GB2208644A - Doped BaTiO3 based compositions - Google Patents

Description

DOPED BaTiOw BASED COHPOSlTIONS The present invention relates to barium titanate based compositions and, more particularly, relates to dispersible, submicron, doped barium titanate coforms with narrow particle size distributions.

Barium titanate based compositions are extensively used in the electronics industry for the production of capacitors, condensers and PTCR (positive temperature coefficient of resistance) devices. Barium titanate is particularly useful and versatile in electronic applications since its electrical properties can be substantially modified by the incorporation of additives and/or dopants. The additives which are frequently employed are HAOs compounds, where M is a diva lent cation and A is a tetravalent cation, having the BaTiOS perovskite structure.

Typical additives include the titanates, zirconates and stannates of calcium, strontium, barium and lead. Since the additive or additives have the same crystal structure as Bats03, they readily form a solid solution during calcination or sintering. In general, additives represent more than 3 mole % of the BaTiOs, based formulation.

Dopants cover a wide range of metal oxides. These, in general, represent less than 5 mole % of the total BaTiOs based formulation. The dopant or dopants employed may be completely or partially miscible in the perovskite lattice or may be immiscible in the lattice. Examples of dopants employed include the oxides of La, the lanthanides, Y, Nb, Ta, Cu, Mo, W, Mn, Fe, Co, Ni, Zn, Al, Si, Sb and Bi.

In commercial practice, barium titanate based formulations are produced either by blending the required pure titanates, zirconates, stannates and dopants or by directly producing the desired powder by a high temperature solid state reaction of an intimate mixture of the appropriate stoichiometric amounts of the oxide or oxide precursors (e.g; carbonates, hydroxides or nitrates) of barium, calcium, titanium, etc. The pure titanates, zirconates, stannates, etc are also, typically, produced by a high temperature solid phase reaction process.

The prior art processes for producing barium titanate and barium titanate based compositions by solid phase reactions are relatively simple; nevertheless, they do suffer from several disadvantages. Firstly, the milling steps serve as a source of contaminants which can adversely affect electrical properties. Secondly, compositional inhomogenieties, resulting from incomplete mixing on a microscale, can lead to the formation of undesirable phases such as barium orthotitanate, Ba2TiO, which can give rise to moisture sensitive properties. Thirdly, during calcination, substantial particle growth and interparticle sintering occur. As a consequence, the milled products consist of irregularly shaped fractured aggregates which have a wide size distribution ranging from about 0.2 to about 10 microns.Moreover, it has been established that green bodies formed from such aggregated powders with broad aggregate size distributions require elevated sintering temperatures and give sintered bodies with broad grain size distributions.

Many approaches have been developed to try to overcome the limitations of the conventional solid state reaction processes. Precipitation of either doped barium titanyl oxalate or doped barium titanyl oxalate with partial substitution of strontium or lead for barium and zirconium for titanium is taught by Gallagher et al., "Preparation of Semiconducting Titanates by Chemical Methods", 46, J.

Amer.Chem.Soc., 359 (1963); Schrey, "Effect of pH on the Chemical Preparation of Barium-Strontium T1tanate," 48, J. Amer.Cer.Soc. 401 (1965) and Vincenzini, "Chemical Preparation of Doped BaTiO3," Proceedings of the Twelfth Ins.1. Conf. Science of Ceramics, Vol.12 p.151 (1983). The oxalates are decomposed at elevated temperature to form the doped barium titanate based compositions. US Patent number 3,637,531 teaches heating a single solution of dopant, titanium compound and alkaline earth salts to form a semisolid mass that is converted to the desired titanate based product by calcination. US Patent Number 4,537,865 discloses combining hydrous oxide precipitates of Ti, Zr,Sn, or Pb and hydrous oxides of the dopants with aqueous slurries of precipitated carbonates of Ba, Sr, Ca or Mg.The solids are calcined to give the required product. Kakegawa et al., "Synthesis of Nb-doped Barium Titanate Semiconductor by a Wet-Dry Combination Technique" 4, J. Mat. Sci. Lets 1266 (1985) describe a similar synthesis procedure.

Mulder, "Preparation of BaTiOs and Other Ceramic Powders by Coprecipitation of Citrates in an Alcohol", 49, Ceramic Bulletin 990-993 (1970) prepares doped BaTiOs and BaTiOs based products by spraying an aqueous solution of citrates or formates of the constituents into an alcohol to effect dehydration and coprecipitation. The products obtained by calcination of the coprecipitated citrate or formate powders consist mostly of compact globules having sizes in the 3 to 10 micron range. US Patent No. 4,061,583 described doped BaTiOZ based compositions prepared by addition of a solution of either the nitrates of chlorides the required constituents to an aqueous alkaline solution containing hydrogen peroxide4.Decomposition of the peroxide containing precipitate at about 100 C results in the formation of an amorphous BaTiO3 based composition.

Calcination of the amorphous product to about 6000C gives crystalline powders. Unfortunately, the primary particle sizes of the products are not characterised. Replication of some of the examples given in the patent indicated that the amorphous powders had primary particle sizes which were substantially smaller than 0.05 microns.

Transmission electron micrographs of the products showed that the primary particles of the 6000C calcined products were aggregated.

In the above examples of typical prior art -processes, calcination is employed to complete the synethesis of the particles of the desired compositions. For reasons already noted, this elevated temperature operation is deleterious as it produces aggregated products which after comminution give smaller aggregate fragments with wide sie distributions.

US Patent numbers 4,233,282, 4,293,534 and 4,487,755 describe synthesising BaT103 and BaTiOs based compositions through a molten salt reaction in which Ba is partially replaced by Sr and Ti is partially replaced by Zr. The products are characterised as being chemically homogeneous and consisting of relatively monodisperse submicron crystallites. Doped BaTlOs based products were not synthesised. Yoon et al., "Influence of the PTCR Effect in Semiconductive BaT103 21 Mat. Res. Bul 1429 (1986) teaches employing a molten salt process to synthesise products having the composition BatO.OO-x,Sr 0. 100 Sb TiO3 where x has the values 0.001,0.002, 0.003 and 0.004.

The bodies produced from the molten salt process exhibited greater effects on the PTCR in their resistivitytemperature characteristics and larger resistivities at room temperature and larger current variations in currenttime characteristics than the comparable specimens formed from powders produced by calcination of a mixture of the oxides and oxide precursors. The differences were attributed to the use of KC1 in the molten salt synthesis process and to the smaller size and size distribution of the grains in the samples derived. Although the molten salt based synthesis process can be used to give submicron doped products with narrow size distributions, the powders are inevitably contaminated with alkali metals, since the molten salts consist of alkali metal salts. Of course, in most electronic applications alkali metals are deleterious contaminants.

Several aqueous based processes have been described for producing BaTiOo as well as BaTiO3 based compositions where Ba is partially replaced by Sr and Ti is partially replaced by Sn or, possibly, by Zr. In the process taught in US Patent number 3,577,487 doped multicomponent alkaline earth and/or Pb (II) titanates, stannates, zirconates and/or hafnates are prepared. In these cases either the coprecipitated hydrogels are treated with alkaline earth hydroxides and subjected to the same treatment steps as those used for producing BaTiO3 or the required gels and alkaline earth hydroxides are added to a preformed BaTiO slurry which is then subjected to fluid energy milling and calcination. Unfortunately, the products prior to fluid energy milling, were not characterised.Hovever, experience would indicate that the doped multicomponents products, prior to milling, should have specific surface areas in excess of 20 m2/g which indicates that the powder primary particle sizes are less than about 0.05 microns. Even after fluid energy milling at outlet temperatures in excess of 427 C, the multicomponent products cited in the examples had specific surface areas in excess of 18 m2/g. Calcination results in a further decrease in specific surface area.

This, for reasons already discussed, will lead to the formation of aggregated products.

In our copending application, UK patent application number 8710244 (2190076A), multicomponent powders having the general formula Ba (1-x-x'-x") M' x@M" x" Ti (1-y-y'-y") A y A' y0A" y"O3 are disclosed where M equals Pb (II), H' equals Ca (II), M" equals Sr (II), A equals Sn (IV), A' equals Zr (IV) and A" equals Hf (IV), x, x', x" and y, y' and y" represent the atom fractions of the divalent and tetravalent cations, respectively, each having independent values ranging from 0 to C).3 so long as the sum of either (x+x'+x") or (y+y'+y") does not exceed 0.4.The products having the above nominal soichiometries were produced in a general hydrothermal process and were termed coforms. Each of the coforms was characterised as being stoichiometric, dispersible, submicron and having a narrow particle size distribution.

Doping of the barium titanate coforms was not investigated in the copending 8710244 application. Hence, there is absent in the prior art any doped coforms of barium titanate which include calcium and/or lead or multiple divalent and tetravalent cation substitutions which are dispersible, spherical and submicron with narrow particle size distributions.

Accordingly the present invention aims to provide a dispersible, submicron doped barium titanate coform with a narrow particle size distribution.

The present invention also aims to provide a wide variety of doped BaTiO3 based compositions of BaTiOa having primary particle sizes in the size range between 0.05 and 0.4 microns.

The doped barium titanate based composition will preferably have equiaxed primary particles, be substantially free of mill media and will further have all constituents intimately mixed on a particle size scale.

Thus the present invention includes a wide variety of dispersible, doped coforms of barium titanate which are substantially spherical, intimately mixed on a particle size scale and submicron with narrow particle size distributions.

In one important embodiment of the present invention, the doped barium titanate based coform is represented by the general formula XBa (1-x ., Cax.O.YTi(i-y-y. -y", SnyZry . Hfy "Oz.ZD, wherein X, Y and Z are coefficients with X and Y having a value between 0.9 and 1. 1 and Z has a value greater than zero and less than 0.1, y, y' and y" have independent values ranging from zero to 0.3, the sum of y+y'+y" is less than 0.4, x' is greater than zero and less than 01.4 and D represents one or more dopant oxides.

In another important embodiment of the present invention, the doped barium titanate coform is represented by the general formula XBa(i-x, PbxO.YTi(1-y-y.-y",SnyZry.Hfy" On.ZD wherein X, Y and Z are coefficients with X and Y having a value between 0.9 and 1.1 and Z has a value greater than zero and less than 0. 1, y, y', and y" have independent values ranging from zero to 0.3, the sum of y+y'+y" is less than 0.4, x is greater than zero and less than 0.4 and D represents one or more dopant oxides.

In a further important embodiment of the present invention, the barium titanate coform is represented by the general formula XBa ( 1 -x-x ' X " ,PbxCax.Srx" O. YTi (1-y-y") SnyZry.Hfy.O2. ZD wherein X, Y and Z are coefficients with X and Y having a value between 0.9 and 1.1 and Z has a value greater than zero and less than 0.1, x, x', x", y, y' and y" each have independent values greater than zero and less than 0.3, the sum of x+x'+x" is less than 0.4, the sum of y+y'+y" is less than 0.4 and D represents one or more dopant oxides.

Each of the doped barium titanate based coforms of the present invention possess the same unique physical properties. The mean primary particle size of the doped barium titanate based coforms is in the range of 0.05 to 0.4 microns. Moreover, the mean particle size as determined by image analysis is comparable to the mean particle size determined by sedimentation demonstrating that the coforms are dispersible. The size distribution curve of the doped coform particles has a quartile ratio less than or equal to 2.0 which establishes that the doped barium titanate based coforms have a fairly narrow particle size distribution.

Additionally, significant is the fact that any of the dispersible, submicron doped barium titanate dielectric compositions of the present invention can be produced by a single, general hydrothermal process.

These and other details and advantages of the invention will be described in connection with the accompanying drawings in which; Fig 1 is a transmission electron micrograph at 50,000x magnification of a dispersible, submicron multiple dopant complex coform according to the present invention having the general formula 1. 02Bao.811Pbo. 105Cao.081Sro. 0003O.Tio.e32Sn0.074Zr0.084O2 0.012CoO. 0.009Mno . 0.005Nb205 and Fig 2 is a transmission electron micrograph at 50,000x magnification of a single dopant complex barium titanate coform, having the general formula 0.998Bao.782Pb0. 104Cao.oseSr0.005O.Ti0.031Sn0.070Zr0.095O2.0.03CoO which exhibits a morphology substantially similar to the morphology of the complex doped coform of Fig 1.

The preferred embodiment of the present invention is a doped coform of the general type XBa(1 -x-x'-x", PbxCax.Srx"O.YTi(1-y-y'-y", SnyZy, Hfy" Oz.ZD wherein X,Y and Z are coefficients for the divalent, tetravalent and dopant cations with X and Y having a value within the range 0.9 to 1. 1 and more preferably within the range 0.95 to 1.05 and Z has a value from greater than O to 0.1 and more preferably from greater than 0. to 0.05, x, x', x" represent the atom fractions of the divalent cations and have independent values ranging from greater than zero to 0.3 and, more.preferably, from greater than zero to 0.2 and the sum x+x'+x" has a value ranging from greater than zero to 0.4 and more preferably from zero to 0.3, y, y' and y" represent the atom fractions of the tetravalent cations and have independent values ranging from greater than zero to 0.3, and more preferably, from greater than zero to 0.25 and the sum of y+y'+y" has a value ranging from greater than zero to 0.4 and, more preferably, from greater than zero to 0.3. D represents the different dopant oxides of the barium titanate coform.

Preferably, the fine, dispersible submicron powder of the present invention consists of a doped barium titanate coform having both a tetravalent and a divalent metal ion substitution of between greater than zero and 30 mole percent. The divalent barium ion can be partially replaced by either lead, calcium1 strontium, or mixtures thereof. Additionally, the tetravalent titanium ion can be partially replaced by tin, zirconium, hafnium or mixtures thereof.

The barium titanate based compositions are doped with small amounts of one or more of a variety of dopants including the oxides of the lanthanides, cobalt, manganese, magnesium, scandium, yttrium, antimony, bismuth, zinc, cadmium, aluminum, boron,tungsten, chromium, nickel, molybdenum, iron, niobium, vanadium, tantalum, copper, silicon and mixtures thereof. The amount of dopant oxide or oxides contained in the coform ranges from between greater than zero to 10 mole percent and preferably from greater than zero to 5 mole percent. Notwithstanding with dopant or combination of dopants is employed in the barium titanate coform, the barium titanate based compositions are uniquely identified by the aforementioned morphological characteristics.Hence, both the single dopant as well as multiple dopant complex coforms of barium titanate consist of substantially spherical, dispersible particles having a primary particle size in the range of 0.05 and 0.4 microns with narrow size distribution.

The preferred approach for producing the doped barium titanate based coforms is to intimately mix the dopant or dopants wit the tetravalent hydrous oxide or oxides.

Intimate mixing can be accomplished by one of a variety of methods. Dopants may be coprecipitated with the tetravalent hydrous oxides. Alternatively, the dopants can be precipitated as high surface area hydrous oxides, washed and then combined with the tetravalent hydrous oxides. Finally, since the dopants can be precipitated in an alkaline media containing Ba(II), their solutions, preferably either as salts of acetates, formates or nitrates or as ammonium salts can be added to the tetravalent hydrous oxides. The slurry of hydrous oxides and dopants is hydrothermally treated with the oxides or hydroxides of lead and/or calcium at temperatures up to 2000C.Thereafter, the slurry is cooled to a temperature between 60 and 1500C. A solution of barium hydroxide4 or barium hydroxide partially replaced by strontium hydroxide, heated to a temperature between 700C and 100"C is added at a constant rate within a time period of 0.1 to 12 minutes to the insoluble divalent cation, tetravalent hydrous oxide and dopant slurry. The slurry is held at the addition temperature for 10 to 30 minutes and then heated to a temperature between 12U C to 225 C to ensure that the required degree of reaction of the hydrous oxide with the soluble diva lent cation hydroxide occurs.

The primary particle size and size distribution of the coforms produced by the hydrothermal process are the same whether the doped barium titanate coforms contain simply a single dopant or instead contain several dopants. This becomes readily apparent from the transmission electron micrograph of the multiple dopant complex coform, 1.02BaO. 011 Pbo. 105Ca0.081Sr 0.030. Tio . 032O.Tio.032 Sn0;074Zro.042O2.

0.012CoO.0.009MnO.0.005Nb20, in Fig 1 which shows the presence of predominantly single, substantially spherical primary particles, having a primary particle size of 0.20 microns with a quartile ratio of 1.29 which indicates that she product has a narrow primary particle size distribution.

A comparison of the multicomponent doped complex barium titanate based coform of Fig 1 with a transmission electron micrograph of a single dopant barium titanate coform, O. 998BaO. 7s2Pbo. lorCaO. ooeSro. ooeÓ. T io. SX Sn0. 070ZrO. 09902 .0.03CoO, in Fig 2 indicates that the morphologies of each of the barium titanate based compositions are very similar.

This is further substantiated by the image analyses results which show that the product of Fig 2 has a primary particle size of 0.18 microns and a quartile ratio of 1.26.

In order to evaluate the physical and chemical properties of the doped barium titanate based coforms for the present invention, a variety of laboratory tests were performed. Reagent grade chemicals or their equivalents were used throughout. The Ba(OH)z 8H2O products employed contained either about 1.0 mole or 0.3 mole percent Sr which tends to concentrate in the product. All solutions of Ba(OH)2, maintained at 70 to 90 C, were filtered prior to use to remove any carbonate present. CaCO3 was calcined at 800"C to give CaO. The latter compound when contacted with water gives Ca(OH)2. Tungsten was introduced as an ammonium tungstate solution.This was prepared by dissolving tungstic acid, W03 HnO, in heated 2M ammonia solution with stirring; the solution formed was metastable and was used shortly after its preparation.

Hydrous oxides of TiO2, SnO2 and ZrOz were prepared by neutralising aqueous solutions of their respective chlorides with aqueous ammonia at ambient temperatures. The products were filtered off and washed until chloride-free filtrates, as determined by Ago3, were obtained. A hydrous Nb(V) oxide was similarly prepared by neutralising a solution of the flouride of Nb (V). Mixed hydrous oxides of TiO2 and Snow, and TiO2 and Sub2O3 were similarly prepared by neutralising solutions containing the chlorides of Ti(IV) and Sn(IV) and Ti(IV and Sb(III), respectively. A coprecipitate of hydrous TiOz and Bi203 was prepared by neutralising an aqueous solution containing Ti(IV) chloride and Bi (III) nitrate. The percent solids present in the washed wet cakes was determined after calcination for one hour at 9000C. Several wet cakes of each product were used during the course of the work.

All synthesis experiments were performed in a 2litre autoclave. To prevent contamination all wetted parts of the autoclave were either made of titanium metal or were coated with Teflon . All synthesis experiments were performed in the absence of Co2. Filtered solutions of Ba(OH) > , maintained at a temperature of about 80"C, were introduced into the autoclave either by means of a high pressure pump or, for rapid addition, by discharging a solution of the Ba(OH)2, contained in a heated bomb, within 10 seconds into the autoclave by means of high pressure Nz. The contents of the autoclave were stirred by means of a one inch diameter turbine type stirrer, operated at 1000 to 1500RPM during the synthesis process.After synthesis, the resulting slurries were transferred to a pressure filter without exposure to the atmosphere (to prevent formation of insoluble BaCOs), filtered and dried in vacuum at 1000C.

Image analysis was used to determine product primary particle size and primary size distribution. These were determined by sizing 500 and 1000 particles in a plurality of TEM fields so as to obtain the equivalent spherical diameters of the primary particles. Two or more touching particles were visually disaggregated and the sizes of the individual primary particles were measured. The equivalent spherical diameters were used to compute the cumulative mass percent distribution as a function of primary particle size.

The median particle size, by weight, was taken to be the primary particle size of the sample. The quartile ratio, QR defined as the upper quartile diameter (by weight) divided by the lower quartile diameter, was taken as a measure of the width of the distribution. Monodisperse products have a QR value of 1.0. Products with QR values in the range of -1.0 to 1.5 were classified as having narrow size distributions; those with QR values ranging from 1.5 to about 2.0 were classified as having moderately narrow size distributions while those with QR values greater than about 2.0 were classified as having broad size distributions.

Experience indicated that the doped coforms could be classified as having narrow, moderately narrow and broad size distributions by a visual examination of the TEM's.

Based on this experience, visual examination was used to classify particle size distribution of the products of this work Since the vast majority of the doped coforms produced had narrow size distributions, in this work, average primary particle size was reliably determined by sizing 20 to 30 particles in the micrographs. Both the quantitive and semiquantitive size measurement procedures indicated that the doped barium titanate based coforms have primary particle sizes in the range between 0.05 and 0.4 microns.

Particle size was also calculated from surface areas determined by nitrogen adsorption. In these calculations the densities of the products were calculated from the composition of the powders and the literature densities of the pure component perovskites. Since the amounts of dopants in the samples are small, and since the dopants, typically, have densities which are not very different from those of the perovskites of interest, the effect of dopant on product density was intentionally ignored. The error introduced by this approximation is small.

It should be mentioned that an exact correspondence between particle size determined by microscopy and by surface area can only be expected for monodisperse spherical powders. As the distribution broadens, the degree of sphericity decreases and particle surface roughness increases, the difference between particle sizes determined by the two techniques increases. Thus, in real systems particle size determined by microscopy is, typically, larger than the size calculated from surface area. In this work, agreement with a factor of two between the two size reasurements.was taken as evidence that the amounts of fine sized precipitates associated with the particles was small.

Produce dispersibility was assessed by comparing the primary particle size and size distributions determined by image analyses with the comparable values determined by sedimentation procedures. The sedimentation process gives the particle Stokes diameter which, roughly, corresponds to the equivalent spherical diameter determined by image analysis. In this work, a Micromeritics Serigraph (Norcross, Georgia) was employed to determine cumulative mass percent distributions in terms of Stokes diameters from which the median Stokes diameters and the QR values were calculated.

Prior to sedimentation, the powders were dispersed by a 15 to 30 minute sonification in isopropanol containing 0.12 weight percent Emphos PS-21A (Witco Organics Division, 520 Madison Ave., New York) as a dispersant.

Particle sizes determined by sedimentation and by image analysis depend on different principles. For this reason an exact correspondence in size by these two methods is not always obtained. Further, as already noted, in image analysis touching particles, some of which may be bonded together, are visually disaggregated. In the sedimentation process both touching and bound particles act as single entities. These entities occur both because of the existence of bonding (e.g; necking) between some of the primary particles forming- cemented aggregates which cannot be readily broken down during the sonification process and because of less than optimum dispersion stability which leads to some flocculation.In this work, agreement within a factor of two between the median weight sizes, determined by image analysis and by sedimentation, was taken as one indication that the products were dispersible. Also,. QR values determined by sedimentation are expected (and found) to be larger than those found by image analysis. It is reasonable to assume that under optimum dispersion conditions the QR value will lie between the values determined by image analysis and by sedimentation. In this work, the additional criterion used to determine dispersibility was that the QR values of the powders obtained by sedimentation was.smaller than 2.0.

A qualitive procedure was also used to assess dispersibility. Experience in this work showed that products could be classified as being dispersible if the bulk of the primary particles in the TEM's were present as single particles. This qualitative assessment of dispersibility will satisfy the quantitative criteria described above for characterising dispersibility.

Sample uniformity was determined by scanning transmission electron microscopy having an energy dispersive X-ray analysis capability. The composition of several primary particles were determined. The product was judged to be uniform, on a particle size scale, if at least 80% of the particles contained all the powder constituents. In practice, where TEM analysis was performed, this criterion was always met. Moreover, the amounts of the various constituents, although not quantified, appeared from peak intensities to be reasonably comparable (within 80%) on a particle to particle basis.

Product composition was determined by elemental analysis, using inductively coupled plasma spectroscopy, IPC, after sample dissolution. The precision of the analyses for the major constituents was about + 2%. The precision of the results for the minor elements was less than this figure. The Ba(II)/Ti(IV) atom ratios of samples which consisted predominantly of BaT 103 were also determined by X-ray fluorescence. These ratios are somewhat more precise than those determined by solution analysis and have been employed where applicable.

The doped barium titanate coforms according to the present invention include coforms having a partial substitution of divalent lead or calcium for the divalent barium. The doped coforms also include coforms in which the divalent barium is partially replaced by mixtures of lead and calcium or mixtures of lead, calcium and strontium. Partial replacement of the tetravalent titanium cation by tin, zirconium and hafnium is also within the scope of the invention. As shown in our copending application regardless of the particular divalent or tetravalent cation substitution, the morphological characteristics of the barium titanate coforms are the same.Consequently, the following non-limiting examples include only the more complex coforms of barium titanate but are intended to provide equally representative teachings of the morphological characteristics of doped coforms of barium lead titanate and barium calcium titanate and tetravalent cation modifications thereof.

EXAMPLE 1 A series of coforms containing a single dopant were prepared by heating, from room temperature to 200"C in about 70 minutes, 0.64 L of a vigorously stirred slurry containing, in moles. 0.167 Tit2, 0.066 Sn, 0.02 ZrO2, 0.022 PbO, 0.022 CaO and 0.006 dopant salt added as a nitrate.

The slurry was cooled to 1200C and 0.46 L of Ba(OH)2 solution heated to between 70 and 90"C, containing about 0.21 moles Ba(OH)2 was added to the slurry in 3.0 + 0.2 minutes. The resulting slurry temperature was held at 1200C for approximately 30 minutes and then heated to 200"C in about 30 minutes. The slurry samples were filtered and dried and then surface area, chemistry and morphological characteristics were determined.

Atom Ratio In Solids Dopant Ba Pb Ca Sr Ti Sn Zr Dopant Co(II) 0.790 0.104 0.098 0.006 0.831 0.070 0.099 0.030 Mn(II) 0.819 0.107 0.105 0.006 0.828 0.072 0.100 0.030 La(III) 0.802 0.100 0.085 0.006 0.841 0.080 0.087 0.027 Cr(III) 0.792 0.102 0.098 0.006 0.830 0.072 0.098 0.032 Dopant X/Y Area Particle Size Size Ratio (m 2/g) Area TEM Distribution Co(II) 0.998 10.8 0.09 0.18 Narrow Mn(II) 1.037 12.4 0.08 0.15 Narrow La(III) 0.993 10.9 0.09 0.19 Narrow Cr(III) 0.998 10.9 0.09 0.19 Narrow The concentrations of the dopants and te tetravalent hydrous oxides in the filtrates were all below the detection limits of the equipment (less than 1 x 10-4 moles/L). It therefore can be assumed that these metals were almost quantitatively incorporated in the solid phase. The TEM's showed that the product particles are substantially spherical and submicron.The particle size determined from the surface area agreed within a factor of two with the particle size determined by microscopy, indicating that little fine-sized material was associated with the particles. All the doped coforms are classified as being dispersible. Visual examination of the TEM's of the cobalt doped product indicated a quartile ratio of 1.26. The TEM's of the other products were visually assessed and found to have similar size distributions. A quantitative assessment of dispersibility, using the sedimentation procedure, was obtained for the manganese doped product.

This procedure shows that the product has a particle size of 0.21 microns, a value which agrees well with the particle size obtained from TEM's, and a quartile ratio of 1.69. These quantitive data confirm that the product is dispersible.

EXAMPLE 2 Complex coforms containing three dopants were prepared by hydrothermal treatment using comparable amounts and sources of the tetravalent hydrous oxides and the alkaline earth and Pb(II) cations employed in Example 1. 0.002 moles of each dopant as nitrate salts or, in the case of niobium, as a Nb (OH)s wet cake, were added to the tetravalent hydrous oxide slurry prior to the barium hydroxide addition. After the hydrothermal treatment process it was found that the solid phase tetravalent and divalent atom ratios were comparable to those in Example 1 and are not reported here.

Dopant Hole Ratio X/Y Area Particle Size Dopant Oxides Ratio (m2/g) (microns) To Perovskite Area TEM Co(II) Nb(V) 0.031 1.035 10.0 0.10 0.20 Mn (I I) Cr (III) Co(II) 0.024 1.034 12.6 0.08 0.15 A1(III) About 50 to 60% of the A1(III) reported to the filtrate and, accordingly, the mole ratio of these dopant oxides to the perovskite, that is, the value of the coefficient Z, for the chromium-cobalt-aluminum doped coform was smaller than that for the cobalt-nidbium-manganese doped coform. All the products appeared to be substantially spherical, dispersible and had narrow size distributions. Image analysis demonstrated that the cobalt-niobium-manganese doped coform had a primary particle size of 0.20 microns and a quartile ratio of 1.29. Sedimentation studies showed that the product has a particle size of 0.25 microns and a quartile ratio of 1.67. These quantitive measurements confirm the qualitative assessments that the products are dispersible and have narrow size distributions. Further indicative of dispersibility is the agreement within a factor of two between the particle size determined by microscopy and from surface area calculations. The cobalt-niobiummanganese doped complex barium titanate coform was subjected to a TEM analysis.

It was found that the particles all contained barium1 lead-, calcium, strontium, titanium, zirconium, tin, manganese, cobalt and niobium in roughly comparable amounts.

The TEM results demonstrate that the multi-doped complex coforms of barium titanate are homogeneous on a particle size scale.

Claims (9)

1. A doped barium titanate based coform comprising substantially spherical particles having the formula XBa(1-x.)Cax. O.YTi ( -y -y. -y", SnyZry.Hfy "O2 . ZD wherein X and Y have values between 0.9 and 1.1, Z has a value greater than zero and less than 0 1, y, y' and y" have independent values ranging from zero to 0.3, the sum of y + y' + y" is less than 0.4, x' is greater than zero and less than 0.4 and D is at least one dopant oxide wherein the mean primary particle size of the doped coform is in the range of 0.05 and 0.4 microns.

2. A doped barium titanate based coform comprising substantially spherical particles having the formula XBat-x,PbxO.YTigl~y~y -y ,SnyZrywHfy 02.ZD wherein X and Y have values between 0.9 and 1.1, Z has a value greater than zero and less than 0. 1, y, y' and y" have independent values ranging from zero to 0.3, the sum of y + y' + y" is less than 0.4, x is greater than zero and less than 0.4 and D is at least one dopant oxide, wherein the mean primary particle size of the doped coform is in the range of 0.05 and 0.04 microns.

3. The doped coform of barium titanate of claim 1 or 2 wherein the doped coform has a narrow particle size distribution as determined by image analysis and the primary particle size distribution curve of the doped coform has a quartile ratio less than or equal to 2.0.

4. A doped barium titanate based coform comprising substantially spherical particles having the formula XBa (1-x-x'-x") PbxCaxzSrx" O.YTi(1-y-y'-y", SnyZry.Hfy".ZD wherein X and Y have values between 0.9 and 1.1, Z has a value greater than zero and less than 0. 1, x, x' and x", y, y' and y" each have independent values greater than zero and less than 0.3, the sum of x + x' + x" is less than 0.4, the sum of y + y' + y" is less than 0.4, and D is at least one dopant oxide, wherein the mean primary particle size of the doped coform is in the range of 0.05 and 0.4 microns.

5. The doped coform of barium titanate of any one of claims 1-4 wherein at least one dopant oxide D is selected from the group consisting of oxides of lanthanides, cobalt, manganese, magnesium, yttrium, bismuth, aluminum, boron, tungsten, niobium, chromium, nickel, molybdenum, iron, antimony, vanadium, tantalum, copper, silver, zinc, cadmium, silicon and mixtures thereof.

6. The doped coform of barium titanate of any one of claims 1-5 wherein the primary particle sizes determined by image analysis and by sedimentation agree within a factor of two.

7. The doped coform of barium titanate of any one of claims 4-6 wherein the doped coform has a narrow particle size distribution as determined by image analysis and the primary particle size distribution curve for the doped coform has a quartile ratio less than or equal to 1.5.

8. The doped coform of barium titanate of any one of claims 1-7 wherein the ratio X/Y is 1.000 + 0.015.

9. The doped coform of barium titanate of any one of claims 1-9 wherein the ratio X/Y is within the range between 0.95 and 1.1.