In vitro contact wear of dental composites

In vitro contact wear of dental composites

Dental Materials (2004) 20, 63–71 http://intl.elsevierhealth.com/journals/dema In vitro contact wear of dental composites Venkata S. Nagarajana,1, S...

354KB Sizes 3 Downloads 356 Views

Dental Materials (2004) 20, 63–71

http://intl.elsevierhealth.com/journals/dema

In vitro contact wear of dental composites Venkata S. Nagarajana,1, Said Jahanmira,*, Van P. Thompsonb a

National Institute of Standards and Technology, Gaithersburg, MD 20899, USA New York University College of Dentistry, New York, NY 10010, USA

b

Received 2 April 2002; received in revised form 20 January 2003; accepted 19 March 2003

KEYWORDS Dental composites; Resin matrix; Ceramic fillers; Wear; Dental restorations

Summary Objective. The aim of this study is to determine the in vitro two-body contact wear mechanisms of three medium filled composites and compare these with a highly filled composite previously investigated. Materials and methods. Three commercial dental composites with filler mass fraction loading of 75– 76% were evaluated. Two of the composites contained Ba – B –Alsilicate glass fillers and fumed silica with different particle sizes and distributions. One of these composites contained a fairly uniform distribution of filler particles ranging in size from 1 to 5 mm, whereas the particle size distribution in the second composite was bimodal consisting of small (less than 1 mm) and large (about 10 mm) particles. The third composite contained Ba –Al-silicate glass and silica with a filler particle size of approximately 1 mm. The composite disks were tested for wear against harder alumina counterfaces. Wear tests were conducted in distilled water using a pin-on-disk tribometer under conditions that represented typical oral conditions (sliding speed of 2.5 mm/s and contact loads ranging from 1 to 20 N). The wear tracks were analyzed by scanning electron microscopy (SEM) and Fourier transform infrared (FTIR) spectroscopy to elucidate the wear mechanisms. The chemical composition of the water solution collected after the tests was determined using an inductively coupled plasma-mass spectrometer (ICP-MS) to detect possible chemical changes, e.g. dissolution of trace elements due to submersion or wear. The wear results were compared with those reported in an earlier study on a highly filled composite containing predominately alumino-silicate glass fillers and alumina at a filler loading of 92%. Results. The differences in two-body wear rates between the three medium filled composites were not statistically significant ðp , 0:05Þ indicating that the variations in filler particle size and slight differences in chemical composition of the glass fillers do not affect the in vitro wear rates of these composites. Wear rates of these medium filled composites, however, were significantly lower than the highly filled composite ðp , 0:05Þ. SEM, FTIR and ICP-MS analyses suggested that wear in the medium filled composites occurs by a complex set of processes involving tribochemical reactions between filler particles and water, formation of surface films containing a mixture offiller fragments and reaction products, and film delamination, as well as dissolution of the reaction products. Significance. This study reveals that subtle changes in the filler particle size and small differences in filler composition do not significantly affect the two-body wear behavior of medium filled composites. However, the chemistry of filler particles plays an important role in altering the wear performance of composites when significant changes are made in the chemical composition of the fillers and when the filler loading is increased. Published by Elsevier Ltd on behalf of Academy of Dental Materials.

*Corresponding author. Present address: MiTiHeart Corporation, P.O. Box 83610, Gaithersburg, MD 20883, USA. Tel.: þ301-869-9720; fax: þ 301-869-9724. E-mail address: [email protected] 1 Present address: Materials, Process and Product Development Engineering Consultants, Duluth, GA 30097, USA. 0109-5641/$ 30.00+0.00 Published by Elsevier Ltd on behalf of Academy of Dental Materials. doi:10.1016/S0109-5641(03)00069-1

64

Introduction Dental composites containing inorganic filler particles in a polymeric resin matrix are gaining acceptance for Class I and II restorations due to their esthetics and durability.1 – 3 These materials are also being utilized for indirectly fabricated inlay, onlay and crown restorations.4 The properties of dental composites depend on the type of resins used, filler content,5 – 7 particle size and distribution of the filler,8,9 and the degree of cure and ageing.9,10 Under clinical conditions wear of the restorations is often classified into two types: contact wear caused by the contact of tooth to the restoration and non-contact wear due to food particles caught in between occlusal surfaces.11 A number of studies have been reported with regard to the wear of composites using various wear test techniques and test conditions.12 – 15 Suzuki et al.15 reported that composites containing harder fillers (e.g. quartz) exhibited less contact and noncontact wear compared to those composites that contained softer filler particles (e.g. Ba-silicate). Suzuki et al.15 also pointed out that composites containing quartz fillers caused greater antagonistic enamel wear than Ba-silicate filled composites. Condon and Ferracane14 evaluated wear of composites using a multi-mode oral simulator to approximate clinical conditions. They noted that antagonistic enamel wear was the greatest for the composites containing the largest filler particle sizes. More recently, two-body wear of a range of composites was compared to dental amalgam.16 Most of the above studies were focused on providing comparative rankings of the dental materials using wear test machines that simulated the oral conditions. Only a few in vitro studies have investigated the fundamental mechanisms that lead to wear of dental composites.17 – 20 Clinical contact wear can be modeled using simple pin-on-disk tribometers,p particularly, to study to the fundamental relationships between the microstructure and the wear mechanisms.21 – 24 In a recent study, Nagarajan et al.20 found that the silica and alumina filler particles in a highly filled composite reacted with water during sliding. The wear of this highly filled composite occurred through a combination of processes that included the formation of aluminum hydroxide and silicon hydroxide by tribochemical reactions, formation of surface films containing a mixture of filler fragments and reaction products, p Standard test method for wear testing with a pin-on-disk apparatus, ASTM Standard G 99, Annual Book of ASTM Standards. Vol. 03.02, American Society for Testing and Materials, West Conshohoken, PA, pp. 401–405.

V.S. Nagarajan et al.

and mechanical detachment of the surface films, as well as dissolution the reaction products. The present study was initiated to determine whether such tribochemical reactions play a role in the wear process of medium filled composites and to evaluate the possible affects of filler particle size and chemical composition on the wear behavior. In this study, three composites with different filler particle size distributions having a slight difference in chemical composition are examined. The results are compared with those presented in the recent study20 conducted on a highly filled composite.

Materials and methods Three medium filled composites were used in the study. Two of the composites, TPH and Ful-Fil (Caulk Division, Dentsply, Milford, DE, USA)† contained inorganic filler particles with a total filler content of 76% mass fraction (approximately 53% by volume). The average size of the filler particles, according to the manufacturer, was about 1 mm for TPH and 6 mm for Ful-Fil with the filler particles in both composites consisting of Ba-containing alumino-boro-silicate glass and about 5% fumed silica with an average particle size of about 50 nm. The resin matrix is a combination of UDMA (urethane dimethacrylate) and TEGDMA (triethylene glycol dimethacrylate) in both composites. The third composite, Artglass (Heraeus Kulzer, Irvine CA, USA), contained a total of 75 mass% fillers consisting of Ba –Al-silicate glass including 20 mass% of fine silica powder. According to the manufacturer, the average filler particle size in this composite is less than 1 mm and a multifunctional methacrylate resin mixture which includes TEGDMA is used, suggesting a highly cross-linked structure. The results from the present study were compared with those on a highly filled composite, HCEstenia (Kuraray, Osaka, Japan) reported recently by Nagarajan et al.20 This composite contained about 92 mass% inorganic fillers in a polymeric resin matrix consisting of a polymerizable tetra functional monomer. The filler particles consisted of 16% ultrafine alumina particles (average particle size of 20 nm) and 76% fine alumino-silicate glass particles (average particle size of 1.5 mm). Disks of 24 mm in diameter and 4 mm in thickness were prepared by molding with commercial pastes (containing powder and resin) supplied by † Information on product names, manufacturers, and suppliers is included in this paper for clarification. This does not imply endorsement by the National Institute of Standards and Technology.

In vitro contact wear of dental composites

the manufacturers. The disk specimens were lightcured for a total of 120 s (90 s on one side and 30 s on the other) using a strobed xenon light source (Dentacolor XS, Heraeus Kulzer, Irvine, CA, USA) with an on – off duty cycle of 20 and 80 ms. The disks were polished sequentially using diamond polishing compounds of 10, 5 and 1 mm in size to an average surface roughness Ra of 0.05 mm. The surface roughness was measured using the profilometer described below. The Vickers hardness of the composites was determined using a standard Tukon microhardness tester (Wilson Instruments, Bridgeport, CT, USA) under a normal load of 1 N. The mean hardness value and one standard deviation were calculated for five repeat tests ðn ¼ 5Þ: Wear tests were conducted using a conventional tribometer (CSEM, Geneva, Switzerland) in a ballon-disk configuration as described elsewhere.25 In this configuration, a stationary ball contacts a rotating disk. If both materials are subjected to wear, the rotation of the disk produces a circular wear pattern on the disk and a relatively flat wear scar on the ball. High-purity alumina balls (AD-995, Coors, Golden, CO, USA) with a hardness of 14.7 GPa were used as the counterface material due to the relatively higher hardness of alumina (by a factor of 15) compared to the composites. The alumina balls (12.7 mm in diameter) had an average surface roughness Ra of about 0.01 mm. Prior to the wear tests, the alumina balls and the composite disks were cleaned in acetone and rinsed in distilled water. The normal load in the tests ranged from 1 to 20 N. The sliding speed at the contact between the ball and the disk was kept constant at 2.5 mm/s. The range of contact loads and the sliding speed used in these experiments were consistent with the average in vivo conditions as reported by DeLong et al.12 A total sliding distance of 3 m (corresponding to approximately 6000 contact cycles assuming a sliding occlusal contact length of 0.5 mm) was used in the wear tests. Wear tests were conducted while immersing the disks in distilled water at room temperature. The balls and the disks were reused in the experiments. Each test, however, was conducted on a new unworn region. The ball was rotated to a new unworn region, and was translated across the disk to produce a new circular wear track. The friction force was monitored with a load transducer during each experiment and was recorded using a data acquisition system (Viewdac, Keithley Instruments, Rochester, NY, USA). Following the experiments the wear volume for each test was directly measured. The average cross-sectional area of each wear track on the disk was determined from surface profiles recorded

65

at three different locations across the wear track using a stylus profilometer (Perthen, Model S-530, Germany). The wear volume for each test was calculated by multiplying the average ðn ¼ 3Þ crosssectional area of the wear track with the circumferential contact length. The employed volume measurement procedure is the most common and direct method for wear calculation and does not require complicated mathematical expressions or simplifying assumptions. The wear tests were repeated three times at each load on three different disk samples of each composite and the average and standard deviation in the wear volume were calculated. The wear volumes were subjected to statistical analysis by ANOVA and the means were compared for significant differences with Scheffe Contrasts ðp , 0:05Þ. In order to obtain a fundamental basis for the contact wear study, microstructure and chemical composition of the composites were characterized by SEM (SEM, Model S-530, Hitachi, Japan) and micro-FTIR (FTIR, Magna-IR 750, Nicolet, Madison, WI, USA). The wear tracks of the composites were examined in the SEM in the back-scattered image mode to assess possible changes in the morphological characteristics of the filler particles, in order to determine the wear mechanisms. The wear tracks were analyzed for possible chemical changes using micro-FTIR spectroscopy directly on the sample surfaces in the reflection mode. Water samples collected after the wear tests for each group of samples were combined for chemical analysis (total of approximately 40 ml). The presence of elements in the water associated with the composite helps to determine whether tribochemical reactions and dissolution may have occurred during the wear process. Water samples were placed in closed polyethylene containers for several days before analysis to allow precipitation of the wear debris, if any. An inductively-coupled plasma spray mass-spectrometer (ICP-MS, Elan 5000, Perkin Elmer-Seiex, Ontario, Canada) was used for the analysis owing to its sensitivity in detection of trace (i.e. few nanograms per milliliter) elements in solution.26 To analyze for the possibility of elemental dissolution by static immersion, polished but unworn specimens were stored in distilled water for 96 h and the supernatant water was analyzed as above. The duration of static immersion tests was much larger than the immersion time during the wear tests by more than a factor of 100 to accentuate possible leaching of elements. Prior to the analysis of the test water samples, distilled water was analyzed as a reference and the results were subtracted from those obtained on the test samples. The relative standard uncertainty

66

V.S. Nagarajan et al.

associated with elemental concentration obtained in the ICP-MS analysis is about ^ 20%.

Results The hardness values for TPH, Ful-Fil, and Artglass were 0.73 ^ 0.01, 0.60 ^ 0.02, and 0.45 ^ 0.01 GPa. In comparison, the hardness of the highly filled HC-Estenia was much higher at 1.67 ^ 0.08 GPa. The hardness of the alumina ball used as the antagonist in the wear tests was 14.7 GPa, higher than the hardness of the composites by a factor of 9 – 32. The SEM micrographs in Fig. 1 show the microstructure of the three medium filled composites as seen on polished surfaces. (The following designation is used herein to identify the composites: A ¼ TPH, B ¼ Ful-Fil, C ¼ Artglass, and D ¼ HCEstenia.) Composite A (Fig. 1(a)) contains a fairly uniform distribution of filler particles ranging in size from less than 1 mm to about 5 mm. It should be noted that these composites also contain very fine filler particles with dimensions much less than 1 mm that cannot be easy resolved in the SEM. Composite B (Fig. 1(b)) has a bimodal distribution of small (about 1 mm) and large (about 10 mm) particles. Composite C (Fig. 1(c)) has a fairly uniform distribution of fine filler particles with a particle size of about 1 mm. The micrograph in Fig. 1(c) shows several large round and irregular particles with dimensions ranging from 3 to 8 mm. The source and nature of these particles are unknown and their potential influence on wear is uncertain.

Figure 1 SEM micrographs showing polished surfaces of medium filled composites (a) composite A, (b) composite B, and (c) composite C.

Figure 2 Wear volume of medium filled composites (A, B and C) as a function of load. Wear data for a highly filled composite (D) is also shown for comparison. The plot includes the mean ^ one standard deviation.

The mean values of the wear volumes for the three medium filled composites (A, B, and C) are shown in Fig. 2 as a function of load. Note an increase in the wear volume as the load is increased from 1 to 10 N. However, the wear volume becomes relatively independent of load as the load is increased from 10 to 20 N. There is no statistically significant difference ðp . 0:05Þ between the wear volumes of the three medium filled composites except at 5 N and the highest load of 20 N. These small differences, though statistically significant, are not of practical importance. The maximum depth of wear for these tests ranged from 0.1 to 10 mm. Examination of the alumina balls used in these experiments as the counterface showed no measurable wear due to the much higher hardness of the alumina compared with the composites. Only a few scratches were seen on the alumina balls where contacts were made during the wear tests. The wear volume data obtained on the highly filled composite D reported in Nagarajan et al.,20 are also included in Fig. 2 for comparison. The relationship between wear volume and the load is similar for the all four composites. The wear volumes of the medium filled composites are nearly an order of magnitude smaller than that of the highly filled composite. The differences in wear volume between each medium filled composite and the highly filled composite are statistically significant ðp . 0:05Þ at all loads. The average coefficients of friction calculated from the friction forces for the three medium filled composites ranged from 0.4 to 0.6 and are shown in Fig. 3. This reveals that the coefficient of friction decreases with increasing load. The coefficient of friction for the highly filled composite, also shown in the figure, follows a similar trend; but it is generally higher than the coefficients of friction for

In vitro contact wear of dental composites

Figure 3 Coefficient of friction of medium filled composites (A, B and C) as a function of load. Data for a highly filled composite (D) is also shown for comparison. The plot includes the mean ^ one standard deviation.

the medium filled composites. Statistical analysis was not conducted on the friction data since the primary focus of this paper was on the relative wear rates and wear mechanisms. The SEM micrographs of the wear tracks at 1 N load for the medium filled composites are shown in Fig. 4. The wear track of composite A (Fig. 4(a)) is covered with a thin film-like substance, but the filler particles are still visible. Several microcracks are seen in the surface film, apparently associated with the interfaces between the filler particles and the resin matrix. In some locations, material has been removed forming shallow pits with a size equivalent to several filler particle diameters. Since the microcracks propagate a short distance effecting the removal of small fragments from the surface, the term ‘microfracture’ is used here to describe the removal process. The structure of the wear track for composite B (Fig. 4(b)) is similar to

Figure 4 SEM micrographs showing the wear tracks at 1 N load for (a) composite A, (b) composite B, and (c) composite C.

67

Figure 5 SEM micrographs showing the wear tracks at 10 N load for (a) composite A, (b) composite B, and (c) composite C.

that of composite A (Fig. 4(a)) with respect to the surface film and microcracks. The wear track of composite C (Fig. 4(c)) is covered with a similar but thicker surface film almost masking the filler particles. Several microcracks and fractured regions are also seen on the wear track. Typical appearance of the wear tracks at a higher load of 10 N is shown in Fig. 5 for the three medium filled composites. Examination of the wear tracks indicated that the surface films were not uniform and that the film had been removed in many locations revealing the filler particles as seen in Fig. 5(a) and (b) for composites A and B. Although the filler particles in composite C are more clearly seen on the wear track at the 10 N load (Fig. 5(c)) as compared to the lower load (Fig. 4(c)), the surface film is still visible in some locations. FTIR analysis was performed to evaluate the chemical nature of the surface films on the wear tracks. This analysis is used as a means to distinguish the surface structural changes that might have taken place as a result of sliding wear and not necessarily to identify the chemical composition of the composites. Therefore, the FTIR spectra obtained from the wear tracks are compared with the spectrum from unworn regions, and only the changes in the spectra that are seen on the wear tracks are analyzed. The spectra obtained on the wear tracks at 1 and 10 N (spectra 1 and 2) and spectra collected from unworn surface regions away from the wear tracks (spectrum 3) are shown in Fig. 6 for the three medium filled composites. Two new peaks, one at 960 cm21 (corresponding to Si– O – H), and the other a broad peak at about 3500 cm21 (corresponding to hydroxyl groups) were identified on the wear tracks of composites A and B.

68

V.S. Nagarajan et al.

Figure 6 FTIR spectra of the three medium filled composites: (a) composite A, (b) composite B, and (c) composite C (spectra 1: wear track at 1 N, spectra 2: wear track at 10 N, and spectra 3: unworn surface).

While the broad peak at 3500 cm21 was also detected on composite C, the peak at 960 cm21 was not resolved due to spectral reflectance on this composite. The peak associated with the hydroxyl groups is also seen on the unworn regions of composite C; however, the peak height is larger on the wear track at 10 N load (spectrum 2, Fig. 6(c)). Composite C also shows a broad peak near 2900 cm21 that could be associated with the organic matrix. The FTIR spectra suggest the possibility of tribochemical reactions between the glass filler particles and water forming hydrated

silicon oxides on the wear tracks. However, the FTIR results alone are not sufficient evidence for such reactions. The ICP-MS results are summarized in Tables 1 – 3 for water collected after the wear tests at two different loads and the water collected after static immersion tests. While the water samples from static immersion and those from the wear tests showed varying amounts of Si, because of possible interference between the signals for Si and N, it is doubtful whether the Si results are significant. The results for static immersion tests in Tables 1 –3 for

Table 1 Concentration of elements in water samples used with composite A after static immersion and wear tests determined by ICP-MS.

Table 2 Concentration of elements in water samples used with composite B after static immersion and wear tests determined by ICP-MS.

Elements

Concentration after static immersion (ng/ml)

Concentration after wear test at 2 N (ng/ml)

Concentration after wear test at 10 N (ng/ml)

Elements

Concentration after static immersion (ng/ml)

Concentration after wear test at 2 N (ng/ml)

Concentration after wear test at 10 N (ng/ml)

Al B Ba Sr

0.5 1.1 918 14.5

0.4 0.1 89 1.6

0.6 0.2 122 2.4

Al B Ba Sr

0.5 0.7 468 8.9

18.1 0.1 16 1.3

16.3 0.2 35 2.2

In vitro contact wear of dental composites

Table 3 Concentration of elements in water samples used with composite C after static immersion and wear tests determined by ICP-MS. Elements

Concentration after static immersion (ng/ml)

Concentration after wear test at 2 N (ng/ml)

Concentration after wear test at 10 N (ng/ml)

Al B Ba Sr

0.5 1.4 225 1.4

0.6 0.1 28 0.3

0.9 0.2 90 0.8

the three medium filled composites indicate abundance of Ba and small amounts of Al, B, and Sr. The concentration of these elements, except for Al, is higher in the static immersion tests than in the water samples from the wear tests because of the much larger immersion times used in the static aging experiments. The presence of Al in water after wear tests, particularly with composite B, could be associated with the slight abrasive wear of alumina balls. The elemental concentrations for B, Ba, and Sr in the water samples from the wear tests have increased by 37 to 125% with increasing the load from 2 to 10 N. Considering the 20% uncertainty associated with this analysis and the fact that the duration of sample immersion in wear tests was the same, these results suggest a general correlation between the load (and therefore the wear volume) and concentrations of these elements in the water samples after wear testing. Note that Ba is used in the glass fillers in all three composites and that Sr is a common element in most glasses. Presence of B in the water samples from the composite C suggests that boron was used in the glass filler particles in this composite, similar to the other two composites.

Discussion The results on wear volume of the three medium filled composites suggested that the two-body wear behavior of these materials depends on the load employed during the test. The wear volume increased by one order of magnitude when the load increased from 1 to 10 N. Examination of the worn surfaces indicated that a thin film was present on the wear track irrespective of the load. While at low loads the wear track was almost completely covered, the film was discontinuous at higher loads and was partially removed. The increase in wear volume was due to film detachment at the higher loads. At loads larger than 10 N,

69

wear was independent of load despite the fact that wear occurred by mechanical detachment of the surface films. This reduction in wear volume with increasing load has been also observed in sinusoidal loading, two-body wear studies of another medium filled composite (Z-100, 3M Dental Products).27 Film detachment at higher loads (18 N) was also noted in that study while cross-sections through the wear track indicated subsurface damage at this load. The wear results, in our study, suggest that the wear rate of the three medium filled composites is not affected by the differences in filler particle size or slight differences in filler chemical composition. The present study is in agreement with the wear test results obtained by Sibelca et al.28 in which a wear tester developed by Suzuki and Leinfelder29 was used and found no statistically significant difference between the wear rates of TPH and Ful-Fil after 25,000 cycles against a steel stylus. The wear volume of the medium filled composites was an order of magnitude smaller than that of the highly filled composite. The wear track on the highly filled composite was covered20 with a relatively thick film composed of a mixture of reaction products (hydrated silica and alumina) and fragments from the base composite (i.e. polymeric resin and small gamma-Al2O3 crystallites). It was hypothesized by Nagarajan et al.20 that fragments from the substrate material were removed by a microfracture process and that the shearing action associated with sliding at the contact mixed the hydrated reaction products with the fragments. The wear process consisted of the formation and removal of the thick surface film.20 The SEM observations on the wear tracks of the three medium filled composites evaluated in this study suggest a similar wear process, i.e. delamination of thin surface films. This process is similar to the delamination wear mechanism.30 According to the delamination wear theory, which was developed to describe the wear process in crystalline materials, thin, sheet-like wear particles form as a result of microcrack generation in the plastically deformed near surface region. The microcracks propagate parallel to the surface following the primary shear stress trajectory before becoming unstable and changing direction towards the surface.31 The delamination process probably occurs more readily in the dental composites, in the present study, due to the discontinuity at the substrate/film interface. Although the differences in wear between the three medium filled composites were not statistically significant, their mean wear volume values appear to follow the hardness. In fact, the harder highly filled composite suffers a much higher wear

70

than the softer medium filled composites. Wear rate is often inversely correlated with hardness, especially for materials in which wear is controlled by mechanical processes, i.e. through deformation and fracture.25 The opposite trend observed between wear and hardness for the composites suggests that other non-mechanical processes may play a role during wear of these materials. The FTIR results on the wear tracks of the medium filled composites suggested the presence of hydrated products, indicating the possibility for tribochemical reactions between the filler particles (silicate glass and fumed silica) and water forming hydrated silica products. The surface film on composite D, however, was composed of hydrated alumina products associated with the gammaalumina particles in that material in addition to hydrated silica.20 Contact wear can occur by material removal through dissolution of reaction products in water.36 As hydrated silicon oxide can readily dissolve in water, it is not surprising to find elemental B, Ba, and Sr (from the glass filler particles) in the water solution after the wear tests and static immersion experiments. Similar observations have been reported by McKinney et al.,19 Shin et al.,32 and So ¨derholm and Mukherjee33 for different glassfilled composites. The results of this study suggest that the dental composites evaluated in this investigation are subjected to a complex set of processes at the contact that include formation of tribochemical reaction products, microfracture of small fragments from the substrate, formation of surface films containing a mixture of wear fragments and reaction products, and film removal by mechanical detachment. Dissolution of reaction products also plays a role in the wear process, but the total contribution of dissolution to the overall wear rate is not known. The observed load-independent wear behavior at high loads may be related to a combination of tribochemical and mechanical processes, since material removal by mechanical action alone should be load-dependent. The processes of tribochemical reactions, film formation, film fracture, and dissolution of reaction product can occur simultaneously. At high loads, the rate of film growth may be sufficiently large to compensate for dissolution or increase in mechanical wear; thus, resulting in a load-independent behavior, as explained by Nagarajan et al.20 The role of resin matrix in tribochemical reactions and film formation needs to be investigated further, since water aging of the composites could result in differences in mechanical properties.10 For example, reduced elastic modulus and fracture

V.S. Nagarajan et al.

toughness as a result of extended submersion in water might change the contributions of the observed wear mechanisms. So ¨lderholm et al.34 have shown, in a well-controlled three-year clinical study of eight composite formulations in Class I and II restorations, that UDEMA-based matrix composites had better wear resistance than the bis-GMA formulations. Filler type and loading had little apparent effect on the wear observed. In the comparisons made in our study only composite D had a bis-GMA formulation. This might be a component in the higher wear rate observed and the tribochemical wear mechanism postulated.20 It should be noted that the present investigation was carried out to assess the fundamental contact wear processes of dental composites. Therefore, no attempt was made to assess the clinical relevance of the results. The salivary films may affect the friction coefficients and the wear rates.35 Biofiolms of organic compounds may interfere with the tribochemical reactions and the reciprocating motion in the oral cavity may reduce the propensity for film formation on the restorations. These issues need to be examined before the present results are used in predicting the in vivo performance of the composites

Conclusions The differences in particle size and distribution and slight differences in chemical composition of the filler particles do not affect the two-body wear mechanism and the wear behavior of medium filled composites tested in vitro conditions. Wear occurs by simultaneous processes consisting of tribochemical reactions between filler particles and water, dissolution of hydrated products, formation of surface films containing a mixture of filler particles and reaction products, and delamination of the surface films. A comparison of the wear behavior of the medium filled composites with a highly filled composite containing alumina fillers in addition to silica suggests that the particle chemistry and filler loading may be more important than the changes in particle size and distribution in altering the two-body wear performance of these composites.

Acknowledgements We thank Lee Yu for the experimental assistance during ICP-MS analysis. We also thank Paul Hammesfahr, John Sibelca (L.D. Caulk/Dentsply, DE)

In vitro contact wear of dental composites

and Sinichi Sato (Kuraray, Japan) for helpful discussions and for providing the composite materials for the study. Assistance of Patricia McGuiggan, Janet Quinn, Piotr Hryniewicz and Lewis Ives (NIST) are gratefully acknowledged. This project was funded by NIH-NIDCR under a contract to the University of Medicine and Dentistry of New Jersey (Contract No. 1 P01DE10976-01A1).

References 1. Sulong MZAM, Azis RA. Wear of materials used in dentistry: a review of the literature. J Prosthet Dent 1990;63:342—9. 2. Mair LM. Wear in dentistry-current terminology. J Dent 1992; 20:140—4. 3. Willems G, Lambrechts P, Braem M, Vanherle G. Composite resins in the 21st century. Quintessence Int 1993;24:641—58. 4. Okada K, Omura I, Yamauchi J. A newly developed composite crown and inlay material. J Dent Res 1997;76:195. Abstr. No. 1452. 5. Johnson WW, Dhuru VB, Brantley WA. Composite microfiller content and its effect on fracture toughness and diametral tensile strength. Dent Mater 1993;9:95—8. 6. Htang A, Ohsawa M, Matsumoto H. Fatigue resistance of composite restorations: effect of filler content. Dent Mater 1995;11:7—13. 7. Jaarada MJ, Wang RF, Lang BR. A regression analysis of filler particle content to predict composite wear. J Prosthetic Dent 1996;77:57—67. 8. Durand JM, Vardavoulias M, Jeandin M. Role of reinforcing ceramic particles in the wear behavior of polymer-based model composites. Wear 1995;181—183:833—9. 9. Condon JR, Ferracane JL. In vitro wear of composite with varied cure, filler level, and filler treatment. J Dent Res 1997;76:1405—11. 10. Indrani DJ, Cook WD, Televantos F, Tyas MJ, Harcourt JK. Fracture toughness of water-aged resin composite restorative materials. Dent Mater 1995;11:201—7. 11. Wilder Jr. AD, Bayne SC, Heymann HO. Long-term clinical performance of direct composites. Acad Dent Mater Trans 1996;9:151—69. 12. DeLong R, Sasik C, Pintado MR, Douglas WH. The wear of enamel when opposed by ceramic systems. Dent Mater 1989; 5:266—71. 13. Bloem TJ, McDowell GC, Lang BR, Powers JM. In vivo wear. Part II: Wear and abrasion of composite restorative materials. J Prosthet Dent 1988;60:242—9. 14. Condon JR, Ferracane JL. Evaluation of composite wear with a new multi-mode oral wear simulator. Dent Mater 1996;12: 218—26. 15. Suzuki S, Suzuki SH, Cox CF. Evaluating the antagonistic wear of restorative materials when placed against human enamel. J Am Dent Assoc 1996;127:74—80. 16. Hu X, Marquis PM, Shortall AC. Two-body in vitro wear study of some current dental composites and amalgams. J Prosthet Dent 1999;82:214—20.

71

17. Prasad SV, Calvert PD. Abrasive wear of particle-filled polymers. J Mater Sci 1980;15:1746—54. 18. Jones DW. Synthesis and evaluation of ceramic, polymer, and composite biomaterials. In: Wise DL, Trantolo DJ, Altobelli DE, Yazemski MJ, Gresser JD, Schwartz ER, editors. Encyclopedic handbook of biomaterials and bioengineering. New York: Marcel Dekker; 1995. p. 141—219. 19. McKinney JE. Environmental damage and wear of dental composite restoratives. In: Vanherle G, Smith DC, editors. International symposium on posterior composite resin dental restorative materials. ; 1985. p. 331—47. 20. Nagarajan VS, Hockey BJ, Jahanmir S, Thompson VP. Wear mechanisms of a dental composite with high filler content. J Mater Sci 2000;35:487—96. 21. McKinney JE. Apparatus for measuring wear of dental restorative materials. Wear 1982;76:337—47. 22. Wu W, McKinney JE. Influence of chemicals on wear of dental composites. J Dent R 1982;61(10):1180—3. 23. Mueller HJ, Bapna MS, Knoeppel R. Human enamel-dental amalgam pin on disc wear. Dent Mater 1985;1:31—5. 24. Georges LA, Eichmiller FC. Wear of enamel against glassceramic, porcelain, and amalgam. In: Gosh A, Barks R, Rusin R, editors. Challenges Ceramic Product Development, Manufacture, and Commercialization, 66. Westerville, OH: American Ceramic Society; 1996. p. 181—8. 25. Nagarajan VS, Jahanmir S. Relationship between microstructure and wear of mica-containing glass-ceramics. Wear 1996;200:176—85. 26. Horlick G. Plasma-source mass spectrometry for elemental analysis. Spectroscopy 1992;7(1):22—9. 27. Marquis PM, Hu X, Shortall AC. Two-body wear of dental composites under different loads. Int J Prosthodont 2000;13: 473—9. 28. Seibelca J et al. Private communication 1997. 29. Suzuki S, Leinfelder KF. Localized wear and marginal integrity of posterior resin composites. Am J Dent 1993;6: 199—203. 30. Suh NP. Delamination theory of wear. Wear 1973;25:111—24. 31. Jahanmir S. On mechanics and mechanisms of laminar wear particle formation. In: Latanision RM, Fischer TE, editors. Advances in the Mechanics and Physics of Surface, 3. Chur, Switzerland: Harwood Academic Publishers; 1986. p. 261—332. 32. Shin MA, Drummondm JI, Botsis J. Atomic absorption studies on leached dental composite fillers. J Dent Res 1997;76(76):499. 33. Soderholm KJ, Mukherjy R. Filler leachability of composites stored in distilled water or artificial saliva. J Dent Res 1996; 75(9):1692—9. 34. Soderholm KJ, Lambrechts P, Sarrett D, Abe Y, Yang MC, Labella R, Yildiz E, Willems G. Clinical wear performance of eight experimental dental composites over three years determined by two measuring methods. Eur J Oral Sci 2001;109:273—81. 35. Tillitson EW, Craig RG, Peyton FA. Friction and wear of restorative materials. J Dent Res 1971;50(1):149—54. 36. Jahanmir S, Fischer TE. Friction and wear of silicon nitride lubricated by numid air, water, hexadecane and hexadecane þ 0.5 percent Stearic acid. STLE Transactions 1987; 31:32—43.