Subsurface compression fatigue in seven dental composites

Dent Mater 10:111-115, March, 1994

Subsurface compression fatigue in seven dental composites Lawrence H. Mair Department of Clinical Dental Sciences, University of Liverpool, Liverpool, England, UK

ABSTRACT

Objectives. The purpose of this study was to evaluate subsurface fatigue in seven dental composites. Methods. Cylindrical test pieces were subjected to 2000 compression cycles with a load of 120 N. The area of stained subsurface was measured, and subsequently, the specimens were examined by scanning electron microscopy. Results. The greatest staining occurred in the composites with large quartz fillers, whereas there was little staining in the microfilled materials. Electron microscopy demonstrated very little evidence of cracks, suggesting that silver nitrate stained a network of crazes and microcracks in the subsurface. The pattern of staining indicated that the microcracks formed within the depth of the subsurface rather than by direct nucleation from the indenter. Significance. The different patterns of subsurface damage occurring as a result of compression should be considered when studying the wear and degradation of these materials. INTRODUCTION The purpose of this study was to investigate the pattern of subsurface fatigue in seven dental composites that were subjected to cyclic compression. In tribology (the study of lubrication, friction and wear), this process is known to cause subsurface damage and is a fundamental mechanism of two-body wear (Zum Gahr, 1987). Strictly speaking, the term "delamination" should be used to describe subsurface damage as a result of sliding wear, and the term "fatigue" reserved for situations of rolling wear (Sarkar, 1980). However, in dentistry the term "fatigue" is more familiar and is used synonymously (Mair, 1992a). Two-body wear occurs at sites of direct contact between the occlusal surfaces during mastication and grinding. The relative contribution of subsurface fatigue to wear in these sites is undecided. Roulet (1987) considered it to be of major importance and demonstrated the presence of fatigue cracks on the composite surface. Mair (1991) stained subsurface fatigue zones under palatal attrition scars in composites which had been removed for replacement but found that staining was not a universal finding under all such scars. The author has suggested that the potential for subsurface fatigue depends upon the material rather than the mechanical forces (Mair 1989a), and this

article describes an experiment to test this hypothesis. The basic process causing subsurface damage during two-body wear is that forces applied to the surface cause plastic shear deformation energy to accumulate in the subsurface. This energy increases with repeated loading, resulting in the formation of cracks in the subsurface (Suh, 1973; 1977). Fatigue can only occur in dental restorations if there is a substantial compressive element to the masticatory load in the occlusal contact area. Compressive loads may be applied by direct contact with the opposing tooth or indirectly during the mastication of hard food. Profit et al. (1983) found that the masticatory load on an adult molar during chewing at 2.5 mm opening was 130 + 104 N. However, during maximum biting effort, loads of300 _+200 N were recorded. In this study, silver nitrate was used to investigate the extent of subsurface damage caused by cyclic compression using a load comparable to the clinical situation. The technique of staining subsurface damage with silver nitrate was developed by Wu and Cobb (1981) and modified by Mair (1991). Silver nitrate penetrates into a damaged subsurface as ionic silver and is subsequently precipitated as colloidal silver particles in microdefects within the material (Mair, 1992b).

MATERIALSAND METHODS Seven dental composites were used in the investigation. These materials are described in Table 1. Small cylindrical test pieces (3.0 mm dia x 2.5 mm) were made by inserting the material into PTFE molds and light curing where indicated. Nine specimens were tested for each material. Each specimen was placed in a cylindrical cavity which had been drilled into the base ofa custommade stainless steel bowl (see B in Fig 1). The diameter of the specimen was 0.5 mm less than that of the retaining cavity to allow for slight lateral sliding during compression. The bowl was clamped onto the platform of a compression tester (J.J. Lloyd, Southampton, UK) with the specimen aligned immediately below the stainless steel indenter. The latter was fixed t o the load cell of the tester; the dimensions are shown in Fig. 1. Al ~r lowering the indenter slowly onto the surface, the tester was programmed to give 2000 compression cycles between 0 and 120 N with a crosshead speed of 1 mm/s. All cycling was pt~rformed under distilled water which was placed in the bowl. Afb ~rremoval from the compressive tester, the specimens were immersed in AgNO3 i Dental Materials~March 1994 111

TABLE 1: MATERIALS MONOMERS

%

FILLERS

MEAN SIZE (p.m

P86112

Bis-GMA TEGDMA

73 24

QUARTZ

50

CAVEX Haarlem, Holland

43116

Bis-GMA TEGDMA

65 35

QUARTZ

80 20

HELIOMOLAR1

VIVADENT Schaan, Lichtenstein

372201

Bis-GMA EGDMA

69 22

COLLOIDAL SILICA

10

OCCLUSIN~

ICI Macclesfield, England, UK

PA61B

TUDMA TEGDMA

46 54

BARIUM GLASS

10

P-301

3M Company

P8701102

Bis-GMA TEGDMA

54 46

ZINC GLASS

3

PROFILE -TLC~

SS WHITE Philadelphia, PA USA

358702

Bis-GMA EGDMA

64 34

STRONTIUM GLASS

8

SILUX3

3M Company

6Wl R

Bis-GMA TEGDMA

40 60

COLLOIDAL SILICA

40

MATERIAL

COMPANY

CONCISE1

3M Company St. Paul, MN, USA

CLEARFIL2 POSTERIOR

Bis-GMA EGDMA TEGDMA TUDMA

: : : :

BATCH NO.

BisphenolA Glycidyl Methacrylate Ethylene-glycolDimethacrylate Tri-ethylene-glycol Dimethacrylate Modified Urethane Dimethacrylate

DATA REFERENCES 1Ruyter(1985) 2Pilliar et al. (1987) 3Eliades etal.(1987)

*CLEARFIL-POSTERIOR has a bimodal size distribution of fillers **Size of prepolymerized particles

(3 mol/L) for 24 h. They were subsequently sectioned through the indentation scar with a carborundum wheel and then embedded in mounting resin (Austenal Dental, Harrow, England, UK). The embedded specimens were viewed with incident crossed-polarized light using the optical system described by Mair (1989b). Each specimen was photographed (in monochrome) and a calibration scale was photographed at the same magnification. Films were processed using stock solutions at controlled temperatures to ensure standard development. The negatives were placed in an enlarger with a standard contrast filter to give a relatively well defined edge to the image of stained area. The films were projected onto graph paper, and the outline of the stained area was traced together with the negative of the calibration scale. The area of stain was "calculated" by counting the graph squares and multiplying by the scaling ratio. After examination by light microscopy, all of the specimens were sputter coated with gold and examined by scanning electron microscopy (JEOL JSM-35, Cambridge, England, UK). Statistical analysis indicated that the data were parametric, but that the variances were unequal. This was corrected using logarithmic transforms, and the data were analyzed by one-way Analysis of Variance with the Scheff~ Multiple Range Test used to determine homogeneous subsets (Kirk, 1982). 112 Mair/Fatigueindentalcomposites

RESULTS Fig. 2 shows the stained zones in Clearfil-Posterior, P-30 and Heliomolar. The mean area of stain for each material, together with the standard deviations, is shown in Fig. 3. (The values for Heliomolar and Silux were 2163 _+587 ~m 2 and 1064 +_818 ~tm2, respectively). The results of the statistical analysis are shown in Table 2. The majority of electron micrographs indicated no signs of degradation in the composite subsurface corresponding to the stained areas. In two specimens ofClearfil-Posterior, defects with fibrils of resin running between the surfaces were observed at high magnification (Fig 4).

DISCUSSION Although Table 2 shows four statistically different subsets, it seems reasonable from Fig. 3 to consider Occlusin and P-30 together. The materials in Subset 1, Heliomolar and Silux, were both microfilled materials. Besides having the lowest area of stain, these materials showed the least evidence of physical deformation, and it was often difficult to see the indentation scar. It appeared that in these microfilled materials, there was elastic rather than plastic deformation. This concurs with the view of Oysaed and Ruyter (1986) who found that Heliomolar had a very

their fatigue limit. In the latter experiment, the composite was compressed between flat platens which causes less local deformation. ANOVA (Log transformed values in i~m2) F=12.86 PF <0.001 An interesting feature was noted during the SUBSET pilot studies for these experiments. Initially the specimens were cycled under AgNO3rather than 1 SILUX HELIOMOLAR (1064) (2163) water and processed for microscopy without further staining. This resulted in a much smaller 2 OCCLUSIN area of stain than when the specimens were (18942) stained after cycling. In the latter case, the area of stain increased with time, reaching a maxi3 P-30 mum at 20 h. This indicates that the staining (54987) occurred through a network of cracks nucleating in the subsurface rather than through an advanc4 PROFILE CONCISE CLEARFIL-P ing series of cracks with their origin at the inden(194612) (219486) (225917) tation. Had that been the case, the silver nitrate would have penetrated with the advancing crack. Values in parentheses are means (l~m2) The two alternative crack mechanisms are illustrated in Fig 5. The wide shallow area of stain for P-30 (Fig. 2B) also supports the concept of a subsurface network LOAD CELL as described by Suh (1973; 1977). The cracks form in the subsurface as a result of rupture of strained polymer around inclusions (Jahanmir and Suh, 1977). An alternative explanation for the difference in staining patterns seen when the materials were cycled in silver nitrate may be that compression causes the accumulation ofstrain energy, but the subsurface remains intact. Only on sectioning is the energy released as cracks. However, this explanation is unlikely because all specimens were stained before sectioning, either during the compression cycles or subsequent to cycling. The lack of specific signs ofdegradation by electron microscopy RADIUS OF indicates that the microcracks which allowed staining were C U RVATU R E extremely small and more akin to crazes. These are planes of = 1.75 mm orientation of the polymer molecules with interspersed voids formed as a result of deformation (Kambour, 1973). Although the appearance of the defect in Fig. 4 is characteristic of a craze having filaments ofmatrix across the crack, its size was far larger COMPRESSION PLATFORM than that usually associated with crazes. Crazes are usually 2040 nm across and are known to be susceptible to silver nitrate Fig 1. The compression jig. I = Indenter, S = Specimen, B = Bowl with staining (Kambour, 1964). The defect shown in Fig. 4 may have depression for specimen, C = Clamp fixing bowl onto compression platform. been caused by crack growth through small air bubbles trapped Inset = Indenter Dimensions during mixing of the two-paste material. The stained zone was always brown, as opposed to the black stain which can result from definite yield point and considered that the material was tough, silver nitrate staining. Mair (1992b) has shown that this color is as opposed to brittle. Occlusin and P-30 both have relatively indicative of much smaller microdefects within the composite, small glass fillers compared with Clearfil and Concise which had which supports the concept of crazes rather than cracks. These results differ from those of Roulet (1987) who found the greatest area of stain. However, particle size alone does not explain the result because the fillers of Profile-TLC are slightly definite wear scars in his experimental microfilled materials. smaller than those of Occlusin. A possible explanation for the However, the indentation scars of Roulet indicate that his indenter had a rough milled surface as opposed to the smooth polished differences between materials is the hardness of the fillers. surface in the present study. A rough surface is predisposed to Clearfil and Concise both have quartz fillers which are harder than those of the various glasses used in other composites (Bolz two-body abrasion rather than fatigue, and this may account for and Tuve, 1970). Harder fillers transmit more of the applied load the different result. There was no relationship between the results for three of the to the composite matrix which increases the potential for materials in the present study and their clinical performance. In microcracking. Recently, Wassell et al. (1992) used silver nitrate to investigate subsurface damage under the indentations of a clinical trial, Mair et al. (1990) found that P-30 had the greatest various hardness testers. They also found that the damaged zone attrition and Clearfil-Posterior the least. The probable reason for was related to the structure of the composite, with the least this disparity is because fatigue is only one aspect of attrition. Recent clinical trials have indicated that, despite their lower filler stainable damage in the microfilled material. These findings conflict with those of Wu et al. (1984) who were unable to cause loading, microfilled posterior composites have a wear rate that is a zone of subsurface damage by compressing dental composites to comparable with modern small particle hybrid materials and TABLE 2: CROSS-SECTIONAL AREA OF STAIN: ANOVA AND HOMOGENOUS SUBSETS

T

Dental Materials~March 1994 113

~'-m" E Z i, 0 ILl

tt X

I

t x

=, A

~

i! i ~il iiI!!I~I[II!I~ iii i~~i~i!!il;ii!!il

~i~ ~

u

~

a.

Fig 3. Graph of the cross-sectional stained areas of the seven composites. x=mean value. Error bars depict standard deviations.

Fig 4. SEM (2000x) of the stained zone in ClearfiI-Posterior.

Fig 2. Silver staining in the subsurface of three composites after 2000 compression cycles. A) Clearfil Posterior; B) P-30; C) Heliomolar; ER=Embedding Resin.

114 Mair/Fatigueindentalcomposites

superior to the older large particle hybrids (Freilich et al., 1992; Mazer et al., 1992). The findings in this study, and those of Wassell et al. (1992), that these materials have the least subsurface damage may explain this result. As yet, the process of clinical wear is not fully understood, and there are no laboratory tests which can predict the clinical performance of materials. The present study has indicated that fatigue varies considerably between materials, and it should certainly be considered in the study of restoration wear. A major problem is that commercial materials vary so widely in the type, shape, size and loading of their fillers. Work is currently proceeding at this institution using custom-made composites in which these variables can be controlled.

INDENTER CENTERED CRACKS

-'.-_-.._ NETWORK OF LAMINATED CRACKS TIME 1

TIME 2

Fig 5. Two systems for crack formation. (Top) Crack nucleating from and centered on indenter; (Bottom) network of laminated cracks in the subsurface.

ACKNOWLEDGEMENTS I am indebted to Miss S. Williams for her excellent technical assistance and to Mr. M. Lockwood for making the indentation head and compression jig. Received April 6, 1992 / Accepted February 20, 1994 Address correspondence and reprint requests to: Lawrence H. Mair Department of Clinical Dental Sciences University of Liverpool P.O. Box 147 Liverpool L69 3BX ENGLAND, UK

REFERENCES Bolz RG, Tuve GL (1970). Properties of glass. In: Handbook of Tables for Applied Engineering Science, CRC. Cleveland: Chemical Rubber Co., 249-266. Eliades GC, Vougiouklakis CJ, Caputo AA (1987). Degree of double bond conversion in light cured composites. Dent Mater 3:236-240. Freilich MA, Goldberg AJ, Gilpatrick RO, Simonsen RJ (1992). Three-yearwear ofposterior composite restoration. Dent Mater 8:224-228. Jahanmir S, Suh NP (1977). Mechanics of subsurface void nucleation in delamination wear. Wear 44:17-38.

Kambour RP (1964). Structure and properties of crazes in polycarbonate and other glassy polymers. Polymer 5:143-155. Kambour RP (1973). Crazing and fracture in thermoplastics. J Polymer Sci (Macromolecular Reviews) 7:1-87. Kirk RE (1982). Multiple comparison tests. In: Experimental Design (Procedures for Behavioural Science). 2nd Ed. California: Brookes Coles, 90-127. Mair LH (1989a). Permeability, degradation and wear of dental composites. Ph.D. Thesis, University of Liverpool. Mair LH (1989b). An investigation into the permeability of composite materials. Dent Mater 5:109-11. Mair LH (1991). Staining of in vivo subsurface degradation in dental composites with silver nitrate. J Dent Res 70:215-220. Mair LH (1992a). Wear in dentistry-Areview ofcurrent terminology. J Dent 140-144. Mair LH (1992b). The colors of silver with silver nitrate staining in dental materials. Dent Mater 8:110-117. Mair LH, Vowles RW, Cunningham J, Williams DF (1990). The clinical wear ofthree posterior composites. BrDent J 169:355360. Mazer RB, Leinfelder KF, Russell CM (1992). Degradation of microfilled posterior composite. Dent Mater 8:185-189. Oysaed H, Ruyter IE (1986). Composites for use in posterior teeth: mechanical properties tested under wet and dry conditions. J Biomed Mater Res 20:261-267. Pilliar RM, Vowles R, Williams DF (1987). The effect of environmental aging on the fracture toughness of dental composites. J Dent Res 66:722-726. Profit WR, Fields HW, Nixon WL (1983). Occlusal forces in normal and long-faced adults. J Dent Res 65:566-571. Roulet JF (1987). Degradation ofDental Polymers. Basel: Karger, 60-90. Ruyter IE (1985). Monomer systems and polymerisation. In: Vanherle G, Smith DC, editors. Posterior Composite Dental Restorative Materials. Amsterdam: Peter Szulc, 109-137. Sarkar AD (1980). Friction and Wear. London: Academic Press, 165-190. Suh NP (1973). The delamination theory of wear. Wear 25:111124. Sub NP (1977). An overview of the delamination theory of wear. Wear 44:1-17. Wassell RW, McCabe JF, Walls AWG (1992). Subsurface deformation associated with hardness measurements of composites. Dent Mater 8:218-223. Wu W, Cobb EN (1981). A silver staining technique for investigating wear ofrestorative dental composites. JBiomed MaterRes 15:343-348. Wu W, Toth EE, MoffaJF, Ellison JA (1984). Subsurface damage layer of in vivo worn dental composite restorations. J Dent Res 63:675-680. Zum Gahr K-H (1987). Microstructure and Wear of Materials. Amsterdam: Elsevier, 80-92.

Dental Materials~March 1994

115