J O U R N A L O F M A T E R I A L S S C I E N C E 28 (1993) 3 7 5 4 - 3 7 6 0
Influence of production conditions on the densification of zirconia powders obtained from acetates
M. D E S C E M O N D , C. B R O D H A G , F. T H E V E N O T
Ecole Nationale SupOrieure des Mines de Saint Etienne, 158 Cours FaurieL 42023 Saint Etienne Cedex 2, France
B. D U R A N D , M. R O U B I N
Laboratoire de Chimie MinOrale 3, URA CNRS no. 116, ISIDT, Universit~ Claude Bernard Lyon I, 43 Boulevard du 11 November 1918, 69622 Villeurbanne Cedex, France
A. S A M D I
Faculte des Sciences Ain Chok, BP 5366, Maarif, Casablanca, Morocco
Tetragonal yttrium-stabilized zirconia samples were obtained by pyrolysis of amorphous acetates containing both metals, at 950~ in air. The dilatometric behaviour of samples
indicates a dedensification when pyrolysis is carried out with a non-linear heating rate causing an absence of fine open porosity. A destabilization when cooling is observed for samples exhibiting a wide granulometry with agglomerate sizes reaching 100 gm. Even when neither one, nor the other phenomenon is noticed, an important aggregation state of the starting powders limits the densification and leads to poor sintered densities.
1. I n t r o d u c t i o n
According to former studies, finely grained pure zirconia may be prepared by pyrolysis of either crys- tallized [1] or amorphous [2] zirconium acetates, whereas yttrium-stabilized zirconia can be obtained only by pyrolysis of amorphous acetates containing both yttrium and zirconium [3], the pyrolysis of crystallized acetates leading to mixtures of zirconia and yttrium oxide [4]. This paper reports a study of the densification of yttrium-stabilized zirconia formed by pyrolysis of amorphous acetates, synthesized fol- lowing the previously proposed procedure [3].
2. Preparation of zirconia p o w d e r s All samples were prepared from commercial zirconyl acetate solutions containing 300 g 1-1 ZrO/(Hoechst or Magnesium Elektron) to which the required amount of yttrium was added in the form of yttrium acetate dissolved in a 1/1 water-ethanoic acid solu- tion, in order to obtain yttrium-stabilized zirconia with a ratio YzO3/(Y203 + ZrO2)= 2.5-3 or 3.5 mol%. The concentrated mixture was then diluted and neutralized to pH 7 with an ammonia solution.
The amorphous acetate was finally obtained, with one exception, by evaporating solutions on a sand bath at 90 ~ According to the dilution and the neutralization procedure, three groups of samples may be distingui- shed (Fig. 1).
It was chosen to carry out pyrolysis at 950~
because an EPR study of amorphous zirconium acet- ates showed that the paramagnetic signal induced by the presence of carbon residues in zirconia disap- peared between 850 and 950 ~ [2].
3. P o w d e r s c h a r a c t e r i z a t i o n 3.1. Granulometric and morphological
characterization
The size of agglomerates was determined with alco- holic suspensions using a laser granulometer. The cumulated frequency curves (Fig. 2) indicate that Samples 4, 5, 6 and 7, exhibiting median diameters of 20 gm, are coarser grained powders than Samples 1, 2, 3, 8 and 9. Sample 8 is the finest powder.
The morphology of powders was also investigated by scanning electron microscopy, dry powders being deposited on a support without preliminary liquid dispersion. Samples 1, 2, 3 and 8 show a wide granu- lometry with a mean agglomerates size close to several tens of micrometres (Fig. 3a). Either very small par- ticles (< 1 gm) or very large agglomerates (up to 100 lam) are observed particularly for Sample 8. These large agglomerates do not appear on the granule- metric curves except for Sample 1. They are probably broken by the liquid dispersion. Samples 4, 5, 6, 7 and 9 present a narrow granulometry with a mean agglom- erates size close to several tens of micrometres in accordance with granulometric analysis (Fig. 3b).
3754 0022 2461 9 1993 Chapman & Hall
Zirconyl acetate [ZrO2] = 300g L-1
I
Dilution in ethanoic acid until [ZrO2] = 22g t-1
I
Neutralization up to p H = 7 with 1 M ammonia solution
I
Evaporatio n at 90~
!
I I
Pyrolysis Linear heating rate
9501' I
L)
k U I
0 #(min) 380
up to 950~
N on linear heating rate
/ j
.,J ',
i I
0 t ( m i n ) 380
Yttrium acetate in
1/1 waterlethanoic acid solution
Samples 1~ 2 and 3 Samples 4, 5 and 6:
3 tool % Y203 2.5-3-3.5 mol % Y203
F i r s t g r o u p
Dilution with dei?nized water up to
Neutralization with concentrated ammonia up to pH= 7
formation of a dense g e l
I
Filtering, washing, drying at 90 ~ C
I
Pyrolysis up to 950 ~ C Non-linear heating rate
I
_ Sample 7 : 3 tool % Y203
S e c o n d g r o u p
[ZrO2] = 35g !-1 (second group) [ZrO2] = 22~ t-1 (third group)
Neutralizati )n up to pH= 7 by I
slow additicn of 1 M ammonia +vigourous stirring
I
Evapoiation at 90~
I
Pyrolys s up to 950~ C Linea ! heating rate
[ Samples 8 and_99 : 3 mol % Y203 I J
T h i r d a r o u o
Figure 1 Zirconia sample preparation procedure. Samples l, 2 and 3 differ in the starting acetate solution. Samples 8 and 9 were prepared by the same procedure at two different time periods.
3.2. Porosity m e a s u r e m e n t s
The porosity of powders was measured on pellets, prepared by uniaxial pressing at 400 MPa, by mercury penetration up to a pressure of 200 MPa. Porous volumes were always in the vicinity of 120 mm 3 g-2, but powders could be separated into two groups according to the pore-size distribution and the specific surface area.
The first group, Samples 1, 2 and 3, was character- ized by specific surface areas close to 5 m 2 g - t and by bimodal size distributions (Table I, Fig. 4). The largest
part of the porous volume was constituted by pores, the size of which is in :luded in the range 0.02-2.50 lain with a maximum at 0 90 gin. The other part was made up of very narrow pores with sizes in the range 4.0-7.5 nm.
The second group, Samples 4, 5, 6 and 7, was characterized by lower specific surface areas, principally for Samples 4, 5 and 6, and regular size distribution (Table I], Fig. 4). The porous volume is constituted by larger pores, the size of which is in the range 0.2-7.5 gm with a maximum frequently ex-
3755
T A B L E I Porosity characterization of Samples 1, 2 and 3
Sample Total porous Specific surface d > 0.1 gm
volume area
(mm 3 g - 1) (m 2 g - 1) Porous volume (mm 3 g - 1)
d < 0 . 1 pm
Pore size Peak Porous Pore size
(pm) maximum volume (10 -3 gm)
(pm) (ram 3 g 1)
1 115 4.2 106
2 127 4.6 120
3 123 6.8 108
0.02 2.50 0.7 9 4.00 7.50
0.02 2.50 0.6 7 4.00-6.00
0.02 2.00 0.9 14 4.00-9.00
cs E
~ J
100 8O 60 40 20
10 z
(o}
\',, k'\,
\ ',,
\ ",
\ ",,
\ ",
\ \",\
\ ' , ,
\ "%
.. "-..
9
, %.
I I ! ! ! ! = ! 1
101
Equivalent spherical diameter (p.m)
100
100
"a 80
o .
6O
E
-~ 40
=~ 2O d
0
10 Z 101
(hi Equivalent spherical diameter (p.rn)
g 80
w
~- 60
E z-- 40
"5
e 20 g
10 o
100
\ \..
\ "...
\ " . .
0 # - . ~. : I I-.4 .-~ . , ~ ~ : I I
10 z 101 10 0
(c) Equivalent spherical diameter (t4m)
Figure 2 Agglomerate sizes of the different samples; cumulative frequency curves. (a) ( ) 1, (---) 2, (- - ) 3; (b) (---) 4, ( ) 5, ( - - ) 6 ; ( c ) ( ) 7 , ( - - - ) 8 , ( - - ~ 9 .
3756
Figure 3 Electron micrographs of powders (a) 1 a n d (b) 4.
ceeding 2 ~tm. For Samples 4, 5 and 6, the distribution is slowly shifted towards large sizes, as the amount of yttrium oxide increases from 2.5 mol % to 3.5 mol %.
Samples 8 and 9 were intermediate between the two groups. They presented only one kind of pore corres- ponding, probably, to an interaggregates porosity as for Samples 4, 5 and 6, but their specific surface area was five-fold greater (1.1 m2g 1 for Sample 9) and their pore size very much smaller (0.75 gm for Sample 9).
3.3. Radiocrystallographical analysis
For Samples 4, 5, 6, 7, 8 and 9, the X-ray diffraction patterns indicate only tetragonal zirconia, whereas for Samples 1, 2 and 3, the peaks of monoclinic zirconia are also present. The ratio of monoclinic phase ob- tained by comparing the intensities of peaks (1 1 1) and (1 1 ]-) for the monoclinic phase and (l 1 1) for the tetragonal one, according to Fillit et al.'s method [5], are given in Table III.
TABLE II Porosity characterization of Samples 4, 5, 6 and 7
Sample Total porous Specific surface
volume area
(ram 3 g t) (m 2 g-1)
Pore size (~m)
Peak maximum (~m)
4 117 0.19
5 125 0.14
6 120 0.17
7 114 0.22
0.30-3.75 2.0
0.30-5.00 2.3
0.30 7.50 3.5
0.25-5.00 2.0
TA BL E I I I X-ray analysis of zirconia samples Sample
1 2 3 4 5 6 7 8 9
Volumic fraction of 24 30 18 0 0 0 0 0 0 monoclinic phase (%)
4O.
30"
I c n
% E 20'
3=
10.
0 10-3 (o)
10-2 10-I 100
Pore radius (lain)
! i | J l , i I
101
0.00 - - -
-0.02-
- 0.04.-
~'~ - O i 0 6 - v ' {C i - -
-o.o8-
- 0 . I 0 -
-0.12-
-0.14
Tempereture (o C)
300 600 900 1200 1500
I I I I I
Figure 5 Dilatometric curves of Samples ( - - ) 1, (-- - - - ) 2 and ( )3.
50 4O 7 30
%
E - - 2 0
10-
0 , i i | 1 1 . , I
10-3 10-2
(hi
J
. . . I1- . . . I , , ,'-I,,,I
10-1 100 101
Pore rodius (Fro)
Figure 4 Pore-size distribution of Samples (a) 1 and (b) 6.
4 . D i l a t o m e t r i c a n d t h e r m o g r a v i m e t r i c b e h a v i o u r o f t h e s a m p l e s
4 . 1 . D i l a t o m e t r i c b e h a v i o u r
As for porosity measurements, the dilatometric beha- viour of samples was investigated on pellets prepared by uniaxial pressing at 400 MPa. The curves were recorded under static air with a heating rate of 150oc h -1.
The powders of the first group exhibit two different types of behaviour.
(i) Curves recorded for Samples 1, 2 and 3 are identical (Fig. 5). A linear expansion is noticed up to 900~ and then a progressive contraction (highest rates at 1160 and 1350~ which is incomplete at 1500~ The cooling curves show a destabilization phenomenon producing a volume increase due to the formation of monoclinic zirconia. The final linear shrinkages do not exceed 11.5%.
(ii) Curves recorded for Samples 4, 5 and 6 (Fig. 6) show a dedensific~tion phenomenon, beginning at l l 0 0 ~ and occur:ing up to 1400~ Its intensity increases at the same time as the proportion of yttrium Oxide increases. The destabilization p h e n o m e n o n is no longer observed on :he cooling curves. The final linear shrinkages fall to 7.5%.
The powder of Lhe second group (Sample 7) is characterized by a dilatometric curve (Fig. 7) very similar to that of Sample 5, the final linear shrinkage being the same.
The dilatometric curves recorded for powders of the third group (SampLes 8 and 9) exhibit neither de- stabilization, nor dedensification (Fig. 8). The densification arises more rapidly and the final linear shrinkage reaches a higher value for Sample 9 (14.5%) than for Sample 8 (10%).
Densities (green and sintered) are given in Table IV.
Values obtained aLer sintering are always very far 3 7 5 7
from the theoretical one (6.1) even for Sample 9, still characterized by an i m p o r t a n t shrinkage.
O b s e r v a t i o n of the fracture of sintered pellets re- veals, for all powders, essentially an intraaggregate
0.000 -0.015-
-0.030.
L-
-0,0~,5- -0.060-
-0.075-
TemperGture (o C)
300 600 900 1200 1500
I I I I I
\ \
t
t
Figure 6 Dilatometric curves of Samples (- - - ) 4, ( - - - ) 5, and ( )6.
sintering (Fig. 9). In Samples 4, 5, 6 and 7, the sizes of sintered aggregates, several tens of micrometres, are greater than in the other samples. M o r e o v e r , blisters and craters are formed at the surface of sintered aggregates.
4.2. P o w d e r t h e r m o g r a v i m e t r i c b e h a v i o u r The thermogravimetric curves of Samples 2, 3, 4, 6 and 8, recorded under the same conditions as the dilato- metric curves (Fig. 10), reveal two successive weight losses. The first, under 300~ shows that small a m o u n t s of water are a d s o r b e d on zirconia powders resulting from the pyrolysis of acetates at 950 ~ The second begins at 900 ~ a n d is finished at 1500 ~ for Samples 2, 3 a n d 8, whereas for Samples 4 a n d 6 it starts a b o v e 900 ~ being later as the yttrium oxide content increases, and is incomplete at 1600~ This second weight loss indicates that zirconia still contains removable species which lead to t h e f o r m a t i o n of gaseous c o m p o u n d s .
5. D i s c u s s i o n
A m o n g the nine powders, Samples 1, 2 and 3 present a
0 . 0 0 0 0.015
~'0.030-
c
o.o45-
0060-
0.075-
Temperature (~
300 600 900 1200 1500
I I I I I
/
J
Figure 7 Dilatometric curve of Sample 7.
Temperature (~
300 600 900 1200 1500
I I I I I
0.00
-0.03-
-~ -0,06-
t-
- 0.0 9
-0.12-
-0.15
\
\
f
Figure 8 Dilatometric curve of Samples (-- - --) 8 and ( ) 9.
TABLE IV Dilatometric behaviour of zirconia samples
Sample Green density Sintering Sintered Shrinkage Expansion
temperature density (%) due to the
(~ phase change
(%)
1 3.28 1500 4.49 11.5 0.5
2 3.2 1500 4.45 12 0.4
3 3.27 1500 4.7 12.5 0.6
4 3.33 1500 4.1 8.3
5 3.32 1550 3.97 7.5 -
6 3.38 1550 4 7.5 -
7 3.38 1500 4.09 7.8 -
8 3.43 1500 4.75 10 -
9 2.86 1500 4.58 14.6
3 7 5 8
Z.O
1 . 5
o
- 1 . 0
C n
3=
0.5
f - - - 3
/ : ... 2
, f
4
/ , /
... ... i, .-~
/ f f . . . j / / " " /
." ... 2 7 " Z ] ] Z - _ . . .
0.0
0 200 400 600 800 1000 1200 1400 1600 Temper0ture (~
Figure 10 Thermogravimetric behaviour of Samples 2, 3, 4, 5, 6 and 8.
Figure 9 Scanning electron micrographs of a fracture of sintered Samples (a) 1, (b, c) 4 and (d) 9.
destabilization during cooling, Samples 4, 5, 6 and 7 a dedensification during heating and Samples 8 and 9 neither one nor the other phenomenon. It is noticeable that the dedensification occurs only for samples,
which were pyrolysed in the muffle furnace with a non- linear heating rate (faster heating at low temperatures, slowing down when approaching 950~ Such a treatment leads to powders characterized, particularly for Samples 4, 5 and 6, by an interaggregate porosity with wide pores (2-3 ~tm). The porous volume is very badly distributed in the bulk of the aggregates, leading to specific surface areas lower than 0.2 m z g 1 and to relatively high" green densities. The aggregates seems to be more and more dense and large as the amount of yttrium oxide increases. The thermogravimetric beha- viour shows that, for these samples the gaseous release shown by the second weight loss appears at higher temperatures than for the other powders. So, it is thought that, owing to the nature of the porosity, the gaseous compound formed locally during heating is trapped inside the sintered aggregates, inducing first an inflation of the sample explaining the dedensific- ation on the dilatometric curves and the presence of blisters on fractures of sintered solids. When the pre- ssure becomes high enough, the blisters burst prod- ucing craters, thus allowing gas release.
The pyrolysis of the other samples, performed with a 150 ~ h - 1 linear heating rate, gives powders charac- terized, especially for Samples 1, 2 and 3, by larger specific surface areas and bimodal pore distributions, including wide pores corresponding to the interagg- regate porosity, and narrower pores due to an open intraaggregate porosity. This fine porosity, well dis- persed in the bulk of the aggregates allows a regular 3759
and earlier elimination of the gaseous compounds formed during sintering, avoiding a dedensification.
As for the destabilization, this is observed for Sam- ples 1, 2 and 3 prepared in the same way as Samples 4, 5 and 6, by at first diluting the commercial starting concentrated zirconyl acetate solution in pure acetic acid, but differing only in the pyrolysis treatment. This no longer happens for Samples 8 and 9 synthesized by diluting zirconyl acetate in water and calcining the acetate powder with a 150 ~ h-1 linear heating rate, as for Samples 1, 2 and 3. Thus the dilution in acetic acid is insufficient to explain by itself the destabil- ization, and the pyrolysis with a non-linear heating rate no more. Both factors must occur together. It is noticeable that the presence of monoclinic zirconia was indicated by X-ray diffraction only in these three samples. This could be explained by an inhomogen- eous repartition of yttrium in the bulk of zirconia. The lower stabilization of powders could also result from crystallites exhibiting a size greater than a critical value, d c, in the range 0.6-1.2/am according to Leriche, et al. [6]. They are liable to induce the growth of large grains during sintering which might undergo the tetragonal to monoclinic transformation during cooling causing, in agreement with Smith and Baum- ard [7"], an expansion.
6. Conclusion
Using the acetate route for yttrium-stabilized zirconia powder synthesis, it is necessary to carry out the pyrolysis with a sufficiently slow linear heating rate, especially at the beginning of the decomposition, in
order to obtain a fine porosity allowing regular gas- eous release and avoiding dedensification. Moreover, it seems preferable to dilute concentrated starting zirconyl acetate solution in water, especially because this procedure reduces the amount of ammonia solu- tion necessary to raise the pH to 7. Under these conditions, a regular shrinkage is noticed during sintering but only poor density values are obtained owing to an absence of interaggregate sintering. Be- fore considering powder milling, the present studies of the acetate production and principally the pyrolysis procedure, should be further developed with the purpose of decreasing the aggregation state of the powders.
7. R e f e r e n c e s
1. A. SAMDI, T. G R O L L I E R BARON, B. D U R A N D and M. R O U B I N , Ann. Chim. Sci. Mater. 13 (1988) 171.
2. ldem, ibid. 13 (1988) 471.
3. ldem, ibid. 13 (1988) 517.
4. ldem, ibid. 13 (1988) 483.
5. R. F I L L I T , P. H O M E R I N , J. S C H A F F E R , H. BRUYAS and F. T H E V E N O T , J. Mater. Sci. 22 (1987) 3566.
6. A. L E R I C H E , G. M O O R T G A T , F. CAMBIER, P. H O M E R I N , F. T H E V E N O T , G. G R A N G E and G. F A N T O Z Z I , in "Advances in Ceramics", Vol. 24 edited by S.
Somiya, N. Y a m a m o t a and H. Yanagida, "Science and Techno- logy of Zirconia III", (American Ceramic Society, 1988) p. 1033.
7. A. S M I T H and J. F. B A U M A R D , Amer. Ceram. Soc. Bull. 66 (1987) 1144.
Received 2 January
and accepted 29 October 1992
3760
