Dent Mater 8:158-161, May, 1992
Bond strength of composite to ,porcelain treated with new porcelain repair agents D.M. Wol~ J.M. Powers ~ K.L. O'Keefe 2 1Oral Biomaterials, 2Occlusion and Fixed Prosthodontics, University of Texas Health Science Center at Houston, Dental Branch, Houston, TX, USA
Abstract. In vitro tensile bond strengths of composite to porcelain were evaluated using three pretreatments (HF etching, sandblasting, diamond abrasion) of the porcelain, four bonding agents (Clearfil Porcelain Bond, Porcelain Liner M, Porcelain Liner M with Super-Bond C&B, and Scotchprime) and two storage conditions (24 h and thermocycling). The overall coefficient of variation was 27%. Significant differences among bond strengths were observed, with storage condition being the most important factor, followed by bonding agent and then pretreatment. Thermocycling decreased the bond strength of all samples, but samples treated with Scotchprime were affected least. For 24 h storage, Clearfil Porcelain Bond and Scotchprime had bond strengths above 23 MN/m 2 to sandblasted porcelain.
INTRODUCTION The increased use of porcelain as an esthetic dental restoration has resulted in the need for a reliable way to repair porcelain fractures. Previous investigations (Diaz-Arnold and Aquilino, 1989; Diaz-Arnold et al., 1989; Gregory et al., 1988; Matsumura et al., 1989; Prattet al., 1989; Stangel et al., 1987) have compared the bond strengths of several porcelain bonding systems which rely on a silane coupling reaction. Silane coupling agents possess the general chemical structure x-(CH2)3Si-(OR)3 and have the ability to bond chemically to both organic and inorganic surfaces. Mixed results, however, have been reported in laboratory studies with typical bond strengths reported between 1 and 15 MN/m2. Bond strength studies have been summarized recently (Council on Dental Materials, Instruments and Equipment, 1991). Recently, an experimental product which utilized a twoliquid primer (3-trimethoxy-silylpropylmethacrylate, methyl methacrylate/ferric chloride, ethanol), adhesive opaque resin (4-methacryloxyethyltrimellitate anhydride, methyl methacrylate with PMMA-coated titanium dioxide), and light-cured composite was described (Matsumura et al., 1989). This new system provided clinically acceptable in vitro values for shear strength of 28 MN/m2after 20,000 cycles. One manufacturer has marketed a porcelain repair system which incorporates a new catalyst to facilitate the silane coupling reaction. This product is claimed to produce adhesive bonds with clinically acceptable tensile bond strengths (18-24 MN/m 2) three times faster than traditional agents. The purpose of this investigation was to measure the in vitro bond strength of a composite bonded to porcelain with three porcelain bonding agents. Three pretreatment conditions and two storage conditions also were evaluated. 158 WolfetaL/Bondsb'engthofcompositetoporcelain
MATERIALS AND METHODS Codes, batch numbers and manufacturers of the bonding agents tested are listed in Table 1. Disks (1 cm in diameter and 2 mm thick) of porcelain (Ceramco II, LNOE519, Ceramco Inc., Burlington, NJ, USA) were prepared in a stainless steel die and fired in a vacuum oven (Jelenko Auto LTII VPF, Jelrus Technical Products Corp., New Hyde Park, NY, USA). The firing cycle was: preheat, 6 min; low set temperature, 649 °C; high set temperature, 974 °C; rate of temperature increase, 32-38 °C/ min. Samples then were screened for surface defects and imbedded in a potting resin. Finally they were ground on a metallurgical wheel sequentially with 240,320, 400, and 600 grit silicon carbide paper (Buehler Ltd., Lake Bluff, IL, USA). The final thickness of the porcelain disks was approximately 1.5 mm. Groups of samples were prepared with three surface pretreatments. The first group was etched with 9.5% hydrofluoric acid gel (Ceram-etch, Gresco Products, Inc., Stafford, TX, USA) for 5 min as recommended by the manufacturer. After etching, the samples were thoroughly rinsed for 30 s using an air/water spray to insure all the acid was removed from the surface. The second group was sandblasted with an abrasive tool (Model ERC, Micro-etcher, Danville Engineering, Danville, CA, USA) using 50 ~m A1203particles at a pressure of 0.48 MPa at 9 L/min for 3 s. Following sandblasting, the samples were cleansed for 60 s using 40% H3PO4 (3M Dental Products, St. Paul, MN, USA) and then rinsed for 30 s using an air/water spray. TABLE1: CODES,BATCHNUMBERSAND MANUFACTURERSOF BONDINGAGENTSTESTED. Batch Code Product Number Manufacturer CPB ClearfilPorcelainBond J. Morita Universal 827 Tustin, CA, USA Catalyst 724 Activator 021 PLM PorcelainLinerM Sun Medical LiquidA 71201 Kyoto,Japan Liquid B 71201 SP Scotchprime 3M DentalProducts Adhesive ODR St. Paul,MN, USA Primer OBR SB Super-BondC&B Sun Medical Powder 91201 Kyoto,Japan Liquid 00502 Catalyst 005012
TABLE 2: BOND STRENGTH (MN/m2) OF COMPOSITETO PORCELAIN TREATED WITH VARIOUSSURFACE PRETREATMENTSAND PRIMERS. Pretreatment CPB PLM PLM&SB SP Sandblast & H3PQ stored 24 h 23.4 (5.1)* 4.9 (1.7) 20.4 (6.0) 23.7 (2.1) Thermocycled 12.5 (2.4) 3.7 (1.0) 12.0 (1.8) 16.3 (6.4) HF Etch stored 24 h 20.0 (5.1) 12.8 (3.6) 16.9 (5.8) 17.2 (4.0) Thermocycled 12.5 (2.4) 8.5 (1.1) 11.0 (1.0) 11.0 (2.9) Diamond & H3PO4 stored 24 h 23.5 (3.6) 7.9 (2.3) 19.1 (5.1) 15.0 (2.7) Thermocycled 12.9 (4.6) 4.6 (1.0) 7.0 (1.1) 11.3 (5.1) *Mean of 5 replications and standard deviation. Tukey intervals (MN/m2) for comparisons among means at the 95% level of confidence were 2.5 among bonding agents, 2.0 among pretreatments, and 1.3 between storageconditions.
The third group was abraded using pear-shaped, coarse diamonds (Two Striper, 285.5C, Abrasive Technology, Inc. Westerville, OH, USA) at high speed followed by 60 s, 40% H3PO4 cleansing and 30 s air/water spray. After pretreatment, four bonding agents (CPB, PLM, PLM & SB, and SP) were applied to the samples following the manufacturers' recommendations. The composite (Herculite XR, #02072, Sybron/Kerr, Romulus, MI, USA) was lightcured on the samples in 1 mm increments in the shape of an inverted, truncated cone with a diameter of 3 mm at the bond interface and a diameter of 5 mm at a height of 5 ram. The composite samples were made using a bonding jig and polytetrafluoroethylene die to insure consistent sample preparation. The porcelain/composite samples were then stored in one of two conditions before testing: 37°C and 100% relative humidity for 24 h or thermocycled (temperature range, 8 to 52°C; dwell time, 30 s; and 1000 cycles). Samples were mounted in a loading jig and debonded in tension as described by Barakat and Powers (1986) using a testing machine (Model 8501, Instron Corp., Canton, MA, USA) at a cross-head speed of 0.05 cm/min. The force at which the bond failed was recorded, and the bond strength was calculated in MN/m 2. The site of bond failure was examined under low-power magnification and recorded. Representative fracture sites were studied using scanning electron microscopy. Five replications for each condition for a total of 120 samples were tested. Analysis of variance (Dalby, 1968) and comparisons of means by a Tukey interval (Guenther, 1964) calculated at the 95% level of confidence were performed. Differences between two means that were greater than the Tukey interval were considered statistically significant.
RESULTS Data for bond strength are listed in Table 2. Analysis of variance showed significant differences in bond strength among the means with storage condition being the most important factor, followed by bonding agent and then surface pretreatment. The overall coefficient of variation was 27%. Tukeyintervals for comparison among means at the 95% level of confidence were 2.5 among bonding agents, 2.0 among pretreatments, and 1.3 MN/m 2 between storage conditions. Figures 1 to 3 show scanning electron micrographs ofthe three
I0 ~m Fig. 1. Scanning electron micrograph (original magnification - 500X) of porcelain surface prepared for bonding by sandblasting,
10 ~m Fig. 2. Scanning electron micrograph (original magnification - 500X) of porcelain surface prepared for bonding by 9.5% hydrofluoric acid etching.
1 0 ~Lm Fig. 3. Scanning electron micrograph (original magnification - 500X) of porcelain surface prepared for bonding by diamond abrasion,
porcelain pretreatments. Thermocycling caused decreased bond strength for sandblasted samples, ranging from a 24% decrease for PLM samples to 46% for CPB samples. Overall, SP samples were least affected by the thermocycling conditions tested. The diamond abraded samples showed decreases in bond strength ranging from 25% for SP samples to 63% for PLM & SB
Dental Materials~May 1992 159
TABLE 3: BOND FAILURESOF COMPOSITETO PORCELAINTREATED
WITH VARIOUSSURFACE PRETREATMENTSAND BONDINGAGENTS. Pretreatment CPB Sandblast& H3PO4 stored 24 h P4, BI* Thermocycled P3, B2
PLM & SB
B3, P2 B5
B3, P2 B3, P2
HF Etch stored 24 h Thermocycled
P3, B2 P4, B1
P3, B2 B3, P2
P4, B1 P4, B1
P3, B2 P5
Diamond & H3PO4 stored 24 h P3, B2 B5 B5 B5 Thermocycled P4, B1 B4, P1 B5 B4, P1 *B is adhesive failure at bond interface. P is cohesive failure in porcelain.
samples. Samples which were etched with hydrofluoric acid gel showed the most consistent decreases in bond strength with a range from 34 to 38%. CPB samples had the highest bond strength, followed by SP, PLM & SB, and then PLM samples. The highest bond strengths were produced by sandblasted porcelain bonded with SP or CPB and by diamond abraded porcelain with CPB. CPB also had the highest bond strength to hydrofluoric acidetched porcelain. PLM samples consistently had the lowest bond strengths. Overall, sandblasting produced the highest bond strengths, followed by hydrofluoric acid-etching and then diamond abrasion. Data for bond failures are listed in Table 3. Two types of bond failure were observed, adhesive failure ofthe bond coded B in Table 3 and shown in Figure 4 and cohesive failure in the porcelain coded P in Table 3 and shown in Figure 5. Adhesive failures (B) were observed with SP (75%), PLM & SB (90%), and PLM (95%) with both sandblasting and diamond bur pretreatments. Cohesive failures of porcelain (P) were observed mostly with CPB (73%) and also in samples pretreated with hydrofluoric acid etching (70%). DISCUSSION This in vitro investigation showed significant differences among bond strength produced using different porcelain bonding agents. CPB, which is a modification of the traditional
1 0 pm Fig. 4. Scanning electron micrograph (original magnification - 1200X) of debonded porcelain surface showing an adhesive failure mode (coded B).
silane coupling system, produced the overall highest bond strengths but showed an increased incidence of cohesive failure of porcelain. Bertolotti et al. (1989) have also reported high bond strengths with CPB. SP, which bonded better to sandblasted porcelain, had a high incidence of adhesive failure and was affected least by thermocycling. Gregory et al. (1988) reported tensile bond strengths of SP to be 42 MN/m2after 24 h and 21 MN/m2after thermocycling. PLM showed low strengths when used byitself with Herculite XR. The manufacturer of PLM recommended the use of SB, an autopolymerized adhesive resin, with PLM to increase the bond strength. Lower bond strength is expected to result from thermocycling conditions because of differences in the coefficient of thermal expansion between porcelain and composite. This study showed all groups to have lower bond strength after thermocycling. Lei et al. (1989) also observed declines in bond strength after thermocycling but conversely reported that CPB had significantly higher bond strength after thermocycling than SP. The increased incidence of cohesive failure in samples pretreated with 9.5% hydrofluoric acid gel may be the result of deep acid penetration. Figure 5 shows a cohesive failure in a sample which was pretreated with the 9.5% hydrofluoricacid gel. The arrows point to voids along the fracture line which were produced by the acid. The manufacturer of this product suggests a 5 rain etch in their instructions. Data from this study suggest that 5 rain may be too long. The highest bond strengths were obtained in this study with sandblasting. Intraoral sandblasting is possible using a small, portable unit (Micro-etcher, Danville Engineering). Care must be taken to avoid abrasion of pre-existing restorations. Diamond abrasion followed by phosphoric acid treatment was the simplest of the clinical pretreatment methods evaluated in this study. Etching the porcelain with hydrofluoric acid was also effective. Alteration of the concentration and/or exposure time of the acid may make hydrofluoric acid etching a more viable alternative clinically. Historically, porcelain repair has been a temporary solution to a difficult problem. Perhaps with newer bonding systems that produce higher bond strengths, porcelain repair with composite resin will become a more long-term treatment modality. Although the present study considered bonding to
10 ~tm Fig. 5. Scanning electron micrograph (original magnification- 1200X) of debonded porcelain surface showing a cohesive failure mode (coded P). The arrow points to voids caused by the acid etching.
porcelain, repair of porcelain often involves composite, metal and tooth surfaces as well. The effects ofhydrofluoric acid and organosilane porcelain repair materials on composites, porcelain-fused-to-metal alloys and tooth structure require further investigation. In summary, 1) Significant differences among bond strengths were observed with storage condition being the most important factor, followed by bonding agent and then pretreatment. 2) For 24-h storage, Clearfil Porcelain Bond and Scotchprime had bond strengths above 23 MN/m2 to sandblasted porcelain. 3) Thermocycling decreased the bond strength ofall samples, but samples treated with Scotchprime were affected least. After thermocycling, Scotchprimebonded better to sandblasted porcelain than Clearfil Porcelain Bond, but they bonded the same to porcelain abraded with a diamond bur or etched. 4) For sandblasting and diamond abrasion, Clearfil Porcelain Bond samples had 70% cohesive porcelain failures, whereas Scotchprime samples had 70% adhesive failure.
ACKNOWLEDGMENTS The authors thank the manufacturers for supplying commercial products and D. Ladd and U. Parikh for technical assistance. The author (D.M. Wolf) thanks the UTHSCH, Dental Branch for a Student Research Scholarship. Received July 19, 1991/Accepted September 12, 1991 Address correspondence and reprint requests to: J.M. Powers Oral Biomaterials UTHSCH, Dental Branch P.O. Box 20068 Houston, TX 77225 USA
REFERENCES Barakat MM, Powers JM (1986). In vitro bond strength of cements to treated teeth. Aust Dent J 31:415-419. Bertolotti RL, Lacy AM, Watanabe LG (1989). Adhesive monomers for porcelain repair. Int J Prosthodont 2:483489. Council on Dental Materials, Instruments and Equipment (1991). Porcelain repair materials. J Am Dent Assoc 122:124-130. Diaz-Arnold AM, Aquilino SA (1989). An evaluation of the bond strengths of four organosilane materials in response to thermal stress. J Prosthet Dent 62:257-260. Diaz-Arnold AM, Schneider RL, Aquilino SA (1989). Bond strengths ofintraoral porcelain repair materials. JProsthet Dent 61:305-309. Dalby J. (programmer) (1968). BMD8V- Analysis ofVariance. Ann Arbor, MI: Statistical Research Laboratory, University of Michigan, 1-8. Gregory WA, Hagen CA, Powers JM (1988). Composite resin repair of porcelain using different bonding materials. Oper Dent 13:114-118. Guenther WC (1964). Analysis ofVariance, Englewood Cliffs, NJ: Prentice-Hall, 1-199. Lei RT, Mittlman G, Thompson VP (1989). Porcelain repair systems: Immediate and thermal cycled tensile bond strength. JDent Res 68:869, Abstr. No. 18. Matsumura H, Kawahara M, Tanaka T, Atsuta M (1989). A new porcelain repair system with a silane coupler, ferric chloride, and adhesive opaque resin. J Dent Res 68:813818. Pratt RC, Burgess MS, Schwartz JH, Smith MS (1989). Evaluation ofbond strength ofsix porcelain repair systems, J Prosthet Dent 61:305-309. Stangel I, Nathanson D, Hsu CS (1987). Shear strength ofthe composite bond to etched porcelain. J Dent Res 66:14601465.
Dental Materials~May 1992 161