAs in single heterostructures. The observed trend is related to the droplet stability on the NW top, favored at large contact angles and hence high group III concentrations. Our strategy can be used for other material systems and ultimately for designing III-V NW heterostructures with the required properties for different applications. For example, Au-catalyzed III-V NWs grown with high


Page 19 of 26

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

V/III ratios are known to have group III poor NPs with small contact angles37 and therefore are not suitable for obtaining straight heterostructures. In contrast, low V/III ratios yield group III rich NPs with large contact angles37 and should be more advantageous for the straight growth.

EXPERIMENTAL DETAILS The NWs were grown in a Riber Compact-21 CBE system. Prior to mounting the substrates into the growth chamber, a thin Au layer was deposited on the substrates at room temperature in a thermal evaporator. InAs NWs and InAs-GaAs NWs were grown on InAs (111)B substrates coated with a 0.5 nm thick Au layer, whereas GaAs and GaAs-InAs NWs were grown on GaAs (111)B substrates with a 0.1 nm thick Au layer. The samples were introduced in the CBE chamber and thermal annealing was carried out before the growth, keeping the samples at the annealing temperature under the As flux for 20 minutes. We used trimethylindium (TMIn), triethylgallium (TEGa) and tertiarybutylarsine (TBAs) as metalorganic (MO) precursors for NW growth. Details of the growth chamber setup are reported elsewhere.38 The InAs growth was performed at 1.6 Torr and 3.6 Torr line pressures for TMIn and TBAs, respectively. For GaAs, we used 0.7 and 1.1 Torr line pressures for TEGa and TBAs, respectively. The substrate temperature was measured by an infrared pyrometer yielding an accuracy of ±10°C. As explained in the main text and in the Supporting information S1 we developed multi-step growth protocols to grow the NW heterostructures in the two material sequences, since InAs and GaAs have different growth temperature windows. Morphological characterization of the NWs was performed using a Zeiss field-emission SEM operated at 5 keV. Structural characterization was carried out using a Zeiss Libra 120 TEM


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 26

equipped with an in-column  filter. Chemical composition of the NWs as well as NPs on their tips was investigated by EDX performed in STEM mode, using a Bruker XFlash® 6T | 60 SDD detector. For chemical analysis, the samples were quickly cooled down after simultaneously terminating all the precursor fluxes. In this way, the EDX analysis should give a good measure of the NP composition at the end of growth, even if it is performed ex-situ and at room temperature.39 The EDX results (expressed in atomic percent units, with an accuracy of 2-3%) were averaged over 10 to 20 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.

ACKNOWLEDGMENT VGD thanks the Russian Foundation for Basic Research for financial support under grants 1552-78057, 15-52-53041, 16-02-00134, 16-29-03129 and 15-02-06525. We gratefully acknowledge the bilateral project CNR/RFBR.


Page 21 of 26

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

Supporting Information available: S1. Growth protocol optimization for InAs-GaAs nanowire heterostructures. S2. Growth interruption tests.

REFERENCES 1. Björk, M. T.; Ohlsson, B. J.; Thelander, C.; Persson, A. I.; Deppert, K.; Wallenberg, L. R.; Samuelson, L. Nanowire Resonant Tunneling Diodes. Appl. Phys. Lett. 2002, 81, 4458–4460. 2. Nilsson, H. A.; Duty, T.; Abay, S.; Wilson, C.; Wagner, J. B.; Thelander, C.; Delsing, P.; Samuelson, L. A Radio Frequency Single-Electron Transistor Based on an InAs / InP Heterostructure Nanowire. Nano Lett. 2008, 8, 872–875. 3. Rossella, F.; Bertoni, A.; Ercolani, D.; Rontani, M.; Sorba, L.; Beltram, F.; Roddaro, S. Nanoscale Spin Rectifiers Controlled by the Stark Effect. Nat. Nanotechnol. 2014, 9, 997– 1001. 4. Lind, E.; Persson, A. I.; Samuelson, L.; Wernersson, L. E. Improved Subthreshold Slope in an InAs Nanowire Heterostructure Field-Effect Transistor. Nano Lett. 2006, 6, 1842–1846. 5. Borgstrom, M. T.; Zwiller, V.; Muller, E.; Imamoglu, A. Optically Bright Quantum Dots in Single Nanowires. Nano Lett. 2005, 5, 1439–1443. 6. Hiruma, K.; Murakoshi, H.; Yazawa, M.; Katsuyama, T. Self-Organized Growth of GaAs/InAs Heterostructure Nanocylinders by Organometallic Vapor Phase Epitaxy. J. Cryst. Growth 1996, 163, 226–231. 7. Gudiksen, M. S.; Lauhon, L. J.; Wang, J.; Smith, D. C.; Lieber, C. M. Growth of Nanowire Superlattice Structures for Nanoscale Photonics and Electronics. Nature 2002, 415, 617–620.


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 26

8. Fahlvik Svensson, S.; Jeppesen, S.; Thelander, C.; Samuelson, L.; Linke, H.; Dick, K. a. Control and Understanding of Kink Formation in InAs-InP Heterostructure Nanowires. Nanotechnology 2013, 24, 345601. 9. Dick, K.; Kodambaka, S.; Reuter, M. C.; Deppert, K.; Samuelson, L.; Seifert, W.; Wallenberg, L. R.; Ross, F. M. The Morphology of Axial and Branched Nanowire Heterostructures. Nano Lett. 2007, 7, 1817–1822. 10. Dick, K. A; Bolinsson, J.; Borg, B. M.; Johansson, J. Nanowires : Beyond the Reservoir Effect. Nano Lett. 2012, 12, 3200–3206. 11. Dick, K. a. A Review of Nanowire Growth Promoted by Alloys and Non-Alloying Elements with Emphasis on Au-Assisted III-V Nanowires. Prog. Cryst. Growth Charact. Mater. 2008, 54, 138–173. 12. Paladugu, M.; Zou, J.; Guo, Y. N.; Auchterlonie, G. J.; Joyce, H. J.; Gao, Q.; Tan, H. H.; Jagadish, C.; Kim, Y. Novel Growth Phenomena Observed in Axial InAs/GaAs Nanowire Heterostructures. Small 2007, 3, 1873–1877. 13. Paladugu, M.; Zou, J.; Guo, Y. N.; Zhang, X.; Kim, Y.; Joyce, H. J.; Gao, Q.; Tan, H. H.; Jagadish, C. Nature of Heterointerfaces in GaAs/InAs and InAs/GaAs Axial Nanowire Heterostructures. Appl. Phys. Lett. 2008, 93, 91–94. 14. Messing, M. E.; Wong-leung, J.; Joyce, H. J.; Zanolli, Z.; Gao, Q.; Tan, H. H.; Reine, L.; Johansson, J.; Jagadish, C. Growth of Straight InAs on GaAs Nanowire Heterostructures. Nano Lett. 2011, 11, 3899–3905. 15. Borgström, M. T.; Immink, G.; Ketelaars, B.; Algra, R.; Bakkers, E. P. a M. Nat. Nanotechnol. 2007, 2 (9), 541–544.


Page 23 of 26

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

16. Krogstrup, P.; Yamasaki, J.; Sorensen, C. B.; Johnson, E.; Wagner, J. B.; Pennington, R.; Aagesen, M.; Tanaka, N.; Nygard, J. Junctions in Axial III - V Heterostructure Nanowires Obtained via an Interchange of Group III Elements. Nano Lett. 2009, 9, 3689–3693. 17. Ghasemi, M.; Sundman, B.; Fries, S. G.; Johansson, J. The Thermodynamic Assessment of the Au-In-Ga System. J. Alloys Compounds. 2014, 600, 178–185. 18. Xiao-Long, L.; Xia, Z.; Xiao-Long, L.; Xin, Y.; Jan-Gong, C.; Jun-Shuai, L.; Yong-Qing, H.; Xiao-Min, R. Growth and characterization of GaAs/InxGa1-xAs/GaAs axial nanowire heterostructures with symmetrical heterointerfaces. Chin. Phys. B 2013, 22 (6), 066101. 19. Ameruddin, A. S.; Caroff, P.; Tan, H. H.; Jagadish, C.; Dubrovskii, V. G. Understanding the growth and composition evolution of gold-seeded ternary InGaAs nanowires. Nanoscale 2015, 7, 16266-16272. 20. Liu, H. S.; Cui, Y.; Ishida, K.; Jin, Z. P. Thermodynamic Reassessment of the Au-In Binary System. Calphad Comput. Coupling Phase Diagrams Thermochem. 2003, 27, 27–37. 21. Tchernycheva, M.; Travers, L.; Patriarche, G.; Glas, F.; Harmand, J. C.; Cirlin, G. E.; Dubrovskii, V. G. Au-Assisted Molecular Beam Epitaxy of InAs Nanowires: Growth and Theoretical Analysis. J. Appl. Phys. 2007, 102, 094313. 22. Bauer, J.; Gottschalch, V.; Paetzelt, H.; Wagner, G. VLS Growth of GaAs/(InGa)As/GaAs Axial Double-Heterostructure Nanowires by MOVPE. J. Cryst. Growth 2008, 310, 5106– 5110. 23. Dubrovskii, V. G.; Sibirev, N. V.; Harmand, J. C.; Glas, F. Growth Kinetics and Crystal Structure of Semiconductor Nanowires. Phys. Rev. B - Condens. Matter Mater. Phys. 2008, 78, 235301.


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 26

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. Hocevar, M.; Immink, G.; Verheijen, M.; Akopian, N.; Zwiller, V.; Kouwenhoven, L.; Bakkers, E. Nat. Commun. 2012, 3, 1266. 26. Nebol’sin, V. A.; Shchetinin, A. A. Role of Surface Energy in the Vapor–Liquid–Solid Growth of Silicon. Inorg. Mater. 2003, 39, 899-903. 27. Glas, F.; Harmand, J. C.; Patriarche, G. Why does wurtzite form in nanowires of III-V zincblende semiconductors ? Phys. Rev. Lett. 2007, 99, 146101. 28. Dubrovskii, V. G.; Cirlin, G. E.; Sibirev, N. V.; Jabeen, F.; Harmand, J. C. Werner, P. New mode of vapor-liquid-solid nanowire growth. Nano Lett. 2011, 11, 1247-1253. 29. Jacobsson, D.; Panciera, F.; Tersoff, J.; Reuter, M. C. Lehmann, S.; Hofmann, S.; Dick, K. A.; Ross F. M. Interface dynamics and crystal phase switching in GaAs nanowires. Nature 2016, 531, 317-322. 30. Wen, C.-Y.; Tersoff, J.; Hillerich, K.; Reuter, M. C.; Park, J. H.; Kodambaka, S.; Stach, E. A.; Ross F. M. Periodically Changing Morphology of the Growth Interface in Si, Ge, and GaP Nanowires. Phys. Rev. Lett. 2011, 107, 025503. 31. Dubrovskii, V. G. Mono- and polynucleation, atomistic growth, and crystal phase of III-V nanowires under varying group V flow. J. Chem. Phys. 2015, 142, 204702. 32. Jasper, J.J. The surface tension of pure liquid compounds. J. Phys. Chem. Ref. Data 1972, I (4), 841-1009. 33. Dubrovskii, V. G.; Sibirev, N. V.; Cirlin, G. E.; Soshnikov, I. P.; Chen, W. H.; Larde, R.; Cadel, E.; Pareige, P.; Xu, T.; Grandidier, B.; Nys, J.-P.; Stievenard, D.; Moewe, M.;


Page 25 of 26

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

Chuang, L. C.; Chang-Hasnain, C. Gibbs-Thomson and diffusion-induced contributions to the growth rate of Si, InP and GaAs nanowires. Phys. Rev. B 2009, 79, 205316. 34. Hilner, E.; Håkanson, U.; Froberg, L. E.; Karlsson, M.; Kratzer, P.; Lundgren, E.; Samuelson, L.; Mikkelsen, A. Direct Atomic Scale Imaging of III−V Nanowire Surfaces. Nano Lett. 2008, 8, 3978–3982. 35. Joyce, H.J.; Wong-Leung, J.; Gao, Q.; Tan, H. H.; Jagadish, C. Phase Perfection in Zinc Blende and Wurtzite III-V Nanowires Using Basic Growth Parameters. Nano Lett. 2010, 10, 908–915. 36. Rosini, M.; Magri. R. Surface Effects on the Atomic and Electronic Structure of Unpassivated GaAs Nanowires. ACS Nano 2010, 4, 6021–6031. 37. Dheeraj, D. L.; Munshi, a M.; Scheffler, M.; van Helvoort, a T. J.; Weman, H.; Fimland, B. O. Nanotechnology 2013, 24, 015601. 38. Gomes, U. P.; Ercolani, D.; Zannier, V.; Beltram, F.; Sorba, L. Controlling the Diameter Distribution and Density of InAs Nanowires Grown by Au-Assisted Methods. Semicond. Sci. Technol. 2015, 30, 115012. 39. Harmand, J. C.; Patriarche, G.; Ṕere-Laperne, N.; Ḿerat-Combes, M. N.; Travers, L.; Glas, F. Analysis of Vapor-Liquid-Solid Mechanism in Au-Assisted GaAs Nanowire Growth. Appl. Phys. Lett. 2005, 87, 1–3.


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 26

Graphical table of contents:

The growth mode of GaAs-InAs NW heterostructures depends on the nanoparticle stability. This is strictly related to the chemical composition (group III/Au ratio) of the alloy nanoparticle which can be finely tuned by the growth parameters.


Catalyst Composition Tuning: The Key for the ... - ACS Publications

Recommend Documents