fatigue resistance of composite restorations: effect of filler content

7
Dent Mater 11:7-i 3, January, 1995 Fatigue resistance of composite restorations: Effect of filler content Aung Htang, Masahiro Ohsawa, Hiroshi Matsumoto Department of Endodontics and Operative Dentistry Nagasaki University School of Dentistry, JAPAN ABSTRACT Objectives. The purpose of this study was to evaluate the effects of filler level on the fatigue impact resistance of resin composite. Methods. A series of experimental composite materials was prepared by incorporating a silanized quartz filler (3-5 pm in size) into a light- cured resin matrix of Bis-GMATTEGDMA. The filler contents in the experimental composites varied from 40 to 85 wt%. The composites were placed in standardized Class I cavities prepared in bovine teeth. The specimens were stressed with a repetitive impact load (1.6 x 10~2 joule) with loading cycles ranging from 50,000 to 150,000 times. The cracks induced by cyclic loading were observed on the sectioned surfaces of the tested specimens. Results. The composites with considerably low or high filler content (~60% or >80% by weight) were significantly low in fatigue resistance. The results revealed that an inverse linear relationship tended to exist between filler level and fatigue resistance of the composite materials beyond a certain level of filler content. Significance. Increased filler level does not necessarily improve the fatigue resistance of a resin composite as determined by applying a repetitive impact load. INTRODUCTION Resin composites were widely used in dental practice as esthetic restorative materials for anterior teeth when they were first developed. The development of several bonding systems and improved mechanical and physical properties gave rise to their use in posterior teeth. Although these composites have many virtues, including certain good mechanical properties, high resistance to dissolution and excellent esthetic qualities, their low fracture and wear resistance have limited their use in stress-bearing areas. However, the growing demand for more esthetic restorations and minimal loss of tooth substance in cavity preparations have made posterior composites an attractive alternative to amalgams (Fusayama, 1990). Several clinical trials have reported that the materials evaluated satisfied established clinical requirements for pos- terior composites (Wilson and Norman, 1991). However, the clinical study of Qvist et al. (1990) in Denmark reported that, of the Class I and Class II restorations, half were replaced because of secondary caries and bulk fracture of the fillings. The clinical survey performed by MjGr and Toffenetti ( 1992) in Italian private practices reported that 14% of the replacement of composite restorations was due to bulk and marginal fracture. Many efforts have been made using conventional test methods to investigate the relationship between filler particles in the resin composites and their mechanical properties. However, the optimum level of f?ller content in a composite for stress-bearing areas is still controversial because the filler size influences the maximum weight percent. High filler loading in composite systems seems to be based on the concept of attaining high mechanical properties as determined by conventional mechanical tests. These conventional tests show that increased filler level results in increased hardness and compressive strength &&main et al., 1985; Li et al., 1985; Chung, 1990) and Young’s modulus (Li et al., 1985; Braem et al., 1986; 1989), while coefficient of thermal expansion (Shderholm, 19841,water sorption (0ysaed and Ruyter, 19861, resistance to toothbrush abrasion, and wear by hydroxy apatite are reduced (Germain et al., 1985; Li et al., 1985). Several studies on fracture toughness and stress intensifi- cation factor (K,,), using single-edge notched specimens, have been made by Lloyd and Ianetta (19821, Lloyd (1982; 1983; 1984), Lloyd and Mitchell (1984), Goldman (19851, Ferracane et al. (1987) and De Groot et al. (1988). The results of these experiments have shown that more heavily filled resin composites have higher fracture toughness than do microfills. However, in all these experiments, a static force was applied. This type of force does not commonly occur during mastica- tion. Also, the materials investigated were proprietary products, which have so many unknown variables. Neverthe- less, Ferracane et al. (1987) suggested that there may be an Dental Materials/January 1995 7

Upload: aung-htang

Post on 28-Aug-2016

216 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Fatigue resistance of composite restorations: Effect of filler content

Dent Mater 11:7-i 3, January, 1995

Fatigue resistance of composite restorations: Effect of filler content

Aung Htang, Masahiro Ohsawa, Hiroshi Matsumoto

Department of Endodontics and Operative Dentistry Nagasaki University School of Dentistry, JAPAN

ABSTRACT

Objectives. The purpose of this study was to evaluate the effects of filler level on the fatigue impact resistance of resin composite. Methods. A series of experimental composite materials was prepared by incorporating a silanized quartz filler (3-5 pm in size) into a light- cured resin matrix of Bis-GMATTEGDMA. The filler contents in the experimental composites varied from 40 to 85 wt%. The composites were placed in standardized Class I cavities prepared in bovine teeth. The specimens were stressed with a repetitive impact load (1.6 x 1 0~2 joule) with loading cycles ranging from 50,000 to 150,000 times. The cracks induced by cyclic loading were observed on the sectioned surfaces of the tested specimens. Results. The composites with considerably low or high filler content (~60% or >80% by weight) were significantly low in fatigue resistance. The results revealed that an inverse linear relationship tended to exist between filler level and fatigue resistance of the composite materials beyond a certain level of filler content. Significance. Increased filler level does not necessarily improve the fatigue resistance of a resin composite as determined by applying a repetitive impact load.

INTRODUCTION

Resin composites were widely used in dental practice as esthetic restorative materials for anterior teeth when they were first developed. The development of several bonding systems and improved mechanical and physical properties gave rise to their use in posterior teeth. Although these composites have many virtues, including certain good mechanical properties, high resistance to dissolution and excellent esthetic qualities, their low fracture and wear resistance have limited their use in stress-bearing areas. However, the growing demand for more esthetic restorations and minimal loss of tooth substance in cavity preparations have made posterior composites an attractive alternative to amalgams (Fusayama, 1990). Several clinical trials have reported that the materials

evaluated satisfied established clinical requirements for pos- terior composites (Wilson and Norman, 1991). However, the clinical study of Qvist et al. (1990) in Denmark reported that, of the Class I and Class II restorations, half were replaced because of secondary caries and bulk fracture of the fillings. The clinical survey performed by MjGr and Toffenetti ( 1992) in Italian private practices reported that 14% of the replacement of composite restorations was due to bulk and marginal fracture.

Many efforts have been made using conventional test methods to investigate the relationship between filler particles in the resin composites and their mechanical properties. However, the optimum level of f?ller content in a composite for stress-bearing areas is still controversial because the filler size influences the maximum weight percent. High filler loading in composite systems seems to be based on the concept of attaining high mechanical properties as determined by conventional mechanical tests. These conventional tests show that increased filler level results in increased hardness and compressive strength &&main et al., 1985; Li et al., 1985; Chung, 1990) and Young’s modulus (Li et al., 1985; Braem et al., 1986; 1989), while coefficient of thermal expansion (Shderholm, 19841, water sorption (0ysaed and Ruyter, 19861, resistance to toothbrush abrasion, and wear by hydroxy apatite are reduced (Germain et al., 1985; Li et al., 1985).

Several studies on fracture toughness and stress intensifi- cation factor (K,,), using single-edge notched specimens, have been made by Lloyd and Ianetta (19821, Lloyd (1982; 1983; 1984), Lloyd and Mitchell (1984), Goldman (19851, Ferracane et al. (1987) and De Groot et al. (1988). The results of these experiments have shown that more heavily filled resin composites have higher fracture toughness than do microfills. However, in all these experiments, a static force was applied. This type of force does not commonly occur during mastica- tion. Also, the materials investigated were proprietary products, which have so many unknown variables. Neverthe- less, Ferracane et al. (1987) suggested that there may be an

Dental Materials/January 1995 7

Page 2: Fatigue resistance of composite restorations: Effect of filler content

1 40 21.0

2 60 37.5

3 70 48.3

4 75 54.5

5 80 61.5

8.5 69 4

optimum particle size and volume fraction which produce a fracture-tough composite. The plane strain fracture toughness of dental composites was determined by Pilliar et al. (1986; 1987) using short-rod fracture toughness test specimens, and revealed no strong correlation between K(, and compositional characteristics of the materials. They considered that filler particle size might be more significant in determining &,.. They also suggested the importance of impact strength and fatigue resistance in appraising the mechanical performance of den- tal composites rather than compressive and tensile strengths because repeated occlusal loads act on these materials.

Since the most common causes of tooth fracture have been identified as high impact forces produced by biting on a hard object or uncontrolled contact of opposing teeth (Burke, 19921, similar phenomena are also expected to affect posterior composite restorations. Therefore, in this study, a repetitive impact load was applied to determine the effect of filler content on the fatigue resistance of resin composite. A bonding system that bonds efficiently to both enamel and dentin was applied because the adhesion of resin to the cavity walls improves the fracture resistance of teeth (Eakle, 1986) and marginal stiffness of the restorations (Kubo, 1987; Morin et al., 1984; 1988).

The purpose of this study was to determine the optimum filler content for composite restorative materials used in stress- bearing areas. A series of experimental composite materials, composed of intermediate-sized filler (3-5 pm) at different weight percentages, was employed in this study A prelimi- nary study was carried out by investigating the fatigue resistance of a commercially available resin composite (Clearfil Photo Posterior) to determine an appropriate impact load.

MATERIALS AND METHODS

Materials. In the preliminary experiment, Clearfil Photo Posterior (Batch No. HPB 5003, Kuraray Co. Ltd., Osaka, Japan) was investigated. It is composed of a combination of intermediate-sized quartz particles (average size <5 pm), barium glass particles (average size <l pm) and microfiller in a Bis-GMA/urethane resin. In the main study, specially prepared experimental composites (Kuraray Co., Ltd., Osaka, Japan) with a matrix consisting of 60% Bis-GMA and 40% TEGDMA, and a filler content ranging from 40 to 85% by weight, were utilized. The filler was silanized crushed quartz (3-5 pm in size), and no microfiller was added. The &me treat- ment was done from an aqueous solution of 1% trimethoxy silane before the filler was incorporated into the matrix. The

Fig. 1. A. Schematic representation of the test set-up: a=specimen embedded in epoxy (b); c-specimen holder; dzplatform for specimen holder. B. A standardized cavity: E:enamel; Dzdentin; DEJ=dentin-enamel junction.

filler content ( wt% 1 is given along with the calculated vol!i in Table 1. The resins utilized a light-curing system containing camphorquinone and tertiary amine.

Specimen preparation. Thawed bovine teeth were used in this study. The labial surface of each tooth was ground flat using 800 grit silicon carbide paper under running water. Approximately 1 mm thickness of the enamel layer was preserved. The teeth were embedded in plastic rings (inside diameter of 1.5 cm) with epoxy resin (Struers, Copenhagen. Denmark) after the roots and the incisal third of the crowns were trimmed off. A standardized cavity 2 mm in diameter and 2 mm in depth, was developed in the center of the labial surface of each tooth, and the margin was beveled 45” (135:’ cave-surface angle) until the diameter of the cavity at the margin was 3 mm (Fig. 1). The cavities were prepared using specially designed diamond burs held in a laboratory handpiece, mounted in a milling machine (EM Parallelometer, Royal MGH Co., Ltd., Tokyo, Japan), and operating at low speed under water coolant. A beveled cavity was chosen in order to induce possible cracks development from the beveled angles (Fig. 1 I.

Light-cured Clearfll Liner Bond system (Batch No. 1114. Kuraray Co., Ltd.), including Protect Liner, was applied on the cavity walls according to the manufacturer’s instructions. The resin composites were placed in the cavity using a plastic instrument and cured incrementally in three layers to minimize the polymerization contraction stress (Lutz et al.. 1986, Wieczkowski et al., 1988) and to obtain the maximum cure (Rueggeberget al., 1993). Each layer was irradiated with a visible light-curing unit (Luxor, ICI, Macclesfield, England I for 20 s, but the fYinal layer, which had been placed in excess, was covered and firmly pressed with a cover glass and irradi- ated for 60 s. The specimens were stored in tap water at room temperature (23 + 2%) for 24 h, and then wet grinding for finishing was carried out using 1500-grit abrasive silicon carbide paper. All the procedures were performed under ambient light and temperature.

Preliminary test for determination of impact load. The fatigue impact testing was conducted with a cyclic impact

8 Hang et a/./Effect of filler content on compos/te fafjgue

Page 3: Fatigue resistance of composite restorations: Effect of filler content

lx R= 6.94491 i

1 2 3 4 51 S 7 s,, = 5.1

Rank of Stress Cycles (1 unit = 25,000 cycles)

Fig. 2. Linear plot for the 75% filler composite.

machine. This apparatus is designed in such a manner that the plungers are lifted up by rotating cams to a set position and dropped freely on the specimens. Cylindrical steel styli having flat-surfaced tips (1.5 mm in diameter) were used as plungers. The weight of each plunger could be adjusted by adding or deducting several weights. Four different loads, namely 250 g, 350 g, 450 g and 550 g, dropped from the height of 3 mm, hit the center of the specimen surfaces with a frequency of 4 Hz. A total of 20 specimens filled with Clearfil Photo Posterior composite were subjected to 100,000 cycles of the impact load (five specimens per load). The surfaces of the stylus tips were examined to ensure that they were flat and would uniformly impact the surfaces of the specimens. An Epitex matrix strip (GC Dental Products Corp., Aichi, Japan), 0.05 mm thick, was placed on each specimen surface as an intermediate substance to prevent subsequent surface deterioration or indentation of the specimens. The tested speci- mens were vertically sectioned into four quarters using a slow speed diamond saw (Buehler Isomet, Lake Bluff, IL, USA) under water coolant; thus, eight cut surfaces were obtained from each specimen. Wet grinding of the sectioned surfaces was performed using 1500-g& silicon carbide paper, and imbibed with an aqueous solution of 15% acid Fuchsin for 60 s to identify the crack lines. The specimens were observed under a light microscope (Nikon, Tokyo, Japan), and the number of cracked (failure) specimens was counted. The specimens with failures were photographed to measure the lengths of the cracks. The relationship between the subjected loads and the number of cracked specimens was examined using a &i-square at the 0.05 significance level.

Main study. The specimens prepared with each experimen- tal composite were subjected to varying cycles of the 550 g impact load (at a frequency of 4 Hz and from 3 mm height). A load of 550 g was prescribed for this test because it was assumed as a stress threshold at which all the tested speci- mens failed at 100,000 stress cycles in the preliminary study. The impact energy of the 550 g load, falling freely from the height of 3 mm, was assumed to apply 1.6 x 1O-2 joules (based on E = mgh, where m is the applied load, g is the acceleration

50,000 2 0 0 4 0

75,000 3 0 0 9 0 100,000 4 1 4 16 1

125,000 5 2 10 25 4

150,000 6 4 24 36 16

xx=20 ,T;y=7 By=38 cx2=90 I+21 L 1 of gravity, and h is the height from which the load is applied). The first five specimens of each experimental composite were tested with 50,000 stress cycles, and the stress cycles for the next five specimens were increased by a fixed increment of 25,000 cycles. At the completion of each stress level, the tested specimens were processed by the same procedures as described in the preliminary study Then the increment of load cycles was carried out until almost every specimen of the last five showed failure. The crack lengths measured in each group of five specimens were pooled and averaged.

The results obtained are shown arranged as inTable 2, and the values of x us. y were plotted on a simple linear regression plot (Fig. 2).Y, on they axis, i.e., the value numerically equal to 50% of the sample population in each observation, was used as an estimator to extrapolate S,, on the X axis using a regres- sion equation (y = b, + b,x). S, is the number of stress cycles related to Y,, on the Y axis, and is assumed as median fatigue limit in this study. Having established that Y is linearly related to X (by the rejection of H,,; /3, = 01, a 95% confidence interval for E(Y) (estimated value on the Y axis) when x = 4 (i.e., 100,000 stress cycles) was constructed. The fatigue impact resistance of each composite may be calculated from the median fatigue limit (stress cycles) multiplied by the impact energy (1.6 x 10~2 joules). Linear regression analysis was applied to analyze the results.

RESULTS

The results of the preliminary study are presented in Table 3. The data analysis with a Cl-n-square showed that there was a relationship between the subjected load and the number of cracked specimens (p < 0.05). The results revealed that the 550 g load had the highest magnitude of stress to produce failure in all tested specimens when they were stressed for 100,000 cycles. It indicated that the capacity of the material to resist the impact stress for 100,000 cycles was less than the 550 g load (1.6 x 1O-2 joules) at the given test conditions.

The results of the main study are shown in Fig. 3. The results of the 40% filler composite were excluded from the measurement of the cracks because it displayed a different crack development behavior. Unlike the other materials in which the cracks initiated from the beveled angles, the specimens with the 40% filler composite showed a catastrophic crack development from the surfaces where the plungers had collided with them. The results in this graphic figure show that the composite with 75% filler loading had the longest

Dental Materials/January 1995 9

Page 4: Fatigue resistance of composite restorations: Effect of filler content

5 498k 134

600 - - + 60% - -o-70%

- 500 : --c 75% E : -0- 80%

,y 4~ I -85%

5

4 300 :

‘: 200 - I!

0 loo -

0 -

I. 1. 1. ’ * ’ 3 ’ - J 25 50 75 100 125 150 175

Number of Load Cycles (x1000)

Fig. 3. Fatigue life of the experimental composites containing various quantities of filler (wt%). Abrupt increase of crack length in composites was observable as they reached certain levels of stress cycles.

30

f

30 40 50 60 70 a0 90

FILLER (W1X)

Fig. 4. Fatigue impact resistance of the experimental resin composites related to filler content.

fatigue life, followed by the 60% or 70% filler composites. whereas the 85% filler composite had the shortest.

The fatigue impact resistance of each composite material. calculated from the median fatigue limit, is shown in Fig. 4. The results revealed that a linear relationship between the> filler level and the fatigue impact resistance is obvious up to the 755? filler content, but an inverse linear relationship apparently exists beyond this filler level. An examination of the results showed the following statistically significant differences. The 40% and 85Q filler composites had a sign.5 cantly lower fatigue resistance than all other tested materials. The 60% and 7Oc;/( filler composites were not statisticall)~ different from either the 75% or 80% filler composites, but the composite with 75% filler loading was significantly different from the 80% filler composite.

The scanning electron microscopic view of the crack devel- opment shown in Fig. 5A demonstrates that the cracks seem to have initiated at the beveled angle (large arrow), and propa- gated toward the top surfaces of the samples. The bonding disruption (small arrow) appears to be propagating toward the cavity floor through the tooth-restoration interface from the same sites where the cracks were initiated. A high magnification of the crack line (Fig. 5B) showed that the crack propagation was basically through the resin matrix or filler- matrix interfaces. Figs. 6 and 7 show SEM views of the fracture surfaces produced by the cyclic impact stresses. The fracture surface of the 8% filler composite in Fig. 6 show* that a certain degree of filler-matrix debonding and that frac- ture through the filler and resin matrix had occurred. Th[r smooth and shiny appearance of the fillers indicates fiacturrs through the filler particles. The fracture surface of the 7% filler composite in Fig. 7 reveals that the failure occurred through the resin matrix and filler-matrix interface. The sparsely coatid appearance of the filler surfaces indicated that the failure was through the filler-matrix interface. and thr> other parts of the fracture surface showed failure in the resin matrix.

DISCUSSION

Bovine teeth were used as a substitute for human teeth in this study because of size and availability. Since there is little or no difference for tensile bond strength tests comparing human and bovine teeth (Nakamichi et al., 1983), the restorations created in the bovine teeth closely simulate restorations in human teeth.

In the present study, concomitant disruptions of tooth- restoration bonding were observed in all failure specimens except in the specimens ofthe 40% filler composite. The initia- tion of the crack commonly occurred at the beveled angles and less commonly at the dentin-enamel junction (DEJ). The cracks tended to propagate toward the top surfaces of the specimens where the collisions occurred (Fig. 5). Here, the beveled angles were the common sites where both the initiation of cracks and bonding disruptions fresuently occurred. Whether both of these failures occurred simultaneously or preceded each other is still uncertain. Before the failures, a certain degree of strain (deformation) caused by the impact stresses should certainly take place in the restorations and the tooth substances. The disruptions of bonding on the lateral walls of the cavities were believed to be caused by a shearing stress that caused strains along the tooth-restoration interface. and/or by the discrep-

10 Htang et a/./Effect of fdler content on composite fatigue

Page 5: Fatigue resistance of composite restorations: Effect of filler content

Fig 5A. Scanning electron micrograph showing crack development in a standardized cavity (bar=300 urn); 6. High magnifica- tion of the crack line squared in A (e=enamel, d=dentin, r=resin composite, p=Protect Liner).

Fig 6. SEM of fatigue fracture surface of the 85 wt% filler composite. The fracture surface shows a certain degree of failure through both filler (arrows) and resin matrix.

ancy of the strains between the tooth substances and the com- posite materials. The initiation of cracks might be related to the configuration of the restorations. The curvature created by the bevel in the restorations might contribute to the initia- tion of failure by concentrating stress at these points. The ex- istence of a stronger bond between the materials and enamel substance might restrain the deformation of the restorations related to the enamel walls, while the remaining parts exhib- ited greater strains, causing stress concentration at the DEJs. As a result, the initiation of failure probably occurred at the DEJs in some cases. Several bonding disruptions found at the lateral walls of some non-failure specimens suggested that a bonding failure might precede crack development in this study

Analysis of the data obtained in this investigation provides some insight on the influence of filler level on the fatigue

endurance of composite ma- terials. One significant tind- ing in this study was that the composites with considerably high or low filler content (> 80% or < 60%) showed lower resistance to repetitive impact loads. This result is qualitatively consistent with that from the previous study on composites made by Lange andRadford(1971).The40% filler composite was ex- tremely weak in fatigue resistance, and its failure pattern was also different from all the other experimen- tal composites. The results presented in Figs. 3 and 4 clearly show that a further increase of filler level beyond 75% resulted in composites

with apparently lower fatigue resistance. A resin composite, according to Bowen et al. (19721, is a

combination of two chemically different materials with a distinct interface separating the components and having properties which could not be achieved by any of the compo- nents acting alone. This implies that the inorganic filler must be present in a sufficiently large amount to bring forth the strengthening and stiffening effect. However, the tendency to strive toward a high filler content, in order to increase modulus of elasticity, wear resistance, and compressive strength and to reduce thermal expansion, water sorption, and poly- merization shrinkage of the composite may present some potential dangers even though this development direction is well justified (Soderholm, 1985). There is a significant risk with high filler content, because some mechanical properties such as tensile strength and compressive strength will deteriorate more rapidly, since the particles are close to, or even in contact with, each other (sijderholm, 1985). In the theo- retical determination of shrinkage stresses in composites by Soderholm ( 1984), it was proposed that a crack will not grow as easily in a composite with low 6ller content as in a highly filled composite because high filler content result in compos- ites with substantially higher tangential tensile stress.

Lloyd and Iannetta (1982) documented that the stress intensification factor (Kc) rises with decreasing interparticle spacing (i.e., increasing 6ller content) until a critical spacing is reached, and after this critical volume fraction has been exceeded, K,, is seen to decrease with further filler addition. In the present experiment, low fatigue fracture resistance was observed in both extremes of percent filler content (Fig. 4). Therefore, an important factor which influences fracture resistance or fatigue endurance is believed to be an inter-particulate distance or interfiller space which is occupied by resin matrix. This is due to the fact that the stress is dissipated on the filler and less on the resin binder (Cross et al., 1983). Moreover, the high susceptibility to crack develop- ment in highly filled composites under impact stresses is probably attributed to the brittle nature of quartz filler particles.

Dental Materials/January 1995 11

Page 6: Fatigue resistance of composite restorations: Effect of filler content

Fig 7. SEM of fatigue fracture surface of the 75 wt% filler composite showing that failure occurred in the resin matrix and resin-filler interface (arrow).

In terms of impact force, the capacity of a material to resist the impact elastically and without permanent deformation is directly proportional to its modulus ofresilience (Phillips, 1973). This modulus of resilience will decrease with an increase in the modulus of elasticity as long as the elastic limit is constant. Since increased incorporation of inorganic filler results in an increased modulus of elasticity in composites (Li et al., 1985, Braem et al., 1986; 1989 1, a highly filled composite will possibly have low resistance against impact stresses. Another inherent mechanical property of composites rendered by increased filler content is brittleness (Fujishima et al., 1988). As brittleness is generally considered to be the opposite oftoughness (Phillips, 1973 1, highly brittle composites will have low fracture resistance against impact stresses.

Nakayama et al. ( 1974) hypothesized that if the elastic properties of composite could be matched to those of tooth substances, marginal separation by mechanical deformation during mastication would be minimal, and the loading stress would be transmitted more uniformly across the tooth- restoration interface. In fact, the 60Q filler composite has a rather low filler content, and bonding disruption found at the cavity margin in some samples of this material might be due to its low elastic modulus, causing high elastic deformation under impact stresses. Even though the 60% filler composite was not significantly different from the 70% or 75% filler composites in fatigue endurance, its extremely low tensile strength (40.6 + 4.6 MPa us. 63.4 +- 5.3 MPa) will render it unsuitable for clinical performance.

The scanning electron micrographs revealed that in the 85% filler composite specimens, the fractured filler particles were clearly visible in the fractured surface. However, it should be noted that only the largest particles were involved in cracks, whereas most of the smaller particles appeared to be covered by resin. The resin coating on the filler particles suggested that the adhesion between the filler and matrix was stronger than the resin matrix itself. The particle fracture that occurred in this composite implied that stress concentration around the filler particles was greater than that in other tested

composites. Microstructural analysis of the flracture surfact- in the 75% and 85% filler composites suggested that then> may be a relation between fatigue fracture, particle size, inter- particle spacing, and fracture surface topography Basically, the crack propagation in dental resin composites is mainly through the matrix. If the filler particle itself is the cause ot low-impact resistance in highly filled resin composites, fur- ther improvements in these materials will likely be achieved through 1) adjustment of filler fraction and size; 2) the use of’ less brittle filler materials; and 3) improvements in the tough- ness of the resin matrix itself and its adhesion to the filler particles.

In summary, the results of this study suggest the existence of an optimal or critical filler level in the formulation of resin composites, especially from the aspect of fatigue endurance tested by a dynamic load. This study also demonstrates hou composite materials behave under impact stresses to natural enamel and dentin. The reported results are specific to thrs materials in the standardized cavity and may not be compa- rable to the results of other test situations.

ACKNOWLEDGMENTS The authors gratemy acknowledge Kuraray Co., Ltd. (Osaka i for supplying the experimental materials used in this studs. The authors would like to express their appreciation to Pro;: K. Yasuda, Department of Dental Materials Science, for his helpful discussion and careful reading of the manuscript. The first author (A. Htang) is supported by a grant from the Japa- nese Ministry of Education for his study in Japan.

Department ofEndodontics and Operative Dcntlatr:\

Nagasaki Liniversity School of Ilentistq

1-7-l Sakamoto

REFERENCES

Bowen RL, Barton JA Jr, Mullineaux AL ( 1972). Compositi restorative materials. In: Dental Materials Research. iV& Bur Stand Special Pub1 354. Gaithersburg, Md.: National Bureau of Standards, 93-100.

Burke FJT i 1992,. Tooth fracture in L$CJO and in L’z~x). ,/ lknt 20: 131-139.

Braem M, Lambrechts P, Van Doren V, Vanherle G ( 1986). The impact ofcomposite structure on its elastic response. JDerlt Res 65:648-653.

Braem M, Finger W, Van Doren VE, Lambrechts P, Vanherle G (1989). Mechanical proper-ties and filler fraction of dental composites. Dent Mater 5:346-349.

Chung KH (1990). The relationship between composition and properties of posterior resin composites. JDent Res 69:852- 856.

Cross M, Douglas WH, Field RP ( 1983). The relationship between filler loading and particle size distribution in composite resin technology. J Dent Res 62:850-852.

De Groot R, Van Elst HC, Peters MCRB 11988). Fracture mechanics parameters for failure prediction of compositi resins. J Dent Res 67:919-924.

12 Htang et a/./Effect of filler content on composde fatigue

Page 7: Fatigue resistance of composite restorations: Effect of filler content

Eakle WS (1986). Fracture resistance of teeth restored with Class II bonded composite resin. JDent Res 65149-153.

Ferracane JL, Antonio RC, Matsumoto H (1987). Variables af%cting the fracture toughness of dental composites.JDent Res 66:1140-1145.

Fujishima A, Miyazaki T, Takatama M, Suzuki E, Miyaji T (1988). Tensile properties of Bis-GMAtri-EDMA compos- ite resin. Jpn J Consen, Dent 31:1589-1596.

FusayamaT (1990). Posterior adhesive composite resin: A his- toric review. J Prosthet Dent 64:534-538.

Germain HST, Swartz ML, PhiIlips RW, Moore BK, Roberts TA ( 1985). Properties of microfilled composite resins as in- fluenced by filler content. J Dent Res 64:155-160.

Goldman M ( 1985). Fracture properties of composite and glass ionomer dental restorative materials. J Biomed Mat Res 19:771-783.

Kubo S (1987). Study on the mechanism of marginal fracture of the posterior composite resin: Part I. Relation between cavity wall adhesion of resins and marginal fracture. Jpn J Conserv Dent 30:13-27.

Lange FF, Radford KC (1971). Fracture energy of an epoxy composite system. J Mater Sci 6:1197-1203.

Li Y, Swartz ML, Phillips RW, Moore BK, Roberts TA (1985). Effect of filler content and size on properties of composites. J Dent Res 64:1396-1401.

Lloyd CH (1982). The fracture toughness of dental composites: II. The environmental and temperature depen- dence of the stress intensif?cation factor. J Oral Rehab 9: 133- 138.

Lloyd CH (1983). Resistance to fracture in posterior compos- ites: Measurement of their fracture toughness and a com- parison with other restorative materials. Br Dent J 155: 411-414.

Lloyd CH (1984). The fracture toughness of dental compos- ites: III. The effect of environment upon the stress intensi- fication factor (K,,) after extended storage. J Oral Rehab 11:393-398.

Lloyd CH, Iannetta KV (1982). The fracture toughness of den- tal composites: I. The development of strength and fracture toughness. J Oral Rehab 9:55-66.

Lloyd CH, Mitchell L (1984). The fracture toughness of tooth color restorative materials. J Oral Rehb 11:257-272.

Lutz F, Krejci I, OldenburgTR (1986). Elimination of polymer- ization stresses at the margin of posterior composite resin restorations: A new restorative technique. Quintessence Int

17:777-784. Mjijr IA, Toffenetti F (1992). Placement and replacement of

resin based composite restoration in Italy. Oper Deli 17: 82-85.

Morin D, De Long R, Douglas WH ( 1984). Cusp reinforcement by the acid etch technique. J Dent Res 63:1075-1078.

Morin D, Douglas WH, Gross M, De Long R (1988). Biophysi- cal stress analysis of restored teeth. Dent Mater 4:41-48.

Nakamichi I, Iwaku M, Fusayama T (1983). Bovine teeth as possible substitutes in the adhesion test. J Dent Res 62: 1076-1081.

NakayamaWT, Hail DR, Grenoble DE, Kartz JR ( 1974). Elastic properties of dental resin restorative materials. J Dent Res 53:1121-1126.

0ysed H, Ruyter IE ( 1986). Water sorption and filler charac- teristics of composites for use in posterior teeth. J Dent Res 65:1315-1318.

Phillips RW ( 1973). Skinner’s Science of Dental Materials, 7th ed. Philadelphia: Saunders, 36-42.

Pilliar RM, Smith DC, Maric B (1986). Fracture toughness of dental composites determined using the short-rod fracture toughness test. J Dent Res 65:1308-1314.

Pilliar RM, Vowles R, Williams DF (1987). The effect of envi- ronmental aging on the fracture toughness of dental com- posites. J Dent Res 66:722-726.

Qvist V, Qvist J, Mjiir IA (1990). Placement and longevity of tooth-colored restorations in Denmark. Acta Odont Stand 48:305-311.

Rueggeberg FA, Caughman WF, Cartis JW, Davis HC (1993). Factors ai&cting cure at depths within light-activated resin composites. Am J Dent 6:91-95.

SGderholm KJM (1984). Influence of &me treatment and fXer fraction on thermal expansion of composite resins. J Dent Res 63:1321-1326.

Stiderholm KJM (1985). Filler system and resin interface. In: Vanherle G, Smith D, editors. International Symposium on Posterior Composite Resin Restorative Materials. The Netherlands: Peter SzuIc Publishing Co., 139-159.

Wieczkowski Jr G, Joynt RB, Klockowski R, Davis EL (1988). Effects of incremental versus bulk fill technique on resistance to cuspal fracture of teeth restored with posterior composites. J Prosthet Dent 60:283-287.

Wilson NHF, Norman RD (1991). Five year findings of a multiclinical trial for a posterior composite. J Dent 19: 153-159.

Dental Materials/January 1995 13