Early Bond Strength of Luting Cements to a Precious Alloy P. MOJON3, E.B. HAWBOLT1, M.I. MacENTEE, and P.H. MA2 Department of Clinical Dental Sciences, University ofBritish Columbia, 2199 Wesbrook Mall, Vancouver, BC V6T 1Z3, Canada; 'Metals and Materials Engineering Department, University ofBritish Columbia, 6530 Stores Road, Vancouver, BC V6T 1W5, Canada; and 2Department of Statistics, University of British Columbia, 2021 West Mall, Vancouver, BC V6T 1Z2, Canada Previous studies have reported that glass-ionomer and adhesive resin cements can bond to various alloys, while zinc phosphate cements lack this adhesive property. This study evaluated the bonding properties of three luting cements during the first seven days after cementation. Thirty cylinders were cast with a highnoble porcelain-fused-to-metal (PFM) alloy and luted in pairs with one ofthe cements. The joints were stored in water at 370C for one, two, or seven days before being fractured in shear. The cylinders were re-used to provide 40 joints within each test group. The data were subjected to a Weibull analysis, a curve-fitting method shown to be appropriate for comparing the bond strengths of dental materials. The results showed that the zinc phosphate cement was the weakest material, whereas the adhesive resin produced the strongest joints. Microscopic observations ofthe fractured samples did not reveal any specific differences between the samples in terms oftheir mechanism of fracture. The glass-ionomer cement reached its maximum bond strength after two days, whereas storage time had no influence on the zinc phosphate cement. The adhesive resin cement was slightly, but not significantly, weaker after one week in water. We suggest that excessive loading of restorations cemented with glass ionomer should be avoided for the first two days after the placement. The use of an adhesive resin cement can be recommended on endodontically treated teeth, but further studies are needed to evaluate its biocompatibility and adhesion to dentin.
J Dent Res 71(9):1633-1639, September, 1992
Introduction. Zinc phosphate cement has been used successfully for several decades for outing artificial crowns on teeth, but it does not bond securely to dental alloys (Phillips, 1973; Moser et al., 1974; Dilts et al., 1985). Glass-ionomer cement, in contrast, will bond to base alloys and, to some extent, to precious alloys (McLean, 1977; Button et al., 1985; Krabbendam et al., 1987). However, the glass-ionomer cement remains weak and sensitive to dissolution in the early stage of a complex setting reaction (Crisp and Wilson, 1974; Crisp et al., 1976; McLean and Wilson, 1977; Mount and MAkinson, 1982), and it is unknown how this maturation process affects the bond strength of the material. Adhesive resins have been developed recently for luting artificial crowns on teeth (Takeyama et al., 1978; Nakabayashi et al., 1982; Yamashita, 1983). They provide a stronger bond to base metals than do glass-ionomer cements (Tanaka et al., 1981; Krabbendam et al., 1987; Yu and Xu, 1987), and they will adhere also to precious alloys when the metal surface is treated to enhance the bond (McLean, 1977; Laufer et al., 1988; Tanaka et al., 1988a,b; Watanabe et al., 1988). It is not clear, however, whether these prosthodontic adhesives will bond to untreated precious metal Received for publication October 21, 1991 Accepted for publication April 6, 1992 3Present address for correspondence: Ecole de M6decine-dentaire, 19, rue Barth6lemy-Menn, 1211 Geneve 4, Switzerland This investigation was supported in part by the British Columbia Health Research Foundation grant #107(90-1) and by Mitsui Petrochemical Industries, Ltd., Japan.
alloys (Krabbendam et al., 1987; Tanaka et al., 1988a; Watanabe et al.,1988). Like all adhesives, dental luting cements are sensitive to the concentration of stress, and their bonding performance is related closely to the test procedures (Lees, 1989). The most common bond tests used in dental research consist of a shear stress or a tensile stress applied to a butt joint. A pure tensile stress is difficult to achieve. A very small deviation in the direction of the load can concentrate stress in a particular area of the joint and lower the resistance to fracture (Lees, 1984). Moreover, even in a perfect experimental setting, the stress may not be distributed evenly along the interface (Lees, 1985; Van Noort et al., 1989). As an alternative to tensile testing, the shear bond test is probably more representative of the stress applied on dental restorations. Van Noort et al. (1989) have shown that the distribution of the shear stress within ajoint depends greatly on the location ofthe loading point. Shearing devices range from a wire loophole attached around a cylindrical sample (Tagami et al., 1990; Sorensen and Dixit, 1991; Prati et al., 1991) to more elaborate devices in which the direction of the stress is more rigidly controlled (Dilts et al., 1985; Krabbendam et al., 1987; Belser et al., 1990). In theory, the rigid devices should provide a more controlled and reproducible load to the joints, which, in turn, should reduce variation in results. Four-point bending tests have also been used for testing bond strength, but the coefficient of variation obtained was no less than that obtained with a tensile test (0ilo and Evje, 1988; Keeny et al., 1990). Whatever the test used, it has been observed that the bond strength data do not follow a normal distribution (McCabe and Carrick, 1986; De Rijk et al., 1986). For this reason, statistical analysis based on the mean and standard deviation could be misleading. The use of a Weibull model has been shown to be a better alternative (De Rijk et al., 1986; McCabe and Walls, 1986; Mojon et al., 1989; Drummond and Miescke, 1991). This statistical technique relates the stress applied to thejoints to the probability offailure. The model is defined by two or three parameters (Weibull, 1951) that
Fig. 1-The cementing jig. a = fixed sample-holder, b = sliding sampleholder, c = spacer (0 and 25 pm), and d = zeroing device.
Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
1633
1634
J Dent Res September 1992
MOJON et al.
characterize the resistance to fracture of the joints and allow test groups to be compared. A minimum of 30 datum points has been recommended if accurate estimates of the Weibull parameters are to be obtained (Davies, 1973). This study was deC signedto evaluate thebond strength ofa glass-ionomer cement to a precious PFM alloy, to determine the influence of the cement as it matured, and to compare the results with the bond strength created by a zinc a phosphate cement and an adhesive resin cement con4taining
F
b
m
Materials and methods. Preparation of the samples.-A high-noble PFM alloy (V-Delta SF, Me'taux precieux SA Me'talor, Neuchatel, Switzerland) was used for casting 30 cylinders (6 mm in diameter by 8 mm in
length). One base of each cylinder was abraded with a to
grade 80 abrasive paper reproduce the typical
surface roughness found at the interface between a tooth and an artificial crown. The surface profile created by a "finishing" diamond bur (Premier Dental Canada, Scarborough, Canada) used for preparation of teeth for crowns was compared with the roughness produced by two abrasive papers in a preliminary study. The average surface roughness (ASV) was measured on six samples with a profilometer (TalysurfV, Taylor and Hobson, Leicester, England), and it was shown that the ASV (mean S.D.) of the 80-grade paper (1.7 0.1 pm) was similar to that ofthe bur (1.8 0.2 pm), while the 120-grade paper produced a smoother surface (0.6 0.1 pm). The abraded surfaces were examined with a reflected-light microscope (20x), and the abrasion was repeated, if necessary, to obtain a surface free of porosity or other defects. The cylinders were thoroughly rinsed with tap water and cleaned with a dental solvent (Siccavit, Pharmachemie AG, Z rich, Switzerland). A pen mark on ±
±
±
0.80
t
0.70 0.60
0 la
0.50 0 40
0.30 0.20 0.10 0.00 0.00
8.00
4.00 0
GIl
12.00
16.00
20.00
24.00
28.00
shear strength (MPa) + o G17 GI2
Fig. 3-Bond strength of GI after 1, 2, and 7 days in water.
ethacryloxyethyl the outer surfaces ofthe cylinders indicated the general direction of
trimellitate anhydride (4META).
Fig. 2-The shear device (Belser et al., 1990). a = hollow cylinder, b = split piston, c = sample cavity, d = space created in the piston to allow fracture of the joint, and F = loading point.
1.00 0.90
±
the grooves created by the abrasive paper. The cylinders were aligned end-to-end in ajig (Fig. 1), and joined in pairs with a glass-ionomer cement, a zinc phosphate cement, or an adhesive resin cement containing 4-META (Table 1). The cementing jig had been machined with a 3-pm tolerance limit, and it was designed to provide a 25-pm gap between the cylinders. Prior to cementation, the cylinders were rotated in the holder until the pen marks matched each other, so the grooves were oriented in the same direction on both surfaces. Excess cement was removed from the surfaces of the joints after 10 min, and two layers of a dental varnish (GC Fuji varnish, G-C Dental Industrial Corp., Tokyo, Japan) were applied around the joints before they were placed for storage in water at 370C. After day 1, 2, or 7 (Table 2), the 15 joints were ruptured in shear, and the preparation process was repeated so that data from 40 joints for each of the seven test groups could be obtained. Shear test.-The shear tests were conducted on an Instron testing machine (Instron Corporation, Canton, MA) equipped with a compressive load cell (90-4500 kg) and a chart recorder running at 5.0 x 101 m/s. The cross-head speed was set at 2.1 x 10 rm/s. The samples were held in a device (Fig. 2) designed to apply a pure shear force to the joints perpendicular to the direction of the grooves (Belser et al., 1990). Fractography.-After each batch of 15 samples had been tested, the broken joints were ranked according to their resistance to shear failure, and they were examined under low (20x) magnification with a reflected-light microscope. Polyvinylsiloxane (Reprosil light, Caulk Division, Dentsply Int., Milford, DE) impressions of the weakest and strongest joints were taken and poured in an epoxy resin (Epo-Tek, Epoxy Technology Inc., Billerica, MA). The replicas were plated with palladium (Hummer IV, Technics, San Jose, CA) for examination with a scanning electron microscope (SEM) (Stereoscan 100, Cambridge Instruments Limited, Cambridge, UK).
TABLE 1 DENTAL LUTING CEMENTS Code
Type
Brand Name
Manufacturer
GI
Glass-ionomer
Aqua Cem
De Trey AG Zurich, Switzerland
ZP
Zinc Phosphate
Fleck's
Mizzi, Inc. Clifton Forge, VA
AR
Adhesive Resin
Super-Bond
Sun Medical Kyoto, Japan
Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
Powder/Liquid g/mL
3.30 2.70 0.35
0
1.00
1.00
0o9o
0.90
0.80
0.80
0.70
0.70
0.60
0.60 A
O
0.
1635
EARLYBOND STRENGTH OF LUTING CEMENTS
Vol. 71 No. 9
0.50
S
R~1
To
da
0.40
0.50 0.40
0.30
0.30
0.20
0.20
0.10
0.10
0.00
0.00
0.00
4.00 0
ZP1
12.00
8.00 +
16.00
shear strength (MPa) ZP7 ° ARI
20.00 a
24.00
28.00
AR7
Fig. 4-Bond strength of ZP and AR after 1 and 7 days in water.
0.00
4.00
8.00
12.00
20.00
16.00
24.00
28.00
shear strength (MPa) 0
G17
+
ZP7
0
AR7
Fig. 5-Bond strength of GI, ZP, and AR at 7 days in water.
A selection offractured joints was also plated and observed directly After one week in water, it was estimated that 90% of the adhesive resinjoints and three-quarters (76%) ofthe glass-ionomer with the SEM. Statistical analysis.-The fracture stress of each joint was di- joints would survive a shear stress of 8 MPa, while most (90%) of vided by the cross-sectional area (28.3 mm2) of the joint so that the the zinc phosphate joints would fail (Fig. 5). After the first day, fracture stress would be obtained. The probability of failure was the characteristic strength and the 10th, 50th, and 90th percenexpressed as a function ofthe fracture stress for each test group, and tiles of zinc phosphate samples were significantly lower than the it was estimated with the following equation (Weibull, 1951): figures obtained for the glass-ionomer samples, which, in turn, were lower than those obtained from the adhesive resin samples P(oJ =1 - exp [- (a -aIa)m] (Table 3). The same ranking was found with samples kept in water for one week, but the adhesive resin and the glass-ionomer samples did not differ statistically at the 10% and 50% chance of where P(ar) = probability of failure at stress a (MPa); m = Weibull modulus; failure. The seven test groups had similar Weibull moduli (m), which indicates that none of the cements was more predictable (s = characteristic strength (MPa); and than the other, and that the storage time did not influence this vu = threshold stress (MPa). The Weibull modulus (m) gives an indication of the predictabil- parameter. Data from the first and seventh days were combined ity ofthe failure. For example, a brittle material, on average, may because the storage time had little effect on the bond strength of withstand high stresses, but some samples willbreak at low stresses, the zinc phosphate or the adhesive resin (Table 4). The same and a few will resist much longer. The Weibull modulus of such a merging was applied to data for the second and seventh days for material will be low compared with that of a ductile material that glass-ionomer cement. Re-analysis of the combined data demonfails at stresses within a narrow range. The characteristic strength strated an increase in the strength of the glass-ionomer bonds (a() relates to the strength ofthe bond. The higher the characteristic after two days, although the difference was still not significant at strength, the stronger the bond for a given value of m and cr . The a 10% chance offailure. The adhesive resin proved stronger than threshold stress is the minimum stress at which a fracture can the glass ionomer at all levels of stress. Fractography.-Fractures occurred predominantly atthe metaloccur. In the models used in this study, we assumed that the threshold stress was zero. The parameters and their respective 95% cement interface, with patches of cement remaining on both sides of confidence limits were evaluated for each test-group by use of a the broken joints. However, the SEM observation revealed that maximum-likelihood estimator technique as described by Cohen (1965) and by Bain and Engelhardt (1981). The 10th, 50th, and 90th TABLE 2 percentiles and their respective 95% confidence limits were calculated following the method described by Nelson (1982). Further, a TEST CONDITIONS Bonferroni correction (Mendenhall et al., 1986) was used to adjust the confidence limits for cumulative error probability. Differences Test No.of Storage between test groups were considered statistically significant when Group Joints Time Cement the confidence intervals did not overlap. 40 GI GI1 1 day Results. 40 2 days G12 GI Shear test.-The plot of shear strength against the probability 40 GI 7 days of failure indicated that the glass-ionomer samples were stron- GI7 ger after maturing for two days (Fig. 3). The characteristic ZP1 1 day 40 ZP strength (ar0) and the 90th percentile increased significantly from day 1 to day 7 with this cement, whereas none of the other ZP7 40 ZP 7days variables proved statistically different within the three GI test groups (Table 3). The zinc phosphate cement and the adhesive AR1 40 1 day AR resin cement were not significantly affected by the storage time 40 7 days AR7 AR (Fig. 4, Table 3). Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
M 1NJ Dent Res Septen 1C636 et al.
r
1992
Fi 6--DreetS obserationofadacturedGjoint bowing sphcrncal PE aoy and If ne of vods (arrows) in the cent. c cement,
o It isprbale thatth tearing Fi 7 hEM ia ofa eturedAl jnt ad peeng occurcd aftr the ini iiffre of the oin ccetent, a = PFM alby and If l of atu
some cementremained in the bottoms ofthe grovs yerted durn
Discussion Cement adeson.The roughness othe bonded surices has been reported prevously (Pre and Sutovw 1988) aid, under these condions he rsin ameit bonded more securely to e precious alloy tan did either the z phsphat or glass-ionomer cement. Al oflhe ceme hats bthan metal or bhvd less predictaly ceramn mcaerals. The EM observations revealed a lack of adhesion between e zin prosphate cement rd the precious alloy but the mechanica retention produced sone resstance to shear. Thi feature m xplaia to a large exten the ong histo y ofclinical satisfaction wit ac phosphate cmeunts Apart You zie piosphate ceme, very lie information has been pahished on the r flu cements to precious ay(Moser et al., 1974,. bond strewn Dils et al. 1985). Sevral authors Hotet al., 1977, Dilts et aci 1985, Tanaka et , 1988a) have reported that there is no adhesion between resin cements and precious metal while others (Krabbendam et al. 1987, Watanabe et l., 1988) have reported a resistance of up to 22.0 MPa. Unfbturately, the devices used for strength tests the small sample sizes and the methods of handling
=
fracture.
he preparation o he samples (Figs 6 and 7) The raking othe samples according to their resistance to shear-ftacture did rot identify a difference between strong or eak samples rd the SEM examinations filed to reveal anything that could be correlated witt the strength of the joints. Storage tme di at influence te ieth e micresopci appearance of te fa re sources
a diffbeit rtfce topography. The ine of fracture witn the zire phosphate was smooth (ig. 8) compared with the irregular, contoured, or fceted fracture line in the gass ionomer (Fig. 6). A gap indicative of a poor bond, was observed invariably between the aznc phosphate and he met at high magnifcation (Fig 9 whereas parts ofthe glass ionome appeared to be in contactaith the alloy (fi. 10) The adhesive resin factured in a more complex pattern Sometimes a small part of the ement remained on one side, leaving a hole in the cement layer on the opposite side, and occasionally edges of the cement were deformed and separated from the metal (Fi 7) Yet in other areas, the resin adhered to the alloy and cracks appeared within the cement (Fig. 11).
cements exhibited
TABLE 3 STATISTICAL RESULTS FOR THE SEVEN TEST GROUPS RANKED ININCREASING ORDER ACCORDING TO THE CHARACTERISTIC STRENGTHl Chane of fiilur
Test Group
S.D. (MPa)
m
ZP1
Man (Ma) 4.7
2.3
22.
ZP7
5.2
2.0
2.8
GI! G12 G17 AR7
8.2 10.6 11.2
2.8
321
3.8 4.1
31 2.8
11.91 12.6
15.7
5-9
249
17.71
AR1
17:0
Weibull modulus, overlapping each other. m
a
strength.
Common
(MPa)
verti
90% (MP) 7,7
50% (MPa)
5_3
10% (MPa) 1.9
59
2.7
521
7.9
4v6 5,7
8.21 10.61
11.8
5.
110
16.8
8.1
15.61
23.6
4,51
10.3 1 17.1 95 fa in dicate values with confidence
18.81
3.7
54
characteristic
a
Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
15.6
2361 intervals
Vol.Vo7lodEARLY 71 I>. 9 BOND SRENGHOLUEN
LEVINT1163
f 9-SEMit IPE al y Fi 8-EM ige far tned ZPjoint e ofa fiared j sho aapa(wb)teven h ine and If f al cement. c -cementa=alloaf=lieofftaer of ftacthe ,
Fte data provide Isults that carnot b e spared Adamis 1991). the adheiv resir We otitcd a decrease in bord st er i sh fh& in clientweiit reac A maimUSu e (Sorensen and Dxt, 1991) manufactr. So as advised oes we cant onfi that the product old iot h used aer he expirati date. e islign 1 et .I sexped tt MVIhanismf ment o iecinders, when cemented and when tsted in s iear, would teducethe variainnrfesul~ts Howeer, the coeents of vacationifr t shear srents obtained in tFis study wer similar t ose prvously reported (Dilts et 1., 1985 Krabbendam et a., 1987; Banraziletal. 1988;Naegelietal 1988). Pdar othe riaion may be an irheret charaersti ofthe points. e urs modri of ie parent ietaL and the adhesives were qite ifren and produced an unven dist io of he stress te ierice (Eley i961)6 fhi phenomenoni$ ror bablyt erse nitve to smal gs strss conctratiet maohi intheappltat oft o toad varied rom on staple t another desp te prse matching of tie apparaus All joints fractured suddenly with no detectabe plasi drmation on the load-time recordings. The teangad peelg observed on some adhesie resin samnples probably cured dung iur when the shear oad was decreasing rapidly to zer Indee a fw tytindes remained loosely connted en they ee roved from tie shear device tir e test I
i
Keny eatal (1990 h leshown hat e actu of joints s ardthe atoy All tree iniated ath at c beten e cements rur study xied surface defeats tiat could have but c id not see a dettha ideified weakened jinnts, the without dout the tini poinr of ture Spherical pores or example werseen the assioner ent bu t path of factur passed o asronal rough Ftres It is also possible a Fre aas of poor hec adsiron and Fat oud ni be seen on e ac d surfces Fna any surface irreguarit~ies on the cylinders ma have zuinitited the crack (Alen 1984)IThis Ferhypothesis is propagation of te ceten re iand te obsevation that sme of re supported byh varni applied aroundte ints ride te isice groovs was partialy dssolved an coud bte peeled of after ne day in t harde .ees notinllenreFresistanctofractureo Fthjoirt water,t so it dirds Maturation time.-Calciumr salts cause te initial hardening of lassionome cements hil auminuw salts crbut tie fial prpertes t he product (C£ p et al., 1974, Crisp and Wison. 19 74 M iLean and Wison, 1977). Ih waium impressive strenF sri haFress aind rsistanc to dissolution of the t one to there dys after the ceent is mixed cement achieved because the trecipitatior the aluminum sa is slo (Crisp et al., 1976; Muit andMakinsn 1982 Mc neyt a. 1987). We ound that the maximum bond strength of the water-activated assinoer oturrd r to days, but ha their was no maturation osered in the zin phosia event tPosser et l(1984d)en-
TABE 4
SATlSTICAL RESULTS FOR THE FOUR COMBINED TEST GRUPS RANKED IN INCREASING ORDER ACCORDING TO THE CHARACTERISTIC SIRENGil Chance of failure Combined in dean MPa) ci (MP'a) 1o0 (MPa) 50% (MPa) 90% (MPa) S.D. (MPs Test Groups 7.9 448 ZPI + ZP7 2.2 5.0 Gil 9.1 8.2 11.8 2.8 4,6 8.2 3.2 16.3 10 8 53 G12 + Gil 10.9: 4.0 12.2 2.9 18 3 23.7 9.1 16:3 5.7 16.3 3.2 AR1 + AR7 95 ne Isg m, Weibull modulus; cO, characteristic strength. Common vertical line-s -nditte tues iapthn each other.
Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
1638
MQJON et al.
J Dent Res September 1992
Fig. 10-hEM image of a factured GI joiht Tie arrows show the integrity oft all ycemcv initerfie. ce cem t al = FM alloy.
Fig. 11 SEM image of a fractured AR joint showing cracks (arrows) wiin the ce nert. e = ceat al PFM alloy
strated that wateractivated and conventional glass ionorers have the same properties after one day; therefore, it is nlikl that a conventional glasswinomer cement would have performed differ ently The adhesive resinjoints were slightly weaker after one week in water, although the differee was not statistically siniicant This fing is n cordne withthat of azArnoldet a. (1989) who found a significant decrease in 4-META bond strength after 30 day in water.
precpiliation reaction. J Dent Res53 420-1425. Des DS (1973) The statistica approach to engineering designs in eramics Proc Br Cerm Soc 22.429-452. De Rijk WJ Ts JA Conner ML (1986) The distribution o failure stresses iv porcelain to meta systems (sact). JDentRes 65217. DiazrArnld AM, Wihiams VD Aquilino SA (1989). Tensile strengths of three luting age for adhesion of fixed partial dentures. ht J Psthdnt 2115-122. Dilts WE, Dicaison MC Jr Miranda FJ Brackett SE (1985). Relative hear bond stregths of outing media with various core materials. J Prosthet Dent 53:505 508. Druvmond JL, Miescke KJ (1991). Weibull models fr the statistical analysis of dentalcomposite data: Aged in physiologic media and cyclicfitigued. Dnt r7*25-29. Ele DD (1961). The distnbution of stress in adhesive joints. In. Eley DD, editor. Adhesio. Oxford Oxford University Press, 207 253. Hotz PM LeanJW Seed I WIsonAD (1977). Theaondingofgiassionomer cements o metal and tooth substrates. Br Dent J 142 41-47. Keevy SM 1:II Sato Y, Tesk JA (1990). Bond strength of resin bonded systems in tension and bending (abstract. J Dent Re 69208. dA HC Duijsters PPE, Davidson CL (1987). learkel Ten Krabbendam Siaer bond strength determinatios on vanous kinds of luting cements h tooth structure and cast allows usian a new testing device. 3 Dent 15.77-81. Laufer B-Z Nicholls J, Towsend D (1988). SiOx C coating: A composite omet lbondngmmechanism. J3Post etDent 60:320-327. Lees WA (984). Design. In: Lees WA, editor Adhesives in engineering design. Berlin: Spnngererlag, 6-12. Lees WA (1985). Stress distribution a bonded joints An exploration within a mthematical niodel Mkterials & Desan> 6 117 123. Lees WA(1989). Design. In: Lees WA editor. Adhesives and the engineer. London Mechanical Engineering Publication, 33-73 McCabe JF Carck TE (1986). A statistical approach to the mechanical testing of dental materials. Dent Mter 2 139 142. McCabe JF Walls AW( 1986). The treatment of results for tensile bond
Acknowledgments We acowledge wihgatitude, Metaux Pre'iex SA Me'talor, Neuchatel Switzerland, for the precious metal the Division of Fixed Prosthodontics and Occlusion, University ofGeieva, Switzerland or the shear-testing device; Dr. H Hawthorne om the National Research Council of Canada, Departmen offribology and Mechanics for providing tie prfilometer and assisting with the analysis; and Dr. M Carius, Vancouver for the cementing jig. REFERENCES
Adams RD (1991). Testirg of adhcsyess eful o rnot Adhrsion 15:1-18. Mien KW (1984). Fundamental aspects of adiheson In.J Le WA, editor. Adhesives in eineerirg desi. Berlin Spriner Verlag 129139. Bain LJU Enelhardt M (1981). Simpe approximate distriutional results ifr confidence and tolerance limits for the WeiblI distribution based on maximum likelihood estimators. Tchrometretr 23:15 20 Barrilay I, Myers ML, Coope LB, Craser CN (1988) Mchaical and chemical rerteion of laboratory cured composite to metal surfics. J Prosthet Dent 59:131 137. Belser UC, LaBar , Bumon M Myer J-M (1990) A new shear store gh test for adhsie prosthodontivs (abstract). J Dnt Res 69:358. Button GL, Barnes RF, Moon PC (1985). Surface preparation and shear bond strength of the casting eemrt nterface. J Piothet Dent 53:34-38. Cohen AC (1965). Maximum likeliood station i the Weilmll dintibution chased on coi plete and on censored samples. hcirhometrs 75739
strength testing JDet 14165-168. McKinneydE, Antonucci JM, Rupp NW (1987). Wear and microhardness of 588. glass-ionmer cements. JDent Res 66:1134-1139. Crisp S, Levis BGC Wilson AD (1976) Characteriza orf si4nmrer McLean JW (1977), A new method of bonding dental cements and porcelain cements. 1 Long term ardness ad compressive strength. J Dent to metal sun aces. Oper Dent 2:130 142. 4:162-166. McLean JW Wilson AD (1977). The clinical development of the glassCnisp S Pringuer MA, Wardleworth D, Wison Al (1974). Reacts in ionoir cevients.I. Formulations and properties. AustDentJ 223136. glass-ionomer cements. 1I. An infred spet3roscopic study. JDentR Men denha1L W, Scaeffr R, Wackerly D (1986). Simultaneous confidence 53.1414 1419. intervals fo more than one parameter. In: Mathematical statistics with Crisp 8, Wilson AD (1974). Reactions in glass-ionomei cements II. The application. Boston: Duxury Press, 559 560. Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
Vol. 71 No. 9
EARLY BOND STRENGTH OF LUTING CEMENTS
Mojon P, Hawbolt EB, MacEntee MI, Belser UC (1989). Maximum bond strength of dental luting cement to amalgam alloy. JDent Res 68:15451549. Moser JB, Brown DB, Greener EH (1974). Short-term bond strengths between adhesive cements and dental alloys. JDentRes 53:1377-1386. Mount GJ, MAkinson OF (1982). Glass-ionomer restorative cements: Clinical implications of the setting reaction. Oper Dent 7:134-141. Naegeli DG, Duke ES, Schwartz R, Norling BK (1988). Adhesive bond of composites to a casting alloy. J Prosthet Dent 60:279-283. Nakabayashi N, Kojima K, Masuhara E (1982). The promotion ofadhesion by the infiltration of monomers into tooth substrates. J Biomed Mater Res 16:265-273. Nelson W (1982). Maximum likelihood analysis of multiple censored data. In: Applied life data analysis. New York: Wiley, 343. 0ilo G, Evje DM(1988). Abendtestformeasuringcement-dentinbond. Dent Mater 4:98-102. Phillips RW (1973). Dental cements for luting and thermal insulation. In: Phillips RW, editor. Skinner's science of dental materials. 7th ed. Philadelphia: W.B. Saunders Company, 466-497.
PratiC,NucciC,MontanariG(1991). Shearbondstrengthandmicroleakage of dentin bonding systems. J Prosthet Dent 65:401-407. Price RB, SutowEJ (1988). Micrographic and profilometric evaluation ofthe finish produced by diamond and tungsten carbide finishing burs on enamel and dentin. JProsthet Dent 60:311-316. Prosser HJ, Powis DR, Brant P, Wilson AD (1984). Characterization ofglassionomer cements. 7. The physical properties ofcurrent materials. JDent
1639
12:231-240. Sorensen JA, Dixit NV (1991). In vitro shear bond strength of dentin adhesives. Int JProsthodont 4:117-125. Tagami J, Tao L, Pashley DH (1990). Correlation among dentin depth, permeability, and bond strength ofadhesive resins. Dent Mater 6:45-50. Takeyama M, Kashibuchi N, Nakabayashi N, Masuhara E (1978). Studies on dental self-curing resin-adhesion of PMMA to bovine enamel or dental alloys. J Jpn Soc Dent Appar Mater 19:179-185. Tanaka T, Atsuta M, Nakabayashi N, Masuhara E (1988a). Surface treatment of gold alloys for adhesion. J Prosthet Dent 60:271-279. Tanaka T, Hirano M, Kawahara M, Matsumura H, Atsuta M (1988b). Anew ion-coating surface treatment ofalloys for dental adhesive resins. JDent Res 67:1376-1380. Tanaka T, Nagata K, Takeyama M, Atsuta M, Nakabayashi N, Masuhara E (1981). 4-META opaque resin-a new resin strongly adhesive to nickelchromium alloy. J Dent Res 60:1697-1706. Van Noort R, Noroozi S, Howard IC, Cardew G (1989). A critique of bond strength measurements. J Dent 17:61-67. Watanabe F, Powers JM, Lorey RE (1988). In vitro bonding of prosthodontic adhesives to dental alloys. J Dent Res 67:479-483. Weibull W (1951). A statistical distribution of wide applicability. JAppl Mechan 18:293-297. Yamashita A (1983). New method of adhesion. J Jpn Dent Assoc 35:10741087. Yu X-Y, XU J-W (1987). The tensile bond strength of various composite resins to alloy. Quint Int 18:145-147.
Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
J Dent Res 71(9):1633-1639, September, 1992
Introduction. Zinc phosphate cement has been used successfully for several decades for outing artificial crowns on teeth, but it does not bond securely to dental alloys (Phillips, 1973; Moser et al., 1974; Dilts et al., 1985). Glass-ionomer cement, in contrast, will bond to base alloys and, to some extent, to precious alloys (McLean, 1977; Button et al., 1985; Krabbendam et al., 1987). However, the glass-ionomer cement remains weak and sensitive to dissolution in the early stage of a complex setting reaction (Crisp and Wilson, 1974; Crisp et al., 1976; McLean and Wilson, 1977; Mount and MAkinson, 1982), and it is unknown how this maturation process affects the bond strength of the material. Adhesive resins have been developed recently for luting artificial crowns on teeth (Takeyama et al., 1978; Nakabayashi et al., 1982; Yamashita, 1983). They provide a stronger bond to base metals than do glass-ionomer cements (Tanaka et al., 1981; Krabbendam et al., 1987; Yu and Xu, 1987), and they will adhere also to precious alloys when the metal surface is treated to enhance the bond (McLean, 1977; Laufer et al., 1988; Tanaka et al., 1988a,b; Watanabe et al., 1988). It is not clear, however, whether these prosthodontic adhesives will bond to untreated precious metal Received for publication October 21, 1991 Accepted for publication April 6, 1992 3Present address for correspondence: Ecole de M6decine-dentaire, 19, rue Barth6lemy-Menn, 1211 Geneve 4, Switzerland This investigation was supported in part by the British Columbia Health Research Foundation grant #107(90-1) and by Mitsui Petrochemical Industries, Ltd., Japan.
alloys (Krabbendam et al., 1987; Tanaka et al., 1988a; Watanabe et al.,1988). Like all adhesives, dental luting cements are sensitive to the concentration of stress, and their bonding performance is related closely to the test procedures (Lees, 1989). The most common bond tests used in dental research consist of a shear stress or a tensile stress applied to a butt joint. A pure tensile stress is difficult to achieve. A very small deviation in the direction of the load can concentrate stress in a particular area of the joint and lower the resistance to fracture (Lees, 1984). Moreover, even in a perfect experimental setting, the stress may not be distributed evenly along the interface (Lees, 1985; Van Noort et al., 1989). As an alternative to tensile testing, the shear bond test is probably more representative of the stress applied on dental restorations. Van Noort et al. (1989) have shown that the distribution of the shear stress within ajoint depends greatly on the location ofthe loading point. Shearing devices range from a wire loophole attached around a cylindrical sample (Tagami et al., 1990; Sorensen and Dixit, 1991; Prati et al., 1991) to more elaborate devices in which the direction of the stress is more rigidly controlled (Dilts et al., 1985; Krabbendam et al., 1987; Belser et al., 1990). In theory, the rigid devices should provide a more controlled and reproducible load to the joints, which, in turn, should reduce variation in results. Four-point bending tests have also been used for testing bond strength, but the coefficient of variation obtained was no less than that obtained with a tensile test (0ilo and Evje, 1988; Keeny et al., 1990). Whatever the test used, it has been observed that the bond strength data do not follow a normal distribution (McCabe and Carrick, 1986; De Rijk et al., 1986). For this reason, statistical analysis based on the mean and standard deviation could be misleading. The use of a Weibull model has been shown to be a better alternative (De Rijk et al., 1986; McCabe and Walls, 1986; Mojon et al., 1989; Drummond and Miescke, 1991). This statistical technique relates the stress applied to thejoints to the probability offailure. The model is defined by two or three parameters (Weibull, 1951) that
Fig. 1-The cementing jig. a = fixed sample-holder, b = sliding sampleholder, c = spacer (0 and 25 pm), and d = zeroing device.
Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
1633
1634
J Dent Res September 1992
MOJON et al.
characterize the resistance to fracture of the joints and allow test groups to be compared. A minimum of 30 datum points has been recommended if accurate estimates of the Weibull parameters are to be obtained (Davies, 1973). This study was deC signedto evaluate thebond strength ofa glass-ionomer cement to a precious PFM alloy, to determine the influence of the cement as it matured, and to compare the results with the bond strength created by a zinc a phosphate cement and an adhesive resin cement con4taining
F
b
m
Materials and methods. Preparation of the samples.-A high-noble PFM alloy (V-Delta SF, Me'taux precieux SA Me'talor, Neuchatel, Switzerland) was used for casting 30 cylinders (6 mm in diameter by 8 mm in
length). One base of each cylinder was abraded with a to
grade 80 abrasive paper reproduce the typical
surface roughness found at the interface between a tooth and an artificial crown. The surface profile created by a "finishing" diamond bur (Premier Dental Canada, Scarborough, Canada) used for preparation of teeth for crowns was compared with the roughness produced by two abrasive papers in a preliminary study. The average surface roughness (ASV) was measured on six samples with a profilometer (TalysurfV, Taylor and Hobson, Leicester, England), and it was shown that the ASV (mean S.D.) of the 80-grade paper (1.7 0.1 pm) was similar to that ofthe bur (1.8 0.2 pm), while the 120-grade paper produced a smoother surface (0.6 0.1 pm). The abraded surfaces were examined with a reflected-light microscope (20x), and the abrasion was repeated, if necessary, to obtain a surface free of porosity or other defects. The cylinders were thoroughly rinsed with tap water and cleaned with a dental solvent (Siccavit, Pharmachemie AG, Z rich, Switzerland). A pen mark on ±
±
±
0.80
t
0.70 0.60
0 la
0.50 0 40
0.30 0.20 0.10 0.00 0.00
8.00
4.00 0
GIl
12.00
16.00
20.00
24.00
28.00
shear strength (MPa) + o G17 GI2
Fig. 3-Bond strength of GI after 1, 2, and 7 days in water.
ethacryloxyethyl the outer surfaces ofthe cylinders indicated the general direction of
trimellitate anhydride (4META).
Fig. 2-The shear device (Belser et al., 1990). a = hollow cylinder, b = split piston, c = sample cavity, d = space created in the piston to allow fracture of the joint, and F = loading point.
1.00 0.90
±
the grooves created by the abrasive paper. The cylinders were aligned end-to-end in ajig (Fig. 1), and joined in pairs with a glass-ionomer cement, a zinc phosphate cement, or an adhesive resin cement containing 4-META (Table 1). The cementing jig had been machined with a 3-pm tolerance limit, and it was designed to provide a 25-pm gap between the cylinders. Prior to cementation, the cylinders were rotated in the holder until the pen marks matched each other, so the grooves were oriented in the same direction on both surfaces. Excess cement was removed from the surfaces of the joints after 10 min, and two layers of a dental varnish (GC Fuji varnish, G-C Dental Industrial Corp., Tokyo, Japan) were applied around the joints before they were placed for storage in water at 370C. After day 1, 2, or 7 (Table 2), the 15 joints were ruptured in shear, and the preparation process was repeated so that data from 40 joints for each of the seven test groups could be obtained. Shear test.-The shear tests were conducted on an Instron testing machine (Instron Corporation, Canton, MA) equipped with a compressive load cell (90-4500 kg) and a chart recorder running at 5.0 x 101 m/s. The cross-head speed was set at 2.1 x 10 rm/s. The samples were held in a device (Fig. 2) designed to apply a pure shear force to the joints perpendicular to the direction of the grooves (Belser et al., 1990). Fractography.-After each batch of 15 samples had been tested, the broken joints were ranked according to their resistance to shear failure, and they were examined under low (20x) magnification with a reflected-light microscope. Polyvinylsiloxane (Reprosil light, Caulk Division, Dentsply Int., Milford, DE) impressions of the weakest and strongest joints were taken and poured in an epoxy resin (Epo-Tek, Epoxy Technology Inc., Billerica, MA). The replicas were plated with palladium (Hummer IV, Technics, San Jose, CA) for examination with a scanning electron microscope (SEM) (Stereoscan 100, Cambridge Instruments Limited, Cambridge, UK).
TABLE 1 DENTAL LUTING CEMENTS Code
Type
Brand Name
Manufacturer
GI
Glass-ionomer
Aqua Cem
De Trey AG Zurich, Switzerland
ZP
Zinc Phosphate
Fleck's
Mizzi, Inc. Clifton Forge, VA
AR
Adhesive Resin
Super-Bond
Sun Medical Kyoto, Japan
Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
Powder/Liquid g/mL
3.30 2.70 0.35
0
1.00
1.00
0o9o
0.90
0.80
0.80
0.70
0.70
0.60
0.60 A
O
0.
1635
EARLYBOND STRENGTH OF LUTING CEMENTS
Vol. 71 No. 9
0.50
S
R~1
To
da
0.40
0.50 0.40
0.30
0.30
0.20
0.20
0.10
0.10
0.00
0.00
0.00
4.00 0
ZP1
12.00
8.00 +
16.00
shear strength (MPa) ZP7 ° ARI
20.00 a
24.00
28.00
AR7
Fig. 4-Bond strength of ZP and AR after 1 and 7 days in water.
0.00
4.00
8.00
12.00
20.00
16.00
24.00
28.00
shear strength (MPa) 0
G17
+
ZP7
0
AR7
Fig. 5-Bond strength of GI, ZP, and AR at 7 days in water.
A selection offractured joints was also plated and observed directly After one week in water, it was estimated that 90% of the adhesive resinjoints and three-quarters (76%) ofthe glass-ionomer with the SEM. Statistical analysis.-The fracture stress of each joint was di- joints would survive a shear stress of 8 MPa, while most (90%) of vided by the cross-sectional area (28.3 mm2) of the joint so that the the zinc phosphate joints would fail (Fig. 5). After the first day, fracture stress would be obtained. The probability of failure was the characteristic strength and the 10th, 50th, and 90th percenexpressed as a function ofthe fracture stress for each test group, and tiles of zinc phosphate samples were significantly lower than the it was estimated with the following equation (Weibull, 1951): figures obtained for the glass-ionomer samples, which, in turn, were lower than those obtained from the adhesive resin samples P(oJ =1 - exp [- (a -aIa)m] (Table 3). The same ranking was found with samples kept in water for one week, but the adhesive resin and the glass-ionomer samples did not differ statistically at the 10% and 50% chance of where P(ar) = probability of failure at stress a (MPa); m = Weibull modulus; failure. The seven test groups had similar Weibull moduli (m), which indicates that none of the cements was more predictable (s = characteristic strength (MPa); and than the other, and that the storage time did not influence this vu = threshold stress (MPa). The Weibull modulus (m) gives an indication of the predictabil- parameter. Data from the first and seventh days were combined ity ofthe failure. For example, a brittle material, on average, may because the storage time had little effect on the bond strength of withstand high stresses, but some samples willbreak at low stresses, the zinc phosphate or the adhesive resin (Table 4). The same and a few will resist much longer. The Weibull modulus of such a merging was applied to data for the second and seventh days for material will be low compared with that of a ductile material that glass-ionomer cement. Re-analysis of the combined data demonfails at stresses within a narrow range. The characteristic strength strated an increase in the strength of the glass-ionomer bonds (a() relates to the strength ofthe bond. The higher the characteristic after two days, although the difference was still not significant at strength, the stronger the bond for a given value of m and cr . The a 10% chance offailure. The adhesive resin proved stronger than threshold stress is the minimum stress at which a fracture can the glass ionomer at all levels of stress. Fractography.-Fractures occurred predominantly atthe metaloccur. In the models used in this study, we assumed that the threshold stress was zero. The parameters and their respective 95% cement interface, with patches of cement remaining on both sides of confidence limits were evaluated for each test-group by use of a the broken joints. However, the SEM observation revealed that maximum-likelihood estimator technique as described by Cohen (1965) and by Bain and Engelhardt (1981). The 10th, 50th, and 90th TABLE 2 percentiles and their respective 95% confidence limits were calculated following the method described by Nelson (1982). Further, a TEST CONDITIONS Bonferroni correction (Mendenhall et al., 1986) was used to adjust the confidence limits for cumulative error probability. Differences Test No.of Storage between test groups were considered statistically significant when Group Joints Time Cement the confidence intervals did not overlap. 40 GI GI1 1 day Results. 40 2 days G12 GI Shear test.-The plot of shear strength against the probability 40 GI 7 days of failure indicated that the glass-ionomer samples were stron- GI7 ger after maturing for two days (Fig. 3). The characteristic ZP1 1 day 40 ZP strength (ar0) and the 90th percentile increased significantly from day 1 to day 7 with this cement, whereas none of the other ZP7 40 ZP 7days variables proved statistically different within the three GI test groups (Table 3). The zinc phosphate cement and the adhesive AR1 40 1 day AR resin cement were not significantly affected by the storage time 40 7 days AR7 AR (Fig. 4, Table 3). Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
M 1NJ Dent Res Septen 1C636 et al.
r
1992
Fi 6--DreetS obserationofadacturedGjoint bowing sphcrncal PE aoy and If ne of vods (arrows) in the cent. c cement,
o It isprbale thatth tearing Fi 7 hEM ia ofa eturedAl jnt ad peeng occurcd aftr the ini iiffre of the oin ccetent, a = PFM alby and If l of atu
some cementremained in the bottoms ofthe grovs yerted durn
Discussion Cement adeson.The roughness othe bonded surices has been reported prevously (Pre and Sutovw 1988) aid, under these condions he rsin ameit bonded more securely to e precious alloy tan did either the z phsphat or glass-ionomer cement. Al oflhe ceme hats bthan metal or bhvd less predictaly ceramn mcaerals. The EM observations revealed a lack of adhesion between e zin prosphate cement rd the precious alloy but the mechanica retention produced sone resstance to shear. Thi feature m xplaia to a large exten the ong histo y ofclinical satisfaction wit ac phosphate cmeunts Apart You zie piosphate ceme, very lie information has been pahished on the r flu cements to precious ay(Moser et al., 1974,. bond strewn Dils et al. 1985). Sevral authors Hotet al., 1977, Dilts et aci 1985, Tanaka et , 1988a) have reported that there is no adhesion between resin cements and precious metal while others (Krabbendam et al. 1987, Watanabe et l., 1988) have reported a resistance of up to 22.0 MPa. Unfbturately, the devices used for strength tests the small sample sizes and the methods of handling
=
fracture.
he preparation o he samples (Figs 6 and 7) The raking othe samples according to their resistance to shear-ftacture did rot identify a difference between strong or eak samples rd the SEM examinations filed to reveal anything that could be correlated witt the strength of the joints. Storage tme di at influence te ieth e micresopci appearance of te fa re sources
a diffbeit rtfce topography. The ine of fracture witn the zire phosphate was smooth (ig. 8) compared with the irregular, contoured, or fceted fracture line in the gass ionomer (Fig. 6). A gap indicative of a poor bond, was observed invariably between the aznc phosphate and he met at high magnifcation (Fig 9 whereas parts ofthe glass ionome appeared to be in contactaith the alloy (fi. 10) The adhesive resin factured in a more complex pattern Sometimes a small part of the ement remained on one side, leaving a hole in the cement layer on the opposite side, and occasionally edges of the cement were deformed and separated from the metal (Fi 7) Yet in other areas, the resin adhered to the alloy and cracks appeared within the cement (Fig. 11).
cements exhibited
TABLE 3 STATISTICAL RESULTS FOR THE SEVEN TEST GROUPS RANKED ININCREASING ORDER ACCORDING TO THE CHARACTERISTIC STRENGTHl Chane of fiilur
Test Group
S.D. (MPa)
m
ZP1
Man (Ma) 4.7
2.3
22.
ZP7
5.2
2.0
2.8
GI! G12 G17 AR7
8.2 10.6 11.2
2.8
321
3.8 4.1
31 2.8
11.91 12.6
15.7
5-9
249
17.71
AR1
17:0
Weibull modulus, overlapping each other. m
a
strength.
Common
(MPa)
verti
90% (MP) 7,7
50% (MPa)
5_3
10% (MPa) 1.9
59
2.7
521
7.9
4v6 5,7
8.21 10.61
11.8
5.
110
16.8
8.1
15.61
23.6
4,51
10.3 1 17.1 95 fa in dicate values with confidence
18.81
3.7
54
characteristic
a
Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
15.6
2361 intervals
Vol.Vo7lodEARLY 71 I>. 9 BOND SRENGHOLUEN
LEVINT1163
f 9-SEMit IPE al y Fi 8-EM ige far tned ZPjoint e ofa fiared j sho aapa(wb)teven h ine and If f al cement. c -cementa=alloaf=lieofftaer of ftacthe ,
Fte data provide Isults that carnot b e spared Adamis 1991). the adheiv resir We otitcd a decrease in bord st er i sh fh& in clientweiit reac A maimUSu e (Sorensen and Dxt, 1991) manufactr. So as advised oes we cant onfi that the product old iot h used aer he expirati date. e islign 1 et .I sexped tt MVIhanismf ment o iecinders, when cemented and when tsted in s iear, would teducethe variainnrfesul~ts Howeer, the coeents of vacationifr t shear srents obtained in tFis study wer similar t ose prvously reported (Dilts et 1., 1985 Krabbendam et a., 1987; Banraziletal. 1988;Naegelietal 1988). Pdar othe riaion may be an irheret charaersti ofthe points. e urs modri of ie parent ietaL and the adhesives were qite ifren and produced an unven dist io of he stress te ierice (Eley i961)6 fhi phenomenoni$ ror bablyt erse nitve to smal gs strss conctratiet maohi intheappltat oft o toad varied rom on staple t another desp te prse matching of tie apparaus All joints fractured suddenly with no detectabe plasi drmation on the load-time recordings. The teangad peelg observed on some adhesie resin samnples probably cured dung iur when the shear oad was decreasing rapidly to zer Indee a fw tytindes remained loosely connted en they ee roved from tie shear device tir e test I
i
Keny eatal (1990 h leshown hat e actu of joints s ardthe atoy All tree iniated ath at c beten e cements rur study xied surface defeats tiat could have but c id not see a dettha ideified weakened jinnts, the without dout the tini poinr of ture Spherical pores or example werseen the assioner ent bu t path of factur passed o asronal rough Ftres It is also possible a Fre aas of poor hec adsiron and Fat oud ni be seen on e ac d surfces Fna any surface irreguarit~ies on the cylinders ma have zuinitited the crack (Alen 1984)IThis Ferhypothesis is propagation of te ceten re iand te obsevation that sme of re supported byh varni applied aroundte ints ride te isice groovs was partialy dssolved an coud bte peeled of after ne day in t harde .ees notinllenreFresistanctofractureo Fthjoirt water,t so it dirds Maturation time.-Calciumr salts cause te initial hardening of lassionome cements hil auminuw salts crbut tie fial prpertes t he product (C£ p et al., 1974, Crisp and Wison. 19 74 M iLean and Wison, 1977). Ih waium impressive strenF sri haFress aind rsistanc to dissolution of the t one to there dys after the ceent is mixed cement achieved because the trecipitatior the aluminum sa is slo (Crisp et al., 1976; Muit andMakinsn 1982 Mc neyt a. 1987). We ound that the maximum bond strength of the water-activated assinoer oturrd r to days, but ha their was no maturation osered in the zin phosia event tPosser et l(1984d)en-
TABE 4
SATlSTICAL RESULTS FOR THE FOUR COMBINED TEST GRUPS RANKED IN INCREASING ORDER ACCORDING TO THE CHARACTERISTIC SIRENGil Chance of failure Combined in dean MPa) ci (MP'a) 1o0 (MPa) 50% (MPa) 90% (MPa) S.D. (MPs Test Groups 7.9 448 ZPI + ZP7 2.2 5.0 Gil 9.1 8.2 11.8 2.8 4,6 8.2 3.2 16.3 10 8 53 G12 + Gil 10.9: 4.0 12.2 2.9 18 3 23.7 9.1 16:3 5.7 16.3 3.2 AR1 + AR7 95 ne Isg m, Weibull modulus; cO, characteristic strength. Common vertical line-s -nditte tues iapthn each other.
Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
1638
MQJON et al.
J Dent Res September 1992
Fig. 10-hEM image of a factured GI joiht Tie arrows show the integrity oft all ycemcv initerfie. ce cem t al = FM alloy.
Fig. 11 SEM image of a fractured AR joint showing cracks (arrows) wiin the ce nert. e = ceat al PFM alloy
strated that wateractivated and conventional glass ionorers have the same properties after one day; therefore, it is nlikl that a conventional glasswinomer cement would have performed differ ently The adhesive resinjoints were slightly weaker after one week in water, although the differee was not statistically siniicant This fing is n cordne withthat of azArnoldet a. (1989) who found a significant decrease in 4-META bond strength after 30 day in water.
precpiliation reaction. J Dent Res53 420-1425. Des DS (1973) The statistica approach to engineering designs in eramics Proc Br Cerm Soc 22.429-452. De Rijk WJ Ts JA Conner ML (1986) The distribution o failure stresses iv porcelain to meta systems (sact). JDentRes 65217. DiazrArnld AM, Wihiams VD Aquilino SA (1989). Tensile strengths of three luting age for adhesion of fixed partial dentures. ht J Psthdnt 2115-122. Dilts WE, Dicaison MC Jr Miranda FJ Brackett SE (1985). Relative hear bond stregths of outing media with various core materials. J Prosthet Dent 53:505 508. Druvmond JL, Miescke KJ (1991). Weibull models fr the statistical analysis of dentalcomposite data: Aged in physiologic media and cyclicfitigued. Dnt r7*25-29. Ele DD (1961). The distnbution of stress in adhesive joints. In. Eley DD, editor. Adhesio. Oxford Oxford University Press, 207 253. Hotz PM LeanJW Seed I WIsonAD (1977). Theaondingofgiassionomer cements o metal and tooth substrates. Br Dent J 142 41-47. Keevy SM 1:II Sato Y, Tesk JA (1990). Bond strength of resin bonded systems in tension and bending (abstract. J Dent Re 69208. dA HC Duijsters PPE, Davidson CL (1987). learkel Ten Krabbendam Siaer bond strength determinatios on vanous kinds of luting cements h tooth structure and cast allows usian a new testing device. 3 Dent 15.77-81. Laufer B-Z Nicholls J, Towsend D (1988). SiOx C coating: A composite omet lbondngmmechanism. J3Post etDent 60:320-327. Lees WA (984). Design. In: Lees WA, editor Adhesives in engineering design. Berlin: Spnngererlag, 6-12. Lees WA (1985). Stress distribution a bonded joints An exploration within a mthematical niodel Mkterials & Desan> 6 117 123. Lees WA(1989). Design. In: Lees WA editor. Adhesives and the engineer. London Mechanical Engineering Publication, 33-73 McCabe JF Carck TE (1986). A statistical approach to the mechanical testing of dental materials. Dent Mter 2 139 142. McCabe JF Walls AW( 1986). The treatment of results for tensile bond
Acknowledgments We acowledge wihgatitude, Metaux Pre'iex SA Me'talor, Neuchatel Switzerland, for the precious metal the Division of Fixed Prosthodontics and Occlusion, University ofGeieva, Switzerland or the shear-testing device; Dr. H Hawthorne om the National Research Council of Canada, Departmen offribology and Mechanics for providing tie prfilometer and assisting with the analysis; and Dr. M Carius, Vancouver for the cementing jig. REFERENCES
Adams RD (1991). Testirg of adhcsyess eful o rnot Adhrsion 15:1-18. Mien KW (1984). Fundamental aspects of adiheson In.J Le WA, editor. Adhesives in eineerirg desi. Berlin Spriner Verlag 129139. Bain LJU Enelhardt M (1981). Simpe approximate distriutional results ifr confidence and tolerance limits for the WeiblI distribution based on maximum likelihood estimators. Tchrometretr 23:15 20 Barrilay I, Myers ML, Coope LB, Craser CN (1988) Mchaical and chemical rerteion of laboratory cured composite to metal surfics. J Prosthet Dent 59:131 137. Belser UC, LaBar , Bumon M Myer J-M (1990) A new shear store gh test for adhsie prosthodontivs (abstract). J Dnt Res 69:358. Button GL, Barnes RF, Moon PC (1985). Surface preparation and shear bond strength of the casting eemrt nterface. J Piothet Dent 53:34-38. Cohen AC (1965). Maximum likeliood station i the Weilmll dintibution chased on coi plete and on censored samples. hcirhometrs 75739
strength testing JDet 14165-168. McKinneydE, Antonucci JM, Rupp NW (1987). Wear and microhardness of 588. glass-ionmer cements. JDent Res 66:1134-1139. Crisp S, Levis BGC Wilson AD (1976) Characteriza orf si4nmrer McLean JW (1977), A new method of bonding dental cements and porcelain cements. 1 Long term ardness ad compressive strength. J Dent to metal sun aces. Oper Dent 2:130 142. 4:162-166. McLean JW Wilson AD (1977). The clinical development of the glassCnisp S Pringuer MA, Wardleworth D, Wison Al (1974). Reacts in ionoir cevients.I. Formulations and properties. AustDentJ 223136. glass-ionomer cements. 1I. An infred spet3roscopic study. JDentR Men denha1L W, Scaeffr R, Wackerly D (1986). Simultaneous confidence 53.1414 1419. intervals fo more than one parameter. In: Mathematical statistics with Crisp 8, Wilson AD (1974). Reactions in glass-ionomei cements II. The application. Boston: Duxury Press, 559 560. Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
Vol. 71 No. 9
EARLY BOND STRENGTH OF LUTING CEMENTS
Mojon P, Hawbolt EB, MacEntee MI, Belser UC (1989). Maximum bond strength of dental luting cement to amalgam alloy. JDent Res 68:15451549. Moser JB, Brown DB, Greener EH (1974). Short-term bond strengths between adhesive cements and dental alloys. JDentRes 53:1377-1386. Mount GJ, MAkinson OF (1982). Glass-ionomer restorative cements: Clinical implications of the setting reaction. Oper Dent 7:134-141. Naegeli DG, Duke ES, Schwartz R, Norling BK (1988). Adhesive bond of composites to a casting alloy. J Prosthet Dent 60:279-283. Nakabayashi N, Kojima K, Masuhara E (1982). The promotion ofadhesion by the infiltration of monomers into tooth substrates. J Biomed Mater Res 16:265-273. Nelson W (1982). Maximum likelihood analysis of multiple censored data. In: Applied life data analysis. New York: Wiley, 343. 0ilo G, Evje DM(1988). Abendtestformeasuringcement-dentinbond. Dent Mater 4:98-102. Phillips RW (1973). Dental cements for luting and thermal insulation. In: Phillips RW, editor. Skinner's science of dental materials. 7th ed. Philadelphia: W.B. Saunders Company, 466-497.
PratiC,NucciC,MontanariG(1991). Shearbondstrengthandmicroleakage of dentin bonding systems. J Prosthet Dent 65:401-407. Price RB, SutowEJ (1988). Micrographic and profilometric evaluation ofthe finish produced by diamond and tungsten carbide finishing burs on enamel and dentin. JProsthet Dent 60:311-316. Prosser HJ, Powis DR, Brant P, Wilson AD (1984). Characterization ofglassionomer cements. 7. The physical properties ofcurrent materials. JDent
1639
12:231-240. Sorensen JA, Dixit NV (1991). In vitro shear bond strength of dentin adhesives. Int JProsthodont 4:117-125. Tagami J, Tao L, Pashley DH (1990). Correlation among dentin depth, permeability, and bond strength ofadhesive resins. Dent Mater 6:45-50. Takeyama M, Kashibuchi N, Nakabayashi N, Masuhara E (1978). Studies on dental self-curing resin-adhesion of PMMA to bovine enamel or dental alloys. J Jpn Soc Dent Appar Mater 19:179-185. Tanaka T, Atsuta M, Nakabayashi N, Masuhara E (1988a). Surface treatment of gold alloys for adhesion. J Prosthet Dent 60:271-279. Tanaka T, Hirano M, Kawahara M, Matsumura H, Atsuta M (1988b). Anew ion-coating surface treatment ofalloys for dental adhesive resins. JDent Res 67:1376-1380. Tanaka T, Nagata K, Takeyama M, Atsuta M, Nakabayashi N, Masuhara E (1981). 4-META opaque resin-a new resin strongly adhesive to nickelchromium alloy. J Dent Res 60:1697-1706. Van Noort R, Noroozi S, Howard IC, Cardew G (1989). A critique of bond strength measurements. J Dent 17:61-67. Watanabe F, Powers JM, Lorey RE (1988). In vitro bonding of prosthodontic adhesives to dental alloys. J Dent Res 67:479-483. Weibull W (1951). A statistical distribution of wide applicability. JAppl Mechan 18:293-297. Yamashita A (1983). New method of adhesion. J Jpn Dent Assoc 35:10741087. Yu X-Y, XU J-W (1987). The tensile bond strength of various composite resins to alloy. Quint Int 18:145-147.
Downloaded from jdr.sagepub.com at Australian National University on March 13, 2015 For personal use only. No other uses without permission.
