VOL. 9, PP. 441-451 (1975)
J. BIOMED. MATER. RES.
Thermodifferential Analysis of Ceramic Implants L. CINI, Faculty of Engineering, University of Bologna, A. PIZZOFERRATO, C. TRENTANI, S. SANDROLINI, and M. PALTRINIERI, Rizzoli Orthopaedic Institute, Bologna, Italy
Summary Ceramic pelletes (in Alsoa 92% and in calcium aluminate) implanted into the muscles of rabbits for varying periods of time, and samples of similar material, not implanted but mixed with organic substances, have been analyzed by means of thermal differential analysis (TDA). Which hydrated calcium aluminates are formed from implants after contact with the tissure fluid, has been established. It has also been ascertained that the ceramic implants absorb organic substances from the tissue fluid, which, from a first evaluation, seem to be composed of simple molecules. The results of this initial inquiry have provided positive indications of the utility of TDA to assist in elucidating the phenomena that occur between implant and tissue.
INTRODUCTION Several experiments dealing with ceramic type implant material have utilized calcium aluminates and products based on porous or nonporous These experiments were performed, mainly using x-rays, electronic and optic microscopy, and an electronic scanning microprobe, to examine implant behavior regarding tissues, morphological aspects of the tissue-implant interface, biocompatibility, etc. In this study, results of the experiments that have been performed by means of the Thermal Differential Analysis (TDA) on samples of calcium aluminate and a 92% porous alumina product after implant in rabbits,' are reported and discussed. 441 @ 1975 by John Wiley & Sons, Inc.
442
CINI ET AL.
EXPERIMENTAL Calcium aluminate was obtained from CaC03 and amorphous A1203. After mixing, a precalcination was performed at 1100°C for 36 hr. The calcined product was ground again, humidified with an aqueous solution containing an alloying element and a lubricant* and granulated. The resultant powder was pressed into disks of 4 cm diam and 1.2 cm thick, these disks were successively heated at 1250°C for 12 hr. Their chemical composition was 43Y0 CaO and 57y0 A1203;the resulting phases (as checked by the Debye method) were 12Ca0. 7Al2O3and Ca0.A1203,which agreed with the phase diagram for A1203-Ca0. On the basis of the phase diagram the calculated percentages of the two phases were 55% Ca0.A1203 and 45% 12Ca0.7Al2O3. The material had a relative density of 1.65 g/cm3, water absorption of 26%, and apparent porosity of 45.8%. The distribution of pores, determined by a mercury porosimeter, is shown in Figure la. From Figure 1, it appears that 60% of the pores had a diameter of between 1 and 2 p , 20% had a diameter of less than 1 p , and 20% had a diameter of greater than 2 p . The ceramics $&?yo A1203 derived from a crystalline alumina powder and clay** and activated by the tensioactive (foam) method. The material obtained was heated at 1450°C: the samples were formed by parallelopipeds of 10 cm X 10 cm X 5 cm. This material had a relative density of 1 g/cc and an apparent porosity of 75%. The distribution of pores, evaluated by cathetometric measurements is shown in Figure lb. From this diagram it appears that the diameter of pores varied almost linearly from 10 to 100 p . From the samples thus prepared, cylindrical pellets were obtained 10 mm long and 5 mm in diam, with a center hole of about 1 mm (Fig. 2). These pellets were implanted in adult male laboratory rabbits (Oryctolagus cuniculus), in their paravertebral muscles. Figure 3 shows an x-ray of the implant. *The alloying element was Rhodoviol; the lubricant, Scurol, was supplied by the Rhone-Poulenc Company. ** The alumina was from the Montecatini Company and the clay from plastic British clay (BWS Brand).
CERAMIC IMPLANTS
443
After a period of 7, 15, and 30 days the animals were sacrificed. The explanted pellets underwent first a manual cleaning, to remove the adhering muscle parts, then were dried at 50"C, following which a second cleaning was performed. The cleaned pellets were then ground and their powder dried at 50°C for 12 hr. On this material the TDA was performed using a Leeds-Northrup Analyzer, with a thermal gradient of lO"C/min, in a static air atmosphere. Figure 4 shows the TDA curves pertinent to the implanted pellets, and to some samples formed by mixtures of ceramic material with coagulated human blood, with glucose, with muscle tissue, and with charcoal.
ENDOTHERMIC EFFECTS It has to be assumed that while the endothermic and exothermic effects that are found in samples of ceramic material with 92% alumina are only due to the adsorption phenomenon of tissue fluids by the implant, for implants containing calcium aluminate, it is necessary to keep in mind that this material is "reactive" with reference t o the tissue fluid. Klawitter and Hulbert2 established by microscopic and microradiographic tests that in implants of calcium aluminate, in the interface, degradation products of about 10 p thickness were formed. These products must have been composed of hydrated calcium aluminates. On the basis of studies by other researchers on hydratation of calcium aluminates, they reported that the only stable phase among the various CAH (C = CaO; A = A1203;H = H20) is C3AH6, which, however, in the presence of organic substances (like sugars) is formed with delay. Hentrich and Graves3 also took into consideration the formation of C3AH6 from anhydrous compounds, a formation that would provide increased porosity of the implant. The TDA curves pertinent to the pellets of calcium aluminate after the implant show three endothermic effects a t (I) 125°C; (2) 190-210°C; and (3) 300°C (curves a, b, c). These effects, on the basis of the literature on dehydratation of hydrated calcium aluminates8p9can be ascribed to (1) C&&, (2) C ~ A H Nand , (3) AH3: the peak corresponding to C3AH6 is not detected between 330°C and 350°C.
444
GIN1 ET AL.
CERAMIC IMPLANTS
445
446
CINI ET AL.
(4 Fig. 2 (continued)
The tissue humors, in contact with the implant, can stabilize “metastable” calcium aluminates. This can be explained by the fact that the implant remains in constant contact with the aqueous solution. Turriziani,’O reporting on the equilibrium of CaOAI2O3-H2O, states that CIAHI, and CzAHs remain stable for long periods of time in aqueous suspension. I n other words, the present conditions exhibit a different water/anhydrous calcium aluminate ratio than that for a true concrete mortar. Curve c corresponds t o our sample of calcium aluminate maintained in water for 15 days at 37°C (temperature of the rabbit), dried a t 50”C, and then mixed with human lyopholized blood. It shows three peaks a t the same temperatures as the implants a and b. The formation of an interface made by these hydrated compounds, not only determines a variation of the porosity, but of the morphology of the contact surface as well. It is sufficient to assume that among
CERAMIC IMPLANTS
447
(b) Fig. 2 . Photograph of ceramic pellets. (a) Calcium aluminates. 92Oj, A1,0,.
(b) Porous
the reaction products formed, aluminium hydroxide takes the form of a “gel.” This new structure governs the process of “diffusion,” from the macroscopic level to the ionic and molecular levels, of the tissue fluid in the implant. It has been established that the diffusion process that controls the bonding between implant and tissue. The endothermic effect that is found in the implants containing 92% alumina a t about 120°C (curves d and e) ‘is obviously determined by substances which give origin to the exothermic effect at high temperature. In the samples of calcium aluminate, this effect is not seen, as it coincides with that of dehydratation.
EXOTHERMIC EFFECTS The interpretation of the exothermic effects is complex unless it is limited to the fact that materials of this typc, during the period of
448
CINI ET AL.
Fig. 3. Roentgenograph of ceramic pellets implanted in the paravertebral muscles of rabbits.
implantation, absorb organic substances from the tissue fluid, which generate heat during oxidation. As is known, the variables that intervene in the kinetics of reactions of oxidation of this type-solid-gas reactions-are many," and they influence the behavior of the TDA curves. In fact, the temperature at which the exothermic peak occurs, and its shape, are influenced by the thermal gradient of raising temperature, by the partial pressure of oxygen, by the conditions of the fluid, by the diffusion of air or oxygen in the sample, etc. These factors explain the reason why the exothermic peaks in the various curves occur at varying temperatures. In order to establish what substances determine the exothermic effects, tests have been made on mixtures formed from ceramic material mixed with coagulated human blood, animal muscle, glucose and charcoal. In the TDA curves, we notice the presence of two exothermic peaks: one a t about 500°C and another a t about 900°C. The peak at 5OO0Cis very evident in calcium aluminate implanted for 7 days and for 15 days, and in the pellet of 920/, alumina implanted for 7 days: it is less evident, but clearly present, in the pellet implanted for 30 days. This peak is completely absent in the samples
Fig. 4. Curves of Thermal Differential Analysis. (a) Implant of calcium aluminate after 7 days. (b) Implant of calcium aluminate after 15 days. (c) Mixture of calcium aluminate in aqueous suspension for 15 days and dried a t 50°C mixed with 10% coagulated human blood. (d) Implant of 92% A1203 after 7 days. (e) Implant of 92% A1203 after 1 month. (f) Mixture of 92% A1203 and 10% coagulated human blood. (9) Mixture of 50% muscle dried at 50°C and 50% of 92% A1z03. (h) Sample of 92% A1203 imbibed in a solution of 22% glucose and dried a t 50°C. (i) Mixture of 92% A1203 with 10% charcoal.
450
CINI ET AL.
of alumina artificially mixed with human lyophilized blood, animal muscle, glucose or charcoal. I n addition, the exothermic peak at about 900°C can be found in all the artificial mixtures of 92% alumina and calcium aluminatc, as well as in 92y0 alumina implanted for 30 days. However, it cannot be detected in 92% alumina implanted for 7 days, or in calcium aluminate after 7 and 15 days implantation. This demonstrates that, undcr the experimental conditions in which the TDA were performed, both the organic substances absorbed by the implanted material and the ones contained in the artificial mixtures give rise t o a carbonic residue which oxidates a t about the same temperature, that is, a t 900°C. It is not yet possible to definitively interpret the peak a t 500°C which appears after a shorter time of implantation. We can interpret the endothermic peaks pertinent t o the artificial mixtures as being connected t o the various proteinic or carbohydratic substances that partially compose the mixture. Neither proteins nor carbohydrates should be present in the implanted materials. However the presence of a n endothermic peak a t about 120"C, can be attributed to water chemically bound to substances that although not having the molecular complexity of protcins or carbohydrates, are obviously present, as demonstrated by the exothermic peak a t 900°C. I n conclusion, i t might be possible to state, hypothetically, that the implanted ceramic materials function, a t least for a short timc, as ultrafilters and allow only low molecular wcight organic substances to pass through the tissue humors. This work was performed under the support of the National Committee of Researches (C.N.R.), Rome, Italy.
References 1. S. F. Hulbert, F. A. Young, R. S. Mathews, J. J. Klawitter, C. I).Talbert, and F. H. Stelling, J . Biomed. Muter. Res., 4(3), 433 (1970). 2. J. J. Klawitter and S. F. Hulbert, J . Biomed. Muter. Res., Symposium 2, 161 (1971). 3. R. L. Hentrich and G. A. Graves, J. Biomed. Muter. Res., 5(1), 25 (1971). 4. S. F. Hulbert, J. J. Klawitter, and R. B. Leonard, Ceramics in Severe Environments Materials Science Research, Vol. 5, Plenum Press, New York, 1972, pp. 417-434. 5. S. F. Hulbert, F. M. King, and J. J. Klawitter, J . Biomed. Muter. Res., Symposium 2, 69 (1971).
CERAMIC IMPLANTS
45 1
6. L. L. Hench, H. A. Paschall, W. C. Allen, G. Piotrowski, Report no. 3, Contract no. 17-70-C-0001, University of Florida, Gainesville, (1972). 7. L. Cini, M. Paltrinieri, A. Pizzoferrato, S. Sandrolini, and C. Trentani, Chir. Org. movimento, 60 (4),423 (1971). 8. H . Lehman and K. J. Leers, Tonind. Zit., 87(2), 29 (1963). 9. E. Crepaz and A. Raccanelli, L'Ind. Ital. Cemento, 519 (1963). 10. R. Turriziani, The Chemistry of Cements, Vol. 2, Academic Press, New York, 1974, pp. 223-283. 11. R. G. Mackenzie, Differential Thermal Analysis, Vols. I and 2, Academic Press, New York, 1973.
Received September 20, 1974
J. BIOMED. MATER. RES.
Thermodifferential Analysis of Ceramic Implants L. CINI, Faculty of Engineering, University of Bologna, A. PIZZOFERRATO, C. TRENTANI, S. SANDROLINI, and M. PALTRINIERI, Rizzoli Orthopaedic Institute, Bologna, Italy
Summary Ceramic pelletes (in Alsoa 92% and in calcium aluminate) implanted into the muscles of rabbits for varying periods of time, and samples of similar material, not implanted but mixed with organic substances, have been analyzed by means of thermal differential analysis (TDA). Which hydrated calcium aluminates are formed from implants after contact with the tissure fluid, has been established. It has also been ascertained that the ceramic implants absorb organic substances from the tissue fluid, which, from a first evaluation, seem to be composed of simple molecules. The results of this initial inquiry have provided positive indications of the utility of TDA to assist in elucidating the phenomena that occur between implant and tissue.
INTRODUCTION Several experiments dealing with ceramic type implant material have utilized calcium aluminates and products based on porous or nonporous These experiments were performed, mainly using x-rays, electronic and optic microscopy, and an electronic scanning microprobe, to examine implant behavior regarding tissues, morphological aspects of the tissue-implant interface, biocompatibility, etc. In this study, results of the experiments that have been performed by means of the Thermal Differential Analysis (TDA) on samples of calcium aluminate and a 92% porous alumina product after implant in rabbits,' are reported and discussed. 441 @ 1975 by John Wiley & Sons, Inc.
442
CINI ET AL.
EXPERIMENTAL Calcium aluminate was obtained from CaC03 and amorphous A1203. After mixing, a precalcination was performed at 1100°C for 36 hr. The calcined product was ground again, humidified with an aqueous solution containing an alloying element and a lubricant* and granulated. The resultant powder was pressed into disks of 4 cm diam and 1.2 cm thick, these disks were successively heated at 1250°C for 12 hr. Their chemical composition was 43Y0 CaO and 57y0 A1203;the resulting phases (as checked by the Debye method) were 12Ca0. 7Al2O3and Ca0.A1203,which agreed with the phase diagram for A1203-Ca0. On the basis of the phase diagram the calculated percentages of the two phases were 55% Ca0.A1203 and 45% 12Ca0.7Al2O3. The material had a relative density of 1.65 g/cm3, water absorption of 26%, and apparent porosity of 45.8%. The distribution of pores, determined by a mercury porosimeter, is shown in Figure la. From Figure 1, it appears that 60% of the pores had a diameter of between 1 and 2 p , 20% had a diameter of less than 1 p , and 20% had a diameter of greater than 2 p . The ceramics $&?yo A1203 derived from a crystalline alumina powder and clay** and activated by the tensioactive (foam) method. The material obtained was heated at 1450°C: the samples were formed by parallelopipeds of 10 cm X 10 cm X 5 cm. This material had a relative density of 1 g/cc and an apparent porosity of 75%. The distribution of pores, evaluated by cathetometric measurements is shown in Figure lb. From this diagram it appears that the diameter of pores varied almost linearly from 10 to 100 p . From the samples thus prepared, cylindrical pellets were obtained 10 mm long and 5 mm in diam, with a center hole of about 1 mm (Fig. 2). These pellets were implanted in adult male laboratory rabbits (Oryctolagus cuniculus), in their paravertebral muscles. Figure 3 shows an x-ray of the implant. *The alloying element was Rhodoviol; the lubricant, Scurol, was supplied by the Rhone-Poulenc Company. ** The alumina was from the Montecatini Company and the clay from plastic British clay (BWS Brand).
CERAMIC IMPLANTS
443
After a period of 7, 15, and 30 days the animals were sacrificed. The explanted pellets underwent first a manual cleaning, to remove the adhering muscle parts, then were dried at 50"C, following which a second cleaning was performed. The cleaned pellets were then ground and their powder dried at 50°C for 12 hr. On this material the TDA was performed using a Leeds-Northrup Analyzer, with a thermal gradient of lO"C/min, in a static air atmosphere. Figure 4 shows the TDA curves pertinent to the implanted pellets, and to some samples formed by mixtures of ceramic material with coagulated human blood, with glucose, with muscle tissue, and with charcoal.
ENDOTHERMIC EFFECTS It has to be assumed that while the endothermic and exothermic effects that are found in samples of ceramic material with 92% alumina are only due to the adsorption phenomenon of tissue fluids by the implant, for implants containing calcium aluminate, it is necessary to keep in mind that this material is "reactive" with reference t o the tissue fluid. Klawitter and Hulbert2 established by microscopic and microradiographic tests that in implants of calcium aluminate, in the interface, degradation products of about 10 p thickness were formed. These products must have been composed of hydrated calcium aluminates. On the basis of studies by other researchers on hydratation of calcium aluminates, they reported that the only stable phase among the various CAH (C = CaO; A = A1203;H = H20) is C3AH6, which, however, in the presence of organic substances (like sugars) is formed with delay. Hentrich and Graves3 also took into consideration the formation of C3AH6 from anhydrous compounds, a formation that would provide increased porosity of the implant. The TDA curves pertinent to the pellets of calcium aluminate after the implant show three endothermic effects a t (I) 125°C; (2) 190-210°C; and (3) 300°C (curves a, b, c). These effects, on the basis of the literature on dehydratation of hydrated calcium aluminates8p9can be ascribed to (1) C&&, (2) C ~ A H Nand , (3) AH3: the peak corresponding to C3AH6 is not detected between 330°C and 350°C.
444
GIN1 ET AL.
CERAMIC IMPLANTS
445
446
CINI ET AL.
(4 Fig. 2 (continued)
The tissue humors, in contact with the implant, can stabilize “metastable” calcium aluminates. This can be explained by the fact that the implant remains in constant contact with the aqueous solution. Turriziani,’O reporting on the equilibrium of CaOAI2O3-H2O, states that CIAHI, and CzAHs remain stable for long periods of time in aqueous suspension. I n other words, the present conditions exhibit a different water/anhydrous calcium aluminate ratio than that for a true concrete mortar. Curve c corresponds t o our sample of calcium aluminate maintained in water for 15 days at 37°C (temperature of the rabbit), dried a t 50”C, and then mixed with human lyopholized blood. It shows three peaks a t the same temperatures as the implants a and b. The formation of an interface made by these hydrated compounds, not only determines a variation of the porosity, but of the morphology of the contact surface as well. It is sufficient to assume that among
CERAMIC IMPLANTS
447
(b) Fig. 2 . Photograph of ceramic pellets. (a) Calcium aluminates. 92Oj, A1,0,.
(b) Porous
the reaction products formed, aluminium hydroxide takes the form of a “gel.” This new structure governs the process of “diffusion,” from the macroscopic level to the ionic and molecular levels, of the tissue fluid in the implant. It has been established that the diffusion process that controls the bonding between implant and tissue. The endothermic effect that is found in the implants containing 92% alumina a t about 120°C (curves d and e) ‘is obviously determined by substances which give origin to the exothermic effect at high temperature. In the samples of calcium aluminate, this effect is not seen, as it coincides with that of dehydratation.
EXOTHERMIC EFFECTS The interpretation of the exothermic effects is complex unless it is limited to the fact that materials of this typc, during the period of
448
CINI ET AL.
Fig. 3. Roentgenograph of ceramic pellets implanted in the paravertebral muscles of rabbits.
implantation, absorb organic substances from the tissue fluid, which generate heat during oxidation. As is known, the variables that intervene in the kinetics of reactions of oxidation of this type-solid-gas reactions-are many," and they influence the behavior of the TDA curves. In fact, the temperature at which the exothermic peak occurs, and its shape, are influenced by the thermal gradient of raising temperature, by the partial pressure of oxygen, by the conditions of the fluid, by the diffusion of air or oxygen in the sample, etc. These factors explain the reason why the exothermic peaks in the various curves occur at varying temperatures. In order to establish what substances determine the exothermic effects, tests have been made on mixtures formed from ceramic material mixed with coagulated human blood, animal muscle, glucose and charcoal. In the TDA curves, we notice the presence of two exothermic peaks: one a t about 500°C and another a t about 900°C. The peak at 5OO0Cis very evident in calcium aluminate implanted for 7 days and for 15 days, and in the pellet of 920/, alumina implanted for 7 days: it is less evident, but clearly present, in the pellet implanted for 30 days. This peak is completely absent in the samples
Fig. 4. Curves of Thermal Differential Analysis. (a) Implant of calcium aluminate after 7 days. (b) Implant of calcium aluminate after 15 days. (c) Mixture of calcium aluminate in aqueous suspension for 15 days and dried a t 50°C mixed with 10% coagulated human blood. (d) Implant of 92% A1203 after 7 days. (e) Implant of 92% A1203 after 1 month. (f) Mixture of 92% A1203 and 10% coagulated human blood. (9) Mixture of 50% muscle dried at 50°C and 50% of 92% A1z03. (h) Sample of 92% A1203 imbibed in a solution of 22% glucose and dried a t 50°C. (i) Mixture of 92% A1203 with 10% charcoal.
450
CINI ET AL.
of alumina artificially mixed with human lyophilized blood, animal muscle, glucose or charcoal. I n addition, the exothermic peak at about 900°C can be found in all the artificial mixtures of 92% alumina and calcium aluminatc, as well as in 92y0 alumina implanted for 30 days. However, it cannot be detected in 92% alumina implanted for 7 days, or in calcium aluminate after 7 and 15 days implantation. This demonstrates that, undcr the experimental conditions in which the TDA were performed, both the organic substances absorbed by the implanted material and the ones contained in the artificial mixtures give rise t o a carbonic residue which oxidates a t about the same temperature, that is, a t 900°C. It is not yet possible to definitively interpret the peak a t 500°C which appears after a shorter time of implantation. We can interpret the endothermic peaks pertinent t o the artificial mixtures as being connected t o the various proteinic or carbohydratic substances that partially compose the mixture. Neither proteins nor carbohydrates should be present in the implanted materials. However the presence of a n endothermic peak a t about 120"C, can be attributed to water chemically bound to substances that although not having the molecular complexity of protcins or carbohydrates, are obviously present, as demonstrated by the exothermic peak a t 900°C. I n conclusion, i t might be possible to state, hypothetically, that the implanted ceramic materials function, a t least for a short timc, as ultrafilters and allow only low molecular wcight organic substances to pass through the tissue humors. This work was performed under the support of the National Committee of Researches (C.N.R.), Rome, Italy.
References 1. S. F. Hulbert, F. A. Young, R. S. Mathews, J. J. Klawitter, C. I).Talbert, and F. H. Stelling, J . Biomed. Muter. Res., 4(3), 433 (1970). 2. J. J. Klawitter and S. F. Hulbert, J . Biomed. Muter. Res., Symposium 2, 161 (1971). 3. R. L. Hentrich and G. A. Graves, J. Biomed. Muter. Res., 5(1), 25 (1971). 4. S. F. Hulbert, J. J. Klawitter, and R. B. Leonard, Ceramics in Severe Environments Materials Science Research, Vol. 5, Plenum Press, New York, 1972, pp. 417-434. 5. S. F. Hulbert, F. M. King, and J. J. Klawitter, J . Biomed. Muter. Res., Symposium 2, 69 (1971).
CERAMIC IMPLANTS
45 1
6. L. L. Hench, H. A. Paschall, W. C. Allen, G. Piotrowski, Report no. 3, Contract no. 17-70-C-0001, University of Florida, Gainesville, (1972). 7. L. Cini, M. Paltrinieri, A. Pizzoferrato, S. Sandrolini, and C. Trentani, Chir. Org. movimento, 60 (4),423 (1971). 8. H . Lehman and K. J. Leers, Tonind. Zit., 87(2), 29 (1963). 9. E. Crepaz and A. Raccanelli, L'Ind. Ital. Cemento, 519 (1963). 10. R. Turriziani, The Chemistry of Cements, Vol. 2, Academic Press, New York, 1974, pp. 223-283. 11. R. G. Mackenzie, Differential Thermal Analysis, Vols. I and 2, Academic Press, New York, 1973.
Received September 20, 1974
