| Sign In to gain access to subscriptions and/or personal tools. |
Long-term Water-aging of Whisker-reinforced Polymer-Matrix CompositesPaffenbarger Research Center, 100 Bureau Dr. Stop 8546, American Dental Association Health Foundation, Building 224, Room A-153, National Institute of Standards and Technology, Gaithersburg, MD 20899-8546; hockin.xu{at}nist.gov
Long-term water exposure may degrade polymer-matrix composites. This study investigated the water-aging of whisker composites. It was hypothesized that whiskers would provide stable and substantial reinforcement, and that whisker type would affect water-aging resistance. Silica-fused Si3N4 and SiC whiskers were incorporated into a resin. The specimens were tested by three-point flexure and nano-indentation vs. water-aging for 1 to 730 days. After 730 days, SiC composite had a strength (mean ± SD; n = 6) of 185 ± 33 MPa, similar to 146 ± 44 MPa for Si3N4 composite (p = 0.064); both were significantly higher than 67 ± 23 MPa for an inlay/onlay control (p < 0.001). Compared with 1 day, the strength of the SiC composite showed no decrease, while that of the Si3N4 composite decreased. The decrease was due to whisker weakening rather than to resin degradation or interface breakdown. Whisker composites also had higher moduli than the controls. In conclusion, silica-fused whiskers bonded to polymer matrix and resisted long-term water attack, resulting in much stronger composites than the controls after water-aging.
Key Words: long-term water-aging • whisker-silica fusion • polymer composite • filler-matrix interfaces • strength • modulus • reinforcement
Significant improvements have been achieved in the fillers, resins, filler-matrix bonding, and cure conditions for dental polymer-matrix composites. Heat-curing increased the degree of conversion and composite strength (Asmussen and Peutzfeldt, 1990; Wendt and Leinfelder, 1990; Ferracane et al., 1995; Loza-Herrero et al., 1998). Short fibers (Krause et al., 1989) and fused fibers (Bayne and Thompson, 1996) have been used to reinforce composites. However, brittle fracture and high failure rates in large stress-bearing restorations (Tyas, 1990; Sakaguchi et al., 1992; Christensen, 1999; Donly et al., 1999) called for further improvements. Recently, whiskers were used to reinforce dental composites (Xu et al., 1999), yielding composite strength and toughness values higher than those of current composites. Water exposure could degrade dental composites due to the degradation of filler particles (Söderholm, 1981; Söderholm et al., 1984; Calais and Söderholm, 1988), the weakening of polymer matrix (Calais and Söderholm, 1988; Ferracane et al., 1995, 1998), or the debonding of filler-matrix interfaces (Pilliar et al., 1986; Ferracane and Marker, 1992). Other studies showed little change for composites after water-aging (Lloyd, 1984; Drummond and Savers, 1993). No long-term water-aging studies have been performed on whisker composites. This study investigated the effects of water-aging on whisker composites, and examined the influence of whisker type and matrix resin, and the mechanisms of degradation. The water-aging time ranged from 1 day to 2 yrs. It was hypothesized that: (1) water-aging would decrease the composite properties due to the weakening of fillers, resin matrix, or interfaces; (2) whisker composites would possess significantly higher strength than the controls after water-aging; and (3) whisker type would significantly affect water-aging resistance.
Filler Treatment Silicon nitride whiskers (β-Si3N4, UBE, New York, NY, USA) were used with diameters ranging from 0.1 µm to 2 µm (mean = 0.4 µm) and lengths from 2 µm to 30 µm (mean = 5 µm). Silicon carbide whiskers (SiC, Advanced Refractory Technologies, Buffalo, NY, USA) that had diameters of 0.1 µm to 3 µm (mean = 0.7 µm) and lengths of 2 µm to 100 µm (mean = 14 µm) were also used. These whiskers were selected because of their small size and high tensile strength (about 50 GPa; Iwanaga and Kawai, 1998) compared with about 3 GPa of glass fibers. Each type of whisker was mixed with silica of 0.04 µm particle size (Aerosil OX50, Degussa, Ridgefield, NJ, USA) at a whisker:silica mass ratio of 2:1 (Xu et al., 1999). The mixed powder was heated in a furnace at 800°C for 30 min to fuse silica onto the whiskers (Xu et al., 2000). The powder was then silanized with mass fractions of 2% n-propylamine and 4% 3-methacryloxypropyltrimethoxysilane.
Specimen Fabrication Following the manufacturer’s instructions, we cured an indirect inlay/onlay composite (ConceptTM, Ivoclar, Amherst, NY, USA) in the Concept Heat Integrated Processor at 120°C for 10 min under a pressure of 0.6 MPa. ConceptTM, denoted as control C, consisted of 76% mass fraction of silicate fillers in a urethanedimethacrylate matrix. An indirect prosthetic composite (ArtglassTM, Heraeus Kulzer GmbH, Wehrheim, Germany) was cured in a Dentacolor XSTM photo-curing unit (Heraeus Kulzer GmbH) for 90 sec on each side of the specimen. ArtglassTM, denoted as control A, contained a mass fraction of 70% barium glass in a polymer matrix with tetra- and hexa-functional groups in addition to conventional bi-functional methacrylates.
Water-aging
Flexural and Nano-indentation Testing To examine if a thin surface layer of the water-aged specimens had degraded more than the strength of the bulk specimen, we used nano-indentation (Nano Instruments, Knoxville, TN, USA) to indent the specimen surfaces. The indentation loads and the corresponding displacements were recorded. The slope of the unloading curve provides a measure of the contact stiffness, which can be used with the contact area to determine the elastic modulus (Oliver and Pharr, 1992; Xu et al., 1998). Twenty-four indentations were made, with 4 indentations in each of 6 specimens, for each material after each water-aging period, yielding a total of 600 indentations. Pmax of 1 N was used to yield indentation contact areas large enough to represent the composite, rather than the individual fillers or the polymer matrix. Selected specimen fracture surfaces were gold-coated and examined with a scanning electron microscope (SEM, JSM-5300, JEOL, Peabody, MA, USA). We performed two-way ANOVA to detect significant difference. We used Tukey’s multiple-comparison test to compare the strength, modulus, and hardness data at a family confidence coefficient of 0.95.
Flexural strengths of the 5 materials vs. water-aging time are shown in two plots (Fig. 1  ), because the 3 materials in (A) had the same polymer composition. Two-way ANOVA for the 5 x 5 design showed significant effects (p < 0.001) of material type and aging time. There was a significant interaction between material type and aging time (p = 0.001). After 1 day, Si3N4 composite had a strength (mean ± SD; n = 6) of 248 ± 23 MPa, significantly higher than 202 ± 26 MPa of SiC composite, 122 ± 44 MPa of unfilled resin, 123 ± 21 MPa of control A, and 120 ± 16 MPa of control C (Tukey’s; family confidence coefficient = 0.95). However, the strength of Si3N4 composite and control C decreased significantly in water-aging (p < 0.001 for Si3N4, and p = 0.015 for control C); that of SiC composite, unfilled resin, and control A showed no significant decrease (p > 0.05). After 730 days, the strength of SiC composite was 185 ± 33 MPa, not significantly different (p = 0.064) from 146 ± 44 MPa of Si3N4 composite. Both of them were higher than 81 ± 23 MPa of unfilled resin and 67 ± 23 MPa of control C; the strength of SiC composite was higher than 108 ± 26 MPa of control A (p < 0.001).
For nano-indentation (Fig. 2 ), two-way ANOVA showed significant effects (p < 0.001) for material type and aging time. After 1 day, the moduli of 19.58 ± 0.69 GPa for Si3N4 composite and 20.10 ± 0.82 GPa for SiC composite were not significantly different; both were significantly higher than 15.82 ± 0.19 GPa for control A, 15.08 ± 0.30 GPa for control C, and 4.61 ± 0.03 GPa for unfilled resin (Tukey’s; family confidence coefficient = 0.95). The amounts of modulus and hardness decreases with increasing aging time were relatively small.
SEM micrographs of fracture surfaces (Fig. 3 ) are for (A) control A, (B) SiC composite, and (C) Si3N4 composite, after 1 day’s immersion. The controls had flat fracture surfaces. In contrast, the whisker composites had rough surfaces with fracture steps (large arrows) and whisker pullout (small arrows). The fracture surfaces of Si3N4 composites at aging times > 200 days had shorter whisker pullout lengths. Examples of this are shown in Fig. 4 for Si3N4 composite after (A) 1 day, (B) 400 days, and (C) 730 days, with shorter whisker pullout at 400 days and 730 days. Resin was observed on the pulled-out whiskers (arrows), indicating good whisker-resin bonding after 730 days of water-aging.
The SiC composite showed no significant strength loss after 730 days of water-aging, and the trend of strength variation mimicked that of the unfilled resin (Fig. 1A  ). This suggests that the strength variation of SiC composite was likely a result of changes in the physical and mechanical behavior of the polymer matrix. Softening of the polymer matrix could occur in water-aging by swelling the polymer network and reducing the frictional forces between the polymer chains (Ferracane et al., 1998). However, this factor, in the present study, appeared to cause only minor decreases in the matrix properties. The water uptake after 7 days’ immersion was measured (mean ± SD; n = 6) in µg/mm3 to be 18.7 ± 2.2 for unfilled resin, 14.0 ± 2.7 for SiC composite, 10.4 ± 1.1 for Si3N4 composite, 9.0 ± 2.4 for control C, and 5.2 ± 2.2 for control A. These values were significantly different from each other (p < 0.05), except for Si3N4 composite and control C, which were not significantly different (p > 0.10). The water uptake values for the whisker composites were within the range reported for different composites in a previous study (seven-day values in Table 3; Øysæd and Ruyter, 1986). The unfilled resin had the highest water uptake, but with no significant strength loss in water-aging (Fig. 1). SiC composite had higher water uptake than control A, but both had no significant strength loss in water-aging (Fig. 1). Hence it appears that water uptake had little effect on composite strength in this study. The Si3N4 composite suffered a significant strength loss in water-aging, with its strength after 730 days being 41% lower than that after 1 day. Most of the strength loss occurred by 200 days, consistent with a previous study observing significant property reduction at 6 months of water-aging, with minimal changes thereafter (Ferracane et al., 1998). Such degradation was attributed to the weakening of polymer matrix in previous studies (Ferracane et al., 1995, 1998). In the present study, however, the strength loss in Si3N4 composite did not appear to be due to a weakening of the polymer matrix, for two reasons. First, the unfilled resin, which was the same as the matrix in Si3N4 composite, did not show a significant strength loss. Second, the SiC composite, which had the same polymer matrix as the Si3N4 composite, did not show a significant strength loss.
Further, the strength loss in Si3N4 composite did not appear to be due to a weakening or breakdown of the filler-polymer matrix interfaces, for two reasons. First, the SiC composite, in which the whiskers had the same silica fusion and silanization as Si3N4, showed no significant strength loss. An interface breakdown would have degraded both Si3N4 and SiC composites. Second, SEM observations of the pulled-out Si3N4 whiskers (Fig. 4 The specimen surfaces did not degrade more than the bulk, since the losses in nanoindentation hardness and modulus in water-aging were slight. The inlay/onlay and prosthetic controls had moduli of 14–16 GPa. The whisker composites had higher moduli of 18–20 GPa, approaching that of dentin (20 GPa) (Xu et al., 1998). A low elastic modulus may cause the restoration to deform and bend excessively during mastication, producing maximum tensile stresses at the internal surface of the restoration, which may lead to failure initiation in vivo (Kelly et al., 1995). Therefore, the high moduli of the whisker composites may be as beneficial as their high strengths. An ongoing three-body wear study showed that, after 4 x 105 wear cycles, the Si3N4 whisker composite had a wear scar diameter (mean ± SD; n = 6) of 643 ± 39 µm and a wear depth of 82 ± 19 µm, significantly less than a wear scar diameter of 1184 ± 34 µm and a wear depth of 173 ± 15 µm of control A (Tukey’s; family confidence coefficient = 0.95). SEM examination revealed that, instead of whiskers protruding from the worn surface, the whiskers were worn with the composite surface, resulting in relatively smooth wear surfaces. The Si3N4 composite, although relatively opaque, had a whitish color similar to that of enamel, and the restorations appeared smooth and shiny after clinical polishing (Xu, 1999). The SiC composites were gray, similar to some commercial core buildup composites (e.g., Ti-CoreTM, Essential Dental Systems, S. Hackensack, NJ, USA). The significant increases in strength and modulus of whisker composites may help improve the performance of composites in large stress-bearing restorations. In summary, reinforcement using both types of whiskers resulted in significantly stronger composites compared with the controls. The SiC composite showed no strength loss in long-term water-aging, while the Si3N4 composite did. The experimental design, including the unfilled resin and two types of whiskers with the same matrix, helped determine the degradation mechanism for Si3N4 composites as being whisker-weakening, rather than degradation of the matrix or the interfaces. After 730 days, the whisker composites possessed strengths 1.4 to 2.8 times those of a prosthetic control and an inlay/onlay control. The elastic moduli of whisker composites were also higher than those of the controls. The results suggest the use of strong and stable fillers as a key microstructural parameter to improve the composite’s resistance to long-term water attack, which may be applicable to improvement in other dental and hard-tissue repair materials.
The author gratefully acknowledges Dr. Douglas T. Smith for help with nano-indentation, Anthony A. Giuseppetti for help with the Instron, Dr. Joseph M. Antonucci for providing the resin monomer, and Drs. Frederick C. Eichmiller and Gary E. Schumacher for discussions. This study was supported by NIDCR grant R29 DE12476, NIST, and the ADAHF.
DISCLAIMER Certain commercial materials and equipment are identified to specify the experimental procedure. This does not imply recommendation or endorsement by NIST or ADAHF or that the material or equipment identified is necessarily the best available for the purpose. Received for publication April 26, 2002. Revision received September 6, 2002. Accepted for publication October 7, 2002.
Journal of Dental Research, Vol. 82, No. 1,
48-52 (2003)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
