Key Engineering Materials, Volume 403 : SiAlONs and Non-oxidesSiAlONs and Non-oxides
Selected, peer reviewed papers from the 2nd International Symposium on SiAlONs and Non-oxides,
December 2nd – 5th, 2007 Ise-shima Royal Hotel, Mie, Japan
Edited by
Katsutoshi Komeya
Yi-Bing Cheng
Junichi Tatami
Mamoru Mitomo
Copyright 2009 Trans Tech Publications Ltd, Switzerland
All rights reserved. No part of the contents of this publication may be reproduced or
transmitted in any form or by any means without the written permission of the publisher.
Trans Tech Publications Ltd
Switzerland
http://www.ttp.net
Volume 403 of Key Engineering Materials ISSN 1013-9826 Full text available online at http://www.scientific.net
Distributed worldwide by and in the Americas by
Trans Tech Publications Ltd. Trans Tech Publications Inc.
Laubisrutistr. 24 PO Box 699, May Street
CH-8712 Stafa-Zurich Enfield, NH 03748
Switzerland USA
Fax: +41 (44) 922 10 33 Fax: +1 (603) 632-5611
e-mail: sales@ttp.net e-mail: sales-usa@ttp.net
Sponsors
Foreword
Nitrides
Preparation and Characterization of MM’Si4N6C Ceramics D.P. Thompson and Y. Zhang 3
The Phase Evolution in the Si3N4-AlN System after High-Energy Mechanical Treatment of the Precursor Powder M. Sopicka-Lizer, T. Pawlik, T. Wodek and M. Tacula 7
New Green Phosphor Ba3Si6O12N2:Eu for White LED: Crystal Structure and Optical Properties M. Mikami, S. Shimooka, K. Uheda, H. Imura and N. Kijima 11
Application of Nitride and Oxynitride Compounds to Various Phosphors for White LED K. Uheda 15
Fabrication of Electrically Conductive Si3N4 Ceramics by Dispersion of Carbon Nanotubes S. Yoshio, J. Tatami, T. Wakihara, K. Komeya and T. Meguro 19
Low Temperature Sintering of Si3N4 Ceramics and its Applicability as an Inert Matrix of the Transuranium Elements for Transmutation of Minor Actinides T. Yano, J. Yamane and K. Yoshida 23
2.45 GHz Microwave Sintering of Silicon Nitride S. Chockalingam, J.P. Kelly, V.R.W. Amarakoon and J.R. Varner 27
Sintering Shrinkage Behavior of Si3N4 Ceramics Prepared by a Post-Reaction Sintering Technique H. Yabuki, T. Wakihara, J. Tatami, K. Komeya, T. Meguro, H. Kita, N. Kondo and K. Hirao 31
Sintering Shrinkage Behavior and Mechanical Properties of HfO2-Added Si3N4 Ceramics D. Horikawa, J. Tatami, T. Wakihara, K. Komeya and T. Meguro 35
Fabrication and Evaluation of AlN–SiC Solid Solutions with p-Type Electrical Conduction R. Kobayashi, J. Tatami, T. Wakihara, K. Komeya, T. Meguro, R. Tu and T. Goto 39
Atomic Resolution and In Situ Characterization of Structural Ceramics Y. Ikuhara 43
Non-Oxide Ceramic Nanocomposites with Multifunctionality T. Kusunose and T. Sekino 45
Electrical Resistivity Control of Hot-Pressed Aluminum Nitride Ceramics N. Yamada, J. Yoshikawa, Y. Katsuda and H. Sakai 49
Fracture Resistance and Wear Properties of Silicon Nitride Ceramics H. Miyazaki, H. Hyuga, Y. Yoshizawa, K. Hirao and T. Ohji 53
Oxidation of Rare Earth Silicon Oxynitride J-Phases J. Takahashi and T. Suehiro 57
Effect of Second Phase After-Heat Treatment on the Thermal Conductivity of AlN Ceramics H.K. Lee and D.K. Kim 61
Thermal Conductivity Measurement of the AlN Ceramics at the Grain Scale Using Thermoreflectance Technique S.K. Lee, I. Yamada, S. Kume, H. Nakano and K. Watari 65
Viscosity Measurement of Molten RE-Mg-Si-O-N (RE=Y, Gd, Nd and La) Glasses N. Saito, D. Nakata, S. Sukenaga and K. Nakashima 69
b SiAlONs and Non-oxides
First Principles Calculations of Advanced Nitrides, Oxides and Alloys I. Tanaka, A. Kuwabara, K. Yuge, A. Seko, F. Oba and K. Matsunaga 73
Advances in Computation of Temperature-Pressure Phase Diagrams of High-Pressure Nitrides P. Kroll 77
SiAlONs
Luminescence Properties of α-SiAlONs and Related Compounds R.J. Xie, M. Mitomo and N. Hirosaki 83
Developments in SiAlON Glasses and their Derivatives: Effects of Chemistry on Properties S. Hampshire and M.J. Pomeroy 87
Controlled Crystallisation of a Y-Si-Al-O-N Glass Typical of Grain Boundary Glasses Formed in Silicon Nitride-Based Ceramics M.J. Pomeroy and S. Hampshire 91
Synthesis and Refinement of β-SiAlON by Nitriding and Post-Sintering of Si Mixture Y.J. Park, E.A. Noh, J.H. Ahn and H.D. Kim 95
Interactions between AlN and SiAlON Ceramics A. Kalemtas, N.C. Acikbas, F. Kara, H. Mandal, K. Krnel and T. Kosma 97
Surface Structuring of α/β-Sialon Ceramics by Plasma-Etching M. Riva, R. Oberacker, M.J. Hoffmann and C. Ziebert 99
SiAlON B-Phase Glass-Ceramic Microstructures L.K.L. Falk, Y. Menke and S. Hampshire 103
Development of α-β SiAlON Ceramics from Different Si3N4 Starting Powders N.C. Acikbas, F. Kara and H. Mandal 107
Effects of Process on Optical Transmittance of Dy-α-Sialon Sintered at Lower Temperatures J.M. Xue, Q. Liu, M. Fang, L.L. Ma, T.P. Xiu and L.H. Gui 109
Mechanical Properties of α- and β-SiAlON Composite Ceramics Using β-SiAlON Powder K. Asakoshi, J. Tatami, K. Komeya, T. Meguro and M. Yokouchi 111
Tribological Performance of Translucent Dy-α-Sialon Ceramics Q. Liu, L.H. Gui, J.H. Meng and Z.F. Li 115
High-Temperature Compressive Deformation of SiAlON Polycrystals Prepared without Additives K. Chihara, D. Hiratsuka, J. Tatami, F. Wakai and K. Komeya 117
Dielectric Properties of β-SiAlON at High Temperature Using Perturbation Method Y.H. Seong, H.N. Kim and D.K. Kim 121
Dielectric Properties of SiAlON Ceramics D.K. Kim, H.N. Kim, Y.H. Seong, S.S. Baek, E.S. Kang and Y.G. Baek 125
Subcritical Crack Growth of α/β-Sialon Ceramics in Distilled Water M. Riva, R. Oberacker, M.J. Hoffmann and T. Fett 129
Corrosion of β-SiAlON in Molten Aluminium, Cryolite and NaCl-KCl Mixture T. Plachký, J. Kesan, M. Korenko, Z. Lenéš and P. Šajgalík 133
Preparation and Corrosion of Mullite Thin Film on ß-SiAlON Ceramics Y. Noritake, H. Kiyono and S. Shimada 135
Thermal Conductivities of β-SiAlONs by Mechanically Activated Combustion Synthesis R. Sivakumar, K. Aoyagi, T. Watanabe and T. Akiyama 139
Synthesis and Characterization of β-SiAlON Phosphor Powder Prepared by Reduction Nitridation of a Zeolite T. Wakihara, Y. Saito, J. Tatami, K. Komeya, T. Meguro, Y. Fukuda, N. Matsuda and H. Asai 141
Carbides
New Ceramic Phases in the Ternary Si-C-N System R. Riedel, E. Horvath-Bordon, H.J. Kleebe, P. Kroll, G. Miehe, P.A. van Aken and S. Lauterbach 147
A New Route of Forming Silicon Carbide Nanostructures with Controlled Morphologies Y.B. Cheng, K. Wang and H.T. Wang 149
Key Engineering Materials Vol. 403 c
TEM Study of SiCN Glasses; Polymer Architecture versus Ceramic Microstructure H.J. Kleebe and H. Störmer 153
Aqueous Processing of Boron Carbide Powders J.X. Zhang, Q.L. Lin and D.L. Jiang 157
Synthesis of Single-Phase, Hexagonal Plate-Like Al4SiC4 Powder J.S. Lee, T. Nishimura, H. Tanaka and S.H. Lee 159
Gelcasting of Carbide Ceramics D.L. Jiang 163
Processing of High Performance Silicon Carbide Y. Hirata, N. Matsunaga, N. Hidaka, S. Tabata and S. Sameshima 165
New Theory of Transformation Induced Grain Growth in Porous SiC H. Tanaka 169
Evaluation of Microstructure of High-Strength Reaction-Sintered Silicon Carbide S. Suyama and Y. Itoh 173
Nano-Grained Microstructure Design of Silicon Carbide Ceramics by SPS Process J. Hojo, H. Matsuura and M. Hotta 177
Macro- and Micro-Scale Thermal Conductivities of SiC Single Crystal and Ceramic I. Yamada, S. Kume, H. Nakano and K. Watari 179
Influence of Additives on Mechanical Properties in Liquid-Phase Sintered Silicon Carbide Ceramics Y.W. Kim, T. Nishimura and M. Mitomo 185
Microstructural and Mechanical Properties of Ti3SiC2 Composites J.L. Huang and H.H. Lu 189
Effect of HfO2 Coating Films on Oxidation Resistance of SiC Ceramics M. Kasajima, T. Akashi and S. Shimada 193
High Temperature Oxidation of SiC Powder in Oxidizing Atmosphere Containing Water Vapor T. Akashi, M. Kasajima, C. Muraoka and H. Kiyono 197
Applying SiC Nanoparticles to Functional Ceramics for Semiconductor Manufacturing Process M. Konishi 201
First-Principles Study of Ceramic Interfaces: Structures and Electronic and Mechanical Properties M. Kohyama and S. Tanaka 205
A New Technology with Porous Materials; Progress in the Development of the Diesel Vehicle Business K. Ohno 207
Synthesis of Nano-Sized SiC Powders by Carbothermal Reduction Y. Yoshioka, H. Tanaka, M. Konishi and T. Nishimura 211
Composites
Passive Oxidation Behavior of ZrB2-SiC Eutectic Composite Prepared by Arc Melting R. Tu, H. Hirayama and T. Goto 217
Fabrication and Mechanical Properties of TiN Nanoparticle-Dispersed Si3N4 Ceramics from Si3N4-Nano TiO2 Composite Particles Obtained by Mechanical Treatment E. Kodama, J. Tatami, T. Wakihara, T. Meguro, K. Komeya and H. Nakano 221
Spark Plasma Sintering of Silicon Nitride-Boron Carbide Composites E. Ayas, A. Kalemtas, G. Arslan, A. Kara and F. Kara 225
In Situ Synthesis and Mechanical Properties of TiN-Si2N2O-Si3N4 Composites H. Kiyono and S. Shimada 227
Dispersion of Carbon Nanotubes in Polysilazanes for the Preparation of Reinforced Si-C-N Composites L. Fernandez, Y. Li, M. Burghard, Z. Burghard, P. Gerstel, J. Bill and F. Aldinger 231
Reaction Bonded Silicon Nitride - Silicon Carbide and SiAlON - Silicon Carbide Refractories for Aluminium Smelting M.I. Jones, R. Etzion, J. Metson, Y. Zhou, H. Hyuga, Y. Yoshizawa and K. Hirao 235
d SiAlONs and Non-oxides
Tribological Characteristics of Carbon Nano-Fiber Dispersed Silicon Nitride Based Composites in High-Temperature Fuel M. Wada, K. Kashiwagi, S. Kitaoka and Y. Fuwa 239
Preparation of β SiAlON-cBN Composites by Spark Plasma Sintering M. Hotta and T. Goto 241
A Novel Approach for Preparing Electrically Conductive SiAlON-TiN Composites by Spark Plasma Sintering E. Ayas, A. Kara and F. Kara 243
Fabrication of AlN Ceramics Using AlN and Nano-Y2O3 Composite Particles Prepared by Mechanical Treatment D. Hiratsuka, J. Tatami, T. Wakihara, K. Komeya and T. Meguro 245
Preparation of Precursors for Aluminum Nitride-Based Ceramic Composites from Cage- Type and Cyclic Building Blocks Y. Sugahara, H. Nakashima, S. Koyama and Y. Mori 249
Pressureless Melt Infiltrated Non-Oxide Ceramic-Metal Composites A. Kalemtas, G. Arslan and F. Kara 251
The Oxidation Behavior of ZrB2-Based Mixed Boride S.J. Lee and D.K. Kim 253
Mechanical and Thermal Properties of Silicon Carbide Composites with Chopped Si-Al-C Fiber Addition K. Itatani, I.J. Davies and H. Suemasu 257
Exergy Analysis on the Life Cycle of Ceramic Parts H. Kita, H. Hyuga, N. Kondo and T. Ohji 261
SiAlON Microstructures L.K.L. Falk 265
Synthesis of Non-Oxide Ceramic Fine-Powders from Organic Precursors T. Nishimura, S. Ishihara, Y. Yoshioka and H. Tanaka 269
Sponsored by The Japan Society for Promotion of Science (JSPS)
Yokohama National University JSPS 124th Committee
Hosokawa Powder Technology Foundation
COMMITTEE MEMBERS
International Advisory Committee
A. Bellosi (CNR-IRTEC, Italy) G. Fantozzi (GEMPPM, France) L. Gao (Shanghai Institute of Ceramics, China) H. T. Hintzen (Eindhoven University of Technology, The Netherlands) J. Hojo (Kyushu University, Japan) M. Ibukiyama (Denki Kagaku Kogyo, Japan) Y. Ito (Toshiba Co., Japan) S. J. Kang (Korea Advanced Institute of Science and Technology, Korea) D. K. Kim (Korea Advanced Institute of Science and Technology, Korea) H. D. Kim (Korea Institute of Machinery & Materials, Korea) H. T. Lin (Oak Ridge National Laboratory, USA) K. MacKenzie (Victoria University of Wellington, New Zealand) Y. Matsuo (Tokyo Institute of Technology, Japan) K. Niihara (Nagaoka University of Technology, Japan) R. Riedel (Technical University Darmstadt, Germany) T. Rouxel (University of Rennes, France) P. Sajgalik (Slovak Academy of Science, Slovakia) H. Sakai (NGK Insulator Ltd, Japan) Z. -J. Shen (Stockholm University, Sweden) M. Singh (NASA Glenn Research Center, U.S.A.) K. Uematsu (Nagaoka University Technology, Japan) F. Wakai (Tokyo Institute of Technology, Japan) A. Yamaguchi (JUTEM, Japan) G. Zhang (Shanghai Institute of Ceramics, China)
Sponsored by The Japan Society for Promotion of Science (JSPS)
Yokohama National University JSPS 124th Committee
Hosokawa Powder Technology Foundation
COMMITTEE MEMBERS
International Advisory Committee
A. Bellosi (CNR-IRTEC, Italy) G. Fantozzi (GEMPPM, France) L. Gao (Shanghai Institute of Ceramics, China) H. T. Hintzen (Eindhoven University of Technology, The Netherlands) J. Hojo (Kyushu University, Japan) M. Ibukiyama (Denki Kagaku Kogyo, Japan) Y. Ito (Toshiba Co., Japan) S. J. Kang (Korea Advanced Institute of Science and Technology, Korea) D. K. Kim (Korea Advanced Institute of Science and Technology, Korea) H. D. Kim (Korea Institute of Machinery & Materials, Korea) H. T. Lin (Oak Ridge National Laboratory, USA) K. MacKenzie (Victoria University of Wellington, New Zealand) Y. Matsuo (Tokyo Institute of Technology, Japan) K. Niihara (Nagaoka University of Technology, Japan) R. Riedel (Technical University Darmstadt, Germany) T. Rouxel (University of Rennes, France) P. Sajgalik (Slovak Academy of Science, Slovakia) H. Sakai (NGK Insulator Ltd, Japan) Z. -J. Shen (Stockholm University, Sweden) M. Singh (NASA Glenn Research Center, U.S.A.) K. Uematsu (Nagaoka University Technology, Japan) F. Wakai (Tokyo Institute of Technology, Japan) A. Yamaguchi (JUTEM, Japan) G. Zhang (Shanghai Institute of Ceramics, China)
Organising Committee
Chairman: K. Komeya (Yokohama National University, Japan) Vice-Chairman: D. P. Thompson (University of Newcastle-upon-Tyne, UK) M. Mitomo (National Institute for Materials Science, Japan) Members: Y. -B. Cheng (Monash University, Australia) T. Goto (Tohoku University, Japan) S. Hampshire (University of Limerick, Ireland) M. J. Hoffmann (University of Karlsruhe, Germany) I. -W. Chen (University of Pennsylvania, U.S.A.) S. Kanzaki (National Institute of Advanced Industrial Science and Technology, Japan) H. Mandal (Anadolu University, Turkey) M. Naito (Osaka University, Japan) S. Shimada (Hokkaido University, Japan) H. Tanaka (National Institute for Materials Science, Japan)
Programme Committee
Chairman: Y. -B. Cheng (Monash University, Australia) Members: Y. Ikuhara (University of Tokyo, Japan) S. Itatani (Sophia University, Japan) K. Hirao (National Institute of Advanced Industrial Science and Technology, Japan) N. Hirosaki (National Institute for Materials Science, Japan) T. Ohji (National Institute of Advanced Industrial Science and Technology, Japan) Y. Sugawara (Waseda University, Japan) J. Tatami (Yokohama National University, Japan) Y. Ukyo (Toyota Central R&D Labs., Inc., Japan) T. Yano (Tokyo Institute of Technology, Japan)
Local Committee
Chairman: J. Tatami (Yokohama National Univ., Japan) Members: T. Akashi (Hokkaido University, Japan) N. Hotta (Niigata University, Japan) H. Kita (National Institute of Advanced Industrial Science and Technology, Japan) T. Nishimura (National Institute for Materials Science, Japan) M. Ooyanagi (Ryukoku University, Japan) J. Takahashi (Tohoku University, Japan) H. Tanaka (National Institute for Materials Science, Japan) P. Xin (Covalent Materials, Japan) H. Wada (Bridgistone Co., Japan) K. Watari (National Institute of Advanced Industrial Science and Technology, Japan) T. Wakihara (Yokohama National University, Japan) N. Yamada (NGK Insulators Ltd., Japan)
Organising Committee
Chairman: K. Komeya (Yokohama National University, Japan) Vice-Chairman: D. P. Thompson (University of Newcastle-upon-Tyne, UK) M. Mitomo (National Institute for Materials Science, Japan) Members: Y. -B. Cheng (Monash University, Australia) T. Goto (Tohoku University, Japan) S. Hampshire (University of Limerick, Ireland) M. J. Hoffmann (University of Karlsruhe, Germany) I. -W. Chen (University of Pennsylvania, U.S.A.) S. Kanzaki (National Institute of Advanced Industrial Science and Technology, Japan) H. Mandal (Anadolu University, Turkey) M. Naito (Osaka University, Japan) S. Shimada (Hokkaido University, Japan) H. Tanaka (National Institute for Materials Science, Japan)
Programme Committee
Chairman: Y. -B. Cheng (Monash University, Australia) Members: Y. Ikuhara (University of Tokyo, Japan) S. Itatani (Sophia University, Japan) K. Hirao (National Institute of Advanced Industrial Science and Technology, Japan) N. Hirosaki (National Institute for Materials Science, Japan) T. Ohji (National Institute of Advanced Industrial Science and Technology, Japan) Y. Sugawara (Waseda University, Japan) J. Tatami (Yokohama National University, Japan) Y. Ukyo (Toyota Central R&D Labs., Inc., Japan) T. Yano (Tokyo Institute of Technology, Japan)
Local Committee
Chairman: J. Tatami (Yokohama National Univ., Japan) Members: T. Akashi (Hokkaido University, Japan) N. Hotta (Niigata University, Japan) H. Kita (National Institute of Advanced Industrial Science and Technology, Japan) T. Nishimura (National Institute for Materials Science, Japan) M. Ooyanagi (Ryukoku University, Japan) J. Takahashi (Tohoku University, Japan) H. Tanaka (National Institute for Materials Science, Japan) P. Xin (Covalent Materials, Japan) H. Wada (Bridgistone Co., Japan) K. Watari (National Institute of Advanced Industrial Science and Technology, Japan) T. Wakihara (Yokohama National University, Japan) N. Yamada (NGK Insulators Ltd., Japan)
FOREWORD This Volume contains the papers presented at the International Symposium on SiAlONS and Non-oxides, held at Ise-Shima, Mie, Japan on December 2-5, 2007. In these fifty years, significant progress has been achieved in SiAlONs and non-oxides. In particular, advances in research and development in the last thirty years have made SiAlONs and non-oxides as important engineering ceramics and new functional ceramics of today. Following the successful 1st International Symposium on SiAlONs in 2001, the 2nd International Symposium on SiAlONs and Non-oxides was held as a broader forum to discuss the most recent developments in research and applications of SiAlONs and non-oxides at Ise-Shima, Mie, Japan, on December 2-5, 2007. Papers presented in this volume are authored by a group of international leading experts in SiAlON and non-oxide materials and give an excellent indication of the current and future directions in the field. All of the papers have been peer-reviewed prior to publication. The Symposium was sponsored by the Japan Society for the Promotion of Science (JSPS), Yokohama National University, Japan, JSPS 124th Committee and Hosokawa Powder Technology Foundation, Japan, and was cooperated by the Ceramic Society of Japan. Supports by the members of the International Advisory Committee, Drs. A. Bellosi, G. Fantozzi, L. Gao, H. T. Hintzen, J. Hojo, M. Ibukiyama, Y. Ito, S. J. Kang, D. K. Kim, H. D. Kim, H. T. Lin, K. MacKenzie, Y. Matsuo, K. Niihara, R. Riedel, T. Rouxel, P. Sajgalik, H. Sakai, Z.-J. Shen, M. Singh, K. Uematsu, F. Wakai, A. Yamaguchi, G. Zhang, and by the members of the Local Organizing Committee and the Program Committee are gratefully acknowledged by the conference organizers. Finally, a special acknowledgement is due to the students and stuff of Yokohama National University and researchers of AIST, Nagoya, and NIMS, Tsukuba for their sustained assistance to the symposium. Katsutoshi Komeya, Yi-Bing Cheng, Junichi Tatami and Mamoru Mitomo Yokohama, Melbourne and Tsukuba, 2008
FOREWORD This Volume contains the papers presented at the International Symposium on SiAlONS and Non-oxides, held at Ise-Shima, Mie, Japan on December 2-5, 2007. In these fifty years, significant progress has been achieved in SiAlONs and non-oxides. In particular, advances in research and development in the last thirty years have made SiAlONs and non-oxides as important engineering ceramics and new functional ceramics of today. Following the successful 1st International Symposium on SiAlONs in 2001, the 2nd International Symposium on SiAlONs and Non-oxides was held as a broader forum to discuss the most recent developments in research and applications of SiAlONs and non-oxides at Ise-Shima, Mie, Japan, on December 2-5, 2007. Papers presented in this volume are authored by a group of international leading experts in SiAlON and non-oxide materials and give an excellent indication of the current and future directions in the field. All of the papers have been peer-reviewed prior to publication. The Symposium was sponsored by the Japan Society for the Promotion of Science (JSPS), Yokohama National University, Japan, JSPS 124th Committee and Hosokawa Powder Technology Foundation, Japan, and was cooperated by the Ceramic Society of Japan. Supports by the members of the International Advisory Committee, Drs. A. Bellosi, G. Fantozzi, L. Gao, H. T. Hintzen, J. Hojo, M. Ibukiyama, Y. Ito, S. J. Kang, D. K. Kim, H. D. Kim, H. T. Lin, K. MacKenzie, Y. Matsuo, K. Niihara, R. Riedel, T. Rouxel, P. Sajgalik, H. Sakai, Z.-J. Shen, M. Singh, K. Uematsu, F. Wakai, A. Yamaguchi, G. Zhang, and by the members of the Local Organizing Committee and the Program Committee are gratefully acknowledged by the conference organizers. Finally, a special acknowledgement is due to the students and stuff of Yokohama National University and researchers of AIST, Nagoya, and NIMS, Tsukuba for their sustained assistance to the symposium. Katsutoshi Komeya, Yi-Bing Cheng, Junichi Tatami and Mamoru Mitomo Yokohama, Melbourne and Tsukuba, 2008
Nitrides Nitrides
D.P. Thompsona and Yue Zhang
Advanced Materials Group, School of Chemical Engineering & Advanced Materials,
University of Newcastle, Newcastle upon Tyne NE1 7RU,UK
a d.p.thompson@newcastle.ac.uk
Abstract. The preparation of high temperature ceramics simultaneously containing silicon, nitrogen
and carbon has only relatively recently become an area of interest for inorganic crystal chemists,
and the recent discovery of a new series of carbonitrides with the general formula MM’Si4N6C is of
interest because of the good high temperature properties they appear to display. On the one hand, M
and M’ can be the same trivalent metal - either rare earth or yttrium; in this case, the resulting
compounds display orthorhombic (pseudo-hexagonal) structures. Alternatively the metals may be a
mix of di- (Ca,Sr, Ba) and tri-valent (Y,Ln) cations, in which case the carbon is replaced by
nitrogen, and the overall symmetry is hexagonal. Other quaternary nitrides of a similar type can be
produced if the two metal cations remain trivalent and one of the silicon atoms is replaced by
aluminium.
The present study describes the preparation of powder samples of Y2Si4N6C and LaYSi4N6C
starting from YH2, La, Si3N4 and carbon precursors, and summarises attempts to achieve a dense
product by hot-pressing at 1700-1800 o C. Some preliminary mechanical property measurements are
included.
Introduction
In the first symposium [1], it was argued that the exploration of new metal sialon compounds had
been very much neglected because of the impressive success of α- and β- Si3N4 and their sialon
equivalents as high-strength engineering materials. As far as α-sialon is concerned, there have been
an increasing range of nitrogen-rich oxynitrides reported in recent years which are likely to have
similar properties as judged by the facts that (a) they consist of a 3D linkage of [SiN4] tetrahedra,
(b) the proportion of large additional metal (M) cations is small relative to Si+Al, and (c) Si+N by
Al+O replacement can generally occur, thereby making preparation/ densification easier and in
some cases allowing different M cations to be incorporated. However, whereas most of these phases
are characterised, with unit cell (and sometimes full crystal structure) information available, very
few attempts have been made to explore ceramic properties. During the last 5 years, Lewis and
coworkers [2] carried out such a study on the sialon S-phase (BaSi5Al2N8O2), and showed that this
was relatively easy to prepare in pure form starting with mixtures of BaCO3, Al2O3, Si3N4 and AlN,
and could be readily densified, but careful TEM work showed that the morphology was needle-
shaped, and therefore unlikely to display high values of fracture toughness. Nevertheless, the
exercise was useful in showing that these compounds could be made into dense materials with
acceptable properties.
The present paper describes a similar exercise applied to the MM’Si4N6C group of ceramics
where M and M’ are typically Y or the rare earths. The first published work on these compounds
was reported by Höppe et al. [3] on the compound LaYbSi4N6C, who showed this to be iso-
structural with the group of compounds of the type ABSi4N7 (A = Ca,Sr, Ba; B = Y or rare earth),
being built up of structural units of the type [Si4N12C], consisting of four tetrahedra meeting at a
point, and stacked on top of the another (but with a 60 o rotation) in the crystallographic z direction.
In these units, a central non-metal atom is linked to four tetrahedra and is therefore an ideal site for
carbon to occupy in the carbonitride derivatives. Concurrent Newcastle work [4,5] showed that
Preparation and Characterization of MM’Si4N6C Ceramics
D.P. Thompsona and Yue Zhang
Advanced Materials Group, School of Chemical Engineering & Advanced Materials,
University of Newcastle, Newcastle upon Tyne NE1 7RU,UK
a d.p.thompson@newcastle.ac.uk
Abstract. The preparation of high temperature ceramics simultaneously containing silicon, nitrogen
and carbon has only relatively recently become an area of interest for inorganic crystal chemists,
and the recent discovery of a new series of carbonitrides with the general formula MM’Si4N6C is of
interest because of the good high temperature properties they appear to display. On the one hand, M
and M’ can be the same trivalent metal - either rare earth or yttrium; in this case, the resulting
compounds display orthorhombic (pseudo-hexagonal) structures. Alternatively the metals may be a
mix of di- (Ca,Sr, Ba) and tri-valent (Y,Ln) cations, in which case the carbon is replaced by
nitrogen, and the overall symmetry is hexagonal. Other quaternary nitrides of a similar type can be
produced if the two metal cations remain trivalent and one of the silicon atoms is replaced by
aluminium.
The present study describes the preparation of powder samples of Y2Si4N6C and LaYSi4N6C
starting from YH2, La, Si3N4 and carbon precursors, and summarises attempts to achieve a dense
product by hot-pressing at 1700-1800 o C. Some preliminary mechanical property measurements are
included.
Introduction
In the first symposium [1], it was argued that the exploration of new metal sialon compounds had
been very much neglected because of the impressive success of α- and β- Si3N4 and their sialon
equivalents as high-strength engineering materials. As far as α-sialon is concerned, there have been
an increasing range of nitrogen-rich oxynitrides reported in recent years which are likely to have
similar properties as judged by the facts that (a) they consist of a 3D linkage of [SiN4] tetrahedra,
(b) the proportion of large additional metal (M) cations is small relative to Si+Al, and (c) Si+N by
Al+O replacement can generally occur, thereby making preparation/ densification easier and in
some cases allowing different M cations to be incorporated. However, whereas most of these phases
are characterised, with unit cell (and sometimes full crystal structure) information available, very
few attempts have been made to explore ceramic properties. During the last 5 years, Lewis and
coworkers [2] carried out such a study on the sialon S-phase (BaSi5Al2N8O2), and showed that this
was relatively easy to prepare in pure form starting with mixtures of BaCO3, Al2O3, Si3N4 and AlN,
and could be readily densified, but careful TEM work showed that the morphology was needle-
shaped, and therefore unlikely to display high values of fracture toughness. Nevertheless, the
exercise was useful in showing that these compounds could be made into dense materials with
acceptable properties.
The present paper describes a similar exercise applied to the MM’Si4N6C group of ceramics
where M and M’ are typically Y or the rare earths. The first published work on these compounds
was reported by Höppe et al. [3] on the compound LaYbSi4N6C, who showed this to be iso-
structural with the group of compounds of the type ABSi4N7 (A = Ca,Sr, Ba; B = Y or rare earth),
being built up of structural units of the type [Si4N12C], consisting of four tetrahedra meeting at a
point, and stacked on top of the another (but with a 60 o rotation) in the crystallographic z direction.
In these units, a central non-metal atom is linked to four tetrahedra and is therefore an ideal site for
carbon to occupy in the carbonitride derivatives. Concurrent Newcastle work [4,5] showed that
Key Engineering Materials Vol. 403 (2009) pp 3-6 © (2009) Trans Tech Publications, Switzerland doi:10.4028/www.scientific.net/KEM.403.3
when M and M’ were different cations in the carbonitride structures, the resulting compounds were
hexagonal, whereas when they were the same, this resulted in a slight distortion to orthorhombic
(pseudo-hexagonal), the distortion being caused by alternate columns of [Si4N12C] being anti-
parallel, in contrast to perfect parallel alignment when the two cations were different.
Experimental
Preparation of oxygen-free nitrides and carbonitrides is difficult, because during initial processing it
is almost impossible to avoid oxygen pick-up especially during addition of the ionic metals used
jointly as sintering aids and as a source of the third metal cation in the formulation (as for example
in α-sialons). Recently Esmaeilzadeh and colleagues at University of Stockholm [6] have shown
that providing the third metal as either a hydride or as the metal powders is an excellent way of
minimising oxygen pick-up. In the present study, it was decided to compare one orthorhombic and
one hexagonal carbonitride; Y2Si4N6C was selected as the orthorhombic derivative and LaYSi4N6C
as the hexagonal variant. YH2 (Aldrich) was used as the source of yttrium, and lanthanum powder
(supplied immersed in oil - Aldrich) as the source of lanthanum. Previous work had shown that
when firing samples in a carbon resistance furnace in a carbon crucible, there was more than
sufficient carbon pick-up to provide the carbon required by the formula; traces of oil present in the
La starting powder probably evaporated from the sample rather than being pyrolysed to carbon.
Si3N4 powder was used as the source of silicon. The starting powders were mixed in isopropanol,
and in the first experiments were compacted and fired in a carbon resistance furnace at 1750 o C in a
nitrogen atmosphere. After only a few experiments it was possible to produce >95% pure samples
of both the Y2 and LaY compounds (see Figure 1). The main impurity was J-phase (Y4Si2O7N2),
stabilised by oxygen introduced either during processing or from impurity in the nitriding gas.
Attempts were then made to prepare dense samples by hot-pressing in graphite dies, and these were
prepared in two ways. In the first case, the already pre-prepared carbonitride powders were finely
ground, and hot-pressed initially without additive at 1750 o C, and then with extra J-phase (to provide
additional liquid phase) at 1800 o C. However, it was found (Figure 2(a)), that this resulted merely in
squeezing the grains together, rather than modifying grain morphology and eliminating pores.
Alternatively, the original powder mixes were used, with the assumption that the oxygen in the air
pockets between the starting powder grains would be removed by the carbon of the die at low
temperatures, leaving the residual nitrogen to be available to nitride the YH2 and La. In fact this
assumption was successful, and the final products were substantially the required phases, even
though in every case there was a slight increase in the percentage of J-phase. However, even
though this second method resulted in slightly better densification, it was still impossible to achieve
full density, and the maximum values achieved were 93% for Y2Si4N6C, and 91% for LaYSi4N6C.
Figure 2(b) shows a sample hot-pressed from the original starting powders, showing regions of J-
phase and even some silicon carbide, in addition to the predominant LaYSi4N6C phase. Regions of
black porosity are also clearly visible. Further work is needed to identify ways of achieving 100%
densification in both the Y2 and the LaY samples
Preliminary measurements of hardness and fracture toughness made on the best hot-pressed
Y2Si4N6C sample yielded values of 14.9GPa and 3.6 MPam ½ respectively, and on similar hot-
pressed samples of LaYSi4N6C of 15GPa and 4.0 MPam ½ . These are typical values for a nitrogen
ceramic, and even though promising, should not be given too much credibility, because the samples
have not yet fully densified, and there is not significant morphological development of product
grains. Further work on microstructure is needed, which in turn will emerge in response to better
densification procedures being developed.
when M and M’ were different cations in the carbonitride structures, the resulting compounds were
hexagonal, whereas when they were the same, this resulted in a slight distortion to orthorhombic
(pseudo-hexagonal), the distortion being caused by alternate columns of [Si4N12C] being anti-
parallel, in contrast to perfect parallel alignment when the two cations were different.
Experimental
Preparation of oxygen-free nitrides and carbonitrides is difficult, because during initial processing it
is almost impossible to avoid oxygen pick-up especially during addition of the ionic metals used
jointly as sintering aids and as a source of the third metal cation in the formulation (as for example
in α-sialons). Recently Esmaeilzadeh and colleagues at University of Stockholm [6] have shown
that providing the third metal as either a hydride or as the metal powders is an excellent way of
minimising oxygen pick-up. In the present study, it was decided to compare one orthorhombic and
one hexagonal carbonitride; Y2Si4N6C was selected as the orthorhombic derivative and LaYSi4N6C
as the hexagonal variant. YH2 (Aldrich) was used as the source of yttrium, and lanthanum powder
(supplied immersed in oil - Aldrich) as the source of lanthanum. Previous work had shown that
when firing samples in a carbon resistance furnace in a carbon crucible, there was more than
sufficient carbon pick-up to provide the carbon required by the formula; traces of oil present in the
La starting powder probably evaporated from the sample rather than being pyrolysed to carbon.
Si3N4 powder was used as the source of silicon. The starting powders were mixed in isopropanol,
and in the first experiments were compacted and fired in a carbon resistance furnace at 1750 o C in a
nitrogen atmosphere. After only a few experiments it was possible to produce >95% pure samples
of both the Y2 and LaY compounds (see Figure 1). The main impurity was J-phase (Y4Si2O7N2),
stabilised by oxygen introduced either during processing or from impurity in the nitriding gas.
Attempts were then made to prepare dense samples by hot-pressing in graphite dies, and these were
prepared in two ways. In the first case, the already pre-prepared carbonitride powders were finely
ground, and hot-pressed initially without additive at 1750 o C, and then with extra J-phase (to provide
additional liquid phase) at 1800 o C. However, it was found (Figure 2(a)), that this resulted merely in
squeezing the grains together, rather than modifying grain morphology and eliminating pores.
Alternatively, the original powder mixes were used, with the assumption that the oxygen in the air
pockets between the starting powder grains would be removed by the carbon of the die at low
temperatures, leaving the residual nitrogen to be available to nitride the YH2 and La. In fact this
assumption was successful, and the final products were substantially the required phases, even
though in every case there was a slight increase in the percentage of J-phase. However, even
though this second method resulted in slightly better densification, it was still impossible to achieve
full density, and the maximum values achieved were 93% for Y2Si4N6C, and 91% for LaYSi4N6C.
Figure 2(b) shows a sample hot-pressed from the original starting powders, showing regions of J-
phase and even some silicon carbide, in addition to the predominant LaYSi4N6C phase. Regions of
black porosity are also clearly visible. Further work is needed to identify ways of achieving 100%
densification in both the Y2 and the LaY samples
Preliminary measurements of hardness and fracture toughness made on the best hot-pressed
Y2Si4N6C sample yielded values of 14.9GPa and 3.6 MPam ½ respectively, and on similar hot-
pressed samples of LaYSi4N6C of 15GPa and 4.0 MPam ½ . These are typical values for a nitrogen
ceramic, and even though promising, should not be given too much credibility, because the samples
have not yet fully densified, and there is not significant morphological development of product
grains. Further work on microstructure is needed, which in turn will emerge in response to better
densification procedures being developed.
4 SiAlONs and Non-oxides
(a)
(b)
Figure 1. XRD patterns of (a) Y2Si4N6C and (b) LaYSi4N6C prepared by nitriding YH2/Si3N4 and
La/YH2/Si3N4 mixtures respectively in a carbon furnace at 1750 o C.
(a)
(b)
Figure 1. XRD patterns of (a) Y2Si4N6C and (b) LaYSi4N6C prepared by nitriding YH2/Si3N4 and
La/YH2/Si3N4 mixtures respectively in a carbon furnace at 1750 o C.
Key Engineering Materials Vol. 403 5
(a) (b)
Figure 2. Microstructures of hot-pressed (a) pre-prepared LaYSi4N6C, and (b) powder mixes of
La/YH2/Si3N4 to produce LaYSi4N6C, both at 1750 o C.
Conclusions
Powder samples of MM’Si4N6C ceramics (M = yttrium; M’ = yttrium or lanthanum) were relatively
easily prepared by reacting YH2/Si3N4 and La/YH2/Si3N4 mixes in nitrogen in a carbon element
furnace at 1750 o C. The main impurity was J-phase (Y4Si2O7N2), present because of the small levels
of oxygen impurity incorporated during powder processing. Attempts to densify these materials
using either the pre-prepared powders, or the original powder starting mix were not fully successful,
probably because the solubility of carbon in the nitride/oxynitride liquids present at high-
temperature in these systems is very low, and even though some rearrangement no doubt occurs
assisted by the applied pressure, there is almost no densification from solution/reprecipitation.
Further work is continuing on related pure nitrides of the type MM’Si4N7, where M is divalent
(Ca,Sr,Ba) and M’ is Y or La, to establish whether these materials densify by traditional liquid
phase sintering methods.
References [1] K. Liddell and D.P. Thompson, Key. Engineering Materials, 237, (2003), 1-10.
[2] M.H. Lewis, B. Basu, M.E. Smith, M. Bunyard and T. Kemp, Silicates Industrielles, Special
Issue, 69(7-8), (2004), 225-32.
[3] H.A. Höppe, G. Kotzyba, R. Pöttgen and W. Schnick, J. Mater. Chem., 11, (2001), 3300-06.
[4] K. Liddell, D.P. Thompson, T. Bräuniger and R.K. Harris, J. Eur. Ceram. Soc., 25, (2005), 37-
47.
[5] K.Liddell, D.P. Thompson and S.J. Teat, J. Eur. Ceram. Soc., 25, (2005), 49-54.
[6] A.S. Hakeem, J. Grins and S. Esmaeilzadeh, J. Eur. Ceram. Oc., 27, (2007), 4783-87.
(a) (b)
Figure 2. Microstructures of hot-pressed (a) pre-prepared LaYSi4N6C, and (b) powder mixes of
La/YH2/Si3N4 to produce LaYSi4N6C, both at 1750 o C.
Conclusions
Powder samples of MM’Si4N6C ceramics (M = yttrium; M’ = yttrium or lanthanum) were relatively
easily prepared by reacting YH2/Si3N4 and La/YH2/Si3N4 mixes in nitrogen in a carbon element
furnace at 1750 o C. The main impurity was J-phase (Y4Si2O7N2), present because of the small levels
of oxygen impurity incorporated during powder processing. Attempts to densify these materials
using either the pre-prepared powders, or the original powder starting mix were not fully successful,
probably because the solubility of carbon in the nitride/oxynitride liquids present at high-
temperature in these systems is very low, and even though some rearrangement no doubt occurs
assisted by the applied pressure, there is almost no densification from solution/reprecipitation.
Further work is continuing on related pure nitrides of the type MM’Si4N7, where M is divalent
(Ca,Sr,Ba) and M’ is Y or La, to establish whether these materials densify by traditional liquid
phase sintering methods.
References [1] K. Liddell and D.P. Thompson, Key. Engineering Materials, 237, (2003), 1-10.
[2] M.H. Lewis, B. Basu, M.E. Smith, M. Bunyard and T. Kemp, Silicates Industrielles, Special
Issue, 69(7-8), (2004), 225-32.
[3] H.A. Höppe, G. Kotzyba, R. Pöttgen and W. Schnick, J. Mater. Chem., 11, (2001), 3300-06.
[4] K. Liddell, D.P. Thompson, T. Bräuniger and R.K. Harris, J. Eur. Ceram. Soc., 25, (2005), 37-
47.
[5] K.Liddell, D.P. Thompson and S.J. Teat, J. Eur. Ceram. Soc., 25, (2005), 49-54.
[6] A.S. Hakeem, J. Grins and S. Esmaeilzadeh, J. Eur. Ceram. Oc., 27, (2007), 4783-87.
6 SiAlONs and Non-oxides
The phase evolution in the Si3N4-AlN system after high-energy mechanical treatment of the precursor powder
Magorzata Sopicka-Lizer, Tomasz Pawlik, Tomasz Wodek, Marta Tacula Silesian University of Technology, 40-019 Katowice, Krasiskiego 8, Poland
amalgorzata.sopicka-lizer@polsl.pl, btomasz.pawlik@polsl.pl, ctomasz.wlodek@polsl.pl, dmarta.tancula@polsl.pl
Keywords: β-sialon, mechanical activation, precursor, nanostructured powder Abstract. The high-energy milling uses the mechanical energy to activate chemical reactions by developing structural changes in the powder particles. High-energy milling with an acceleration of 28g was applied for the mechanical activation of the aluminium and silicon nitrides mixture with yttria additive. The activated powders showed the significant damage of the crystal structure and limited formation of a solid solution. Sintering of the activated precursor demonstrated higher ability for densification and started at 300 ºC lower temperature in comparison to the standard mixture. The phase evolution during sintering was dependent on the starting composition and degree of powder activation.
Introduction
High-energy milling of β-SiAlON precursor has been reported lately as a new method of the nitride-based powder preparation for subsequent densification at lower temperature and/or nanoceramic manufacturing [1- 4]. The nanostructured precursor powder offers new possibilities in tailoring the microstructure and properties of silicon nitride ceramics and pressureless sintering techniques could be applied for manufacturing of the fully dense silicon nitride ceramics. However, little is known about the phase transformation and densification in the Si-Al-O-N system with the highly defective crystal structure. The previous studies on densification of nanosized β-Si3N4 showed dependence of α→β transformation and the features of the resultant microstructure on the amount and size of the β-nuclei [5]. The aim of this work is to compare the low-temperature behavior and the phase evolution of the Si3N4-AlN mixture after high-energy activation in a planetary mill.
Materials and methods
The initial powders were α-Si3N4 (H.C.Starck-B7), 10-15 wt% of β-Si3N4), β-Si3N4 (Aldrich, -325 mesh, 5 wt% of Si2N2O contamination), AlN (H.C.Starck-C), Y2O3 (H.C.Starck-C). The composition of the batches was chosen to be close to z=0.4 in the β-SiAlON solid solution: 89,3 wt% of Si3N4, 5.7 wt% of AlN and 5 wt% of Y2O3. Alumina was eliminated from the precursor mixture because of the expected oxidation during milling. There were prepared two batches for activation milling: with α-Si3N4 (A) and β-Si3N4 (B) as a source of silicon nitride. Both batches were activated in a MPP-1 (TTD, Russia) planetary mill for 30 min or 60 min with silicon nitride balls. The ball-to-powder ratio was 6:1 and the acceleration of the centrifugal field attained the value of 28g (g is the gravitational acceleration). For the comparison reason the reference A-RB and B-RB batches were prepared by mixing the components with isopropanol on the roller bench for 24 hours. The powders after activation or mixing were dispersed in an aezotropic MEK+EtOH solvent with hypermer KD1 dispersant.
Subsequently the suspensions were dried at room temperature, the powders were cold uni-axially pressed into form of tablets and isostatically pressed with 250 MPa pressure. The resultant tablets were placed in a BN crucible filled with Si3N4/BN powder bed. Sintering was performed in a graphite
The phase evolution in the Si3N4-AlN system after high-energy mechanical treatment of the precursor powder
Magorzata Sopicka-Lizer, Tomasz Pawlik, Tomasz Wodek, Marta Tacula Silesian University of Technology, 40-019 Katowice, Krasiskiego 8, Poland
amalgorzata.sopicka-lizer@polsl.pl, btomasz.pawlik@polsl.pl, ctomasz.wlodek@polsl.pl, dmarta.tancula@polsl.pl
Keywords: β-sialon, mechanical activation, precursor, nanostructured powder Abstract. The high-energy milling uses the mechanical energy to activate chemical reactions by developing structural changes in the powder particles. High-energy milling with an acceleration of 28g was applied for the mechanical activation of the aluminium and silicon nitrides mixture with yttria additive. The activated powders showed the significant damage of the crystal structure and limited formation of a solid solution. Sintering of the activated precursor demonstrated higher ability for densification and started at 300 ºC lower temperature in comparison to the standard mixture. The phase evolution during sintering was dependent on the starting composition and degree of powder activation.
Introduction
High-energy milling of β-SiAlON precursor has been reported lately as a new method of the nitride-based powder preparation for subsequent densification at lower temperature and/or nanoceramic manufacturing [1- 4]. The nanostructured precursor powder offers new possibilities in tailoring the microstructure and properties of silicon nitride ceramics and pressureless sintering techniques could be applied for manufacturing of the fully dense silicon nitride ceramics. However, little is known about the phase transformation and densification in the Si-Al-O-N system with the highly defective crystal structure. The previous studies on densification of nanosized β-Si3N4 showed dependence of α→β transformation and the features of the resultant microstructure on the amount and size of the β-nuclei [5]. The aim of this work is to compare the low-temperature behavior and the phase evolution of the Si3N4-AlN mixture after high-energy activation in a planetary mill.
Materials and methods
The initial powders were α-Si3N4 (H.C.Starck-B7), 10-15 wt% of β-Si3N4), β-Si3N4 (Aldrich, -325 mesh, 5 wt% of Si2N2O contamination), AlN (H.C.Starck-C), Y2O3 (H.C.Starck-C). The composition of the batches was chosen to be close to z=0.4 in the β-SiAlON solid solution: 89,3 wt% of Si3N4, 5.7 wt% of AlN and 5 wt% of Y2O3. Alumina was eliminated from the precursor mixture because of the expected oxidation during milling. There were prepared two batches for activation milling: with α-Si3N4 (A) and β-Si3N4 (B) as a source of silicon nitride. Both batches were activated in a MPP-1 (TTD, Russia) planetary mill for 30 min or 60 min with silicon nitride balls. The ball-to-powder ratio was 6:1 and the acceleration of the centrifugal field attained the value of 28g (g is the gravitational acceleration). For the comparison reason the reference A-RB and B-RB batches were prepared by mixing the components with isopropanol on the roller bench for 24 hours. The powders after activation or mixing were dispersed in an aezotropic MEK+EtOH solvent with hypermer KD1 dispersant.
Subsequently the suspensions were dried at room temperature, the powders were cold uni-axially pressed into form of tablets and isostatically pressed with 250 MPa pressure. The resultant tablets were placed in a BN crucible filled with Si3N4/BN powder bed. Sintering was performed in a graphite
Key Engineering Materials Vol. 403 (2009) pp 7-10 © (2009) Trans Tech Publications, Switzerland doi:10.4028/www.scientific.net/KEM.403.7
furnace (Thermal Technology) in a nitrogen flow of 3l/h in the temperature range of 1200-1600 ºC for 60 min. One tablet of the A-30 batch was hot-pressed with 250 bar pressure at 1500 ºC for 60 min. The activated powders were measured for their specific surface area by BET method (ASAP 2010) and observed in scanning electron microscope (Hitachi S-4200) as well as transmission microscope (TEM). The oxygen level was controlled (ELTRA ON) and XRD studies (X’PERT) with Rietveld refinement were applied for measurement of the phase composition after activation.
The density of sintered tablets was determined by the Archimedes method. XRD studies were performed on the polished cross section of the tablets. The phase composition was calculated and lattice constants for silicon nitride phases were measured.
Results and discussion
The specific surface area of the initial mixtures was 5.5 m2g-1. The activation milling for 30 min increased the specific surface area (SSA) of the precursor to 18,7 m2g-1. Prolongation of the activation milling to 60 min enlarged SSA to 21,9 m2g-1. 2.4 wt% of oxygen was measured in the non-activated A-type mixture while activation milling for 30 min led to a slight increase of oxygen content (3,9 wt%). Estimation of the mean particle size from BET measurement showed the value of 80 nm. However, the microscopic studies revealed rather nanostructured particles than the nanosized ones (Fig.1). On the other hand, XRD results demonstrate substantial changes of the phase composition after mechanical treatment: decrease or decay of yttria and aluminum nitride; thus increase of silicon nitride content after milling is apparent. Decline of Y2O3 and AlN component in the activated batches is apparent as well since they are not identified by XRD. It is assumed that their crystal lattice has been destroyed to such extent they were not able to produce the well defined diffraction lines. The several crystal defects (dislocations, plastic deformation) were observed under TEM studies.
Thermal treatment of the specimens showed a different behavior of the activated batches derived samples in comparison to the standard precursor: full density (98,7%) was obtained for the sintered tablets if the precursor was activated for 60 min. The similar results (relative density=97.2 %) were obtained for the A-type batch but activated for 30 min. Increase of the density was observed after sintering the activated precursor at 1200 ºC whereas densification start of the reference tablets from non-activated powders occurred at 1500 ºC (Fig.2). Densification was accompanied by the changes of the phase composition of the thermal treated samples. The sequence of the phase evolution was dependent on the starting composition and time of the precursor
Figure 1. The powder of the B-type precursor after activation with 28g for 30 min, a-SEM, the microstructure of the green tablet; b-TEM
50
60
70
80
90
100
Sintering temperature [oC]
activated
RB
Figure 2. The effect of sintering temperature on a relative density of the B-type mixture. Homogenized (RB) and activated for 60 min. precursors are compared
furnace (Thermal Technology) in a nitrogen flow of 3l/h in the temperature range of 1200-1600 ºC for 60 min. One tablet of the A-30 batch was hot-pressed with 250 bar pressure at 1500 ºC for 60 min. The activated powders were measured for their specific surface area by BET method (ASAP 2010) and observed in scanning electron microscope (Hitachi S-4200) as well as transmission microscope (TEM). The oxygen level was controlled (ELTRA ON) and XRD studies (X’PERT) with Rietveld refinement were applied for measurement of the phase composition after activation.
The density of sintered tablets was determined by the Archimedes method. XRD studies were performed on the polished cross section of the tablets. The phase composition was calculated and lattice constants for silicon nitride phases were measured.
Results and discussion
The specific surface area of the initial mixtures was 5.5 m2g-1. The activation milling for 30 min increased the specific surface area (SSA) of the precursor to 18,7 m2g-1. Prolongation of the activation milling to 60 min enlarged SSA to 21,9 m2g-1. 2.4 wt% of oxygen was measured in the non-activated A-type mixture while activation milling for 30 min led to a slight increase of oxygen content (3,9 wt%). Estimation of the mean particle size from BET measurement showed the value of 80 nm. However, the microscopic studies revealed rather nanostructured particles than the nanosized ones (Fig.1). On the other hand, XRD results demonstrate substantial changes of the phase composition after mechanical treatment: decrease or decay of yttria and aluminum nitride; thus increase of silicon nitride content after milling is apparent. Decline of Y2O3 and AlN component in the activated batches is apparent as well since they are not identified by XRD. It is assumed that their crystal lattice has been destroyed to such extent they were not able to produce the well defined diffraction lines. The several crystal defects (dislocations, plastic deformation) were observed under TEM studies.
Thermal treatment of the specimens showed a different behavior of the activated batches derived samples in comparison to the standard precursor: full density (98,7%) was obtained for the sintered tablets if the precursor was activated for 60 min. The similar results (relative density=97.2 %) were obtained for the A-type batch but activated for 30 min. Increase of the density was observed after sintering the activated precursor at 1200 ºC whereas densification start of the reference tablets from non-activated powders occurred at 1500 ºC (Fig.2). Densification was accompanied by the changes of the phase composition of the thermal treated samples. The sequence of the phase evolution was dependent on the starting composition and time of the precursor
Figure 1. The powder of the B-type precursor after activation with 28g for 30 min, a-SEM, the microstructure of the green tablet; b-TEM
50
60
70
80
90
100
Sintering temperature [oC]
activated
RB
Figure 2. The effect of sintering temperature on a relative density of the B-type mixture. Homogenized (RB) and activated for 60 min. precursors are compared
8 SiAlONs and Non-oxides
activation but was independent on the technique of sintering (Table 2). α→β-Si3N4 transformation is believed to occur in the presence of the liquid phase. If there is a sufficient amount of the Y-Si-Al-O-N eutectic liquid phase then β-phase crystallization/precipitation follows. The amount of the eutectic liquid phase depends on temperature if the same chemical composition is considered, but could also be affected by amorphization of one or more components. The amount and composition of the liquid phase at 1500 ºC in the non-activated precursor must have been insufficient for β-Si3N4 crystallization because only less than 1/4 of the initial silicon nitride was transformed to β-phase and YAG presence shows silicon lacking in the liquid at 1500 ºC. On the other hand, significantly more liquid phase was present at 1500 ºC in the activated precursor: the amount of β-phase was two times higher in comparison to the non-activated precursor. Application of the hot-pressing technique can not change the amount of the liquid phase but can accelerate densification: thus the activated specimens were fully dense (3,20 gcm-3) after hot pressing at 1500 ºC but degree of α→β-Si3N4 transformation was similar to that from the powder bed sintering method (Table 2).
The phase evolution of β-Si3N4-based precursor after heat treatment in powder bed was different from α-Si3N4-based ones, because the initial β-Si3N4 powder was contaminated by Si2N2O and their susceptibility for activation was different. Moreover, increase of Si2N2O was observed after heat treatment of the activated precursor and/or mullite formation. The unexpected deviation and mullite or Si2N2O presence must be due to the oxidation or crystallization of the highly defective Si-Al-O-N phase. The final amount of β-Si3N4 in the resultant ceramic was comparable after heat treatment at 1600ºC despite the starting composition. However, β-SiAlON formation was found if measured by changes of the β-Si3N4 unit cell parameters (Fig. 3). That
2.906
2.908
2.91
2.912
2.914
Temperature [oC]
RB
Activated
Figure 3. Development of c-parameter in β-Si3N4 unit cell vs sintering temperature Table 1 The phase composition of the initial A and B mixtures as batched and measured after activation milling
Phase composition of β-sialon precursor [wt%] Precursor α-Si3N4 β-Si3N4 AlN Y2O3 Si2N2O A-initial 73 16 6 5 0
A-30 min 81,6 14,4 3,2 0,8 0 B-initial 0 84 6 5 5
B-60 min 0 96,4 0 0 3,6 Table 2 Phase assemblage [wt%] of β-sialon precursor after sintering at 1500 ºC and 1600 ºC in powder bed. The HP denoted specimen was hot pressed.
Temperature of sintering:1500oC α-Si3N4 β-Si3N4 Si2N2O YAG mullite A-RB 64,2 30,4 0 5,4 0 A-30 min 35,9 64,1 0 0 0 A-30 min, HP 38,2 61,8 0 0 0 B-RB 1,3 89,7 0 9 0 B-30 min 1 94,1 1,1 0,3 3,4 B-60 min 0 91,9 8,1 0 0
Temperature of sintering: 1600oC α-Si3N4 β-Si3N4 Si2N2O YAG mullite A-30 min 11,4 88,9 0 0 0 B-RB 0 92,4 0 7,6 0 B-30 min 0 96,5 0 0 3,5 B-60 min 0 90,8 9,2 0 0
activation but was independent on the technique of sintering (Table 2). α→β-Si3N4 transformation is believed to occur in the presence of the liquid phase. If there is a sufficient amount of the Y-Si-Al-O-N eutectic liquid phase then β-phase crystallization/precipitation follows. The amount of the eutectic liquid phase depends on temperature if the same chemical composition is considered, but could also be affected by amorphization of one or more components. The amount and composition of the liquid phase at 1500 ºC in the non-activated precursor must have been insufficient for β-Si3N4 crystallization because only less than 1/4 of the initial silicon nitride was transformed to β-phase and YAG presence shows silicon lacking in the liquid at 1500 ºC. On the other hand, significantly more liquid phase was present at 1500 ºC in the activated precursor: the amount of β-phase was two times higher in comparison to the non-activated precursor. Application of the hot-pressing technique can not change the amount of the liquid phase but can accelerate densification: thus the activated specimens were fully dense (3,20 gcm-3) after hot pressing at 1500 ºC but degree of α→β-Si3N4 transformation was similar to that from the powder bed sintering method (Table 2).
The phase evolution of β-Si3N4-based precursor after heat treatment in powder bed was different from α-Si3N4-based ones, because the initial β-Si3N4 powder was contaminated by Si2N2O and their susceptibility for activation was different. Moreover, increase of Si2N2O was observed after heat treatment of the activated precursor and/or mullite formation. The unexpected deviation and mullite or Si2N2O presence must be due to the oxidation or crystallization of the highly defective Si-Al-O-N phase. The final amount of β-Si3N4 in the resultant ceramic was comparable after heat treatment at 1600ºC despite the starting composition. However, β-SiAlON formation was found if measured by changes of the β-Si3N4 unit cell parameters (Fig. 3). That
2.906
2.908
2.91
2.912
2.914
Temperature [oC]
RB
Activated
Figure 3. Development of c-parameter in β-Si3N4 unit cell vs sintering temperature Table 1 The phase composition of the initial A and B mixtures as batched and measured after activation milling
Phase composition of β-sialon precursor [wt%] Precursor α-Si3N4 β-Si3N4 AlN Y2O3 Si2N2O A-initial 73 16 6 5 0
A-30 min 81,6 14,4 3,2 0,8 0 B-initial 0 84 6 5 5
B-60 min 0 96,4 0 0 3,6 Table 2 Phase assemblage [wt%] of β-sialon precursor after sintering at 1500 ºC and 1600 ºC in powder bed. The HP denoted specimen was hot pressed.
Temperature of sintering:1500oC α-Si3N4 β-Si3N4 Si2N2O YAG mullite A-RB 64,2 30,4 0 5,4 0 A-30 min 35,9 64,1 0 0 0 A-30 min, HP 38,2 61,8 0 0 0 B-RB 1,3 89,7 0 9 0 B-30 min 1 94,1 1,1 0,3 3,4 B-60 min 0 91,9 8,1 0 0
Temperature of sintering: 1600oC α-Si3N4 β-Si3N4 Si2N2O YAG mullite A-30 min 11,4 88,9 0 0 0 B-RB 0 92,4 0 7,6 0 B-30 min 0 96,5 0 0 3,5 B-60 min 0 90,8 9,2 0 0
Key Engineering Materials Vol. 403 9
behavior is in contrary to the non-activated batches as any changes of the β-Si3N4 unit cell parameters were observed after heat treatment at the tested temperature range. It is interesting to note that formation of β-SiAlON solid solution and degree of substitution in the specimens from the activated powders was closely related to temperature of the heat treatment and grew smoothly. The final z-value was close to 0.3 which is slightly lower than the designed 0.4 value.
Summary
Nanustructured silicon and aluminum nitrides mixture can be successfully produced by high-energy planetary milling. XRD studies showed substantial changes of activated precursor’s diffraction picture. The resultant activated powder could be sintered to full density at 1600 ºC without external nitrogen (gas) pressure. The densification and phase transformation occurred via transient liquid phase and was related to the extend of the precursor amorphization during the activation milling. Silicon oxynitride contamination should be avoided in the initial mixture.
References
[1] X. Xu, T. Nishimura, N. Hirosaki, R-J. Xie, Y. Yamamoto, H. Tanaka : Nanotechnology, Vol. 16 (2005), p. 1569
[2] X. Xu, T. Nishimura, T. Hirosaki, R.-J. Xie, Y. Yamamoto, H. Tanaka, J. Am. Ceram. Soc.,Vol. 88 (2005), p. 934
[3] M. Sopicka-Lizer, M. Tacula, T. Wodek, K. Rodak, M. Hüller, V. Kochnev, E. Fokina, K. MacKenzie, J. Eur. Ceram. Soc., Vol. 28 (2008), p. 279
[4] M. Sopicka-Lizer, M. Tacula, T. Pawlik, V. Kochnev, E. Fokina, Mat. Sci. Forum, Vol. 554 (2007), p. 59
[5] M. Herrmann, I. Schulz, I. Zalite, J. Eur. Ceram. Soc., Vol. 24 (2004), p. 3327
behavior is in contrary to the non-activated batches as any changes of the β-Si3N4 unit cell parameters were observed after heat treatment at the tested temperature range. It is interesting to note that formation of β-SiAlON solid solution and degree of substitution in the specimens from the activated powders was closely related to temperature of the heat treatment and grew smoothly. The final z-value was close to 0.3 which is slightly lower than the designed 0.4 value.
Summary
Nanustructured silicon and aluminum nitrides mixture can be successfully produced by high-energy planetary milling. XRD studies showed substantial changes of activated precursor’s diffraction picture. The resultant activated powder could be sintered to full density at 1600 ºC without external nitrogen (gas) pressure. The densification and phase transformation occurred via transient liquid phase and was related to the extend of the precursor amorphization during the activation milling. Silicon oxynitride contamination should be avoided in the initial mixture.
References
[1] X. Xu, T. Nishimura, N. Hirosaki, R-J. Xie, Y. Yamamoto, H. Tanaka : Nanotechnology, Vol. 16 (2005), p. 1569
[2] X. Xu, T. Nishimura, T. Hirosaki, R.-J. Xie, Y. Yamamoto, H. Tanaka, J. Am. Ceram. Soc.,Vol. 88 (2005), p. 934
[3] M. Sopicka-Lizer, M. Tacula, T. Wodek, K. Rodak, M. Hüller, V. Kochnev, E. Fokina, K. MacKenzie, J. Eur. Ceram. Soc., Vol. 28 (2008), p. 279
[4] M. Sopicka-Lizer, M. Tacula, T. Pawlik, V. Kochnev, E. Fokina, Mat. Sci. Forum, Vol. 554 (2007), p. 59
[5] M. Herrmann, I. Schulz, I. Zalite, J. Eur. Ceram. Soc., Vol. 24 (2004), p. 3327
10 SiAlONs and Non-oxides
New Green Phosphor Ba3Si6O12N2:Eu for White LED: Crystal Structure and Optical Properties
Masayoshi Mikami1,a, Satoshi Shimooka1, Kyota Uheda1, Hiroyuki Imura1 and Naoto Kijima1
1 Mitsubishi Chemical Group Science and Technology Research Center, Inc. 1000, Kamoshida-cho, Aoba-ku, Yokohama, 227-8502, Japan
amikami.masayoshi@mv.m-kagaku.co.jp
Abstract. A new oxynitride, Ba3Si6O12N2, has been synthesized. The crystal structure has been successfully determined by close collaboration between experiment and first-principles calculation. This compound doped with Eu exhibits intense green photoluminescence with high color purity under near-ultraviolet to blue light excitation; in particular, it has much less thermal quenching than (Ba,Sr,Eu)2SiO4. Thus (Ba,Eu)3Si6O12N2 appears promising green phosphor for white LED backlight for display. The atomic/electronic structure is discussed in comparison with Ba3Si6O9N4, which could not become efficient phosphor by doping Eu due to strong thermal quenching at room temperature.
Introduction
In the white LED market, the share of green and red phosphors has been gradually increasing to achieve good color reproducibility for liquid crystal display (LCD) backlight. Promising candidates that satisfy high color purity are, e.g., (M,Eu)2SiO4[1] and β -sialon:Eu[2] for green phosphor, and (M,Eu)AlSiN3[3] for red phosphor (M: alkaline-earth element). The green phosphors still confront some difficulties; (M,Eu)2SiO4 has strong thermal quenching[1], whereas the synthesis of efficient β -sialon:Eu is not facile. We have thus explored M-Si-O-N system for such green phosphors. Although MSi2O2N2[4] and Ba3Si6O9N4[5] have been known, it has not been reported that the both compounds doped with rare-earth element work as efficient green phosphors with high color purity.
In the present work, we have successfully synthesized a new green phosphor, (Ba,Eu)3Si6O12N2[6]. We have also identified the crystal structure by a new protocol combining X-ray/neutron powder diffraction analysis with first-principles study. Although Ba3Si6O9N4 looks similar to Ba3Si6O12N2 from the viewpoint of crystal structure and chemical formula, their optical properties are quite different; (Ba,Eu)3Si6O9N4 exhibits blue-green photoluminescence (PL) only at low temperatures (i.e. little luminescence at room temperature (R.T.)), whereas (Ba,Eu)3Si6O12N2 exhibits intense green PL at R.T. with thermal quenching smaller than (Ba,Sr,Eu)2SiO4. In this paper, the crystal structure and optical properties of the new green phosphor are described. The interpretation of the origin of the different optical properties between (Ba,Eu)3Si6O12N2 and (Ba,Eu)3Si6O9N4 are briefly explained.
Experimental and theoretical procedure
A single phase of the new compound was prepared by conventional method [6]. The X-ray diffraction (XRD) indicated that the compound should be a new crystal phase. The crystal symmetry (trigonal) and approximate lattice constants (a=7.48(1)Å, c=6.47(1)Å) were obtained by TEM. From the chemical analysis (O-rich) and weight density (= 4.13 g/cm3) indicating Z=1, we assumed that the compound has crystal structure close to Ba3Si6O9N4 (P3, a = 7.249(1)Å, c = 6.784(2)Å, Z=1). Among possible compositions, Ba3Si6OxNy (x > y) with (x, y)=(6,6),(9,4),(12,2), only Ba3Si6O12N2 had not been reported. Thus the crystal structure of Ba3Si6O12N2 was approximately solved as P3 by a direct method from the XRD. The geometry was then optimized by ab initio calculation based on density
New Green Phosphor Ba3Si6O12N2:Eu for White LED: Crystal Structure and Optical Properties
Masayoshi Mikami1,a, Satoshi Shimooka1, Kyota Uheda1, Hiroyuki Imura1 and Naoto Kijima1
1 Mitsubishi Chemical Group Science and Technology Research Center, Inc. 1000, Kamoshida-cho, Aoba-ku, Yokohama, 227-8502, Japan
amikami.masayoshi@mv.m-kagaku.co.jp
Abstract. A new oxynitride, Ba3Si6O12N2, has been synthesized. The crystal structure has been successfully determined by close collaboration between experiment and first-principles calculation. This compound doped with Eu exhibits intense green photoluminescence with high color purity under near-ultraviolet to blue light excitation; in particular, it has much less thermal quenching than (Ba,Sr,Eu)2SiO4. Thus (Ba,Eu)3Si6O12N2 appears promising green phosphor for white LED backlight for display. The atomic/electronic structure is discussed in comparison with Ba3Si6O9N4, which could not become efficient phosphor by doping Eu due to strong thermal quenching at room temperature.
Introduction
In the white LED market, the share of green and red phosphors has been gradually increasing to achieve good color reproducibility for liquid crystal display (LCD) backlight. Promising candidates that satisfy high color purity are, e.g., (M,Eu)2SiO4[1] and β -sialon:Eu[2] for green phosphor, and (M,Eu)AlSiN3[3] for red phosphor (M: alkaline-earth element). The green phosphors still confront some difficulties; (M,Eu)2SiO4 has strong thermal quenching[1], whereas the synthesis of efficient β -sialon:Eu is not facile. We have thus explored M-Si-O-N system for such green phosphors. Although MSi2O2N2[4] and Ba3Si6O9N4[5] have been known, it has not been reported that the both compounds doped with rare-earth element work as efficient green phosphors with high color purity.
In the present work, we have successfully synthesized a new green phosphor, (Ba,Eu)3Si6O12N2[6]. We have also identified the crystal structure by a new protocol combining X-ray/neutron powder diffraction analysis with first-principles study. Although Ba3Si6O9N4 looks similar to Ba3Si6O12N2 from the viewpoint of crystal structure and chemical formula, their optical properties are quite different; (Ba,Eu)3Si6O9N4 exhibits blue-green photoluminescence (PL) only at low temperatures (i.e. little luminescence at room temperature (R.T.)), whereas (Ba,Eu)3Si6O12N2 exhibits intense green PL at R.T. with thermal quenching smaller than (Ba,Sr,Eu)2SiO4. In this paper, the crystal structure and optical properties of the new green phosphor are described. The interpretation of the origin of the different optical properties between (Ba,Eu)3Si6O12N2 and (Ba,Eu)3Si6O9N4 are briefly explained.
Experimental and theoretical procedure
A single phase of the new compound was prepared by conventional method [6]. The X-ray diffraction (XRD) indicated that the compound should be a new crystal phase. The crystal symmetry (trigonal) and approximate lattice constants (a=7.48(1)Å, c=6.47(1)Å) were obtained by TEM. From the chemical analysis (O-rich) and weight density (= 4.13 g/cm3) indicating Z=1, we assumed that the compound has crystal structure close to Ba3Si6O9N4 (P3, a = 7.249(1)Å, c = 6.784(2)Å, Z=1). Among possible compositions, Ba3Si6OxNy (x > y) with (x, y)=(6,6),(9,4),(12,2), only Ba3Si6O12N2 had not been reported. Thus the crystal structure of Ba3Si6O12N2 was approximately solved as P3 by a direct method from the XRD. The geometry was then optimized by ab initio calculation based on density
Key Engineering Materials Vol. 403 (2009) pp 11-14 © (2009) Trans Tech Publications, Switzerland doi:10.4028/www.scientific.net/KEM.403.11
functional theory(DFT)[7]. It turned out that the calculated geometry has inversion symmetry, which could not be detected by the XRD analysis. The optimized crystal structure was used for initial inputs for Rietveld analysis of the X-ray/neutron diffraction. The structure was finally obtained as P3 .
Results and Discussions
The crystal structure of Ba3Si6O12N2 is illustrated in Fig. 1. The crystallographic data are summarized in Table 1. The calculated lattice constants overestimate the experimental data due to the approximation, generalized-gradient correction (GGA), in the DFT. It is noted that the calculated parameters were used for the Rieltveld analysis of XRD/neutron diffraction (not given here).
Table 1: Crystallographic data of Ba3Si6O12N2 (experimental and theoretical data) System (Space group, No.) Trigonal (P 3 , No. 147)
Experiment (from XRD) Theory Lattice parameters / Å a=7.5046(8), c=6.4703(5) a=7.59684, c=6.57487
Atomic coordinates Label Wyckoff-position x y z x y z Ba1 1a 0 0 0 0 0 0 Ba2 2d 1/3 2/3 0.1039(2) 1/3 2/3 0.10060 Si1 6g 0.2366(6) 0.8310(6) 0.6212(8) 0.23594 0.82847 0.60978 N1 2d 1/3 2/3 0.568(3) 1/3 2/3 0.56311 O1 6g 0.356(2) 0.295(2) 0.173(1) 0.36096 0.29646 0.17190 O2 6g 0.000(1) 0.681(1) 0.589(2) -0.01462 0.68138 0.58846
The fused rings-sheet, 2 ∞ [(Si ]4[
6 O ]2[ 6 N ]3[
2 )O ]1[ 6 ] −6 , is composed of 8-membered Si-(O,N) and
12-membered Si-O rings. The compound is built up of corner sharing SiO3N tetrahedra forming corrugated layers between which the Ba2+ ions are located. The Ba2+ ions occupy two different crystallographic sites; one is trigonal anti-prism (distorted octahedron) with six oxygen atoms, and the other is trigonal anti-prism with six oxygen atoms, further capped with a nitrogen atom (Fig.1(c)).
The crystal structure and chemical formula of Ba3Si6O12N2 appears close to Ba3Si6O9N4. We review the crystal structure of Ba3Si6O9N4 by following Ref. 5. The compound is composed by corner sharing SiO2N2 tetrahedra forming corrugated layers between which the Ba2+ ions are located. The Ba2+ ions occupy three different crystallographic sites; two of them are trigonal anti-prisms with six oxygen atoms, and the other is trigonal anti-prism with six oxygen atoms, further capped with a nitrogen atom (Fig.2). The Ba clusters in Ba3Si6O9N4 looks similar to those in Ba3Si6O12N2. Still, a main difference is the Ba-N distance: about 3.2 Å in Ba3Si6O9N4 whereas about 3.0 Å in Ba3Si6O12N2. It is also noteworthy that the host absorption band of the Ba3Si6O9N4 appears lower in energy than that of Ba3Si6O12N2, although the both absorption were observed below 300 nm from the diffuse reflectance spectra. This is also implied by our band structure calculation; the computed band gap of Ba3Si6O12N2 is 4.63 eV whereas that of Ba3Si6O9N4
is 4.37 eV. Although it is known that band gap calculated within GGA will be underestimated, computed band gaps can be qualitatively compared.
Figure 1: Projections of the unit cell of Ba3Si6O12N2 viewed along the c axis (a) and b axis (b), and two coordination environments around Ba2+ ion (c); the clusters are defined within 3.2 Å.
functional theory(DFT)[7]. It turned out that the calculated geometry has inversion symmetry, which could not be detected by the XRD analysis. The optimized crystal structure was used for initial inputs for Rietveld analysis of the X-ray/neutron diffraction. The structure was finally obtained as P3 .
Results and Discussions
The crystal structure of Ba3Si6O12N2 is illustrated in Fig. 1. The crystallographic data are summarized in Table 1. The calculated lattice constants overestimate the experimental data due to the approximation, generalized-gradient correction (GGA), in the DFT. It is noted that the calculated parameters were used for the Rieltveld analysis of XRD/neutron diffraction (not given here).
Table 1: Crystallographic data of Ba3Si6O12N2 (experimental and theoretical data) System (Space group, No.) Trigonal (P 3 , No. 147)
Experiment (from XRD) Theory Lattice parameters / Å a=7.5046(8), c=6.4703(5) a=7.59684, c=6.57487
Atomic coordinates Label Wyckoff-position x y z x y z Ba1 1a 0 0 0 0 0 0 Ba2 2d 1/3 2/3 0.1039(2) 1/3 2/3 0.10060 Si1 6g 0.2366(6) 0.8310(6) 0.6212(8) 0.23594 0.82847 0.60978 N1 2d 1/3 2/3 0.568(3) 1/3 2/3 0.56311 O1 6g 0.356(2) 0.295(2) 0.173(1) 0.36096 0.29646 0.17190 O2 6g 0.000(1) 0.681(1) 0.589(2) -0.01462 0.68138 0.58846
The fused rings-sheet, 2 ∞ [(Si ]4[
6 O ]2[ 6 N ]3[
2 )O ]1[ 6 ] −6 , is composed of 8-membered Si-(O,N) and
12-membered Si-O rings. The compound is built up of corner sharing SiO3N tetrahedra forming corrugated layers between which the Ba2+ ions are located. The Ba2+ ions occupy two different crystallographic sites; one is trigonal anti-prism (distorted octahedron) with six oxygen atoms, and the other is trigonal anti-prism with six oxygen atoms, further capped with a nitrogen atom (Fig.1(c)).
The crystal structure and chemical formula of Ba3Si6O12N2 appears close to Ba3Si6O9N4. We review the crystal structure of Ba3Si6O9N4 by following Ref. 5. The compound is composed by corner sharing SiO2N2 tetrahedra forming corrugated layers between which the Ba2+ ions are located. The Ba2+ ions occupy three different crystallographic sites; two of them are trigonal anti-prisms with six oxygen atoms, and the other is trigonal anti-prism with six oxygen atoms, further capped with a nitrogen atom (Fig.2). The Ba clusters in Ba3Si6O9N4 looks similar to those in Ba3Si6O12N2. Still, a main difference is the Ba-N distance: about 3.2 Å in Ba3Si6O9N4 whereas about 3.0 Å in Ba3Si6O12N2. It is also noteworthy that the host absorption band of the Ba3Si6O9N4 appears lower in energy than that of Ba3Si6O12N2, although the both absorption were observed below 300 nm from the diffuse reflectance spectra. This is also implied by our band structure calculation; the computed band gap of Ba3Si6O12N2 is 4.63 eV whereas that of Ba3Si6O9N4
is 4.37 eV. Although it is known that band gap calculated within GGA will be underestimated, computed band gaps can be qualitatively compared.
Figure 1: Projections of the unit cell of Ba3Si6O12N2 viewed along the c axis (a) and b axis (b), and two coordination environments around Ba2+ ion (c); the clusters are defined within 3.2 Å.
12 SiAlONs and Non-oxides
Next, we have examined the optical properties of (Ba,Eu)3Si6O12N2 and (Ba,Eu)3Si6O9N4. In the
both compounds, Eu2+ ions are supposed to occupy the Ba-sites (Fig.1(c) and Fig.2). The spectra of PL and PL excitation (PLE) have been measured from liquid-He temperature to R.T. The details will be discussed elsewhere [8], so that we briefly introduce the results here. Under the near-UV to blue light irradiation, (Ba,Eu)3Si6O12N2 exhibited intense broad green emission spectrum (PL peak at about 530 nm, with full width at half maximum of 68nm) at R.T. As for the PLE of (Ba,Eu)3Si6O12N2, the broad excitation bands were observed in wavelengths ranging from 200 to 500 nm. The PL/PLE should originate from the allowed transition from 4f7 grand state to 5d state of Eu2+, because the host absorption is below 300 nm. It is underscored that the phosphor will be suitable for LED backlight in LCD owing to its high green color purity with the CIE color coordinates (x,y)=(0.274, 0.644) similar to those of (Ba,Sr,Eu)2SiO4. It was confirmed that thermal quenching of (Ba,Eu)3Si6O12N2 is much smaller than (Ba,Sr,Eu)2SiO4; the emission intensity of (Ba,Eu)3Si6O12N2 at 100ºC was about 90% of that measured at R.T., whereas the emission intensity of (Ba,Sr,Eu)2SiO4 at 100ºC was about 75% of that at R.T. Thus (Ba,Eu)3Si6O12N2 appears a promising green phosphor for LCD backlight use.
On the other hand, the PL of (Ba,Eu)3Si6O9N4, broad blue-green emission spectra, was observed only at low temperature (lower than RT) due to its strong thermal quenching. The PL peak was about 480 nm, shorter than the PL peak of (Ba,Eu)3Si6O12N2 (about 530 nm). The excitation bands ranging from 200 to 440 nm was narrower than that of (Ba,Eu)3Si6O12N2. The PL and PLE also originate from the allowed transition from 4f7 grand state to 5d state of Eu2+. We did not see large difference in Stokes shift for the both compounds. The question is why the PL/PLE spectra and the thermal quenching behaviors were quite different for (Ba,Eu)3Si6O12N2 and (Ba,Eu)3Si6O9N4, even with the similar structure of the Ba-clusters (i.e. the trigonal symmetry at the Ba centers, Fig.1(c) and Fig.2).
Figure 2: Coordination environments around Ba2+ ion (defined within 3.2 Å) in Ba3Si6O9N4 [5].
Figure 3: Schematic illustration of level energies in Eu2+doped barium silicon oxynitrides [conduction band (C.B.) , valence band (V.B.) and photoluminescence (PL)] . Dotted lines and broken lines denote crystal field splitting due to Eu-O ligand and Eu-N ligand, respectively. The autoionization process (Eu2+àEu3++e-) is denoted in Eu-doped Ba3Si6O9N4 (right). Stokes shifts and details of the trigonal distortion splits of Eu-d states are not drawn for simplification.
Next, we have examined the optical properties of (Ba,Eu)3Si6O12N2 and (Ba,Eu)3Si6O9N4. In the
both compounds, Eu2+ ions are supposed to occupy the Ba-sites (Fig.1(c) and Fig.2). The spectra of PL and PL excitation (PLE) have been measured from liquid-He temperature to R.T. The details will be discussed elsewhere [8], so that we briefly introduce the results here. Under the near-UV to blue light irradiation, (Ba,Eu)3Si6O12N2 exhibited intense broad green emission spectrum (PL peak at about 530 nm, with full width at half maximum of 68nm) at R.T. As for the PLE of (Ba,Eu)3Si6O12N2, the broad excitation bands were observed in wavelengths ranging from 200 to 500 nm. The PL/PLE should originate from the allowed transition from 4f7 grand state to 5d state of Eu2+, because the host absorption is below 300 nm. It is underscored that the phosphor will be suitable for LED backlight in LCD owing to its high green color purity with the CIE color coordinates (x,y)=(0.274, 0.644) similar to those of (Ba,Sr,Eu)2SiO4. It was confirmed that thermal quenching of (Ba,Eu)3Si6O12N2 is much smaller than (Ba,Sr,Eu)2SiO4; the emission intensity of (Ba,Eu)3Si6O12N2 at 100ºC was about 90% of that measured at R.T., whereas the emission intensity of (Ba,Sr,Eu)2SiO4 at 100ºC was about 75% of that at R.T. Thus (Ba,Eu)3Si6O12N2 appears a promising green phosphor for LCD backlight use.
On the other hand, the PL of (Ba,Eu)3Si6O9N4, broad blue-green emission spectra, was observed only at low temperature (lower than RT) due to its strong thermal quenching. The PL peak was about 480 nm, shorter than the PL peak of (Ba,Eu)3Si6O12N2 (about 530 nm). The excitation bands ranging from 200 to 440 nm was narrower than that of (Ba,Eu)3Si6O12N2. The PL and PLE also originate from the allowed transition from 4f7 grand state to 5d state of Eu2+. We did not see large difference in Stokes shift for the both compounds. The question is why the PL/PLE spectra and the thermal quenching behaviors were quite different for (Ba,Eu)3Si6O12N2 and (Ba,Eu)3Si6O9N4, even with the similar structure of the Ba-clusters (i.e. the trigonal symmetry at the Ba centers, Fig.1(c) and Fig.2).
Figure 2: Coordination environments around Ba2+ ion (defined within 3.2 Å) in Ba3Si6O9N4 [5].
Figure 3: Schematic illustration of level energies in Eu2+doped barium silicon oxynitrides [conduction band (C.B.) , valence band (V.B.) and photoluminescence (PL)] . Dotted lines and broken lines denote crystal field splitting due to Eu-O ligand and Eu-N ligand, respectively. The autoionization process (Eu2+àEu3++e-) is denoted in Eu-doped Ba3Si6O9N4 (right). Stokes shifts and details of the trigonal distortion splits of Eu-d states are not drawn for simplification.
Key Engineering Materials Vol. 403 13
From the observation of the Ba-cluster models, we may suppose that the excited states of Eu is composed as the superposition of the Eu-O and the Eu-N crystal field splitting; if E-N ligand would not affect PL/PLE, (Ba,Eu)3Si6O12N2 and (Ba,Eu)3Si6O9N4 should have exhibited similar PL/PLE properties due to the same local symmetry with the similar Ba(Eu)-O lengths. Since we see the difference in the above, the Eu-N ligand should be effective. Thus we focus on the trigonal anti-prisms with six O atoms plus one N atom (Fig.3). Our PL/PLE results suggest that the Eu d-states from Eu-N ligand determine the lowest excited state, because of the nephelauxetic effect originating from anion polarizability; note that N3- is more polarizable/covalent than O2-. Since the Ba-N length in (Ba,Eu)3Si6O12N2 (about 3.0Å) is shorter than that in (Ba,Eu)3Si6O9N4 (about 3.2Å), the Eu-N crystal field splitting in (Ba,Eu)3Si6O12N2 is expected to be wider than that in (Ba,Eu)3Si6O9N4.
With the above assumption in mind, we are ready to explain our experimental results. The blue shift of the PL peak of 530 nm ((Ba,Eu)3Si6O12N2) to 480 nm ((Ba,Eu)3Si6O9N4) originates from the higher Eu-d states of (Ba,Eu)3Si6O9N4 than that of (Ba,Eu)3Si6O12N2. Remembering the smaller band gap of (Ba,Eu)3Si6O9N4, we suppose that the lowest Eu-d states are so close to the conduction bands in (Ba,Eu)3Si6O9N4 that the photo-excited 5d-electrons of Eu2+ can be thermally ionized to the conduction bands at R.T.[9] The broadness of the excitation bands may be related to the superposition of the Eu-O and the Eu-N crystal field splittings. Incidentally, the narrowness of the PL may be reflected by the anisotropic structure of the trigonal anti-prisms with six O atoms plus one atom.
Summary
A new oxynitride, Ba3Si6O12N2, has been synthesized and its crystal structure has been determined. It means that we have a series of composition, Ba3Si6O6N6(=BaSi2O2N2), Ba3Si6O9N4, Ba3Si6O12N2, and Ba3Si6O15(=BaSi2O5), by substituting N2 with O3 formally. (Ba,Eu)3Si6O12N2 exhibits efficient green photoluminescence with high color purity under InGaN diode irradiation; in particular, it has much less thermal quenching than the other green phosphor, (Ba,Sr,Eu)2SiO4. Although the crystal structure and chemical formula appears close to Ba3Si6O9N4, their optical properties and thermal quenching behaviors are quite different. Stronger thermal quenching in (Ba,Eu)3Si6O9N4 may be ascribed to smaller band gap and longer Ba-N distance (i.e. smaller crystal field splitting).
Acknowledgment: We wish to thank Dr. M. Takashima for the TEM analysis, Dr. Y. Sasaki for the PL/PLE measurement (liquid-He temperature to room temperature), and Mr. K. Horibe for LED fabrication. The neutron diffraction experiment was performed at Japan Atomic Energy Agency.
References
[1] T. L. Barry : J. Electrochem. Soc. Vol. 115 (1968), p. 1181.
[2] N. Hirosaki et al., Appl. Phys. Lett.86 (2005), p.211905.
[3] K. Uheda et al., Electrochem. Solid State Lett. Vol. 9, (2006), p. H22.
[4] R. Müller-Mach et al., phys. stat. sol. (a) Vol. 202 (2005), p. 1727.
[5] F. Stadler and W. Schnick, Z. Anorg. Allg. Chem. Vol. 632 (2006), p. 949.
[6] S. Shimooka, K. Uheda, M. Mikami, N. Kijima, H. Imura, and K. Horibe, PCT WO2007/088966.
[7] The ABINIT code is a common project of the Université Catholique de Louvain, Corning Incorporated, and other contributors (URL http://www.abinit.org).