Influence of Silicon on the Nucleation Rate of GaAs Nanowires on

Subscriber access provided by UNIV OF DURHAM

C: Physical Processes in Nanomaterials and Nanostructures

Influence of Silicon on the Nucleation Rate of GaAs Nanowires on Silicon Substrates Hadi Hijazi, Vladimir G. Dubrovskii, Guillaume Monier, Evelyne Gil, Christine Leroux, Geoffrey Avit, Agnès Trassoudaine, Catherine Bougerol, Dominique Castelluci, Christine Robert-Goumet, and Yamina André J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.8b05459 • Publication Date (Web): 24 Jul 2018 Downloaded from http://pubs.acs.org on July 30, 2018

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

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

Page 1 of 24 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

The Journal of Physical Chemistry

Influence of Silicon on the Nucleation Rate of GaAs Nanowires on Silicon Substrates Hadi Hijazi †,*, Vladimir G. Dubrovskii □,*, Guillaume Monier †, Evelyne Gil †,□, Christine Leroux‡ , Geoffrey Avit †, Agnès Trassoudaine †, Catherine Bougerol ◊, Dominique Castellucci †, Christine Robert-Goumet †,Yamina André †,□ †

Université Clermont Auvergne, CNRS, SIGMA Clermont, Institut Pascal, F-63000 ClermontFerrand, France. □ ITMO University, Kronverkskiy pr. 49, 197101 St. Petersburg, Russia ‡ Université du Sud Toulon-Var, IM2NP, Bât.R, B.P.20132, 83957, La Garde Cedex, France. ◊ Univ. Grenoble Alpes, CNRS, Institut Néel, 38000 Grenoble France.

KEYWORDS. HVPE Growth, GaAs Nanowires, Silicon Substrate, Silicon Dioxide, Chemical Potential, Nucleation Rate.

ABSTRACT

Despite the unavoidable presence of silicon atoms in the catalyst alloy droplets during the vaporliquid-solid growth of III-V nanowires on silicon substrates, it remains unknown how the nucleation of nanowires is affected by these foreign atoms. In this work, we present the first attempt to quantify the nanowire nucleation rate versus the silicon concentration in the droplet. We calculate the chemical potential difference per Ga-As pair in the quaternary Au-Ga-As-Si liquid alloy droplet and in solid, and compare it to the ternary Au-Ga-As droplet without silicon. This allows us to compute the nucleation rates of GaAs nanowires versus the silicon concentration under different conditions. We find that the presence of silicon in the droplet decreases the nucleation probability of GaAs nanowires for gallium-rich droplets (with the gallium contents greater than 0.6) and increases it for gold-rich droplets ( < 0.6). The

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 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 24

model is used to explain our experimental data for hydride vapor phase epitaxy of gold-catalyzed GaAs nanowires, which easily nucleate on Si(111) covered with different SiO2 layers but do not grow on the bare Si(111). In the latter case, more silicon is etched from the substrate and enters the gallium-rich droplets, which suppresses the nanowire nucleation. We discuss other relevant data, including the known difficulties in obtaining self-assisted GaAs NWs on silicon by chemical epitaxy techniques. These results may be useful for the fine tuning of III-V nanowire properties and integrating them with silicon electronics.

INTRODUCTION

III-V semiconductor nanowires (NWs) grown by the vapor-liquid-solid (VLS) method1 are promising for fundamental studies as well as applications in electronic2, photovoltaic3, photonic4, thermoelectric5, and sensing6 devices. The VLS growth can be promoted by either a foreign metal catalyst, often gold1-3,7-12, or else a group III metal constituting the NW itself (gallium in the case of GaAs NWs) in the self-assisted (or-self-catalyzed) approach13-17. For GaAs, one of the most important among the entire range of III-V materials, gold-catalyzed GaAs NWs have been grown on GaAs substrates by different techniques including molecular beam epitaxy (MBE) 7-9, metal-organic vapor phase epitaxy (MOVPE) 2,3,10 and hydride vapor phase epitaxy (HVPE)

11,12

. Each of these techniques operates only within a certain temperature

domain. For example, gold-catalyzed GaAs NWs on GaAs substrates by MBE are grown at temperatures from 400 to 620 °C (Ref. [9]), while gold-catalyzed HVPE GaAs NWs are grown at temperatures higher than 700 oC (Refs. [11,12]). In this wide range of temperatures, the atomic interdiffusion between the substrate and the catalyst always serves as an additional supply of gallium and should be favorable for the NW nucleation on GaAs substrates7,9,18.


2

Page 3 of 24 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


The picture becomes different when the VLS growth of GaAs NWs takes place on silicon substrates covered/or not with an oxide layer, which is paramount for monolithic integration of III-V photonics with silicon electronic platform. It has been shown that the presence of an oxide layer at the interface between gold and silicon substrate blocks the interexchange between the two materials19. Therefore, the interdiffusion of silicon atoms into the gold alloy droplets occurs only after the decomposition of SiO2 underneath the droplets, which has been long known for temperatures above 400 °C20. The Si(s) +SiO2(s) 2SiO (g) reaction is enhanced in the presence of the gold atoms that reach the SiO2/Si interface by thermally activated diffusion into SiO221,22. Obviously, this diffusion is easier through thinner SiO2 layers. A similar enhancement is anticipated when the VLS growth on silicon is assisted by the gallium droplets14-17. Additionally, the SiO2 layer can be decomposed at high enough temperatures via the reaction with gallium23. A fraction of the gallium droplet is evaporated as Ga2O and the remaining part pumps the silicon atoms away from the substrate, leading to enhanced silicon evaporation. Therefore, in both gold-catalyzed and self-assisted VLS growth of GaAs NWs on silicon substrates, the diffusion or re-adsorption of the silicon atoms into the droplet prior to nucleation of GaAs NWs is unavoidable. The strength of this effect may vary depending on the deposition conditions and is anticipated to be larger for thinner oxide layers. Also, silicon should diffuse into the catalyst droplets more easily at higher temperatures and in chemical epitaxy techniques where different precursors are able to etch out some silicon from the substrate. Consequently, here we present an approach to quantify how the presence of silicon atoms in the catalyst droplet affects the nucleation rate of GaAs NWs on silicon substrates. We then use the model to interpret the data on gold-catalyzed HVPE growth of GaAs NWs on silicon. Our approach is general as it treats chemical potentials in a quaternary Au-III-V-Si liquid alloy over a wide range of


3


Page 4 of 24

temperatures and irrespective of a particular epitaxy technique employed. Therefore, any VLS III-V NWs grown on silicon with different catalysts (gold, gallium, and indium) may be treated using the proposed scheme. Silicon is widely used as dopant for GaAs NWs. Here, we only consider the influence of the Si concentration in the droplet on the nucleation probability of pure GaAs NWs on different SiOx/Si substrates, that is, not allowing for Si to enter the solid state. The growth of Si-doped GaAs NWs and ultimately a ternary Si-GaAs solid alloy will be studied elsewhere.

MODEL We will compare two VLS systems, (I) quaternary Au-Ga-As-Si liquid droplet and (II) ternary Au-Ga-As droplet without silicon, both producing the first monolayer of GaAs NW on a silicon substrate. Assuming that silicon is added to an initially ternary Au-Ga-As droplet at the fixed total numbers of the gold, gallium, and arsenic atoms N Au = const , N Ga = const . N As = const , we first compare the atomic concentrations in the initial droplet without silicon, ci = N i /( N Au + N Ga + N As ) for i = Au, Ga, As and in the droplet with NSi Si atoms, c~i = N i /( N Au + N Ga + N As + N Si ) including Si. It is easy to find that in this case c~i = ci (1 − cSi ) for i = Ga, As (and Au), with c Si = N Si /( N Au + N Ga + N As + N Si ) . We will now consider how adding a given number of the silicon atoms changes the chemical potential value of GaAs in the droplet having the initial composition cGa and cAs . In our approach, we consider Si influencing the initial nucleation step of GaAs NWs. This is supported by the data of Ref. [24], where it has been shown that Si starts to incorporate into solid NWs after the NW growth is being initiated.


4

Page 5 of 24 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


Regular solution model25-28 for chemical potentials of the gallium and arsenic atoms in a quaternary Ga-As-Au-Si liquid is given by 0 2 µGa = µGa + kBT ln cGa + ωGaAscAs + ωGaSicSi2 + ωGaAu(1 − cGa − cAs − cSi )2 + (ωGaSi + ωGaAs − ωAsSi)cAscSi + (ωGaSi + ωGaAu − ωAuSi)cSi (1 − cGa − cAs − cSi ) + (ωGaAs + ωGaAu − ωAsAu)cAs (1 − cGa − cAs − cSi ) ,

0 2 µ As = µAs + kBT ln cAs + ωGaAscGa + ωAsSicSi2 + ωAsAu(1 − cGa − cAs − cSi )2 + (ωAsSi + ωGaAs − ωGaSi)cGacSi + (ωAsSi + ωAsAu − ωAuSi)cSi (1 − cGa − cAs − cSi ) + (ωGaAs + ωAsAu − ωGaAu)cGa (1 − cGa − cAs − cSi ) .

(1) Here, µ i0 are the reference chemical potentials in pure liquids, ωik are the binary interaction constants between atoms i and k , T is the absolute temperature and k B is the Boltzmann constant. We omit the higher order interactions28,29 as less essential for our analysis. The chemical potential difference per GaAs pair in the droplet with and without Si is given by f = ∆µ Si − ∆µ 0 .

(2)

Here ∆µ Si and ∆µ 0 are, respectively, the difference of chemical potentials with and without Si atoms in the droplet. Assuming that the solid state is (almost) pure GaAs without silicon, the same equation is valid for the chemical potential difference ∆µ with respect to the solid state (with chemical potential

µGaAs ) , that is, f = µGa [cGa (1 − cSi ), c As (1 − cSi ), cSi ] + µ As [cGa (1 − cSi ), c As (1 − cSi ), cSi ] − µGa (cGa , c As , cSi = 0) − µ As (cGa , c As , cSi = 0) After simple calculations, the f can be obtained from Eqs. (1) and (2) in the form f = 2 k BT ln(1 − cSi ) + η cSi + κ c Si2 .

(3)

The coefficients η and κ are the second order polynomials of cGa and cAs given by


5


Page 6 of 24

η = ωGaSi + ωAsSi − ωGaAu − ωAsAu − 2ωAuSi + (5ωGaAu + ωAsAu + 2ωAuSi − 2ωGaSi − ωGaAs )cGa +(5ωAsAu + ωGaAu + 2ωAuSi − 2ωAsSi − ωGaAs )cAs + 4(ωGaAs − ωGaAu − ωAsAu )cGa cAs , 2 2 −4ωGaAu cGa − 4ωAsAu cAs

κ = 2ωAuSi + 2(ωGaSi − ωGaAu − ωAuSi )cGa + 2(ωAsSi − ωAsAu − ωAuSi )cAs 2 2 +2(ωGaAu + ωAsAu − ωGaAs )cGa cAs + 2ωGaAu cGa + 2ωAsAu cAs

.

(4)

For small enough c Si , we can expand the logarithmic function in Eq. (3), yielding f ≅ Ac Si + Bc Si2 ,

with A = η − 2k BT and B = κ − k BT . In the following, we will use the composition-independent interaction constants ωik tabulated for different temperatures in Refs. [28-30] for different binaries. This is different from Ref. [25] where the ωik were taken inversely proportional to the sum Vi ci + Vk c k according to Ref. [31], with V as the elementary volume of the corresponding element in liquid. For the composition-independent interaction constant, the f function is approximately parabolic and its sign at small c Si is determined by the sign of A . Negative A correspond to lower chemical potentials and hence more difficult nucleation of GaAs NWs from the droplets with silicon. Note that both A and B depend on the droplet composition, most importantly, the gallium content cGa , so the nucleation probabilities of GaAs NWs on silicon should be highly sensitive to whether the alloy is gold or gallium-rich. According to Ref. [32], the nucleation probability for each NW monolayer in the mononuclear vapor-liquid-solid growth, including the very first GaAs monolayer on the substrate, is given by p = πR 2 J , with R as the droplet base radius and J the Zeldovich nucleation rate given by J=

2

 a2  h . DAs c As   e µGaAs ∆µ 1 / 2 exp a − 4∆µ  π Ω 

33 / 4

(5)


6

Page 7 of 24 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


Here, D As is the diffusion coefficient of As in liquid, h =0.326 nm in the monolayer height of GaAs, Ω = 0.0452 nm3 is its elementary volume in solid, µ denotes chemical potentials in the units of k BT and a = 2 × 33 / 4

γ k BT

(Ωh)1 / 2

(6)

is the energetic constant related to the effective surface (or edge) energy γ of two-dimensional island with regular triangle shape. At T =715 oC, this constant equals 40.5 × γ if γ is measured in J/m2 and therefore the a 2 / 4 value under the exponent of the Zeldovich nucleation rate is very large, typically on the order of several tens33. Analyzing Eqs. (5) and (6), it is reasonable to assume that the island surface energy remains the same at low enough silicon concentrations in the droplet. This is also supported by the fact that gallium is the lowest surface energy element among the droplet constituencies and hence should segregate at the droplet surface. In this case, the most important part of the ratio between the nucleation probabilities of GaAs from the droplets with and without Si is given by an extremely steep exponential term exp[ a 2 / 4∆µ ] , while all the unknown terms cancel in the ratio. Therefore, we can write

 p Si a2 f    = exp α= p0 4 ∆ µ ∆ µ k T 0 Si B  

(7)

where f is the chemical potential difference defined in Eq. (3). This can be further simplified to

=

 p Si f   , ≅ exp ic p0 k T B  

(8)

with


7


a2 ic = 4∆µ02

Page 8 of 24

(9)

as the critical size of classical nucleation theory [32] for the droplet without silicon. Here, we account only for the differences in the chemical potential value by adding Si to the liquid droplet. It should be noted that dissolving more Si also increases the contact angle of the droplet resting on the substrate surface. This effect may additionally decrease the nucleation rate and ultimately suppress the NW growth according to Ref. [14].

RESULTS AND DISCUSSION Figure 1 (a) shows the evolution of as a function of for the lowest and highest growth temperatures of GaAs NWs, 450 oC and 715

o

C (as in our HVPE growth experiments) and

different ranging from 0.5 to 0.99. The curves reveal that the chemical potential difference

in the droplets with and without silicon depends drastically on for all in the entire range of temperatures. Although the temperature domain is very wide, the curves at 450 oC and 715 oC are not so different and follow the same trends. For gallium-rich droplets with cGa ≥ 0.6, the chemical potential in the droplets with silicon is lower than without it for low enough c Si , with the difference becoming larger toward higher cGa . Therefore, adding silicon to gallium-rich AuGa droplets makes the initial nucleation of GaAs NWs more difficult, and this effect is stronger for more gallium-rich droplets. The limiting case of gallium-assisted VLS growth ( cGa → 1 ) is the most sensitive to the presence of silicon even for its very small fractions in the droplet. The chemical potential difference becomes positive after excessing a certain critical c Si , which depends upon cGa . This critical silicon concentration is greater than 0.35 for cGa = 0.99 and about


8

Page 9 of 24 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


0.1 for cGa = 0.6 at any temperature. For gold-rich droplets ( cGa < 0.6), the chemical potential difference in systems with and without silicon becomes monotonically increasing with c Si , meaning that adding any amount of silicon to such droplets only helps to nucleate GaAs NWs. Figure 1 (b) shows the dependence of on for different and at a fixed temperature of 715 oC. As expected, the curves are not very sensitive to c As (in the plausible range between 0.005 and 0.06) due to the smallness of the interaction terms with arsenic.


9


Page 10 of 24

Figure 1. The chemical potential difference per Ga-As pair in quaternary Au-Ga-As-Si liquid alloy versus ternary Au-Ga-As alloy as a function of the silicon concentration for (a) different

from 0.5 to 0.99, with the arsenic concentration fixed at 0.01, at T = 450 oC (dashed lines) and 715 oC (solid lines), and (b) different and at T=715 °C.

We now consider our experimental data on the gold-catalyzed HVPE GaAs NWs grown at 715 o

C on Si(111) substrates using the above model. Figure 2 depicts the typical behavior when the

silicon surface is covered with 500 nm of thermal SiO2 layer (S1) or with a native SiO2 (S2) [Figure 2 (a) and 2 (b), respectively]. We observe the simultaneous formation of almost cylindrical NWs and pyramidal-like “scales”, while we can only see fewer but larger scales and some parasitic islands on the bare Si(111) after HF etching of the oxide layer (S3) [Figure 2 (c)]. According to the analysis of Ref. [12], pure zincblende crystal structures of HVPE GaAs NWs requires gallium-rich Au-Ga-Si droplets, with cGa definitely higher than 0.65. According to Figures 1 (a), such droplets should produce fewer GaAs NWs for larger silicon concentrations. As mentioned above, more silicon should arrive to the droplets resting on the bare Si(111) substrates than on Si(111) covered by an oxide layer, because the latter suppresses the silicon etching. Ignoring the influence of highly volatile arsenic, the data of Refs. [34, 35] give a silicon solubility of about 0.3 in liquid gold and 0.05 in liquid gallium. Therefore, for any Au-Ga alloy, we should not anticipate to have more than 0.3 of silicon dissolved in the droplets. Looking again


10

Page 11 of 24 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


at the curves of Figure 1 (b), the condition c Si < 0.3 at cGa > 0.65 guarantees that the presence of silicon suppresses nucleation of GaAs NWs on Si(111), while they perfectly grow under the same conditions on GaAs substrates12. Therefore, our model is very well consistent with the experimental observations. (c)

(b)

(a)

20 µm

20 µm

20 µm

Figure 2. Scanning electron microscopy (SEM) images of the GaAs structures grown at 715 oC after depositing 1 nm of gold on Si(111) (a) with 500 nm of thermal SiO2, (b) native SiO2 and (c) without SiO2. More quantitatively, Figure 3 shows high resolution transmission electron microscopy (HRTEM) images and the corresponding energy dispersive spectroscopy (EDS) compositional data given in Table 1 for the catalyst droplets of GaAs structures grown on S1, S2, and S3 substrates at 715 o

C and under otherwise identical conditions. Of course, the EDS data are obtained for the

droplets after growth, which have lost some of their gallium atoms in the cooling down stage under the arsenic flux. This process leads to a droplet shrinking and the corresponding diminishing of the NW top radius16. Assuming that the number of gold and silicon atoms present in the droplet during growth remains the same after growth, we can roughly estimate the and

from geometry by comparing the initial volume of the droplet with gallium and its final volume without gallium with the measured initial and final radii of the NW top [see Figure 3 (d)]. These estimates yield the values of silicon concentrations given in Table 1. In terms of , the calculations show the values close to 0.95 for all samples [we reiterate that the influence of arsenic on the shapes of the chemical potential curves is weak according to Figure 1 (b)].


11


Page 12 of 24

Figure 3. HRTEM images of the gold-catalyzed GaAs structures with droplets on top, grown at 715 oC on different substrates – (a) S1, (b) S2, and (c) S3. The technique used to estimate the concentrations of gallium and silicon during growth is represented schematically in (d), with =130o and β f =95o reported in Ref. [16].

Table 1. EDS compositional data for the silicon concentrations in the droplets catalyzing GaAs structures on S1, S2 and S3 substrates after growth, with the corresponding estimations of during growth

Sample

after growth

during growth

S1

0.02

0.001

S2

0.05

0.003

S3

0.21

0.031

Assuming that the critical size ic in Eq. (9) equals six GaAs pairs, and using the chemical potential curve at a fixed cGa of 0.95, the surface energy γ equals 0.178 J/m2 (corresponding to a = 7.21). This is close to the estimates of Refs. [17] and [36] for purely gallium-assisted and


12

Page 13 of 24 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


gold-catalyzed growths of GaAs NWs, respectively. With this surface energy, the ratio of nucleation probabilities for GaAs NWs emerging from the droplets with and without silicon is plotted in Figure 4 and gives the excellent quantitative correlation with the data. Indeed, we can see that adding a small amount of silicon to the droplets (0.001 and 0.003 in samples S1 and S2) does not have any significant impact on the nucleation probability ( α ≅ 1). On the other hand, the nucleation probability decreases by three times when the silicon concentration is increased to ~ 0.03, as in sample S3. Therefore, the droplets on the bare Si(111) do not produce any NWs and instead inflate with the arriving gallium, which makes the NW nucleation even less probable. This explains why almost no regular NWs are seen in sample S3, but only large scales catalyzed by the swelling droplets.

Figure 4. Ratio of nucleation probabilities of GaAs structures at 715°C with and without silicon at a fixed cGa of 0.95. The labels show the positions of the three samples (S1, S2 and S3) on the curve corresponding to their during growth estimated in Table 1. The obtained results call for a more general discussion. It is well known that galliumassisted growth of GaAs NWs on silicon is extremely difficult to achieve in chemical epitaxy techniques such as MOVPE and HVPE. For example, our own results37 reveal some gallium-


13


Page 14 of 24

catalyzed HVPE GaAs NWs on silicon, but with very low density which should be due to difficult nucleation in the presence of silicon according to our Figure 1. Similar difficulties have been described for gallium-assisted GaAs NWs on silicon in MOVPE technique38,39, where the rate of silicon etching from the substrate under the gallium droplets is enhanced by the presence of Trimethylgallium (TMGa). On the other hand, much less etching should be present during MBE, which is why the silicon concentration is lower and gallium-catalyzed GaAs NWs can easily be grown on silicon by this technique14-16. Our calculated results reveal that having more gold in the droplets helps to nucleate GaAs NWs on silicon. This property has been long known for MOVPE (see, for example, Ref. [40, 41]). Very high concentrations of silicon dissolved in either pure gallium droplet or an Au-Ga alloy may lead to the formation of ternary GaAsSi or even pure silicon NWs. We have observed such NWs in some samples grown on the bare Si(111). This requires a generalization of the model to enable silicon crystallization in the solid state.

CONCLUSION In conclusion, we have developed a model and presented the supporting experimental data showing that adding silicon to the Au-Ga-As droplets has a drastic effect on chemical potentials and the nucleation probabilities of VLS GaAs NWs on silicon substrates. Most importantly, silicon suppresses nucleation from the gallium-rich droplets as long as a certain critical concentration is not reached, with the most difficult nucleation of GaAs NWs occurring in the gold-free self-assisted VLS growth. Adding silicon to the gold-rich droplets only enhances the NW nucleation. We speculate that more silicon should enter the droplets in MOVPE and HVPE due to enhanced silicon etching by chemical precursors. This explains the known problem of


14

Page 15 of 24 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


gallium-catalyzed growth of GaAs NWs by these techniques. It has been demonstrated that HVPE growth of GaAs NWs always occurs from very gallium-rich droplets due to a high material input of gallium. An oxide layer on Si(111) is then required which blocks out the silicon diffusion into the droplet and helps to start the NW growth. Overall, our model correlates with prior experimental findings and may be further used to understand and control the VLS growth of other III-V NWs on silicon, silicon doping of III-V NWs, and even hybrid III-V-Si NWs.

AUTHOR INFORMATION

Corresponding Authors * Hadi Hijazi, Université Clermont Auvergne, CNRS, SIGMA Clermont, Institut Pascal, F63000 Clermont-Ferrand, France. E-mail: [email protected], [email protected] * Vladimir G. Dubrovskii, ITMO University, Kronverkskiy pr. 49, 197101 St. Petersburg, Russia. Email: [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 This work has been supported financially by the CPER MMASYF of Region Auvergne-Rhone Alpes that we acknowledge gratefully. This work was also funded by the program "Investissements d'avenir" of the French ANR agency, the French governement IDEX-SITE initiative 16-µIDEX-0001 (CAP20-25), the European Commission (Auvergne FEDER Funds) and the Region Auvergne in the framework of the LabEx IMobS3 (ANR-10-LABX-16-01) and


15


Page 16 of 24

CPER. We would like also to thank Mr. Pierre Perreau from CEA-LETI, Grenoble, France, for offering Si substrates. VGD thanks the Ministry of Education and Science of the Russian Federation for financial support under grant 14.587.21.0040 (project ID RFMEFI58717X0040). We

thank

Jonas

Johansson

(Lund

University,

Sweden)

for

help

with

the

interaction constants for some binaries.

REFERENCES (1) Wagner, R. S.; Ellis, W. C. Vapor‐Liquid‐Solid Mechanism of Single Crystal Growth. Appl. Phys. Lett. 1964, 4, 89-90. (2) Bryllert, T.; Wernersson, L. E.; Lowgren, T.; Samuelson, L. Vertical Wrap-Gated Nanowire Transistors. Nanotechnology 2006, 17, S227-S230. (3) Wallentin, J.; Anttu, N.; Asoli, D.; Huffman, M.; Åberg, I.; Magnusson, M. H.; Siefer, G.; Fuss-Kailuweit, P.; Dimroth, F.; Witzigmann, B. et al. InP Nanowire Array Solar Cells Achieving 13.8% Efficiency by Exceeding the Ray Optics Limit. Science 2013, 339, 1057-1060. (4) 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, 451, 617-620. (5) Ali, A.; Chen, Y.; Vasiraju, V.; Vaddiraju, S. Nanowire-Based Thermoelectrics. Nanotechnology 2017, 28, 282001. (6) Yang, P.; Yan, R. Fardy, M. Semiconductor Nanowire: What’s Next? Nano Lett. 2010, 10(5), 1529–1536. (7) Dubrovskii, V. G.; Cirlin, G. E.; Soshnikov, I. P.; Tonkikh, A. A.; Sibirev, N. V.; Samsonenko, Yu. B.; Ustinov, V. M. Diffusion-Induced Growth of GaAs Nanowhiskers: Theory and Experiment. Phys.Rev. B 2005, 71, 205325. (8) Harmand, J. C.; Patriarche, G.; Péré-Laperne, N.; Mérat-Combes, M. N.; Travers, L.; Glas, F. Analysis of Vapor-Liquid-Solid Mechanism in Au-Assisted GaAs Nanowire Growth. Appl. Phys. Lett. 2005, 87, 203101.


16

Page 17 of 24 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


(9) Harmand, J. C.; Tchernycheva, M.; Patriarche, G.; Traversa, L. ; Glas, F.; Cirlin, G. GaAs Nanowires Formed by Au-Assisted Molecular Beam Epitaxy: Effect of Growth Temperature. J. Cryst. Growth 2007, 301–302, 853–856. (10) Seifert, W.; Borgstrom, M.; Deppert, K.; Dick, K. A.; Johansson, J.; Larsson, M. W.; Martensson, T.; Skold, N.; Svensson, C. P. T.; Wacaser, D. A. et al. Growth of One-Dimensional Nanostructures in MOVPE. J. Cryst. Growth 2004, 272, 211-220. (11) Ramdani, M. R.; Gil, E.; Leroux, Ch.; André, Y.; Trassoudaine, A. ; Castelluci, D.; Bideux, L.; Monier, G.; Robert-Goumet, C.; Kupka, R. Fast Growth Synthesis of GaAs Nanowires with Exceptional Length. Nano Lett. 2010, 10 (5), 1836-1841. (12) Gil, E.; Dubrovskii, V. G. ; Avit, G. ; André, Y. ; Leroux, C. ; Lekhal, K. ; Grecenkov, J. ; Trassoudaine, A. ; Castelluci, D. ; Monier, G. et al. Record Pure Zincblende Phase in GaAs Nanowires Down to 5 nm in Radius. Nano Lett. 2014, 14 (7), 3938−3944. (13) Colombo, C.; Spirkoska, D.; Frimmer, M.; Abstreiter, G.; Fontcuberta i Morral, A. GaAssisted Catalyst-Free Growth Mechanism of GaAs Nanowires by Molecular Beam Epitaxy. Phys. Rev. B 2008, 77, 155326. (14) Matteini, F.; Dubrovskii, V. G.; Rüffer, D.; Tütüncüoğlu, G.; Fontana, Y.; Fontcuberta i Morral, A. Tailoring the Diameter and Density of Self-Catalyzed GaAs Nanowires on Silicon. Nanotechnology 2015, 26, 105603. (15) Dubrovskii, V. G.; Xu, T.; Álvarez, A. D.; Plissard, S. R.; Caroff, P.; Glas, F. Grandidier, B. Self-Equilibration of the Diameter of Ga-Catalyzed GaAs Nanowires. Nano Lett. 2015, 15 (8), 5580–5584. (16) Kim, W.; Dubrovskii, V. G.; Vukajlovic-Plestina, J.; Tütüncüoglu, G.; Francaviglia, L.; Güniat, L.; Potts, H.; Friedl, M.; Leran, J. B.; Fontcuberta i Morral, A. Bistability of Contact Angle and its Role in Achieving Quantum-Thin Self-Assisted GaAs Nanowires. Nano Lett.

2018, 18, 49–57. (17) Glas, F.; Ramdani, M. R.; Patriarche, G.; Harmand, J. C. Predictive Modeling of SelfCatalyzed III-V Nanowire Growth. Phys. Rev. B 2013, 88, 195304. (18) André, Y.; Lekhal, K.; Hoggan, P.; Avit, G.; Cadiz, F.; Rowe, A.; Paget, D.; Petit, E.; Leroux, C.; Trassoudaine, A. et al. Vapor-liquid-Solid Hydride Vapor Phase Epitaxy (VLSHVPE) Growth of Ultra-Long Defect Free GaAs Nanowires: AB Initio Simulations Supporting Center Nucleation. J. Chem. Phys. 2014, 140, 194706.


17


Page 18 of 24

(19) Rath, A.; Dash, J. K.; Juluri, R. R. ; Rosenauer, A.; Schoewalter, M.; Satyam, P. V. Growth of Oriented Au Nanostructures: Role of Oxide at the Interface. J. Appl. Phys. 2012, 111, 064322. (20) Hiraki, A.; Lugujjo, E.; Mayer, J. W. Formation of Silicon Oxide over Gold Layers on Silicon Substrates. J. Appl. Phys. 1972, 43, 3643-3649. (21) Dallaporta, H.; Liehr, M.; Lewis, J. E. Silicon Dioxide Defects Induced by Metal Impurities. Phys. Rev. B 1990, 41, 5075-5083. (22) Leroy, F.; Passanante, T.; Cheynis, F.; Curiotto, S.; Bussmann, E. B. and Müller P. Catalytically Enhanced Thermal Decomposition of Chemically Grown Silicon Oxide Layers on Si(001). Appl. Phys. Lett. 2016, 108, 111601. (23) Tauchnitz, T.; Nurmamytov, T. ; Hubner, R.; Engler, M.; Facsko, S.; Schneider, H.; Helm, M.; Dimakis, E. Decoupling the Two Roles of Ga Droplets in the Self-Catalyzed Growth of GaAs Nanowires on SiOx/Si(111) Substrates. Cryst. Growth Des. 2017, 17 (10), 5276−5282. (24) Dufouleur, J.; Colombo, C.; Garma, T.; Ketterer, B.; Uccelli, E.; Nicotra, M.; Fontcuberta i Morral, A. Nano Lett. 2010, 10 (5), pp 1734–1740. (25) Glas, F. Chemical Potentials for Au-Assisted Vapor-Liquid-Solid Growth of III-V Nanowires. J. Appl. Phys. 2010, 108, 073506. (26) Dubrovskii, V. G. Refinement of Nucleation Theory for Vapor-Liquid-Solid Nanowires. Cryst. Growth Des. 2017, 17, 2544–2548. (27) Dubrovskii, V. G.; Koryakin, A. A.; Sibirev, N. V. Understanding the Composition of Ternary III-V Nanowires and Axial Nanowire Heterostructures in Nucleation-Limited Regime. Mat. Design 2017, 132, 400–408. (28) Leshchenko, T. D.; Ghasemi, M.; Dubrovskii, V. G.; Johansson, J. Nucleation-Limited Composition of Ternary III-V Nanowires Forming from Quaternary Gold Based Liquid Alloys. CrystEngComm. 2018, 20, 1649 – 1655. (29) Chevalier, P. Y. A Thermodynamic Evaluation of the Au-Ge and Au-Si Systems. Thermochimica Acta 1989, 141, 217-226. (30) Mostafa, A.; Medraj, M. Binary phase diagrams and thermodynamic properties of silicon and essential doping elements (Al, As, B, Bi, Ga, In, N, P, Sb and Tl), Materials 2017, 10, 676. (31) Stringfellow, G. B. Calculation of Ternary Phase Diagrams of III-V Systems. J. Phys. Chem. Solids 1972, 33, 665-677.


18

Page 19 of 24 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


(32) Dubrovskii, V. G.; Grecenkov, J. Zeldovich Nucleation Rate, Self-Consistency Renormalization, and Crystal Phase of Au-Catalyzed GaAs Nanowires. Cryst. Growth Des.

2015, 15, 340–347. (33) Dubrovskii, V. G. Theory of VLS Growth of Compound Semiconductors, In: A. Fontcuberta i Morral, S. A. Dayeh and C. Jagadish, editors, Semiconductors and Semimetals, Burlington: Academic Press 2015, 93, 1-78. (34) Okamoto, H.; Massalski, T. B. The Au-Si (Gold-Silicon) System. Bull. Alloy Phase Diagrams 1983, 4, No. 2. (35) Olesinski, R. W.; Kanani, N.; Abbaschian, G. J. Ga-Si (Gallium-Silicon) System. Bull. Alloy Phase Diagrams 1985, 6, 362-364. (36) 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. (37) Dong, Z.; André, Y.; Dubrovskii, V. G.; Bougerol, C.; Leroux, C.; Ramdani, M R.; Monier, G.; Trassoudaine, A.; Castelluci, D.; Gil, E. Self-Catalyzed GaAs Nanowires on Silicon by Hydride Vapor Phase Epitaxy. Nanotechnology 2017, 28, 125602. (38) Dadgar, A.; Poschenrieder, M.; Blasing, J.; Contreras, O.; Bertram, F.; Riemann, T.; Reiher, A.; Kunze, Daumiller, M. I.; Krtschil, A.; et al. MOVPE Growth of GaN on Si(1 1 1) Substrates. J. Crys. Growth 2003, 248, 556–562. (39) Werner, K.; Beyer, A.; Oelerich, J. O.; Baranovskii, S. D.; Stolz, W.; Volz, K. Structural Characteristics of Gallium Metal Deposited on Si(001) by MOCVD. J. Cryst. Growth 2014, 405, 102–109. (40) Khorenko, V.; Regolin, I.; Neumann, S.; Prost, W.; Tegude, F. J. Photoluminescence of GaAs Nanowhiskers Grown on Si Substrate. Appl. Phys. Lett. 2004, 85, 6407-6408. (41) Chuang, L. C.; Moewe, M.; Crankshaw, S.; Chase, C.; Kobayashi, N. P.; Chang-Hasnain, C. Critical Diameter for III-V Nanowires Grown on Lattice-Mismatched Substrates. Appl. Phys. Lett. 2007, 90, 043115.


19


Page 20 of 24

Table of Content Graphic:


20

Page 21 of 24 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


Figure 1: The chemical potential difference per Ga-As pair in quaternary Au-Ga-As-Si liquid alloy versus ternary Au-Ga-As alloy as a function of the silicon concentration c_Si for (a) different c_Ga from 0.5 to 0.99, with the arsenic concentration c_As fixed at 0.01, at T= 450 °C (dashed lines) and 715 °C (solid lines), and (b) different c_Ga and c_As at T=715 °C. 82x39mm (300 x 300 DPI)



Figure 2: Scanning electron microscopy (SEM) images of the GaAs structures grown at 715 °C after depositing 1 nm of gold on Si(111) (a) with 500 nm of thermal SiO2, (b) native SiO2 and (c) without SiO2. 621x140mm (150 x 150 DPI)


Page 22 of 24

Page 23 of 24 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


Figure 3: HRTEM images of the gold-catalyzed GaAs structures with droplets on top, grown at 715 °C on different substrates – (a) S1, (b) S2, and (c) S3. The technique used to estimate the concentrations of gallium and silicon during growth is represented schematically in (d), with β_i=130° and β_f =95° reported in Ref. [16]. 194x191mm (120 x 120 DPI)



Figure 4: Ratio of nucleation probabilities of GaAs structures at 715°C with and without silicon at a fixed c_Ga of 0.95. The labels show the positions of the three samples (S1, S2 and S3) on the curve corresponding to their c_Si during growth estimated in Table 1. 254x190mm (96 x 96 DPI)


Page 24 of 24

Influence of Silicon on the Nucleation Rate of GaAs Nanowires on

Recommend Documents