A New, Low-Temperature Polymorph of O -SiAlON Mark E. Bowden, Glen C. Barris, and Ian W. M. Brown Industrial Research Limited, Lower Hutt, New Zealand David A. Jefferson Department of Chemistry, University of Cambridge, Cambridge CB2 1EW, United Kingdom A new phase in the Si-Al-O-N system has been identified, following syntheses based on the nitridation of silicon/clay mixtures at low temperatures (<1350°C). The structure of the new phase was determined using a combination of dif- fraction and high-resolution imaging techniques, and this new phase possessed the same sheet structure as O-SiAlON (Si 2-x Al x O 1+x N 2-x ) but with a different stacking arrange- ment. It is considered to be a low-temperature polymorph of O-SiAlON and transforms to conventional O-SiAlON at temperatures greater than ∼1350°C. I. Introduction T HE Si-Al-O-N system has become the subject of widespread research since the initial discovery 1,2 that aluminum and oxygen can be substituted into the Si 3 N 4 lattice. One reason for the intensity of the research effort is the many different SiAlON phases that can be prepared, each of which presents different physical properties. The Si-Al-O-N phase diagram was explored in the early history of SiAlON research, and by the time Jack performed his review in 1976, 3 all the presently known phases had been discovered. In this communication, we describe the structure of the first new SiAlON to be found since that times: a low-temperature polymorph of O'-SiAlON (Si 2-x Al x O 1+x N 2-x ). 4 II. Experimental Procedure Mixtures of clay (light kaolin, BDH Chemicals Pty., Ltd., Poole, U.K.), silicon (grade 4D, Permascand AB, Ljungaverk, Sweden), and silica (Superfine Quartz, Commercial Minerals Ltd., Auckland, New Zealand) were blended for 20 h in a ball mill using Si 3 N 4 media and hexane solvent. Major impurities (>0.05%) in this clay are 0.75% K 2 O, 0.45% Fe 2 O 3 , 0.35% MgO, 0.21% P 2 O 5 , and 0.14% Na 2 O. The stoichiometry for each mixture was calculated according to the following reac- tion scheme, illustrated for the x  0.2 O'-SiAlON composition: 0.1Si 2 Al 2 O 5 (OH29 4 + 1.35Si + 0.25SiO 2 + 0.9N 2 → Si 1.8 Al 0.2 O 1.2 N 1.8  0.2H 2 O After drying, the powder was lightly pressed (8 MPa) into disks 15 mm in diameter and heated under flowing H 2 (10%) in N 2 for 8 h at 1270°C. X-ray diffractometry (XRD) was performed using a system (Model PW1700, Philips Research Laboratories, Eindhoven, The Netherlands) with a post-diffraction graphite monochro- mator and CoK radiation. High-resolution electron micros- copy (HREM) utilized a 200 keV instrument (Model 200CX, JEOL, Tokyo, Japan) equipped with a high-resolution (C s  0.41 mm) pole piece. Simulated HREM images were calculated with the EMS software system, 5 using values of 0.5 mrad for beam divergence and 100 Å for focal spread. III. Results and Discussion The first indications of the formation of a new phase arose when the XRD patterns of reacted mixtures could not be matched to any known compounds. The unmatched XRD peaks were particularly apparent in samples reacted at tempera- tures <1300°C, which is a temperature considerably lower than those normally used for solid-state syntheses of O'-SiAlON (e.g., 1800°C 6 ). The ability to prepare SiAlON phases at low temperature was a consequence of the synthesis method used and, in particular, the use of clay and silicon as reactants. Adding small amounts of iron (1.5% as Fe 2 O 3 ) improved the reaction rate, as has been observed for the nitridation of silicon to Si 3 N 4 . 7 The results obtained here do not show whether the iron or other impurities derived from the clay have any role in stabilizing the new phase, in a manner similar to the additional cations used to stabilize -SiAlON. However, we do note that the level of impurities in the present material is considerably lower than that required to stabilize -SiAlON. The composition of the new phase approximated that of O'-SiAlON, as reacted mixtures with this composition pro- duced the highest yields. In addition, the new phase trans- formed to O'-SiAlON at 1350°C without a significant weight change and without the appearance or disappearance of sec- ondary phases. All these factors have led us to believe that the low-temperature conditions used here permitted the synthesis of a new, low-temperature polymorph of O'-SiAlON. Although the crystallites prepared were too small for single- crystal analysis, the structure was solved by using a combina- tion of HREM and selected-area electron diffraction. Electron diffraction patterns taken from different zone axes suggested an orthorhombic unit cell, and preliminary structural coordinates were proposed on the basis of the HREM images. Simulation of these images and the XRD pattern confirmed the proposed structure, which was subsequently refined from the observed XRD data using the Rietveld technique. Some small discrep- ancies between the observed and calculated XRD patterns re- mained after the refinement; these discrepancies were ascribed to the presence of planar defects in the specimen (this point will be discussed later in the paper). However, the overall agree- ment (Bragg factor (R) of 5.37%, goodness of fit is 10.3) was sufficient to give us considerable confidence in the validity of the model. Structural coordinates that resulted from these re- finements are given in Table I, and tabulated XRD pattern data are provided in Table II. The structure is best described by comparing it with O'- SiAlON, 6 and projections of both structures are shown in Fig. 1. For simplicity, we will consider the x  0 composition (Si 2 N 2 O), although the actual specimen that was examined contained 10 mol% of aluminum and oxygen, substituted on T. Ekstro ¨m—contributing editor Manuscript No. 190375. Received February 16, 1998; approved April 27, 1998. J. Am. Ceram. Soc., 81 [8] 2188–90 (1998) J ournal 2188