You are on page 1of 5

JO U R N AL O F M A TE R IA LS S CI E NC E 3 2 (1 9 9 7) 3 86 3 38 6 7

Synthesis and characterization of lanthanum


phosphate sol for fibre coating
PE N G C HE N
Department of Chemistry, University of North Texas, PO Box 5068, Denton, TX 76203 USA
TAI - I L MA H
UES, Inc., 4401 Dayton-Xenia Rd. Dayton, OH 45432 USA

Various routes to synthesize lanthanum phosphate sols (LaPO4) were explored. Prepared
lanthanum phosphate sols were characterized using differential thermal analysis/
thermogravimetric analysis (DTA/TGA) and powder X-ray diffraction (XRD) after heat
treatments. Appreciable crystallization of lanthanum phosphate was observed above 400 C
and the average crystallite sizes were calculated to be 15 nm and 46 nm after heat treatment
at 700 and 1100 C, respectively. Sapphire fibres were dip coated in LaPO4 sol, and
subsequently heated to 1100 C. The thickness of the coating ranged from 60 to 150 nm
depends upon the sol concentration and the number of times of dip coatings. XRD of coated
fibres revealed the formation of LaPO4 after heat treatment above 700 C.

1. Introduction 2. Experiments
To achieve tough behaviour of ceramic matrix com- 2.1. Synthesis and characterization
posites (CMC), the fibre/matrix interlayer should be of lanthanum phosphate sol
weak to promote crack deflection and fibre pull-out Lanthanum phosphate (LaPO , Monazite) can be
4
[1, 2]. Various interlayers have been employed to synthesized through various reactions and/or combi-
achieve this goal: in situ formed carbon interlayer, nations of them [1524]. Several of these reaction
fugitive carbon or molybdenum coating, easily cleav- routes were investigated in this study and the experi-
able oxides such as beta alumina and mica, etc. mental details of each route were as follows.
[312]. Another requirement for CMC is that the
composite constituents, as well as the interlayer(s),
should not only be compatible with each other 2.1.1. Method A
but should also be oxidation resistant at elevated The first method involves using an equal molar ratio
temperatures. of LaPO and H PO [24]
4 3 4
Recently, Morgan and Marshall [7] studied poten-
LaCl #H PO " LaPO #3HCl (1)
tial interfacial coating materials for sapphire/alumina 3 3 4 4
composites, with mechanical weakness, high tem- An approximately 20 wt % phosphoric acid solution
perature chemical compatibility and stability in oxidiz- was reacted with lanthanum chloride. White precipi-
ing environments criteria. SnO , LaPO , CaF , and
2 4 2 tates were immediately formed. The precipitates were
various aluminates having the magnetoplumbite-b- washed with water three times and dried at room
alumina structure were studied. They used the candi- temperature.
date interlayer materials as matrices for sapphire
fibres or plates. Based upon the observations of their
preliminary assessment study, they concluded LaPO
4 2.1.2. Method B
would be a good candidate for coating material. Syn- Lanthanum phosphate precipitates were prepared
thesis of LaPO sol, which is suitable for fibre coating,
4 by the neutral reaction of lanthanum nitrate with
was not reported. ammonium phosphate at room temperature (reaction
The objective of the present research effort is to 4). It was reported that large, high purity LaPO
study various routes to synthesize lanthanum phos- 4
crystals could be produced by this reaction [25]. Dis-
phate (LaPO ) sols that are suitable for continuous
4 advantages of this reaction include the cumbersome
single fibre and fibre tow coating [13, 14], and charac- procedures and the various by-products that are
terize these samples by various techniques including formed
differential thermal analysis/thermogravimetric analy-
sis (DTA/TGA), scanning electron microscopy (SEM) H PO #NH OH " NH H PO #H O
3 4 4 4 2 4 2
and X-ray diffraction (XRD). (2)
00222461 ( 1997 Chapman & Hall 3863
La O [or 2La(OH) ]#6HNO sample was stirred for 10 h to ensure the completion of
2 3 3 3
the reaction. This was followed by centrifugal separ-
P 2La(NO ) #3H O (3)
33 2 ation and washing the sample with water three times.
La(NO ) #NH H PO A large-scale batch (2 moles of LaPO ) of the sol was
33 4 2 4 4
easily prepared using this reaction route. The coating
" LaPO #NH NO #2HNO (4)
4 4 3 3 experiments were performed using this sol (diluted to
The ammonium phosphate was prepared by gradually various concentrations).
adding 85 wt % phosphoric acid to ammonium hy- The ratio between phosphate and lanthanum in the
droxide and water solution until a pH of 7&8 was reaction mixture is important in terms of the final
reached. It was then cooled to allow NH H PO to lanthanum phosphate formed [28]. An experiment
4 2 4
crystallize and then vacuum filtered. was carried out to confirm the formation of LaP O
3 9
The lanthanum nitrate was prepared by dropwise according to the following reaction equation
addition of concentrated nitric acid into a solution of
La(OH) #3H PO " LaP O #6H O (7)
lanthanum oxide and water. The mixture was stirred 3 3 4 3 9 2
until all the solids disappeared. The pH of the solution In this case, the sol was prepared with a molar ratio of
was measured between 6 and 7. Since lanthanum ni- 1:3 instead of 1:1. The dried sol was heat treated for
trate is extremely difficult to crystallize, due to its high 2 h at 700 C before examination using XRD.
solubility in water [26], the resulting solution was
allowed to cool at 4 C to crystallize. The crystals
were separated by vacuum filtration and recrystallized 2.2. Fibre dip coating process
again to obtain purified lanthanum nitrate. The crys- The coating experiments were carried out using sim-
tals were then dried in air before use. ilar technique reported in the literature [2931]. The
Equimolar amounts of lanthanum nitrate and am- sapphire fibres, obtained from Advanced Crystal
monium phosphate were dissolved into water to make Products Corporation, Woburn, MA, was used for the
0.1 M solutions. During the vigorous mixing of these dip coating experiments. The fibres were cut into 7.62-
solutions, white insoluble LaPO precipitates were cm long pieces and cleaned with a 85% H PO solu-
4 3 4
formed. These precipitates were separated using a cen- tion. After the acid cleaning, the surface of the fibre
trifuge, washed with water three times and finally was rinsed with water.
dried in air. Four different concentrations (0.1, 0.05, 0.01 and
0.001 M LaPO ) were prepared using the sol obtained
4
by Method D, reaction 6. The cleaned sapphire fibres
2.1.3. Method C were dip coated 20 times with drying (30 s) using an
The direct reaction between lanthanum oxide and infrared lamp between each coating cycle. The coated
phosphoric acid is a clean reaction with no by-prod- fibres were calcined at 700 C and 1100 C for 2 h. The
ucts other than water. coated fibre surfaces were examined using polarized
light microscopy and SEM.
La O #2H PO " 2LaPO #3H O (5)
2 3 3 4 4 2
Due to the hygroscopic nature of lanthanum oxide
[27], the starting powder was calcined at 700 C for 3. Results and discussion
2 h. The La O powder was slowly added to 85 wt % It was found from the current study that LaPO can
2 3 4
H PO . The reaction was vigorous and exothermic. be synthesized through four chemical reaction routes
3 4
Large precipitates were immediately formed at the described in the experimental section. However,
reaction site, and then gradually broke into fine par- methods A and B are abandoned at the early stage of
ticles. The sol was extremely dense and stable and did this work due to the formation of HCl as a by-product
not separate for a long period of time. If the sol and the cumbersome multiple step reactions,
concentration was less than 0.1 M, visible separation respectively.
between the sol and water was observed. The sol was The DTA/TGA curves of the room temperature
stirred for 30 min and then separated by centrifuging. dried LaPO sol, prepared through reaction 6 are
4
Traces of excess H PO were removed by washing shown in Fig. 1. Most of the weight loss takes place
3 4
samples three times with water. below 300 C and no further weight loss was observed
between 400 and 1100 C. More weight loss was ob-
served between 1100 and 1200 C, which corresponds
2.1.4. Method D to an endothermic peak in TGA curve. The exother-
The synthesis of LaPO by the reaction of La(OH) mic peak around 400 C corresponds to the crystalli-
4 3
and H PO has been studied and is reproducible zation of LaPO . The XRD pattern of the sol calcined
3 4 4
at 700 C for 2 h is shown in Fig. 2. All the peaks in the
La(OH) #H PO " LaPO #3H O (6)
3 3 4 4 2 diffraction pattern matched with those of LaPO
4
An approximately 20 wt % H PO solution was used (JCPDS file d32-0493). The XRD pattern of the
3 4
to prepare a lanthanum phosphate sol. The reaction LaPO sol calcined at 1100 C for 2 h is similar to
4
did not complete within several hours when the con- Fig. 2, except the peaks became much sharper. Despite
centration of phosphoric acid used was less than the weight loss between 1100 to 1200 C, the XRD
10 wt %. The reaction was initiated by slowly adding pattern (Fig. 3) of the sol after the calcination at
La(OH) powder into a 20 wt % H PO solution. The 1200 C for 2 h show only peaks for LaPO . The
3 3 4 4
3864
t " 0.9k/MBcosh N, where t is particle size, k is wave-
B
length of X-ray used (k " 0.154 nm), B is the line
width to half the maximum intensity, in radians, and
h is a half of 2h in degrees. The calculated particle
B
sizes after 700 and 1100 C calcinations were about
15 nm and 46 nm, respectively.
The molar ratio between lanthanum and phosphate
during the reaction is important in terms of the forma-
tion of final products. As listed in reactions 6 and 7,
the La to P ratio has to be 1:1 in order to achieve
LaPO as a final product. When the reactant molar
4
ratio was 1:3, LaP O (JCPDS d33-717) was the
3 9
only crystalline phase observed by powder XRD after
700 C calcination (Fig. 4). When the molar ratio was
adjusted between 1:1 and 1:3, mixtures of LaPO and
4
La PO were observed. This effect was also observed
3 9
in the fibre coatings.
Four concentrations of the LaPO sols were pre-
4
pared (0.1, 0.05, 0.01 and 0.001 M) for fibre coating
experiments. The concentration variation was two or-
Figure 1 DTA/TGA curves of the dried LaPO sol at 1200 C. ders of magnitude, which provides a wide range of sol
4
concentration for good fibre coatings. The coating
was not easily described by light microscopy on the
fibres coated with the 0.001 M LaPO sol. The coating
4
was too thick and showed signs of cracking when the
0.1 M LaPO sol was used. Reasonably good coatings
4
were obtained for the 0.01 and 0.05 M LaPO sols.
4
SEM micrographs of the fractured sapphire fibre (with
the 0.01 M sol coating) are shown in Fig. 5. The top
micrograph (Fig. 5a) shows secondary electron image
(SEI) and back-scattered electron (BSE) images of the
fractured fibre. Since the fibre and the coating have
vastly different average atomic weights, the coating
was easily detectable by BSE imaging (Fig. 5a, right).
The higher magnification micrograph (Fig. 5b) of the
periphery of the fibre shows the coating with a uni-
form thickness of about 60 nm. For the same dip
Figure 2 XRD pattern of LaPO after 700 C calcination, below is coating and heat treatment, the coating thickness in-
4
the JCPDS file. creased to &150 nm when the 0.05 M LaPO sol was
4
used (Fig. 6).
As in the case of the monolithic powder, the coat-
ings on the sapphire fibres were crystalline LaPO
4
after heat treatment at 1100 C for 2 h. The XRD
pattern of the coated fibre surface is shown in Fig. 7.
A strong (1 1 0) reflection of corundum is shown along
with the LaPO diffraction peaks. The intensities
4

Figure 3 XRD pattern of LaPO sol after 1200 C calcination.


4

reasons for the weight loss and the endothermic reac-


tion between 1100 to 1200 C are not clear at this
moment.
The crystallite sizes of the calcined sols were
calculated according to the Scherrer formula [32], Figure 4 XRD pattern of LaP O sol after 700 C calcination.
3 9
3865
Figure 6 SEM photomicrograph of LaPO sol (0.05 M) dip-coated
4
sapphire fibre after 1100 C calcination.

Figure 5 SEM photomicrograph of LaPO sol (0.01 M) dip-coated


4
sapphire fibre after 1100 C calcination.

Figure 7 XRD pattern of LaPO sol (0.01 M) coated sapphire fibre


4
after 1100 C calcination. The strongest peak is (1 1 0) reflection of
corundum.
match reasonably well with the powder XRD pattern,
which implies that the LaPO crystals in the coating
4
are not strongly textured.
treatment. The endothermic reaction accompanied by
weight loss between 1100 to 1200 C can not be ex-
4. Conclusions plained at present. The calculated crystallite sizes of
Various routes to synthesize lanthanum phosphate the sols after 700 and 1100 C calcinations were about
sols for potential CMC interlayer coatings were 15 and 46 nm, respectively.
studied. It was relatively simple to synthesize Coating experiments on sapphire fibres were carried
stoichiometric LaPO sol by the reaction of lan- out using various sol concentrations. Based upon the
4
thanum oxide or lanthanum hydroxide and phos- dip coating technique used in this study, the average
phoric acid, with water as the only by-product. It was coating thicknesses were 60 and 150 nm for the sol
found that any excess phosphoric acid over the concentrations of 0.01 and 0.05 M LaPO , respectively.
4
stoichiometric reaction led to the formation of an- The XRD analysis of the coated fibre surface con-
other lanthanum phosphate (LaP O ) phase, along firmed that the coating was LaPO .
3 9 4
with LaPO . The as-reacted sol could be diluted to
4
a wide range of concentrations and the resulting sols
were stable. Acknowledgements
The crystallization behaviour of the LaPO sol was The authors thank Drs John Woodhouse and Peter
4
studied using DTA/TGA. The crystallization of Brown and Mr Marlin Cook for their help in Hitachi
LaPO started around 400 C and well-developed 900 high resolution SEM, Leica 360FE, and DTA/
4
crystals were observed by XRD after a 700 C, 2 h heat TGA analysis, respectively. One of the authors (PC)
3866
acknowledges AFOSR for the financial support while Conference), 1977, edited by G. J. McCarthy, J. Gregory and
he was working for 10 weeks as summer faculty and J. J. Rhyne (Plenum, New York), p. 209.
graduate student programme at WPAFB, OH. 18. M. KIZ ILY AL L I, J. Pure Appl. Sci. 8 (1975) 179.
19. I. V . T AN A N AE V , V. P. RE P KO , V . P . O R LO VS K II, G. M.
SAF R O NO V, K H . M . K U R BA N OV and A. G.
KH AI M O VA , Akad. Nauk SSSR, Neorg, Mater. 7 (1971)
References 1769.
1. R . J. KE RA N S, R . S . HA Y , N . J . PA GA N O and T. A . P AR - 20. E. G . T S ELE B RO V SK AY A, B. F. D Z H UR IN S KII , G . V .
TH ASA RA TH Y , Amer. Ceram. Soc. Bull. 68 (1989) 429. LYS A N OV A and M . G . K O M O VA , Neorg. Khim. 36 (1991)
2. A . G . EV AN S and D . B. M A RS H A LL , Acta Metall. 37 (1989) 2467.
2567. 21. Y . YU , M . L I , N . H U , uji Cailiao Xuebao 4 (1989) 2532.
3. J. J. BR EN N AN , in Tailoring multiphase and composite 22. R . C. M OO N E Y, Acta Crystallogr. 3 (1950) 337.
ceramics, edited by R. E. Tressler and G. L. Messing 23. V . P. O R LO V S K II , H . S C HA FE R, V . P. R EP K O, G. M.
(Pantano, and Newnham, Plenum, NY, 1986) p. 549. SAF R O NO V and I. V . TAN A N AE V, Inorg. Chem. 7 (1971)
4. T. M A H , M. G . M E N DI RA TT A, A. P. K A TZ and K . S. 860.
MA Z D IY AS N I, Amer. Ceram. Soc. Bull. 66 (1987) 304. 24. M. IV A NO V , Sov. Phys. Crystallogr. (Engl. Transl.) 13 (1969)
5. R . F. C OO P ER and K . C H YU N G, J. Mater. Sci. 22 (1989) 786.
3148. 25. E. J A RO SE WI C H , L. A . B O AT NE R, Geostand. Newsl. 15
6. J. L LO RC A and R . N. SI N G H, J. Amer. Ceram. Soc. 74 (1991) (1991) 397.
2882. 26. J.- C . G . B U N ZL I, E. M O RE T and J. -R . Y ER S I N, Helv.
7. P. E . D . M OR G AN and D . B. M AR S H AL L, Mater. Sci. Eng. Chim. Acta 61 (1978) 762.
A162 (1993) 15. 27. R . R O Y and H . A. Mc K IN S T RY , Acta Crystallogr. 6 (1953)
8. M. K. C IN I BU L K , Ceram. Eng. Sci. Proc. 15 (1994) 721. 365.
9. T. M A H , K . K EL LE R , T. A. PA RTH A S AR A TH Y and 28. I. BO TT O , J. Appl. Crystallogr. 12 (1979) 257.
J. G UT H, ibid. 12 (1991) 1802. 29. A . M A K I SH I MA , H . K UB O , K . W A D A, Y . K I TAM I and T.
10. J. B . D A VIS , J. P. A. LO F VA N D ER and A. G. E VA N S, SH I MOH I RA , J. Amer. Ceram. Soc. 69 (1986) 127.
J. Amer. Ceram. Soc. 76 (1993) 1249. 30. S. P. M U KH E RJE E , in Ultrastructure processing of ceram-
11. K. CH Y UN G and S . B . D AW E S , Mater. Sci. Eng. A162 (1993) ics, glass, and composites, edited by L. L. Hench and D. R.
27. Ulrich (John Wiley & Sons, New York, 1984) p. 178.
12. R . S . H AY , T . M AH and C . CO O K E, Ceram. Eng. Sci. Proc. 15 31. K. O K AD A and O. N O ZO MU , in Ceramic transactions, Vol.
(1994) 760. 6, Mullite and mullite composites, edited by S. Somiya, R. F.
13. R . S . H AY and E. E. H E RM E S , ibid. 11 (1990) 1526. Davis & J. A. Pask (American Ceramics Society, Westville,
14. Idem., US Patent No. 5, 217, 533 (1993). OH, 1990) p. 425.
15. M. I N O UE , T. N A KA M UR A, H . O T SU , H . K O M I N AM I 32. B. D . C UL LITY , Elements of X-ray diffraction (Addison-
and T. I N U I, Nippon Kagaku Kaishi 5 (1993) 612. Wesley Pub. Co., Inc. Reading, MA, 1956).
16. A . N . D H AS and K. C . P ATI L, J. Alloys Compd. 202 (1993)
137.
17. M. K I ZIL Y ALL I and A . J. E . W E LCH , in Rare earths mod. Received 22 June 1995
sci. technol. (Rare Earth in Modern Science and Technology and accepted 13 November 1996

3867