Low temperature synthesis of LiSi₂N₃ nanobelts via molten salt nitridation and their photoluminescence properties

Abstract

LiSi₂N₃ nanobelts were synthesized using a novel low temperature molten salt nitridation technique using silicon and melamine as starting materials, and lithium chloride and sodium fluoride to form a reaction medium. The as-synthesized nanobelts were characterized by XRD, FESEM, HRTEM and SAED. The amount of LiSi₂N₃ increased with temperature. The optimal synthesis temperature for phase pure LiSi₂N₃ was about 1200 °C, which was about 200 °C lower than that required for the conventional solid-state reaction routes. LiSi₂N₃ nanobelts about a few hundred nanometers long and 50–200 nm in width were distributed uniformly in the final products. A possible growth mechanism was proposed based on the experimental results. Their photoluminescence emission at 459 nm (2.70 eV) at room temperature suggested that they could be potentially used in light-emitting nano-devices.

---

1. Introduction

Replacement of traditional energy-wasting light sources such as incandescent lamps with energy-efficient LED lamps is considered as a major step towards reduction of electrical energy consumption worldwide.¹ In recent years, nitrides have received remarkable attention due to their nontoxicity, interesting luminescence properties, and potential applications as phosphors and pigments.^2,3

Ternary lithium silicon nitrides are one of such interesting materials for luminescent applications.^4,5 Several ternary phases including LiSi₂N₃,⁶ Li₂SiN₂,⁷ Li₅SiN₃ (ref. 8) and Li₈SiN₄ (ref. 9) exist in the Li–Si–N ternary system, of which LiSi₂N₃ with a well-defined wurtzite type structure (space group Cmc2₁)¹⁰ is of particular interest because of its high stability for practical applications. It is generally synthesized via the conventional solid-state reaction route using Li₃N and Si₃N₄ as the starting materials. Unfortunately, this synthesis method suffers from several disadvantages including requirements of high pressure and temperature,^6,11,12 and formation of heavily agglomerated LiSi₂N₃ with large particle size and spheroid morphology, which limits its functional applications.¹³ To overcome these and to prepare high quality LiSi₂N₃ nanomaterials with novel morphologies and photoluminescence (PL) properties, an alternative processing route needs to be developed.

In response to this, LiSi₂N₃ nanobelts have been synthesized for the first time by using a novel low temperature molten salt nitridation (MSN) technique from relatively cheap Si powders and lithium chloride in this work. The effects of reaction temperature on phase composition and morphology of the final products were investigated, and PL properties of as-synthesized LiSi₂N₃ nanobelts examined.

2. Experimental procedure

Silicon (Si, purity ≥ 99 wt%, particle size ≤ 2 μm, Aladdin) and melamine (C₃N₃(NH₂)₃, purity ≥ 99 wt%, particle size ≤ 5 μm, Aladdin) powders were used as starting materials. They were pre-mixed and further combined with a LiCl–NaF binary salt containing 20 wt% NaF. The mixed powder batch was placed in an alumina crucible with a lid and heated at 5 °C min⁻¹ in a N₂ protected alumina-tube furnace to a temperature between 900 and 1200 °C for 3 h. After furnace-cooling to room temperature, the reacted mass was washed repeatedly with distilled water to remove the residual medium salt. The resultant powders were oven-dried overnight at 110 °C prior to further characterization.

Phases in product powders were identified by using an X-ray diffractometer (XRD, X'Pert Pro, Philips, Netherlands) with Cu-Kα radiation (λ = 1.5406 Å) operated at 40 kV and 40 mA, with a scanning rate of 2° (2θ) min⁻¹ and a step size of 0.02° (2θ). Morphologies of product phases were examined using a field emission scanning electron microscope (FESEM, Novo 400, FEI Co., USA), a transmission electron microscope (TEM, 2000F, Jeol Ltd., Japan) and selected area electron diffraction (SAED). UV-vis and room-temperature PL spectra were recorded respectively using a UV-vis spectrophotometer (Shimadzu UV-3600, Japan), and a fluorescence spectrophotometer (PerkinElmer LS 55, USA).

3. Results and discussion

Fig. 1 shows XRD patterns of product samples resultant from 3 h firing at various temperatures. At 900 °C, no LiSi₂N₃ was detected and only unreacted Si remained. At 1000 °C, LiSi₂N₃ started to appear and Si decreased, suggesting the formation reaction of LiSi₂N₃ just occur. On increasing the firing temperature to 1100 °C, LiSi₂N₃ peaks increased whereas those of Si decreased, implying more LiSi₂N₃ was formed at the expense of Si. On further increasing the temperature to 1200 °C, Si disappeared and only LiSi₂N₃ was identified, indicating the complete conversion from Si to LiSi₂N₃. This synthesis temperature was about 200 °C lower than that (1400 °C) required by the conventional solid-state reaction method,⁴ indicating that the MSN technique introduced here can efficiently reduce the synthesis temperature of LiSi₂N₃. The composition of salt may play the key role in lowering the temperature, the reason was discussed in our previous literature.^14–16

Fig. 1 XRD patterns of product samples resultant from 3 h firing at various temperatures.

Illustrated in Fig. 2a–c are morphologies of samples resultant from 3 h firing at various temperatures in LiCl–NaF. At 1000 °C, many granular LiSi₂N₃ particles formed on the surface of Si (Fig. 2a). Upon increasing the temperature to 1100 °C, large amounts of nanobelt-like LiSi₂N₃ phases appeared. On further increasing the temperature to 1200 °C, LiSi₂N₃ nanobelts with a high aspect ratio of ∼10 were formed (Fig. 2b and c). They were a few hundred nanometers in length and 50–200 nm in width, after epitaxial growth on the surface of Si (Fig. 2c).

Fig. 2 SEM images showing morphologies of product phases in samples resultant from 3 h firing in LiCl–NaF at various temperatures: (a) 1000 °C, (b) 1100 °C, (c) 1200 °C (inset shows a high-magnification image), and (d) schematic of the growth mechanism of LiSi₂N₃ nanobelts.

The possible growth mechanism of these LiSi₂N₃ nanobelts can be schematically illustrated in the Fig. 2d and described as follows: in the initial stage, NaF in the LiCl–NaF binary salt diffused onto the Si surface and then reacted to form SiF₄(g) according to reaction (1). On the other hand, melamine (C₃N₃(NH₂)₃) decomposed in situ to produce NH₃ according to reaction (2) (step 1). In addition, LiCl interacted with Si(s), N₂(g) and NH₃(g), forming eutectic liquid droplets of Li–Si–N via reaction (3) (step 2). Upon oversaturation of the liquid droplets with LiSi₂N₃, the nucleation of LiSi₂N₃ would occur, followed by the growth of LiSi₂N₃ from the droplets (step 3). With increasing the temperature to 1200 °C, more gaseous species (e.g. LiCl, SiF₄, and NH₃) were generated and dissolved in the droplets, sustaining the growth of LiSi₂N₃ via reaction (4) (step 4). Owing to the orthorhombic structure nature (i.e., cell parameters a ≠ b ≠ c), the different planes of LiSi₂N₃ possess different surface energy, and crystal surfaces with lower energies tend to serve as the enclosure surfaces. During the preparation process, the Li, Si and N elements preferred to deposit on the high energy surfaces, finally resulting in simultaneous formation of nanobelt structures. On the basis of the reaction mechanism mentioned above and the observation of LiSi₂N₃ small particles (Fig. 2a), we proposed that the growth of the as-synthesized LiSi₂N₃ nanostructures was controlled by the classic VLS mechanism.¹⁷

4NaF(l) + Si(s) → SiF₄(g) + 4Na(g) (1)

2C₃N₃(NH₂)₃(s) → 3(CN)₂(g) + 4NH₃(g) + N₂(g) (2)

3LiCl(l) + 6Si(s) + 4N₂(g) + NH₃(g) → 3LiSi₂N₃(s) + 3HCl(g) (3)

LiCl(g) + 2SiF₄(g) + 3NH₃(g) → LiSi₂N₃(s) + 8HF(g) + HCl(g) (4)

TEM images were further performed along with the SAED patterns to assist identifying the crystalline structure of LiSi₂N₃ nanobelts resultant from 3 h firing at 1200 °C. Fig. 3a shows a low-magnification TEM image of a representative individual nanobelt, revealing that its aspect ratio was >10 and it had a smooth surface and a uniform width (about 150 nm). Fig. 3b further presents a high-resolution TEM (HRTEM) image of its edge area, showing that the interplaner distance was 0.33 nm which matched with the (111) plane of LiSi₂N₃. This result indicated that the LiSi₂N₃ nanobelt grew along the [111] direction, as suggested by the XRD results (Fig. 1). Moreover, the SAED pattern (inset in Fig. 3b) confirmed that the LiSi₂N₃ nanobelt was single crystalline in nature.

Fig. 3 TEM images of as-synthesized LiSi₂N₃ nanobelts. (a) A low-magnification TEM image of a representative individual nanobelt, (b) a high-resolution TEM image taken from the edge region of the nanobelt shown in (a). The inset was the representative SAED pattern of the nanobelt.

The UV-vis absorption spectrum of as-synthesized LiSi₂N₃ nanobelts (Fig. 4a) shows nearly zero absorbance in the visible range but significant absorbance in the UV region. A narrow absorption peak centered at 221 nm appeared, which corresponded to a band gap of ∼5.61 eV. This value was smaller than that (at ∼6.40 eV) reported previously for bulk LiSi₂N₃.⁴ Such a shift in the present sample could be attributed to the saddle point transition in the band structure.¹⁸ The optical band gap (OBG) estimated¹⁹ from the UV-vis absorption spectrum of as-synthesized LiSi₂N₃ nanobelts was ∼5.25 eV (see the inset in Fig. 4a). To further understand optoelectronic properties of as-synthesized LiSi₂N₃ nanobelts, their PL properties were also examined. Ultraviolet light used to excite the nanobelts was obtained from xenon lamp and its excitation wavelength was set at 221 nm. As seen from the room-temperature PL emission spectrum (Fig. 4b), intense luminescence in the violet-blue spectral range from 350 to 500 nm, with a main emission peak at around 459 nm (2.70 eV), occurred. Similar phenomena were also observed for BN nanoplates and AlN nanoneedles,^20,21 which are believed to arise from the surface effect (increased surface-to-volume ratios) and defect concentrations. Such explanation could also be used for present observed optical behavior of the as-synthesized LiSi₂N₃ nanobelts. However, detailed mechanisms on the PL properties of the LiSi₂N₃ nanobelts are not fully understood and require more systematic investigation. The intensive PL emission spectrum indicated that LiSi₂N₃ nanobelts prepared in this work could be potentially used in optical and optoelectronic devices such as LEDs, blue-light source, and UV detector.

Fig. 4 (a) UV-vis absorption spectrum and optical band gap analysis (inset), and (b) emission photoluminescence (PL) spectrum (λ_ex = 221 nm) of as-synthesized LiSi₂N₃ nanobelts.

4. Conclusions

A novel molten salt nitridation technique was successfully developed and used for the first time to synthesize single crystalline LiSi₂N₃ nanobelts. The synthesis temperature for phase pure LiSi₂N₃ was about 1200 °C, which was about 200 °C lower than that required by the conventional solid-state method. As-synthesized LiSi₂N₃ nanobelts were a few hundred nanometers long and 50–200 nm in width. The VLS growth mechanism is believed to have dominated the formation process of the LiSi₂N₃ nanobelts. The room-temperature PL spectrum suggested that as-synthesized LiSi₂N₃ nanobelts could be a promising candidate material for optical and optoelectronic applications.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (51502216, 51472184, 51472185), Natural Science Foundation of Hubei Province (China, Contract No. 2013CFA086), Foreign cooperation projects in Science and Technology of Hubei Province (Contract No. 2013BHE002) and China Postdoctoral Science Foundation (2014M560631).

References

1. C. Che and R. S. Liu, J. Phys. Chem. Lett., 2011, 2, 1268–1277.
2. R. J. Xie, J. Am. Ceram. Soc., 2013, 96, 665–687.
3. P. Pust, V. Weiler, C. Hecht, A. Tücks, A. S. Wochnik, A. K. Henß, D. Wiechert, C. Scheu, P. J. Schmidt and W. Schnick, Nat. Mater., 2014, 13, 891–896.
4. Y. Q. Li, N. Hirosaki, R. J. Xie, T. Takeka and M. Mitomo, J. Solid State Chem., 2009, 182, 301–311.
5. Q. Wu, Y. Li, X. Wang, Z. Zhao, C. Wang, H. Li, A. Mao and Y. Wang, RSC Adv., 2014, 4, 39030–390306.
6. E. Narimatsu, Y. Yamamoto, T. Nishimura and N. Hirosaki, J. Ceram. Soc. Jpn., 2010, 118, 837–841.
7. A. J. Anderson, R. G. Blair, S. M. Hick and R. B. Kaner, J. Mater. Chem., 2006, 16, 1318–1322.
8. A. T. Dadd and P. Hubberstey, J. Chem. Soc., 1981, 77, 1865–1870.
9. H. Yamane, S. Kikkawa and M. Koizumi, Solid State Ionics, 1987, 25, 183–191.
10. M. Orth and W. Schnick, Z. Anorg. Allg. Chem., 1999, 625, 1426–1428.
11. J. Ding, Q. Wu, Y. Li, Q. Long, C. Wang and Y. Wang, J. Am. Ceram. Soc., 2015, 98, 2523–2527.
12. N. Tapia-Ruiz, M. Segalés and D. H. Gregory, Coord. Chem. Rev., 2013, 257, 1978–2014.
13. X. G. Wan, J. M. Dong and D. Y. Xing, Phys. Rev. B: Condens. Matter Mater. Phys., 1998, 58, 6756.
14. J. Ye, S. Zhang and W. E. Lee, J. Eur. Ceram. Soc., 2013, 33, 2023–2029.
15. Z. Huang, H. Duan, J. Liu and H. Zhang, Ceram. Int., 2016, 42, 10482–10486.
16. Z. Huang, F. Li, C. Jiao, J. Liu, J. Huang, L. Lu, H. Zhang and S. Zhang, Ceram. Int., 2016, 42, 6221–6227.
17. Z. Peng, N. Zhu, X. Fu, C. Wang, Z. Fu, L. Qi and H. Mao, J. Am. Ceram. Soc., 2010, 93, 2264–2267.
18. A. Zunger, A. Katzir and A. Halperin, Phys. Rev. B: Solid State, 1976, 13, 5560–5573.
19. Y. Stehle, H. M. Meyer, R. R. Unocic, M. Kidder, G. Polizos, P. G. Datskos, R. Jackson, S. N. Smirnov and I. V. Vlassiouk, Chem. Mater., 2015, 27, 8041–8047.
20. L. Ye, L. Zhao, F. Liang, X. He, W. Fang, H. Chen, S. Zhang and S. An, Ceram. Int., 2015, 41, 14941–14948.
21. S. H. Shah, G. Nabi, W. S. Khan, A. Majid, C. Cao, S. Ali, M. Hussain, A. Nabi, S. Ishaq and F. K. Butt, Mater. Lett., 2013, 107, 255–258.

---