**3. Results and discussion**

Indium metal has a low melting point, and the eutectic temperature of the In–Si binary system coincides with the melting temperature of indium at 157°C [20, 38]. It is also worthy to note that the In–Si eutectic alloy exhibits a steep liquidus line, such that the liquid alloy can promptly be supersaturated with Si in a wide range of temperatures (at least up to 800°C) and In-Si eutectic alloy has an extremely low Si solubility approximately ∼10−4 at.% Si [26].

In this work the In-NDs were grown with the optimized growth condition, where the In-deposition time were increased from 20 min to 30 min in the plasma assisted crystal growth reactor as shown in **Figure 2a** and **b**. By decreasing the *T*S from 600 to 530°C in the high vacuum along with H2-plasma treatment, the diameter of the In-NDs were varied from 45 nm and to 500 nm. However, in some In-NDs, got an exceptionally low contact angle as shown in **Figure 2b**, which is quite interesting in the context of NWs verticality control for better Si-precipitation via In-NDs. On the other hand, we observed exceedingly high contact angle approximately 140o , as shown in **Figure 2a**. In this case the liquid NDs acts as a facilitator site for Si deposition (precipitation). It was found that In-NDs have higher sticking coefficient than the solid surfaces [37]. As shown in the schematic view of **Figures 1** and **2c**, where the supersaturation of the In-NDs, induced by the continuous gas phase supply of Si species (Vapor), leads to the precipitation of Si nanowires (Solid) at the interface of Si-substrate and In-NDs (Liquid). Si NWs growth was initiated, when a steady-state condition between the flux of the Si through the particle and the precipitation of Si on the substrate via In-NDs was reached, shown in **Figure 1** [26, 37]. Later the growth condition for Si NWs was improved from sample-Na to sample-Nw. The contact angle along with the In-NDs size together can define the diameter of the Si NWs as depicted in the **Figures 2c** and **3a, b**. It is all about the growth condition as well as growth in the reactor with and without air-breaking. By using HR-SEM, we investigated the interface condition between the In-NDs and the Si substrate prior to the crystal growth of Si NWs, shown in **Figure 3a** and **b**. In the case of thermally evaporated In-NDs on Si-substrate, a spherical shaped In-NDs were observed with different sizes in the range of 30–100 nm, as well as with quite large θC of 140°, shown in **Figure 3a**.

#### **Figure 2.**

*(a) SEM images of In-NDs (high contact angle) grown by sputtering of In-target at RT, for 20 min and then treated by plasma treatment at 530°C for 3 min, (b) SEM images of In-NDs (low contact angle) grown by sputtering of In-target at RT, for 20 min and then treated by H2-plasma at 600°C for 3 min, and (c) Si precipitation mechanism in the VLS growth mode via well wetted In-NDs to grow Si NWs.*

**123**

**Figure 3.**

*Indium (In)-Catalyzed Silicon Nanowires (Si NWs) Grown by the Vapor–Liquid–Solid (VLS)…*

The spherical shape of In-NDs (due to low wettability) on the Si substrate can be explained by the interactions mechanism of oxidized In-NDs (In2O3) in the context of surface free energy (*E*f) and θC in a qualitative manner on Si substrate in the conventional physical evaporator. Previously, the surface free energy (*E*f) for pure In droplet and In2O3, respectively, were found to be 525 mN/m and 500–520 mN/m at 850 K [40]. During the growth of In-NDs in the air-breaking scenario, where the *E*f of In-NDs on Si might be reduced due to the presence of the thin oxide layer around the In-NDs on substrate and subsequently increased to the θC up to ~140° of In-NDs on Si substrate, shown in **Figure 3a**. As a result, extremely low density of vertically aligned (111)-Si NWs at the supersaturation phase were grown, shown in **Figure 3c**. In the grown Si NWs, one can see quite big used cap of In-NDs on the top

*with the permission from ref [39]. Copyright 2015 the Royal Society of Chemistry (RSC).*

*SEM images of In-NDs grown by sputtering in the reactor, (a) with air-breaking (sample-Na), (b) without air-breaking (sample-Nw). TEM micrograph of Si NWs grown in the reactor, (c) with air-breaking, (d) without air-breaking. HR-TEM micrograph of the as-grown Si NWs (e) taken at point "A1" in the sample-Na, and the SAED pattern taken at point "A3" and "A2" are shown in the inset, and (f) taken at point "B2" in the sample-Nw, and the SAED pattern taken at point "B3" and "B1" are shown in the inset. Figure 3 reproduced* 

*DOI: http://dx.doi.org/10.5772/intechopen.97723*

*Indium (In)-Catalyzed Silicon Nanowires (Si NWs) Grown by the Vapor–Liquid–Solid (VLS)… DOI: http://dx.doi.org/10.5772/intechopen.97723*

#### **Figure 3.**

*Nanowires - Recent Progress*

**3. Results and discussion**

solubility approximately ∼10−4 at.% Si [26].

contact angle approximately 140o

as via cross-sectional view by high-resolution transmission electron microscope (HR-TEM, Model: H-9000NAR). Migration and trapping of In-Nanoparticles (NPs) on the side wall of Si NWs were deeply investigated by energy-dispersive X-ray spectroscopy (EDX) and dark field scanning of scanning-TEM (STEM).

Indium metal has a low melting point, and the eutectic temperature of the In–Si binary system coincides with the melting temperature of indium at 157°C [20, 38]. It is also worthy to note that the In–Si eutectic alloy exhibits a steep liquidus line, such that the liquid alloy can promptly be supersaturated with Si in a wide range of temperatures (at least up to 800°C) and In-Si eutectic alloy has an extremely low Si

In this work the In-NDs were grown with the optimized growth condition, where the In-deposition time were increased from 20 min to 30 min in the plasma assisted crystal growth reactor as shown in **Figure 2a** and **b**. By decreasing the *T*S from 600 to 530°C in the high vacuum along with H2-plasma treatment, the diameter of the In-NDs were varied from 45 nm and to 500 nm. However, in some In-NDs, got an exceptionally low contact angle as shown in **Figure 2b**, which is quite interesting in the context of NWs verticality control for better Si-precipitation via In-NDs. On the other hand, we observed exceedingly high

NDs acts as a facilitator site for Si deposition (precipitation). It was found that In-NDs have higher sticking coefficient than the solid surfaces [37]. As shown in the schematic view of **Figures 1** and **2c**, where the supersaturation of the In-NDs, induced by the continuous gas phase supply of Si species (Vapor), leads to the precipitation of Si nanowires (Solid) at the interface of Si-substrate and In-NDs (Liquid). Si NWs growth was initiated, when a steady-state condition between the flux of the Si through the particle and the precipitation of Si on the substrate via In-NDs was reached, shown in **Figure 1** [26, 37]. Later the growth condition for Si NWs was improved from sample-Na to sample-Nw. The contact angle along with the In-NDs size together can define the diameter of the Si NWs as depicted in the **Figures 2c** and **3a, b**. It is all about the growth condition as well as growth in the reactor with and without air-breaking. By using HR-SEM, we investigated the interface condition between the In-NDs and the Si substrate prior to the crystal growth of Si NWs, shown in **Figure 3a** and **b**. In the case of thermally evaporated In-NDs on Si-substrate, a spherical shaped In-NDs were observed with different sizes in the range of 30–100 nm, as well as with quite large θC of 140°, shown in **Figure 3a**.

*(a) SEM images of In-NDs (high contact angle) grown by sputtering of In-target at RT, for 20 min and then treated by plasma treatment at 530°C for 3 min, (b) SEM images of In-NDs (low contact angle) grown by sputtering of In-target at RT, for 20 min and then treated by H2-plasma at 600°C for 3 min, and (c) Si* 

*precipitation mechanism in the VLS growth mode via well wetted In-NDs to grow Si NWs.*

, as shown in **Figure 2a**. In this case the liquid

**122**

**Figure 2.**

*SEM images of In-NDs grown by sputtering in the reactor, (a) with air-breaking (sample-Na), (b) without air-breaking (sample-Nw). TEM micrograph of Si NWs grown in the reactor, (c) with air-breaking, (d) without air-breaking. HR-TEM micrograph of the as-grown Si NWs (e) taken at point "A1" in the sample-Na, and the SAED pattern taken at point "A3" and "A2" are shown in the inset, and (f) taken at point "B2" in the sample-Nw, and the SAED pattern taken at point "B3" and "B1" are shown in the inset. Figure 3 reproduced with the permission from ref [39]. Copyright 2015 the Royal Society of Chemistry (RSC).*

The spherical shape of In-NDs (due to low wettability) on the Si substrate can be explained by the interactions mechanism of oxidized In-NDs (In2O3) in the context of surface free energy (*E*f) and θC in a qualitative manner on Si substrate in the conventional physical evaporator. Previously, the surface free energy (*E*f) for pure In droplet and In2O3, respectively, were found to be 525 mN/m and 500–520 mN/m at 850 K [40]. During the growth of In-NDs in the air-breaking scenario, where the *E*f of In-NDs on Si might be reduced due to the presence of the thin oxide layer around the In-NDs on substrate and subsequently increased to the θC up to ~140° of In-NDs on Si substrate, shown in **Figure 3a**. As a result, extremely low density of vertically aligned (111)-Si NWs at the supersaturation phase were grown, shown in **Figure 3c**. In the grown Si NWs, one can see quite big used cap of In-NDs on the top of Si NWs. The non-uniform diameter and length of Si NWs ranging for 30–100 nm and 150–200 nm, respectively, were observed by TEM and SEM observations, shown in **Figure 3c**. It can be speculated that the precipitation of Si atomic migration through the liquid phase of In-NDs to the Si substrate at the eutectic phase might be hindered by the thin oxide layer that exists around the majority of In-NDs and on the Si substrate. This kind of low wettability condition with extremely large θC of In-NDs on Si substrate does not support to the crystal growth of high-density Si NWs, shown in **Figure 3c**. Due to the small contact area of wet surface of In-NDs on the Si substrate may not be symmetric too, which ultimately give rise to the tilted Si NWs with big In-NPs cap, shown in **Figure 3c**. The (111)-oriented crystallinity at point "A1" was confirmed by the HR-TEM image (**Figure 3e**). The SAED at points "A3" and "A2" (shown in the inset of **Figure 3e**) corresponding to Si NWs and Si substrate, respectively, exhibit a spotty pattern, which indicating that the VLS growth was happened successfully for few Si NWs (with extremely low density of 2.5 μm−2) using crystal growth in the air-breaking condition (sample-Na).

In the second experiment (without air-breaking), we optimized the growth conditions to obtain good interface between In-NDs and Si substrate by eliminating the oxide layer to enhance the wettability. All the in-situ growth steps starting from In-NDs growth, plasma treatment, and up to the crystal growth of Si NWs were performed in a relatively clean and high vacuum chamber without airbreaking (sample-Nw). The θC of In on Si (111) was reported to be approximately 125° at 350°C by Mattila *et al.* [41] After improving the wettability condition, the θC of In-NDs on Si substrate was remarkably decreased from our previous value of 140° to 80°, shown in the inset of **Figure 3b**. Ultimately the In-NDs on the Si substrate were suppressed to hemispherical shaped with uniform sizes in the range of 70–100 nm (without air-breaking). The wettability of the In-NDs and increase of the *E*f of In-NDs, are attributed to the combined action of the H2-plasma treatment at *T*S of 600°C, and clean growth environment by performing the entire growth process in the same chamber (without air-breaking). The verticality of (111)-oriented Si NWs was nicely controlled with a uniform top diameter and length, respectively, of approximately 18 nm and 100 nm, shown in **Figure 3d** and **f**. The top diameter value approximately ~18 nm of Si NWs is the smallest value ever achieved for the applications of nano-scale devices application. The controllability of the verticality of Si NWs are attributed to the cylindrically symmetric flow of the precipitated Si atoms via highly wettable In-NDs toward the Si substrate contact points at an incredibly low θC, shown in **Figure 3d**–**f**. The density of the grown Si NWs at *T*S = 600°C was also remarkably enhanced by 28 time from 2.5 μm−2 (sample-Na) to 70 μm−2 (sample-Nw). The Si NWs were found to be cone shaped after tapering, which is caused by the migration of In-NPs from the top of NWs, shown in **Figure 3a** and **b**. Based on the HR-TEM observation at point "B2", the (111)-orientation of Si NWs was confirmed (as shown in the inset of **Figure 3f**). Also, spotty pattern both in Si NWs and Si substrate at point "B1" and "B3", respectively, were confirmed by SAED observation (shown in the inset of **Figure 3f**). Energy dispersive X-ray spectroscopy (EDX) was taken near the cap of the Si NWs to investigate the compositional information. It was found that NW only contains a Si peak, while In was detected for the spheres only. The top diameter value approximately 18 nm was found quite small (without air-breaking) when compared to the previously obtained vertically aligned Si NWs grown with In catalyst in the VLS mode (with air-breaking condition). We strongly need to decrease the tapering phenomena and also to increase the length of the In-catalyzed Si NWs. These are still an open research challenges and it can be speculated that the control of the diameter of In-NDs both at the top and base area might be possible after suppressing the trapping rate of In NPs by Si NWs. Further reduction of In-NDs θC on Si substrate

**125**

**Figure 4.**

*Physics (JSAP).*

*Indium (In)-Catalyzed Silicon Nanowires (Si NWs) Grown by the Vapor–Liquid–Solid (VLS)…*

Alet *et al.* grown In-catalyzed crystalline Si NWs at low temperature on Indium doped Tin Oxide (ITO), which were not vertically aligned and detail investigation about the planar as well as stacking fault were not given [42]. In our case a typical cross-section TEM image of the cone like Si NWs arrays grown at 600°C has been shown in **Figure 4b**. The HR-TEM shown in **Figures 3d** and **4b** shows that the Si NWs have an average diameter of ≅ 18 nm at the top and average diameter≅ 30 nm at the bottom, lengths longer than 100 nm and grow at a right angle with respect

can further enhance the length of NWs. We need to further reduce the size and wettability of In-NDs and then can further reduce the diameter of the vertically aligned Si NWs to realize the quantum size effect for wide band gap applications of solar cells. In-catalyzed Si NWs may be a potential candidate material for nanoscale

devices, because of its shallow acceptor levels and its vertical alignment.

*(a) Cross-sectional TEM image of the grown Si NWs on Si (111), (b) HR-cross-sectional TEM image of an individual Si NW from fig. 4a (In the inset of (b), the corresponding SAED pattern taken along the [110] zone axis at point "P1", "P2", "P3", and "P4" respectively are given). Different planar defects along the axial segments and twining of planar defects has been shown at point S-I and S-II, respectively. Figure 4 reproduced with the permission from ref [43]. Copyright 2016 the Japan Society of Applied* 

*DOI: http://dx.doi.org/10.5772/intechopen.97723*

*Indium (In)-Catalyzed Silicon Nanowires (Si NWs) Grown by the Vapor–Liquid–Solid (VLS)… DOI: http://dx.doi.org/10.5772/intechopen.97723*

can further enhance the length of NWs. We need to further reduce the size and wettability of In-NDs and then can further reduce the diameter of the vertically aligned Si NWs to realize the quantum size effect for wide band gap applications of solar cells. In-catalyzed Si NWs may be a potential candidate material for nanoscale devices, because of its shallow acceptor levels and its vertical alignment.

Alet *et al.* grown In-catalyzed crystalline Si NWs at low temperature on Indium doped Tin Oxide (ITO), which were not vertically aligned and detail investigation about the planar as well as stacking fault were not given [42]. In our case a typical cross-section TEM image of the cone like Si NWs arrays grown at 600°C has been shown in **Figure 4b**. The HR-TEM shown in **Figures 3d** and **4b** shows that the Si NWs have an average diameter of ≅ 18 nm at the top and average diameter≅ 30 nm at the bottom, lengths longer than 100 nm and grow at a right angle with respect

#### **Figure 4.**

*Nanowires - Recent Progress*

of Si NWs. The non-uniform diameter and length of Si NWs ranging for 30–100 nm and 150–200 nm, respectively, were observed by TEM and SEM observations, shown in **Figure 3c**. It can be speculated that the precipitation of Si atomic migration through the liquid phase of In-NDs to the Si substrate at the eutectic phase might be hindered by the thin oxide layer that exists around the majority of In-NDs and on the Si substrate. This kind of low wettability condition with extremely large θC of In-NDs on Si substrate does not support to the crystal growth of high-density Si NWs, shown in **Figure 3c**. Due to the small contact area of wet surface of In-NDs on the Si substrate may not be symmetric too, which ultimately give rise to the tilted Si NWs with big In-NPs cap, shown in **Figure 3c**. The (111)-oriented crystallinity at point "A1" was confirmed by the HR-TEM image (**Figure 3e**). The SAED at points "A3" and "A2" (shown in the inset of **Figure 3e**) corresponding to Si NWs and Si substrate, respectively, exhibit a spotty pattern, which indicating that the VLS growth was happened successfully for few Si NWs (with extremely low density of

2.5 μm−2) using crystal growth in the air-breaking condition (sample-Na).

In the second experiment (without air-breaking), we optimized the growth conditions to obtain good interface between In-NDs and Si substrate by eliminating the oxide layer to enhance the wettability. All the in-situ growth steps starting from In-NDs growth, plasma treatment, and up to the crystal growth of Si NWs were performed in a relatively clean and high vacuum chamber without airbreaking (sample-Nw). The θC of In on Si (111) was reported to be approximately 125° at 350°C by Mattila *et al.* [41] After improving the wettability condition, the θC of In-NDs on Si substrate was remarkably decreased from our previous value of 140° to 80°, shown in the inset of **Figure 3b**. Ultimately the In-NDs on the Si substrate were suppressed to hemispherical shaped with uniform sizes in the range of 70–100 nm (without air-breaking). The wettability of the In-NDs and increase of the *E*f of In-NDs, are attributed to the combined action of the H2-plasma treatment at *T*S of 600°C, and clean growth environment by performing the entire growth process in the same chamber (without air-breaking). The verticality of (111)-oriented Si NWs was nicely controlled with a uniform top diameter and length, respectively, of approximately 18 nm and 100 nm, shown in **Figure 3d** and **f**. The top diameter value approximately ~18 nm of Si NWs is the smallest value ever achieved for the applications of nano-scale devices application. The controllability of the verticality of Si NWs are attributed to the cylindrically symmetric flow of the precipitated Si atoms via highly wettable In-NDs toward the Si substrate contact points at an incredibly low θC, shown in **Figure 3d**–**f**. The density of the grown Si NWs at *T*S = 600°C was also remarkably enhanced by 28 time from 2.5 μm−2 (sample-Na) to 70 μm−2 (sample-Nw). The Si NWs were found to be cone shaped after tapering, which is caused by the migration of In-NPs from the top of NWs, shown in **Figure 3a** and **b**. Based on the HR-TEM observation at point "B2", the (111)-orientation of Si NWs was confirmed (as shown in the inset of **Figure 3f**). Also, spotty pattern both in Si NWs and Si substrate at point "B1" and "B3", respectively, were confirmed by SAED observation (shown in the inset of **Figure 3f**). Energy dispersive X-ray spectroscopy (EDX) was taken near the cap of the Si NWs to investigate the compositional information. It was found that NW only contains a Si peak, while In was detected for the spheres only. The top diameter value approximately 18 nm was found quite small (without air-breaking) when compared to the previously obtained vertically aligned Si NWs grown with In catalyst in the VLS mode (with air-breaking condition). We strongly need to decrease the tapering phenomena and also to increase the length of the In-catalyzed Si NWs. These are still an open research challenges and it can be speculated that the control of the diameter of In-NDs both at the top and base area might be possible after suppressing the trapping rate of In NPs by Si NWs. Further reduction of In-NDs θC on Si substrate

**124**

*(a) Cross-sectional TEM image of the grown Si NWs on Si (111), (b) HR-cross-sectional TEM image of an individual Si NW from fig. 4a (In the inset of (b), the corresponding SAED pattern taken along the [110] zone axis at point "P1", "P2", "P3", and "P4" respectively are given). Different planar defects along the axial segments and twining of planar defects has been shown at point S-I and S-II, respectively. Figure 4 reproduced with the permission from ref [43]. Copyright 2016 the Japan Society of Applied Physics (JSAP).*

to the Si (111) surface as shown in **Figures 3d**–**f** and **4b**. The shape of the Si NWs were found to be cone like, and this can be explained by the similar analogy of tapering of the NWs grown by Au-catalyzed [44, 45]. Sharma *et al*. found that, when NWs elongate from the In-NDs (or In- droplets), the base area of the NWs remains exposed for a longer time to the reactive radicals when compared to the newly grown upper part of the NWs [46]. In the case of Au-catalyzed Si NWs grown in the 〈112〉 direction, where single twin planes and arrays of {111} stacking faults were reported [14], whereas Lopez *et al.* found the same structures for 〈111〉-oriented Au-catalyzed grown Si NWs [36]. These defects orientation is parallel to the 〈112〉 NW axis, often extending throughout the entire length of the wire. There is no report about the crystalline twining defects, planar defects and stacking fault along 〈112〉 direction in case of In-catalyzed vertically aligned, and (111)-oriented Si NWs.

We can see the planar defects as well as twining defects appearing in many segments of the Si NWs (sample-Nw) as shown in the inset of **Figure 5b** by green color arrow as well as red color arrow along the 〈112〉 direction. The (111)-oriented Si NWs at segments "P1–4" were also confirmed by HR-TEM image, as shown in **Figure 4b**. The corresponding selected area diffraction (SAED) shown in the in-set of **Figure 4b** has been recorded along the [110] zone axis from the Si NWs and which confirm the single crystal nature and its axial direction. But you can see twining of planes along 〈112〉 direction, which are distributed along the axial direction marked as S-I and S-II, as shown in **Figure 4b** for single (111)-oriented Si NW.

It can be established that the HR-TEM image of single Si NWs along with the SAED at "P1" and "P3" of Si NWs (taken from **Figure 5b**), marked by green color

#### **Figure 5.**

*(a) High-resolution cross-sectional TEM image around the Si-substrate and deposited In-NDs surrounded by Si NWs at the interface, (b) the corresponding SAED pattern taken along the [1* 1 *0] zone axis at point "P5" and "P6" respectively. In the inset of (b) is a SAED along [* 1 *10] zone axis at point "P5", where the red arrow intensity originates from 1/2{111} spots associated with the 2H polytype and the green intensity is originating from 1/3{111} spots associated with 9R stacking. Figure 5 reproduced with the permission from ref [43]***.**  *Copyright 2016 the Japan Society of Applied Physics (JSAP).*

**127**

*Indium (In)-Catalyzed Silicon Nanowires (Si NWs) Grown by the Vapor–Liquid–Solid (VLS)…*

arrow, follow the same crystallinity of Si-substrate orientation marked at "P6" in **Figure 5b**. The diamond structure of the Si NWs with a lattice constant = 0.543 nm was confirmed by HR-TEM image and the SAED pattern at segments "P1" and "P3", which corresponds to the space group Fd3m grown along the 〈111〉 direction. Unfortunately, the SAED at "P2" and "P4" (taken from **Figure 4b**) of Si NW shows small tilt with respect to orientation of Si-substrate, and the same have been marked by red color arrow, as shown in HR-TEM image given in **Figure 5b**. Spotty pattern both at "P2" and "P4" segments of the NW were confirmed by SAED observation, as shown in the inset of **Figure 5b** and same value of small angular tilt of both "P2" and "P4" segments were observed. We concluded that the Si NWs are vertically aligned except the twining of planar defects, which might be caused by the faster cooling rate 100°C/6 min, as well as due to the longer exposer time of downside wall to the reactive radicals as compared to the upper side of Si NWs [46]. We have to consider about the thermal conductivity mechanism, which is different

Here we focus to discuss about the Si NWs crystallinity at the interface of emanating Si NWs from the Si-substrate, as shown in **Figure 5a**. **Figure 5b** gives the SAED pattern at point "P5", which confirm the stacking fault in the grown Si NWs. The red arrow intensity originates from 1/2{111} spots is related to the 2H-polytype (stacking ABAB…) and the green arrow intensity originates from 1/3{111} spots is

Such structures are either attributed to scattering phenomena from two overlapping crystals with a stepped {111} twin boundary (parallel to the electron beam) [48, 49], or it might be the direct evidence of a 9R-polytype [23]. Furthermore, it has been established, when the crystallographic direction of the lattice abruptly changes in the In-Si material system then stacking fault may generated. Especially, when two crystals parts begins to grow separately and then meet at certain point, where the crystallographic direction remains the same, but each side of the boundary has an opposite phase. These kinds of stacking fault can be formed due to the complex dynamics of the In-NPs migration as well as mixing of the Si-atoms from the top of the Si NWs as well as precipitated Si atoms too. Such complex scenario

*DOI: http://dx.doi.org/10.5772/intechopen.97723*

for Si NWs as compared to the bulk Si-wafer [35].

related to 9R (stacking ABCBCACAB…) [47].

will be explained later with the help of model given in **Figure 7**.

We realized that the In-NPs migration from the top of Si NWs toward the unused In-NDs and Si-substrate interface, where Si NWs are emanating from Si-substrate may cause to the stacking fault. We also know that an isolated defect like {111} faults have been observed to trap Au-atoms [34]. Therefore one cannot negate the possibility of In-atoms trapping by the planar defects of Si NWs during the VLS mode growth. **Figure 6a** gives the dark field-STEM image of the Si NWs, where one can clearly see the white spherical contrast of In-NPs around the side wall of the Si NWs. The compositional investigation by EDX taken at top of the Si NWs as well as taken from the pure In-NPs around the side wall of the Si NWs has been shown in the **Figure 6b**. The Kα X-ray energies for the In is 24.21 keV, and Lα X-ray energies for the same elements is 3.287 keV. Lα lines of the In-atoms can be separated, and this technique can be quantitatively used in a SEM. As shown in **Figure 6c** and **d**, we observed a reasonable mapping of Si and In sources, which originate from the Si NWs top and from the pure spherical In-NPs on the side wall of the Si NWs structure, respectively. The same phenomenon for Au NPs on the side of the Si NWs was observed by Krylyuk *et al.* [44]. Migration of In-NDs from the top of NWs was observed and confirmed by both dark field and bright field-STEM images. The coalescence of the migrated In-NPs with the unused In-NDs having high contact angle may resulted in relatively large size (~ 500 nm) of In-island as shown in the red color dotted marked line in **Figure 6a** and **d**, respectively. The generation mechanism will be explained in the proposed model given in **Figure 7**.

*Indium (In)-Catalyzed Silicon Nanowires (Si NWs) Grown by the Vapor–Liquid–Solid (VLS)… DOI: http://dx.doi.org/10.5772/intechopen.97723*

arrow, follow the same crystallinity of Si-substrate orientation marked at "P6" in **Figure 5b**. The diamond structure of the Si NWs with a lattice constant = 0.543 nm was confirmed by HR-TEM image and the SAED pattern at segments "P1" and "P3", which corresponds to the space group Fd3m grown along the 〈111〉 direction. Unfortunately, the SAED at "P2" and "P4" (taken from **Figure 4b**) of Si NW shows small tilt with respect to orientation of Si-substrate, and the same have been marked by red color arrow, as shown in HR-TEM image given in **Figure 5b**. Spotty pattern both at "P2" and "P4" segments of the NW were confirmed by SAED observation, as shown in the inset of **Figure 5b** and same value of small angular tilt of both "P2" and "P4" segments were observed. We concluded that the Si NWs are vertically aligned except the twining of planar defects, which might be caused by the faster cooling rate 100°C/6 min, as well as due to the longer exposer time of downside wall to the reactive radicals as compared to the upper side of Si NWs [46]. We have to consider about the thermal conductivity mechanism, which is different for Si NWs as compared to the bulk Si-wafer [35].

Here we focus to discuss about the Si NWs crystallinity at the interface of emanating Si NWs from the Si-substrate, as shown in **Figure 5a**. **Figure 5b** gives the SAED pattern at point "P5", which confirm the stacking fault in the grown Si NWs. The red arrow intensity originates from 1/2{111} spots is related to the 2H-polytype (stacking ABAB…) and the green arrow intensity originates from 1/3{111} spots is related to 9R (stacking ABCBCACAB…) [47].

Such structures are either attributed to scattering phenomena from two overlapping crystals with a stepped {111} twin boundary (parallel to the electron beam) [48, 49], or it might be the direct evidence of a 9R-polytype [23]. Furthermore, it has been established, when the crystallographic direction of the lattice abruptly changes in the In-Si material system then stacking fault may generated. Especially, when two crystals parts begins to grow separately and then meet at certain point, where the crystallographic direction remains the same, but each side of the boundary has an opposite phase. These kinds of stacking fault can be formed due to the complex dynamics of the In-NPs migration as well as mixing of the Si-atoms from the top of the Si NWs as well as precipitated Si atoms too. Such complex scenario will be explained later with the help of model given in **Figure 7**.

We realized that the In-NPs migration from the top of Si NWs toward the unused In-NDs and Si-substrate interface, where Si NWs are emanating from Si-substrate may cause to the stacking fault. We also know that an isolated defect like {111} faults have been observed to trap Au-atoms [34]. Therefore one cannot negate the possibility of In-atoms trapping by the planar defects of Si NWs during the VLS mode growth. **Figure 6a** gives the dark field-STEM image of the Si NWs, where one can clearly see the white spherical contrast of In-NPs around the side wall of the Si NWs. The compositional investigation by EDX taken at top of the Si NWs as well as taken from the pure In-NPs around the side wall of the Si NWs has been shown in the **Figure 6b**. The Kα X-ray energies for the In is 24.21 keV, and Lα X-ray energies for the same elements is 3.287 keV. Lα lines of the In-atoms can be separated, and this technique can be quantitatively used in a SEM. As shown in **Figure 6c** and **d**, we observed a reasonable mapping of Si and In sources, which originate from the Si NWs top and from the pure spherical In-NPs on the side wall of the Si NWs structure, respectively. The same phenomenon for Au NPs on the side of the Si NWs was observed by Krylyuk *et al.* [44]. Migration of In-NDs from the top of NWs was observed and confirmed by both dark field and bright field-STEM images. The coalescence of the migrated In-NPs with the unused In-NDs having high contact angle may resulted in relatively large size (~ 500 nm) of In-island as shown in the red color dotted marked line in **Figure 6a** and **d**, respectively. The generation mechanism will be explained in the proposed model given in **Figure 7**.

*Nanowires - Recent Progress*

ented Si NWs.

to the Si (111) surface as shown in **Figures 3d**–**f** and **4b**. The shape of the Si NWs were found to be cone like, and this can be explained by the similar analogy of tapering of the NWs grown by Au-catalyzed [44, 45]. Sharma *et al*. found that, when NWs elongate from the In-NDs (or In- droplets), the base area of the NWs remains exposed for a longer time to the reactive radicals when compared to the newly grown upper part of the NWs [46]. In the case of Au-catalyzed Si NWs grown in the 〈112〉 direction, where single twin planes and arrays of {111} stacking faults were reported [14], whereas Lopez *et al.* found the same structures for 〈111〉-oriented Au-catalyzed grown Si NWs [36]. These defects orientation is parallel to the 〈112〉 NW axis, often extending throughout the entire length of the wire. There is no report about the crystalline twining defects, planar defects and stacking fault along 〈112〉 direction in case of In-catalyzed vertically aligned, and (111)-ori-

We can see the planar defects as well as twining defects appearing in many segments of the Si NWs (sample-Nw) as shown in the inset of **Figure 5b** by green color arrow as well as red color arrow along the 〈112〉 direction. The (111)-oriented Si NWs at segments "P1–4" were also confirmed by HR-TEM image, as shown in **Figure 4b**. The corresponding selected area diffraction (SAED) shown in the in-set of **Figure 4b** has been recorded along the [110] zone axis from the Si NWs and which confirm the single crystal nature and its axial direction. But you can see twining of planes along 〈112〉 direction, which are distributed along the axial direction marked as S-I and S-II, as shown in **Figure 4b** for single (111)-oriented Si NW. It can be established that the HR-TEM image of single Si NWs along with the SAED at "P1" and "P3" of Si NWs (taken from **Figure 5b**), marked by green color

*(a) High-resolution cross-sectional TEM image around the Si-substrate and deposited In-NDs surrounded by Si NWs at the interface, (b) the corresponding SAED pattern taken along the [1* 1 *0] zone axis at point "P5" and "P6" respectively. In the inset of (b) is a SAED along [* 1 *10] zone axis at point "P5", where the red arrow intensity originates from 1/2{111} spots associated with the 2H polytype and the green intensity is originating from 1/3{111} spots associated with 9R stacking. Figure 5 reproduced with the permission from ref [43]***.** 

*Copyright 2016 the Japan Society of Applied Physics (JSAP).*

**126**

**Figure 5.**

#### **Figure 6.**

*(a) Dark-field-STEM images of Si NWs along the [1* 1 *0] zone axis to show the In-NPs as a white spherical contrast on the side wall of the Si NWs, (b) Elemental analysis of sample-Nw, (c) EDX intensity mapping of pure Si NWs, and (d) EDX intensity mapping of precipitated In-composition. Figure: 6 reproduced with the permission from ref [43]. Copyright 2016 The Japan Society of Applied Physics (JSAP).*

#### **Figure 7.**

*Schematic overview of the proposed model to explain step by step about the In-NPs migration from the top of the Si NWs and tapering as well as stacking fault formation in Si NWs, which in return restricted the length of the Si NWs. Figure 7 is reproduced with the permission from ref [43]. Copyright 2016 the Japan Society of Applied Physics (JSAP).*

**129**

*Indium (In)-Catalyzed Silicon Nanowires (Si NWs) Grown by the Vapor–Liquid–Solid (VLS)…*

Si-substrate during the crystal growth of Si NWs may not be clearly known yet but we can anticipate the mechanism of In-NPs coalescence with In-NDs and mixing

The In-catalyzed Si NWs grown by VLS mechanism confronted with planar defects, twining and stacking fault, which were observed by HR-STEM, as shown in **Figures 4a, b** and **5a, b**. To explain about the In-NPs migration to Si-substrate, Si NWs tapering, stacking fault, as well as limiting of NWs length by In-NPs migration, one simple model was anticipated. In this model, as a first step (Step-I), the In-NPs on Si(111), investigated, where most of the In-NDs got good wettability after Ar/H2-plasma treatment at 600°C with semispherical shape and few In-NDs were got spherical shape as shown in **Figure 2a** as well as in **Figure 3a** and the same is depicted schematically in **Figure 7**, Step-I. The spherical shape In-NDs may be confronted with thin oxide problems at the interface, shown in **Figure 7** (Step-I). During the Si NWs growth as a second step (Step-II) we observed that the In-NPs migration from the top of Si NWs toward the Si-substrate are taking place and mixing with already grown In-NDs having thin oxide interface layer between Si-substrate and In-NDs, as shown in **Figure 6a**–**d** and the same is depicted schematically in **Figure 7**. Due to the migration of In-NPs toward the Si-substrate some new contact points on Si-substrate were created and the already In-NDs size were

The supersaturation of the droplet, induced by the continuous gas phase supply of Si species, leads to the precipitation of Si nanowhiskers at the interface between the particle and the substrate, shown in **Figure 1**. Growth is obtained, as shown in **Figure 7**, Step-II, when a steady-state condition is reached between the flux of the Si through the In-NDs and the precipitation of Si on the Si-substrate [26, 37]. At the same time Si-atoms from sputtering source are also adsorbed on the same surface of In-NDs already deposited In-NDs on substrate having weak oxide interface layer, where the super saturation limit may be exceeded and now the precipitation of In-NPs were initiated as depicted in **Figure 7**, Step-II and the same has been confirmed in **Figure 6a**–**d**. Previously, the boron precipitation limit in BaSi2 were studied by in-plane and out of plane XRD, HR-STEM and TEM measurement and then successfully overcome the boron precipitation issue in bulk thin film *p*-BaSi2 layer [50, 51]. Altogether XRD out of-plane and in-plane characterization, SEM and HR-STEM observation could be a useful approach to investigate the precipitation of In-NPs in Si NWs. In Step-III, tapering of NWs are started due to the decreasing of In-NDs size at the top of Si NWs, which is caused by the In-NPs migration/ trapping from the top of the Si NWs as well as due to the longer exposer time of downside wall to the reactive radicals as compared to the upper side of Si NWs, as shown in **Figure 6a** and **c** [46]. At the same time the Si atoms adsorption by In-NDs are increasing because the In-NDs at the top of the Si NWs are decreasing. During such situation, the adsorbed Si-atoms by In-NPs cannot be precipitated to grow Si NWs due the partial oxide thin layer between the In-NDs and Si-substrate but instead support to sidewise expansion of the two neighboring In-NDs, shown in **Figure 6a**–**d** and the same has been depicted in Step-III, of **Figure 7**. At last, the two In-NDs coalesce on the surface of Si-substrate surrounded by Si NWs, as depicted in schematically in Step-III of **Figure 7**. Also, the deposition of the Si-atoms is increasing and at the same time the In-NPs migration are decreasing, which initiated the stacking fault shown in the inset of **Figures 4a, b** and **5a, b**. In Step-IV, finally the In-NPs migration is stopped as shown in **Figure 6a** as well as depicted in **Figure 7**. Subsequently, highly precipitated of In-NPs/NDs in the liquid phase are mixed with Si-atoms then finally give rise to stacking fault at the Si-substrate surrounded by

The reason of the In-NPs migration from the top of the Si NWs toward the

with precipitated Si-atoms via proposed model shown in **Figure 7**.

*DOI: http://dx.doi.org/10.5772/intechopen.97723*

enhanced or elongated.

*Indium (In)-Catalyzed Silicon Nanowires (Si NWs) Grown by the Vapor–Liquid–Solid (VLS)… DOI: http://dx.doi.org/10.5772/intechopen.97723*

The reason of the In-NPs migration from the top of the Si NWs toward the Si-substrate during the crystal growth of Si NWs may not be clearly known yet but we can anticipate the mechanism of In-NPs coalescence with In-NDs and mixing with precipitated Si-atoms via proposed model shown in **Figure 7**.

The In-catalyzed Si NWs grown by VLS mechanism confronted with planar defects, twining and stacking fault, which were observed by HR-STEM, as shown in **Figures 4a, b** and **5a, b**. To explain about the In-NPs migration to Si-substrate, Si NWs tapering, stacking fault, as well as limiting of NWs length by In-NPs migration, one simple model was anticipated. In this model, as a first step (Step-I), the In-NPs on Si(111), investigated, where most of the In-NDs got good wettability after Ar/H2-plasma treatment at 600°C with semispherical shape and few In-NDs were got spherical shape as shown in **Figure 2a** as well as in **Figure 3a** and the same is depicted schematically in **Figure 7**, Step-I. The spherical shape In-NDs may be confronted with thin oxide problems at the interface, shown in **Figure 7** (Step-I). During the Si NWs growth as a second step (Step-II) we observed that the In-NPs migration from the top of Si NWs toward the Si-substrate are taking place and mixing with already grown In-NDs having thin oxide interface layer between Si-substrate and In-NDs, as shown in **Figure 6a**–**d** and the same is depicted schematically in **Figure 7**. Due to the migration of In-NPs toward the Si-substrate some new contact points on Si-substrate were created and the already In-NDs size were enhanced or elongated.

The supersaturation of the droplet, induced by the continuous gas phase supply of Si species, leads to the precipitation of Si nanowhiskers at the interface between the particle and the substrate, shown in **Figure 1**. Growth is obtained, as shown in **Figure 7**, Step-II, when a steady-state condition is reached between the flux of the Si through the In-NDs and the precipitation of Si on the Si-substrate [26, 37]. At the same time Si-atoms from sputtering source are also adsorbed on the same surface of In-NDs already deposited In-NDs on substrate having weak oxide interface layer, where the super saturation limit may be exceeded and now the precipitation of In-NPs were initiated as depicted in **Figure 7**, Step-II and the same has been confirmed in **Figure 6a**–**d**. Previously, the boron precipitation limit in BaSi2 were studied by in-plane and out of plane XRD, HR-STEM and TEM measurement and then successfully overcome the boron precipitation issue in bulk thin film *p*-BaSi2 layer [50, 51]. Altogether XRD out of-plane and in-plane characterization, SEM and HR-STEM observation could be a useful approach to investigate the precipitation of In-NPs in Si NWs. In Step-III, tapering of NWs are started due to the decreasing of In-NDs size at the top of Si NWs, which is caused by the In-NPs migration/ trapping from the top of the Si NWs as well as due to the longer exposer time of downside wall to the reactive radicals as compared to the upper side of Si NWs, as shown in **Figure 6a** and **c** [46]. At the same time the Si atoms adsorption by In-NDs are increasing because the In-NDs at the top of the Si NWs are decreasing. During such situation, the adsorbed Si-atoms by In-NPs cannot be precipitated to grow Si NWs due the partial oxide thin layer between the In-NDs and Si-substrate but instead support to sidewise expansion of the two neighboring In-NDs, shown in **Figure 6a**–**d** and the same has been depicted in Step-III, of **Figure 7**. At last, the two In-NDs coalesce on the surface of Si-substrate surrounded by Si NWs, as depicted in schematically in Step-III of **Figure 7**. Also, the deposition of the Si-atoms is increasing and at the same time the In-NPs migration are decreasing, which initiated the stacking fault shown in the inset of **Figures 4a, b** and **5a, b**. In Step-IV, finally the In-NPs migration is stopped as shown in **Figure 6a** as well as depicted in **Figure 7**. Subsequently, highly precipitated of In-NPs/NDs in the liquid phase are mixed with Si-atoms then finally give rise to stacking fault at the Si-substrate surrounded by

*Nanowires - Recent Progress*

**128**

**Figure 7.**

**Figure 6.**

*Applied Physics (JSAP).*

*Schematic overview of the proposed model to explain step by step about the In-NPs migration from the top of the Si NWs and tapering as well as stacking fault formation in Si NWs, which in return restricted the length of the Si NWs. Figure 7 is reproduced with the permission from ref [43]. Copyright 2016 the Japan Society of* 

*(a) Dark-field-STEM images of Si NWs along the [1* 1 *0] zone axis to show the In-NPs as a white spherical contrast on the side wall of the Si NWs, (b) Elemental analysis of sample-Nw, (c) EDX intensity mapping of pure Si NWs, and (d) EDX intensity mapping of precipitated In-composition. Figure: 6 reproduced with the* 

*permission from ref [43]. Copyright 2016 The Japan Society of Applied Physics (JSAP).*

**Figure 8.** *Schematic Si NWs/c-Si tandem solar cells with expected efficiency of 30% (1sun).*

the neighboring Si NWs and the same has been confirmed in the inset of **Figure 5b** at "P5". Probably at this situation the sticking coefficient of Si-atoms are reduced, and no further Si-atoms can be absorbed by In-NDs island. This is the time where all In-NDs at the top of Si NWs might be disappeared and give rise to stacking fault on the surface of Si-substrate surrounded by Si NWs, shown in **Figure 6a** as a white contrast (red dotted line). As a result, Si NWs growth were stopped, and we can safely say that the tapering of the Si NWs and length of the Si NWs were restricted by In-NDs disappearance or migration from the top of the In-catalyzed Si NWs.

Suppression of the In-NPs migration from the top of the Si NWs are essential to grow longer Si NWs as well as to avoid NWs tapering and also to fix the stacking fault in Si NWs. We need to find new growth condition to suppress the In-NPs migration to find the lower optimal substrate temperature without compromising on the verticality control of Si NWs. The ultra-clean interface between Si-substrate and In-NDs is essential to get smaller θC of the In-NDs to increase its wettability on Si-substrate, without any oxidation issues prior to Si NWs growth. We have to try with longer Ar/H2 plasma treatment time, by using low plasma power to avoid any kind of surface damage to the Si-substrate to further reduce the θC. Our new growth condition can be used to grow vertically aligned Si NWs using In-catalyzed in VLS mode for many electronic and nano device applications [52]. Solar cell architecture of the 4-terminal based wide bandgap top cell (Si NWs) and narrow bandgap bottom cell for best matching efficiency with 10% Ge in SiGe active layer is possible, where the photocurrent limits under the solar spectrum for varying band gap of SiGe materials due to the Ge composition for bottom Heterojunction solar cell applications can be achieved [53]. Heterojunction light-emitting diodes (LEDs) comprising p-type Si nanowires (p-Si NWs) and n-type indium gallium zinc oxide (n-IGZO) were successfully fabricated [54]. Band gap energy of Si NWs can be controlled around 1.7 eV by changing the diameter of the NW [3]. Quite high conversion efficiency around 30% is expected in the Si NWs/c-Si tandem solar cells structure, as shown in **Figure 8**.

#### **4. Summary**

Using In-catalyzed VLS mode growth, we have successfully controlled the verticality and (111)-orientation of Si NWs and ultimately scaled down the diameter of NWs to 18 nm. The density of vertically aligned Si NWs was enhanced from 2.5 μm−2 to 70 μm−2. During the *in situ* sequential deposition of In-NDs catalyzed Si NWs in high vacuum environment has successfully blessed us with vertically aligned and (111)-oriented Si NWs arrays using VLS mode. Using the HR-TEM, HR-SEM, and EDX, the planar defects as well as twining defect structure, which is grown perpendicularly to the Si-substrate (along 〈112〉 Si-NW direction) to

**131**

**Author details**

M. Ajmal Khan1

m\_ajmal\_khan@yahoo.com

provided the original work is properly cited.

\* and Yasuaki Ishikawa<sup>2</sup>

\*Address all correspondence to: muhammad.khan@riken.jp;

1 RIKEN Cluster for Pioneering Research (CPR), Wako, Saitama, Japan

2 Department of Electrical Engineering and Electronics, College of Science and Engineering, Aoyama Gakuin University, Sagamihara-shi, Kanagawa, Japan

© 2021 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/3.0), which permits unrestricted use, distribution, and reproduction in any medium,

*Indium (In)-Catalyzed Silicon Nanowires (Si NWs) Grown by the Vapor–Liquid–Solid (VLS)…*

the axial direction (111)-orientation in many parts of the individual Si NWs were found. Stacking fault were confirmed, where the intensity originates from 1/2{111} spots associated with the 2H polytype and also the intensity originates from 1/3{111} spots associated with 9R stacking in the emanating NW crystal structure at the interface of Si-substrate. These kinds of studies could greatly influence the future understanding about the phase purity, crystallinity and controlled growth of In-catalyzed Si NWs, especially the length and periods of vertically aligned Si NWs

This work was partially supported by the MEXT, FUTURE-PV Innovation (FUkushima Top-level United Center for Renewable Energy Research–Photovoltaics

for the Nano device applications including solar cells and LEDs.

*DOI: http://dx.doi.org/10.5772/intechopen.97723*

**Acknowledgements**

Innovation) Project.

*Indium (In)-Catalyzed Silicon Nanowires (Si NWs) Grown by the Vapor–Liquid–Solid (VLS)… DOI: http://dx.doi.org/10.5772/intechopen.97723*

the axial direction (111)-orientation in many parts of the individual Si NWs were found. Stacking fault were confirmed, where the intensity originates from 1/2{111} spots associated with the 2H polytype and also the intensity originates from 1/3{111} spots associated with 9R stacking in the emanating NW crystal structure at the interface of Si-substrate. These kinds of studies could greatly influence the future understanding about the phase purity, crystallinity and controlled growth of In-catalyzed Si NWs, especially the length and periods of vertically aligned Si NWs for the Nano device applications including solar cells and LEDs.
