Experimental

The AlN nanotips and nanorods were coated with silver by the ion beam sputtering (IBS) process. A base pressure of 1×10^-6 Torr was achieved in the

sputtering chamber that contained a high purity silver target, mounted on a rotating stage, and the AlN NTs and NRs substrates held about 10-15 cm away from the target.

An argon (Ar) ion beam (Ar flow rate of 2 sccm) was accelerated toward the silver target using a Commonwealth Scientific (CS) IBS controller. The sputtered silver was

collected on the substrate, kept at room temperature, for different durations of time.

The working pressure was about 5×10^-4 Torr.

Commonly used molecule for SERS experiment such as Rhodamine 6G was then

solution and dried. They were then used for the Raman measurements using a Renishaw-2000 micro Raman spectroscope. Single accumulations with integration

times of 5s were used to collect the spectra in a backscattered mode. A 532 nm laser with a source intensity of 100 mW was used with a probe beam diameter of 1 µm.

7.2 Results and discussion

Scanning electron microscope (SEM) images of the deposit reveal AlN nanorods to be sharply faceted quasi-aligned nanorods with a hexagonal cross-section. Figure 7.1a shows the SEM images of the as-grown AlN nanorods. Fig. 7.1b displays the SEM image of Ag-coated AlN nanorods. Fig. 7.1c shows the Ag-coated AlNNTs, and Fig.

7.1d exhibit the high magnification SEM image of Ag nanoparticles coated on AlN nanotips.

FIG. 7.1 (a) Cross section and top view (inset) SEM image of quasi aligned AlN nanorods; (b) Top view SEM image of Ag-coated AlN nanorod; (c) Cross section view of Ag-coated AlN nanotips; (d) High magnification SEM image of Ag-coated AlN nanotips; (e) X-ray diffraction spectrum of as-grown and Ag coated AlN

The XRD spectra of the as-grown AlN nanotips and nanorods shows strong signatures of h-AlN phases ((100), (002) and (101)) (JCPDS 25-1133) along with unreacted pure Al (JCPDS 04-0787) residue. In addition to the XRD spectrum, the selected area electron diffraction (SAED) pattern (top inset, Fig. 7.2a), and electron energy loss spectroscopy (EELS) (bottom inset, Fig. 7.2a), carried out during the transmission electron microscopy (TEM) measurements conclusively prove that they are AlN in composition having hexagonal symmetry. Figure 7.2a shows a TEM image of an Ag-coated AlN nanorod. Figure 7.2b shows the lattice image of the as grown AlN nanorods with a d₍₀₀₂₎ = 0.252 nm, matching that of h-AlN. The growth direction of the single crystalline h-AlN nanorods were evidently [002].

The h-AlN nanotips and nanorods provide a surface, with suitable interface energy and lattice mismatch that promotes the generation of nc-Ag, instead of a thin film, under an incident flux of Ag vapors in a sputtering chamber with specified deposition

parameters. Argon ion beam sputtering [209] of a Ag target for 10 minutes under room temperature and 5×10^-4 Torr pressure, used in this case, produced a single

overcoat of nc-Ag on the entire surface of the h-AlN nanotips and nanorods. The sizes of the nc-Ag are mostly ~5 nm in diameter but not exceeding 10 nm, having an extremely high packing density such that they touch each other (Fig. 7.1b and 1d).

Note that the packing density of the nc-Ag, per unit probe volume, is further enhanced

due to the rough nanostructured template compared to a flat template surface. Figure 7.2c shows the SAED pattern of the AlN nanorods with a post synthesis coating of sputtered Ag. Clear and sharp diffraction spots, which are signatures of h-AlN, were observed in the SAED pattern along with sharp ring patterns arising from nc-Ag (Fig.

7.2c). The sharp ring features present in the SAED pattern (Fig. 7.2c), but absent in Fig. 7.2a (top inset), confirms the presence of nc-Ag in the Ag coated samples.

However, crystalline Ag signal cannot be detected unambiguously from the XRD spectrum of the Ag coated h-AlN nanostructures shown in Fig. 7.1c mainly due to a small cross-section of the nc-Ag to the incident X-rays and the close proximity of the Ag (JCPDS 03-0921) and unreacted Al signals.

FIG. 7.2 (a) TEM image of a single Ag-coated AlN nanorod (the arrows showing the nc-Ag); top and bottom inset shows the SAED pattern and the electron energy loss spectroscopy (EELS) spectrum of the as-grown AlN nanorod, respectively; b) HRTEM image of the as-grown AlN nanorod lattice; c) SAED pattern of the Ag-coated AlN nanorod.

The Raman spectrum of the as-grown h-AlN nanostructures excited by a 532 nm laser is shown in Fig. 7.3a (AlNNRs) and 3d (AlNNTs). Measured quantities (2-5 µL)

of micro-molar (10^-6 M) solutions of Rhodamine 6G (R6G) in methanol when dropped on these as-grown h-AlN nanostructures show only a broad and strong fluorescence characteristic of R6G as shown in Fig. 7.3b (AlNNRs) and 3e (AlNNTs). The nc-Ag coated h-AlN nanorod surface demonstrates SERS of 10^-6 M R6G solution. The generally fluorescent R6G molecule has its fluorescence quenched when adsorbed on the nc-Ag present on the h-AlN nanostructures template and the Raman signals are promoted as demonstrated in Fig. 7.3c (AlNNRs) and 3f (AlNNTs).

Assuming 100% adsorption of the R6G molecules on the nc-Ag, a minimum enhancement in the range of ~2×10⁶ was calculated [220] for the Raman

cross-section.

FIG 7.3 Raman spectrum of (a) the as-grown AlN nanorods, (b) as-grown AlN

nanorods with R6G, and (c) SERS spectrum of Ag coated AlN nanorods with R6G, (d) as-grown AlN nanotips, (e) as-grown AlN nanotips with R6G, and (c) SERS spectrum of Ag coated AlN nanotips with R6G.

7.3 Summary

In final, we have demonstrated the synthesis of quasi-aligned h-AlN nanostructures from pure Al powders heated to 900 and 1200°C in presence of ammonia. Ion beam sputtered silver self-assembles as nanocrystals on these h-AlN nanostructures and readily exhibits surface enhanced Raman scattering. Minimum SERS cross-section enhancement factors in the range of 2 × 10⁶ have been obtained.

Chapter 8

Field emission from quasi-aligned AlNNTs

One-dimensional (1-D) nanostructures have recently brought significant research activities for its unique properties in different applications such as atomic-resolution scanning tunneling microscopy [221], near field optical microscopy [222], and high-resolution atomic force microscopy [223]. At the same time, wide band-gap materials have been promising as cold-cathode materials [224,225] due to their low electron affinities as well as their excellent thermal, chemical, and mechanical stability. Recently, the field emission characteristics of few 1-D wide bandgap semiconductors, such as well-aligned ZnO nanowires [226], SiC nanowires [227,228], SiCN nanorods [229], GaN nanowires [138], and diamond-like nanostructure [230], have been widely investigated and reported. Superior field emission performance can be achieved by adjusting the aspect ratio and spacing of the nanomaterials to optimize the enhancement of the local electric field at the emitting sites. It is clear that sharper tips give higher field enhancement factor which provides higher field emission current.

One dimensional nanostructures including nanotips are interesting systems to investigate the field emission phenomenon [145,146,156]. The inherent current or

concern over the stability of the process and the nanostructures itself. Wide bandgap materials such as AlN may provide important clues or answers to this problem and that remains the focus of this report. According our previous observations, the hexagonal and faceted AlNNTs always reveal a sharp end with point tip. This geometrical feature and the intrinsically low electron affinity nature make AlNNTs an emitter material of potential. In this chapter, the field emission properties of the quasi-aligned AlNNTs grown on p⁺-, p-, n-, and n⁺- type Si substrates were studied.

Section 8.1 will introduce the experimental section including the sample and parameters for field emission measurement. The results of geometrical feature and field emission of AlNNTs are discussed in section 8.2. Finally, section 8.3 will make a summary.

8.1. Experimental

The sample of AlNNTs and AlNNRs for field emission measurement were grown on gold coated p⁺-type Si (p⁺-Si) (resistivity ~ 2-5 mΩ cm), p- (resistivity ~ 10 Ω cm), n- (resistivity ~10 Ω cm), and n⁺-Si (resistivity ~ 2-5 mΩ cm) Si substrates, respectively via APCVD. Field emission measurements were carried out using parallel-plate configuration under a base pressure of ∼1 × 10^-7 Torr. The schematic drawing of FE is shown in Fig. 8.1. A Keithley 237 high voltage source-measure unit

was used for sourcing the voltage (up to 1100 V) and measuring the current (with pico-ampere sensitivity). The quasi-aligned AlN nanostructures served as the cathode and an indium tin oxide (ITO) coated glass was used as the anode for our FE measurement. The effective emission area is defined by the patterned ITO anode (emitting area 0.06 cm²). The distance between these two electrodes could be controlled, and kept at 70 µm.

8.2. Results and discussion

Figure 8.2 a-b display the field emission scanning electron microscope image of the AlN nanotips grown on p⁺-type Si (p⁺-Si) substrate, and AlN nanorod grown on n⁺-type Si (p⁺-Si) substrate (Fig. 8.2 c-d). All AlN nanostructures prepared on differently doped Si substrates have very similar morphologies to each other. It was found that AlN nanostructures could be successfully synthesized depending only on the growth parameters, such as temperature, reaction time and gas flow rate but independent of the dopants in the Si substrate. Typical cross-sectional SEM image of AlN nanotips and nanorods are shown in Fig. 8.2(b) and (d). This photograph displays a high density of quasi-aligned nanotips and well aligned nanorods uniformly distributed over the entire substrate. It should be noted that all the nanotips exhibited the same self-selective apex angle, indicating the formation of nanotips was subjected to a surface energy controlled process [166].

FIG. 8.2 Typical FESEM (a) top view, (b) cross section images of AlN nanotips ; (c) top view, (d) cross section images of AlN nanorods grown on p⁺-Si substrate.

The typical field emission current density (J), which is defined as the current collected per unit area of the anode, vs. applied field (E) characteristic of AlNNTs and AlNNRs is depicted in Fig. 8.3. The quasi-aligned AlN nanotips served as the cathode and an indium tin oxide coated glass was used as the anode for our FE measurement.

The anode area is approximately around 0.25 cm². The anode-cathode distance was kept at 70µm. If a turn-on field (Von) for field emission is defined as the field required in extracting a current density of 10 µA/cm² [232] then Von for samples grown on p⁺ (resistivity ~ 2-5 mΩ cm) and p type (resistivity ~ 10 Ω cm) Si substrates were 6.0 and 8.5 V/µm, respectively. In contrast, it is difficult to obtain any field emission current of the AlN nanotips grown on n- (resistivity ~10 Ω cm) and n⁺-Si (resistivity ~ 2-5 mΩ cm) substrates as well as the AlNNRs grown on p⁺-Si. No emission current observed from AlNNRs may be due to the surface field declining attributed to high density of nanorods and flat surface. Emission currents higher than 0.22 A/cm² are observed on the case of AlNNTs grown on p⁺-Si while the applied field exceeds 10 V/µm. The obtained turn-on field of 6.0 V/µm is comparable to the lowest Eto values reported so far for CNTs (0.6 − 2.8 V/µm) [232-234], N-doped diamond film (~1.5 V/µm) [235] and amorphous carbon film (6.0 V/µm) [236]. According to the F-N plot (see the inset), a linear relation between ln(J/E²) and 1/E is observed, indicating that the field emission is intrinsically driven by the electric field. The reproducibility of

the FE properties was verified for the AlN nanotips grown on p⁺- and p-Si substrates by performing repeated cycles of measurements and is shown in Fig. 8.3(b). For practical application as an emitter material in flat panel display, emission current stability of AlNNTs has been tested under high vacuum conditions for extended periods of time. The stability study was done by extracting a constant current density of 100 µA/cm², which is 10 times larger than the minimal FED operation current of 10 µA/cm², continuously over a period of 10 hours and monitoring the flat plate electric field required for the purpose. Figure 8.3(b) demonstrates the stability of electron emission from the AlN nanotips grown on p⁺ and p-Si substrates. For AlN nanotips grown on p⁺ and p-Si substrates, the applied electric field varied by only about less than 5% and 10%, respectively, of its original value to maintain the 100 µA/cm² current density, it is manifesting the highly durable and robust capability of the AlNNTs. An acceptable stability of the quasi-aligned AlN nanotips was demonstrated.

However, the instability in the applied field at 4 hrs into the experiment probably indicates some damage, under such high current extraction (current stressing), in the substrate-nanotip interfacial area resulting in a reduced density of the emitters. The reduced density of the nanotips can reduce the effective field screening at the tips and bring down the required field for a specific current extraction.

FIG. 8.3 (a) Field emission characteristic curves for quasi-aligned AlN nanotips emitter (emitting area 0.06 cm²) grown on (▲) p⁺-type Si, (▽) p-type Si, (▓) n⁺-type Si and (○) n-type Si, respectively. (★) is the AlNNRs grown on p⁺-type Si. Inset shows the F-N plot for field emission of the AlN nanotips grown on (▲) p⁺ and (▽) p-type Si; (b) Emission stability of AlN nanotips grown on (▲) p⁺ and (▽) p-type Si substrates, where the emission current was kept at a constant value of 100µA/cm².

Field emission mechanism from supported nanostructures should consider electron injections from the two interfaces namely the substrate-emitter interface and the emitter-vacuum interface. Whereas the emitter-vacuum interface is widely studied and reported, the substrate-emitter interface was rarely studied. In 1996, a simplified electron emission model, proposed by Choi et al. [237], on a metal-diamond-vacuum system, which considered tunneling of electrons through the Mo-diamond Schottky barrier into the conduction band by applying an electric field. A similar model was proposed by Geis et al [238]. Furthermore, Snow et al. [239], proposed the model of electron emission mechanism based on transport across both interfaces, one is substrate-diamond interface, and the other is diamond-vacuum interface. It has been reported that the emission performance of diamond is often limited by the injection of electrons into the diamond through the metal-diamond interface. Therefore, the quantity of electrons injected from the substrate to the wide bandgap material is the determining step for the field emission of the wide bandgap materials.

To explain the electron transport phenomenon across the substrate-emitter interface, we revert to semiconductor heterojunctions. The analogy of the present system, AlN on Si, with that of a semiconductor heterojunction is valid owing to the large difference in their bandgaps and resistivities. The AlN material has a high electrical

resistivity and low bandgap (1.1 eV) of the Si substrates. The work function for AlN was found to be 4.3 eV [241] whereas that for n-type Si is ~ 4.15 eV from an independent ab-initio calculation [242]. The electron affinity (χ) of wide bandgap AlN is 0.6 eV [241] and less than that of the low bandgap Si (χ = 4.01 eV) [242]. Hence for Si-AlN junction, a “straddled” heterojunction was obtained wherein the bandgap of the AlN completely overlaps the bandgap of Si. After Si and AlN forms a junction and thermal equilibrium is established, band bending occurs at the interface (Fig. 8.4).

Due to a very large separation in the Fermi levels of n-type Si (EFN) and AlN (EF) a large number of electrons flow into AlN for Fermi level alignment across a Si (n-type)-AlN junction. This will bring about a severe band bending at the interface and a higher barrier for further electron movement from Si to AlN. On the other hand, for Si (p-type)-AlN heterojunction, the Fermi level in Si (EFP) is slightly lower than that in AlN (EF), causing a flow of holes from p-type Si to AlN to obtain Fermi level alignment. Therefore, the band banding at the Si(p-type)-AlN interface results in a

“well”, instead of a barrier, allowing an easier electron flow across this junction under an applied field. This effect will be more pronounced for a p⁺-type Si-AlN interface since an even narrower separation exist between the EFP and EF and will lead to a more efficient electron tunneling into AlN. Therefore, AlN nanotips grown on p-type and p⁺-type substrates have higher emission currents than those grown on n-type and

n⁺-type Si substrates. This phenomenon is very similar to that of the field emission properties of one dimensional wide bandgap SiCN nanostructure where no gold catalyst was used [243]. The field emission results we obtained on differently doped substrates were not dominated by any possible gold doping which, if present, would have been identical for all the samples. The contrasting field emission results can only result from the donor/acceptor doping in the substrates which was the only distinguishing factor for the samples studied.

FIG. 8.4 Representative band diagram for (a) Si (n-type)-AlN and (b) Si (p-type)-AlN heterojunctions, before and after thermal equilibration. EC, EV, EF stands for the conduction band, valence band and Fermi level, respectively, of AlN. ECN, EVN, EFN stands for the conduction band, valence band and Fermi level, respectively, of n-type Si and ECP, EVP, EFP stands for the conduction band, valence band and Fermi level, respectively, of p-type Si.

8.3 Summary

In final, FE properties of AlN nanotips and nanorods were studied. The effect of the p- and n-Si substrates in enhancing or limiting the field emission from the Si-AlN interface was demonstrated. The results of improved field emission of AlN on p⁺- and p-Si substrates were explained based on band diagram considerations in semiconductor heterojunctions. The existence of an energy barrier at the interface of n-/n⁺-type Si and AlN attenuated the electron flow whereas the absence of the same eliminated any threshold for electron flow across the p-/p⁺-type Si and AlN interface.

A stable field emission from the AlN nanotips was observed over a period of 10 hours.

Due to the nanosize structure and stable nature, the AlN nanotips show great potential for field emission device application.

Chapter 9

Conclusion

Via the technique of horizontal-flow hot-wall atmosphere pressure chemical vapor deposition (APCVD) using highly pure aluminum and ammonia gas as the CVD precursor, conductive AlN 1D nanocrystals, including NTs and NRs. A detailed characterization focusing on the morphology, structures, orientations and compositions of the various AlN samples have been carried out by means of FESEM, HRTEM, SAED, EELS, XRD, and Raman scattering.

The vertical quail-aligned and highly packing density AlNNTs are prepared via the APCVD technique. The AlNNTs have six-sided hexagonal geometry and sharp tip.

The effects of variety types and thickness of metal coating on the morphology and size have been observed via FE-SEM. The size of AlNNTs can be controlled by adjusting the thickness of the Au layer. For deposition on thin (7 nm) Au-coated Si substrate, AlNNTs exhibited a mean diameter of 10 nm at the apex, 80 nm at the base, and 250 nm in length whereas NTs deposited at thicker Au layers (15-50 nm) produced longer nanotips (300-3000 nm) with wider apex (20-100 nm) and base (100-700 nm) diameters. Furthermore, AlNNTs can be prepared on Al and Pt coated substrates, even on bare Si (without metal coating). It indicated that metal coating is

not indispensable for the growth, but it does help in controlling the morphology of the AlNNTs. The Al nanocrystals nucleated within the first 20 minutes of growth and served as the base for the AlNNTs formation. AlNNTs start to germinate at the first 25 minutes of the growth and complete formation in 25-30 minutes. TEM and SAD measurements indicate that these hexagon tips have their long-axis along [002]

direction. XRD patterns show that AlNNTs exhibit hexagonal crystal symmetry with a preferred growth direction of (002). The growth mechanism of AlNNTs is supposed to be different from ordinary vapor-liquid-solid or catalyst-mediated growth mechanisms.

AlNNTs possess a stable plane (namely (221)) with 81° angle with respect to the basal plane. The “nongrowth” surface model was proposed to be the main mechanism leading to the tip formation.

By designing a series of APCVD experiments, a morphological evolution from tip-like AlN to rod-like NRs has been observed as increasing the reaction temperature from 950 to 1200℃. The structural characterization by TEM, SAED and XRD shows the single-crystalline quality and an identical preferred c-axis growth direction for the AlN NTs and NRs. The diffusion mediated growth model incorporating an Ehrlich-Schwoebel barrier was proposed to explain the formation of high aspect ratio nanotips and nanorods.

observed in optical properties measurements. Spontaneously absorption from host-lattice and a serious of oxygen-related defect levels were obtained from excitation spectra and UV absorption technique. Luminescence model was adapted to describe the luminescence features of AlNNTs.

The use of AlN nanostructures as nano-templates for SERS experiments were carried out by Ag coating. Nc-Ag particles are well dispersed on the h-AlN nanotips

and nanorods with the mean diameter of 5-10 nm, exhibitng minimum SERS

and nanorods with the mean diameter of 5-10 nm, exhibitng minimum SERS