Synthesis of red-emitting nanocrystalline phosphor CaAlSiN₃:Eu²⁺ derived from elementary constituents

Abstract

A highly efficient red-emitting nitride phosphor of CaAlSiN₃:Eu²⁺ is prepared via one-pot ammonothermal synthetic route starting from elementary constituents of Ca, Al, Si, and Eu. Respective elements are able to be dissolved in the supercritical solution of ammonia, transformed into intermediates of metal amides, and consecutively converted to metal nitride of nanocrystalline CaAlSiN₃:Eu²⁺.

---

In the past few decades, nitridosilicates (or nitrides) as well as oxonitridosilicates (or oxynitrides) have attracted a great deal of research interest because they have excellent mechanical, chemical and thermal stabilities as well as structural diversities. In this regard, much attention has been paid to investigate the Eu²⁺-doped nitride and oxynitride phosphor materials such as α-and β-SiAlON,^1,2 M₂Si₅N₈ (M = Ca, Sr, Ba),³ MSi₂O₂N₂ (M = Ca, Sr, Ba),⁴ and CaAlSiN₃⁵ because they are the promising phosphor candidates used for industrial applications in white light-emitting diodes (LEDs). Among them, CaAlSiN₃:Eu²⁺ showed superb red-emitting luminescent performance due to its high quantum efficiency (∼80%), radiation stability, and remarkable thermal quenching behavior.⁵

In terms of the synthesis of CaAlSiN₃:Eu²⁺ (shortly CASN:Eu²⁺) phosphors, a wide range of conventional methods have been reported: (1) solid state reaction of constituent nitrides at high temperature (SSR: 1600–1800 °C),⁶ (2) carbothermal reduction and nitridation of oxide (CRN: 1400–1500 °C),⁷ and (3) self-propagating high-temperature synthesis from a CaAlSi alloy (SHS: 1450–1550 °C).⁸ Recently, a synthesis from alloy-derived ammonometallates in ammonia solution has been reported.^9,10 However, this method had serious difficulty controlling the composition of products because alloy materials of Ca_1−xAlSi:Eu_x with a fixed composition rate were pre-synthesized deliberately in order to change the composition rate of CASN:Eu²⁺. In addition, it hampered the mechanistic understanding of the formation of multinary nitrides since it did not follow the dissolution–crystallization like sol–gel process as the starting materials were from alloy materials, not from metal constituents.

To the best of our knowledge, there has been no report that multinary nitrides can be synthesized in supercritical ammonia solution from elementary constituents. Therefore, to improve the limitations in the previous method, for the first time, we have developed and systematically investigated the preparation of Eu²⁺-doped nitridosilicates of CASN:Eu²⁺ through the dissolution–crystallization process starting from elementary constituents such as Ca, Al, Si, and Eu at a lower reaction temperature of 580 °C with an energy-efficient and benign process. This novel solution-based one-pot approach provides us with not only a convenient way to control the composition rate of products, but also a deep insight into how respective metals convert to amides, to imides, and ultimately to multinary nitrides in the ammonia system.

It is well known that oxide materials are synthesized in aqueous media and nitride materials are also able to be synthesized in ammonia solution in the same sol–gel-like process. In the ammonia system, the elementary constituents can be dissolved to amides, and then condensed to imides and finally to nitrides in a sequence with releasing ammonia.¹¹ This is highly dependent on the reaction temperature and pressure of the ammonia system. In the meantime, in order to increase the solubility of constituents (metals) as well as increase the speed of crystallization, mineralizers are typically used as they can play a significant role in facilitating the synthesis by the formation of the intermediate of metal complex.¹² In our system, sodium azide is employed as a mineralizer (or flux) because sodium azide is chemically more stable than sodium amide. Sodium amide is easily contaminated by oxygen due to its high reactivity toward oxygen or moisture and sodium azide can generate highly purified sodium and nitrogen when thermally decomposed between 250–300 °C, generating sodium amide and hydrogen simultaneously.¹²

It has been reported that metal elements of Ca,^13,14 Al,¹⁵ Si (semimetal),¹⁶ and Eu^17,18 are allowed to be soluble in sodium amide–ammonia solution. Many studies have been conducted on the formation of a variety of metal amides, indicating that all elements employed in our synthetic system such as Ca, Al, Si, and Eu are able to react with fused alkali metal amide (e.g. NaNH₂) or ammonia solution. It was well known that metal elements in ammonia are able to form metal amides. Zeuner et al. reported that metal amides could be used as reactive precursors prepared by dissolving metals in liquid and/or supercritical ammonia because of the low decomposition temperature of metal amides.^19,20 In addition, in the presence of sodium flux, Li et al. referred metal amides or its polymerized products as ammonometallates, converted from pure metals.⁹ Thus, in our synthetic conditions of mixed sodium amide–ammonia solution, respective metals are easily dissolved in the sodium amide–ammonia solution because metals can be converted to sodium ammonometallates by forming the adducts of metal amide and sodium amide, and these precursors could decompose to multinary metal imide and nitride in order.^11,12 After the complete dissolution of metals in ammonia, the crystallization process occurred for the synthesis of multinary nitride. The schematic description of the synthesis is presented in Scheme 1; Eu was excluded for simplicity. Among each metal of Ca (Eu), Al, and Si, the solubility in sodium amide–ammonia system is in the order: Si < Al < Ca (Eu). This is mainly attributed to the reactivity of each metal and the number of amide bonds to be made between metal cation and amide ion. For calcium, two bonds between the calcium cation and amide ion should be made in a sodium amide–ammonia system like Ca(NH₂)₂·Na(NH)₂ while aluminum and silicon should have each structure like Al(NH₂)₃·Na(NH)₂ and Si(NH₂)₄·Na(NH)₂ according to the valence charge of metals. Consequently, silicon is reported as the least reactive metal between them and is only slightly dissolved in sodium amide–ammonia solution at 350–400 °C.¹⁶

Scheme 1 Schematic diagram for the synthesis of multinary nitride of CaAlSiN₃:Eu²⁺, starting from the elements of Ca, Al, Si, and Eu.

First of all, the preliminary experiments were conducted for the synthesis of binary Si₃N₄, starting from an element of Si and NaN₃ in the supercritical ammonia solution in an attempt to check the feasibility of the synthesis of multinary nitride. It was interesting to note that the precursor of sodium ammonosilicate started to decompose to form NaSi₂N₃ with sodium remaining in the host crystal structure.²¹ However, it was suggested that CASN:Eu²⁺ can be synthesized when Ca and Al are introduced to the synthesis of NaSi₂N₃ system. This is because NaSi₂N₃ and CASN:Eu²⁺ had the same space group, Cmc2₁, and orthorhombic lattice structure with the close lattice parameters from ICDD database: PDF# 01-081-1098 and PDF# 97-016-1796. The XRD patterns showed that well crystallized NaSi₂N₃ having the same crystal structure with CASN was successfully synthesized, indicating that the least soluble element of Si was able to be dissolved and reacted in sodium amide–ammonia solution at the low temperature of 580 °C (Fig. 1). This strongly encouraged us that Si can be dissolved and converted to the intermediate of sodium ammonosilicate, and condensed to sodium silicon nitride through the ammonothermal process after that.

Fig. 1 XRD patterns of NaSi₂N₃ under the reaction of Si and NaN₃ at the reaction condition of 580 °C for 10 days.

Since the least reactive element of Si could be dissolved in the melted sodium amide–ammonia system, we firmly believed that the intermediates of respective sodium ammonometallates of Ca, Al, Si, and Eu can be formed after they were completely dissolved in the ammonia system and simultaneous conversion to nitride occurred because the solubility of other materials is better than that of silicon.

Second, we investigated the intermediate products of CaAlSiN₃:Eu²⁺ obtained at the reaction temperature of 500 °C for 50 hours to examine the intermediate structure and the dissolution of pure metals. The XRD patterns of the as-prepared sample showed that any strong peaks from pure metals were not observed except for the peaks from Na₂CaSiO₄, Na₂SiO₃, and AlN. These were suggested to be oxidized from the binary or ternary intermediates. It has been reported that binary metal amides adducted with NaNH₂ are existed such as KEu(NH₂)₃,¹⁷ NaCa(NH₂)₃,²² NaAl(NH₂)₄²³ and, NaSi(NH₂)₅²¹ Therefore, metal elements were completely dissolved in ammonia with a mineralizer and transformed into the intermediates of amide, and then condensed to imide or nitride to some extent. However, after washing treatment, XRD patterns from CaAlSiN₃ and unreacted Si were observed (Fig. 2). This is attributed to not homogeneously stirring the sample during the reaction. As a result, a large amount of unreacted Si element staying at the bottom of basket were analyzed. However, more importantly, it was suggested that the least reactive element, Si was not fully converted to silicon amide at that reaction condition, resulting in strong peaks of raw Si in the XRD patterns. Nevertheless, some part of Si was already dissolved and reacted to form CASN because the color of the acquired sample changed to gray in comparison with silver and black of starting powders. Moreover, collected samples had a size of a few hundred nanometers. As shown in Fig. S1, ESI,† samples before washing showed a nanobar-like shape, but after washing they were supposed to be dissolved in a mixed solvent of water or ethanol, with only round-shaped plates remaining. Compared with SEM images of the raw starting materials with a relatively large size in few mircometers or millimeters, depicted in Fig. S2, ESI,† it was shown that the elementary constituents of metals were definitely dissolved in the sodium amide–ammonia solution and then the crystallization process occurred to generate nano-sized plates of CASN:Eu²⁺ phosphor.

Fig. 2 XRD patterns of intermediate products of CaAlSiN₃:Eu²⁺ phosphors collected at the reaction temperature of 500 °C for 50 h: (a) as-synthesized and (b) after washing treatment.

Finally, we studied the crystal phase, morphologies, compositions, and optical properties of the CaAlSiN₃:Eu²⁺ phosphors synthesized under the reaction at 580 °C for 20 days with various treatment conditions: (1) as-synthesized, (2) after washing treatment with water and ethanol, (3) after treated with 1 M HCl acid for 5 min, and (4) for 20 min, respectively. The XRD patterns shown in Fig. 3 indicated that CASN was successfully synthesized, consistent with the PDF card# 97-016-1796 of CaAl_0.54Si_1.38N₃. For the sample before washing, the impurity phases of AlN and CaO were detected, but they were removed after washing treatment. This is due to the facts that poorly crystallized powders were hydrolysed in water and washed out and the heaviest CASN powders were precipitated more rapidly than others during centrifugation: (density of CASN: 3.7919 g cm⁻³).⁸ When the sample was treated with 1 M HCl acid, the crystallinity of CASN:Eu²⁺ was slightly improved, but an impurity phase of Al₂O₃ appeared with increasing acid treatment time up to 20 min, showing that the residue of Al in the sample were most likely oxidized to Al₂O₃. In order to measure the atomic concentration of samples, EDS analysis was carried out. It was shown that both samples before and after washing contained a high oxygen content (∼40%). This is due to the fact that unreacted ammonometallates were oxidized to amorphous metal oxide when the reactor was unsealed to air. However, after treatment with acid, the oxygen content in the sample drastically decreased as these amorphous oxide residues were removed by acid treatment, while nitride components remained because they were resistant enough to the acid treatment (Table S1, ESI†).

Fig. 3 XRD patterns of synthesized CaAlSiN₃:Eu²⁺ phosphors with various treatment conditions: (a) as-synthesized and (b) after washing treatment, (c) after acid treatment 1 M HCl for 5 min, and (d) for 20 min, respectively.

SEM images of the CASN:Eu²⁺ phosphors shown in Fig. 4 showed that plate-like nanocrystals were synthesized before and after washing, which was consistent with previous reports.¹⁰ It was speculated that, in the prolonged reaction time, nano-bars or nano-particle-like crystals seemed to aggregate to form nano-plate-like crystals in comparison with the shape of sample acquired at the early reaction period at 500 °C for 50 h (i.e. nanorods or nanoparticles). Accordingly, a small amount of remaining nanorod or nanoparticle crystals are also observed in Fig. 2(a and b). However, the research on the detailed mechanism for the formation of nano-plates of CASN:Eu²⁺ has not been conducted yet. Despite the acid treatment, the size and morphology of the samples remain the same.

Fig. 4 SEM images of synthesized CaAlSiN₃:Eu²⁺ phosphors with various treatment conditions: (a) as-synthesized and (b) after washing treatment, (c) after acid treatment 1 M HCl for 5 min, and (d) for 20 min, respectively.

The room-temperature PL excitation and emission spectra of CASN:Eu²⁺ were shown in Fig. 5. The excitation spectra covered the region from near UV to visible with the strongest two peaks located at 325 nm and 450 nm due to the excitation of Eu²⁺ ion transition from 4f⁷ to 4f⁶5d¹. The 5d level was strongly dependant on the outer crystal field and split by the ligand field strength of the local symmetry around the Eu²⁺ ions. In this case, the 5d orbital of Eu²⁺ is split into two levels such as T_2g and E_g due to the tetrahedral symmetry of surrounding rigid network of [AlN₄] and [SiN₄]. Under excitation of 450 nm, strong red emission peak centered at 650 nm was observed. After washing treatment with water and acid, the both PL excitation and emission intensity increased because non-emitting amorphous materials were removed by washing treatment. Acid treatment enabled an slight increase in the PL emission intensity. However, they showed a relatively low PL efficiency of ∼5% compared to conventionally prepared phosphors (Fig. S3, ESI†) due to an extremely low reaction temperature of 580 °C instead of 1600 °C, but had an excellent thermal stability. The temperature-dependent PL emission showed that PL intensity remained at 75% of the initial value with heating sample up to 180 °C, shown in Fig. S4, ESI.†

Fig. 5 PL excitation and emission spectra of CaAlSiN₃:Eu²⁺ from bottom to top: as-synthesized, after washing treatment, after acid treatment 1 M HCl for 5 min, and for 20 min, respectively.

Additionally, in order to evaluate the effect of a mineralizer and reaction time, the amount of sodium azide and reaction time were altered respectively. The more sodium azide employed in the ammonia system, the more crystallized CASN:Eu²⁺ powders were synthesized, with the improved XRD patterns shown in Fig. S5, ESI.† This is because sodium azide played an important role in the facilitation of the dissolution of raw materials, nucleation and growth of multinary crystals. When an insufficient amount of mineralizer was used, such as the molar ratio of Na/(Ca + Eu) = 0.5, a large amount of unreacted Si was found, indicating that a mineralizer was really important to dissolve the pure metal solutes, especially Si due to the least reactive element, in ammonia solution. In a prolonged reaction period of up to 30 days, more converted CASN:Eu²⁺ powders were prepared with the conversion yield of up to 85% (Fig. S6, ESI†).

Overall, all elements were dissolved in sodium amide–ammonia solution to form ammometallates of Ca, Al, Si and Eu at 400 °C, and these precursors were condensed together to form multinary intermediates with increasing the reaction temperature from 400 °C to 580 °C: from the binary amide of M–(NH₂)_x (M = Ca(Eu), Al, Si) to the ternary amide imides of Ca–(NH)–Al, Ca–(NH)–Si, and Al–(NH)–Si, and the quaternary imide nitrides of Ca–N–(Si)Al.^11,24–27 An indication of the adduct type with NaNH₂ were omitted for the simplicity. At the critical temperature, pressure, and concentration of multinary nitrides containing Ca, Al, Si, and N, they were suggested to decompose to a seed of CASN:Eu²⁺ accompanying simultaneous a nucleation and a following crystallization growth. Unfortunately, at present, it is uncertain whether the intermediate of ammonometallates or their condensed forms (i.e. imides, imide nitrides, nitrides) exist at the molecular level or partially polymerized level depending on their size. However, it was thought that each ammonometallate existing at the molecular level may condense each other promptly to the certain polymerized products containing quaternary elements, considering the reactivity of amides, which would be used a seed of CASN:Eu²⁺ later. It is difficult to characterize intermediate complex because the ammonometallates are so reactive in air that they transformed to an amorphous phase with a low crystallinity immediately after the reactor was unsealed under atmospheric conditions. Thus, we are in the process of systematically investigating the intermediate complexes to fully understand the formation mechanism of multinary nitrides, especially from elementary constitutents.

In conclusion, the novel ammonothermal synthesis of CASN:Eu²⁺, starting from respective metal constituents was investigated. This synthetic method is an easy and convenient way to control the composition ratio of products, and provides insight into understanding the formation mechanism of the multinary nitride of CASN:Eu²⁺ under a supercritical ammonia system with a sequence of metal, sodium ammonometallates, imide, and multinary nitride at a lower reaction temperature of 580 °C. It allows us to synthesize various nitrides via elementary metals-driven ammonothermal process.

Acknowledgements

This work is financially supported by Korea Research Council Industrial Science and Technology (KK-1307-B9).

Notes and references

1. R.-J. Xie, N. Hirosaki, M. Mitomo, Y. Yamamoto, T. Suehiro and K. Sakuma, J. Phys. Chem. B, 2004, 108, 12027–12031.
2. N. Hirosaki, R.-J. Xie, K. Kimoto, T. Sekiguchi, Y. Yamamoto, T. Suehiro and M. Mitomo, Appl. Phys. Lett., 2005, 86, 211905.
3. R. Mueller-Mach, G. Mueller, M. R. Krames, H. A. Höppe, F. Stadler, W. Schnick, T. Juestel and P. Schmidt, Phys. Status Solidi A, 2005, 202, 1727–1732.
4. Y. Q. Li, A. C. A. Delsing, G. de With and H. T. Hintzen, Chem. Mater., 2005, 17, 3242–3248.
5. K. Uheda, N. Hirosaki, Y. Yamamoto, A. Naito, T. Nakajima and H. Yamamoto, Electrochem. Solid-State Lett., 2006, 9, H22–H25.
6. Y. Li, N. Hirosaki, R. Xie, T. Takeda and M. Mitomo, Chem. Mater., 2008, 20, 6704–6714.
7. H. S. Kim, T. Horikawa, H. Hanzawa and K.-i. Machida, J. Phys.: Conf. Ser., 2012, 379, 012016.
8. X. Piao, K.-i. Machida, T. Horikawa, H. Hanzawa, Y. Shimomura and N. Kijima, Chem. Mater., 2007, 19, 4592–4599.
9. J. Li, T. Watanabe, H. Wada, T. Setoyama and M. Yoshimura, Chem. Mater., 2007, 19, 3592–3594.
10. J. Li, T. Watanabe, N. Sakamoto, H. Wada, T. Setoyama and M. Yoshimura, Chem. Mater., 2008, 20, 2095–2105.
11. T. M. Richter and R. Niewa, Inorganics, 2014, 2, 29–78.
12. B. Wang and M. J. Callahan, Cryst. Growth Des., 2006, 6, 1227–1246.
13. J. Senker, H. Jacobs, M. Müller, W. Press, P. Müller, H. M. Mayer and R. M. Ibberson, J. Phys. Chem. B, 1998, 102, 931–940.
14. R. Juza and H. Schumacher, Zeitschrift für anorganische und allgemeine Chemie, 1963, 324, 278–286.
15. D. Peters, J. Cryst. Growth, 1990, 104, 411–418.
16. R. Levine and W. C. Fernelius, Chem. Rev., 1954, 54, 449–573.
17. H. Jacobs and U. Fink, Zeitschrift für anorganische allgemeine Chemie, 1978, 438, 151–159.
18. H. Imamura, Y. Sakata, Y. Tsuruwaka and S. Mise, J. Alloys Compd., 2006, 408, 1113–1117.
19. M. Zeuner, F. Hintze and W. Schnick, Chem. Mater., 2008, 21, 336–342.
20. M. Zeuner, P. J. Schmidt and W. Schnick, Chem. Mater., 2009, 21, 2467–2473.
21. H. Jacobs and H. Mengis, Eur. J. Solid State Inorg. Chem., 1993, 30, 45–53.
22. H. Jacobs and U. Fink, J. Less-Common Met., 1979, 63, 273–286.
23. H. Jacobs and B. Nöcker, Zeitschrift für anorganische und allgemeine Chemie, 1993, 619, 381–386.
24. K. N. Semenenko, B. M. Bulychev and E. A. Shevlyagina, Russ. Chem. Rev., 1966, 35, 649.
25. R. Juza, Angew. Chem., Int. Ed. Engl., 1964, 3, 471–481.
26. E. Wiberg and A. May, Naturforsch. B, 1955, 10, 229–230.
27. M. Blix and W. Wirbelauer, Berichte der deutschen chemischen Gesellschaft, 1903, 36, 4220–4228.

---

Footnote

† Electronic supplementary information (ESI) available: Details of experimental procedure, characterization, and figures. See DOI: 10.1039/c4ra02550d

---