Document G6oRgKmaLnDBjy7VO4BVVvR9x

34S Journal of The A merican Ceramic Society--Bailey Vol. 46, No. 7 based on works25-*1"'1 involving the calculation of the change in tlie ionic polarizability of tie tetravalent cation. Kralnik11 indicates that it follows from these works that the relative stability of a parallel and antiparallel distribution of dipole moments in a lattice of the CsCl type is determined by the relative magnitude of the dipole moments of the ions of the different translatory lattices. When the difference in these dipole moments is sufficiently large, the antiparallel dis tribution is favored and conversely, if the difference is in sufficiently large, the parallel distribution is favored. A qualitative application of this influence seems applicable to the systems (Pb,Ba)HfOj and PbZrOj-PVTiOj. In PbHfOj, there is a large dipole moment attributed to the lead ions and a relatively small dipole moment due to the Hf ion. The difference between the dipole moments is apparently large enough to lead to the antiparallel arrangement of dipole moments and therefore the antiferroelectric character is observed in PbHfOj. Replacement of Pb ions by the less polarizable Ba ions apparently reduces the difference be tween the dipole moments of the tetra- and bivalent cations sufficiently to favor the parallel- arrangement of dipole moments, leading to the ferroelectric phase observed in the system (Pb.Ba)HfOj. The same reasoning leads to a rough explanation of the results observed by replacing the Zr ion in PbZrOi by the ' .. In this system, the smaller Ti ion has a larger ionic zability than does the larger Zr ion. Therefore, inng amounts of Ti ions would reduce the difference ,-en the dipole moments of the tetra- and bivalent ions aereby favor the parallel arrangement of dipoles, rursory examination of the analogous systems (Pb.Ba)and (Pb.Ba)ZrOj shows that, for a given temperature, ;es more Ba ions to induce a ferroelectric phase in the 'Oj system than in the PbZrOj system. Applying the ning used in the previous paragraphs, this is interpreted waning that it takes more Ba ions in the PbHfOj system in the PbZrOj system to reduce the difference between ;etra- and bivalent cations to a level which favors the del arrangement of dipoles. This can be explained by ____ning that the Hf*+ ion has a lower polarizability than the Zr<+ ion. The Hf ion and Zr ion are very nearly equal in size so the ionic polarizabilities would be essentially the same. Therefore, the lower polarizability of the Hf ion would have to be attributed to its lower electronic polariz*. bilitv. Thus, it seems reasonable to assume that differences which are observed between the analogous hafnate and zirconate systems are due to the lower electronic polarua. bility of the hafnium ion. V. Summary of Results (1) The intermediate antiferroelectric phase observed in PbHfO* was observed in both the systems (Pb.Ba)HfQ, and (Pb.Sr)HfOj. (2) A rhombohedral ferroelectric phase was observed ia the system (Pb.Ba)HfOj. The angle a between the rbombohedral axes for ferroelectric Pb0.ioBa<i.joHfOj was less than 90 by 6 minutes at 25C. (3) A phase transition, in addition to the transition in PbHfOj, was observed in the system (Pb,Sr)HfO. The transition occurs at a temperature which is above the upper antiferroelectric phase transition of lead hafnate. (4) The explanation of the appearance of the ferroelectric phase in the system (Pb.Ba)HfOj was based on a qualitative discussion of the influence of the electrostatic energy on the relative stability of the ferroelectric and antiferroelectric states. The less polarizable Ba ion reduces the difference between the dipole moments of the tetra- and bivalent ions and this reduction favors the parallel arrangement of dipoles. (5) The differences between the analogous hafnate and zirconate systems appear to be due to the smaller electronic polarizability of the hafnium ion. Acknowledgment The writers wish to acknowledge the fellowship grant of the Lead Industries Association which made this research possible. 11J. A. Sauer, "Magnetic Energy Constants of Dipolar Lat tices.'' Pkys. Ren., 57, 142-46 (1940). 14 Yutaka Takagi, "Ferroelectricity and AntiferToelectridty of a Crvstal Containing Rotatable Polar Molecules." Rhys. Rev., 85 [2] 315-24 (1952). ,jul fri: Microscopy of Steam-Cured Asbestos Portland Cement Products by DONALD BAILEY Retearch and Engineering Center, Jahnt-Manville Product* Corporation, Monvitle, New Jeney A method is presented for the microscopic exam ination of 8sbestos-cement products. The use - of petrographic thin sections and polished. sec tions in determining composition, texture, grain size, and cement-silica reaction is described. The preparation of thin sections and polished sections, using epoxy resin as an impregnant, is outlined. I. Introduction lthough the gel matrix of hydrated portland cement A pastes consists of poorly crystalline or amorphous material too fine in structure to be resolved by the light microscope, much of significant interest remains in these otherwise fine-gTained products which can be observed with the light microscope when properly applied.1 Pozzolaruc materials, fillers, asbestos fibers, and grains of unhy drated portland cement (present even in well-cured products) are among those features which may be brought to light at magnifications of X100 to X500. < Received October 22, 1962; revised copy received March 2". 1963. . The writer is microscopist, Research Center, Johns-Manvu' Products Corporation. 1 L. S. Brown, "Petrography of Cement and Concrete.'' J Develop. Lob., Portland Cement Assoc., 1 [3J 23-34 (Septefn*>er 1959). .4 II. Specimen Preparation The petrographic rhi-n section frequently affords a compre hensive means of evaluating hydrated portland cement } r products!4 Specjmens Which do not require impregnation before sawing may be sectioned as follows: A wafer cut from l the material is surfaced on a glass lap or plate with 600 or | 800 silicon carbide in ethylene glycol. After washing and * overnight drying, the wafer is impregnated under vacuum in t Epon 820 (Shell Chemical Company) plus two drops of ni benzyl dimethylamine and two drops of dibutyl phthalate * per milliliter of resin. The wafer is then mounted on a glass ? slide and left in the oven at 65 to 70C without vacuum for 1 several hours or preferably overnight. When fully cured, the i resin forms a hard, transparent solid, possessing excellent i mechanical strength and adhesiveness. The refractive index, 1.535, is sufficient to give quartz moderate relief and permits better resolution of portland cement than is possible with Canada balsam. Epoxy resins develop strain bire fringence from the impact of abrasive panicles, but this is evident only in holes or around edges. The mounted wafer .. may be reduced to the proper thickness by grinding on a ' cast-iron lap with 100 silicon carbide, followed by 600 silicon r carbide. 320 silicon carbide Wet-or-Dry paper mounted on ' a rotating wheel affords a rapid alternative method for the " final grinding of thin sections and greatly reduces the in~ cadence of strain birefringence in epoxy mounting media. A metal block machined to accommodate the hand and with a flat undersurface into which has been cut a fiat trough to take the glass slide will serve as a simple holding device. The same resin formulation used to mount the wafer also serves as _ a cover glass cement. The examination of polished sections of hydrated portland cement by reflected vertical illumination will frequently supply specific information not readily obtained through a S study of thin sections alone. The two methods in fact supple st ment each other. Impregnation of the specimen before # polishing is necessary, and for this purpose various resins, % including methyl methacrylate, have been found to be satisfactory'. The section shown later in Fig. 3 was impregnated 3? with the following mixture: methyl methacrylate monomer, jjt 5 ml; Laminae 4116, 2 drops; methyl ethyl ketone peroxide, > 2 drops. This mixture cures in from 12 to 24 hours at 65 7 to 70 C. Equally good results have been obtained with f epoxy resins as outlined in the foregoing for thin sections, except that the impregnated wafer is placed on a sheet of aluminum foil instead of a glass slide while curing. After :% curing of the resin, the aluminum foil is peeled away, and the T impregnated wafer is mounted face down in a block of plastic. This mount may be formed of Laminae 4116, plus 1 to 3% % methyl ethyl ketone peroxide.5 The combination sets within % a few hours at room temperature and requires only two Lshaped metal blocks as embedding equipment. After the s* Laminae has .cured, excess epoxy resin is removed from the `g face of the specimen by buffing cm a wheel armored with 320 silicon carbide Wet-or-Dry paper. Since only shallow penetra# lion by the epoxy resin will have taken place, care must S' be used to remove only the excess resin. The specimen is 'S-- tiea polished on a canvas lap with Linde A powder in ethylene g glycol, washed in alcohol, and etched for 3 seconds in 1% total. III. Results _ Texture, grain size, shape characteristics, and size distribu 3?: tion are among the factors which may.be deduced from micro- J Herbert Insley and V. D. Frechette, Microscopy of Ceramics lad Cements Including Glasses, Slags, and Foundry Sands. f-cadtmic Press, Inc.. New York, 1955. 286 pp.; Ceram. Abslr., f May. p. 108a. . 1 E- P. Cadwell, "Color Microscopy for the Mill Man," Eng. lining J.t 160 jl] 81-90 (1959). Fig. 1. Avtocisved-cured asbestos cement. Arrows indicate vnbydrotod cement; Q quartz grains; C chrysolite. Note oiso the croeidolite asbestos. Light-colored mottling shows carbonation within the set. (Transmitted light; crossed Ntcois with analyzer rotated from crossed position. X500.) scopic study of thin sections. By means of cross-polariza tion, mineral analysis is also achieved. Owing to overlap ping of grains, only semiquaatitative results are possible. However, the method supplies information not obtainable by other means. In sections impregnated with epoxy resin, quartz is readily seen even without cross-polarization, since both indices of the mineral are sensibly lower than the medium. Quartz grains in steam-cured portland cement products frequently show thin rims of reaction. These show distinctly in epoxy resin, since the reaction product is lower' in refractive index than quartz. The rims usually are not'strongly birefringent, except in regions of considerable carbonation. A few of these rims may be seen in Fig. 1, where they appear as thin double lines along the edges of quartz grains. Otherwise, quartz may be recognized in the usual manner by its low birefringence, uniaxial character, and absence of cleavage. Chrysotile asbestos is easily recognized by its fibrous struc ture, its low refringence and birefringence, and its positive elongation. A small fiber bundle of chrysotile is indicated in Fig. 1. The dark-colored fibers in the photomicrograph are croddolite. This mineral is easily recognized by its blue color, high relief, and strong pleochroism. Unhydrated portland cement is almost always present in hydrated portland cement products, whether cured at room temperature or in the autoclave. It occurs as oversized grains which apparently have been walled in by gel and thus pro tected from further action. The presence of a moderate amount of unhydrated cement is not in any way detrimental, but is in fact a good indication that the product is sound and not excessively porous. It is easily recognized in thin section by its high relief and low to moderate birefringence and by the fact that a single grain usually is comprised of many smaller crystals. Since 3CaO-SiOi has a birefringence July 16 350 Journal of The American Ceramic Society--Bailey Vol. 46, No. 7 ated in : ; rials, a providec to the ( sections in the f< either t\ care am ' pressure in plasti mined. ' proper h Evide may be 1 have gr< well-pre FTg. 4. Crystols of aragonite (CoCOj) filling a crack in autoclave-cured osbestos cement kept 12 years in a corrosive solution in the presence of CO]. (Transmitted light, crossed Nicois, analyzer rotated 4 0 from crossed position. XI00.) Fig. 2. Portland cement clinker. Grey oblong crystal* are 3CaO SiOj, twinning structure is 2CoO-SiO, and the white interstitial material is 4CoO- AljOj-FejOjin o matrix of darker glass. (Reflected light; X600.) of 0.005 and 2CaO-SiO* a birefringence of 0.018, it is some times possible to differentiate the two in a thin section. Be cause, however, of overlapping and small size of the individual crystals, this is not usually practical. Quantitative determinations usually are possible with properly etched polished sections. Since in this method only those grains exposed at the surface are viewed, the problem of overlapping is eliminated. Moreover, the polysynthetic twinning in 2CaO-SiOj serves to distinguish it from 3CaO-SiOj, whereas 4CaO AljOj FejO, also present in portland cement, is recognized by its high reflectance. Glass and 3CaO AljOj, present as interstitial components, are recognized by their intermediate reflectance. These features are re produced in Fig. 2, showing a polished section of portland cement clinker. A comparison with Fig. 3, showing a pol ished section of steam-cured asbestos cement, will illustrate the superiority of the polished section over the thin section for the identification of portland cement. On the other hand, substances of comparable hardness and similar shape characteristics, such as quartz and glass, cannot be differenti- ;s- ' Fig. 3. Autoelove-eured osbestos cement showing several grain* of un hydrated cement. Eoch of these grains is comprised of individual crystols of 2CaO-SiO>, 3CoO SiO;, ond 4CoO AI;Oj Fe;Oj in glass. The lightcolored groins surrounded by dork border* ore quartz. A bundle of Chrysotile asbestos moy be fointly discerned ot the lower left. (Reflected light; X500.) Fig. 5. Thin section of an experimental mixture containing perlite, l`me' portland cement, and quartz. Note reaction rim (arrow) obout a pe'i'le groin. (Transmitted light; X500.) he Is Tdirecficuh however, tion dist mens bs From su relates I ofmeasu The p i- a grade graphic L needles) I*, trustor/: ft"/? gra^ ot /tjT-r-aphic .S'* me grain siz' Younj V were m. Poisson: &. Young's i ated in a polished section. For the distinction of such matei;als, a polished thin section offers definite advantages, provided a means of changing from one type of illumination .o the other has been built into the optical system. Thin sections impregnated and mounted in Epon 820, as outlined in the foregoing, have been polished enough to be viewed by either type of illumination. The polishing operation requires care and patience, since a thin section will not tolerate the pressures commonly'applied in polishing specimens mounted in plastic blocks. Once the correct pressure has been deter mined, the task is not unduly time-consuming, provided a proper holding device is used. Evidence of chemical attack in a portland cement product may be seen in the appearance of secondary crystals which have grown at the expense of the host material. For example, well-preserved microcrystals of aragonite (CaCO*) were found filling a crack within a specimen of steam-cured asoestos cement which had been kept 12 years in a corrosive solution in the presence of COt in the laboratory. These are shown in Fig. 4 and are a good example of the use of epoxy resin as an impregnant for friable specimens. The occurrence of reaction rims about quartz in steamcured portland cement products has been noted. An even more striking example of zonal reaction about a pbzzolanic substance is indicated in Fig. 5, showing a thin section of a steam-cured mixture containing portland cement, lime, quartz, and perlite. The reaction rim, surrounding a grain of perlite, appears to consist of minute fibers growing per pendicular to the edge of the grain. Acknowledgment The invaluable guidance of Herbert Insley is gratefully ac knowledged. Discussions and Notes t j j Single-Crystal Elastic Constants of BeO from Polycrystalline Measurements by L. H. SJODAHl and B. A. CHANDLER he lack of BeO single crystals of sufficient size has made the The degree of preferred orientation was calculated from meas Tdirect measurement of single-crystal elastic constants dif ficult. Young's modulus for single crystals can be deduced, urement of refiected Intensities from six crystallographic planes by X-ray diffraction on a transverse section, normal to the axis I however, from measurements of Young's modulus and orieota-of extrusion. The intensities I for the six planes of the transverse r tion distribution functions of three polycrystalline BeO speci- section were compared with the corresponding reflections I% i mens having different degrees of preferred grain orientation. from randomly disposed material (7* values were normalized w From such measurements an equation has been derived which so that /<m> 1.00). The ratios 1/1% represent the amount : relates Young's modulus for single-crystal BeO to the angle relative to random of BeO oriented with c axes at the. angles 6 of measurement from the crystallographic c axis. between the reflecting plane and the basal plane, which are The polycryst :iine specimens used were rods extruded from equivalent to the angles between the c axis and the extrusion . a grade of BeO that contains needlelike particles (crystallo- direction. A plot of 1/1% versus 6 produces a smooth curve from graphic c a-ds in the direction of the long dimension of the which the 1/1% ratio can be interpolated for ten angles corre 7 aeedles)'-which tend to hecome aligned in the direction of ex* sponding to the midpoints of ten arbitrarily selected zones about trusi<j and whose growth in sintering enhances the preferred CO- g'ain orientation o( the rods. Orientation of the crystallo- i^graphic c axes in the direction of extrusion as high as 80% has " bfen measured for rods formed by extrusion and sintered to ^ grain sizes of about 100;x. 7 Young's modulus measurements used in the calculations I. were made by dynamic methods, corrected for size, shape, Foisson ratio effect, and corrected to zero porosity by established J*; Young's modulus-porosity correlations. Received March 14, 1963. This work was performed under United States Atomic Energy Commission Contract No. AT(40-1 )-2S47. The writers are principal engineers, Nuclear Materials & Pro pulsion Operation, Advanced Technology Sendees, General Elec tric Company, Cincinnati, Ohio. &- li: