High-Temperature Stability of the Al 2 O 3 –LaPO 4 System David B. Marshall, * Peter E. D. Morgan, * Robert M. Housley, * and Jeffrey T. Cheung Rockwell Science Center, Thousand Oaks, California 91360 The compatibility of Al 2 O 3 and LaPO 4 at temperatures up to 1600°C is examined. Provided the ratio of La to P was close to 1:1, no reactions were observed after 200 h at 1600°C. Moreover, the Al 2 O 3 /LaPO 4 interface remained sufficiently weakly bonded to cause deflection of cracks, as reported previously. In the presence of excess P or La, reactions occurred as expected, forming AlPO 4 in the case of excess P, and LaAlO 3 and LaAl 11 O 18 in the case of ex- cess La. I. Introduction L ANTHANUM PHOSPHATE (La-monazite) was recently sug- gested as a potential interphase in oxide composites to allow debonding between the matrix and reinforcement, as re- quired for damage tolerance. 1–3 The characteristics of LaPO 4 that make it appealing for this purpose are the following: the low fracture energy of the interface between it and several other oxides (Al 2 O 3 , ZrO 2 , mullite); its refractory nature 4 (melting point 2074°C); its stability in oxidizing environments; and its expected compatibility with other oxides (natural mona- zite is known to be stable with many other minerals in a wide variety of geological environments). 5 La-monazite is a stoichi- ometric La:P, 1:1 compound. The stability of the system LaPO 4 /Al 2 O 3 in an air environ- ment has been examined at temperatures up to 1600°C in sev- eral previous studies. In one, the presence of impurities of alkali metals or divalent elements (K, Mg, Ca, and others) promoted formation of La-containing -alumina–magneto- plumbite phases, especially near free surfaces. 3 In the absence of these elements, no reactions were observed after periods up to 24 h at 1400° and 1600°C. However, apparently conflicting results were reported in another study, 6 in which no reactions were observed at 1550°C, but the formation of LaAl 11 O 18 be- tween LaPO 4 and Al 2 O 3 was observed at 1600°C. The purpose of this paper is to examine the long-term compatibility of the LaPO 4 /Al 2 O 3 system, and in particular the role of the La/P ratio, in an air environment. II. Experiments Two sources of monazite were used to fabricate composites containing Al 2 O 3 and LaPO 4 . One was a precipitated slurry of hexagonal rhabdophane (LaPO 4  1 /2H 2 O) precursor for mona- zite, synthesized in our laboratory. This precursor required ex- haustive washing with water to remove excess phosphorus. The other source was coatings deposited by laser ablation, using a target fabricated from the rhabdophane precursor. One of us (J.T.C.) has found conditions under which fibers may be coated uniformly around their circumference without rotation, thus allowing multiple fibers to be coated together. The coatings used in this study were from an early batch, with an inhomo- geneous columnar growth morphology and radial variations of density and composition. Although the inhomogeneity made quantitative analysis difficult, EDS analysis indicated a net deficiency in phosphorus. In the first experiments, a slurry of rhabdophane precursor that had been washed insufficiently to remove the excess phos- phorus was mixed with a small amount of Al 2 O 3 (<2 vol%, particle diameters up to 10 m) and consolidated by filter pressing. After drying, the consolidated body was cold isostati- cally pressed at 300 MPa, then sintered at 1300°C for 6 h. After further washing of this rhabdophane slurry to achieve the 1:1 ratio of La:P (and addition of 1 wt% AlO(OH) as a buffer), the slurry was used to form coatings on individual sapphire fibers by dip-coating. After allowing the coatings to dry, the coated fibers were embedded in high-purity -aluminia powder (Sumitomo AKP30) and hot-pressed for 1 h at 1400°C. Also included in the same composite were other sapphire fibers that had been coated with LaPO 4 by laser ablation. The LaPO 4 coating in that case contained excess La. The fibers were aligned approximately parallel. After hot pressing, the composite was cut into slices ∼2 mm thick normal to the fibers and each slice was heat-treated in air for 2, 22, or 200 h at 1400° or 1600°C. After heat treatment, the surface normal to the fibers was polished using diamond me- dia, removing a depth of approximately 100 m, and examined using scanning electron microscopy (SEM) with energy dis- persive spectroscopy (EDS). Vickers indentations (100 N load) were positioned in the polished polycrystalline alumina matrix near selected fibers so as to generate cracks that impinged upon the fibers and tested the debonding characteristics of the coat- ing. The slices were also fractured in three-point bending using indentations adjacent to selected fibers to initiate fracture and thereby allow observation of the debonded surfaces of the fi- bers and coatings. III. Results (1) Sintered Rhabdophane (P-rich)–Alumina After firing at 1300°C for 6 h, the rhabdophane had con- verted to monazite with a La:P ratio of 1:1, as measured by EDS on a polished cross section. However, all Al 2 O 3 particles were surrounded by a coating containing Al and P in the correct proportions for AlPO 4 with no detectable La (Fig. 1). The thicknesses of the AlPO 4 layers varied with the location of the Al 2 O 3 particles: near the center of the sintered pellet the layer thicknesses were ∼1 to 2 m, whereas near the surfaces of the pellet, where excess phosphorus could escape, the layer thick- nesses were smaller (∼0.2 to 0.4 m). After thermal etching of the polished surface at 1300°C for 1 h, the grain structure of the monazite was visible (Fig. 2(b)), while the exposed AlPO 4 layers had undergone severe etching, presumably associated with loss of phosphorus. (2) Coated Fibers (A) Heat Treatment at 1400°C: Fiber coatings formed from the more thoroughly washed and buffered slurry and from D. J. Green—contributing editor Manuscript No. 191204. Received February 13, 1997; approved June 4, 1997. Supported by the U.S. Office of Naval Research under Contract No. N00014-95- C-0057 monitored by Dr. S. G. Fishman. * Member, American Ceramic Society. J. Am. Ceram. Soc., 81 [4] 951–56 (1998) J ournal 951