- Email: [email protected]
Factors influencing metal-resin tensile bond strength to filled composites
Factors influencing metal-resin tensile bond strength to filled composites H. Lfithy* C.P. Marinello P. Schiirer Department of Fixed and Removable Prosthodontics and Dental Materials Science, Dental Institute, University of 7erich, Plattenstrasse 11, CH-8028 Z(Jrich, Switzerland Received November 30, 1988 Accepted December 14, 1989 *Corresponding author Dent Mater 6:73-77, April, 1990 Abstract-This study evaluated the effect of metal surface conditioning, application of a silicon layer, water storage, and resin filling on tensile bond strength of a metal-resin system using three experimental composites (un-, micro-, and macrofilled) having the same selfcuring resin composed of Bis-GMA and TEGDMA (2:1 wt%). Test specimens were prepared by bonding the resin between pairs of Ni-Cr-Be alloy cast disks (diameter, 8 mm) previously subjected to heat treatments simulating porcelain firing procedures. A specially constructed apparatus facilitated the absolutely parallel alignment and orientation of the disk faces to each other, maintaining a constant resin thickness of 100 ~m. Before being bonded, the sand-blasted metal surfaces were either electrolytically etched and/or silicoated. Prior to being tested, assemblies were stored in water at 37°C for one and 30 days. Thereafter, the specimens were processed in a universal testing machine at a cross-head speed of 2 ram/rain until failure. Bond strengths ranged from 4.2 to 20.5 MPa. Data were analyzed by ANOVA with a factorial design (conf. level = 99%). The results showed that: (i) bond strength was increased when the metal was silicoated, (ii) the combination of sandblasting/silicoating produced the best values, and (iii) the 30-day water storage combined with silicoating enhanced the strength of the bond. The resin filling had no significant effect, indicating that neither its presence nor type affects bonding strengths to metal. m
~
t is accepted that fillers increase not only the mechanical properties of dental resins (Phillips, 1982; Craig et al., 1983; Asmussen, 1985) but also the bond strength to acidetched enamel, based essentially on mechanical retention. In this respect, the type of filler plays an important role for fixing resin-bonded bridges (Zidan et al., 1980; Boyer et al., 1982; Craig et al., 1983). Zidan et al. (1980) and Boyer et al. (1982) have shown that macrofflled composites had a higher bond strength to etched enamel than microfflled composites. They found a positive correlation between mechanical properties of the dental composites and their bonding strength to enamel. As far as the bond between metal and resin is concerned, its strength also relies on mechanical retention provided by electrolytical etching of the base metal (Thompson et al., 1981; Livaditis and Thompson, 1982). Recently, a new technique was introduced based on the chemical bonding of resin to alloy. This technique, called silicoating, was developed by Musit and Tiller (1984a) and involves applying a silicate layer (SiOx-C) on the metal by flamespraying. The aim of the present study was to evaluate the effects of metal surface conditioning (sand-blasting/ electrolytical etching), silicoating, water storage, and filler type on the metal-resin bond strength. The goal was to verify whether the acid-etching and/or the silicoating technique discriminated among three experimental composites (un-, micro-, and macrofflled) with the same resin matrix, as when acid-etched enamel is bonded to resin. MATERIALS AND METHODS Three composites (un-, micro-, and macrofflled) with the same resin matrix were tested using combinations of differently treated metal surfaces
(A = sand-blasted; B = electrolytically etched; C = non-silicoated; D = silicoated) subsequently stored in water for one and 30 days after being bonded. The combinations tested were AC, AD, BC, and BD (Table 1). In this study, 240 metal disks (diameter, 8 mm) of Ni-Cr-Be alloy (Rexillium III, Jeneric Gold Co., Wallingford, CT, USA), with a loop on one face (Fig. 1), were cast in an induction casting machine (Linn GmbH Elektronik Induktherm, HFSVac, Hirschbach, FRG) according to TABLE 1 EXPERIMENTAL PROCEDURE
Casting of Disks Porcelain Firing Simulation Sand-blasting
[
I
GroupAC I
t
I
I
Group BC GroupAD Group BD Etching Silicoating Etching I I Silicoating
I
Bonding
]
P
Microfilled Composite
Macrofilled Composite
r
f
Unfilled Composite
I
Water Storage
1 day
I
30 days
I Tensile testing
Dental Materials~April 1990 73
1
1 mm
2 3
i
J 10.5 mm
~
9
Fig. 2. Schematic appearance of the specially constructed apparatus used for bonding disks: (1) wing screw; (2) upper block; (3) locking screw; (4) vertical mobile piece; (5) upper specimen support; (6) threaded shaft; (7) guiding shaft; (8) lower block; and (9) lower specimen support. ¢~ 1.5
mm ::
::
O
3
::1
~,Smm
Fig. 1. Dimension of the disks showing the loop on one face.
the manufacturer's instructions. The loops facilitated attachment by means of hooks to the grips of a tensile testing machine. All castings were subjected to fore- heat-treatment cycles (960°C) simulating porcelain firing (Lfithy and Marinello, 1986), and then sand-blasted (50 ~m Al208 particles at 0.4 MPa). Two metal disks formed a test assembly. The 120 assemblies were scheduled for one (N = 60) and 30 (N = 60) days' storage in water, each further subdivided into three groups for the three composites tested, Ze., N = 20 per composite. Thus, five assemblies were tested for each condition (metal surface t r e a t m e n t , composite type, duration of water storage). Disks selected for electrolyrical etching were treated using 10 vol. % sulfuric acid at a current density of 300 mA/cm2 for three rain, then cleaned using 18% hydrochloride acid in an ultrasonic bath for 10 min (Livaditis and Thompson, 1982).
The specimens selected for silicoating were t r e a t e d by Kulzer's technique (Kulzer & Co. GmbH, F r i e d r i c h s d o r f , FRG) by being cleaned with e t h y l - a c e t a t e (Siliclean, Kulzer) and coated with a SiOx-C layer (Siliflam, Kulzer) for four min. The specimens were then cooled for two min, coated with a silane adhesive agent (Silicoup, Kulzer), and a light-cured opaque bonding resin (Dentacolor XS, Kulzer). The entire silicoatSng procedure was done according to the manufacturer's instructions, except that the disks were s a n d - b l a s t e d prior to being silicoated, with A1203 of mean particle size 50 ~m instead of 250 Ixm, for practical reasons. A specially constructed apparatus was used to bond two disks with identically treated surfaces (Fig. 2). Basically, the apparatus consists of u p p e r and lower stainless steel blocks. The upper block moves only vertically along two guiding shafts, and its position in relation to the lower block is further stabilized by the two wing screws fitted to the two threaded shafts. The upper block has five pieces which are mobile in the vertical direction. Each piece can be secured at an exactly pre-selected height by the tightening of its locking screw. The lower block has five pairs of specimen supports placed exactly below the five corresponding pairs attached to the mobile pieces
in the upper block. The disks were secured by silicone strips passed through the attached loops (Fig. 3). The configuration of the upper and lower specimen supports enables a pre-selected distance to be maintained between the disk surfaces, by placement of exact gauges between the disks. In this study, a distance of 100 ~m was selected. This ensured that the thickness of composite between the disks was exactly 100 ~m in each case. Excess composite was always placed in the space between the disks to compensate for the polymerization shrinkage, and the excess was removed after setting. The composites used (Ivoclar AG, Schaan, Liechtenstein) had the same self-curing resin composed of BisGMA and TEGDMA (2:1 wt%). The silane-treated fillers were silica particles (34 v%) with 40-nm diameter or glass particles (49 v%) with 2-~m diameter for the macro- and microfilled types, respectively. One h after h a r d e n i n g , the assemblies w e r e stored for either one or 30 days in water at 37°C before being tensiletested. Each assembly unit was attached by hooks to hydraulic parallel-acting grips of a universal testing machine (RM 50, Schenck-Trebel GmbH, Ratingen, FRG), ensuring that the alignment was in a perfectly straight vertical line. The hooks were of the same alloy as the disks, and all tests were done at a cross-head speed of 2 mm/min until failure. Data were analyzed by ANOVA with a factorial design 24, i.e., four factors (X1 to X4), each factor being taken at two levels. The factors were Xz-metal-surface conditioning (parameters A and B); X2-non- and silicoated surface (C and D); X.g-water storage (one and 30 days); and X 4 composite filling (either un- and microfilled, or un- and macrofilled, or micro- and macrofilled). RESULTS
The average tensile strength values for the three composites are graphically presented in Fig. 4 for all permutations. The statistical evaluation (conf. level 99 and 99.9%) shows (Table 2) that two factors and two interactions significantly influenced the tensile bond strength to composites. The two factors were the silicoating
74 LOTHY et a~/FACTORS INFLUENCING METAL-RESIN TENSILE BOND STRENGTH TO FILLED COMPOSITES
p r o c e d u r e and the 30-day w a t e r storage. The two interactions were the combination of silicoating/water storage and of metal-surface conditioning/silicoating (Figs. 5a and b). The best bonding results were obtained by the combination of parameters AD (sand-blasting/silicoating) after prolonged water storage (Fig. 4). Non-silicoated specimens, etched or sand-blasted, yielded to inferior bond strengths (AC and BC, Fig. 4). An additional important result was the total absence of the effect of filler on bonding s t r e n g t h . The same bonding s t r e n g t h was r e c o r d e d whether the resin was un-, micro, or macrofilled, provided that the metal surface conditioning prior to bonding was comparable (Fig. 4). In summary, the presence and type of filler did not contribute to the metal-resin bonding strength.
DISCUSSION A review of the literature (Musil and Tiller, 1984a, b; Barquins et al., 1985; Hero et al., 1987; Laufer et al., 1988; Jakob and Marx, 1988; Van der Veen, 1988) on adhesion of resin by the silicoating technique is controversial, probably depending on the metal chemistry, kind of resin, testing procedures, etc. However, t h e r e is agreement on one point: After dlT or one-day water storage, the bond strength to silicoated metal is significantly higher than that to other differently treated metal surfaces, e.g., sand-blasted or electrolytically etched. The present findings verified these results (Laufer et al., 1988). The 30-day water storage affected the bonding strength (Fig. 4), while water storage with silicoating reinforced the bond strength. Fig. 5a illustrates this phenomenon for a sandblasted surface. The slope of the curve after silicoating is about three times greater than that without silicoating. Musil and Tiller (1984a, b), Laufer et al. (1988), and Jakob and Marx (1988) all found a significant increase or at least no weakening of the bond strength, even after a prolonged storage time in water. The present results confirmed this trend. According to Jakob and Marx (1988), the silicate layer does not swell due to the humidity and forms a waterresistant bond with the adhesive.
The above observations are not in accordance with those of Hero et al. (1987), who registered a 30% reduction in bond strength to Ag-Pd alloys after 90 days. It is important to specify that in the present study no experiment was conducted to determine whether the centers of the test specimens were in equilibrium for water sorption after 30 days' storage. Given a 4-mm diffusion distance
.
(edge to center) and a 30-day water storage at 37°C, a calculation for a resin of the same composition gives a saturation of about 80% (Kalachandra and Turner, 1987). Further longterm investigations are needed. As for the rest, a significant interaction was found between metal surface conditioning (sand-blasting/ etching) and silicoating (Table 2 and Fig. 5b). It indicates that the sand-
Fig. 3, Details of the silicone strips placed through the loops securing both disks in position. Water storage: 20 -
15-~
30 days
1 day
E I •
unfilled microfilled macrofilled
Q. ° et-
m
10 ~ tO
ID I--
5 ~
0 J.
AC
BC
BD
AD
I11Ii AC
BC
BD
AD
Combination of parameters
Fig, 4. Mean tensile bond strengths (--. SD) of three composites (un-, micro-, and macrofilled) to differently treated metal surfaces after one- and 30-day water storage. A = sand-blasting; B = electrolytical etching; C = no sllicoating; and D = with silicoating.
Dental Materials~April 1990 75
TABLE 2 FACTORS SIGNIFICANTLY INFLUENCING METAL-RESIN BOND STRENGTH (CONF. LEVEL = 99 and 99.9%) Factors Main effects Silicoating 30-day water storage
F Values Fb
Fa 207.4*** 42.3**
396.6*** 96.3***
Fc 578.9*** 108.7"**
Interactions 20.9** 36.0** 52.5*** Silicoating/water storage 29.0"* 64.4*** 46.0** Metal surface conditioning/silicoating Fa Fb Fo are F values calculated with the composite filling X, for un-/microfilled, un-/macrofilled, and micro-/macrofilled, respectively. Conf. level F 1,5.0.99 = 16.3 F 1,s.0.999 = 47.2 **Significant at 99.0% level. ***Significant at 99.9% level.
blasted, gave the best adhesion results. They attributed these good results to the surface chemistry of the metal, which was more modified by the selective anodic etching than by the sand-blasting. Consequently, the oxides formed were different, leading to different bonding modes with the silicate layer. Such an hypothesis can also be applied to the present findings, namely, that the surface chemistry of the Ni-Cr-Be alloy used was differently changed by the twosurface conditioning processes. However, in this case, the oxides formed after sand-blasting were more favorable for the chemical bond to the SiOx-C layer than those after etching. The final and important point in
blasted surface reacted more to silicoating than did the electrolytically etched surface. Fig. 5b illustrates this behavior after water storage of one day. Average tensile bond s t r e n g t h tripled (from 4.5 MPa to 14 MPa) by silicoating after sand-blasting, while it increased to a lesser extent by silicoating after etching (from 9 MPa to 11 MPa). The present results confn~ned those of the inventors (Musil and Tiller, 1984a, b) of the silicoating technique, who recommend sandblasting prior to silicoating. However, the results are not in agreement with those of Barquins et al. (1985). The latter showed that a sillcoated Co-Cr ~lloy, previously electrolytically etched rather than sandMetal surface: sandblasted
Water storage: 1 day
20-
microfilled • ~
macrofilled •
16
with s,i ati t-
cO
12 8
CONCLUSIONS This study has shown that the metalresin tensile bond strength of three composites (un-, micro-, and macrofilled) with the same experimental resin matrix was positively and significantly influenced by: othe silicoating, ethe water storage (30 days), othe combination of sand-blasting and silicoating, and othe combination of silicoating and water storage. The difference between the composites was not statistically significant, indicating that the presence and type of filler had no effect on the bonding strength to metal.
unfilled •
: n~
the present study is that the same metal-surface processing did not discriminate between (i) unfilled and filled resins and (ii) micro- and macrofilled resins, with the same resin matrix (Fig. 4). The presence of filler did not increase the bond strength to m e t a l , a l t h o u g h t h e tensile strengths of the composites alone are ranked from unfilled to microfilled to macrofilled types (Phillips, 1982; Craig et al., 1983; Asmussen, 1985). Consequently, there is no correlation between the mechanical properties of the composites used alone and their bond strength to metal. This behavior differs basically from the bonding behavior b e t w e e n acidetched enamel and dental composite, where the presence and type of filler have a significant influence (Zidan et al., 1980; Boyer et al., 1982; Craig et al., 1983).
•
with silicoating
10-
ACKNOWLEDGMENTS
The authors wish to thank E. Egli, M. Morandini, P. Elmiger, and O. Loeffel for their laboratory assistance. This paper was presented in part at the 66th General Session of the IADR, Montreal, Canada, March, 1988 (Abstract No. 884).
(D t-
~
4
0
I
I
1
30
Water storage: days
0
I
I
sandblasted
Metal surface conditioning
Fig. 5. Influence on tensile bond strength of: (a) combination of silicnating and water storage, and (b) combination of metal surface conditioning and silicoating.
76
etched
REFERENCES ASMUSSEN, E. (1985): Clinical Relevance of Physical, Chemical and Bonding Properties of Composite Resins, Oper Dent 10:61-73. BARQUINS, M.; OUDARD,B.; DEGRANGE, M.; and ROQUES-CARMES,C. (1985): Application de la M~canique de la Rupture ~ la Tenue de Joints Coll~s, J Biomat Dent 1:309-320.
LUTHY et aL/FACTORS INFLUENCING METAL-RESIN TENSILE BOND STRENGTH TO FILLED COMPOSITES
BOYER, D.; CHALKLEY,Y.; and CHAN, K.C. (1982): Correlation between Strength of Bonding to Enamel and Mechanical Properties of Dental Composites, J Biomed Mat Res 16:775-783. CRAIG, R.; O'BRIEN, W.J.; and POWERS, J.M. (1983): Dental Materials, 3rd ed., St. Louis: Mosby Company, pp. 70-71. HER0, H.; RUYTER, L.E.; WAARLI,M.I.; and HULTQUIST, G. (1987): Adhesion of Resins to Ag-Pd Alloys by Means of the Silicoating Technique, J Dent Res 66:1380-1385. JAI(OS, E. and MARX,R. (1988): DaN Silicoaterverfahren ffir die Klebebrficke, Dtsch Zahn(~rztl Z 43:461-464. KALACHANDRA, S. and TURNER, D.T. (1987): Water Sorption of Polymethacrylate Networks: Bis-GM_A/rEGDM Copolymers, J Biomed Mater Res 21:329-338.
LAUFER, B.Z.; NICHOLLS, J.I.; and TOWNSEND, J.D. (1988): SiOx-C Coating: A Composite-to-Metal Bonding Mechanism,J Prosthet Dent 60:320327. LIVADITIS, G.J. and THOMPSON, V.P. (1982): Etched Castings: an Improved Retentive Mechanism for Resin Bonded Retainers, J Prosthet Dent 47:52-58. LOTHY, H. and M-ARINELLO,C.P. (1986): Durch W~irmebehandlung (Porzellanbrandsimulation) bedingte Ver~inderung elektrochemisch ge~itzter, aus einer Ni-Cr-Be-Aufbrennlegierung (Rexillium III) hergestellter Adh~isivbrfickenhalteelemente, Schweiz Mschr Zahnmed 96:1217-1224. MUSIL, R. and TILLER, J. (1984a): DaN Kulzer-Silicoater Verfahren. Wehrheim, FRG: Kulzer & Co. GmbH. MUSIL, R. and TILLER, J. (1984b): Die
molekulare Kopplung der Kunststoffverblendung an die Legierungsoberfi~iche,Dent Labor 32:1155-1161. PHILLIPS, R.W. (1982): Skinner's Science of Dental Materials, 8th ed, Philadelphia: Saunders Company, p. 219. THOMPSON, V.P.; LIVADITIS, G.S.; and DEL-CASTILLO,E. (1981): Resin Bond to Electrolytically Etched Non-precious Alloys for Resin Bonded Prostheses, J Dent Res 60:377, Abstr. No. 265. VAN DER VEEN, H. (1988): Resin-bonded Bridges, Thesis, University of Groningen, The Netherlands. ZIDAN, 0.; ASMUSSEN, E.; and J@RGENSEN, K.D. (1980): Correlation between Tensile and Bond Strength of Composite Resins, Scand J Dent Res 88:348-351.
Dental Materials~April 1990 77
~
t is accepted that fillers increase not only the mechanical properties of dental resins (Phillips, 1982; Craig et al., 1983; Asmussen, 1985) but also the bond strength to acidetched enamel, based essentially on mechanical retention. In this respect, the type of filler plays an important role for fixing resin-bonded bridges (Zidan et al., 1980; Boyer et al., 1982; Craig et al., 1983). Zidan et al. (1980) and Boyer et al. (1982) have shown that macrofflled composites had a higher bond strength to etched enamel than microfflled composites. They found a positive correlation between mechanical properties of the dental composites and their bonding strength to enamel. As far as the bond between metal and resin is concerned, its strength also relies on mechanical retention provided by electrolytical etching of the base metal (Thompson et al., 1981; Livaditis and Thompson, 1982). Recently, a new technique was introduced based on the chemical bonding of resin to alloy. This technique, called silicoating, was developed by Musit and Tiller (1984a) and involves applying a silicate layer (SiOx-C) on the metal by flamespraying. The aim of the present study was to evaluate the effects of metal surface conditioning (sand-blasting/ electrolytical etching), silicoating, water storage, and filler type on the metal-resin bond strength. The goal was to verify whether the acid-etching and/or the silicoating technique discriminated among three experimental composites (un-, micro-, and macrofflled) with the same resin matrix, as when acid-etched enamel is bonded to resin. MATERIALS AND METHODS Three composites (un-, micro-, and macrofflled) with the same resin matrix were tested using combinations of differently treated metal surfaces
(A = sand-blasted; B = electrolytically etched; C = non-silicoated; D = silicoated) subsequently stored in water for one and 30 days after being bonded. The combinations tested were AC, AD, BC, and BD (Table 1). In this study, 240 metal disks (diameter, 8 mm) of Ni-Cr-Be alloy (Rexillium III, Jeneric Gold Co., Wallingford, CT, USA), with a loop on one face (Fig. 1), were cast in an induction casting machine (Linn GmbH Elektronik Induktherm, HFSVac, Hirschbach, FRG) according to TABLE 1 EXPERIMENTAL PROCEDURE
Casting of Disks Porcelain Firing Simulation Sand-blasting
[
I
GroupAC I
t
I
I
Group BC GroupAD Group BD Etching Silicoating Etching I I Silicoating
I
Bonding
]
P
Microfilled Composite
Macrofilled Composite
r
f
Unfilled Composite
I
Water Storage
1 day
I
30 days
I Tensile testing
Dental Materials~April 1990 73
1
1 mm
2 3
i
J 10.5 mm
~
9
Fig. 2. Schematic appearance of the specially constructed apparatus used for bonding disks: (1) wing screw; (2) upper block; (3) locking screw; (4) vertical mobile piece; (5) upper specimen support; (6) threaded shaft; (7) guiding shaft; (8) lower block; and (9) lower specimen support. ¢~ 1.5
mm ::
::
O
3
::1
~,Smm
Fig. 1. Dimension of the disks showing the loop on one face.
the manufacturer's instructions. The loops facilitated attachment by means of hooks to the grips of a tensile testing machine. All castings were subjected to fore- heat-treatment cycles (960°C) simulating porcelain firing (Lfithy and Marinello, 1986), and then sand-blasted (50 ~m Al208 particles at 0.4 MPa). Two metal disks formed a test assembly. The 120 assemblies were scheduled for one (N = 60) and 30 (N = 60) days' storage in water, each further subdivided into three groups for the three composites tested, Ze., N = 20 per composite. Thus, five assemblies were tested for each condition (metal surface t r e a t m e n t , composite type, duration of water storage). Disks selected for electrolyrical etching were treated using 10 vol. % sulfuric acid at a current density of 300 mA/cm2 for three rain, then cleaned using 18% hydrochloride acid in an ultrasonic bath for 10 min (Livaditis and Thompson, 1982).
The specimens selected for silicoating were t r e a t e d by Kulzer's technique (Kulzer & Co. GmbH, F r i e d r i c h s d o r f , FRG) by being cleaned with e t h y l - a c e t a t e (Siliclean, Kulzer) and coated with a SiOx-C layer (Siliflam, Kulzer) for four min. The specimens were then cooled for two min, coated with a silane adhesive agent (Silicoup, Kulzer), and a light-cured opaque bonding resin (Dentacolor XS, Kulzer). The entire silicoatSng procedure was done according to the manufacturer's instructions, except that the disks were s a n d - b l a s t e d prior to being silicoated, with A1203 of mean particle size 50 ~m instead of 250 Ixm, for practical reasons. A specially constructed apparatus was used to bond two disks with identically treated surfaces (Fig. 2). Basically, the apparatus consists of u p p e r and lower stainless steel blocks. The upper block moves only vertically along two guiding shafts, and its position in relation to the lower block is further stabilized by the two wing screws fitted to the two threaded shafts. The upper block has five pieces which are mobile in the vertical direction. Each piece can be secured at an exactly pre-selected height by the tightening of its locking screw. The lower block has five pairs of specimen supports placed exactly below the five corresponding pairs attached to the mobile pieces
in the upper block. The disks were secured by silicone strips passed through the attached loops (Fig. 3). The configuration of the upper and lower specimen supports enables a pre-selected distance to be maintained between the disk surfaces, by placement of exact gauges between the disks. In this study, a distance of 100 ~m was selected. This ensured that the thickness of composite between the disks was exactly 100 ~m in each case. Excess composite was always placed in the space between the disks to compensate for the polymerization shrinkage, and the excess was removed after setting. The composites used (Ivoclar AG, Schaan, Liechtenstein) had the same self-curing resin composed of BisGMA and TEGDMA (2:1 wt%). The silane-treated fillers were silica particles (34 v%) with 40-nm diameter or glass particles (49 v%) with 2-~m diameter for the macro- and microfilled types, respectively. One h after h a r d e n i n g , the assemblies w e r e stored for either one or 30 days in water at 37°C before being tensiletested. Each assembly unit was attached by hooks to hydraulic parallel-acting grips of a universal testing machine (RM 50, Schenck-Trebel GmbH, Ratingen, FRG), ensuring that the alignment was in a perfectly straight vertical line. The hooks were of the same alloy as the disks, and all tests were done at a cross-head speed of 2 mm/min until failure. Data were analyzed by ANOVA with a factorial design 24, i.e., four factors (X1 to X4), each factor being taken at two levels. The factors were Xz-metal-surface conditioning (parameters A and B); X2-non- and silicoated surface (C and D); X.g-water storage (one and 30 days); and X 4 composite filling (either un- and microfilled, or un- and macrofilled, or micro- and macrofilled). RESULTS
The average tensile strength values for the three composites are graphically presented in Fig. 4 for all permutations. The statistical evaluation (conf. level 99 and 99.9%) shows (Table 2) that two factors and two interactions significantly influenced the tensile bond strength to composites. The two factors were the silicoating
74 LOTHY et a~/FACTORS INFLUENCING METAL-RESIN TENSILE BOND STRENGTH TO FILLED COMPOSITES
p r o c e d u r e and the 30-day w a t e r storage. The two interactions were the combination of silicoating/water storage and of metal-surface conditioning/silicoating (Figs. 5a and b). The best bonding results were obtained by the combination of parameters AD (sand-blasting/silicoating) after prolonged water storage (Fig. 4). Non-silicoated specimens, etched or sand-blasted, yielded to inferior bond strengths (AC and BC, Fig. 4). An additional important result was the total absence of the effect of filler on bonding s t r e n g t h . The same bonding s t r e n g t h was r e c o r d e d whether the resin was un-, micro, or macrofilled, provided that the metal surface conditioning prior to bonding was comparable (Fig. 4). In summary, the presence and type of filler did not contribute to the metal-resin bonding strength.
DISCUSSION A review of the literature (Musil and Tiller, 1984a, b; Barquins et al., 1985; Hero et al., 1987; Laufer et al., 1988; Jakob and Marx, 1988; Van der Veen, 1988) on adhesion of resin by the silicoating technique is controversial, probably depending on the metal chemistry, kind of resin, testing procedures, etc. However, t h e r e is agreement on one point: After dlT or one-day water storage, the bond strength to silicoated metal is significantly higher than that to other differently treated metal surfaces, e.g., sand-blasted or electrolytically etched. The present findings verified these results (Laufer et al., 1988). The 30-day water storage affected the bonding strength (Fig. 4), while water storage with silicoating reinforced the bond strength. Fig. 5a illustrates this phenomenon for a sandblasted surface. The slope of the curve after silicoating is about three times greater than that without silicoating. Musil and Tiller (1984a, b), Laufer et al. (1988), and Jakob and Marx (1988) all found a significant increase or at least no weakening of the bond strength, even after a prolonged storage time in water. The present results confirmed this trend. According to Jakob and Marx (1988), the silicate layer does not swell due to the humidity and forms a waterresistant bond with the adhesive.
The above observations are not in accordance with those of Hero et al. (1987), who registered a 30% reduction in bond strength to Ag-Pd alloys after 90 days. It is important to specify that in the present study no experiment was conducted to determine whether the centers of the test specimens were in equilibrium for water sorption after 30 days' storage. Given a 4-mm diffusion distance
.
(edge to center) and a 30-day water storage at 37°C, a calculation for a resin of the same composition gives a saturation of about 80% (Kalachandra and Turner, 1987). Further longterm investigations are needed. As for the rest, a significant interaction was found between metal surface conditioning (sand-blasting/ etching) and silicoating (Table 2 and Fig. 5b). It indicates that the sand-
Fig. 3, Details of the silicone strips placed through the loops securing both disks in position. Water storage: 20 -
15-~
30 days
1 day
E I •
unfilled microfilled macrofilled
Q. ° et-
m
10 ~ tO
ID I--
5 ~
0 J.
AC
BC
BD
AD
I11Ii AC
BC
BD
AD
Combination of parameters
Fig, 4. Mean tensile bond strengths (--. SD) of three composites (un-, micro-, and macrofilled) to differently treated metal surfaces after one- and 30-day water storage. A = sand-blasting; B = electrolytical etching; C = no sllicoating; and D = with silicoating.
Dental Materials~April 1990 75
TABLE 2 FACTORS SIGNIFICANTLY INFLUENCING METAL-RESIN BOND STRENGTH (CONF. LEVEL = 99 and 99.9%) Factors Main effects Silicoating 30-day water storage
F Values Fb
Fa 207.4*** 42.3**
396.6*** 96.3***
Fc 578.9*** 108.7"**
Interactions 20.9** 36.0** 52.5*** Silicoating/water storage 29.0"* 64.4*** 46.0** Metal surface conditioning/silicoating Fa Fb Fo are F values calculated with the composite filling X, for un-/microfilled, un-/macrofilled, and micro-/macrofilled, respectively. Conf. level F 1,5.0.99 = 16.3 F 1,s.0.999 = 47.2 **Significant at 99.0% level. ***Significant at 99.9% level.
blasted, gave the best adhesion results. They attributed these good results to the surface chemistry of the metal, which was more modified by the selective anodic etching than by the sand-blasting. Consequently, the oxides formed were different, leading to different bonding modes with the silicate layer. Such an hypothesis can also be applied to the present findings, namely, that the surface chemistry of the Ni-Cr-Be alloy used was differently changed by the twosurface conditioning processes. However, in this case, the oxides formed after sand-blasting were more favorable for the chemical bond to the SiOx-C layer than those after etching. The final and important point in
blasted surface reacted more to silicoating than did the electrolytically etched surface. Fig. 5b illustrates this behavior after water storage of one day. Average tensile bond s t r e n g t h tripled (from 4.5 MPa to 14 MPa) by silicoating after sand-blasting, while it increased to a lesser extent by silicoating after etching (from 9 MPa to 11 MPa). The present results confn~ned those of the inventors (Musil and Tiller, 1984a, b) of the silicoating technique, who recommend sandblasting prior to silicoating. However, the results are not in agreement with those of Barquins et al. (1985). The latter showed that a sillcoated Co-Cr ~lloy, previously electrolytically etched rather than sandMetal surface: sandblasted
Water storage: 1 day
20-
microfilled • ~
macrofilled •
16
with s,i ati t-
cO
12 8
CONCLUSIONS This study has shown that the metalresin tensile bond strength of three composites (un-, micro-, and macrofilled) with the same experimental resin matrix was positively and significantly influenced by: othe silicoating, ethe water storage (30 days), othe combination of sand-blasting and silicoating, and othe combination of silicoating and water storage. The difference between the composites was not statistically significant, indicating that the presence and type of filler had no effect on the bonding strength to metal.
unfilled •
: n~
the present study is that the same metal-surface processing did not discriminate between (i) unfilled and filled resins and (ii) micro- and macrofilled resins, with the same resin matrix (Fig. 4). The presence of filler did not increase the bond strength to m e t a l , a l t h o u g h t h e tensile strengths of the composites alone are ranked from unfilled to microfilled to macrofilled types (Phillips, 1982; Craig et al., 1983; Asmussen, 1985). Consequently, there is no correlation between the mechanical properties of the composites used alone and their bond strength to metal. This behavior differs basically from the bonding behavior b e t w e e n acidetched enamel and dental composite, where the presence and type of filler have a significant influence (Zidan et al., 1980; Boyer et al., 1982; Craig et al., 1983).
•
with silicoating
10-
ACKNOWLEDGMENTS
The authors wish to thank E. Egli, M. Morandini, P. Elmiger, and O. Loeffel for their laboratory assistance. This paper was presented in part at the 66th General Session of the IADR, Montreal, Canada, March, 1988 (Abstract No. 884).
(D t-
~
4
0
I
I
1
30
Water storage: days
0
I
I
sandblasted
Metal surface conditioning
Fig. 5. Influence on tensile bond strength of: (a) combination of silicnating and water storage, and (b) combination of metal surface conditioning and silicoating.
76
etched
REFERENCES ASMUSSEN, E. (1985): Clinical Relevance of Physical, Chemical and Bonding Properties of Composite Resins, Oper Dent 10:61-73. BARQUINS, M.; OUDARD,B.; DEGRANGE, M.; and ROQUES-CARMES,C. (1985): Application de la M~canique de la Rupture ~ la Tenue de Joints Coll~s, J Biomat Dent 1:309-320.
LUTHY et aL/FACTORS INFLUENCING METAL-RESIN TENSILE BOND STRENGTH TO FILLED COMPOSITES
BOYER, D.; CHALKLEY,Y.; and CHAN, K.C. (1982): Correlation between Strength of Bonding to Enamel and Mechanical Properties of Dental Composites, J Biomed Mat Res 16:775-783. CRAIG, R.; O'BRIEN, W.J.; and POWERS, J.M. (1983): Dental Materials, 3rd ed., St. Louis: Mosby Company, pp. 70-71. HER0, H.; RUYTER, L.E.; WAARLI,M.I.; and HULTQUIST, G. (1987): Adhesion of Resins to Ag-Pd Alloys by Means of the Silicoating Technique, J Dent Res 66:1380-1385. JAI(OS, E. and MARX,R. (1988): DaN Silicoaterverfahren ffir die Klebebrficke, Dtsch Zahn(~rztl Z 43:461-464. KALACHANDRA, S. and TURNER, D.T. (1987): Water Sorption of Polymethacrylate Networks: Bis-GM_A/rEGDM Copolymers, J Biomed Mater Res 21:329-338.
LAUFER, B.Z.; NICHOLLS, J.I.; and TOWNSEND, J.D. (1988): SiOx-C Coating: A Composite-to-Metal Bonding Mechanism,J Prosthet Dent 60:320327. LIVADITIS, G.J. and THOMPSON, V.P. (1982): Etched Castings: an Improved Retentive Mechanism for Resin Bonded Retainers, J Prosthet Dent 47:52-58. LOTHY, H. and M-ARINELLO,C.P. (1986): Durch W~irmebehandlung (Porzellanbrandsimulation) bedingte Ver~inderung elektrochemisch ge~itzter, aus einer Ni-Cr-Be-Aufbrennlegierung (Rexillium III) hergestellter Adh~isivbrfickenhalteelemente, Schweiz Mschr Zahnmed 96:1217-1224. MUSIL, R. and TILLER, J. (1984a): DaN Kulzer-Silicoater Verfahren. Wehrheim, FRG: Kulzer & Co. GmbH. MUSIL, R. and TILLER, J. (1984b): Die
molekulare Kopplung der Kunststoffverblendung an die Legierungsoberfi~iche,Dent Labor 32:1155-1161. PHILLIPS, R.W. (1982): Skinner's Science of Dental Materials, 8th ed, Philadelphia: Saunders Company, p. 219. THOMPSON, V.P.; LIVADITIS, G.S.; and DEL-CASTILLO,E. (1981): Resin Bond to Electrolytically Etched Non-precious Alloys for Resin Bonded Prostheses, J Dent Res 60:377, Abstr. No. 265. VAN DER VEEN, H. (1988): Resin-bonded Bridges, Thesis, University of Groningen, The Netherlands. ZIDAN, 0.; ASMUSSEN, E.; and J@RGENSEN, K.D. (1980): Correlation between Tensile and Bond Strength of Composite Resins, Scand J Dent Res 88:348-351.
Dental Materials~April 1990 77