Langmuir 1988,4, 31-37 y-Fe203,respectively. After iron particles are removed from toluene, the Fe content remains essentially unchanged for at least 1month, indicating that the thin oxide layer prevents further oxidation. On reduction of uncoated hematite to iron the particle size did not change while the specific surface area decreased, showing less porosity, despite the loss of oxygen from the crystal lattice. This effect must be obviously due to a change in the crystal structure, as substantiated by an increase in the internal crystallite size (as seen from Tables I and 11). I t has been reported that silica coating gives iron particles a high coercivity by inhibiting their sintering.17*’* Tables I and I1 show that the size of these particles is indeed quite close to that of the silica-coated precursor a-Fez03. However, the reduction causes a decrease in the crystallite size of iron and an increase in the specific surface area. When silica coating amounted to more than 3 wt % Si02,the resulting iron particles were amorphous. These results suggest that the described treatment not only inhibits sintering of iron particles but also crystallization of metal iron. The heterogeneous coverage of hematite powder with Co304is illustrated in Figure 7b. Since no independent cobalt oxide particles are detected in the course of the interaction with cobalt acetate solution, it appears that the deposition occurs by direct crystal growth on the a-Fe203 surface. Furthermore, the adhered Co304cannot be removed by application of ultrasonic energy on particles suspended in water. Both metallic iron and cobalt are
31
produced by interaction with hydrogen at temperature lower than needed for pure and silica-coated a-Fe203, respectively. This finding is in agreement with previously reported data showing that adsorbed Co and Ag promote reduction of a-Fe203to iron.8 The coercivity values of all samples listed in Table I1 are larger than the maximum calculated values of 500 Oe for single-domain iron particles,l which may be due to a large shape anisotropy contribution. Our data indicate that the higher coercivity is related to small crystallite size. Sample V is an exception, in which case the crystallite dimension is less than the calculated single-domain size of 24 nm.19 Similar results have been observed with iron particles obtained from ferric oxyhydroxides.m*21Sample 111 (Table 11),the crystallites of which are of nearly single-domain size, exhibits the highest coercivity. The saturation magnetization of iron decreases generally by oxidation, owing to the formation of iron oxides of low u8,22 Values of a, in Table I1 are smaller than that of pure iron (210 emu g-l) because, as discussed above, the surface of these iron particles is partially oxidized. Acknowledgment. We are greatly indebted to Louis C. Nanna (Clarkson University) and Dr. Sonia Sikorav (RhSne-Poulenc Recherches, Aubervilliers, France) for the measurements of magnetic properties.
Registry No. Fe, 7439-89-6;FenOa,1309-37-1. (19)Brown, W.F. Phys. Rev. 1957,105, 1479. (20) Suzuki. S.:Sakumoto. H.: Minenishi. J.: Omote. Y. IEEE Trans. ” Magn:, Mag17 1981, 3017. (21) Sueyoshi, T.;Tashita, K.; Hirai, S.; Kishimoto, M.; Hayashi, Y.; Amemiya, M. J. Appl. Phys. 1982,53, 2570. (22) Schnitker, W.; Rau, H. IEEE Trans. Magn., Mag-16 1980,14. I
(17)Sueyoshi, T.; Naono, H.; Kawanami, M.; Amemiya, M.; Hayama, H.IEEE Trans. Magn., Mag-20 1984,42. (18)Homola, A. M.: Lorenz, M. R.; Tilbury, D. L. IEEE Trans. Mum., Mag-22 1986,716.
,
~I
Preparation and Properties of Monodispersed Colloidal Particles of Lanthanide Compounds. 2. Cerium(1V)t Wan Peter Hsu, Lena Ronnquist, and Egon Matijevie* Department of Chemistry, Clarkson University, Potsdam, New York 13676 Received July 10, 1987. In Final Form: August 25, 1987 The preparation of colloidal dispersions containing spherical cerium(1V) oxide sols of narrow size distribution can be accomplished by hydrolytic deprotonation of cerium(1V) ions in the presence of diluted sulfuric acid. At higher sulfate ion content, rodlike particles of ceric basic sulfate are formed. It was found that the particle morphology could be altered by reactant concentrations and the nature of anions in solution. These particles show definite crystallinity by X-ray diffraction analysis yet consist of much smaller uniform subunits. The materials so obtained were characterized in terms of size distribution, electrokinetics,surface area, and optical properties. Introduction
It is highly desirable to establish a technique to generate well-defined, finely dispersed particles of “ceria” (Ce02) powder because of its unique optical properties and oxidizing power. This material finds varied applications in such things as glass polishers, decolorizers, ceramic opacifiers, phosphors, and capacitors. In addition, cerium(IV) oxide has been considered as one of the most universal Supported by the Air Force Contract F49620-85-C-0142.
0743-7463/88/2404-0031$01.50/0
catalysts used in petroleum technology, catalytic converters, etc.1>2 Numerous studies3 have dealt with the precipitation phenomena of cerium(1V) hydroxide, usually obtained by the addition of an oxidizing agent (e.g., H202)to cerium(1)Maestro, P.J. Less-Common Met. 1985, I l l , 43. (2)Encyclopedia of Chemical Technology, 2nd ed.;John Wiley: New York, 1968; Vol. 17, pp 143-168 3rd ed.; 1982;Vol. 19, pp 833-854. (3)Ryabchikov, D.I.; Ryabukhin, V. A. Analytical Chemistry of Yttrium and the Lanthanide Elements; Ann Arbor-Humphrey Sci.: Ann Arbor, MI, 1970; Chapter 3,pp 50-139.
0 1988 American Chemical Society
32 Langmuir, Vol. 4, No. 1, 1988
t
' '
I
'
""I
Hsu et al. I / /
'
Table I. Effect of the Concentrations of Reactants and of Aging Time at 90 "Con the Size of Particles Obtained from
U n " I
[ce(so4),l/
Ce(S04)2-H2S04 Solutions modal particle diameter (nm) at [HzS04]/moldm-3
IO3 mol dm-3 0.5
0.04 30 (40)"
70 (110)
0.8 1.0 1.2 1.6
40 (50) 50 (50) 60 (60)
70 (120) 80 (140) 80 (140)
0.06
0.08 100 (130) 130 (150) 170 (220) 220 (230) 170 (240)
0.10 110 (140)
140 (200) 190 (210) 150 (210)
"First number refers to aging for 12 h, while the number in parentheses is for 48 h of aging.
3m
w
5
1
"
mmmo 0
x x
0
O
x
o o a a /
x
x
x
x
/ /
0
2
/
Experimental Section
/
e/ 1
0 O
preparation of such uniform dispersions is described in this paper, and a possible mechanism of condensation reactions which yield such sols is offered. The materials so obtained were characterized in terms of their composition, surface area, and electrokinetic and optical properties.
/ I
0.02
H,SO,
,
l l 1 I l I
0.05
0.1
I
0.2
,
,
I
/
,
l
,
0.5
10
CONCENTRATION (mol dm'3)
Figure 1. Precipitation domain for solutions containing Ce(S04)2 and H2SO4 aged at 90 "C for 12 h. Symbols designating different kinds of particles: 0,spheres; 0,rods, un, rods mixed with spheres; @, a very small amount of spheres; x, no particle formation.
(111) salt solutions. As a rule, these procedures yielded bulky, amorphous precipitates of Ce(OH)4. Recently, dispersible ceria particles (about 5 nm in diameter) have been obtained by a suspension/deaggregation treatment of either precipitated Ce(OH), or commercial ceria produ c t ~ Other ~ * ~ methods involved the preparation of a ceric oxide precursor, such as cerous cerous valerate,8 or thiourea-coordinated cerium p r ~ p i o n a t ewhich ,~ were then converted to ceric oxide at temperatures above 400 "C. The morphologies of the latter solids depended on the nature of the precursors, and were mostly ill defined. In a previous publication,1° it was shown that "monodispersed" sols could be prepared by aging at elevated temperatures solutions of lanthanide(II1) salts in the presence of urea. The resulting colloidal particles consisted of spherical or elongated lanthanide(II1) basic carbonates of narrow size distribution. Cerium(1V) ion is much less basic than lanthanide(II1) ions, and its high charge causes strong hydration. The hydrated Ce4+ion is readily hydrolyzed to give polymeric ~ p e c i e s , l l -which ~ ~ are intermediates to precipitation of the (hydrous) oxide. It should then be possible to generate, in principle, uniform ceria particles in situ by kinetically controlled deprotonation of the hydrated Ce4+ions at elevated temperatures ("forced hydrolysis")1b17in the absence of urea. The method of (4)Woodhead, J. L.U.S.Patent 4 231 893, 1980. (5)Woodhead, J. L.; Raw, G. Fr. Patent 2482075,1981. (6)Brittain, H.G.;Gradeff, P. S. J.Less-Common Met. 1983,94,277. (7)Ishii, E.;Miyake, Y. Osaka Kogyo Gijutsu Shikensho Kiho 1973, 24,99. (8) Lazareva, L. S.; Ambrozhii, M. N.; Dvornikova, L. M. Zh. Neorg. Khim. 1970,15,354. (9) Sakharova, Yu. G.; Borisova, G. M.; Loginov, V. I. Zh. Neorg. Khim. 1978,23,376. (10)MatijeviE, E.;Hsu, W. P. J.Colloid Interface Sci. 1987,118,506. (11)Hardwick, T.J.; Robertson, E. Can. J. Chem. 1951,29,818. (12)Hardwick, T.J.; Robertson, E. Can. J. Chem. 1951,29,828. (13)Lundgren, G. Recl. Trav. Chirn. Pays-Bas 1956,75,585. (14)Baes, C. F.;Mesmer, R. E. The Hydrolysis of Cations; John Wiley: New York, 1976;pp 138-146.
Materials and Techniques. Cerium(1V) sulfate [Ce(S0&.4H20, E M Industries], ammonium cerium(1V) sulfate [ (NH4)4Ce(S04)4.2H20,Aldrich], and ammonium cerium(1V) nitrate [ (NH4)2Ce(N03)6,Aldrich] were of laboratory purity (>99%). Inorganic salts and acids of "Dilute-It" concentrates were used without further purification. The structure of the solids was determined by X-ray diffraction using a Philips diffractometer with Ni-filtered Cu K a radiation. The infrared spectra were obtained with a Perkin-Elmer 1430 spectrophotometer. Samples were prepared by mixing the powder with potassium bromide (1:20 by weight) in a steel die. Differential thermal analysis (DTA) with a Perkin-Elmer Model 1700 Analyzer was employed to follow the phase transformation in ceria powders. The particle shape and size were elucidated by transmission and scanning electron microscopy. The size distribution of the suspended particles was confirmed by light scattering employing the polarization-ratio method.'* This procedure also yielded the refractive index of the dispersed particles. A Coulter N4MD model particle size analyzer coupled with a size distribution processor (SDP) was used for comparison purposes. The run time to collect the autocorrelation function was set to 200 s in the SDP analysis. Sample preparation and the correct input value of refractive index of the solid were found to be critical for reliable results. The specific surface area of the dried powder was obtained by a multipoint BET method using a Quantasorb apparatus equipped with a linear flow controller. Electrokinetic measurements were made with the PenKem System 3000. The particles were dispersed in a 1.0 X mol dm-3 NaN03 solution in which the pH was adjusted with HNO, or NaOH. Sols were aged for 3-5 h then treated in an ultrasonic bath before the measurements were taken. Sol Preparation. All sols to be described were obtained by heating solutions of ceric salts for different periods of time. The concentration of these salts, the ionic strength, and the temperature were carefully controlled. One important requirement to generate uniform particles by forced hydrolysis involves homogeneous ionic solutions. Since ceric sulfate and ammonium ceric sulfate are only slightly soluble in water and in mineral acids other than sulfuric, the precipitation experiments were carried out in the presence of the latter. The stock solutions of ceric salts containing predetermined amounts of sulfuric acid were filtered through 0.2-pm pore size Nuclepore membranes. It is essential to use a freshly prepared stock solution as precipitates form on storage at room temperature. Pyrex test tubes containing diluted stock solutions, with pH varied between 0.5 and 2.0, were tightly closed with Teflon-lined (15)MatijeviE, E.Acc. Chem. Res. 1981,14,22. (16)MatijeviE, E.Annu. Rev. Mater. Sci. 1985,15,483. (17)Matijevie, E.Langmuir 1986,2,12. (18)Kerker, M. The Scattering of Light and Other Electromagnetic Radiation; Academic: New York, 1969;pp 351-373.
Monodispersed Colloidal Cerium(IC? Compounds
Langmuir, V06. 4, No. I , 1988 33
0
Figure 2. Transmission (TEM) (A) and scanning (SEM) (B)electron micrograph of CeOpparticles obtained by aging at 90 OC for 48 h a solution 1.2 X mol d d in Ce(SO,), and 8.0 x lo-, mol dm" in HBO,.
A
B
c . l
0.2$m Figure 3. TEM of precipitates obtained by aging at 90 'C for 12 h the following solutions: (A) 2.5 X 0.1 mol dm4 in H,SO,: (B) 9.0 X mol d w 3 in Ce(SO,), and 5.0 X IO-? mol dmJ in H,SO,.
caps and placed in a forced convection oven at 90 OC. After aging
for various periods of time, the tubes were quenched by cooling to room temperature. The systems were then centrifuged for 20 min at 2600 rpm, the supernatant solutions discarded, and the solids resuspended in water by using an ultrasonic bath. This washing process was repeated 4-5 times, and the dispersions so obtained were designated as Sol A. A further step in purification consisted of removing sulfate ions from the sol particles. This was accomplished by dispersingthe washed sols in 2.0 X lW3 mol dm-3NaOH, followed by treatment in an ultrasonic bath, centrifugation, and decantation of the supernatant solution. The separated solids were then redispersed in distilled water. The same repeptization process was repeated 4 times. The final powder was washed 10 times with distilled water, and the resulting dispersion was designated as Sol B. Finally, Sol C was obtained by calcining at 600 OC for 2 h the particles separated from Sol B and redispersed in aqueoua media.
lo* mol d w 3in Ce(SO,), and
Results Particle Formation. A systematic study of the effect of concentrations of reacting solutions on the natures of precipitated solids has resulted in a complete domain shown in Figure 1;data refer to systems aged at 90 "C for 12 h. Within the dashed lines the solid phase is formed by homogeneous precipitation due to hydrolysis. Above the upper dashed line solids appear on mixing the reacting components at r m m temperature, while below the lower dashed line the solutions remain clear even after prolonged aging at temperatures up to 100 O C . Spherical particles of narrow size distribution were formed at lower concentrations of ceric sulfate and H$04 (see Figure 1, circles). Depending on acidity and aging time, the particle diameter ranged from 30 to 240 nm
34
Langmuir, Vol. 4, No. I , I988
Hsu et al.
B
i
C
57z A
D
b
-
0.5,um
Figure 4. Electron micrographs of rodlike particles obtained by aging at 90 “C for 12 h the following solutions: (A, B) 2.5 X 10-9 mol dm4 in Ce(SO,),, 1.0 x 10.’ mol dm-3 in HBO,, and 0.4 mol dm-, in Na,SO,; (C, D) 4.5 x mol dm” in Ce(SO,)p,2.5 x mol dm” in H,SO,, and 0.45 mol dmJ in Na2S04.
(Table I). As a rule, with increasing cerium and acid concentrations the particles became larger. A typical example of such dispersions is illustrated in the transmiasion and scanning electron micrographs of Figure 2. In solutions of higher pH and ceric sulfate concentrations,irregular rodlike or cylindrical particles were formed designated by rectangles in Figure 1. At intermediate concentrations of H,S04 and Ce(S04),, spherical CeOz particles of - 4 0 nm in diameter coexist with the rods (Figure 3A). In the presence of less acid the spheres adhere to the rods (Figure 3B). Experiments were also carried out to investigate the effect of the addition of sodium salts on the precipitate formation in solutions of ceric sulfate a t varied pH. For the dilute solution of [Ce(S04),I = 4.5 X 1W mol dm-3 and [H2SOl] = 5.0 X lo-, mol dm-3, the addition of Na2S04, NaC1, and NaC104 in concentrations ranging from 0.01to 0.10 mol dm-3 had little effect on the average particle size, although agglomeration did occur. However, NaNO, caused a decrease in the particle diameter from 120 to -80 nm when 0.01 mol dm-3 of this electrolyte was added.
Over a narrow range of intermediate cerium(1V) ion concentrations, the addition of Na2S0, generated either single rods (Figure 4, A and B)or composite rods (Figure 4, C and D). The production of these particles in uniform state is extremely sensitive to the working parameters (pH and temperature). A t still higher concentrations of ceric sulfate (6 X to 1.5 X 10.’ mol dm”), the addition of Na2S04caused varied precipitation phenomena as illustrated in the electron micrograph of Figure 5A. A t sulfate concentrations exceeding 0.6 mol d m 3 no precipitation occurred, while irregular clusters were formed over the range 0.1-0.5 mol dm-3 Na2S04. At [NaaO,] < -0.05 mol dm”, rodlike particles separated out. The same aging procedure, applied to ammonium ceric sulfate in the presence of H,SO, solutions, yielded on heating at 90 ‘C from 6 to 48 h small spheres (0.06 pm), rods, or their mixtures, 88 summarized in Table 11. It should he noted that the diameters of spheres are smaller than the diameters of particles obtained from comparable ceric sulfate systems, and their size cannot he altered
Langrnuir, Vol. 4, No. I , I 9 8 8 35
Monodispersed Colloidal Cerium(1VJ Compounds
D
.P P
c (
0.5..urn mol dm-3 in Ce(SO,),, 3.2 Figure 5. T E M of precipitates obtained by agin at 90 "C for 12 h the following solutions: (A) 7.1 X X 10-2moldm-l in H,SO,, and 8.0 X IO-, mol dm- in NaaO,, pH 1.4; (BJ1.5 X 10" mol d d in (NH,),Ce(NO),, 6.4 X 10.' mol d w 9 in H,SO,. and 1.6 X mol dm" in Na,SO., pH 1.2; (C) 1.5 x IO-, mol d d in (NHJ2Ce(NO&. 6.4 x IO-?mol dm-' in H,SO,, and 1.6 10'' mol dm" in Na2S04,pH 1.1; ID) TEM of a magnified particle aggregate selected from (C).
3.
x
significantly by adjusting the reactant concentrations. Limited experiments have been done by using ammonium ceric nitrate as a starting reactant. However, either no solids or some geUie precipitates were formed. Rodlike mixed particles of other morphologies were formed only when Na2SOl was added (Figure 5, B-D). Particle Characterization. The X-ray diffraction data of the collected spherical particles displayed four major lines (d = 3.12, 2.72, 1.92, and 1.63) characteristic of face-centered cubic CeO,, confirming particle crystallinity. Differential thermal analysis of a powder of Sol A below loo0 "C shows only one weak endothermic peak at 750 OC due to the presence of SO2-ion as contaminant. The infrared spectra of the powder revealed two broad hands centered at 410 and 500 cm-' identified before as cerium(IV)oxide's and one u ( S - 0 ) shoulder at 1000-1200 em-'. The last band was not detectahle in the deionized sol (Sol (19) McDevitt, N. T.;Baux, W.L. Speetmchim. Acto 1964,20,799.
B). No traces of ,9012- could be found in the supernatant solution of this sol, treated with 2 mol dm-3 "OB. The infrared analysis of the rodlike particles (Figure 4) indicated that the composition of these solids is cerium(W basic sulfate. The BET analysis of a sample of spherical particles with a modal diameter of 0.10 pm dried at 110 "C for 12 b yielded a specific surface area of 19.3 m2gl, which is larger than the calculated value of 8.3 m2 g-l, assuming smooth surface and a densityMof 7.22 g ~ m - ~The . TEM indeed indicates a rough surface of such particles. Figure 6 (top) gives the size distribution of CeO, (Sol A) particles shown in Figure 2 as determined by light scattering (436- and 546-nm wavelengths) using the polarization ratio method as compared to the histogram ob(20) Encyclopedio of Minerals; Roberta,W.L., FIapp, G.R, Weber, J., Eds.; Van Nostrand Reinhold New York, 1974; p 115. (21) Espemheid, W. F.; Kerker, M.; Matijevif, E. J. P h p . Chem.
1964,68,3093.
36 Langmuir, Vol. 4, No. 1, 1988
Hsu et al. CERIUM ( I V ) OXIDE NoNO, 0001 mol dm-3 so1
I J
6O I 2 l A
O08i( ,
21
1
W 22
23
Figure 6. Size distribution analysis of the CeOz sol shown in Figure 2. Top: A comparison of the histogram obtained by electron microscopy with the distributions from light scattering at two wavelengths (436 and 546 nm). Bottom: Contour error map for the light-scattering data at 436 nm. The contour levels (A) correspond to the root square mean deviation between the calculated and measured values. A unique fit is found at aM = 2.24 and uo = 0.11. Table 11. Results of Aging Ammonium Cerium(1V) Sulfate Solutions at 90 O C for 12 h [HzSO,l/mol dm-3 [Ce(IV)l/ lo4 mol dm-3 solids 2 10-1 5-100 no precipitation 2.5 X 5-10 spheres (0.06 r m ) 10-100 spheres and rods 5.0 X 5-10 spheres (0.06 rm) 10-100 spheres and rods 1.0 x 10-2 8-100 spheres and rods
tained from electron micrographs. The optical size aM= rrd/X, where d is the particle diameter and X the wavelength in the medium. The width of the distribution parameter, a,, is as defined elsewhere.z1 The inner contour lines in the ~ M - u , domain in the bottom part of Figure 6 represent root mean square deviations, A, between the experimental (at 436 nm) and calculated values.18 The so-obtained modal diameter is 0.23 pm (uo = 0.11) at 436 nm and 0.22 pm (a, = 0.07) at 546 nm. The corresponding refractive indices for the solid CeOz are 2.05-0.034i at 436 nm and 2.04-0.013i at 546 nm. The size distribution of the same sol determined with a Coulter N4MD counter gave a mean diameter of 0.24 pm with a standard deviation u = 0.04 pm for the same diluted dispersion of CeOzin the presence of 3 X mol dm-3 NaOH. A suitable concentration of the sample was established by analyzing the intensity of the signal as a function of the concentration. The best reproducibility was noted in the range (1-3) X lo5 counts/s. The dispersions were treated for 3-5 min in an ultrasonic bath before the measurements were taken. Figure 7 gives the electrokinetic mobilities of CeOz particles as a function of pH. The isoelectric points (iep) for Sols A, B, and C were found to be 5.9, 6.1, and 5.2, respectively,which are lower than the previously reported values of 6.75zz and 7.6,z3respectively. (22) Matson, S.; Pugh, A. J. Soil Sci. 1934, 38, 229. (23) Ray, K. C.; Sengupta, P. K.; Roy, S. K. Indian J. Chem. 1979, I7A, 348.
-
PH Figure 7. Electrokinetic mobilities of CeOz sols as a function of pH in 1.0 X loT3mol dms NaN03. Sols: A (o), B (O), C (A).
While at pH > iep the three systems give mobilities in reasonable agreement, the values for Sol A at low pH are much lower than for the other two kinds of particles.
Discussion Cerium has two common oxidation states, cerous, Ce(III), and ceric, Ce(1V). Cerous compounds resemble the other trivalent lanthanides,"'O while ceric compounds are like those of zirconium and thorium.z Complexes of the type Ce(S04)n4-2n (n = 1 , 2 , 3 ) are present in sulfate solutions at 25 OC.12 A t 200 OC, solutions of Ce(1V) sulfate apparently result in the Ce~04(0H)4'z+and chainlike (Ce0),2n+polynuclear species.13 Anhydrous ceric sulfate heated to 200 "C is transformed by the loss of SO3 to Ce302(S04)4, and on continued heating additional SO3 is released, eventually yielding CeOFZ4 These findings show that the solids precipitated in cerium(1V) sulfate solutions vary with the experimental conditions. Indeed, in this work particles of different chemical composition and morphology are generated from solutions of the same reactants by changing their concentrations. Spherical particles prepared by forced hydrolysis consist of CeOz with sulfate ion as contaminant, which can be removed by washing. Obviously, in this case the anions are not bound in a stoichiometrically well-defined compound as is, for example, the case with alunite^^^*^^ prepared under similar conditions. The intriguing observation is that spherical particles prepared by homogeneous precipitation exhibit crystalline characteristics, as revealed by X-ray diffraction analysis. It is most unlikely that such dispersed solids could consist of single crystals. Indeed, Figure 8 clearly demonstrates the composite nature of the final particles. The latter are formed by aggregation of small spherical subunits. A similar observation was reported earlier for hydrous ferric oxide.z7 The low-angle X-ray analysis of spherical hematite particles obtained by forced hydrolysisz8indicated these to be composed of tiny globules 30 A in diameter.z9 Furthermore, spherical magnetite particles prepared by (24) Sneed, M. C.; Brasted, R. C. Comprehensioe Inorganic Chemistry; Van Nostrand: Princeton, NJ, 1955; Vol. IV, p 177. (25) MatijeviE, E.; Sapieszko, R. S.; Melville, J. B. J. Colloid Interface Sci. 1975, 50, 567. (26) Sapieszko, R. S.; Patel, R. C.; MatijeviE, E. J.Phys. Chem. 1977, 81, 1061. (27) Towe, K. M.; Bradley, W. F. J. Colloid Interface Sci. 1967, 24, 384. (28) MatijeviE, E.; Scheiner, P. J. Colloid Interface Sci. 1978,63,509. (29) Hosemann, R., private communication. (30) Sugimoto, T.; MatijeviE, E. J.Colloid Interface Sci. 1980, 74, 227.
Longmuir, Vol. 4, No. I , 1988 31
Monodispersed Colloidal Cerium(IV) Compounds
H
0.05 prn Figure 8. Formation of CeO, particles by forced hydrolysis of an acidic (4.0 X 1(r2mol dm4 HSOJ solution of Ce(SOJ2 (1.0 X l@ mol dm-? heated at 90 OC during a 6-h period.
oxidation crystallization of a ferrous hydroxide gel in the nresence of sulfate ions also consisted of much smaller brimary particles. Heller31has pointed out that the primary condition for ordered m_ _e a-t i o n of colloidal particles is their uniformity and that the second requirement is a high collision number leading to mutual fixation of colliding particles in three dimensions. It would seem that the formation of spherical CeO, particles of narrow size distribution, consisting of a large number of small subunits, is then caused by the uniformity of the Latter and the chemical surface reactions (condensation and sulfate bridging) taking place on their collision. I t is also noteworthy that CeOz dispersions of hexagonal particles (Figure 5D) consist of solids with subunits of spherical shape. The narrow size distribution of the final spheres generated by such a process requires an adequate theoretical interpretation, which is now being studied. The discrepancy between specific surface areas, determined experimentally and calculated from strictly geometric parameters, can easily be understood in view of the composite nature of the spheres. Elongated particles (Figures 3 and 4) are of different chemical composition and include sulfate ion corresponding to a basic cerium(IV)sulfate. AH such, this anion cannot be removed simply by washing. This example clearly demonstrates how a small change in the concentration of the reactants can alter particle composition and shape. The reason for these effects must be in the complex chemistry of the solute precursors which affect the solidDhase formation. However. at mesent there is no understanding of the relationships between resulting morphol@ea and the variation of cornpaition of solutions in which such solids are formed.
Electrokinetic measurements show that the presence of the sulfate ion (Sol A) has a pronounced effect on the positively charged particles (Le., at pH < iep). The decrease in the mobility as the pH is lowered is most likely due to leaching of sulfate ions from the particles in acidic media. The shift in the isoelectric point (iep) of calcined particles has been observed before. It was interpreted in t e r n of the dependence of the acid strength of surface metal hydroxide groups on the 02-/OH- ratio or on the increased bond strength due to improved The value of the refractive index of CeO, (n = 2.05) is much higher than the value of trivalent rare-earth oxides, such as Gd203(n = 1.79), as previously reported.'O Since the ion radii of Ce" and Gd3* (0.92 and 0.94 A) are comparable in size,the differences in refractive index are likely to be due to their crystalline structures. For the high opacity of ceramics, the coated particles should have high vdue of relative refractive index and a particle size of the same magnitude as the radiation wa~elength.~ The ~ monodispersed, submicron particles of CeO, powder meet these criteria and, therefore, may be a suitable opacifier for advanced ceramic
a
Acknowledgment. We are indebted to Pmfwor Maria Hepel (State University of New York, Potsdam, NY)for some X-ray analyses. Registry No. CeO,, 1306-383; Ce,O,(SO,),, 11137867-7:
ce(s0,),,13590-82-4;H z ~ ~7664-93-9, ,,
(32) ParLe, G. A. Chern. Rev. 1965,ffi. 177. (33) Park&G. A. Eouilibrium ConreDts in Natural Water System; Am.eriean Ch&nieal G i e t y : Washingtbn, D.C.,1967; pp 121-160. (34) Healy, T. W.; Fuerstenau, D. W. J. Colloid Sei. 1965,20, 376. 1.99 Kinzerv. ~ W. D.: Bowen. H.~K.: Uhlmann. D.~R. Introduction ,.., ~ ~ ~ . ~ to Ceramics, 2nd ad.; John Wiley: New'York, 1976; pp 666+72 (36) Kilbourn, B. T. J . Less-Common Met. 1985. 111, 1. ~
(31) Heller, W. In Polymer Colloids II; Fitch, R. M., Ed.; Plenum: New York, 1980; pp 153-207.
~~.
~
,