PROPERTIES OF SOME PHOSPHATE CEMENTS JOHN RUSSELL GREEN THESIS FOR THE DEGREE OF BACHELOR OF SCIENCE IN CERAMIC ENGINEERING COLLEGE OF ENGINEERING UNIVERSITY OF ILLINOIS 1922 Digitized by the Internet Archive in 2016 https://archive.org/details/propertiesofsomeOOgree ■ PROPERTIES OF SOME PHOSPHATE CEMENTS CONTENTS Page I. Introduction 1 II. Scope and Results of Former Work ..... 1-2 III. Basis of Study 2-4 IV. Preliminary Study 4-13 A. Objects ........ 4 B. Selection for Study ............. 5 C. Preparation of Powders 6-S D. Results ... ......... 9-12 E. Conclusions . 12-14 V. Final Study 14-22 A. Objects 14 B. Preparation of Powders ........... 14-17 C. Results ................... 17-21 D. Conclusions 21-22 VI, Summary 22-23 PROPERTIES OF SOME PHOSPHATE CEMENTS I Introduction The so-called silicate cements commonly used in the dental pro- fession have been subjects of considerable study in the past , with particular regard being paid to their physiological properties as tooth fillers. These cements consist of finely ground powders, set into a cement by solutions of phosphoric acid, often containg modi- fiers of aluminum phosphate or related compounds. These mixtures harden quite rapidly, develop considerable strength, and do not show a great amount of shrinkage. The main constituents are Lime, Alumina, and Silica, or their mineral compounds, and various other oxides such as Bertflliuni, Zinc , and Boron. These oxides are calcined or fused at high temperatures in the course of their preparation as powders . The present work is based on, and a continuation of, the ex- periments carried on in 1916 as a thesis by Joseph William Wright, The results of this former work makes it possible to materially re- duce the field of investigation, and to specialize in the determin- ation of the various physical properties and compositions of phos- phate cements. In Wright’s work the compositions of the powders used, varied to cover nearly all of the ternary system. Lime, Alumina, and Sili- ca. The liquids used for setting vary in strength of Phosphoric II Scope and Results of Former Work Acid from include also some solutions con- T • . 4 .4 " 4 ' - . . T . ' ' . r\ *C. m taining modifiers. The powders were calcined at various tempera- tures and the crushing strength, time of set, and heat developed at setting were determined for each variation of heat treatment, material and acid concentration. The results show that mixtures of the compositions CaO.SiOg and CaO.Al^O^ , when calcined at a high temperature .harden readily with phosphoric acid solutions. The calcium aluminate develops considerable strength. Calcined mixtures of lime and clay showed better working properties, set well and developed excellent strength Increase of acid concentration decreases the rate of setting, gives less heat in setting and lowers the porosity of the hardened ce- ment. The addition of A1 2 (P0^) 2 to the solution retards setting. an The presence of hygroscopic water in the powder seems to have im- A portant effect on the setting properties. As commercial cements, the composition C . A . s 2 » and CA^S^ mad « from clay (N. Carolina Kaolin), and Whiting, and mixed with liquids containing 60 and 75 1 ° of phosphoric acid were found to give the most satisfactory results . Ill Basis of Study These mixtures form the foundation upon which this study is built. Both of these mixtures lie within the field, on the tri- axial diagram of Lime, Alumina, and Silica, which is bounded by S, AS and CA 2S. If these mixtures were entirely melted and al- lowed to cool, they would therefore crystalize out into the three compositions of SiC>2 » AI2O3 , Si02 * and CaO .AI2O3 ,2Si02 > but if they are not cooled in equilibrium or have not all gone to the liquid phase, there may be other compounds present on cooling which ' •• v . .. ■ V *• ■ . j - . •• - -■ ’ V- • .. i ♦ I*.; ' ; . ■ i 1 ■ ! i. . ■ v, ■ . . J ■ '• . : * ' - * > ■ C ' ' . - 4 ~ would fall in some of the adjoining fields. The limits set for si- licate cements by Dr*s. W. and D. Asch 1 are : CaO 6 to 12 percent A1 2 0^ “to 50 percent SiC >2 40 to 44 percent These limits also fall almost entirely within this area on the tri- axial as shown by Figure 1« This diagram of the triaxial field shows the areas of crystalization near the compositions studied. The limits as shown above are plotted, and it can be seen that a small area (shown shaded) is within all limits found in commercial cements. This area shall hereafter be referred to as the "most probable area*. In Figure 1, the compositions studied by Wright are designated by email red circles, the compositions showing particular promise are crossed. The boundaries of the areas of crystalization are shown, as well as the area in which are found the commercial Port- land cements. All compounds are marked, and the compositions studied in the present investigation are designated by small black circles . to IV Preliminary Study A, Objects. Preliminary studying the specific properties of any cement to A a large degree it was considered advisable to determine the most satisfactory? (l) Body Composition. ^ Silicates in Chemistry and Commerce. - 5 ' (2) Clay material. (3) Temperature of calcination. (4) Percentage of Phosphoric acid for short time tests. B. Selection for Study. (l) In order to study the effect of composition within a fairly small range, four bodies of the following compositions were select- ed: Body No. 1 2 3 4 CaO 20$ 6$ 6$ 10$ A1 2°3 42$ 56$ 42$ 4S$ Si0 2 3S$ 3S$ 52$ 42$ As is shown in the diagram , No. 1 is located in the field CaO .AlgO^ .2Si0g - 2Ca0 .AlgO^ .SiOg - AlgO^ , and contains the most lime} No. 2 is located on the boundary line between the fields SiO ? , CaO .AlgOj .2SiC 2 - Al^O^.SiOg and CaO. AlgO-^ ,2Si0g- Al^O^. SiOg - AlgO^, and contains the greatest amount of A1 2 0^; No. 3 con- tains the most SiOg and is located in the field SiOg - AlgO-^.SiOg- Cao. Alg0^.2Si0g; and No, 4 is located in the center of the "most probable area". These bodies are made from the following mater- ials! Whiting CaCO-? J Clay (as far as possible) Aluminum Hydroxide Alg(OH)^ Flint Si Op (2) In the former study only one clay substance was used. As clays vary considerable in their physical properties, it was thought advisable to test the effect of a variation in the clay ingredients - 6 - by making each body with the following clays? A. North Carolina Kaolin B. Florida Kaolin C. English Kaolin (3) As former mixtures had been calcined only at high tempera- tures, it was decided to test the range of the temperature of cal- cine by firing each body to the following temperatures? (a) 1200 Deg, C (b) 1000 Deg, C (c) S00 (d) 600 (4) The two strengths of acid concentration found to be most desirable for developing strength, and as possibilities for further study are? Per Cent, H-,P0^ Specific Gravity by Wt, 75 1.57 60 1.4l C. Preparation of Powders. (1) Calculating and Weighing? The batches were calculated and weighed according to the ceramic formulas of pure substances. The deviations of actual analyses on the clays used from theoretical composition is not sufficiently great to warrant their use, (2) Molding and Drying of Raw Materials? Each batch wae mixed thoroughly on oilcloth, with suffi- cient distilled water added to make plastic, and molded in brass molds 1" x l n x 6", The resulting bars were dried at ( \ > ( ■ -7- room conditions for a day and then were placed in a drier at 60 Deg. C., and kept there until placed in the kiln for cal- cination. (3) Calcining: The bars were calcined in covered saggars in an open gas fired kiln. The temperature was measured by a platinum - platinum rhodium thermocouple and galvanometer. The maximum temperature was maintained for half an hour and then the burners were turned off. (4) Grinding and Screening? After removing from the kiln the pieces were kept in a drier at 60 Deg. C. until ready to be ground. In grinding the material was crushed progressively in (l) jaw crushers, (2) roll crushers and (3) small porcelain ball mills (dry). The grinding in the mills was continued only until about 90 $ or over of the material could pass a 150 mesh seive. After removal from the ball mills, the powders were screened in a Ro-Tap machine to pass 150 mesh. The residue on the screen was returned to a ball mill and ground to pass 150 mesh. (5) Mixing and Molding: Because of the large quantities of materials to be mixed, the practice of the dental profession could not be strictly followed. Instead, the powdered materials were mixed frith the phosphoric acid solution in a porcelain mortar with a porcelain pestle. The entire portion of powder (about 60 grams’ was placed in the mortar and then covered with the desired amount of acid. By stirring with the pestle, more acid being added if needed, the mixture was brought to the consistency V * f N — . . J • ♦ f ' * ' ) - . . . . ■ . * ' ■l'' : ' , ■■■, . . . . -g- of a stiff paste o This method did not lead to a great amount of symmetry between the separate mixes. The molding was done by forcing the plastic material into wooden molds, forming cylinders 3/^* in diameter and 1 1/4" high. Each mold made five cylinders, measured S" x 1 1/4" x 1 5/8* and was care- f fully coated with paraiine before each filling. The molds A were split thru the center, forming a perfect half on each side, in order to aid in removing the cylinders. During the time of molding and until the next mold was to be filled, the molds were held together by two small clamps over the ends and were placed on glass plates to form perfect ends to the cylin- ders. The material was forced into the molds with a porcelain spatula, covered with another glass plate, and subjected to a load of 10 lbs. until the next mold was ready (about 10 min.) (6) Setting! The test pieces were allowed to set in the molds under room conditions for the first 24 hours. After this period they were entirely exposed to the open air for the remainder of the seven days which was allowed for setting before crush- ing strength tests were made. (7) Crushing Strength! The crushing strength was determined in a 10000 lb. Tinius Olsen testing machine located in the Fatigue of Metals Labora- tory at the University of Illinois. The cylinders were sup- ported on both ends by small bearing plates (l/2" steel balls) and padded with two layers of blotting paper. Five test pieces were used for each mix. All were not perfect, however, and many were discarded. •f * ’ . . . : < f ' * . ' i ’■ r . . ■ .. f . ' ... • _ . 1 . rr.iv •. . .. . . . ) . . . - 9 “ Do Results o (1) Working properties: When in a plastic condition, the Florida Kaolin gives the best working properties, the English Kaolin is next in order, followed by the N. Carolina Kaolin, The mixture of Feldspar barely held together, (2) Calcination: The bars of Feldspar mixture were the only examples of deformation of any kind. These bars showed some signs of fusion over surface. Actual temperature attained: (a) 1210 Degrees C. (b) 1050 (c) S00 (d) 600 (3) Grinding and Screening: Observations of the length of time required to grind the powders in the ball mills, and to screen them showed that the mixtures made from Florida Kaolin calcined into the hardest mass at each temperature, the English Kaolin next, and the N. Carolina Kaolin broke up very easily into seemingly large grains , The average number of revolutions in ball mills required to grind the Florida Kaolin mixtures was 225,000 while the number of revolutions for N. Carolina Kaolin mixtures was 125,000, The Florida Kaolin gave all evidences of being the finest grained material and the most easily vitrified. ( \ ( \ / . . -10 ( 4 -) Setting: The amount of setting shrinkage seem3 to be slightly greater for powders made from Florida Kaolin. This shrinkage was estimated from the ease with which the cylinders were re- moved from the molds. ( 5 ) Crushing Strength! (a) Powders calcined at 1210 Degrees C. Powder Acid Con- Average Characteristics No, centration Crushing Strength 75 1 A 60 Set immediately 75 IB 60 H » 75 1 C 60 H It 75 2 A 60 1100 (1 test) 75 2 B 60 75 2 C 60 — 3 A 75 — 60 — 75 1130 (3 tests) 3 B 60 1215 (4 tests) 3 C 75 — 60 7&0 (l test) 4 A 75 — 60 I ' ' - ' ■ Powder Acid Con- -11- Average Characteristics No. centrat ion Crushing Strength 4B 75 665 (4 tests) 60 1750 (2 tests) 75 600 kc 60 75 600 Set immediately Spar 60 (b) Powders calcined at 1050 Degrees C, Powder Acid Con- Average Characteristics No, centration Crushing Strength 1A 75 60 Set immediately IB 1C 75 60 75 60 75 It « « « Soft 2A 6o — Stiff 75 — Soft 2B 60 1200 Stiff (2 tests) 75 Soft 2C 60 500 Stiff 75 Set too quickly 3A 60 soo (l test) 75 3B 6o 75 3C 6o - 12 - Powder Acid Con- Average Characteristics No. centration Crushing Strength 4A 75 Set too quickly 60 4b 75 * w it 60 4C 75 * « H 60 75 Spar 60 Set immediately (c) Powders calcined at 800 and 600 Degrees C. All of these powders set too quickly for placing in molds and so gave no information as to their character- istics,. E 0 Conclusions* (a) A survey of the results obtained by variations in the body compositions indicates? (1) The speed with which the initial reactions of setting take place varies directly with the percentage of CaO in the mixture. This is shown by the fact that No. 1 composition shows the first signs of such initial set and that No. 4 con- taining the next largest percentage of CaO is the next in order „ (2) The addition of 1^3 part feldspar as a fluxing mater- ial greatly increases the speed of initial set. (3) The compositions located entirely in the field SiOg “ AlgO-^.SiOg - Cao. Al^O-^ . SiQ 2 are the most satisfactory. The lower percentage of lime in No. 3 makes it a slower setting * ( ' _ ■ ► * % ( ' ' . . . . < I . - 13 - mixture « ( 4 ) The highest crushing strength is found in composition No. k - 9 located in the center of the "most probable area"* Selection for further study; Composition No. (b) Results of the variations of clay material indicate; (1) A more satisfactory white color is obtained from Flori- da Kaolin and English Kaolin than from North Carolina Kaolin. This difference becomes more noticeable as the temperature of calcination is decreased. (2) The best working properties both in mixing the materials before calcination and for mixing with the acid solutions, as well as the highest crushing strength are found with the mix- tures made with Florida Kaolin, Selection for further study; Florida Kaolin. (c) The results of variations in the temperature of calcina- tion indicate; (1) The speed of the reaction of initial set increases with a decrease in the temperature of calcination. (2) The crushing strengths obtainable are higher with a higher temperature of calcination. Selection for further study; 1250 Degrees C. (d) Variations in the strength of acid solutions indicate; (l) The concentration of 6o$> phosphoric acid gives the highest results in crushing strength for a period of setting of one week. Selection for further study; A 60 $ solution of phosphoric acid in distilled water. . -14- V Final Study A. Object® In a final study it was desired to determine the effect of: (1) Slight variations in the proportion of dry material to acid. (2) Moisture and vapor pressures during period of initial set as determined by crushing strength. (3) Moisture and vapor pressure after initial set as deter- mined by crushing strength. It was also desired to obtain by the use of somewhat larger cylinders, more reliable data upon the crushing strnegth obtain- able with this type of cement. B. Preparation of Powders (l) Composition and Materials: (a) A body of the following composition was prepared: CaO 10fo A1 2°3 SiOg (b) Materials: Material Whiting Florida Kaolin Approximate formula CaCO^ A1 2 0^. 2Si0 2 .6H 2 0 Aluminum Hydroxide Al p (OH)^ Batch Wt, 537 gram 2709 564 Total 3S10 gram (c) Acid Solution: The acid solution oontained 60% by weight of H^PO^, . « * % ( -15- and a small quantity of Al^PO^Jg or£ *® r to retard the rate of setting. The specific gravity of the liquid be- fore the addition of the Al^ (P0^) 2 was 1.4-1. (2) Mixing and Calcinings The finely powdered materials from stock were mixed on an oilcloth, passed thru a 20 mesh seive, and mixed again. This assured a thorough mixing and a satisfactory fineness of material. The powder was then placed in new clean saggars, wadded tightly and fired in a coal fired test kiln. A plati- num-platinum rhodium pyrometer registered 1150 Degrees C. on a galvanometer . Orton pyrometric cones placed in the kiln showed that a heat treatment of Cone 5 was attained. (3) Grinding and Screening? The calcined material was ground in small porcelain ball mills as before, and screened through a 150 mesh seive in Ro-Tap machine. The residues on the screens were small and gave the same results with acid tests as were found with the screened materials. (4-) Mixing; A definite proportion of powder to acid was attained by weighing out exactly 75 grams of the powder for each batch. The acid solution was measured to l/2 cubic centimeter in a 50 c.c. graduate. In order to make the mixing more uniform throughout the short period of stirring, a small quantity of the powder at a time was added to the mortar and partly covered with acid solution. In order to keep the initial set as low as possible, the acid solution was kept in a bath of cold water. Two consistencies were used for each test. The . ■ - . . * \ . ' . . . . -16- first contained 37 cc. of acid solution, and had the consis- tency of plastic putty. The other contained 40 c.c. of acid solution, and worked into a creamy paste, which flowed quite readily from a porcelain spatula. In commercial practice it would be very easy to remain within even smaller limits by an observation of the consistency of the paste. (5) Molding! The molding was done in the wooden molds, as in the preliminary work. The interiors were waxed as before, but the surfaces of the glass plates used to form the ends, were covered with a thin film of oil to avoid sticking to the plates. After the pasty material had been forced into molds with porcelain spatula, a pressure of approximately 5 lbs. was applied by a small wooden piston, forcing the material more compactly into the mold. The mold was again filled to the top and compressed by the spatula. During the period of preparation of another mold, a force of 10 lbs. was applied to the upper glass plate. Ten test pieces were prepared for each test, five being of 37 c.c. mix and five of 4-0 c.c. mix. (6) Setting: The period of initial set is assumed as the first 24- hours. 2 Seven variations in the conditions of setting were made as follows: 1. The first 24- hours in the mold under atmospheric conditions, followed by 6 days entirely exposed to atmos- pheric conditions. p Morgenstern, Osterr .-Ungas . Vierteljahrschr . f. Zahnheilk 1905 p.535 - - r . ! * . , t- i . . . -17- 2, The first 24- hours in the mold under atmospheric conditions, followed by the final period of set completely exposed to an atmosphere of high vapor pressure, and 100 % relative humidity, 3 o The first 24- hours in the mold under atmospheric conditions, followed by 6 days of final set completely immersed in distilled water, 4-, The first 24- hours in the mold under atmospheric conditions, followed by a final setting period of 6 day% entirely exposed to an atmosphere of low vapor pressure, 5, The first 24- hours in the mold under conditions of high vapor pressure, followed by a period of 6 days completely exposed to atmospheric conditions, 6* The first 24- hours in the mold under conditions of low vapor pressure, followed by a period of 6 days, compdetely exposed to atmospheric conditions, 7. The first 24- hours in the mold and immersed in distilled water, followed by a period of 6 days complete- ly exposed to atmospheric conditions, (7) Crushing Test: The crushing strength was determined in the same manner as in the preliminary work. The only difference being that for the final work, the cylinders for each test were arranged in the order of their visible perfection before crushing. ■ i • « • - » V « 1 t • ( ' -IS- C. Results (a) General Atmospheric Data: Day Wet Bulb Dry Bulb Relative Vapor Pressure Humidity (Carrier) 1 60° C 72i° C 50$ 0.400 in Hg. 2 6l 70 60 0,425 3 620c 72 56 0.420 4 64 71 2 6S 0 .480 5 63 72 60 0,450 6 63 72 60 0,450 7 63 73 5$ 0.450 g 66 73 66 0.520 9 69 7&i 70 0.5S0 10 68 74 75 0.575 11 69 75 79 0.610 (b) Conditions for each test: (1) Atmosphere - Atmosphere Temperature of Initial Set 22*5° c Vapor Pressure of Initial Set 0.400 in Hg. Ave. Temp. Final Set 22.6 Ave. Vapor Pressure Final Set 0.45 (2) Atmospheric - High Vapor: Temperature of Initial Set 22,50 c Vapor Pressure of Initial Set 0.400 in Hg. Temperature of Final Set 22,6 Ave. Vapor Pressure Final Set 0.675 ( 3 ) Atmosphere - Immersed: Temp, of Initial Set 22.22 Vapor Pressure Initial Set 0.420 - 19 ' Ave, Temp. Water Final Set 21.0 (4) Atmosphere - Low Vapor Pressure Temperature of Initial Set 22.22 Vapor Pressure of Initial Set Ave. Temperature of Final Set 22,6 Ave. Vapor Pressure of Final Set (Vapor pressure over 95 i° HgSO^. at 22° C as determined by inter- polation of values found in Physical Chemical Tables * Landolt & Bbrnsteln.) ( 5 ) High Vapor Pressure - Atmospheric 0.4-20 0.0Q09S5 Temperature of Initial Set 22.22 Vapor Pressure of Initial Set 0.630 Ave. Temperature of Final Set 23.5 Ave. Vapor Pressure of Final Set 0.531 (6) Low Vapor Pressure - Atmospheric Temperature of Initial Set 22.22 Vapor Pressure of Initial Set 0.Q009&5 Ave. Temperature of Final Set 23.5 ( 7 ) Immersed - Atmosphere Temperature of Water 21 . 5 Ave. Temperature of Final Set 23 »5 Ave, Vapor Pressure of Final Set 0.533- ■ l \ - 20 - (c) Compression Test: Setting Acid Actual Load Average Conditions c .c . in lbs. (1) Atm. - Atm. 40 2280 ,2500 ,2500 ,2700 ,2500 2496 37 2720 , 3000 , 1920 , 1840 ,Bend 2537 (2) Atm. - High Vap, 40 2200 ,2500,2800,2200,2900 2520 37 2860 ,Bend , 2800 , 2700 ,1800 2727 ( 3 ) Atm. - Immersed 40 2300 ,1800 ,1750,1700 ,2100 1930 37 2500,1800 2150 ( 4 ) Atm. - Low Vap. 40 3400 , 3100 , 3260,3300 3265 37 2000,2000,(1900,1800 short] ) 1925 (5) High Vap. - Atm. 40 2300 ,2200 ,2000 ,2500 ,2800 2360 37 2500,2000,1500 2000 (6) Low Vap. - Atm. 40 3500 , 3300,3250 3350 37 3200,3200,2500,2900 2950 ( 7 ) Immersed - Atm. 40 1400 , 1700 , 1400,1000 1375 37 1100, 700,1700 (short) 1166 - 21 - (d) Average strength according to proportion of acid to Powder; Average strength of mixtures containing 40 c.c, 24/1 lbs " " « « « 37 c.c. 2216 (e) Highest Average Crushing Strength expressed in lbs / sq.in. is 6130 #/ sq. in* Do Conclusions,, (a) Variation in the proportion of dry material to acid solu- tion indicate that; (1) Very small changes in the proportion of powder to acid solution give marked differences in the plasticity of the mixture . (2) The results obtained are not consistent enough to warrant any conclusions as to the effect upon the compressive strength • (b) Variations in the moisture and vapor pressure treatment during the period of initial set indicate that; (1) The highest crushing strength is developed under con- ditions with the lowest vapor pressure, (2) The setting of the cement under water does not pro- duce cements of sound structure or of high crushing strength. This indicates that these cements are not hydraulic in their reactions, . (c) Variations of the moisture and vapor pressure treatment during the period of final set indicate that; (l) The highest crushing strength is developed under con- ditions of low vapor pressure. • * ' > * * - 22 - (2) The final setting under water gives lower values in crushing strength than are found with the same cements set under moist or dry conditions, (d) The study with variations both in the period of initial set (24 hours) and in final set (6 days) indicate that! (1) The period of initial set is the period containing the principle chemical change, as shown by variations in values under these conditions, (2) The chemical changes are not complete, however, in the initial period but give all evidence of undergoing pro- gressive change over a period of days or months after being mixed, VI Summary . Compositions of Lime, Alumina and Silica producing workable phosphate silicate cements were found to lie in the field of crys- talization S, AS, CA2S. The most promising cement is developed from the composition! CaO 10fo Al^O-^ 4S Si0 2 42 The character of raw materials used has a very marked effect upon properties and behavior of the cement. Florida Kaolin as the clay material gives the highest crushing strength, and the most desirable working properties, as well as good color. The speed of the initia.l setting reaction increases with in- crease in percentage of CaO in the composition. A high temperature of calcination is the most satisfactory ' . ■ . -23“ for the preparation of the powders, and indicates that fusion of the materials may be desirable. The proportion of powder to acid solution has no particular relation to the physical properties of the hardened cement. The principal chemical change takes place within the first 24 hours of setting, but a progressive chemical action continues for some days. The cements with the highest crushing strength result from setting, both during the initial and final period, under conditions of low vapor pressure. This is in agreement with statements made by Dr*s W. and D. Asch3. The highest crushing strength obtained on a 7 day test was slightly over 6000 lbs. / sq. in. This was on a cylinder made from the composition: CaO 10$ A1 2°3 48 SiG 2 42 The clay material was Florida Kaolin, and the conditions of setting were: Initial set under low vapor pressure. Final set under atmospheric conditions. 3 Silicates in Chemistry and Commerce. Dr*s W. and D. Asch Hardening of Dental Cements p.218.