Dent Mater 8:185-189, May, 1992
Degradation of microfilled posterior composite R.B. Mazer ~, K.F. Leinfelder 2, C.M. Russell 3
~Restorative Dentistry/Biomaterials, 2 Biomaterials, School of Dentistry, University of Alabama at Birmingham, Birmingham, AL, USA 3Biostatistician, Center for Disease Control, Atlanta, GA, USA
Abstract. The substantial improvement in the chemical, physical and mechanical characteristics of posterior composites has contributed to their increased use in recent years. However, some troubling characteristics of these materials are their susceptibilityto wear, marginal leakage, and recurrent caries. Numerous studies have dealt with the wear resistance of posterior composites. Only a few have investigated the mechanisms of failure, particularly those containing submicron-sized fillers. The purpose of this study, therefore, was to analyze the clinical characteristics of a posterior composite to determine the mechanisms responsible for marginal degradation. Using a series of optical standards, it was determined that the generalized wear-rate was linear, averaging 8 pm/year. Furthermore, it was shown thatthe marginal defectwas cohesive in nature and that this type of defect, which is inherent in submicron-type posterior composites, was probably due to tensile fatigue failure. INTRODUCTION The use of posterior composites has continued to increase over the last several years. The growing use of this important material may be attributed to a number of factors. These include the potential for excellent esthetics, concern over the issue of mercury toxicity, and the ability of these polymers to bond to tooth structure. In spite of their growing popularity, composites continue to exhibit a number of undesirable characteristics. One of the more significant problems is their degradation potential when subjected to high levels of stress (Powers et al., 1981;Wu and Cobb, 1981; Wu and Mc Kinney, 1982; Lambrechts, 1983; Roulet, 1987). Depending upon the size and concentration of the filler particle, posterior composites may exhibit localized or generalized wear patterns, bulk fracture and marginal deterioration. In general, posterior composites containing filler particles less than 1 ~m exhibit distinctly different wear characteristics than those in which the mean particle size is greater than 1 ~m (J~rgensen and Asmussen, 1978; J~rgensen et al., 1979; Leinfelder, 1987b). As a rule, the submicron-sized posterior composites are more resistant to generalized wear, but they are more prone to localized wear (Lutz et al., 1979; Lutz et al., 1984; Roulet, 1985). Furthermore, those composites consisting entirely of colloidal silica or filler particles averaging less than 1 ~m generally exhibit marginal fracturing or "ditching" (McComb, 1985; Leinfelder, 1987b). Such types of defects normally do not occur with materials containing substantially larger-sized filler particles. Still one more difference between the two different types of
composites has been identified. As a rule, the microfilled composites, as well as those in which the mean filler particle is less than 1 ~m, exhibit a linear rate of generalized wear (Mitchem and Gronas, 1982; Lambrechts et al., 1985). Heliomolar RO (Mazer et al., 1989) and Herculite XR (Wisniewski et al., 1988) are typical examples of this kind of wear-rate. Posterior composites containing supramicron sized filler particles on the other hand exhibit a decreasing rate of wear (Goldberg et al., 1984; Leinfelder et al., 1986). P-30 (Teixeira et al., 1987), P-50 (Mazer and Leinfelder, 1988), Fulfil (Vann et al., 1986) and Occlusin (Gerbo et al., 1990) are typical examples of this wear modality. It was the purpose of this study to: 1) characterize in detail the clinical behavior ofa microfilled posterior composite, and 2) to determine the physical and/or mechanical properties responsible for the problem of marginal degradation normally associated with this type of restorative material. Both clinical and laboratory phases were included in the analysis of this degradation process.
METHODS AND MATERIALS Sixty-eight composite restorations (Heliomolar RO, Vivadent, Schaan, Liechtenstein, Batch No. 336501) were inserted into Class II and Class I cavity preparations. The ratio of the two different types was 2:1. Both molars and premolars were included and also followed a 2:1 ratio. A total of 17 patients (18 - 35 years of age) was treated in the study. All restored teeth were in occlusion and at least one proximal surface area of the Class II restorations was in contact. Each cavity preparation was standardized to conform to an amalgam preparation. None of the occlusal cavosurface margins were beveled. Instead, a well defined butt joint was established between the preparation and the restoration. This technique was employed to facilitate detection of any material loss or degradation at or adjacent to the margins. A slight bevel, however, was placed on the gingival margin for the purpose of increasing the surface area of the enamel in this region, thereby optimizing bonding ofthe composite. As a rule, a thin layer of calcium hydroxide (Dycal; L. D. Caulk/Dentsply, Milford, DE, USA) was placed over the prepared dentinal surfaces in order to protect the dentinal tubules and pulp prior to etching of the enamel with a 37% phosphoric acid solution. All operative procedures were carried out under a rubber dam. In each case the restorative material was inserted into the preparation in increments of two or more with each not exceeding 2 mm in thickness. Upon insertion each portion was photocured for at least 40 s. All occlusal surfaces were conDental Materials/May 1992 185
toured with an appropriate size 12 fluted carbide bur in the presence of water. Final polishing was carried out under slow speed, utilizing the same types of burs and a series of discs. In addition to the clinical phase, Class I cavity preparations were generated in a series of extracted teeth. Restorative procedures were carried out in the same manner as previously described. After storing the specimens in deionized water for 24 h at 37°C, the occlusal surfaces were subjected to cyclic fatigue testing as diagrammatically illustrated in Fig. 1. Using a universal testing machine (Model 3111, Instron Corp., Canton, MA, USA), the occlusal surface of the restoration was subjected to a load of 50 lbs at a rate of one cycle every 3 s for 10,000 cycles. The load was imparted by means of a stainless steel stylus. The tip of the stylus approximated the shape of a sphere approximately 1.5 mm in diameter. At the end of every 2,500 cycles the occlusal surfaces of the restored teeth were replicated with a vinyl polysiloxane impression material (Reprosil; L. D. Caulk/Dentsply) and then cast in epoxy resin (Epoxy-Die; Vivadent). After sputter coating with Au/Pd, the specimens were examined by means of scanning electron microscopy (ISI-100B, International Scientific Instruments, Tokyo, Japan). All restored teeth were evaluated for clinical performance in accordance with the criteria established by the United States Public Health Service (Cvar and Ryge, 1971). Characteristics evaluated included 1) color match, 2) interfacial staining, 3) caries, 4) wear or loss of anatomic form, 5) marginal adaptation, 6) surface texture, and 7) bulk fracture. Evaluations were carried out by two clinicians and were conducted at the time of insertion and then each three months during the first year. Evaluations were also conducted at 18 months and two years. An interexaminer agreement of at least 85% was obtained throughout the two-year period for each characteristic evaluated. All restorations were photographed at each recall period at an original magnification of 1.5x using colored film transparencies. In addition, all restored teeth were impressed with vinyl polysiloxane and then cast with die stone. The die stone casts were visually compared to a series of optical standards (Moffa-Lugassy standards) for the purpose of determining loss of material at the cavosurface margin. The die stone casts were also used to establish the degree of marginal integrity along the cavosurface margin. Procedurally, each cast was observed under an optical bench microscope (Model SFB-2,
TABLE: MEANWEAR(p_m)VALUESOF HELIOMOLAR RO Time
Mean
SD
p
0
0.0
5.8
0.6
3
1.0
6.3
0.8
6
3.0
7.2
0.2
9
5.0
7.8
0.9
12
8.0
8.6
0.9
18
10.0
9.0
0.8
24
15.0
13.1
*
* single evaluatorwear values
Bausch & Lomb, Inc., Rochester, NY, USA) at 1.5x magnification. The site of initiation and subsequent propagation of the marginal defect along the cavosurface margin were recorded on a schematic representation of the restored tooth. Finally, a number of specimens were selected for replication and evaluation with scanning electron microscopy throughout the study periods. The selection criteria was based on extension of the preparation (both Class I and II), occlusal and proximal contacts RESULTS Direct. During the first 24 months of service, the percent of restorations exhibiting an ideal color match never decreased below 98%. In all cases, the translucency and opacity of the restoration closely resembled that of the adjacent enamel. During the course of the first 24 months, no secondary caries were detected. Two of the restorations exhibited evidence ofinterfacial staining, but in both cases the extent was superficial. Also, during this time, only two restorations were judged to exhibit clinical evidence of wear and were rated as Bravo. No localized or bulk fractures were detected and only one restoration exhibited a surface texture less smooth than that of the adjacent enamel. Indirect. The results ofwear measurements as determined by the M-L optical standards are presented in the Table. As can be seen, mean values in micrometers are given for all time periods up to 24 months. These values were generated by subtracting the baseline values from those determined at each recall period. Standard deviations for each mean value are also presented. No significant differences between the wear readings of the two evaluators were demonstrated (p < 0.05) throughout the entire two years. A comparison of the wear
120
E ~L
100 ~
Ful Fil
80 -~
~
Herculite
40 20
I
I
I
I
0.5
1.0
2.0
3.0
Time (yrs) Fig. 1. Schematic illustration of cyclic loading test used to produce marginal degradation,
186 Mazer et aL/Degradation of microfilled posterior composite
Occlusin
60
F~g.2. In vivo wear rates of several proprietary posterior composites.
3 Mo
6 Mo
9 Mo
12 Mo
18 Mo
24 Mo
Fig. 3. Schematic illustration demonstrating the initiation of propagation of marginal degradation. Cross.hatched markings on occlusal surface indicate the location of occlusal contact areas. Solid black areas along margins of restoration indicate the location of the interfecial defects (see arrows), Note that the defects (3 months) generally initiate in the areas closest to contact areas occurring on the composite and/or tooth site.
resistance of Heliomolar RO with three other posterior composites tested under the same conditions are presented in Fig. 2. A schematic illustration of a molar tooth restored with the posterior resin composite is presented in Fig. 3. Depicted are the location of the occlusal centric holding areas, both on the restoration as well as the tooth structure, and the areas in which marginal fracture was initiated. The illustration also demonstrates the rate at which the degradation process occurs. The data were derived from die stone replicas which were observed under the optical microscope. Initial marginal fracture is indicated by the location of arrows. Its horizontal extension is schematically represented by a thickening of the margin at the tooth/restoration interface. In each case, the marginal fracture originated in locations nearest the centric holding areas positioned on the restorative material. The degradation process steadily continued as a function of time. At the end of 24 months, the defect
Fig. 5. In vivo scanning electron micrograph (12 months) of interracial defect. Note that the propagating crack is cohesive in nature. Original magnification 1,000x. Arrows indicate junction between etched enamel and restoration.
had propagated to nearly the entire length of the occlusal cavosurface margin. Interestingly, it should be pointed out that this marginal defect was not readily detected clinically. However, it became obvious when viewed on die stone casts and scanning electron micrographs. The rate at which this marginal defect propagated along the margins of the restoration is illustrated in Fig. 4. The time scale is represented on the horizontal axis, whereas the extent of breakdown along the margin is represented on the vertical axis. It can be seen that the degradation process is approximately linear and required nearly two years to complete. A scanning electron micrograph of one of the clinical restorations portraying the marginal defect is shown in Fig 5. For purposes of orientation, the enamel surface is located at the bottom of the micrograph. Although the fracture occurred along the margin, it is basically cohesive in nature. Fig. 6 is a scanning electron micrograph of one of the specimens sub-
Rate of Margin Deterioration 100
"~
80
a r--
60
L..
40 c~ L..
~
2o
I 0
I 1
I 2
Time In Years Fig.4. Schematic illustration demonstrating the rate at which the intedacial defect progresses along the margins of restoration.
Fig. 6. In vitro scanning electron micrograph of interfacial defect (10,000cycles). Note that fracture is cohesive in nature. Original magnification 1,800x. Dental Materials~May 1992 187
Etched Enamel
Fig. 7. Schematic illustration demonstrating location and type of fracture (in vivo and in vitro) following occlusal loading.
jected to mechanical stressing in vitro. A series of parallel fracture lines occurred inside the restorative material, resulting in an eventual loss of composite. All in vitro specimens subjected to mechanical cyclic stress underwent the same type of marginal deterioration as was observed clinically. After approximately 5,000 cycles, the margins of the restoration began to exhibit a series of parallel microcracks. Like the clinical restorations, these defects were all restricted to the restorative material. None of the margins actually debonded from the walls of the etched enamel. A typical cohesive fracture pattern is that illustrated in Fig 6.
DISCUSSION The clinical results of this study indicate that Heliomolar RO is a highly wear resistant posterior composite. The total mean loss of material on the occlusal surface, for example, was only 15.3 ~m at the end of 24 months. In addition, no evidence of localized or bulk fracture was detected on any of the restorations over the two-year period. The marginal crevicing associated with this material as well as other submicron-sized posterior composites appears to be quite common. It should be pointed out that subsequent to polymerization, the matrix absorbs water. However, this volumetric expansion (1.0%, according to manufacturer) is not sufficient to offset the shrinkage attributed to the curing process (2.8 vol. %). Based upon the careful mapping of this defect in conjunction with controlled laboratory experiments the mechanism of failure appears to be associated with one or more mechanical properties which characterize this type of material. Based upon the in vivo and in vitro aspects of this study, it is proposed that the marginal degradation on the occlusal surface is due to tensile fatigue failure. The possible mechanism responsible for this is schematically illustrated in Fig. 7. During the polymerization process, the restorative materials undergo a volumetric shrinkage of approximately 2.8% after 24 h (Feilzer et al., 1988). Consequently, tensile stresses are generated in the composite in areas immediately adjacent to the tooth/restoration interface. During mastication, the restoration is deflected pulpally and physiologic cuspal movements also take place. Both contribute to increased tensile stresses along the margins. Since the submicron-sized posterior composites posses a
188 Mazer et aL/Degradation of microfilled posterior composite
lower elastic modulus than those with larger-sized fillers, the elastic strain is somewhat higher. Also, because the tensile strength for this type of material is less, small localized fractures would be expected. Interestingly, the microscopic defects originated on the margins closest to the centric stops. Assuming that the tensile stresses should be greatest in these areas, it is not surprising that these marginal defects occurred where they did. A careful examination of the scanning electron micrographs of both the in vitro as well as the in vivo restorations revealed that the defect is entirely cohesive. In other words, the defect is initiated along the margins but within the restorative material itself. In all cases, the tooth structure adjacent to the microstructural defect remained covered or sealed with the composite. An examination of other posterior composites in which the filler particles are greater than 1 ~m does not always reveal this type ofmicrostructural defect. It is entirely possible then, that the small particle size and the resultant decrease in certain tensile properties are responsible for this type of behavior. Based upon limited SEM observations, it is possible that the degradation process of all posterior composite restorations is initiated by submicroscopic marginal fractures. However, since the microfilled composites are generally more wear resistant, the marginal defects are more readily detectable. In summary, Heliomolar RO is highly resistant to generalized wear on the occlusal surface. At the end of2 y, for example, the total loss of substance averaged only 15.3 ram. At the end of 2 y, 95% of the restorations exhibited an ideal color match. Also, no evidence of bulk fracture was observed. Finally, the marginal degradation, although not readily observed clinically, probably was related to tensile fatigue. Received September, 17, 1991/AcceptedDecember20, 1991 Address correspondenceand reprint requests to: R. B. Mazer School of Dentistry Box 49 University of Alabama at Birmingham Birmingham,AL 35294 USA
REFERENCES Cvar JF, Ryge G (1971). Criteria for the clinical evaluation of dental restorative materials. Washington, DC: United States Department of Health, Education, and Welfare. Feilzer AJ, De Gee AJ, Davidson CL (1988). Curing contraction of composites and glass-ionomer cements. J Prosthet Dent 59(3): 297-300. Gerbo LR, Leinfelder KF, Mueninghoff LA, Russell C (1990). Use of optical standards for determining wear of posterior composite resins. JEsthetic Dent 2:5, 148-152. Goldberg AJ, Rydinge E, Santucci EA, Racz, WB (1984). Clinical evaluation methods for posterior composite restorations. JDent Res 63(12): 1387-1391. J~rgensen KD, Asmussen E (1978). Occlusal abrasion of a composite restorative resin with ultra-fine filler. An initial study. Quint Int 9: 73-79.
J~rgensen KD, Horsted P, Janum O, Krogh J, Schultz J (1979). Abrasion of Class I restorative resins. Scand JDent Res 87: 140-145. Lambrechts P (1983). Properties of dental composites and their impact on clinical performance. Thesis, Leuven, Belgium. Lambrechts P, Braem M, Vanherle G (1985). Accomplishments and expectations with posterior composite resins. In: International Symposium on Posterior Composite Resin, Dental Restorative Materials 521-540, The Netherlands, Peter Szule. Leinfelder KF, Wilder AD, Jr, Teixeira LC (1986). Wear rates of posterior composite resins. J A m Dent Assoc 112: 829833. Leinfelder KF (1987a). Evaluation of criteria used for assessing the clinical performanceofcomposite resins in posterior teeth. Quint Int 18:531-536. Leinfelder KF (1987b). Wear patterns and rates of posterior composite resins. Int Dent J 37:152-157. Lutz F, Imfeld T, Meier C, Firestone A. (1979). Composite versus amalgam - measurements of in vivo wear resistance: one-year report. Quint Int 10:77-87 Lutz F, Phillips RW, Roulet J-F, Setcos JC (1984). Potential posterior composites. An in vitro and in vivo comparison for wear. JDent Res 63: 914-920. Mazer RB, Leinfelder KF (1988). Clinical evaluation of a posterior composite resin containing a new type of filler particle. J Esthetic Dent 1: 66-70. Mazer RB, Leinfelder KF, Russell CM (1989). Mechanism of failure in a microfilled composite resin. JDent Res 68: 233,
Abstr. No. 418. McComb D (1985). Evaluation of clinical wear of posterior composite resins. In: International Symposium on Posterior Composite Resin, Dental Restorative Materials pp.511-517, The Netherlands, Peter Szule. Mitchem J, Gronas DG (1982). In vivo evaluation ofthe wear of restorative resin. J A m Dent Assoc 104: 333-335. Powers JM, Fan PL, Marcotte M (1981). In vitro accelerated aging of composites and a sealant. J Dent Res 60: 16721677. Roulet J-F (1985). In vivo wear measurements of composite resins. In: International Symposium on Posterior Composite Resin, Dental Restorative Materials, The Netherlands, Peter Szule, 365-372. Roulet J-F (1987). Degradation of Dental Polymers. First Edition. New York: Karger, 77-79. Teixeira LC, Isenberg BP, Leinfelder KF (1987). Effect of silane treatment on wear of a posterior composite resin. J Dent Res 66: 167, Abstr. No. 482. Vann WF, Barkmeier WW, Oldenburg TE, Leinfelder KF (1986). Quantitative wear assessments for composite resin restorations in primary molars. J Pediatric Dent 8: 7-10. Wisniewski JF, McCartha CD, Leinfelder KF, Isenberg BP (1988). Two year clinical investigation of a fine particle posterior composite resin. JDentRes 67:120, Abstr. No. 62. Wu W, Cobb EN (1981). A silver staining technique for investigating wear of restorative dental composites. J Biota Mater Res 15: 343-348. Wu W, Mc Kinney JE (1982). Influence of chemicals on wear of dental composites. JDent Res 61: 1180-1183.
Dental Materials~May 1992 18£
Degradation of microfilled posterior composite R.B. Mazer ~, K.F. Leinfelder 2, C.M. Russell 3
~Restorative Dentistry/Biomaterials, 2 Biomaterials, School of Dentistry, University of Alabama at Birmingham, Birmingham, AL, USA 3Biostatistician, Center for Disease Control, Atlanta, GA, USA
Abstract. The substantial improvement in the chemical, physical and mechanical characteristics of posterior composites has contributed to their increased use in recent years. However, some troubling characteristics of these materials are their susceptibilityto wear, marginal leakage, and recurrent caries. Numerous studies have dealt with the wear resistance of posterior composites. Only a few have investigated the mechanisms of failure, particularly those containing submicron-sized fillers. The purpose of this study, therefore, was to analyze the clinical characteristics of a posterior composite to determine the mechanisms responsible for marginal degradation. Using a series of optical standards, it was determined that the generalized wear-rate was linear, averaging 8 pm/year. Furthermore, it was shown thatthe marginal defectwas cohesive in nature and that this type of defect, which is inherent in submicron-type posterior composites, was probably due to tensile fatigue failure. INTRODUCTION The use of posterior composites has continued to increase over the last several years. The growing use of this important material may be attributed to a number of factors. These include the potential for excellent esthetics, concern over the issue of mercury toxicity, and the ability of these polymers to bond to tooth structure. In spite of their growing popularity, composites continue to exhibit a number of undesirable characteristics. One of the more significant problems is their degradation potential when subjected to high levels of stress (Powers et al., 1981;Wu and Cobb, 1981; Wu and Mc Kinney, 1982; Lambrechts, 1983; Roulet, 1987). Depending upon the size and concentration of the filler particle, posterior composites may exhibit localized or generalized wear patterns, bulk fracture and marginal deterioration. In general, posterior composites containing filler particles less than 1 ~m exhibit distinctly different wear characteristics than those in which the mean particle size is greater than 1 ~m (J~rgensen and Asmussen, 1978; J~rgensen et al., 1979; Leinfelder, 1987b). As a rule, the submicron-sized posterior composites are more resistant to generalized wear, but they are more prone to localized wear (Lutz et al., 1979; Lutz et al., 1984; Roulet, 1985). Furthermore, those composites consisting entirely of colloidal silica or filler particles averaging less than 1 ~m generally exhibit marginal fracturing or "ditching" (McComb, 1985; Leinfelder, 1987b). Such types of defects normally do not occur with materials containing substantially larger-sized filler particles. Still one more difference between the two different types of
composites has been identified. As a rule, the microfilled composites, as well as those in which the mean filler particle is less than 1 ~m, exhibit a linear rate of generalized wear (Mitchem and Gronas, 1982; Lambrechts et al., 1985). Heliomolar RO (Mazer et al., 1989) and Herculite XR (Wisniewski et al., 1988) are typical examples of this kind of wear-rate. Posterior composites containing supramicron sized filler particles on the other hand exhibit a decreasing rate of wear (Goldberg et al., 1984; Leinfelder et al., 1986). P-30 (Teixeira et al., 1987), P-50 (Mazer and Leinfelder, 1988), Fulfil (Vann et al., 1986) and Occlusin (Gerbo et al., 1990) are typical examples of this wear modality. It was the purpose of this study to: 1) characterize in detail the clinical behavior ofa microfilled posterior composite, and 2) to determine the physical and/or mechanical properties responsible for the problem of marginal degradation normally associated with this type of restorative material. Both clinical and laboratory phases were included in the analysis of this degradation process.
METHODS AND MATERIALS Sixty-eight composite restorations (Heliomolar RO, Vivadent, Schaan, Liechtenstein, Batch No. 336501) were inserted into Class II and Class I cavity preparations. The ratio of the two different types was 2:1. Both molars and premolars were included and also followed a 2:1 ratio. A total of 17 patients (18 - 35 years of age) was treated in the study. All restored teeth were in occlusion and at least one proximal surface area of the Class II restorations was in contact. Each cavity preparation was standardized to conform to an amalgam preparation. None of the occlusal cavosurface margins were beveled. Instead, a well defined butt joint was established between the preparation and the restoration. This technique was employed to facilitate detection of any material loss or degradation at or adjacent to the margins. A slight bevel, however, was placed on the gingival margin for the purpose of increasing the surface area of the enamel in this region, thereby optimizing bonding ofthe composite. As a rule, a thin layer of calcium hydroxide (Dycal; L. D. Caulk/Dentsply, Milford, DE, USA) was placed over the prepared dentinal surfaces in order to protect the dentinal tubules and pulp prior to etching of the enamel with a 37% phosphoric acid solution. All operative procedures were carried out under a rubber dam. In each case the restorative material was inserted into the preparation in increments of two or more with each not exceeding 2 mm in thickness. Upon insertion each portion was photocured for at least 40 s. All occlusal surfaces were conDental Materials/May 1992 185
toured with an appropriate size 12 fluted carbide bur in the presence of water. Final polishing was carried out under slow speed, utilizing the same types of burs and a series of discs. In addition to the clinical phase, Class I cavity preparations were generated in a series of extracted teeth. Restorative procedures were carried out in the same manner as previously described. After storing the specimens in deionized water for 24 h at 37°C, the occlusal surfaces were subjected to cyclic fatigue testing as diagrammatically illustrated in Fig. 1. Using a universal testing machine (Model 3111, Instron Corp., Canton, MA, USA), the occlusal surface of the restoration was subjected to a load of 50 lbs at a rate of one cycle every 3 s for 10,000 cycles. The load was imparted by means of a stainless steel stylus. The tip of the stylus approximated the shape of a sphere approximately 1.5 mm in diameter. At the end of every 2,500 cycles the occlusal surfaces of the restored teeth were replicated with a vinyl polysiloxane impression material (Reprosil; L. D. Caulk/Dentsply) and then cast in epoxy resin (Epoxy-Die; Vivadent). After sputter coating with Au/Pd, the specimens were examined by means of scanning electron microscopy (ISI-100B, International Scientific Instruments, Tokyo, Japan). All restored teeth were evaluated for clinical performance in accordance with the criteria established by the United States Public Health Service (Cvar and Ryge, 1971). Characteristics evaluated included 1) color match, 2) interfacial staining, 3) caries, 4) wear or loss of anatomic form, 5) marginal adaptation, 6) surface texture, and 7) bulk fracture. Evaluations were carried out by two clinicians and were conducted at the time of insertion and then each three months during the first year. Evaluations were also conducted at 18 months and two years. An interexaminer agreement of at least 85% was obtained throughout the two-year period for each characteristic evaluated. All restorations were photographed at each recall period at an original magnification of 1.5x using colored film transparencies. In addition, all restored teeth were impressed with vinyl polysiloxane and then cast with die stone. The die stone casts were visually compared to a series of optical standards (Moffa-Lugassy standards) for the purpose of determining loss of material at the cavosurface margin. The die stone casts were also used to establish the degree of marginal integrity along the cavosurface margin. Procedurally, each cast was observed under an optical bench microscope (Model SFB-2,
TABLE: MEANWEAR(p_m)VALUESOF HELIOMOLAR RO Time
Mean
SD
p
0
0.0
5.8
0.6
3
1.0
6.3
0.8
6
3.0
7.2
0.2
9
5.0
7.8
0.9
12
8.0
8.6
0.9
18
10.0
9.0
0.8
24
15.0
13.1
*
* single evaluatorwear values
Bausch & Lomb, Inc., Rochester, NY, USA) at 1.5x magnification. The site of initiation and subsequent propagation of the marginal defect along the cavosurface margin were recorded on a schematic representation of the restored tooth. Finally, a number of specimens were selected for replication and evaluation with scanning electron microscopy throughout the study periods. The selection criteria was based on extension of the preparation (both Class I and II), occlusal and proximal contacts RESULTS Direct. During the first 24 months of service, the percent of restorations exhibiting an ideal color match never decreased below 98%. In all cases, the translucency and opacity of the restoration closely resembled that of the adjacent enamel. During the course of the first 24 months, no secondary caries were detected. Two of the restorations exhibited evidence ofinterfacial staining, but in both cases the extent was superficial. Also, during this time, only two restorations were judged to exhibit clinical evidence of wear and were rated as Bravo. No localized or bulk fractures were detected and only one restoration exhibited a surface texture less smooth than that of the adjacent enamel. Indirect. The results ofwear measurements as determined by the M-L optical standards are presented in the Table. As can be seen, mean values in micrometers are given for all time periods up to 24 months. These values were generated by subtracting the baseline values from those determined at each recall period. Standard deviations for each mean value are also presented. No significant differences between the wear readings of the two evaluators were demonstrated (p < 0.05) throughout the entire two years. A comparison of the wear
120
E ~L
100 ~
Ful Fil
80 -~
~
Herculite
40 20
I
I
I
I
0.5
1.0
2.0
3.0
Time (yrs) Fig. 1. Schematic illustration of cyclic loading test used to produce marginal degradation,
186 Mazer et aL/Degradation of microfilled posterior composite
Occlusin
60
F~g.2. In vivo wear rates of several proprietary posterior composites.
3 Mo
6 Mo
9 Mo
12 Mo
18 Mo
24 Mo
Fig. 3. Schematic illustration demonstrating the initiation of propagation of marginal degradation. Cross.hatched markings on occlusal surface indicate the location of occlusal contact areas. Solid black areas along margins of restoration indicate the location of the interfecial defects (see arrows), Note that the defects (3 months) generally initiate in the areas closest to contact areas occurring on the composite and/or tooth site.
resistance of Heliomolar RO with three other posterior composites tested under the same conditions are presented in Fig. 2. A schematic illustration of a molar tooth restored with the posterior resin composite is presented in Fig. 3. Depicted are the location of the occlusal centric holding areas, both on the restoration as well as the tooth structure, and the areas in which marginal fracture was initiated. The illustration also demonstrates the rate at which the degradation process occurs. The data were derived from die stone replicas which were observed under the optical microscope. Initial marginal fracture is indicated by the location of arrows. Its horizontal extension is schematically represented by a thickening of the margin at the tooth/restoration interface. In each case, the marginal fracture originated in locations nearest the centric holding areas positioned on the restorative material. The degradation process steadily continued as a function of time. At the end of 24 months, the defect
Fig. 5. In vivo scanning electron micrograph (12 months) of interracial defect. Note that the propagating crack is cohesive in nature. Original magnification 1,000x. Arrows indicate junction between etched enamel and restoration.
had propagated to nearly the entire length of the occlusal cavosurface margin. Interestingly, it should be pointed out that this marginal defect was not readily detected clinically. However, it became obvious when viewed on die stone casts and scanning electron micrographs. The rate at which this marginal defect propagated along the margins of the restoration is illustrated in Fig. 4. The time scale is represented on the horizontal axis, whereas the extent of breakdown along the margin is represented on the vertical axis. It can be seen that the degradation process is approximately linear and required nearly two years to complete. A scanning electron micrograph of one of the clinical restorations portraying the marginal defect is shown in Fig 5. For purposes of orientation, the enamel surface is located at the bottom of the micrograph. Although the fracture occurred along the margin, it is basically cohesive in nature. Fig. 6 is a scanning electron micrograph of one of the specimens sub-
Rate of Margin Deterioration 100
"~
80
a r--
60
L..
40 c~ L..
~
2o
I 0
I 1
I 2
Time In Years Fig.4. Schematic illustration demonstrating the rate at which the intedacial defect progresses along the margins of restoration.
Fig. 6. In vitro scanning electron micrograph of interfacial defect (10,000cycles). Note that fracture is cohesive in nature. Original magnification 1,800x. Dental Materials~May 1992 187
Etched Enamel
Fig. 7. Schematic illustration demonstrating location and type of fracture (in vivo and in vitro) following occlusal loading.
jected to mechanical stressing in vitro. A series of parallel fracture lines occurred inside the restorative material, resulting in an eventual loss of composite. All in vitro specimens subjected to mechanical cyclic stress underwent the same type of marginal deterioration as was observed clinically. After approximately 5,000 cycles, the margins of the restoration began to exhibit a series of parallel microcracks. Like the clinical restorations, these defects were all restricted to the restorative material. None of the margins actually debonded from the walls of the etched enamel. A typical cohesive fracture pattern is that illustrated in Fig 6.
DISCUSSION The clinical results of this study indicate that Heliomolar RO is a highly wear resistant posterior composite. The total mean loss of material on the occlusal surface, for example, was only 15.3 ~m at the end of 24 months. In addition, no evidence of localized or bulk fracture was detected on any of the restorations over the two-year period. The marginal crevicing associated with this material as well as other submicron-sized posterior composites appears to be quite common. It should be pointed out that subsequent to polymerization, the matrix absorbs water. However, this volumetric expansion (1.0%, according to manufacturer) is not sufficient to offset the shrinkage attributed to the curing process (2.8 vol. %). Based upon the careful mapping of this defect in conjunction with controlled laboratory experiments the mechanism of failure appears to be associated with one or more mechanical properties which characterize this type of material. Based upon the in vivo and in vitro aspects of this study, it is proposed that the marginal degradation on the occlusal surface is due to tensile fatigue failure. The possible mechanism responsible for this is schematically illustrated in Fig. 7. During the polymerization process, the restorative materials undergo a volumetric shrinkage of approximately 2.8% after 24 h (Feilzer et al., 1988). Consequently, tensile stresses are generated in the composite in areas immediately adjacent to the tooth/restoration interface. During mastication, the restoration is deflected pulpally and physiologic cuspal movements also take place. Both contribute to increased tensile stresses along the margins. Since the submicron-sized posterior composites posses a
188 Mazer et aL/Degradation of microfilled posterior composite
lower elastic modulus than those with larger-sized fillers, the elastic strain is somewhat higher. Also, because the tensile strength for this type of material is less, small localized fractures would be expected. Interestingly, the microscopic defects originated on the margins closest to the centric stops. Assuming that the tensile stresses should be greatest in these areas, it is not surprising that these marginal defects occurred where they did. A careful examination of the scanning electron micrographs of both the in vitro as well as the in vivo restorations revealed that the defect is entirely cohesive. In other words, the defect is initiated along the margins but within the restorative material itself. In all cases, the tooth structure adjacent to the microstructural defect remained covered or sealed with the composite. An examination of other posterior composites in which the filler particles are greater than 1 ~m does not always reveal this type ofmicrostructural defect. It is entirely possible then, that the small particle size and the resultant decrease in certain tensile properties are responsible for this type of behavior. Based upon limited SEM observations, it is possible that the degradation process of all posterior composite restorations is initiated by submicroscopic marginal fractures. However, since the microfilled composites are generally more wear resistant, the marginal defects are more readily detectable. In summary, Heliomolar RO is highly resistant to generalized wear on the occlusal surface. At the end of2 y, for example, the total loss of substance averaged only 15.3 ram. At the end of 2 y, 95% of the restorations exhibited an ideal color match. Also, no evidence of bulk fracture was observed. Finally, the marginal degradation, although not readily observed clinically, probably was related to tensile fatigue. Received September, 17, 1991/AcceptedDecember20, 1991 Address correspondenceand reprint requests to: R. B. Mazer School of Dentistry Box 49 University of Alabama at Birmingham Birmingham,AL 35294 USA
REFERENCES Cvar JF, Ryge G (1971). Criteria for the clinical evaluation of dental restorative materials. Washington, DC: United States Department of Health, Education, and Welfare. Feilzer AJ, De Gee AJ, Davidson CL (1988). Curing contraction of composites and glass-ionomer cements. J Prosthet Dent 59(3): 297-300. Gerbo LR, Leinfelder KF, Mueninghoff LA, Russell C (1990). Use of optical standards for determining wear of posterior composite resins. JEsthetic Dent 2:5, 148-152. Goldberg AJ, Rydinge E, Santucci EA, Racz, WB (1984). Clinical evaluation methods for posterior composite restorations. JDent Res 63(12): 1387-1391. J~rgensen KD, Asmussen E (1978). Occlusal abrasion of a composite restorative resin with ultra-fine filler. An initial study. Quint Int 9: 73-79.
J~rgensen KD, Horsted P, Janum O, Krogh J, Schultz J (1979). Abrasion of Class I restorative resins. Scand JDent Res 87: 140-145. Lambrechts P (1983). Properties of dental composites and their impact on clinical performance. Thesis, Leuven, Belgium. Lambrechts P, Braem M, Vanherle G (1985). Accomplishments and expectations with posterior composite resins. In: International Symposium on Posterior Composite Resin, Dental Restorative Materials 521-540, The Netherlands, Peter Szule. Leinfelder KF, Wilder AD, Jr, Teixeira LC (1986). Wear rates of posterior composite resins. J A m Dent Assoc 112: 829833. Leinfelder KF (1987a). Evaluation of criteria used for assessing the clinical performanceofcomposite resins in posterior teeth. Quint Int 18:531-536. Leinfelder KF (1987b). Wear patterns and rates of posterior composite resins. Int Dent J 37:152-157. Lutz F, Imfeld T, Meier C, Firestone A. (1979). Composite versus amalgam - measurements of in vivo wear resistance: one-year report. Quint Int 10:77-87 Lutz F, Phillips RW, Roulet J-F, Setcos JC (1984). Potential posterior composites. An in vitro and in vivo comparison for wear. JDent Res 63: 914-920. Mazer RB, Leinfelder KF (1988). Clinical evaluation of a posterior composite resin containing a new type of filler particle. J Esthetic Dent 1: 66-70. Mazer RB, Leinfelder KF, Russell CM (1989). Mechanism of failure in a microfilled composite resin. JDent Res 68: 233,
Abstr. No. 418. McComb D (1985). Evaluation of clinical wear of posterior composite resins. In: International Symposium on Posterior Composite Resin, Dental Restorative Materials pp.511-517, The Netherlands, Peter Szule. Mitchem J, Gronas DG (1982). In vivo evaluation ofthe wear of restorative resin. J A m Dent Assoc 104: 333-335. Powers JM, Fan PL, Marcotte M (1981). In vitro accelerated aging of composites and a sealant. J Dent Res 60: 16721677. Roulet J-F (1985). In vivo wear measurements of composite resins. In: International Symposium on Posterior Composite Resin, Dental Restorative Materials, The Netherlands, Peter Szule, 365-372. Roulet J-F (1987). Degradation of Dental Polymers. First Edition. New York: Karger, 77-79. Teixeira LC, Isenberg BP, Leinfelder KF (1987). Effect of silane treatment on wear of a posterior composite resin. J Dent Res 66: 167, Abstr. No. 482. Vann WF, Barkmeier WW, Oldenburg TE, Leinfelder KF (1986). Quantitative wear assessments for composite resin restorations in primary molars. J Pediatric Dent 8: 7-10. Wisniewski JF, McCartha CD, Leinfelder KF, Isenberg BP (1988). Two year clinical investigation of a fine particle posterior composite resin. JDentRes 67:120, Abstr. No. 62. Wu W, Cobb EN (1981). A silver staining technique for investigating wear of restorative dental composites. J Biota Mater Res 15: 343-348. Wu W, Mc Kinney JE (1982). Influence of chemicals on wear of dental composites. JDent Res 61: 1180-1183.
Dental Materials~May 1992 18£
