Enhanced field emission properties based on In–In₂O₃ composite nanopagodas

Abstract

In–In₂O₃ composite nanopagodas demonstrate excellent field emission properties with an enhanced kinetics factor (saturation ratio). The turn-on fields and β of the In–In₂O₃ nanopagodas were 2.0 V μm⁻¹ and 3590. The indium of the In–In₂O₃ nanopagodas accumulates electrons and forms electron tunnels, which enhances the field emission properties.

---

Introduction

Recently, In₂O₃ has attracted much attention due to its wide direct bandgap energy of 3.4–3.8 eV,^1,2 and it is now widely used in field emission, gas sensors and solar cells.^3–5 It is well known that the optical and electronic properties of metal oxide semiconductors can be modulated by varying the morphology or doping with a variety of elements. For the morphology of In₂O₃, various one-dimensional (1D) In₂O₃ nanostructures have been synthesised, including nanotubes, nanowires (NWs), nanobelts and nanorods, etc.^6–8

1D nanostructures (NSs) have attracted considerable interest for their potential application in field emission (FE) because of their sharp tips and high aspect ratios, which cause their high field enhancement factors (β).⁹ Li et al. reported electric-field-aligned vertical and random growth of In₂O₃ NWs.¹⁰ Kar et al. reported nanopyramids, nanocolumns and NWs of In₂O₃.¹¹ Thus, investigation of the FE properties of In₂O₃ 1D nanoforms is a key factor.

According to Huang et al., four kinds of novel layered In₂O₃ NS were synthesized by controlling the kinetics factor (saturation ratio).¹² In this paper, two kinds of In₂O₃ NS were synthesized by changing the direction of the substrate. In addition, by decreasing the distance between the metal source and the substrate, we demonstrate the first successful fabrication of In–In₂O₃ conformable nanopagodas (NPs). The FE properties of the In–In₂O₃ NPs was better than those of In₂O₃ NWs.

Experimental

A 500 nm-thick silicon dioxide (SiO₂) insulating layer was deposited onto a Si (111) substrate by wet oxidation. A 3 nm-thick Au film as the catalyst layer was deposited onto the SiO₂/Si substrate using an electron-beam evaporator. Thermal vapor deposition was performed for the vapor–liquid–solid (VLS) growth of the In₂O₃ NSs. About 0.3 g of In powder (99.993%) was poured into an alumina bowl. The bowl, filled with powder, was placed in the center of an alumina boat. The Au/SiO₂/Si substrate was placed next to the alumina bowl, and the substrate that was positioned upon the alumina boat was reversed. The argon flow rate was maintained at 54.4 sccm and the oxygen flow rate was 0.3 sccm. During the growth of In₂O₃, the quartz tube pressure, growth temperature, and growth time were maintained at 8 Torr, 1000 °C, and 30 min, respectively. For an enhanced saturation ratio, we changed the placement of the substrate beside and upon the indium metal source bowl in a horizontal quartz furnace.

The morphologies and size distributions of the as-grown In₂O₃ NSs were characterized using a field-emission scanning electron microscope (FESEM) (JSM-7000F), which was operated at 10 keV. The crystallographic and structural properties of the as-grown In₂O₃ NSs were then determined using a MACMXP18 X-ray diffractometer. The field emission (FE) properties of these In₂O₃ NSs were measured at room temperature in a vacuum chamber at a pressure of 4 × 10⁻⁶ Torr. During the FE measurements, the anode was a tungsten probe with a diameter of 1.8 mm, while the inter electrode distance (d) between the anode and the In₂O₃ NSs was precisely maintained at 150 μm. A lamp was placed above the Si substrate as an excitation source for the In₂O₃ NSs. The current–voltage (I–V) characteristics were determined using a Keithley 237 high-voltage source and measuring unit, which measured in femtometers and provided a maximum voltage of 1000 V. The FE properties were extracted from the obtained I–V curves using the Fowler–Nordheim (F–N) model.

Results and discussion

Fig. 1(a) shows a cross-sectional FESEM image of the front of the Au/SiO₂/Si substrate with the morphology of In₂O₃ NWs. The diameter and length of the In₂O₃ NWs were 100 nm and 12–18 μm. To enhance the vapor pressure of the metal, we reversed the location of the substrate above the indium power, and got the In–In₂O₃ NPs, as shown in Fig. 1(b). The diameter and length of the In–In₂O₃ NPs were 0.1–1 μm and 3–6 μm. According to the energy-dispersive X-ray (EDX) spectrum of the In₂O₃ NWs and In–In₂O₃ NPs, the In : O ratios were 1.8 : 3 and 2.3 : 3, as shown in Fig. 1(c) and (d). This showed that the content of In in the In–In₂O₃ NPs was higher than that in the In₂O₃ NWs.

Fig. 1 FESEM images and EDS of In₂O₃ NSs synthesized at 1000 °C with the position of the substrate ((a) and (c)) beside and ((b) and (d)) upon the precursor.

Lattice structure analysis was performed using XRD measurements. Fig. 2 shows the XRD 2θ-scan spectra. The black line represents the XRD result of the In₂O₃ NWs we synthesized before and reveals a body-centred cubic (bcc) structure with a lattice constant of a = 10.09 Å, as characterized by JCPDS card no. 65-3170. Other peaks that are also assigned to the In₂O₃ structure according to the JCPDS database are present and indicated in the spectrum. These weaker peaks were thought to be the result of thin layers or particles depositing on the surface during the initial growth process. The red line represents the XRD result of the In–In₂O₃ NPs. The (222) peak is strengthened in In–In₂O₃ NPs. The noticeable strengthening of the (222) peak is ascribed to lateral 0-D growth of the In–In₂O₃ NPs, as described in the previous growth process. As shown in the spectra, the other difference is that there is a peak assigned to pure indium by JCPDS card no. 85-1409. Indium metal really existed in the In–In₂O₃ NPs, and it is worth pointing out that the other peaks present for the In–In₂O₃ NPs have no shift compared with the database, which leads us to the conclusion that the In₂O₃ NPs fabricated in this study are characteristic of the fine crystal structure.

Fig. 2 X-Ray diffraction spectra for the In₂O₃ NWs and In–In₂O₃ NPs.

During the production process of the In₂O₃ NWs, the substrate was placed next to the indium powder boat, and the Au film on the SiO₂/Si substrate would form Au nanoparticles at around 500 °C. The Au nanoparticles acted as the catalysts for VLS growth. Due to the large amount of indium vapor reacting with oxygen, the oxidized indium vapors had a large supersaturation ratio. Research has shown that the supersaturation ratio is inversely proportional to the critical core radius.¹³ Additionally, eutectic droplets acted as absorption sites for the oxidized indium vapors and confined these precipitates to one-dimensional growth. At higher temperatures, indium vapors would be notably generated and abundant reagent species would be adsorbed onto the droplets. This could lead to the rapid precipitation of solid In₂O₃ NSs, and the reaction would take place at the liquid–solid interface maintaining the 1-D growth.

The synthesis of the In–In₂O₃ NPs during the growth process was carried out as illustrated in Fig. 3, and three different mechanisms were ascribed to the formation of the In–In₂O₃ NPs.

Fig. 3 Growth process of In–In₂O₃ NPs.

(I) To begin with, with the evaporation of liquid indium, some indium vapors would directly deposit and alloy with the Au nanoparticles before oxidizing due to the short distance between the substrate and the precursor. As a result, the composition of the eutectic droplet would mainly be indium, which plays a catalytic role in the growth mechanism. The abundant In vapors react with oxygen as the temperature rises and continuously dissolve into the alloy droplet. Once the vapor pressure of these adsorbed species in the droplet reaches supersaturation vapor pressure, the consequent 1-D growth mechanism of solid-state In₂O₃ would start, as shown in Fig. 3(a).

(II) As the growth time increases, the gradual decrease in the saturation ratio of the reagent species would slow down the rate of 1-D growth. It is at this stage that the continuous indium vapor that evaporates from the precursor would be oxidized and adsorbed onto the surface of the 1-D In₂O₃ NS. Considering the low saturation ratio of the adsorbed species in the droplet, this is too low to maintain the mechanism of 1-D growth. As a result, these abundant oxidized indium vapors would favor 0-D growth, followed by a vapor–solid (V–S) mechanism.¹⁴ Subsequently, lateral growth would carry on, as shown in Fig. 3(b).

(III) During the 0-D growth process, there were some oxidized indium vapors dissolved into the catalytic droplet at the top. Moreover, the reagent species adsorbed on the lateral surface might also have diffused into the droplet. As a result, the saturation ratio would rise gradually and reach the supersaturation vapor pressure. The VLS mechanism then dominated the growth process once again. As a result, most of the adsorbed species would transfer to maintain the 1-D growth, as Fig. 3(c) shows.

(IV) Again, if the saturation ratio was gradually lowered, it would convert back to the 0-D growth mechanism as described in step II, and lateral growth would restart as shown in Fig. 3(d). Based on the varied saturation ratio of the droplet, the whole growth process showed a competitive phenomenon between longitudinal 1-D growth and radial 0-D growth, as shown in Fig. 3(e). It is worth mentioning that even when 1-D growth was dominant there were still some reagent species diffusing to the lateral surface. Consequently, this caused uninterrupted lateral growth during the whole growth process, so if the lateral layer of the NPs was closer to the bottom it would have a large diameter because of the longer growth time.

Fig. 4 plots the FE current density (J) of the In₂O₃ NWs as a function of the applied electric field (E). The turn-on field (E_to) is defined as an electric field producing a current density of 10 μA cm⁻². The E_to of the In₂O₃ NWs and In–In₂O₃ NPs are 2.9 and 2.0 V μm⁻¹. The ln(J/E²) was plotted as a function of 1/E in the F–N plot in the inset of Fig. 4. The FE properties were analyzed by F–N theory to understand the emission characteristics, using the following equation:

where E is the applied field; V represents the applied voltage; Φ is the work function; β is the so-called field enhancement factor, which is principally related to the emitter geometry; and A and B are constants with the values of 1.56 × 10⁻⁶ A eV V⁻² and 6.83 × 10³ eV^−3/2 V μm⁻¹, respectively.^15–17 The slope of the FN plot was a function of both β and Φ, expressed as:
which could be used to calculate the field enhancement factor or the work function. Assuming that the work function is 5.0 eV for In₂O₃, the field enhancement factor of the In₂O₃ NWs was estimated to be about 2123. When the saturation ratio increased, the In₂O₃ NS transformed into In–In₂O₃ NPs and the value of β was enhanced to 3590.

Fig. 4 The J–E field emission plots of the In₂O₃ NWs and In–In₂O₃ NPs. The inset displays the F–N plots of ln(J/E²) versus 1/E.

In Table 1, the In₂O₃ nanotowers displayed the highest E_to value, which was dependant on the tip diameter.¹⁸ Lin et al. reported that β values are related to the emitter geometry, crystal structure, vacuum gaps, and spatial distribution of the emitting centers.²¹ This shows that β obviously depends on d, with a relationship of 1/β being proportional to 1/d. According to the above two viewpoints, in this work, the E_to and β values of the In₂O₃ NWs and In–In₂O₃ NPs are lower and higher than those in the references in Table 1. However, the In₂O₃ NWs and In–In₂O₃ NPs in this report have the same tip diameter and d, so the metal indium in the In–In₂O₃ NPs must enhance the transmission of electrons.

Table 1 Comparative key FE performance parameters of the present In–In₂O₃ NPs and conventional In₂O₃ NSs reported in the literature (the E_to is defined as an electric field producing a current density of 10 μA cm⁻²)

Material	Length (μm)	Tip diameter (nm)	d (μm)	Eto (V μm^−1)	β	Ref.
In–In2O3 nanopagodas	3–6	100	150	2.0	3590	This work
In2O3 nanowires	12–18	100		2.88	2123
In2O3 nanoneedles	2	50	300	4.9	3695	18
Nanohooks	Several tens	50		7.5	1770
Nanorods	Several tens	125		7.7	1374
Nanotowers	Several tens	300		9.5	458
In2O3 nanotrees	Several tens	Few micrometers	220	6.45	1993	19
Nanopins	>300	1000		4.58	2320
Nanobrushes	>400	600		3.94	2749
Sn–In2O3 nanowires	5	50	—	6.6	1857	20
In2O3 nanowires – Ga2O3 nanobelts	60	300	170	4.23	1142	21

---

Fig. 5(a) schematically depicts the band diagrams of the tips of the In₂O₃ NWs. In₂O₃ is an n-type material. The Fermi level (E_f) of the depletion region was close to the conduction band E_c. When in a high electric field, the conduction band bends to E_f, and the electrons will tunnel to the vacuum level. In accordance with the indium metal enhancing the FE performance, the band diagrams of the In–In₂O₃ NPs are shown in Fig. 5(b). In Fig. 5(c), the electrons came from electrical power and were then transported from the In₂O₃ NPs to the metal In region. The In clusters of the In₂O₃ NPs assembled many transfer paths for enhanced conductivity. Ultimately the electrons were collected at the Au–In alloy nanoparticles (the tips of the In–In₂O₃ NPs) by the electric field effect. The work functions of In₂O₃, Au and In are 5.0 eV, 5.1 eV and 4.0 eV, respectively.²² The tunnelling of electrons from the In region to the vacuum level will be easier than for Au and In₂O₃. The enhanced FE properties of the In–In₂O₃ NPs come from the In cluster assembled conduction paths and reduced work function (Φ). This work reveals the lowest E_to and the best FE performance compared with past In₂O₃ nanostructure reports.

Fig. 5 Band diagrams of (a) the In₂O₃ NWs and (b) the In–In₂O₃ NPs. (c) A schematic diagram of the electronic transmission channel in the In–In₂O₃ NPs.

Conclusions

The morphology of In₂O₃ NSs was transformed from NWs into NPs when the In vapor pressure increased. The In–In₂O₃ conformable NPs were synthesized with ultrahigh concentrations of indium vapor. The turn-on fields of the In₂O₃ NWs and In–In₂O₃ NPs were 2.9 and 2.0 V μm⁻¹, respectively. The In–In₂O₃ NPs had an increased β value of 3590, and the β of the novel In₂O₃ NWs was 2123. The In clusters in the In–In₂O₃ NPs act as electron transmission channels and FE emitters due to their lower work function (Φ). The In–In₂O₃ NPs present the best FE properties of the In₂O₃ NSs reported, and they are a suitable candidate for FE sensitive switching elements in nanoelectronic device applications.

Acknowledgements

The authors would like to thank the Ministry of Science and Technology of the Republic of China, Taiwan, for financially supporting this research under Contract no. MOST 103-2221-E-024-016.

Notes and references

1. H. A. R. Aliabad, Y. Asadi and I. Ahmad, Opt. Mater., 2012, 34, 1406.
2. Y. Fan, M. Jin, F. Xianjin and K. Lingyi, J. Cryst. Growth, 2008, 310, 4054.
3. S. Q. Li, Y. X. Liang and T. H. Wang, Appl. Phys. Lett., 2005, 87, 143104.
4. P. S. Khiabani, E. Marzbanrad, H. Hassani and B. Raissi, J. Am. Ceram. Soc., 2013, 96, 2493.
5. D. H. Kim, S. Lee, J. H. Park, J. H. Noh, I. J. Park, W. M. Seong and K. S. Hong, Sol. Energy Mater. Sol. Cells, 2012, 96, 276.
6. L. Xu, B. A. Dong, Y. Wang, X. Bai, Q. Liu and H. W. Song, Sens. Actuators, B, 2010, 147, 531.
7. G. Z. Shen, J. Xu, X. F. Wang, H. T. Huang and D. Chen, Adv. Mater., 2011, 23, 771.
8. J. Yang, C. K. Lin, Z. L. Wang and J. Lin, Inorg. Chem., 2006, 45, 8973.
9. K. F. Huo, X. M. Zhang, L. S. Hu, X. Sunl, J. Ful and P. K. Chu, Appl. Phys. Lett., 2008, 93, 013105.
10. S. Q. Li, Y. X. Liang and T. H. Wang, Appl. Phys. Lett., 2005, 87, 143104.
11. S. Kar, S. Chakrabarti and S. Chaudhuri, Nanotechnology, 2006, 17, 3058.
12. Y. Huang, K. Yu, Z. Xu and Z. Zhu, Phys. E, 2011, 43, 1502.
13. Y. Hao, G. Meng, C. Ye and L. Zhang, Cryst. Growth Des., 2005, 5, 617.
14. Y. G. Yan, Y. Zhang, H. B. Zeng and L. D. Zhang, Cryst. Growth Des., 2007, 7, 940.
15. R. H. Fowler and L. Nordheim, Proc. R. Soc. London, Ser. A, 1928, 119, 173.
16. C. J. Edgcombe, Philosophical Magazine Part B, 2002, 82, 1009.
17. V. Filip, D. Nicolaescu, M. Tanemura and F. Okuyama, Ultramicroscopy, 2001, 89, 39.
18. B. Wang, X. Jin, Z. B. Ouyang and P. Xu, J. Nanopart. Res., 2012, 14, 1008.
19. Y. Wang, Y. Li, K. Yu and Z. Zhu, J. Phys. D: Appl. Phys., 2011, 44, 105301.
20. B. Wang, Z. Zheng, H. Wu and L. Zhu, Nanoscale Res. Lett., 2014, 9, 111.
21. J. Lin, Y. Huang, Y. Bando, C. C. Tang, C. Li and D. Golberg, ACS Nano, 2010, 4, 2452.
22. S. Kar, S. Chakrabarti and S. Chaudhuri, Nanotechnology, 2006, 17, 3058.

---