J. BIOMED. MATER. RES.
VOL. 10, PP. 777-788 (1976)
Polymer-Ceramic Composite for Tooth-Root Implant* A. R. GREENBERG and IHAB KAMEL, Department of Materials Engineering and the Biomedical Engineering Program, Drexel University, Philadelphia, Pennsylvania 19104
Summary A new polymer-ceramic composite suitable for tooth-root implants has been developed in this study. This material exhibited the desirable combination of good mechanical properties, controlled porosity, and ease of processing. A thermal processing technique was utilized to polymerize acrylic acid (AA) in the presence of either 0.3 or 0.05 @ alumina particles. Porosity and pore size distribution were influenced by the alumina particle size and the processing technique. For a 50 vol % AA solution, the composite had an average compressive strength of 18,000 psi and 38% porosity when 0.3 I.( filler particles were used. I n comparison, the 0.05 alumina-filled composite had an average compressive strength of 28,OOO.psi with a 15% porosity. Data on the physical and structural characteristics of the composite are presented in this study. Based on these results, the composite material shows good potential for use in tooth-root implants as well as other orthopedic implant applications.
INTRODUCTION Although various workers have adopted somewhat different criteria for the successful tooth-root implant, it has generally been that such an implant must be nontoxic, be noncarcinogenic, provide for stable retention, and display durability in the body environment so that adequate physical and mechanical properties are retained. In the past, various materials and implant designs have been tried in an attempt to meet the above The problems associated with each of these implants have been well documented in the literature.3s10 Some of the more serious dis-
* Presented a t the Seventh Annual International Biomaterials Symposium, Clemson University, Clemson, South Carolina, April 26-30, 1975. 777 @ 1976 by John Wiley & Sons, Inc.
778
GREENBERG AND KAMEL
advantages associated with each of these approaches have included leaching of unreacted monomer into surrounding tissues from porous l2 the brittleness self-curing poly(methy1 methacrylate) systems, and high notch sensitivity of ~ e r a m i c s ,reaction ~ ~ ~ ~ of , ~metals ~ with the body environment,13 and loosening of screws and blades in V ~ V O . ~ I n a n attempt t o overcome some of these problems, a polymerceramic composite is introduced in this study which meets the four general criteria for implantable materials. The composite which has been developed consists of an aluminum oxide filler in a poly(acrylic acid) matrix. I n regard t o the basic compatibility of the composite, it will suffice t o say that previous investigators have shown alumina and the polycarboxylates to be both nontoxic and noncar~inogenic.'~-'~ Recent work has demonstrated that a porous structure which allows tissue ingrowth into the tooth-root implant can provide the needed supportive retention and eliminate epithelial invagination and fibrous encapsulation of the implant.4t12 This investigation has shown that not only can the poly(acry1ic acid) [PAAI-alumina composite be produced with a fine interconnected porosity, but that the material has a combination of physical and mechanical properties that merit its testing in implant applications. The purpose of this paper is to present some of the physical and mechanical properties as well as the structural characteristics of the PAA-alumina system.
MATERIALS AND METHODS Materials Commercially available acrylic acid monomer (Polysciences Inc., Warrington, Pa.) inhibited with 200 ppm MEHQ was used without further purification. Metallurgical-grade alumina powder (Buehler Ltd., Evanston, Illinois) of 0.05 and 0.3 p average particle size was used as received from the manufacturer.
Composite Preparation Aqueous solutions of 50 vol % acrylic acid containing 0.2 vol yo MEK-peroxide were mixed with the powdered alumina t o form a homogeneous slurry. The slurry was poured into glass ampoules, covered, and heated in a 70°C oven for 4 hr t o polymerize the acrylic
POLYMER-CERAMIC COMPOSITE FOR TOOTH-ROOT IMPLANT 779
acid. Two methods were employed to ensure the homogeneity of the sample and to change the outcome of the final porosity. These procedures included 1) allowing the filler to settle by gravity, and 2) rotating the samples in the vertical position during polymerization. Following polymerization, the ingots were removed from the test tubes and dried for 20 hr a t 70°C. In addition, some samples were further heat-treated at 150°C for 12-15 hr in order to remove the tightly bound water remaining in the sample.
Mechanical Testing The dry samples were machined into cylinders of 0.500 in. in diameter and 1.00 in. in length. These specimens were tested in compression on an Instron Universal testing machine. All tests were conducted at crosshead speeds of 0.005 in./min. In accordance with ASTM specification D695, the yield point was taken as the first point on the stress-strain diagram a t which a n increase in strain occurred without an increase in stress. I n addition, the modulus of elasticity was taken as the ratio of stress to corresponding strain below the proportional limit.
Solubility and Swelling Measurements Thin wafers were sectioned from each ingot and immersed in either boiling water or 1 N KOH a t 80°C. Initial estimates of solubility were determined by measuring the percent weight loss of the specimen as a function of immersion time. In addition, swelling was determined by measuring the percent weight change of the wafers as a result of soaking in a water bath at room temperature. Furthermore, the volume changes which occurred during swelling were also measured. Estimates of the initial sample porosity were then determined by calculating the difference between the weight and volume increases (AW - AV) of the sample:
% porosity
=
lOO(AW - AV)/Vo
(1)
where V ois the dry volume of the sample.
Microscopic Examination Specimens were immersed in liquid nitrogen until thermal equilibration and then fractured. The fracture surfaces were made conductive by application of a chromium coating by the vapor
780
GREENBERG AND KAMEL
deposition method. The microstructures of the coated specimens were then examined under the scanning electron microscope.
Thermal Analysis The thermal decomposition pattern of the polymeric matrix was determined by heating a small sample of the composite (10-15 mg) in a du Pont 950 thermogravimetric analyzer at a rate of 5"C/min. Either nitrogen or air was used as the purge medium. The decomposition of the polymer was completed before lOOO"C, and the ratio of polymer/ceramic in the sample was obtained directly.
RESULTS Because of the intimate contact between an implanted material and the body fluids, water sorption and polymer solubility are important factors in determining material performance. Swelling data for the various composites are plotted in Figure 1. It is quite
0 03 0 0 05 (unheat-treated)
48
96
144
TIME ( h r J
Fig. 1. Effect of heat treatment and particle size on swelling of PAA-alumina composites in water at 20°C.
POLYMER-CERAMIC COMPOSITE FOR TOOTH-ROOT IMPLANT 781
apparent from these curves that the 0.3 p alumina composites swelled much faster than and reached about three times the swelling level of the 0.05 p alumina composites. These results would be expected inasmuch as water sorption is dependent on both the pore size and the total porosity and, as will be demonstrated, the 0.3 I.C material is greater in both of these characteristics. In addition, the nonheat-treated 0.05 p composites evidenced a significantly higher degree of swelling than their heat-treated counterparts. Although PAA is completely water-soluble, the composites were not. After 24 hr in boiling water, the 0.05 p specimens evidenced a weight loss of only about 3%, while the larger particle-filled composites had a loss of approximately 12%. This difference between the two particle sizes is, however, somewhat magnified because the measurements also included the loss of the relatively dense alumina. When the samples were placed in the KOH a t 8O"C, the 0.3 p specimens fell apart and partially dissolved after 24 hr ; however, the 0.05 p samples, both heat-treated and untreated, were converted into a gel-like condition and yet still maintained the original sample shape even after 72 hr. This indicates that the particle/matrix interaction is of a secondary rather than primary nature. However, the magnitude of this secondary bonding is large enough to enable the filler to act as a crosslinking agent. It is to be expected that this interaction should be inversely related to the filler particle size. Estimates of initial composite porosity were calculated by the use of eq. (1). The large filler samples were found to have a porosity of about 38% while the 0.05 p samples evidenced a porosity of approximately 15%. The processing technique was varied in order to determine whether the structure of the 0.05 p samples could be modified. This was accomplished by altering the settling rate. Two techniques were used: settling by gravity, and settling by rotation of the slurry. One measure of the effectiveness of these treatments was the polymerceramic ratio as determined by thermal analysis. The results showed that the specimens prepared by rotation and those produced without rotation had final ceramic concentrations of 44 and 36 w t %, respectively. No variation in processing technique was conducted for the 0.3 p specimens. All of the larger filler samples were prepared without rotation. Thermal analysis of these samples indicated a ceramic concentration of 53 wt yo.
782
GREENBERG AND KAMEL
Differences in particle size, porosity, and processing technique were reflected in the structural characteristics of the respective samples. Figure 2 shows scanning electron micrographs of the 0.3 I.( sample. The surface evidences many elevations and depressions, and the
(b)
Fig. 2. SEM micrographs of fracture surface of 0.3 p alumina-PAA composite. A: 300 X ; B: 1000 X.
POLYMER-CERAMIC COMPOSITE FOR TOOTH-ROOT IMPLANT 783
structure can be characterized as homogeneous but highly porous with a n average pore size of 25 p. Figure 3 shows the structure of the 0.05 p sample produced by the rotation technique. The structure is quite different than that of the
(b)
Fig. 3. SEM micrographs of fracture surface of 0.05 p alumina-PAA composite. (a) 300 X; (b) 1000 X.
GREENBERG AND KAMEL
784
0.3 p specimen. Although the structure still contains elevations and depressions, the effects of closer packing are evident. Most of the porosity is in the form of large shallow craters, with relatively few small but deeper pores. The latter was taken to be the minimum interconnecting porosity with a measured diameter of approximately 10 p. Examination of the 0.05 p samples produced without rotation indicated that the structure had been altered. These samples had more of the wider and shallower craters and less of the deep interconnecting pores. Compared to the rotated samples the structure was considerably looser, and the interconnecting pores were somewhat larger, having an average diameter of about 15 p . The mechanical properties of the composites are summarized in Table I and plotted in the form of stress-strain curves in Figure 4. These results clearly show that the heat treatment removed the plasticizing effect of the water such that the yield strength and elastic modulus are increased while the elongation at yield is significantly decreased. The yield point of the nonheat-treated samples could not be defined according to ASTM specification D695 because the stress did not level off and the samples did not fracture. The particle size of the filler is also seen to influence the mechanical properties. The 0.05 p filler had about a 65% increase in compressive yield strength compared to the 0.3 p particle composite. However, some of these differences in the mechanical behavior are influenced by both the increased porosity and filler content of the 0.3 p specimens. TABLE I Mechanical Properties of the PAA-Alumina Composites
Composite type (I.r)
0.3 0.05 (nonheat-treated) 0.05 a
Mean compressive yield stress (psi) X10-3
Mean strain at yield
Mean elastic modulus (psi)
(%I
x 10-5
18.10 9.80. 27.80
8.7 11.0 4.0
4.25 2.50 8.20
All stresses determined a t 11% strain since no definite yield point was observed.
POLYMER-CERAMIC COMPOSITE FOR TOOTH-ROOT IMPLANT 785
'
A005
I
0 0 3
25
0 0 05 (unheot-treoted)
20 ?
D
__-a
15
rn
v) v)
U
10
tv)
5
10
5 STRAIN
1x1
Fig. 4. Effect of heat treatment and particle size on stress-strain behavior of PAA-alumina composites.
DISCUSSION Significant differences in composite behavior have been shown to occur when either the 0.05 or the 0.3 p alumina particles were used. The 0.3 p composites evidenced a substantial increase in swelling rates and levels over the 0.05 p composites. In addition, a twofold increase in porosity occurred with the larger filler using the same processing technique. Furthermore, although PAA is infinitely soluble in water, the results of the solubility tests have shown only minimal weight loss by the composites in boiling water. This indicates that the filler must act as a strong crosslinking agent.in this system. The polymer-filler bond, although of a secondary chemical nature, becomes very effective when the ceramic particles are small. This can be explained on the basis of the average distance between particles at a constant volume fraction of filler. If the particle
786
GREENBERG AND KAMEL
diameter is substantially reduced, the number concentration of particles will be very large; hence, the average distance between them very small. If this distance is of the same order of magnitude as the polymer chain length, it can result in a polymer molecule bonding t o more than one filler particle and therefore a very strong filler-topolymer crosslinking effect. It was shown that the 0.05 p aluminafilled composite has a much denser structure and was very difficult t o swell in water, when compared t o the larger filler composites. It is proposed, therefore, that these differences in swelling, solubility, and microstructure can be related to differences in the number concentration of filler and the resulting effectiveness of the filler-polymer network structure. Mechanical testing of the composites has shown that one effect of thermally removing the tightly bound water is to produce glassy behavior with a corresponding increase in yield strength and decrease in elongation at yield. This suggests that plastic deformation in the nonheat-treated composites occurs by the relatively easy slippage of macroscopic regions past one another. Apparently, once the bound water is removed, the average distance between particles is decreased and the magnitude of the particle crosslinking effect is increased such that these macroscopic regions become difficult to separate. This view gains support from the shown swelling and solubility behavior of the composites. Although differences in the solubility of the heat-treated and untreated matrix of the 0.05 p composites in boiling water were negligible, a much greater degree of swelling occurred in the nonheat-treated samples. This can only be attributed to the ease of conformation of the polymer chains when they are relatively unrestricted by crosslinks. These properties of the PAA-alumina system give it certain advantages in implant applications. The fact that the polymer is insoluble in the monomer while the monomer is soluble in water ensures that all of the monomer can be removed with the water; hence, no problems with residual monomer leaching will occur in vivo. The hydrophillic nature of the polymer resulting in a slight swelling can produce a tight fit of a n implant against the walls of the prepared socket. Recent evidence20 has suggested that the presence of compressive stress across the implant-tissue interface can be an aid to tissue ingrowth.
POLYMER-CERAMIC COMPOSITE FOR TOOTH-ROOT IMPLANT 787
The composites have been shown to have a generally porous structure with fine interconnecting porosity. Klawitter and HulbertZ1have demonstrated that there is a minimum interconnecting pore size necessary for fibrous tissue ingrowth on the order of 5-15 p. This is of the same order as that measured in the denser 0.05 p composites. It should be noted, however, that the hydrophillic nature of the polymer should allow for an increase of this critical dimension when the implant is placed in contact with body fluids.
CONCLUSIONS A new poly(acry1ic acid)-alumina composite has been developed for use in tooth-root implant application. This composite has many advantages as a n implantable prosthesis, such as : 1) biocompatible components, 2) ease of processing, 3) good mechanical properties, 4) controlled interconnecting porosity, and 5 ) controlled swelling in aqueous media. The mechanical properties of the composite are dependent on the alumina particle size, plasticizing effects of water, and the polymer/ ceramic ratio. The physical properties of the composites indicate strong interaction between polymer and filler. This interaction was significantly increased with the 0.05 p alumina particles. The scanning electron microscopy showed that the composite possessed homogeneous and porous microstructure with maximum porosity attained with the 0.3 p filler particles. This investigation was supported by NIDR grant # DE-0199-08.
References 1. J. E. Hamner, 0. M. Reed, and A. R. Hand, Oral Surg., 30, 555 (1970). 2. J. E. Hamner, 0. M. Reed, and It. C. Greulich, J . Biomed. Mater. Res. S y m p . No. 3, 1 (1972). 3. W. E . Cook, C. R. Manning, J. C. Hurt, and D. F. Taylor, J . Riomed. Mater. Res. Symp. No. 9 (Part a), 443 (1972). 4. M. Hodosh, M. Povar, and G. Shklar, J . Periodont., 39, 187 (1968). 5 . M. Hodosh, M. Povar, and G. Shklar, J . Prosth. Dent., 24, 456 (1970). 6. M. Hodosh, G. Shklar, and M. Povar, Oral Surg., 33, 1022 (1972). 7. J. E. Hamner and 0. M. Reed, J . Biomed. Mater. Res. S y m p . No. 4, 217 (1973). 8. F. A. Young, J . Biomed. Mater. Res. Symp. N o . 2 (Part a), 281 (1972).
788
GREENBERG AND KAMEL
9. A. N. Cranin, T. A. Dennison, P. Schnitman, S. Piliero, and L. Pentel, J . Biomed. Muter. Res. Symp. N o . 4, 235 (1973). 10. D. D. Moyle, J. J. Klawitter, and S. F. Hulbert, J . Biomed. Muter. Res. Symp. N o . 4, 363 (1973). 11. J . E. Hamner and 0. M. Reed, Ariz. Dent. J . , 17, 12 (1971). 12. J. E. Hamner and 0. M. Reed, J . Biomed. Muter. Res. Symp. N o . 2 (Part 2), 297 (1972). 13. It. G. Topaxian, W. B. Hammer, C. D. Talbert, and S. F. Hulbert, J . Biomed. Muter. Res. Symp. N o . 2 (Part 2), 311 (1972). 14. T. D. Driskell, M. J. O’Hara, H. D. Skeets, G. W. Green, and J. R. Natiella, J . Biomed. Muter. Res. Symp. N o . 2 (Part 2), 345 (1972). 15. D. C. Smith, Biomed. Eng., 8, 108 (1973). 16. S. F. Hulbert, S. J. Morrison, and J. J. Klawitter, J . Biomed. Muter. Res., 6, 347 (1972). 17. S. F. Hulbert, F. W. Cooke, J. J . Klawitter, R. B. Leonard, B. W. Sauer, D. D. Moyle, and H. B. Skinner, J . Biomed. Mater. Res. Symp. N o . 4, 1 (1973). 18. W. J. Peters, R. W. Jackson, and D. C. Smith, J . Biomed. Muter. Res., 8, 53 (1974). 19. J. H. Main, D. Mock, G. S. Beagne, and D. C. Smith, J . Biomed. Muter. Res., 9, 69 (1975). 20. E. F. Rybicki, K. R. Wheeler, M. T. Karagianes, F. A. Simonen, C. R. Hassler, and L. E. Hulbert, paper presented a t the 27th Annual Conference on Engineering in Medicine and Biology, Philadelphia, Pa., October 6-10, 1974. 21. J. J. Klawitter and 9. F. Hulbert, J . Biomed. Muter. Res. Symp. N o . 2 (Part l ) , 161 (1971).
Received November 17, 1975 Revised December 8, 1975
VOL. 10, PP. 777-788 (1976)
Polymer-Ceramic Composite for Tooth-Root Implant* A. R. GREENBERG and IHAB KAMEL, Department of Materials Engineering and the Biomedical Engineering Program, Drexel University, Philadelphia, Pennsylvania 19104
Summary A new polymer-ceramic composite suitable for tooth-root implants has been developed in this study. This material exhibited the desirable combination of good mechanical properties, controlled porosity, and ease of processing. A thermal processing technique was utilized to polymerize acrylic acid (AA) in the presence of either 0.3 or 0.05 @ alumina particles. Porosity and pore size distribution were influenced by the alumina particle size and the processing technique. For a 50 vol % AA solution, the composite had an average compressive strength of 18,000 psi and 38% porosity when 0.3 I.( filler particles were used. I n comparison, the 0.05 alumina-filled composite had an average compressive strength of 28,OOO.psi with a 15% porosity. Data on the physical and structural characteristics of the composite are presented in this study. Based on these results, the composite material shows good potential for use in tooth-root implants as well as other orthopedic implant applications.
INTRODUCTION Although various workers have adopted somewhat different criteria for the successful tooth-root implant, it has generally been that such an implant must be nontoxic, be noncarcinogenic, provide for stable retention, and display durability in the body environment so that adequate physical and mechanical properties are retained. In the past, various materials and implant designs have been tried in an attempt to meet the above The problems associated with each of these implants have been well documented in the literature.3s10 Some of the more serious dis-
* Presented a t the Seventh Annual International Biomaterials Symposium, Clemson University, Clemson, South Carolina, April 26-30, 1975. 777 @ 1976 by John Wiley & Sons, Inc.
778
GREENBERG AND KAMEL
advantages associated with each of these approaches have included leaching of unreacted monomer into surrounding tissues from porous l2 the brittleness self-curing poly(methy1 methacrylate) systems, and high notch sensitivity of ~ e r a m i c s ,reaction ~ ~ ~ ~ of , ~metals ~ with the body environment,13 and loosening of screws and blades in V ~ V O . ~ I n a n attempt t o overcome some of these problems, a polymerceramic composite is introduced in this study which meets the four general criteria for implantable materials. The composite which has been developed consists of an aluminum oxide filler in a poly(acrylic acid) matrix. I n regard t o the basic compatibility of the composite, it will suffice t o say that previous investigators have shown alumina and the polycarboxylates to be both nontoxic and noncar~inogenic.'~-'~ Recent work has demonstrated that a porous structure which allows tissue ingrowth into the tooth-root implant can provide the needed supportive retention and eliminate epithelial invagination and fibrous encapsulation of the implant.4t12 This investigation has shown that not only can the poly(acry1ic acid) [PAAI-alumina composite be produced with a fine interconnected porosity, but that the material has a combination of physical and mechanical properties that merit its testing in implant applications. The purpose of this paper is to present some of the physical and mechanical properties as well as the structural characteristics of the PAA-alumina system.
MATERIALS AND METHODS Materials Commercially available acrylic acid monomer (Polysciences Inc., Warrington, Pa.) inhibited with 200 ppm MEHQ was used without further purification. Metallurgical-grade alumina powder (Buehler Ltd., Evanston, Illinois) of 0.05 and 0.3 p average particle size was used as received from the manufacturer.
Composite Preparation Aqueous solutions of 50 vol % acrylic acid containing 0.2 vol yo MEK-peroxide were mixed with the powdered alumina t o form a homogeneous slurry. The slurry was poured into glass ampoules, covered, and heated in a 70°C oven for 4 hr t o polymerize the acrylic
POLYMER-CERAMIC COMPOSITE FOR TOOTH-ROOT IMPLANT 779
acid. Two methods were employed to ensure the homogeneity of the sample and to change the outcome of the final porosity. These procedures included 1) allowing the filler to settle by gravity, and 2) rotating the samples in the vertical position during polymerization. Following polymerization, the ingots were removed from the test tubes and dried for 20 hr a t 70°C. In addition, some samples were further heat-treated at 150°C for 12-15 hr in order to remove the tightly bound water remaining in the sample.
Mechanical Testing The dry samples were machined into cylinders of 0.500 in. in diameter and 1.00 in. in length. These specimens were tested in compression on an Instron Universal testing machine. All tests were conducted at crosshead speeds of 0.005 in./min. In accordance with ASTM specification D695, the yield point was taken as the first point on the stress-strain diagram a t which a n increase in strain occurred without an increase in stress. I n addition, the modulus of elasticity was taken as the ratio of stress to corresponding strain below the proportional limit.
Solubility and Swelling Measurements Thin wafers were sectioned from each ingot and immersed in either boiling water or 1 N KOH a t 80°C. Initial estimates of solubility were determined by measuring the percent weight loss of the specimen as a function of immersion time. In addition, swelling was determined by measuring the percent weight change of the wafers as a result of soaking in a water bath at room temperature. Furthermore, the volume changes which occurred during swelling were also measured. Estimates of the initial sample porosity were then determined by calculating the difference between the weight and volume increases (AW - AV) of the sample:
% porosity
=
lOO(AW - AV)/Vo
(1)
where V ois the dry volume of the sample.
Microscopic Examination Specimens were immersed in liquid nitrogen until thermal equilibration and then fractured. The fracture surfaces were made conductive by application of a chromium coating by the vapor
780
GREENBERG AND KAMEL
deposition method. The microstructures of the coated specimens were then examined under the scanning electron microscope.
Thermal Analysis The thermal decomposition pattern of the polymeric matrix was determined by heating a small sample of the composite (10-15 mg) in a du Pont 950 thermogravimetric analyzer at a rate of 5"C/min. Either nitrogen or air was used as the purge medium. The decomposition of the polymer was completed before lOOO"C, and the ratio of polymer/ceramic in the sample was obtained directly.
RESULTS Because of the intimate contact between an implanted material and the body fluids, water sorption and polymer solubility are important factors in determining material performance. Swelling data for the various composites are plotted in Figure 1. It is quite
0 03 0 0 05 (unheat-treated)
48
96
144
TIME ( h r J
Fig. 1. Effect of heat treatment and particle size on swelling of PAA-alumina composites in water at 20°C.
POLYMER-CERAMIC COMPOSITE FOR TOOTH-ROOT IMPLANT 781
apparent from these curves that the 0.3 p alumina composites swelled much faster than and reached about three times the swelling level of the 0.05 p alumina composites. These results would be expected inasmuch as water sorption is dependent on both the pore size and the total porosity and, as will be demonstrated, the 0.3 I.C material is greater in both of these characteristics. In addition, the nonheat-treated 0.05 p composites evidenced a significantly higher degree of swelling than their heat-treated counterparts. Although PAA is completely water-soluble, the composites were not. After 24 hr in boiling water, the 0.05 p specimens evidenced a weight loss of only about 3%, while the larger particle-filled composites had a loss of approximately 12%. This difference between the two particle sizes is, however, somewhat magnified because the measurements also included the loss of the relatively dense alumina. When the samples were placed in the KOH a t 8O"C, the 0.3 p specimens fell apart and partially dissolved after 24 hr ; however, the 0.05 p samples, both heat-treated and untreated, were converted into a gel-like condition and yet still maintained the original sample shape even after 72 hr. This indicates that the particle/matrix interaction is of a secondary rather than primary nature. However, the magnitude of this secondary bonding is large enough to enable the filler to act as a crosslinking agent. It is to be expected that this interaction should be inversely related to the filler particle size. Estimates of initial composite porosity were calculated by the use of eq. (1). The large filler samples were found to have a porosity of about 38% while the 0.05 p samples evidenced a porosity of approximately 15%. The processing technique was varied in order to determine whether the structure of the 0.05 p samples could be modified. This was accomplished by altering the settling rate. Two techniques were used: settling by gravity, and settling by rotation of the slurry. One measure of the effectiveness of these treatments was the polymerceramic ratio as determined by thermal analysis. The results showed that the specimens prepared by rotation and those produced without rotation had final ceramic concentrations of 44 and 36 w t %, respectively. No variation in processing technique was conducted for the 0.3 p specimens. All of the larger filler samples were prepared without rotation. Thermal analysis of these samples indicated a ceramic concentration of 53 wt yo.
782
GREENBERG AND KAMEL
Differences in particle size, porosity, and processing technique were reflected in the structural characteristics of the respective samples. Figure 2 shows scanning electron micrographs of the 0.3 I.( sample. The surface evidences many elevations and depressions, and the
(b)
Fig. 2. SEM micrographs of fracture surface of 0.3 p alumina-PAA composite. A: 300 X ; B: 1000 X.
POLYMER-CERAMIC COMPOSITE FOR TOOTH-ROOT IMPLANT 783
structure can be characterized as homogeneous but highly porous with a n average pore size of 25 p. Figure 3 shows the structure of the 0.05 p sample produced by the rotation technique. The structure is quite different than that of the
(b)
Fig. 3. SEM micrographs of fracture surface of 0.05 p alumina-PAA composite. (a) 300 X; (b) 1000 X.
GREENBERG AND KAMEL
784
0.3 p specimen. Although the structure still contains elevations and depressions, the effects of closer packing are evident. Most of the porosity is in the form of large shallow craters, with relatively few small but deeper pores. The latter was taken to be the minimum interconnecting porosity with a measured diameter of approximately 10 p. Examination of the 0.05 p samples produced without rotation indicated that the structure had been altered. These samples had more of the wider and shallower craters and less of the deep interconnecting pores. Compared to the rotated samples the structure was considerably looser, and the interconnecting pores were somewhat larger, having an average diameter of about 15 p . The mechanical properties of the composites are summarized in Table I and plotted in the form of stress-strain curves in Figure 4. These results clearly show that the heat treatment removed the plasticizing effect of the water such that the yield strength and elastic modulus are increased while the elongation at yield is significantly decreased. The yield point of the nonheat-treated samples could not be defined according to ASTM specification D695 because the stress did not level off and the samples did not fracture. The particle size of the filler is also seen to influence the mechanical properties. The 0.05 p filler had about a 65% increase in compressive yield strength compared to the 0.3 p particle composite. However, some of these differences in the mechanical behavior are influenced by both the increased porosity and filler content of the 0.3 p specimens. TABLE I Mechanical Properties of the PAA-Alumina Composites
Composite type (I.r)
0.3 0.05 (nonheat-treated) 0.05 a
Mean compressive yield stress (psi) X10-3
Mean strain at yield
Mean elastic modulus (psi)
(%I
x 10-5
18.10 9.80. 27.80
8.7 11.0 4.0
4.25 2.50 8.20
All stresses determined a t 11% strain since no definite yield point was observed.
POLYMER-CERAMIC COMPOSITE FOR TOOTH-ROOT IMPLANT 785
'
A005
I
0 0 3
25
0 0 05 (unheot-treoted)
20 ?
D
__-a
15
rn
v) v)
U
10
tv)
5
10
5 STRAIN
1x1
Fig. 4. Effect of heat treatment and particle size on stress-strain behavior of PAA-alumina composites.
DISCUSSION Significant differences in composite behavior have been shown to occur when either the 0.05 or the 0.3 p alumina particles were used. The 0.3 p composites evidenced a substantial increase in swelling rates and levels over the 0.05 p composites. In addition, a twofold increase in porosity occurred with the larger filler using the same processing technique. Furthermore, although PAA is infinitely soluble in water, the results of the solubility tests have shown only minimal weight loss by the composites in boiling water. This indicates that the filler must act as a strong crosslinking agent.in this system. The polymer-filler bond, although of a secondary chemical nature, becomes very effective when the ceramic particles are small. This can be explained on the basis of the average distance between particles at a constant volume fraction of filler. If the particle
786
GREENBERG AND KAMEL
diameter is substantially reduced, the number concentration of particles will be very large; hence, the average distance between them very small. If this distance is of the same order of magnitude as the polymer chain length, it can result in a polymer molecule bonding t o more than one filler particle and therefore a very strong filler-topolymer crosslinking effect. It was shown that the 0.05 p aluminafilled composite has a much denser structure and was very difficult t o swell in water, when compared t o the larger filler composites. It is proposed, therefore, that these differences in swelling, solubility, and microstructure can be related to differences in the number concentration of filler and the resulting effectiveness of the filler-polymer network structure. Mechanical testing of the composites has shown that one effect of thermally removing the tightly bound water is to produce glassy behavior with a corresponding increase in yield strength and decrease in elongation at yield. This suggests that plastic deformation in the nonheat-treated composites occurs by the relatively easy slippage of macroscopic regions past one another. Apparently, once the bound water is removed, the average distance between particles is decreased and the magnitude of the particle crosslinking effect is increased such that these macroscopic regions become difficult to separate. This view gains support from the shown swelling and solubility behavior of the composites. Although differences in the solubility of the heat-treated and untreated matrix of the 0.05 p composites in boiling water were negligible, a much greater degree of swelling occurred in the nonheat-treated samples. This can only be attributed to the ease of conformation of the polymer chains when they are relatively unrestricted by crosslinks. These properties of the PAA-alumina system give it certain advantages in implant applications. The fact that the polymer is insoluble in the monomer while the monomer is soluble in water ensures that all of the monomer can be removed with the water; hence, no problems with residual monomer leaching will occur in vivo. The hydrophillic nature of the polymer resulting in a slight swelling can produce a tight fit of a n implant against the walls of the prepared socket. Recent evidence20 has suggested that the presence of compressive stress across the implant-tissue interface can be an aid to tissue ingrowth.
POLYMER-CERAMIC COMPOSITE FOR TOOTH-ROOT IMPLANT 787
The composites have been shown to have a generally porous structure with fine interconnecting porosity. Klawitter and HulbertZ1have demonstrated that there is a minimum interconnecting pore size necessary for fibrous tissue ingrowth on the order of 5-15 p. This is of the same order as that measured in the denser 0.05 p composites. It should be noted, however, that the hydrophillic nature of the polymer should allow for an increase of this critical dimension when the implant is placed in contact with body fluids.
CONCLUSIONS A new poly(acry1ic acid)-alumina composite has been developed for use in tooth-root implant application. This composite has many advantages as a n implantable prosthesis, such as : 1) biocompatible components, 2) ease of processing, 3) good mechanical properties, 4) controlled interconnecting porosity, and 5 ) controlled swelling in aqueous media. The mechanical properties of the composite are dependent on the alumina particle size, plasticizing effects of water, and the polymer/ ceramic ratio. The physical properties of the composites indicate strong interaction between polymer and filler. This interaction was significantly increased with the 0.05 p alumina particles. The scanning electron microscopy showed that the composite possessed homogeneous and porous microstructure with maximum porosity attained with the 0.3 p filler particles. This investigation was supported by NIDR grant # DE-0199-08.
References 1. J. E. Hamner, 0. M. Reed, and A. R. Hand, Oral Surg., 30, 555 (1970). 2. J. E. Hamner, 0. M. Reed, and It. C. Greulich, J . Biomed. Mater. Res. S y m p . No. 3, 1 (1972). 3. W. E . Cook, C. R. Manning, J. C. Hurt, and D. F. Taylor, J . Riomed. Mater. Res. Symp. No. 9 (Part a), 443 (1972). 4. M. Hodosh, M. Povar, and G. Shklar, J . Periodont., 39, 187 (1968). 5 . M. Hodosh, M. Povar, and G. Shklar, J . Prosth. Dent., 24, 456 (1970). 6. M. Hodosh, G. Shklar, and M. Povar, Oral Surg., 33, 1022 (1972). 7. J. E. Hamner and 0. M. Reed, J . Biomed. Mater. Res. S y m p . No. 4, 217 (1973). 8. F. A. Young, J . Biomed. Mater. Res. Symp. N o . 2 (Part a), 281 (1972).
788
GREENBERG AND KAMEL
9. A. N. Cranin, T. A. Dennison, P. Schnitman, S. Piliero, and L. Pentel, J . Biomed. Muter. Res. Symp. N o . 4, 235 (1973). 10. D. D. Moyle, J. J. Klawitter, and S. F. Hulbert, J . Biomed. Muter. Res. Symp. N o . 4, 363 (1973). 11. J . E. Hamner and 0. M. Reed, Ariz. Dent. J . , 17, 12 (1971). 12. J. E. Hamner and 0. M. Reed, J . Biomed. Muter. Res. Symp. N o . 2 (Part 2), 297 (1972). 13. It. G. Topaxian, W. B. Hammer, C. D. Talbert, and S. F. Hulbert, J . Biomed. Muter. Res. Symp. N o . 2 (Part 2), 311 (1972). 14. T. D. Driskell, M. J. O’Hara, H. D. Skeets, G. W. Green, and J. R. Natiella, J . Biomed. Muter. Res. Symp. N o . 2 (Part 2), 345 (1972). 15. D. C. Smith, Biomed. Eng., 8, 108 (1973). 16. S. F. Hulbert, S. J. Morrison, and J. J. Klawitter, J . Biomed. Muter. Res., 6, 347 (1972). 17. S. F. Hulbert, F. W. Cooke, J. J . Klawitter, R. B. Leonard, B. W. Sauer, D. D. Moyle, and H. B. Skinner, J . Biomed. Mater. Res. Symp. N o . 4, 1 (1973). 18. W. J. Peters, R. W. Jackson, and D. C. Smith, J . Biomed. Muter. Res., 8, 53 (1974). 19. J. H. Main, D. Mock, G. S. Beagne, and D. C. Smith, J . Biomed. Muter. Res., 9, 69 (1975). 20. E. F. Rybicki, K. R. Wheeler, M. T. Karagianes, F. A. Simonen, C. R. Hassler, and L. E. Hulbert, paper presented a t the 27th Annual Conference on Engineering in Medicine and Biology, Philadelphia, Pa., October 6-10, 1974. 21. J. J. Klawitter and 9. F. Hulbert, J . Biomed. Muter. Res. Symp. N o . 2 (Part l ) , 161 (1971).
Received November 17, 1975 Revised December 8, 1975
