- Email: [email protected]
Formation of Sn nano-template by Hot-Wire Chemical Vapor Process on different substrates
Accepted Manuscript
Formation of Sn nanotemplate by Hot-wire Chemical Vapor Process on different substrates Ankur Soam , Nitin Arya , Nagsen Meshram , Alka Kumbhar , Rajiv Dusane PII: DOI: Reference:
S2468-0230(17)30069-X 10.1016/j.surfin.2017.06.005 SURFIN 110
To appear in:
Surfaces and Interfaces
Received date: Revised date: Accepted date:
23 March 2017 26 May 2017 19 June 2017
Please cite this article as: Ankur Soam , Nitin Arya , Nagsen Meshram , Alka Kumbhar , Rajiv Dusane , Formation of Sn nanotemplate by Hot-wire Chemical Vapor Process on different substrates, Surfaces and Interfaces (2017), doi: 10.1016/j.surfin.2017.06.005
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Formation of Sn nanotemplate by Hot-wire Chemical Vapor Process on different substrates Ankur Soam1,2, Nitin Arya1, Nagsen Meshram1, Alka Kumbhar1 and Rajiv Dusane1 1
Semiconductor Thin Films and Plasma Processing Laboratory, Department of Metallurgical Engineering and Materials Science, Indian Institute of Technology Bombay, Mumbai-400076, India 2
CR IP T
Nanomaterials and Device Fabrication Laboratory, Center for Nanoscience and Nanotechnology, Siksha O Anusandhan University, Bhubaneswar-751030, Odisha. Email: [email protected]
Abstract
AN US
We report here a primer on the formation of Sn nano-template by Hot-wire Chemical Vapor Process on different substrates namely Stainless steel (SS 304), polished Si wafer and corning glass. The morphology of the as-deposited Sn film is found to be different on different substrates. On SS substrate, interestingly, the as-deposited Sn film is a naturally grown template of Sn nano-particles without any annealing required whereas on Si wafer and glass substrates it
M
is in a quasi-continuous film form of Sn rather than a well defined nano-template (well separated nano-particles) as in the former case. However, the as-deposited films on glass and Si wafer also turn into a defined template after annealing. It has been detected that all these Sn nano-particles
ED
are covered with a thin layer of SnO that has to be removed to enable the growth of SiNWs.
dependent.
PT
Therefore, the morphology and the growth of SiNWs are found to be necessarily substrate
CE
Keywords: Hot-wire CVP, Sn nano-template, stainless steel, silicon nanowires
Introduction
AC
Uniform distribution and appropriate size of the metal nano-template or nano-particles are of significant importance for growing the SiNWs with well-controlled dimensions via VLS mechanism [1-3]. Hot-wire Chemical Vapor Process (HWCVP) grown SiNWs mostly have a core-shell structure [4]. The microstructure properties of the shell depend on the HWCVP parameters during SiNWs growth whereas the size of the core is governed by the size of used nano-template which can in turn be controlled by the thickness of the catalyst film or the annealing time. It is thus necessary to study the synthesis conditions of the Sn nano-template in
ACCEPTED MANUSCRIPT
order to grow SiNWs with well-controlled dimensions to be integrated in a device. Nano-particle formation using different metals such as Al, Au, In and Sn has been reported earlier by annealing in vacuum and hydrogen plasma treatment [1, 2, 5, 6]. P. J. Alet et al., have reported Au nanoisland formation by annealing the as-deposited film in the range of substrate temperature of 150°C-500°C [7]. The island formation process was attributed to the coalescence of the metal
CR IP T
particles resulting in the formation of Au aggregates. M. Jeon and K. Kamisako [8] have reported metal indium nanocrystals formation by hydrogen plasma treatment wherein the particles formed at 400 oC have size from several nm to 150 nm. In both the cases the treatment temperature used was high (≥400 ºC). Gold catalyst is widely employed to grow the SiNWs with an eutectic temperature of 363 ºC [6, 9]. SiNWs grown with Gold as catalyst have deep states in silicon
AN US
which degrade the electronic properties of the SiNWs [6]. In this regard Sn can be considered as a good alternative to Au as it has good electronic compatibility with Si and it can also bring down the growth temperature to as low as 240 oC. SiNWs synthesized at such low temperature can be used in systems where high-temperature process is not allowed. Though a lot of work has been reported on the fabrication of SiNWs using Sn nano-particles, not much has been said about
M
the formation of Sn nano-template. In the present work we have studied the Sn nano-particle formation on different substrate in HWCVP. The role of atomic hydrogen generated by the hot
Experimental details
ED
filament in HWCVP on the Sn nano-particles is also discussed.
PT
Prior to deposition the Sn film, the substrate of Si wafer, glass and stainless steel were cleaned ultrasonically in DI water and finally in methanol. The Sn film was deposited in thermal
CE
evaporation unit at room temperature by taking different amount of the Sn powder in a Tantalum boat. The hydrogen treatment on the as-deposited film was performed for 10 min in a HWCVP
AC
chamber where the hydrogen dissociation was carried out over a resistive heated Tantalum filament at 1700 oC. The flow of hydrogen was fixed at 50 sccm and the chamber pressure was adjusted to 50 mTorr. During the hydrogen treatment the substrate temperature was kept at constant temperature of 400 oC. To study the effect of the substrate on the SiNWs growth, the Sn nano-particles were then exposed to the Si species in a different HWCVP chamber in which the silane (SiH4) was dissociated over the Ta filament heated at 1700 oC [10]. The chamber pressure during the growth of SiNWs was set to 20 mTorr. The substrate temperature and the flow were
ACCEPTED MANUSCRIPT
set to 400 oC and 1 sccm, respectively. The Sn nano-particles and SiNWs morphology was observed by Field emission gun scanning electron microscopy (FEG-SEM, Model Jeol, JSM7600F, operated voltage 0.1 to 30 kV as well as Dual beam FIB, from Carl Zeiss).
Results and discussion
CR IP T
Fig. 1 shows the morphology of as-deposited film on different substrates at the room temperature. The as-deposited Sn film has different morphology on different substrate. It consists of nano-particles of spherical or oval like shapes on SS substrate. On Si wafer the film has island type structure with oval like shapes whereas it seems to be a continuous film on a glass substrate. Y. Y. Wei et al. [11] have observed the islands formation when 5-40 nm thick films of Fe and Ni
AN US
on a silicon substrate were annealed in a vacuum at a substrate temperature of 660 oC.
Formation of these islands is caused by the stress induced between the substrate and the film which breaks the film apart. In the case of SS substrate, no annealing step was required to break the film into small islands as the deposited material was naturally found in nano-particles form.
M
However, the as-deposited film is found to be almost continuous on a glass substrate instead of having nano-particles as shown in the Fig. 1 (c). Therefore, it is possible that the dewetting is
ED
higher on SS substrate as compared to the glass. The film on Si wafer has different structure showing the irregular Sn nano-particles. Though it is yet not clear why the as-deposited film of Sn on SS substrate is in the form of nano-particles rather than a continuous film? We think that
PT
binding energy of Sn-Sn system may be stronger than the Sn and SS substrate which causes dewetting of Sn film on SS substrate resulting in the formation of the droplets. The hydrogen
CE
may also affect the surface of the SS substrate. It is also possible that the hydrogen can increase the wetability of Sn on SS substrate, therefore deviation from the spherical particle. Moreover, it
AC
has also been also reported that the atomic hydrogen etches Sn in form of Sn hydride [12] which can also affect the morphology of as-deposited Sn film as observed in SEM images.
(b)
After annealing the as-deposited Sn film on wafer and glass in presence of atomic hydrogen, it changes to spherical particles (Fig. 2). S. Gbordzoe et al. [13] have also observed the same behaviour when the as-deposited film of Au on Si wafer was annealed at 800 oC to transform it into nano-particles. In our case, the particles are in liquid state as the substrate temperature was
ACCEPTED MANUSCRIPT
set above the melting point of Sn, therefore their shape changes into spherical from irregular as found on glass and Si substrates [14]. However, in our case slight increase in the size of the droplets and further deformation in their original shape are also observed on SS substrate. The particle density after annealing does not seem to be changed much on SS substrate as compared to that on glass and Si wafer. The density of particles is greatly decreased on wafer and glass as
CR IP T
the small particles coalesce to bigger clusters; therefore, we found spherical particles of Sn with a larger size on wafer and glass substrates. However, this effect is not seen on SS substrate. It has been established that the formation of nanostructures is also influenced by the roughness of underlying substrate [15]. As the energy barriers for migration of clusters on a rough surface is higher than the smooth surface small clusters of Sn atoms will not migrate on the SS substrate
AN US
easily and as a result no notable change in the size of the particles is observed [16]. Therefore formation of bigger particles on smoother surfaces is more likely.
In Fig. 2 (b) and Fig. 2 (c), the "dirt" surrounding the Sn particles is also visible. To find out the elemental composition of Sn nanoparticles on glass substrate, EDX study was done and the
M
results are shown in Fig. 3. It indicates the presence of Sn, silicon and oxygen on the Sn nanoparticles on the glass substrate. It shows the major peak of Sn. Some amount of Si is also
ED
present which could come from the deposition chamber itself since the environment in the chamber is predominantly Si rich as the same chamber is used to deposit Si films and to grow SiNWs. We can also find the oxygen peak in the EDX. Actually, after hydrogen treatment the
PT
sample is exposed to the atmosphere for characterization and this causes the oxidation of Sn nanoparticles. This is the reason why we don’t get SiNWs growth on as-deposited Sn film on SS
CE
substrate without hydrogen treatment. To grow the SiNWs, the H-treated Sn films were exposed to SiH4 instantly in the same chamber without breaking vacuum. The dirt between the particles
AC
could be of remaining Sn on the substrate left out after the entire process of being etched away by atomic hydrogen in form of metal hydride and annealing. To control the size of the Sn nano-particles on SS substrate, Sn films with different average particle sizes were deposited by placing the different amount of Sn, 1.5 mg, 3.0 mg, 6 mg, and 9.0 mg inside the Tantalum boat during thermal evaporation. The as-deposited film consists of nano-particles (Fig. 4). Their size distribution is shown in Fig. 5 and the number density of the Sn particles is summarized in Table 1. The size of these particles is found to increase linearly
ACCEPTED MANUSCRIPT
with Sn quantity (Fig. 6). Another important observation is that the density of small particles reduces while the number of bigger particles increases with the increment in the quantity of Sn. The increase in the particle size is accompanied by a decrease in the particle density. This effect is more pronounced for a large quantity of Sn. For 9.0 mg of Sn, the separation between the particles is also larger. If the Sn quantity is increased, particles of bigger size appear over the
CR IP T
substrate since more particles of smaller size coalesce together forming large particles. This is clearly seen in SEM images of Fig. 4, where the density of small sized particles is large and the spread over the entire substrate for 1.5 mg as compared to that of 9.0 mg of Sn.
Table 1 Observed Sn particle size and density for different quantity of Sn during thermal evaporation
Size of particles (nm) Particle Density (x1010 cm-2)
AN US
Amount of Sn (mg) 1.5 3.0 6.0
15-20
27
5-6
57
2-3
85
0.7-1.2
M
9.0
15
These Sn nano-particles on SS substrate were then exposed directly to the Si flux generated by
ED
the hot filament in HWCVP to synthesize the SiNWs. Initially, SiNWs could not be obtained and the particles were found to be covered with amorphous Si film as shown in Fig. 7 (a). SiNWs growth could only be facilitated when these particles were exposed to atomic hydrogen
PT
generated by the hot filament in HWCVP prior to the growth of SiNWs as shown in Fig. 7 (b). To identify the phases present in the as-deposited film of Sn, the Sn film was analyzed by X-ray
CE
diffraction shown in Fig. 8. In the X-ray diffraction spectrum, a peak of SnO (JCPDS file no. 01077-2296) is identified. Tin oxide acts as the barrier for Si diffusion into the Sn metallic phase.
AC
One can then conclude that as-deposited particles must be consisting of an upper thin layer of SnO which may reduce the catalytic activity of Sn preventing the growth of SiNWs and allowing the deposition of a-Si over the outer surface of the catalyst. This layer of SnO is removed by atomic hydrogen treatment making the Sn particles active for the growth of SiNWs.
Byung-Su Kim et al. [17] have studied the reduction of tin oxide by hydrogen gas. In that study the commercial powder of Sn (particle size of 45 µm) was exposed to the hydrogen gas and tin
ACCEPTED MANUSCRIPT
metal phase has been produced after the reaction. They have concluded that the reaction rate of hydrogen with SnO2 is initially rapid and then decreases because of pores being filled by molten tin. Therefore, in our case, the atomic hydrogen removes the SnO from the Sn particles (Fig. 9) and makes it active for the growth of SiNWs.
CR IP T
After the study of the formation of Sn nano-particles on different substrates, the effect of the growth of SiNWs on them is also studied. The typical morphology of SiNWs synthesized by HWCVP using Sn catalyst on Si wafer and glass substrates is illustrated in FEG-SEM image shown in Fig. 10. The growth time of the SiNWs was adjusted to 10 min. Prior to the growth, the as-deposited Sn film was treated by atomic hydrogen for 10 min to reduce the SnO to Sn [14]
AN US
and thus making it ready for SiNWs growth. The as-grown SiNWs on glass substrate are found to be having different morphology than the ones on SS and the wafer. Wires on SS and wafer have a needle-shaped structure whereas on glass they are observed to have noodles type structure. H.F. Al-Taay et al. [1] have also observed the same morphology of SiNWs grown on ITO coated glass using Sn catalyst. The difference in morphology could be due to different
M
thermal conductivities of the substrates. It has been reported that the SiNWs morphology is greatly dependent on the growth temperature [18, 19]. Glass has a poor thermal conductivity
ED
which causes the temperature at the growth surface to be lower than in the other two cases and hence resulting in a different morphology on glass substrate. In VLS growth, the diameter of SiNWs should be equal to the size of Sn nano-particles, which is not observed here. It is seen
PT
here that the diameter of the resulting SiNWs (100-200 nm) is larger than the size of Sn nanotemplate (20-40 nm). The simultaneous deposition of a-Si:H on SiNWs during the growth results
CE
in higher diameters [4, 20]. The diameter of HWCVP grown SiNWs using Sn as catalyst is found to be non-uniform along the length of the wire. The diameter is larger at the bottom part of the
AC
wire and it decreases towards the end of the wire as shown in Fig. 11. The bottom part of the wires is exposed for longer duration therefore larger diameter at the bottom. The size of the Sn particle also decreases gradually with the length of the wire, which is one more reason for tapering of SiNWs and it is completely removed after a certain length. The decrease in the size of Sn particles could be due to 1) the etching of Sn in the form of metal hydride by atomic hydrogen [12] or 2) the molten Sn could flow along the periphery of the NW due to its low surface tension [21].
ACCEPTED MANUSCRIPT
Conclusions We observe that it is possible to obtain Sn nano-particles on SS substrate without any annealing required. On other hand annealing is essential in order to get the Sn nano-template from the as-
CR IP T
deposited Sn film on a glass and wafer substrates. Moreover, Sn nano-particles are found to have a tin oxide layer on them which has to be removed in order to make them active for the growth of SiNWs. Controlling the size of Sn nano-particles is possible by controlling the amount of Sn during evaporation. Based on the present observations it appears that the Sn nano-template formation critically depends on the substrate morphology. Therefore, SiNWs morphology in turn
AN US
is also greatly influenced by the substrate naturally. However, deeper investigations are still required to fully understand the mechanism of template formation on different substrates.
References
4.
5. 6. 7. 8.
M
AC
9.
ED
3.
PT
2.
Al-Taay, H.F., et al., Controlling the diameter of silicon nanowires grown using a tin catalyst. Materials Science in Semiconductor Processing, 2013. 16(1): p. 15-22. Cui, Y., et al., Diameter-controlled synthesis of single-crystal silicon nanowires. Applied Physics Letters, 2001. 78(15): p. 2214-2216. Sharma, S., T.I. Kamins, and R.S. Williams, Diameter control of Ti-catalyzed silicon nanowires. Journal of Crystal Growth, 2004. 267(3–4): p. 613-618. Soam, A., et al., Controlling the shell microstructure in a low-temperature-grown SiNWs and correlating it to the performance of the SiNWs-based micro-supercapacitor. Applied Nanoscience, 2016. 6(8): p. 11591165. Weigand, C.C., et al., Effects of substrate annealing on the gold-catalyzed growth of ZnO nanostructures. Nanoscale Research Letters, 2011. 6(1): p. 566-566. Schmidt, V., et al., Silicon Nanowires: A Review on Aspects of their Growth and their Electrical Properties. Advanced Materials, 2009. 21(25-26): p. 2681-2702. Alet, P.-J., et al., Transition from thin gold layers to nano-islands on TCO for catalyzing the growth of onedimensional nanostructures. physica status solidi (a), 2008. 205(6): p. 1429-1434. Minsung Jeon and Koichi Kamisako, Synthesis of silicon nanowires after hydrogen radical treatment. Materials Letters, 2008. 62: 3903–3905. Kwak, D.W., H.Y. Cho, and W.C. Yang, Dimensional evolution of silicon nanowires synthesized by Au–Si island-catalyzed chemical vapor deposition. Physica E: Low-dimensional Systems and Nanostructures, 2007. 37(1–2): p. 153-157. Ankur Soam, Nitin Arya and Rajiv O. Dusane, Fabrication of silicon nanowires based on-chip microsupercapacitor, Chemical Physics Letters, 2017. 678: 46–50. Wei, Y.Y., et al., Effect of catalyst film thickness on carbon nanotube growth by selective area chemical vapor deposition. Applied Physics Letters, 2001. 78(10): p. 1394-1396. S. J. Rathi, et al., Tin-Catalyzed Plasma-Assisted Growth of Silicon Nanowires. The Journal of Physical Chemistry C, 2011. 115: 3833-3839. Gbordzoe, S., et al., Effect of substrate temperature on the microstructural properties of titanium nitride nanowires grown by pulsed laser deposition. Journal of Applied Physics, 2014. 116(6): p. 064310.
CE
1.
10. 11. 12. 13.
ACCEPTED MANUSCRIPT
15 16. 17. 18. 19. 20.
AC
CE
PT
ED
M
AN US
21.
S. Ya. Khmel, E. A. Baranov, A. V. Zaikovskii, A. O. Zamchiy, E. A. Maximovskiy, D. V. Gulyaev, and K. S. Zhuravlev, Synthesis of silicon oxide nanowires by the GJ EBP CVD method using different diluent gases, Phys. Status Solidi A, 2016. 7: 1774-1782. Ho, S.-T., et al., Catalyst-free selective-area growth of vertically aligned zinc oxide nanowires. Chemical Physics Letters, 2008. 463(1–3): p. 141-144. Roozbehi, M., et al., The effect of substrate surface roughness on ZnO nanostructures growth. Applied Surface Science, 2011. 257(8): p. 3291-3297 Kim, B.-S., et al., Reduction of SnO2 with Hydrogen. MATERIALS TRANSACTIONS, 2011. 52(9): p. 1814-1817. Ramanujam, J., D. Shiri, and A. Verma, Silicon Nanowire Growth and Properties: A Review. Materials Express, 2011. 1(2): p. 105-126. Schmidt, V., J.V. Wittemann, and U. Gösele, Growth, Thermodynamics, and Electrical Properties of Silicon Nanowires. Chemical Reviews, 2010. 110(1): p. 361-388. Linwei Yu, Benedict O’Donnell, Pierre-Jean Alet, S Conesa-Boj, F Peiro, J Arbiol and Pere Roca i Cabarrocas, Plasma-enhanced low temperature growth of silicon nanowires and hierarchical structures by using tin and indium catalysts. Nanotechnology, 2009. 20: 225604. S. Misra, et al., Wetting Layer: The Key Player in Plasma-Assisted Silicon Nanowire Growth Mediated by Tin. The Journal of Physical Chemistry C, 2013. 117: 17786-17790
CR IP T
14.
ACCEPTED MANUSCRIPT
Figure captions Fig. 1 SEM micrographs of as-deposited film of Sn on (a) SS substrate and (b) Si wafer (c) and glass. Fig. 2 Morphology of Sn films after annealing them to atomic hydrogen generated by the hot filament in HWCVD chmaber for 10 min at 400 oC substrate temperature (a) on SS (b) on wafer and (c) on glass substrate. Fig 3 Figure 1 EDX spectra of H treaded Sn film on glass substrate.
mg (d) 9.0 mg.
CR IP T
Fig. 4 SEM images of as-deposited film on SS substrate for different quantities of Sn (a) 1.5 mg (b) 3.0 mg (c) 6.0
Fig. 5 Size distribution of Sn nano-particles for as-deposited Sn films on SS substrate with different amount of Sn (a) 1.5 mg (b) 3.0 mg (c) 6.0 mg (d) 9.0 mg.
Fig. 6 Variation of the size of nano-particles as a function of the amount of Sn.
Fig. 7 (a) SEM images of SiNWs grown on as-deposited Sn film on SS substrate without Hydrogen treatment (nogrowth is observed) and (b) the growth of SiNWs after hydrogen treatment.
AN US
Fig. 8 XRD Spectra of as-deposited film of Sn on SS substrate, which shows SnO peak. The presence of a SnO on Sn metallic nano-particles prevents the growth of SiNWs.
Fig. 9 A schematic showing the role of atomic hydrogen on as-deposited Sn film (a) after H treatment (b). Fig. 10 SEM images of SiNWs grown on (a) wafer (b) glass substrate.
CE
PT
ED
M
Fig. 11 TEM image of a SiNW showing a Sn particle sitting at the end of the wire.
AC
List of Figures
(b)
(a)
100 nm Figure 1
(c)
ACCEPTED MANUSCRIPT
(b)
100 nm
100 nm
AC
CE
PT
ED
M
AN US
Figure 2
(c)
CR IP T
(a)
Figur 3
100 nm
ACCEPTED MANUSCRIPT
(b)
AN US
(c)
CR IP T
(a)
AC
CE
PT
ED
M
Figure 4
(d)
100 nm
AC
CE
PT
ED
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
Figure 5
AN US
CR IP T
ACCEPTED MANUSCRIPT
M
Figure 6
(a)
AC
CE
PT
ED
(b)
1 µm
Figure 7
Figure 8
PT
ED
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
(a)
AC
CE
(b)
Figure 9
ACCEPTED MANUSCRIPT
(b)
1 µm
AC
CE
PT
ED
M
AN US
Figure 10
CR IP T
(a)
Figure 11
1 µm
Formation of Sn nanotemplate by Hot-wire Chemical Vapor Process on different substrates Ankur Soam , Nitin Arya , Nagsen Meshram , Alka Kumbhar , Rajiv Dusane PII: DOI: Reference:
S2468-0230(17)30069-X 10.1016/j.surfin.2017.06.005 SURFIN 110
To appear in:
Surfaces and Interfaces
Received date: Revised date: Accepted date:
23 March 2017 26 May 2017 19 June 2017
Please cite this article as: Ankur Soam , Nitin Arya , Nagsen Meshram , Alka Kumbhar , Rajiv Dusane , Formation of Sn nanotemplate by Hot-wire Chemical Vapor Process on different substrates, Surfaces and Interfaces (2017), doi: 10.1016/j.surfin.2017.06.005
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Formation of Sn nanotemplate by Hot-wire Chemical Vapor Process on different substrates Ankur Soam1,2, Nitin Arya1, Nagsen Meshram1, Alka Kumbhar1 and Rajiv Dusane1 1
Semiconductor Thin Films and Plasma Processing Laboratory, Department of Metallurgical Engineering and Materials Science, Indian Institute of Technology Bombay, Mumbai-400076, India 2
CR IP T
Nanomaterials and Device Fabrication Laboratory, Center for Nanoscience and Nanotechnology, Siksha O Anusandhan University, Bhubaneswar-751030, Odisha. Email: [email protected]
Abstract
AN US
We report here a primer on the formation of Sn nano-template by Hot-wire Chemical Vapor Process on different substrates namely Stainless steel (SS 304), polished Si wafer and corning glass. The morphology of the as-deposited Sn film is found to be different on different substrates. On SS substrate, interestingly, the as-deposited Sn film is a naturally grown template of Sn nano-particles without any annealing required whereas on Si wafer and glass substrates it
M
is in a quasi-continuous film form of Sn rather than a well defined nano-template (well separated nano-particles) as in the former case. However, the as-deposited films on glass and Si wafer also turn into a defined template after annealing. It has been detected that all these Sn nano-particles
ED
are covered with a thin layer of SnO that has to be removed to enable the growth of SiNWs.
dependent.
PT
Therefore, the morphology and the growth of SiNWs are found to be necessarily substrate
CE
Keywords: Hot-wire CVP, Sn nano-template, stainless steel, silicon nanowires
Introduction
AC
Uniform distribution and appropriate size of the metal nano-template or nano-particles are of significant importance for growing the SiNWs with well-controlled dimensions via VLS mechanism [1-3]. Hot-wire Chemical Vapor Process (HWCVP) grown SiNWs mostly have a core-shell structure [4]. The microstructure properties of the shell depend on the HWCVP parameters during SiNWs growth whereas the size of the core is governed by the size of used nano-template which can in turn be controlled by the thickness of the catalyst film or the annealing time. It is thus necessary to study the synthesis conditions of the Sn nano-template in
ACCEPTED MANUSCRIPT
order to grow SiNWs with well-controlled dimensions to be integrated in a device. Nano-particle formation using different metals such as Al, Au, In and Sn has been reported earlier by annealing in vacuum and hydrogen plasma treatment [1, 2, 5, 6]. P. J. Alet et al., have reported Au nanoisland formation by annealing the as-deposited film in the range of substrate temperature of 150°C-500°C [7]. The island formation process was attributed to the coalescence of the metal
CR IP T
particles resulting in the formation of Au aggregates. M. Jeon and K. Kamisako [8] have reported metal indium nanocrystals formation by hydrogen plasma treatment wherein the particles formed at 400 oC have size from several nm to 150 nm. In both the cases the treatment temperature used was high (≥400 ºC). Gold catalyst is widely employed to grow the SiNWs with an eutectic temperature of 363 ºC [6, 9]. SiNWs grown with Gold as catalyst have deep states in silicon
AN US
which degrade the electronic properties of the SiNWs [6]. In this regard Sn can be considered as a good alternative to Au as it has good electronic compatibility with Si and it can also bring down the growth temperature to as low as 240 oC. SiNWs synthesized at such low temperature can be used in systems where high-temperature process is not allowed. Though a lot of work has been reported on the fabrication of SiNWs using Sn nano-particles, not much has been said about
M
the formation of Sn nano-template. In the present work we have studied the Sn nano-particle formation on different substrate in HWCVP. The role of atomic hydrogen generated by the hot
Experimental details
ED
filament in HWCVP on the Sn nano-particles is also discussed.
PT
Prior to deposition the Sn film, the substrate of Si wafer, glass and stainless steel were cleaned ultrasonically in DI water and finally in methanol. The Sn film was deposited in thermal
CE
evaporation unit at room temperature by taking different amount of the Sn powder in a Tantalum boat. The hydrogen treatment on the as-deposited film was performed for 10 min in a HWCVP
AC
chamber where the hydrogen dissociation was carried out over a resistive heated Tantalum filament at 1700 oC. The flow of hydrogen was fixed at 50 sccm and the chamber pressure was adjusted to 50 mTorr. During the hydrogen treatment the substrate temperature was kept at constant temperature of 400 oC. To study the effect of the substrate on the SiNWs growth, the Sn nano-particles were then exposed to the Si species in a different HWCVP chamber in which the silane (SiH4) was dissociated over the Ta filament heated at 1700 oC [10]. The chamber pressure during the growth of SiNWs was set to 20 mTorr. The substrate temperature and the flow were
ACCEPTED MANUSCRIPT
set to 400 oC and 1 sccm, respectively. The Sn nano-particles and SiNWs morphology was observed by Field emission gun scanning electron microscopy (FEG-SEM, Model Jeol, JSM7600F, operated voltage 0.1 to 30 kV as well as Dual beam FIB, from Carl Zeiss).
Results and discussion
CR IP T
Fig. 1 shows the morphology of as-deposited film on different substrates at the room temperature. The as-deposited Sn film has different morphology on different substrate. It consists of nano-particles of spherical or oval like shapes on SS substrate. On Si wafer the film has island type structure with oval like shapes whereas it seems to be a continuous film on a glass substrate. Y. Y. Wei et al. [11] have observed the islands formation when 5-40 nm thick films of Fe and Ni
AN US
on a silicon substrate were annealed in a vacuum at a substrate temperature of 660 oC.
Formation of these islands is caused by the stress induced between the substrate and the film which breaks the film apart. In the case of SS substrate, no annealing step was required to break the film into small islands as the deposited material was naturally found in nano-particles form.
M
However, the as-deposited film is found to be almost continuous on a glass substrate instead of having nano-particles as shown in the Fig. 1 (c). Therefore, it is possible that the dewetting is
ED
higher on SS substrate as compared to the glass. The film on Si wafer has different structure showing the irregular Sn nano-particles. Though it is yet not clear why the as-deposited film of Sn on SS substrate is in the form of nano-particles rather than a continuous film? We think that
PT
binding energy of Sn-Sn system may be stronger than the Sn and SS substrate which causes dewetting of Sn film on SS substrate resulting in the formation of the droplets. The hydrogen
CE
may also affect the surface of the SS substrate. It is also possible that the hydrogen can increase the wetability of Sn on SS substrate, therefore deviation from the spherical particle. Moreover, it
AC
has also been also reported that the atomic hydrogen etches Sn in form of Sn hydride [12] which can also affect the morphology of as-deposited Sn film as observed in SEM images.
(b)
After annealing the as-deposited Sn film on wafer and glass in presence of atomic hydrogen, it changes to spherical particles (Fig. 2). S. Gbordzoe et al. [13] have also observed the same behaviour when the as-deposited film of Au on Si wafer was annealed at 800 oC to transform it into nano-particles. In our case, the particles are in liquid state as the substrate temperature was
ACCEPTED MANUSCRIPT
set above the melting point of Sn, therefore their shape changes into spherical from irregular as found on glass and Si substrates [14]. However, in our case slight increase in the size of the droplets and further deformation in their original shape are also observed on SS substrate. The particle density after annealing does not seem to be changed much on SS substrate as compared to that on glass and Si wafer. The density of particles is greatly decreased on wafer and glass as
CR IP T
the small particles coalesce to bigger clusters; therefore, we found spherical particles of Sn with a larger size on wafer and glass substrates. However, this effect is not seen on SS substrate. It has been established that the formation of nanostructures is also influenced by the roughness of underlying substrate [15]. As the energy barriers for migration of clusters on a rough surface is higher than the smooth surface small clusters of Sn atoms will not migrate on the SS substrate
AN US
easily and as a result no notable change in the size of the particles is observed [16]. Therefore formation of bigger particles on smoother surfaces is more likely.
In Fig. 2 (b) and Fig. 2 (c), the "dirt" surrounding the Sn particles is also visible. To find out the elemental composition of Sn nanoparticles on glass substrate, EDX study was done and the
M
results are shown in Fig. 3. It indicates the presence of Sn, silicon and oxygen on the Sn nanoparticles on the glass substrate. It shows the major peak of Sn. Some amount of Si is also
ED
present which could come from the deposition chamber itself since the environment in the chamber is predominantly Si rich as the same chamber is used to deposit Si films and to grow SiNWs. We can also find the oxygen peak in the EDX. Actually, after hydrogen treatment the
PT
sample is exposed to the atmosphere for characterization and this causes the oxidation of Sn nanoparticles. This is the reason why we don’t get SiNWs growth on as-deposited Sn film on SS
CE
substrate without hydrogen treatment. To grow the SiNWs, the H-treated Sn films were exposed to SiH4 instantly in the same chamber without breaking vacuum. The dirt between the particles
AC
could be of remaining Sn on the substrate left out after the entire process of being etched away by atomic hydrogen in form of metal hydride and annealing. To control the size of the Sn nano-particles on SS substrate, Sn films with different average particle sizes were deposited by placing the different amount of Sn, 1.5 mg, 3.0 mg, 6 mg, and 9.0 mg inside the Tantalum boat during thermal evaporation. The as-deposited film consists of nano-particles (Fig. 4). Their size distribution is shown in Fig. 5 and the number density of the Sn particles is summarized in Table 1. The size of these particles is found to increase linearly
ACCEPTED MANUSCRIPT
with Sn quantity (Fig. 6). Another important observation is that the density of small particles reduces while the number of bigger particles increases with the increment in the quantity of Sn. The increase in the particle size is accompanied by a decrease in the particle density. This effect is more pronounced for a large quantity of Sn. For 9.0 mg of Sn, the separation between the particles is also larger. If the Sn quantity is increased, particles of bigger size appear over the
CR IP T
substrate since more particles of smaller size coalesce together forming large particles. This is clearly seen in SEM images of Fig. 4, where the density of small sized particles is large and the spread over the entire substrate for 1.5 mg as compared to that of 9.0 mg of Sn.
Table 1 Observed Sn particle size and density for different quantity of Sn during thermal evaporation
Size of particles (nm) Particle Density (x1010 cm-2)
AN US
Amount of Sn (mg) 1.5 3.0 6.0
15-20
27
5-6
57
2-3
85
0.7-1.2
M
9.0
15
These Sn nano-particles on SS substrate were then exposed directly to the Si flux generated by
ED
the hot filament in HWCVP to synthesize the SiNWs. Initially, SiNWs could not be obtained and the particles were found to be covered with amorphous Si film as shown in Fig. 7 (a). SiNWs growth could only be facilitated when these particles were exposed to atomic hydrogen
PT
generated by the hot filament in HWCVP prior to the growth of SiNWs as shown in Fig. 7 (b). To identify the phases present in the as-deposited film of Sn, the Sn film was analyzed by X-ray
CE
diffraction shown in Fig. 8. In the X-ray diffraction spectrum, a peak of SnO (JCPDS file no. 01077-2296) is identified. Tin oxide acts as the barrier for Si diffusion into the Sn metallic phase.
AC
One can then conclude that as-deposited particles must be consisting of an upper thin layer of SnO which may reduce the catalytic activity of Sn preventing the growth of SiNWs and allowing the deposition of a-Si over the outer surface of the catalyst. This layer of SnO is removed by atomic hydrogen treatment making the Sn particles active for the growth of SiNWs.
Byung-Su Kim et al. [17] have studied the reduction of tin oxide by hydrogen gas. In that study the commercial powder of Sn (particle size of 45 µm) was exposed to the hydrogen gas and tin
ACCEPTED MANUSCRIPT
metal phase has been produced after the reaction. They have concluded that the reaction rate of hydrogen with SnO2 is initially rapid and then decreases because of pores being filled by molten tin. Therefore, in our case, the atomic hydrogen removes the SnO from the Sn particles (Fig. 9) and makes it active for the growth of SiNWs.
CR IP T
After the study of the formation of Sn nano-particles on different substrates, the effect of the growth of SiNWs on them is also studied. The typical morphology of SiNWs synthesized by HWCVP using Sn catalyst on Si wafer and glass substrates is illustrated in FEG-SEM image shown in Fig. 10. The growth time of the SiNWs was adjusted to 10 min. Prior to the growth, the as-deposited Sn film was treated by atomic hydrogen for 10 min to reduce the SnO to Sn [14]
AN US
and thus making it ready for SiNWs growth. The as-grown SiNWs on glass substrate are found to be having different morphology than the ones on SS and the wafer. Wires on SS and wafer have a needle-shaped structure whereas on glass they are observed to have noodles type structure. H.F. Al-Taay et al. [1] have also observed the same morphology of SiNWs grown on ITO coated glass using Sn catalyst. The difference in morphology could be due to different
M
thermal conductivities of the substrates. It has been reported that the SiNWs morphology is greatly dependent on the growth temperature [18, 19]. Glass has a poor thermal conductivity
ED
which causes the temperature at the growth surface to be lower than in the other two cases and hence resulting in a different morphology on glass substrate. In VLS growth, the diameter of SiNWs should be equal to the size of Sn nano-particles, which is not observed here. It is seen
PT
here that the diameter of the resulting SiNWs (100-200 nm) is larger than the size of Sn nanotemplate (20-40 nm). The simultaneous deposition of a-Si:H on SiNWs during the growth results
CE
in higher diameters [4, 20]. The diameter of HWCVP grown SiNWs using Sn as catalyst is found to be non-uniform along the length of the wire. The diameter is larger at the bottom part of the
AC
wire and it decreases towards the end of the wire as shown in Fig. 11. The bottom part of the wires is exposed for longer duration therefore larger diameter at the bottom. The size of the Sn particle also decreases gradually with the length of the wire, which is one more reason for tapering of SiNWs and it is completely removed after a certain length. The decrease in the size of Sn particles could be due to 1) the etching of Sn in the form of metal hydride by atomic hydrogen [12] or 2) the molten Sn could flow along the periphery of the NW due to its low surface tension [21].
ACCEPTED MANUSCRIPT
Conclusions We observe that it is possible to obtain Sn nano-particles on SS substrate without any annealing required. On other hand annealing is essential in order to get the Sn nano-template from the as-
CR IP T
deposited Sn film on a glass and wafer substrates. Moreover, Sn nano-particles are found to have a tin oxide layer on them which has to be removed in order to make them active for the growth of SiNWs. Controlling the size of Sn nano-particles is possible by controlling the amount of Sn during evaporation. Based on the present observations it appears that the Sn nano-template formation critically depends on the substrate morphology. Therefore, SiNWs morphology in turn
AN US
is also greatly influenced by the substrate naturally. However, deeper investigations are still required to fully understand the mechanism of template formation on different substrates.
References
4.
5. 6. 7. 8.
M
AC
9.
ED
3.
PT
2.
Al-Taay, H.F., et al., Controlling the diameter of silicon nanowires grown using a tin catalyst. Materials Science in Semiconductor Processing, 2013. 16(1): p. 15-22. Cui, Y., et al., Diameter-controlled synthesis of single-crystal silicon nanowires. Applied Physics Letters, 2001. 78(15): p. 2214-2216. Sharma, S., T.I. Kamins, and R.S. Williams, Diameter control of Ti-catalyzed silicon nanowires. Journal of Crystal Growth, 2004. 267(3–4): p. 613-618. Soam, A., et al., Controlling the shell microstructure in a low-temperature-grown SiNWs and correlating it to the performance of the SiNWs-based micro-supercapacitor. Applied Nanoscience, 2016. 6(8): p. 11591165. Weigand, C.C., et al., Effects of substrate annealing on the gold-catalyzed growth of ZnO nanostructures. Nanoscale Research Letters, 2011. 6(1): p. 566-566. Schmidt, V., et al., Silicon Nanowires: A Review on Aspects of their Growth and their Electrical Properties. Advanced Materials, 2009. 21(25-26): p. 2681-2702. Alet, P.-J., et al., Transition from thin gold layers to nano-islands on TCO for catalyzing the growth of onedimensional nanostructures. physica status solidi (a), 2008. 205(6): p. 1429-1434. Minsung Jeon and Koichi Kamisako, Synthesis of silicon nanowires after hydrogen radical treatment. Materials Letters, 2008. 62: 3903–3905. Kwak, D.W., H.Y. Cho, and W.C. Yang, Dimensional evolution of silicon nanowires synthesized by Au–Si island-catalyzed chemical vapor deposition. Physica E: Low-dimensional Systems and Nanostructures, 2007. 37(1–2): p. 153-157. Ankur Soam, Nitin Arya and Rajiv O. Dusane, Fabrication of silicon nanowires based on-chip microsupercapacitor, Chemical Physics Letters, 2017. 678: 46–50. Wei, Y.Y., et al., Effect of catalyst film thickness on carbon nanotube growth by selective area chemical vapor deposition. Applied Physics Letters, 2001. 78(10): p. 1394-1396. S. J. Rathi, et al., Tin-Catalyzed Plasma-Assisted Growth of Silicon Nanowires. The Journal of Physical Chemistry C, 2011. 115: 3833-3839. Gbordzoe, S., et al., Effect of substrate temperature on the microstructural properties of titanium nitride nanowires grown by pulsed laser deposition. Journal of Applied Physics, 2014. 116(6): p. 064310.
CE
1.
10. 11. 12. 13.
ACCEPTED MANUSCRIPT
15 16. 17. 18. 19. 20.
AC
CE
PT
ED
M
AN US
21.
S. Ya. Khmel, E. A. Baranov, A. V. Zaikovskii, A. O. Zamchiy, E. A. Maximovskiy, D. V. Gulyaev, and K. S. Zhuravlev, Synthesis of silicon oxide nanowires by the GJ EBP CVD method using different diluent gases, Phys. Status Solidi A, 2016. 7: 1774-1782. Ho, S.-T., et al., Catalyst-free selective-area growth of vertically aligned zinc oxide nanowires. Chemical Physics Letters, 2008. 463(1–3): p. 141-144. Roozbehi, M., et al., The effect of substrate surface roughness on ZnO nanostructures growth. Applied Surface Science, 2011. 257(8): p. 3291-3297 Kim, B.-S., et al., Reduction of SnO2 with Hydrogen. MATERIALS TRANSACTIONS, 2011. 52(9): p. 1814-1817. Ramanujam, J., D. Shiri, and A. Verma, Silicon Nanowire Growth and Properties: A Review. Materials Express, 2011. 1(2): p. 105-126. Schmidt, V., J.V. Wittemann, and U. Gösele, Growth, Thermodynamics, and Electrical Properties of Silicon Nanowires. Chemical Reviews, 2010. 110(1): p. 361-388. Linwei Yu, Benedict O’Donnell, Pierre-Jean Alet, S Conesa-Boj, F Peiro, J Arbiol and Pere Roca i Cabarrocas, Plasma-enhanced low temperature growth of silicon nanowires and hierarchical structures by using tin and indium catalysts. Nanotechnology, 2009. 20: 225604. S. Misra, et al., Wetting Layer: The Key Player in Plasma-Assisted Silicon Nanowire Growth Mediated by Tin. The Journal of Physical Chemistry C, 2013. 117: 17786-17790
CR IP T
14.
ACCEPTED MANUSCRIPT
Figure captions Fig. 1 SEM micrographs of as-deposited film of Sn on (a) SS substrate and (b) Si wafer (c) and glass. Fig. 2 Morphology of Sn films after annealing them to atomic hydrogen generated by the hot filament in HWCVD chmaber for 10 min at 400 oC substrate temperature (a) on SS (b) on wafer and (c) on glass substrate. Fig 3 Figure 1 EDX spectra of H treaded Sn film on glass substrate.
mg (d) 9.0 mg.
CR IP T
Fig. 4 SEM images of as-deposited film on SS substrate for different quantities of Sn (a) 1.5 mg (b) 3.0 mg (c) 6.0
Fig. 5 Size distribution of Sn nano-particles for as-deposited Sn films on SS substrate with different amount of Sn (a) 1.5 mg (b) 3.0 mg (c) 6.0 mg (d) 9.0 mg.
Fig. 6 Variation of the size of nano-particles as a function of the amount of Sn.
Fig. 7 (a) SEM images of SiNWs grown on as-deposited Sn film on SS substrate without Hydrogen treatment (nogrowth is observed) and (b) the growth of SiNWs after hydrogen treatment.
AN US
Fig. 8 XRD Spectra of as-deposited film of Sn on SS substrate, which shows SnO peak. The presence of a SnO on Sn metallic nano-particles prevents the growth of SiNWs.
Fig. 9 A schematic showing the role of atomic hydrogen on as-deposited Sn film (a) after H treatment (b). Fig. 10 SEM images of SiNWs grown on (a) wafer (b) glass substrate.
CE
PT
ED
M
Fig. 11 TEM image of a SiNW showing a Sn particle sitting at the end of the wire.
AC
List of Figures
(b)
(a)
100 nm Figure 1
(c)
ACCEPTED MANUSCRIPT
(b)
100 nm
100 nm
AC
CE
PT
ED
M
AN US
Figure 2
(c)
CR IP T
(a)
Figur 3
100 nm
ACCEPTED MANUSCRIPT
(b)
AN US
(c)
CR IP T
(a)
AC
CE
PT
ED
M
Figure 4
(d)
100 nm
AC
CE
PT
ED
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
Figure 5
AN US
CR IP T
ACCEPTED MANUSCRIPT
M
Figure 6
(a)
AC
CE
PT
ED
(b)
1 µm
Figure 7
Figure 8
PT
ED
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
(a)
AC
CE
(b)
Figure 9
ACCEPTED MANUSCRIPT
(b)
1 µm
AC
CE
PT
ED
M
AN US
Figure 10
CR IP T
(a)
Figure 11
1 µm