Title Molten salt-based synthesis of metal nitride semiconductors and their photoelectrochemical properties( 本文(Fulltext) )
Author(s) GANI PURWIANDONO
Report No.(Doctoral
Degree) 博士(工学) 甲第602号
Issue Date 2021-03-25
Type 博士論文
Version ETD
URL http://hdl.handle.net/20.500.12099/81589
※この資料の著作権は、各資料の著者・学協会・出版社等に帰属します。
Molten Salt-based Synthesis of Metal Nitride Semiconductors and Their Photoelectrochemical
Properties
㔠ᒓ❅≀༙ᑟయࡢ⁐⼥ሷ࣮࣋ࢫࡢྜᡂග㟁
ẼᏛ≉ᛶGANI PURWIANDONO
Environmental and Renewable Energy Systems Graduate School of Engineering
Gifu University, Japan
2021
2
TABLE OF CONTENTS
Thesis outline.
Abbreviations.
Chapter 1.
General Introduction.
1.1. Background ... 7
1.2. Semiconductor as photocatalyst ... 8
1.3. III-nitride semiconductor as photocatalyst ... 10
1.4. Molten salt-based synthesis of photocatalyst...12
1.5. The aim of the study ... 14
References. Chapter 2. Photoelectrochemical Property of 2D Hexagonal-shape GaN Nanoplates Synthesized using Solid Nitrogen Source in Molten Salt. 2.1. Introduction ... 17
2.2. Experimental ... 19
2.3. Result and discussion ... 21
2.4. Conclusions ... 33 References.
3 Chapter 3.
A Molten Salt-based Nitridation Approach for Synthesizing Nanostructured InN Electrode Materials.
3.1. Introduction ... 36
3.2. Experimental ... 38
3.3. Results and discussion ... 41
3.4. Conclusions ... 55
References. Chapter 4. General Conclusions...………..58
4
Molten Salt-based Synthesis of Metal Nitride Semiconductors and Their Photoelectrochemical Properties
(
㔠ᒓ❅≀༙ᑟయࡢ⁐⼥ሷ࣮࣋ࢫࡢྜᡂග㟁 ẼᏛ≉ᛶ)
Gani Purwiandono
Thesis Outline
The efficient utilization of renewable solar energy can be a key solution for environmental and energy issues. Photocatalysis has expected to make a great contribution to both environmental treatment and renewable energy. Over the past few decades, some applications based on photocatalysis has been developed. In the development of materials photocatalyst, III-nitrides are apparent due to unexpected properties such as wide-ranged direct band gap (ranging from deep UV AlN (~6.01 eV) to far-infrared InN (~0.7 eV)), high thermal stability and high electron mobility. III-nitride material family can cover almost all of the usable solar emission range (0.5–3.0 eV) due to their unique property of band gap tuning which can be achieved by varying the composition of the metal.
To date, many semiconductor photocatalysts have been created by a conventional method. In conventional synthesis, the semiconductor photocatalyst often required gas-phase nitridation of the precursors at relatively high temperatures. It was necessary to find a new synthesis route in the synthesis of semiconductor photocatalysts with low temperature, which would open up new
5
avenues to create functional material photocatalysts.
This thesis, which is composed of four chapters, focuses on the molten salt-based synthesis of GaN and InN using the mixture of metal chloride and solid nitrogen source and its photoelectrochemical property. Chapter 1 is the introduction, which described the background and objectives of this study.
Chapter 2 investigated the molten salt-based synthesis of 2D hexagonal-shape GaN nanoplates using gallium chloride and lithium nitride as a metal and nitrogen source. LiCl as molten salt could interact with the GaN surface, leading to the specific crystal orientation, producing hexagonal-shape GaN at a nanoscale. The formation of hexagonal-shape GaN instead of random-shaped crystals could enhance the charge transport in the photoelectrochemical cells.
Chapter 3 examined the effect of molten salt on the nitridation synthesis of single-phase InN nanocrystals. The photoelectrochemical cell fabricated using InN electrodes exhibited a photoresponse under visible and near-IR light irradiation.
Chapter 4 summarizes the major conclusions of this study. This thesis proposed a new approach in the molten salt-based synthesis of metal nitride semiconductor. The particular importance of our molten salt-based synthesis lies in the size control of metal nitride particles at a nanoscale, which is a new concept for synthesizing metal nitride. Further research of these new approaches will be able to create many semiconductor photocatalysts which could be applied in harvesting visible light for solar water splitting.
6 Abbreviations
1. Valence band (VB) : The band is occupied with electrons (in solid) 2. Conduction band (CB) : The lower level is barely occupied with
electrons
3. Bandgap (Eg) : The energy gap between the valence band and the conduction band
4. eV : Electron volt
5. nm : Nanometer
6. m : Mol
7. M : Mol Lˉ1
8. N2 : Nitrogen
9. h : Hour
10. mL : Milliliter
11. UV : Ultraviolet
12. μA : Micro ampere
13. mA : Mili ampere
14. HR-TEM : High-Resolution Transmission electron microscopy
15. SEM : Scanning Electron Microscopy
16. XRD : X-Ray Diffraction
17. BET : Brunauer-Emmet-Teller
18. XPS : X-Ray Photoelectron Spectroscopy
7 CHAPTER I General Introduction
1.1. Background
Global energy demand is growing exponentially as a result of exponential demographic growth, with social and economic change at the same time. Clean and green energy is a significant problem in recent years due to global energy concerns. Fossil fuel is often used as a means of energy supply. Consumption of carbon fuels raises greenhouse gasses in the atmosphere, which have been found to be the primary cause of climate change along with other significant environmental and health problems, including water pollution, human and animal health consequences.1-2
In this sense, the quest for renewable and reliable energy sources is becoming increasingly relevant in resolving energy demand and climate change.
Several research experiments have been introduced over the past few decades to discuss the potential decline of conventional energy supplies and global energy demand in conjunction with environmental pollution.3 Many of these reports concentrate on the influence of energy conservation on economics and population development by introducing the use of renewables or sustainable energy sources and sequestration of the greenhouse effect such as carbon dioxide.4–7
Hydrogen is known to be one of the most viable energy carriers to substitute fossil fuels for the world’s energy needs. In the context of producing clean hydrogen, photosynthesis is a natural process that is both environmentally benign and renewable, where water and carbon dioxide are converted to glucose
8
and oxygen. The mechanism of photosynthesis may be defined as an equation (1.1).
ଶ ଶ୮୦୭୲୭୬ሱۛۛۛሮ ଵଶ ଶ (1.1)
Thus, photosynthesis can serve as inspiration for the design of chemical processes for splitting water into H2 and O2. In this regard, photocatalytic water splitting for hydrogen production is seen as a type of artificial photosynthesis, where it utilizes solar energy to convert water into H2 and O2, as shown in equation (1.2).8,9
ʹଶ୮୦୭୲୭୬ୟ୬ୢୡୟ୲ୟ୪୷ୱ୲
ሱۛۛۛۛۛۛۛۛۛۛۛۛۛሮ ʹଶ ଶ (1.2) Semiconductor photocatalyst is a type of heterogeneous catalyst accelerated by photon energy absorption. In this type of catalyst, the photocatalyst is typically a solid material capable of producing an electron-hole pair upon the light irradiation, involved in redox reactions. Honda and Fujishima introduced the first photoelectrochemical cell (PEC) that could separate water into hydrogen and oxygen by using ultraviolet (UV) irradiation in 1972.10 The works of Honda and Fujishima inspired many researchers to develop semiconductor-based materials for photocatalytic water splitting.
1.2. Semiconductor as photocatalyst
Semiconductor materials are particularly useful for the photocatalytic process because of a favourable combination of electronic structure, light adsorption properties, charge transport characteristics, and a lifetime of the excited state.11 The semiconductor can be used as a photocatalyst since it has a particular electronic configuration of a filled valence band (VB) and an empty conduction band (CB). The valence band (VB) is the band made up of the completely occupied molecular orbital, low in energy. On the other hand, the conduction band
9
(CB) is the band of the molecular orbital that is high in energy, sufficient to make the electrons free to move from atom to atom under the influence of applied energy. The energy gap between the conduction band and the valence band is called bandgap energy, Eg. It requires applied energy to promote electrons from the valence band to the conduction band, and since the bandgap energy is different for different semiconductors, the required energy is also different for different semiconductors.12,13
When a semiconductor catalyst is illuminated with photons whose energy is equal to or greater than the bandgap of the semiconductor (Eg), the electron (e-) is promoted from the VB to the CB, leaving a hole (h+) in the VB (Figure 1.1, I).
The excited state conduction band electrons and valence band holes are kinetically as well as thermodynamically favoured to recombine and dissipate the input energy in the form of heat or emitted light as shown in Figure 1.1, II. However, since the electrons and holes can migrate to the surface of the semiconductor without recombination (Figure 1.1, III), they can be involved in electrochemical processes with species adsorbed on the semiconductor surface. Photogenerated electrons act as reducing agents and the holes act as oxidizing agents. The redox ability of the electron/hole pairs can be used for photocatalytic water/air remediation and photocatalytic hydrogen production.14,15
10
Fig. 1.1 The basic concept in semiconductor-based photocatalysis.14,15 1.3. III-nitride semiconductor as photocatalyst
Since the first report of photocatalytic and photoelectrochemical (PEC) water splitting on semiconductor materials under ultraviolet (UV) irradiation by Honda and Fujishima in the early 1970s, may researchers have begun to investigate new material semiconductors.10 In the development of materials photocatalyst, III-nitrides are apparent due to unexpected properties such as wide-ranged direct band gap (ranging from deep UV AlN (~6.01 eV) to far-infrared InN (~0.7 eV)), high thermal stability and high electron mobility.16-21 III-nitride material family can cover almost all of the usable solar emission range (0.5–3.0 eV) due to their unique property of band gap tuning which can be achieved by varying the composition (ratio of indium, x to gallium, 1-x). Just by adjusting the composition, the band gap of InGaN can be tuned from 0.7 eV (InN;
x=1) to 3.4 eV (GaN; x=0) covering almost the entire solar spectrum.22
Jung et al.23 showed a comparative study of photocatalytic activity of GaN nanostructure, submicron dots, and thin film. They observed that the photochemical activity of GaN was much higher than that of GaN submicron dots
11
and thin films by measuring the amount of photodegraded dye solution. The enhanced photochemical performance of the GaN was due to the large surface area and high crystallinity of the nanostructure. Subsequently, substantial attempts have been devoted to the study of the photochemical performance of group III-nitride nanostructures.
In order to more effectively utilize the longer wavelength (!500 nm) into deep-visible and NIR solar spectrum, higher In composition of InGaN nanostructure is an alternative approach to extend the absorption edge. However, the development of high-quality InGaN with high In content (!50 %) was extremely difficult due to the large lattice mismatch (a11 %) between InN and GaN, spinodal decomposition, In phase separation, and In surface segregation.24
Recently, Sugiura et al.25,26 have demonstrated the nitridation method of synthesis GaN and InN as semiconductor photocatalyst by the reaction of metal oxide and solid nitrogen source in metal flux at the temperature of 400°C-600°C under N2 atmosphere. Particulate InN and GaN crystals with a diameter of approximately 800 nm were obtained. Metal flux could act as a media of nitridation reaction. Similar to metal flux, molten salts are often used as a media to synthesize metal nitride. They can effectively control the crystallinity at relatively low temperatures. Moreover, molten salts have been shown to reducing the temperature as well as producing phase-pure nanoscale materials during the nitridation process.27-29 One of the challenges is to synthesize a single-phase metal nitride nanostructure under molten salt-based nitridation reaction.
1.4. Molten salt-based synthesis of photocatalyst
The term molten salt is a liquid matrix ionic.30 The molten salt compounds can be classified as single, binary, ternary and quaternary mixtures
12
based on the number of salts included in the system, as shown in Table 1.1.
Table 1.1 Molten salt-type examples
Types Examples
Single LiCl, KCl, NaCl
Binary KCl-LiCl, NaCl-KCl
Ternary LiCl-KCl-KF
Quarternary NaCl-KCl-LiCl-CaCl2
Molten salt synthesis is a well-established low-temperature synthesis technique that has gained much interest in recent years for its potential use in fabricating a variety of materials such as carbides31, oxides32, aluminates33, and titanates34. The key point of this synthesis method is that a considerable volume of water-soluble salt is applied to the reactants and the mixture is heated above the melting point of the salt to produce a large quantity of liquid phase in the synthesis system. The selected salt typically possesses a low melting point and provides a liquid phase at a relatively low temperature. The obtained liquid bath acts as a reaction solvent that defines the product characteristics (e.g., sizes, shapes). The given molten salt facilitates the dissolution of the reactants in the molten salt, leading to a homogeneous mixture and increased contact opportunities between the reactants in the melt. In addition, the liquid medium provides accessible routes for the dissolved species to transport through it, resulting in an increased diffusion rate of those species and thus the completion of reactions at a relatively low temperature and in a short reaction period.35,38
The molten salt-based method, which determines the product's morphologies, typically includes two primary reaction mechanism:
template-growth and dissolution-precipitation. The relative dissolution rates of the
13
reactants in the molten salt system determine the principal reaction mechanism. Li et al.36,37 and Kimura38 summarised the different characteristics of the two reaction mechanisms.
The dissolution-precipitation mechanism governs the synthesis formation when the dissolution rates of reactants A and B in the molten salt are comparable (Fig. 1.3a). As a result, all reactants are soluble in the molten salt, and the resulting product precipitates in the melt under an extreme degree of supersaturation. The final product gained from this formation mechanism shows a different morphology than the starting materials.
The template-growth mechanism is dominant when the dissolution rate of reactant. A is slightly higher than that of reactant B. A dissolves in the molten salt and diffuses onto the surface of B (Fig. 1.3b) and B functions as the reaction template. The reaction subsequently takes place on the surface and forms the layer of product phase in situ (P) (Fig. 1.3b). The formed layer prevents the further dissolution of B. The dissolution of A in the melt increases with time (Fig. 1.3b) while the product layer prohibits the dissolution of B. More dissolved A diffuses deeper to the unreacted zone of B, thereby increasing the product layer. This phenomenon is repeatedly processed until the reaction is completed. Thus, the morphologies (shapes and sizes) of the synthesised product (P) obtained via the template growth reaction are similar to those of the less soluble reactant (B).
14
Fig. 1.3 Schematic diagrams the formation of product C from reactants A and B via (a) dissolution-precipitation and (b) template growth mechanisms.36-38
1.5. The aim of the study
The purpose of this thesis is to synthesize gallium nitride (GaN) nanoplates and indium nitride (InN) nanocrystals as a metal nitride photocatalyst.
These photocatalysts were synthesized by using molten salt-assisted nitridation reaction as a new approach of the synthesis method. The photoelectrochemical properties of those materials electrode were also reported.
15 References
1. C. McGlade and P. Ekins, Nature, 2015, 517, 187-190.
2. T. Abbasi and S. A. Abbasi, Renew. Sustain. Energy Rev., 2011, 15, 1828-1834.
3. L. Wagner, I. Ross, J. Foster and B. Hankamer, PLoS ONE, 2016, 11, e0149406.
4. T. R. Karl and K. E. Trenberth, Science, 2003, 302, 1719-1723.
5. A. M. Omer, J. Renew. Sustain. Energy, 2009, 1, 53101.
6. N. R. Council, The Hydrogen Economy: Opportunities, Costs, Barriers, and R& D Needs, The National Academies Press, 2004.
7. N. S. Lewis and D. G. Nocera, Proc. Natl. Acad. Sci., 2006, 103, 15729-15735.
8. S. V. Mohan, G. Mohanakrishna and S. Srikanth, Biofules: alternative feedstocks and conversion processes, 2011, 499-524.
9. H. Uchida and M. R. Harada, Science and engineering of hydrogen-based energy technology, 2019, 201-220.
10. A. Fujishima and K. Honda, Nature, 1972, 238, 37-38.
11. Y. Nosaka and A. Nosaka, Royal Society of Chemistry, 2019, 51-52.
12. X. Chen, S. Shen, L. Guo, and S. S. Mao, Chem. Rev., 2010, 110, 6503-6570.
13. R. Abe, J. Photochem. Photobiol. C, 2010, 11, 179-209.
14. S. Dong, J. Feng, M. Fan, Y. Pi, L. Hu, X. Han, M. Liu, J. Sun and J. Sun, Rsc Adv., 2015, 5(19), 14610-14630.
15. P. Riente and T. Noël, 2019. Catal. Sci. Technol., 2019, 9(19), 5186-5232.
16. F. Chen, X. Ji and S. P. Lau, Mat. Sci. Eng., 2020, 142, 100578.
17. A. D. L. Bugallo, L. Rigutti, G. Jacopin, F. H. Julien, C. Durand, X. J. Chen, D.
Salomon, J. Eymery, M. Tchernycheva, Appl. Phys. Lett., 2011, 98, 233107.
18. H. P. T. Nguyen, S. Zhang, K. Cui, X. Han, S. Fathololoumi, M. Couillard, G.
A. Botton, Z. Mi, Nano Lett., 2011, 11, 1919-24.
19. W. Guo, M. Zhang, A. Banerjee, P. Bhattacharya, Nano Lett., 2010, 10(9), 3356-9.
20. R. Calarco, Materials, 2012, 5, 2137-50.
21. Y. Taniyasu, M. Kasu, T. Makimoto, Nature, 2006, 441, 325-8.
22. U. Chatterjee, J-H. Park, D-Y. Um, C-R. Lee, Renew. Sust. Energ. Rev., 2017, 79, 1002-1015.
23. H. S. Jung, Y. J. Hong, Y. Li, J. Cho, Y. J. Kim, G. C. Yi, ACS Nano., 2008, 2(4), 637-642.
16
24. F. E. Osterloh, Chem. Soc. Rev., 2013, 2013, 42(6), 2294-2320.
25. T. Zhang, A. Kouyama and T. Sugiura, J. Ceram. Soc. Jpn., 2012, 120(1397), 25-29.
26. T. Zhang, T. Sugiura, W. Lu, F. Wu, J. Mao, P. Qiu and Hugejile, J. Ceram.
Soc. Jpn., 2017, 125, 371-374.
27. L. Gan, Z.-Y. Mao, F.-F. Xu, Y.-C. Zhu, X.-J. Liu, Ceram. Int., 2014, 40, 5067-5071.
28. L. A. Yolshina, A. G. Kvashinchev, Mater. Des., 2016, 105, 124-132.
29. J. Zhang, W. Li, Q. Jia, L. Lin, J. Huang, S. Zhang, Ceram. Int., 2015, 41, 12614-12620.
30. J. Sato, S. Saito, H. Nishiyama, Y. Inoue, J. Phys. Chem. B, 2003, 107, 7965.
31. J.Sato, S. Saito, H. Nishiyama, Y. Inoue, J. Phys. Chem. B, 2001, 105, 6061.
32. H. Yamashita, M. Takeuchi and M. Anpo, Encyclopedia of nanoscience and nanotechnology, 2004, 10, 639-654.
33. D. G. Lovering and R. J. Gale, Molten Salt Techniques vol. 1, New York and London: Plenum Press, 1983.
34. R. Yang, L. Cui, Y. Zheng and X. Cai, Mat. Lett., 2007, 61(26), 4815-4817.
35. X. Liu, N. Fechler and M. Antonietti, Chem. Soc. Rev., 2013, 42(21), 8237-8265.
36. Z. Li, S. Zhang and W. E. Lee, J. Eur. Ceram. Soc., 2007, 27(12), 3407-3412.
37. Z. Li, W. E. Lee and S. Zhang, J. Am. Ceram. Soc., 2007, 90(2), 364-368.
38. T. Kimura, Advances in ceramics-synthesis and characterization, processing and specific applications, 2011, 75-100.
17 CHAPTER II
Photoelectrochemical Property of 2D Hexagonal-shape GaN Nanoplates Synthesized using Solid Nitrogen Source in Molten Salt
Abstract
Two dimensional (2D) hexagonal-shape gallium nitride (HS-GaN) was synthesized by reacting GaCl3 and LiNH2 with LiCl as the molten salt. We demonstrated that the size and shape of 2D plate-like GaN nanoplates could be controlled under a relatively low temperature from 600°C to 700°C. TEM analysis revealed that the single crystal of HS-GaN has a plate-like morphology with diameter and thickness of 450 nm and 80 nm, respectively. The photoelectrochemical cell fabricated using the HS-GaN electrode showed a higher current by a factor of up to 2 compared to that using random-shaped GaN particles under UV light irradiation. The precise growth control of 2D GaN crystals will facilitate new methods of developing high-performance photoenergy conversion materials.
2.1 Introduction
Gallium nitride (GaN), is a group III nitride semiconductor. At 20-25°C, the hexagonal and cubic forms of GaN show a direct energy bandgap of about 3.4 and 3.2 eV, respectively.1 Thin films of these forms of GaN have beenwidely adopted for both commercial and fundamental research, such as in light-emitting diodes (LEDs) and UV light water-splitting.2-5 Much interest has also been paid to the creation of nanostructures of GaN including the morphology of wires, rods, sheets, seeds and hollows.6-11 The nitridation reactions are often conducted using gas flow techniques for growing GaN crystals. In this process, a nitrogen-rich gas, such as ammonia, is usually used as a source of nitrogen. GaN single crystals with a diameter of approximately 300 Pm have been grown at temperatures ranging between 700°C to 900°C. For the preparation of GaN crystals, Ga is treated under N2 atmosphere.12-14 Typically, these reactions are conducted under high-pressure
18
and by using the N2 gas as the nitrogen source. Zhang et al. successfully synthesized GaN by using LiNH2 as a solid nitrogen source.15 Here, the Ga metal was reacted with LiNH2 at a temperature of 750°C under ammonia atmosphere.
GaN crystals with a diameter of approximately 100 Pm were obtained. However, none of these studies have investigated crystal growth control for achieving the 2D morphology of GaN at a nanoscale.
Molten salts are often used as reaction media to increase the crystallinity of ceramics. They can effectively control the crystallinity at relatively low temperatures.16-18 Liang et al. reported the use of LiCl-NaF as the molten salt for the synthesis of LiSi2N3 nanobelts under N2 atmosphere at 1200°C.19 The reaction was performed at a lower temperature of as much as 200°C than those used for the conventional solid-state reaction routes. On the other hand, Liu et al. used KCl-KF as the molten salt in the nitridation process to synthesize VN nanopowder with a particle size of 130 nm at 700°C under N2 atmosphere.20 Therefore, molten salts have been shown to play a vital role in reducing the temperature as well as producing phase-pure nanoscale materials during the nitridation process.
In this chapter, the synthesis of 2D GaN nanoplates has been established for the first time by using LiCl as the molten salt at relatively low temperatures.
Also, we anticipated that the formation of 2D GaN nanoplates instead random-shaped crystals could enhance the charge transport in the photoelectrochemical cells. Here, we present the effect of the molten salt on the crystal growth of the 2D HS-GaN microstructure and the photochemical properties of the GaN electrodes.
19 2.2 Experimental section
2.2.1 Preparation of GaN powder
GaN was synthesized using GaCl3 (Yamanaka Hutech, 99.99%) as gallium and LiNH2 (Merck, 95%) as a nitrogen source. LiCl (Kanto, 99.95%) was used as the molten salt. For the synthesis of the desired materials, 1 mole of GaCl3 was reacted with 6 moles of LiNH2 in 1 mole of LiCl. The starting materials (GaCl3, LiNH2) and the molten salt (LiCl) were mixed inside the glove box in a graphite crucible (inner diameter: 55 mm, length: 309 mm, SUS 316) under N2 atmosphere with 15% humidity. The mixture was initially preheated at 473 K for 1 h.21 After this, the nitridation reactions were carried out at different temperatures such as 550qC, 600°C, 650°C, 700°C, and 800qC for 2 h under N2 atmosphere. The ramping rate for nitridation reaction was set at 10qC/min and the cooling rate was set at 2qC/min. After cooling down, the products were washed with 0.1 M HCl and distilled water to obtain GaN powders.
2.2.2 Preparation of GaN electrodes
GaN/titanium paste electrodes were prepared to examine the photoelectrochemical properties of GaN crystals with CoOx as co-catalyst. To prepare GaN electrodes, the as-synthesized GaN powder was mixed with the 2-butanol solution (20 wt%, Wako, 99%) for 3 hours. The optimal paste was then applied on the Ti substrate (Nilaco) by doctor blade technique using scotch tape as a spacer. Before using, the Ti substrate was cleaned by etching solution (5 mL distilled water, 5 mL HNO3 (Nacalai Tesque, 60%) and 1 mL HF (Wako, 46%).22 The heat treatment was performed at 500qC for 2 h under N2 atmosphere. CoOx of
20
Co(NO3)2.6H2O (5 mM, Wako, 99.5%) was added to NaOH (5 mM, Kanto, 97%) solution until the solution becomes neutral (pH = 7). After this, the heat treatment of the co-catalyst was initially performed at 600qC for 2 h under N2 atmosphere and then at 200qC for 1 h under the ambient condition.
The Mott-Schottky analysis was applied to determine the donor density (ND) of GaN electrodes, using the equation (1). E, EFB,and c indicated the applied potential, flat band potential and space charge capacitance in the electrode, respectively. T, k, e, εo, and ε are the absolute temperature, Boltzmann constant, elemental charge, vacuum permittivity, and relative permittivity, respectively. The relative permittivity of 8.9 was used for obtaining the ND value using equation (2).
The Mott-Schottky plots were obtained using a VersaSTAT3 potentiostat. The measurements were performed using 0.1 M Na2SO4 solution at a given bias potential under dark condition. The measured frequency was 1 kHz.
ͳ
ଶ ൌ ሺ െ െ ൗ ሻ
ୈɂ୭ɂ
൘ (1)
ୈ ൌ ʹሺ
ሻ
ሺͳ ଶሻ
൙
ɂ୭ɂ
(2)
2.2.3 Structure characterization of GaN and photoelectrochemical measurement of the electrode
The crystallinity and crystallographic structure of GaN were examined by X-ray diffraction (XRD, Rigaku Ultima II/PC) with CuKD radiation and transmission electron microscopy (TEM, JEOL, JEM-2100), respectively. The surface morphology of GaN was observed by scanning electron microscopy
21
(SEM, Hitachi, S-4800). The elemental analysis of samples was examined by X-ray photoelectron spectrocopy (XPS, ULVAC, Quantera SXM), Photoelectrochemical measurements were performed with a three-electrode cell combining the GaN photoanode with a saturated calomel reference electrode and a platinum counter electrode in the aqueous Na2SO4 (Chameleon, 99%) electrolyte.
The current-voltage characteristics were obtained using a potentiostat under intermittent UV light irradiation (Xe lamp with a light intensity of 86.6 mWcmˉ2).
2.3 Results and discussion
The XRD patterns of synthesized GaN powders with LiCl as the molten salt at 550°C, 600°C, 650°C, 700°C, and 800°C are shown in Fig. 2.1. The major diffraction peaks of the products can be indexed to HS-GaN, which agreed well with the JCPDS data (89-8624). No peaks of the by-products were detected, suggesting the production of single-phase GaN crystals. The controlled reaction at 550°C, where the temperature was lower than the melting point of LiCl (600°C), was also carried out. The diffraction peaks, in this case, became significantly lower in intensity as compared to those observed for the 600°C reaction (Fig. 2.1).
This indicated that the molten salt plays a crucial role in facilitating crystal growth. At higher reaction temperatures, the intensity of the diffraction peaks increased. However, this increasing trend was observed only till 700°C of reaction temperature, after which (e.g., at 800°C), the intensity of the peaks corresponding to GaN decreased. This result was also inferred from the decreased size of samples at 800°C as shown in Table 2.1.15 To understand the role of molten salt, we have synthesized GaN without LiCl and have also measured by XRD and SEM (Figs. 2.3 and 2.4). The intensity of GaN peaks in XRD pattern were
22
increased with increasing temperaure up to 700°C. For 700°C sample, the peaks of byproduct Li3GaN2 appeared. SEM images indicated that the crystal size became larger with increasing the reaction temperature. The hexagonal-shape crystals formed for 650°C and 700°C samples. However, the size and shape of crystals were not homogeneous for both conditions.
Table 2.1 Reaction yields and crystallite sizes of GaN synthesized at different temperatures.
Temperature (°C) Reaction Yields (%) Crystallite sizes for each plane (nm)
(100) (002) (101)
550 12 - - -
600 47 79 83 110
650 66 101 92 184
700 86 156 169 177
800 41 65 48 88
The reaction yields and crystallite sizes of GaN synthesized at different temperatures shown in Table 2.1, suggested that both the factors increased with a rise in the reaction temperature from 550°C to 700°C. The largest crystallite size of 184 nm for the (101) plane was obtained from GaN synthesized at 650°C. In the SEM images of GaN crystals synthesized at various temperatures, similar hexagonal-plate morphologies were observed (Fig. 2.2). For the synthesis at 600°C, 650°C, and 700°C, GaN formed with the particle size smaller than 500 nm (Fig. 2.2). These GaN nanoplates are termed as 2D HS-GaN. Notably, by increasing temperature up to 700°C, particle sizes increased to be more than 300 nm and the crystal shape changed from granular to hexagonal. However, at 800°C, the crystal size decreased and the particles adopted a random plate-like shape. The thickness of the 2D HS-GaN nanoplates varied from 50 to 80 nm, as shown in Fig.
2.2f, 2.2g, and 2.2h.
23
Fig. 2.1 XRD patterns of GaN synthesized using GaCl3, LiNH2, and LiCl. The reactions were carried out at 550°C, 600°C, 650°C, 700°C, and 800°C.
(a) (b) (c)
(d) (e) (f)
24
Fig. 2.2 SEM images of GaN nanoplates prepared by the reaction of GaCl3, LiNH2, and LiCl at (a) 550°C, (b) 600°C, (c) 650°C, (d) 700°C, (e) 800°C. Images (f), (g), and (h) show the edge part of 2D plate-like structure corresponding to the samples prepared at 600°C, 650°C, and 700°C.
Fig. 2.3 XRD patterns of GaN synthesized using GaCl3 and LiNH2 without LiCl.
The reactions were carried out at 550°C, 600°C, 650°C, 700°C, and 800°C.
(g) (h)
25
Fig. 2.4 SEM images of GaN nanoplates prepared by the reaction of GaCl3 and LiNH2 without LiCl at (a) 550°C, (b) 600°C, (c) 650°C, (d) 700°C, and (e) 800°C.
It can be expected that the LiCl assisted synthesis facilitated a more homogeneous reaction than that of the solid-state synthesis at 550°C. In the molten state, the mobility of the reactants GaCl3 and LiNH2 is higher than that in the solid state. The increased mobility causes the reactants to be uniformly diffused within a short period.23 Thus, at temperatures, more than 600°C, GaCl3
and LiNH2 in the molten salt allowed the reaction to proceed at an atomic level.
The magnified HR-TEM images of an individual GaN nanoplate synthesized at 650°C showed the hexagonal morphology with a dimension of around 450 nm (Fig. 2.5a). The lattice fringe of 0.276 nm corresponding to the spacing of the (100) plane was also observed in the TEM image (Fig. 2.5b).
1.00 Pm
(e)
(a) (b) (c)
(d)
1.00 Pm 1.00 Pm
1.00 Pm 1.00 Pm
26
Fig. 2.5 TEM images of (a) GaN crystals synthesized at 650°C for 2 h and (b) selected area diffraction (SAD) pattern.
Durations for the nitridation reactions were varied from 1 to 3 h at 650°C to investigate the crystal growth mechanism. The XRD patterns of the products synthesized with LiCl as the molten salt at different reaction times suggested that the crystal growth of HS-GaN started to occur during the first 1 h of nitridation and the products were completely converted to hexagonal particles at 2 h (Figs.
2.2, 2.5 and 2.6). For 1 h reaction, the formation of GaN crystals from Li3GaN2
was not completed as shown in equations (3) and (4). Therefore, the Li3GaN2
peaks in the XRD pattern (Fig. 2.6) were still observed. For 3 h reaction, the sizes (a)
(b)
27
of hexagonal-shape GaN became smaller when they were compared to those of 2 h reaction (Table 2.2, Fig. 2.8). The size decrease was inferred from the decreased crystallinity as observed in XRD pattern (Fig. 2.6). These observations indicated that the reaction time significantly affected the shape and size of GaN crystals.
The mixture of reactants and the LiCl salt were uniformly heated, resulting in the decomposition of LiNH2 at a low temperature of 370°C to form Li3GaN2
particles. This was the first step (3), which led to the growth of the HS-GaN nanoplates in the melted LiCl and Li3GaN2 as the second step (4). Thus, the mechanism for the growth of the HS-GaN crystals started by the decomposition of GaCl3 and LiNH2 to form Li3GaN2 at 1 h (Fig. 2.6).
Fig. 2.6 XRD patterns of nitridation sample synthesized using GaCl3, LiNH2, and LiCl at 650°C. The XRD patterns of 1 h and 3 h reactions are presented.
28
2GaCl3 + 6LiNH2 Æ2Li3GaN2 + 3Cl2 + 6H2 + N2 (3) 2Li3GaN2 + 3Cl2 Æ 2GaN + 6LiCl + N2 (4) At 2 and 3 h of reaction times, no peak for Li3GaN2 was observed in the XRD patterns (Fig. 2.1 and Fig. 2.6). The increase in the reaction time will cause the decomposition of Li3GaN2 to form GaN particles. The Clˉ anions would interact with the (002) surface, which would further suppress the growth of GaN, leading to the nanoscale HS-GaN formation. XPS spectra of GaN samples prepared at various temperatures were shown in Fig. 2.7. We found that no Li was detected in our GaN samples.
Fig. 2.7 XPS spectra of samples prepared by the reaction of GaCl3, LiNH2, and LiCl for 2 h at (a) 700°C and (b) 800°C. Li1s spectra are also presented in the figure.
Table 2.2 Crystallite sizes of GaN synthesized at 650°C at various reaction times.
Reaction Time (h)
Crystallite sizes for each plane (nm)
(100) (002) (101)
1 75 54 98
2 101 92 184
3 86 71 122
29
Fig. 2.8 SEM images GaN prepared by using the mixture of GaCl3, LiNH2, and LiCl at 650°C for (a) 1 h, (b) 2 h, and (c) 3 h.
Based on the observations of SEM and TEM images, the crystal growth mechanism of GaN using the LiCl molten salt for 2 h reaction time can be proposed, as shown in Fig. 2.9. At the temperature lower than the melting point of LiCl (550°C), the small granular particles of GaN formed. At higher temperatures, the nucleation and crystal growth of HS-GaN occurred in the nitridation reaction.
Based on TEM analysis, it was clear that Ga atoms were only exposed at the surface of (002) plane, whereas nitrogen atoms were mostly occupied at the surfaces of (100) and (101) planes. Thus, Clˉ anions could preferentially interact with Ga cations at the surface of the (002) plane. This probably led to the suppression of crystal growth in this direction, producing the 2D plate-like morphology as shown in Fig. 2.10. By increasing the reaction temperature, the GaN grains grow along the perpendicular direction to the c-axis, and the (002) crystalline planes expanded, due to which the diameter of the nanoplates increased.
At a temperature of 800°C, Clˉ anions probably interact with different crystal phases in a random manner, resulting in the non-homogeneous shaped particles with decreased crystallinity.
(a) (b) (c)
30
Fig. 2.9 Schematic diagram of the crystal growth mechanism of HS-GaN nanoplates using LiCl molten salt.
Fig. 2.10 The palussible mechanism of crystal growth hexagonal GaN plate-like morphology using LiCl as molten salt.
GaN paste electrode was prepared to examine the photoelectrochemical properties of the synthesized 2D HS-GaN. The Ti plate was used for the substrate.
We have checked stability of the photocurrent at a constant potential for HS-GaN and random-shaped GaN electrodes (Fig 2.11). Up until 2.5 hours, slight decrease of photocurrent was observed for both electrodes. This photocurrent change was due to not only the partial photocorrosion of GaN but also water oxidation during
31
light irradiation.24,25 The photocurrents became constant after 2.5 hours. The results indicated that the higher photocurrent for HS-GaN came from more efficient water oxidation compared to that of random-shaped GaN.
Fig. 2.11 Stability of the photocurrent density at a bias of 1.0 V.
As shown in Fig. 2.12, the onset potentials of anodic photocurrents were -0.87 V vs SCE. The highest anodic photocurrent was 0.54 mAcm-2 for HS-GaN and 0.26 mAcm-2 for random-shaped GaN at 1.0 V. The anodic photocurrent for the hexagonal particles was about two times higher than that of the random-shaped ones.
Fig. 2.12 Current-potential curves of the GaN electrodes under intermittent UV irradiation. Samples were prepared at 700°C and 800°C for 2 h.
32
To understand the carrier transport, we have compared the donor density of HS-GaN and random-shape GaN electrodes determined by Mott-Schottky plots and the results were presented in Fig. 2.13. The donor density for the HS-GaN sample was 7.15×1018 cm-3, which was higher than that of the random-shaped GaN (8.90×1017cm-3). GaN is normally doped by oxygen and silicon impurities, which make it n-type.26, 27 Oxygen dopant can be an effective and controllable impurity for obtaining n-type GaN.28, 29 An uniform reaction in the Li molten salt may promote incorporation of oxygen atoms, leading to the higher donor density for HS-GaN. These results indicated that the carrier transport was more effective for HS-GaN. The photoelectrochemical performance of 600°C, 650°C, and 700°C samples were also shown in Fig. 2.14. The anodic photocurrents of GaN electrodes synthesized at 600°C, 650°C, and 700°C were 0.47 mAcm-2, 0.49 mAcm-2, and 0.51 mAcm-2, at 1.0 V, respectively.
Fig. 2.13 Mott-Schottky plots of GaN electrodes for 700°C and 800°C samples.
Donor densities (ND) are also presented in the figure.
33
Fig. 2.14 Current-potential curves of the GaN electrodes under intermittent UV irradiation. Samples were prepared at 600°C, 650°C, and 700°C for 2 h.
2.4 Conclusions
In summary, we have successfully synthesized the 2D HS-GaN nanoplates through the reaction of GaCl3 and LiNH2 using LiCl as the molten salt. The use of LiCl as the molten salt can accelerate the homogeneous formation of nanoscale 2D HS-GaN. Our synthetic analogy of LiNH2-assisting nitridation with the molten salt provides a generalized approach to create versatile group III nitride semiconductors. From photoelectrochemical properties of GaN paste electrode, the highest anodic photo-current was obtained for the 2D HS-GaN particles. The photoelectrochemical data revealed that the hexagonal structure gave a higher anodic photocurrent by a factor of up to 2 compared to the random-shaped GaN obtained at a higher temperature. To our belief, the photoelectrochemical properties of 2D HS-GaN electrode indicate that it has significant potential for photocatalytic water splitting.
34 References
1. M. Drygas, M. Sitarz, J. F. Janik, RSC Adv., 2015, 5, 106128-106140.
2. S.-H. Kang and T.-H. Fang, J. Alloys Compd., 2014, 597, 72.
3. H. Gu, G. Ren, T. Zhou, F. Tian, Y. Xu, Y. Zhang, M. Wang, Z. Zhang, D.
Cai, J. Wang, K. Xu, J. Alloys Compd., 2016, 674, 218-222.
4. M. Tchernycheva, A. Messanvi, A. D. Bugallo, G. Jacopin, P. Lavenus, L.
Rigutti, H. Zhang, Y. Halioua, F. H. Julien, J. Eymery, C. Durand, Nano Lett.,.
2014, 14, 3515-3520.
5. T. J. Flack, B. N. Pushpakaran, S. B. Bayne, J. Electron. Mater., 2016, 45, 2673-2682.
6. X. Sun and Y. Li, Angew. Chem., Int. Ed. 43., 2004, 29, 3827-3831.
7. P. Huang, H. Zhang, J.-J. Shi, M. Zhang, X.-H. Jiang, H.-X. Zhong, Y-.M.
Ding, Y.-P. He, J. Lu, X.-D. Hu, ACS Catal., 2016, 5, 3180-3192.
8. P. Tyagi, C. Ramesh, B.S. Yadav, S. S. Kushvaha, M. S. Kumar, CrystEngComm., 2019, 36, 5448-5454.
9. B. Liu, W. Yang, J. Li, X. Zhang, P. Niu, X. Jiang, Nano Lett., 2017, 5, 3195-3201.
10. T. Li, G. Ren, X. Su, J. Yao, Z. Yan, X. Gao, K. Xu, CrystEngComm., 2019, 33, 4874-4879.
11. C. Blumberg, F. Wefers, F. J. Tegude, N. Weimann, W. Prost, CrystEngComm., 2019, 48, 7476-7488.
12. F. Kawamura, M. Morishita, M. Tanpo, M. Imade, M. Yoshimura, Y. Kitaoka, Y. Mori, T. Sasaki, J. Cryst. Growth., 2008, 310, 3946-3949.
13. F. Kawamura, M. Morishita, N. Miyoshi, M. Imade, M. Yoshimura, Y.
Kitaoka, Y. Mori, T. Sasaki, J. Cryst. Growth., 2009, 311, 4647-4651.
14. T. Yamada, H. Yamane, Y. Yao, M. Yokoyama, T. Sekiguchi, Mater. Res.
Bull., 2009, 44, 594-599.
15. T. Zhang, T. Sugiura, W. Lu, F. Wu, J. Mao, P. Qiu, Hugejile, J. Ceram. Soc.
Jpn., 2017, 125, 371-374.
16. L. Gan, Z.-Y. Mao, F.-F. Xu, Y.-C. Zhu, X.-J. Liu, Ceram. Int., 2014, 40, 5067-5071.
17. L. A. Yolshina and A. G. Kvashinchev, Mater. Des., 2016, 105, 124-132.
18. J. Zhang, W. Li, Q. Jia, L. Lin, J. Huang, S. Zhang, Ceram. Int., 2015, 41, 12614-12620.
19. F. Liang, L. Tian, H. Zhang, F. Liang, S. Lu, R. Cheng, S. Zhang, RSC Adv., 2016, 6, 68615-68618.
35
20. H. Liu, C. Deng, X. Wang, C. Yu, J. Ding, H. Zhu, Ceram. Int., 2019, 45, 16638-16644.
21. A. Amir, K. Manseki, T. Sugiura, Trans. Mater. Res. Soc. Jpn., 2017, 42, 89-92.
22. T. Zhang, A. Kouyama, M. Miura, T. Sugiura, Trans. Mater. Res. Soc. Jpn., 2011, 36, 513-516.
23. L. Lei and D. He, Cryst. Growth Des., 2009, 9, 1264-1266.
24. H. Bae, H. Kim, J. W. Ju, D. W. Jeon, S. W. Ryu, Y. Moon, J. S. Ha, J.
Electrochem. Soc., 2019, 166, 103-107.
25. J. Kamimura, P. Bogdanoff, F. F. Abdi, J. Lähnemann, R. van de Krol, H.
Riechert, L. Geelhaar, J. of Phys. Chem. C., 2017, 121(23), 12540-12545.
26. A. J. Ptak, L. J. Holbert, L. Ting, C. H. Swartz, M. Moldovan, N. C. Giles, T.
H. Myers, P. V. Lierde, C. Tian, R. A. Hockett, S. Mitha, Appl. Phys. Lett., 2001, 79, 2740-2742.
27. C. G. Van de Walle, C. Stampfl, J. Neugebauer, J. Cryst. Growth., 1998, 189, 505-510.
28. Y. Wang, T. Teramoto, K. Ohkawa, Phys. Status Solidi (b)., 2015, 252(5), 1116-1120.
29. F. Lu, L. Liu, S. Xia, Y. Diao, S. Feng, Superlattice Microst., 2018, 118, 160-169.
36 CHAPTER III
A Molten Salt-based Nitridation Approach for Synthesizing Nanostructured InN Electrode Materials
Abstract
Single-phase InN nanocrystals were synthesized for the first time by a molten salt-based nitridation approach using InCl3 and LiNH2 as indium and nitrogen sources, respectively.
A molten salt, KCl-LiCl, during nitridation, enabled us to obtain InN nanocrystals at relatively low temperatures ranging from 400°C to 500°C. SEM and HR-TEM measurements coupled with XRD data revealed that InN nanocrystals were formed with average grain sizes of approximately 50–60 nm. Notably, the photoelectrochemical cell fabricated using the InN nanocrystals synthesized at 450°C exhibited a photocurrent response under light irradiation from 400 nm to 880 nm. The precise control of the growth of InN particles using our synthetic approach provides opportunities for developing versatile nitride nanocrystals.
3.1 Introduction
Indium nitride (InN) has been regarded as an attractive semiconductor for applications such as light-emitting diodes (LEDs) and infrared detectors.1-5 Moreover, the narrow bandgap, typically from 0.6 eV to 0.9 eV, enables the detection of its spectral range in the near-infrared region.6-13 In growing pure InN, the main difficulty lies in its thermal decomposition, which becomes significant above 600°C. In addition, high-temperature treatment causes the inclusion of impurities, especially oxygen atoms.14-15
Much attention has been given to the synthesis of single-phase InN using several methods, such as molecular beam epitaxy (MBE), metal organic chemical vapor deposition (MOCVD), and plasma and sputtering techniques.16-23 However,
37
few reports are available on nitridation reactions for growing nanoscale InN crystals.
Jung et al.24 reported the gas-phase nitridation synthesis of InN with NH3
gas at 600°C. InN particles with a diameter of approximately 300 nm were obtained. Our group has previously demonstrated the formation of InN at 400°C using In2O3 as an indium source in indium flux under N2 gas flow. Particulate InN crystals with a diameter of approximately 800 nm were obtained.25 These studies have demonstrated the effect of raw materials (In2O3 and LiNH2) on the morphology of InN crystals. However, none of the above-mentioned studies have investigated the morphology control of InN particles at the nanoscale and their photoelectrochemical properties.
We have successfully synthesized 2D hexagonal-shaped GaN nanoplates through the reaction of GaCl3 and LiNH2 using LiCl as the molten salt.26 The use of LiCl as the molten salt was found to accelerate the homogenous GaN formation at the nanoscale. Photoelectrochemical analysis of the GaN electrode revealed that the hexagonal-shaped GaN electrode exhibited an anodic photocurrent that was higher by a factor of 2 compared to that of the random-shaped GaN. With regards to the InN photo-electrode, Lindgren et al.27 reported an incident photon-to-electron conversion efficiency (IPCE) of InN film of up to 2% at 350 nm, whereas no photocurrent response was observed in the visible and near-IR regions.
The particular importance of our molten salt-based synthesis lies in the size control of InN particles at a nanoscale, which is a new concept for synthesizing InN nanocrystals. Our previous study on LiCl-based GaN synthesis showed that the Clˉ anions of LiCl can suppress the crystal growth of GaN through the Clˉ coordination to Ga(III).26 Based on the results, we anticipated that
38
InN nanocrystals could be synthesized using LiCl-containing molten salts. In this paper, we use a KCl-LiCl mixture with eutectic composition of 60% mol LiCl and 40% mol KCl that has a melting point of 353°C28, taking into consideration that the molten salts for nitridation reactions of InN should have a melting point that is lower than the synthesis temperature at 400°C-500°C.25,29
Herein, we present, for the first time, the effect of molten salt on the nitridation synthesis of single-phase InN nanocrystals. The InN nanocrystals were successfully obtained at relatively low temperatures ranging from 400°C to 500°C.
In addition, the photoelectrochemical cell fabricated using the InN electrodes exhibited a photocurrent response under visible and near-IR light irradiation.
Importantly, our methodology for controlling the growth of InN nanocrystals using molten salt-assisted nitridation provides a new approach to create various nitride nanocrystals and their electrodes for photo-energy conversion.
3.2 Experimental section
3.2.1 Synthesis of InN nanocrystals
InCl3 (0.664 g, 1 mol), LiNH2 (0.482 g, 7 mol), LiCl (0.153 g, 0.8 mol), and KCl (0.179 g, 1.2 mol) were mixed in a graphite crucible (inner diameter: 55 mm, length: 309 mm, SUS 316) under N2 atmosphere (50 mL/min). The mixture was preheated at 150°C for 30 min. The nitridation reactions were then carried out at 350°C, 400°C, 450°C, 500°C, and 550°C for 2 h under N2 atmosphere. The ramping rate for the nitridation reaction was set to 10°C/min, and the cooling rate was set to 2°C/min. After cooling, the products were washed with 1 M HCl and ethanol to obtain InN powder samples.
39 3.2.2 Preparation of InN Photoelectrodes
InN/titanium paste electrodes were prepared to examine the photoelectrochemical properties of InN films with a thickness of 0.5 Pm. To prepare InN electrodes, the as-synthesized InN powder was mixed with a 2-butanol solution (20 wt%, Wako, 99%) for 24 h. The paste was deposited on a Ti substrate (Nilaco) by the doctor blade technique using scotch tape as a spacer.
Prior to deposition, the Ti substrate was cleaned with an etching solution (a mixture of 7 mL distilled water, 5 mL HNO3 (Nacalai Tesque, 60%), and 1 mL HF (Wako, 46%)).The heat treatment was performed at 350°C for 2 h under a N2
atmosphere. Then, 5 mL of a NaOH (5 mM, Kanto, 97%) solution was added to 20 mL of a Co(NO3)2∙6H2O (5 mM, Wako, 99.5%) solution, and a portion of this mixture was dropped onto the InN film. Subsequently, heat treatment was performed at 350°C for 2 h under a N2 atmosphere.
3.2.3 Structural characterization of InN powder
The structure of the InN powder was analyzed by X-ray diffraction (XRD, Rigaku Ultima II/OC) with CuKD radiation and transmission electron microscopy (TEM, JEOL, JEM-2100). The morphology of the InN particles was observed by scanning electron microscopy (SEM, Hitachi, S-4800). The elemental analysis of the samples was performed via X-ray photoelectron spectroscopy (XPS, ULVAC, Quantera SXM). The specific surface areas of the InN powders were determined using a gas sorption analyzer (Micromeritics Tristar II 3020). The absorbance measurements were carried out on a UV–vis spectrophotometer (Hitachi, U-4000).
The Debye-Scherrer equation was applied to determine the crystallite size of InN powder using equation (1). D, O, Eand Ʌ correspond to the crystallite size,
40
wavelength of X-ray radiation, full width at half maximum (FWHM), and diffraction angle, respectively.
ൌǤଽସO
E Ʌ (1)
3.2.4 Photoelectrochemical measurements
Photoelectrochemical measurements were performed with a three-electrode cell consisting of the InN photoanode, saturated calomel reference electrode, and platinum counter electrode using 1 M Na2SO4 (50 mL). H2O2 (1 mL, 30% concentration) was also added to the electrolyte.28 The current-voltage characteristics were obtained using a potentiostat under intermittent UV light irradiation with a light intensity of approximately 120 mWcm·2 (Xe lamp with a 64R filter). Mott–Schottky analysis was applied to determine the donor density (ND) of the InN electrodes using equation (2). E, EFB,and c correspond to the applied potential, flat band potential, and space charge capacitance in the electrode, respectively. T, k, e, εo, and ε are the absolute temperature, Boltzmann constant, elemental charge, vacuum permittivity, and relative permittivity, respectively. A relative permittivity of 15.3 Fm·1 was used to obtain the ND value using equation (3). The Mott–Schottky plots were obtained using a VersaSTAT3 potentiostat. The measurements were performed using a 1 M Na2SO4 solution at a given bias potential under dark conditions. The measured frequency was 1 kHz.
ଵ
ୡమ ൌሺ െ െ ൗ ሻ
ୈɂ୭ɂ
൘ (2)
ୈ ൌ
ଶሺౚు
ౚሻ
ୢሺౙమభሻ
൙
ୣகக (3)
41
The IPCE measurements were performed using a three-electrode configuration with an InN working electrode, a Ag/AgCl reference electrode, and a platinum counter electrode. The electrodes were immersed in the same electrolyte as mentioned above. The quantum efficiency was recorded at an applied potential of 1.0 V vs. Ag/AgCl. A metal halide lamp was used with bandpass cut filters (400, 450, 500, 550, 600, 650, 700, and 880 nm). The incident light intensities were 3.0–8.0 mWcm·2 for each filter to obtain the IPCE using equation (4).
ሺΨሻ
ൌ ͳʹͶͲ˘ ሺɊ ିଶሻ
ሺሻ˘ሺ ିଶሻ
(4)
3.3 Results and discussion
3.3.1 Synthesis of InN nanocrystals
Fig. 3.1 shows the XRD pattern of the synthesized InN powders using LiCl-KCl as the molten salt at various temperatures. The diffraction peaks of the InN products can be indexed to the JCPDS data (PDF: 79-2498). At temperatures of 400°C, 450°C, and 500°C, no peaks of byproducts were detected, indicating the formation of single-phase InN crystals for all samples. A control reaction at 350°C, which is lower than the melting point of LiCl-KCl (353°C), was also carried out.
Peaks of InN and InCl3 were observed as byproducts, suggesting that LiCl-KCl could act as a reactant at temperatures below the melting point. Furthermore, at higher reaction temperatures (400°C and 450°C), the intensity of all the InN diffraction peaks increased when compared to that at 350°C. These observations indicate that LiCl-KCl plays a crucial role in facilitating the single-phase crystal
42
growth of InN. However, this trend was only observed up to 500°C. Above 550°C, the peak corresponding to the (101) plane significantly decreased, and Li3InN2
was concomitantly formed. This result was probably due to the decomposition of InN. The crystallite sizes along the (101) plane of InN are presented in Table 3.1.
The largest crystallite size of 53 nm was obtained from the sample prepared at 450°C.
Fig. 3.1 XRD patterns of InN synthesized using InCl3, LiNH2, and LiCl-KCl.
The reactions were carried out at 350°C, 400°C, 450°C, 500°C, and 550°C.
To characterize the product, XPS spectra were obtained (Fig. 3.2). For the reaction temperatures of 400°C, 450°C, and 500°C, the binding energies of In 3d5/2 and N 1s were 443.6 and 396.5 eV, respectively. The binding energies of In 3d5/2 and N 1s were consistent with the reported values for InN, and no peaks
43
corresponding to Cl 2p and Li 1s originating from byproducts were observed.29 For samples prepared at 350°C and 550°C, peaks at 57.5 eV and 198.9 eV corresponding to Li 1s and Cl 2p, respectively, were detected.
Fig. 3.2 XPS spectra of InN samples: (a) whole region, (b) In 3d, (c) Li 1s, (d) Cl 2p, and (e) K 2p.
(a) (b)
(c) (d)
(e)
44
SEM images of InN powder samples formed at 350, 400, 450, 500, and 550°C are shown in Fig. 3.3a and Fig. 3.4. The SEM images clearly indicated that with the use of LiCl-KCl as the molten salt, InN nanocrystals with grain sizes smaller than 100 nm were formed at 400°C, 450°C, and 500°C (Table 3.1). The largest grain size was 59 nm, which was measured for the sampled prepared at 450°C. The BET specific surface areas of these three samples had comparable values of approximately 12-14 m2/g, as summarized in Fig. 3.5.
Fig. 3.3 SEM images of InN samples prepared at 450°C by the reaction of (a) InCl3, LiNH2 in LiCl-KCl and (b) InCl3 and LiNH2 without LiCl-KCl.
(a)
(b)
45
Fig. 3.4 SEM images of InN crystals prepared by the reaction of InCl3, LiNH2 and LiCl-KCl at (a) 350°C, (b) 400°C, (c) 500°C, and (d) 550°C.
Fig. 3.5 Brunauer-Emmett-Teller (BET) isotherms and Barret-Joyner-Halenda (BJH) pore-size distribution (inset) of InN synthesized at 400°C, 450°C, and 500°C.
(a)
(d) (c)
(b)
46
The reaction yields and crystallite sizes for the (101) plane of InN synthesized at different temperatures are also shown in Table 3.1. The crystallite sizes along the (101) plane of InN synthesized at 400, 450, and 500°C were 41 nm, 53 nm, and 50 nm, respectively. The reaction yields increased with increasing reaction temperature from 400°C to 450°C. The highest reaction yield (72%) for the (101) plane was calculated for InN synthesized at 450°C.
Table 3.1 Reaction yields and structure characterization of InN nanocrystals.
InN sample
Reaction yields (%)
Specific surface area (m2/g)
Average grain size (nm)
Crystallite size*
(nm)
400°C 42 13 49 41
450°C 72 12 59 53
500°C 62 14 51 50
*Determined using the XRD data of InN corresponding to the (101) plane.
Fig. 3.6 XRD patterns of InN synthesized using InCl3 and LiNH2 without LiCl-KCl. The reactions were carried out at 400°C, 450°C, 500°C, and 550°C.
47
To understand the role of molten salt in crystal growth, we synthesized InN without LiCl-KCl (the XRD pattern and SEM images are presented in Figs.
3.6 and 3.7). In contrast to that of the molten salt synthesis, the XRD pattern exhibited peaks that were assigned to the byproduct InLi (Fig. 3.6). The SEM images indicated that the average grain size of InN synthesized without molten salt was significantly larger than that obtained using the molten salt (Fig. 3.7); Fig.
2b shows one of such SEM images (450°C sample). Considering that a random-shaped morphology with an average grain size of 200 nm was obtained, it is likely that the LiCl-KCl molten salt contributed to the suppression of InN growth, leading to the size reduction in the InN formation.
Fig. 3.7 SEM images of InN synthesized using InCl3 and LiNH2 without LiCl-KCl at (a) 400°C, (b) 450°C, (c) 500°C, and (d) 550°C.
(a) (b)
(c) (d)
48
The high resolution transmisson electron microscope (HR-TEM) images of InN nanocrystals synthesized at 450°C are also presented in Fig. 3.8a. A lattice fringe of 0.30 nm corresponding to the spacing of the (100) plane was observed (Fig. 8b).
Fig. 3.8 TEM images of (a) InN nanocrystals synthesized at 450°C for 2 h and (b) lattice fringe of (100) plane.
The plausible reactions for the formation of InN nanocrystals are shown below.
ʹଷ ଶ ՜ ʹଷଶ ͵ଶ ଶ ଶ (5) ʹଷଶ ͵ଶ ՜ ʹ ଶ (6) In the first step (equation (5)), the mixture of reactants (InCl3 and LiNH2) and the LiCl-KCl salt were uniformly heated, resulting in the decomposition of LiNH2 at
(a)
(b)
49
370°C to form Li3InN2 particles. It is presumed that the InN nanocrystals were formed by the decomposition of Li3InN2 in the second step (equation (6)).
Single-phase InN without any byproduct was observed at temperatures from 400°C to 500°C. Upon increasing the reaction temperature to 550°C, Li3InN2 was produced as a byproduct, as shown in the XRD pattern. The decreased intensity of InN indicated that InN started to decompose at approximately 550°C.
In a previous study, we successfully synthesized GaN nanoplates using LiCl as the molten salt. The molten salt could accelerate the homogenous formation at the nanoscale during the nucleation and crystal growth of GaN particles.26 The effect of LiCl-KCl on the crystal growth of InN can be explained as follows (Fig. 9): excess Cl· anions in the nitridation process (equation 6) can interact with the (100), (002), and (101) surfaces. This phenomenon probably leads to the suppression of crystal growth, producing InN nanocrystals with a narrow size distribution. However, without the use of LiCl-KCl as the molten salt, it is likely that the lack of Cl· anions resulted in the formation of random-shaped particles. In addition, we have synthesized the InGaN synthesis as a preliminary experiment using GaCl3 and InCl3 as metal sources in LiCl-KCl at a temperature of 400°C. XRD pattern showed that the major diffraction peaks of the products can be indexed to InGaN. However, it turned out that the InGaN product contained only 10% composition of In. A more through study for optimizing InN synthesis is necessary. Nevertheless, we believe that our nitridation-based molten salt reaction provides a generalized approach to create versatile group III-nitride semiconductors.
