Ronald P. Lubovich, D.D.S., M.S.,* Nashville, Tenn., and Minneapolis,
and Richard Minn.
eramic-metal restorations combine the beauty of porcelain and the strength of a metal substructure. A modified dental gold alloy has been used clinically as a ceramicmetal substructure for two decades with success. However, during the past few years, the technologic parameters of dental ceramic alloys have incurred sweeping changes. Precious metal/ceramic alloys are now being challenged by manufacturers of numerous nonprecious alloys who claim superior physical properties for their products. Nonprecious nickel-chromium alloys present: (1) a density that is one half that of gold alloy, (2) a modulus of elasticity that is 2 to 2$/s times that of the gold alloy, (3) a sag resistance that is 9 times greater, (4) a yield strength that is approximately 20,000 p.s.i. greater, (5) elongation values of between 3 and 22 per cent, and (6) a cost that is one fifth that of the gold alloy. I-5 Nonprecious alloy substructures ( 1) may be finished thinner than precious alloys, (2) provide greater resistance to deformation, and therefore, (3) provide increased strength to the ceramic-metal restorations. Even though nonprecious alloys have superior mechanical properties, there appears to be an investigative controversy about the bond strength of nonprecious alloy systems. Moffa and associates1 determined that the bond shear strength of two nonprecious alloys fused to Ceramco porcelaint was between 13,500$ and 139009 p.s.i., which was comparable to the bond of a gold-ceramic system. McClearF demonstrated that nickel and chromium oxide decreased the thermal coefficient of expansion of Read before the Midwest Academy Condensed from a thesis presented of Master of Science at the University *Assistant Professor, Nashville, Tenn.
**Professor, Director of Dentistry, Minneapolis, Minn.
of Prosthodontics, Minneapolis, Minn. in partial fulfillment of the requirements for the degree of Minnesota, School of Dentistry, Minneapolis, Minn.
Prosthodontics, Graduate City,
J. F. Jelenko & Company, Inc., New Ultratek Metals for Modern Dentistry,
N. Y. Rochelle, N. Y. Inc., Danville,
of Dentistry, School
Fig. 1. Dimensions of test samples.
Table I. Alloys Alloy name S.M.G. III S.M.G. W Nobil-Ceram Permabond Victory
tested Classification Precious Semiprecious Nonprecious Nonprecious Nonprecious
Manufacturer The J. M. Ney Company, Hartford, Conn. The J. M. Ney Company, Hartford, Conn. Nobillium Company, Chicago, Ill. Permabond Company, Walnut Creek, Calif. Unitek Corp. Monrovia, Calif.
Vita porcelain* and suggested that this may induce stresses that could cause failure of nonprecious ceramic-metal restorations. The preceding controversy provided the stimulus and guidance for this investigation. The objectives were (1) to determine whether the bond shear strength for a representative sample of nonprecious alloys was comparable to that of the traditional gold-based alloys, (2) to determine the bond strength and compatibility of Vita and Ceramco porcelains with precious, semiprecious, and nonprecious alloys, and (3) to evaluate the manufacturers’ recommended techniques for surface preparation of metals prior to porcelain application.
MATERIALS AND METHODS Five alloys designed for ceramic-metal restorations were investigated : one precious alloy, one semiprecious alloy, and three nonprecious alloys (Table I). The two porcelains investigated were Ceramco and Vita. The method for testing was patterned after cylindrical pull-through tests performed by Shell and Nielson’, 8 and modified by Anthony and Burnett.9 The alloy test sample consisted of a cast cylinder which measured 0.080 inch in diameter by 3 inches in length. A porcelain donut of 0.070 to 0.090 inch thickness was fused 0.5 inch from one end of the alloy sample (Fig. 1) . The test specimen was lubricated superior and inferior to the porcelain donut with silicone grease. The end of the test sample containing the porcelain donut was then embedded within a gypsum stone cylinder measuring 1.5 inches in diameter by 1.25 inches in height. A Perma-Flex Black Tuffy moldt with an 0.082 inch metal *Unitek Corp., Monrovia, Calif. +Perma-Flex Mold Company, Columbus, Ohio.
j. Prosthet. March.
Fig. 2. Testing
sleeve held each sample at the correct height for pouring the gypsum stone cylinder. One-hundred test alloy samples were cast and prepared for porcelain according to the technique recommended by each manufacturer. Ceramco and Vita porcelain donuts of 0.070 to 0.090 inch thickness were fused to 50 test samples. Each sample was fired three times: ( 1) opaque porcelain application, (2) body porcelain application, and (3) glazing. An Instron testing machine” was used to load each sample in tension until failure occurred. The completed test sample, contained within the gypsum stone cylinder, was mounted in a steel sling suspended between the superior and inferior elements of the tensile testing device (Fig. 2). An Instron X head speed of 0.02 inch per minute was selected. The applied load in tension was converted to a shearing force at the donut-alloy interface via the cylindrical pull-through test design. A load deflection stress-strain curve resulted. The load at failure was used to calculate the pounds per square inch necessary to shear the porcelain-metal bond. Scanning electron microscopy was used to evaluate the surface topography of each alloy after the manufacturer’s porcelain preparation procedures were performed. The electron microscopet was also employed to examine nonprecious alloy test specimens *Instron tS.E.M. England.
*Ten samples failure.
III W N-C, N-C, N-C, were
A lloy S.M.G. III S.M.G. W Nobil-Ceram Permabond Victory bond
‘Ten samples failure.
N* 8 10 10 IO 9 per
tThe adjusted means represent formed via the analysis of variance
S.M.G..III S.M.G. W Nob&Ceram Permabond Victory bond
III W N-C, N-C, N-G
7 9 10 10 10
+The adjusted means represent formed via the analysis of variance
of high and low bond strengths alloys’ microstructural constituents
Adjusted Xt (p.s.i.)
for Vita X
the results of a statistical missing N values.
13,041 17,165 8,582 17,335 14,366 occurred
porcelain s (p2.i.)
Adjusted Xf (p.s.i.)
1,376 621 2,798 1,989 2.306
16,085 17,159 7,588 15,245 11.452
15,847 17,249 7,588 15,245 11.452 In
1,486 1,624 1,777 2,136 794
the results of a statistical for missing N values.
13,053 17,165 8,582 17,335 14,882 In
after shear failure. The chemical analysis of the was performed by the use of x-ray spectrograph.*
RESULTS The raw data demonstrated a wide range of bond strengths. Values as low as 5,149 p.s.i. were recorded for Nobil-Ceram (N-C,), whereas values as high as 20,182 p.s.i. were recorded for Permabond (N-C2) (Tables II and III). Tables II and III present the mean alloy-bond shear strengths for Ceramco and Vita porcelains. Each mean represents values obtained from 10 test replications. An analysis of variance was performed and demonstrated that there was a significant difference in bond strength among the five alloys tested. Significance was reflected by a probability of this occurring of less than 0.0005. A test was performed to determine the difference between Ceramco and Vita porcelains. The resulting P value was greater than 0.10, which indicated that there was no significant difference between Ceramco and Vita porcelains with regard to over-all bond strength to the five alloys tested. There was, however, a ceramic-alloy interaction. Vita porcelain produced greater bond strengths with the traditional gold-based alloy than Ceramco porcelain (P < 0.001). A cross“Kevex
1. I’rosthet. March.
Solid bar = Ceramco Lined bar = Vita
III Fig. 3. Bar
over of greater bond strengths was then observed with Ceramco porcelain when applied to nonprecious alloys (P < 0.005). A n over-all estimate of this trend may be seen in Fig. 3. During application of porcelain to alloy samples, an unexpected result occurred. Seven of the Vita porcelain donuts incurred check-line fractures. The check lines did not result in separation of the two halves but were visible in the body porcelain mass. No fracture lines occurred in Ceramco porcelain test specimen. A Fisher exact statistical analysis was used to determine the significance of this unexpected result. The probability was determined to be less than 0.0125 and termed “significant.” An electron probe analysis was performed on five alloys identical to those subjected to the bond strength tests. Scan lines were recorded for the characteristic wavelengths emitted and are recorded in Table IV. Scanning electron micrographs at 2,000x of alloys identical to those tested are represented in Figs. 4 to 8. Individual alloys were prepared per the manufacturer’s instructions and examined for surface texture. Alloys W and N-C, demonstrated the greatest degree of surface roughness, depicted by craters and sharp angular concavities. This roughness was created by a 50 p aluminous oxide/air abrasive. Alloy III demonstrated the smoothest surface, as would be expected of rubber wheel preparation. N-C, presented an angular, parallel coarse, rough surface characteristic of the heatless-stone preparation. N-C, depicted a moderate roughness in comparison to N-C, ; it was first abraded by 50 ,.L aluminous oxide/air and then surface prepared with a medium-grit brown stone. Figs. 9 and 10 represent topographic, interfacial bond area, surface roughness micrographs of previously tested samples of N-C, and N-C,. Alloy N-C, presented the least bond strength, and N-Ce presented the greatest bond strength. When the speci-
Fig. 4. Alloy III-prepared by rubber wheel. (Scanning electron micrograph. Original magnification X2,000.) Fig. 5. Alloy W-50 JI aluminous oxide. (Scanning electron micrograph. Original magnification x2,000.) Table
Niobium Palladium Silver Tin Iron I Gold
Aluminum Rubidium Silver Tin Iron Gold Palladium
Aluminum Nickel Zirconium Chromium Iron
Aluminum Nickel Zirconium Chromium Iron
Aluminum Nickel Zirconium Chromium Iron
mens were examined microscopically, no remaining porcelain particles were readily visible between the inferior and superior borders of the porcelain bond site. Upon magnification at 1,000x, alloy N-C, exhibited a moderate degree of roughness with porcelain particles infrequently noted on the scanned surface. By contrast, scanning electron micrographs of alloy N-C, presented numerous minute porcelain particles scattered throughout the surface craters and concavities. DISCUSSION The success of ceramic-metal restorations depends, to a certain extent, upon the adherence of porcelain to the ceramic alloy. Four theories of enamel adherence have been proposed.7, lo They are: ( 1) mechanical, (2) wetting (van der Waals) , (3) chemical bonding, and (4) compressive bonding forces. Various researchers have attempted to relate ceramic-metal bonding to potential mechanical forces which actually interlock porcelain to the metal alloy.11-‘5 Research electron micrographs demonstrated the degree of roughness as a result of various methods of pre-porcelain treatment recommended by manufacturers (Figs. 4 to 8). Of those methods of pre-porcelain surface preparation used in this study, a 50 p aluminous oxide/air abrasive produced maximum sample roughness.* This was the *Interstate
Dental, Buffalo, N. Y.
Fig. 6. Alloy magnification
7. Alloy X2,000.)
8. Alloy X2,000.)
Fig. 9. Alloy magnification
method of preparation recommended for alloys W and N-C,. These two alloys consistently outperformed all others in measured and calculated bond shear strengths, exhibiting values of approximately 17,000 p. s.i. This might relate to the research of Lavine and CusteP who concluded that the roughening of gold castings before the addition of porcelain resulted in 13 to 15 per cent greater bond strengths as compared to nonroughened castings. A contrasting method of preparation was recommended for alloy III. The manufacturer’s technique consisted of polishing the surface with a white Dedeco Flexie” (Fig. 4). Determined bond shear strengths for the polished alloy III averaged 13,000 to 16,000 p.s.i. depending upon the ceramic system tested. This observation correlated with the research of King and associate?’ *Dedeco
Fig. 10. Alloy magnification
which stated that surface roughness would not alter or replace the chemical conditions required for good bonding in enamels. The polished surface and relatively high bond shear strength of alloy III do not, however, preclude the contributory effect of surface roughening. The question might be posed, “What effect would roughening the surface of alloy III have upon the resultant bond strength?” Numerous researchers believe the ceramic-metal bond to be a direct result of oxide formation at the ceramic-metal interface .I* 7* 8, 18eZ2 The bonds that hold porcelain and metals together were reported to be a direct result of electron attraction. The bonds were divided into ionic, covalent, and metallic groups and result from a sharing or transfer of electrons at the ceramic-metal interface.23 Oxide formation was found to be increasingly visible after repeated firings of the test alloys. The alloy oxide that was least observable clinically was found on the N-C, surface, which may account for its relatively poor bond strength. Alloy N-C, produced mean bond strengths of 8,500 and 7,500 p.s.i. (Tables II and III). The chemical components of this alloy were found to be similar to those of N-C&, whereas the N-C2 produced bond strengths that were two times greater than those of N-C,, or 17,000 p.s.i. Another noted difference between N-C1 and N-C, was that of surface preparation. N-C, was prepared with a Mizzy heatless stone* and degassed, and porcelain was applied.3 N-C, was degassed and then abraded with 50 p aluminous oxide/air to remove the oxide and create surface roughness .4 The scanning electron micrograph demonstrated that the method for alloy N-C, produced greater surface roughness (compare Figs. 6 and 7). Although most authoritieP suggest that surface roughness alone could not contribute 8,000 p.s.i. to the calculated bond strength of alloy N-Cl, Gilman indicated that roughness might increase fracture resistance by a factor of 2. Further, more-complete quantitative chemical analysis might suggest reasons for the relative lack of oxide formation and lower bond strength values of alloy N-C,.
More recently, Bergerl” and Vickery and Bandinellilo suggested that the major ceramic-metal bond attachment was derived from compressive forces. Crane specified that this was the proposed theory of Vita porcelain formulation.” This corresponded to research results of Nallyz5 who demonstrated that Vita porcelain incurred two times the residual contraction of Ceramco porcelain. In comparing the preceding results with the present study, it was noted that no craze lines appeared on the Ceramco samples; whereas, seven craze lines were noted on the samples of Vita fused to nonprecious alloys. It might be hypothesized that the greater rigidity of modulus of the nonprecious alloys prevented this compressive phenomenon from occurring, thus promoting fractures in samples. However, the check lines did not produce a significant difference in bond strength of the two porcelains, according to statistical analysis. The analysis of variance suggested that alloys had a ceramic preference which was pointed to by the different bond strengths observed. Differential tests” performed for Ceramco porcelain indicated that there was no difference between alloys W and N-C,, with means of 17,000 p.s.i. No difference was noted between alloys III and with means of 13,000 and 14,000 p.s.i., respectively. N-C, was found to have a N-C,, significantly lower bond strength of 8,000 p.s.i., deficient by 5,000 p.s.i. when compared to the traditional gold-based alloy. In general, the Vita alloys formed three groups. The statistical comparison again placed N-C, at the 8,000 p.s.i. level. Surprisingly, N-C, alloy, which was specifically formulated for Vita porcelain, performed better with Ceramco porcelain. The trend of superior performance by Ceramco porcelain in comparison to Vita porcelain when applied to nonprecious alloys is depicted by a bar graph of representative bond strengths (Fig. 3). The bar depicts the crossover of a superior bond from Vita with alloy III (precious) to Ceramco with nonprecious alloys. McClean” suggested that although nonprecious alloys may produce good bond strengths, this phenomenon may be due to the residual stresses produced at the bond by the porcelain’s decreased thermal coefficient of expansion. Chromium oxide and nickel oxide were found to decrease the thermal coefficient of expansion of Vita porcelain. Most of McClean’s research was performed with Vita porcelain. The effect of chromium and nickel oxide upon the thermal coefficient of Ceramco has not been reported in the literature. Perhaps the residual stresses reported by McClean” have an effect on Vita porcelain and not Ceramco, or perhaps the residual stresses are responsible for the Vita fracture lines that appeared in the nonprecious alloy samples in this investigation. In comparing Ceramco and Vita porcelains, it appears that Ceramco porcelain’s thermal coefficient of expansion and physical characteristics are more compatible to the nonprecious alloys (Fig. 3 ) . The accurate measurement of ceramic-metal bond strengths has promoted numerous test designs and research instruments. Careful evaluation revealed that no particular test was totally free of inherent error due to the complexity of the ceramicmetal bond. The cylindrical pull-through test of Shell and Nielson,T, R modified by Anthony and Burnett,” was used in this investigation for the following reasons: (I) *Tukey’s
the failure at the ceramic-metal bond interface could be repeatedly reproduced, and (2) this investigation’s results could be compared to other research,‘, 2, 6-s thus verifying the reliability of the testing method. Previously reported research test values for the traditional gold-ceramic alloys with Ceramco porcelain compared well with the means for bond strength observed in this study. The following researchers employed cylindrical pull-through instruments and tested the bond shear strengths of traditional gold-ceramic alloys with Ceramco porcelain: Shell and Nielson73 8 reported a bond strength range of 10,000 to 13,000 p.s.i., Moffa and associate+ 26 reported a mean of 10,600 p.s.i., and Anthony and Burnett” reported a mean of 13,800 p.s.i. The test alloy of Anthony and Burnett was identical of alloy III (S.M.G. III), and their results compared favorably with this investigation-13,800 to 13,041 p.s.i. It was felt that the closely approximated results somewhat verified the reliability of the cylindrical pull-through test. McCleans suggested that the cylindrical pull-through test design incorporated residual thermal mismatch compressive forces similar to those in a shrink-fit metal collar. Since substructure design research2” 28 indicated that increased porcelain coverage caused increased fracture resistance, it was felt that the wraparound donut would create clinical conditions somewhat similar to those when porcelain covered most of a crown surface. Vickery and Bandinellii” suggested that compressive forces were inclusive and part of the ceramic-metal bond complex. Removal of these stresses from a test piece could possibly remove a portion of the contributing ceramic-metal bond complex. Assuming that the cylindrical pull-through test is reliable, it might be hypothesized that the bond strengths of the nonprecious alloys in this investigation represent accurate estimates of each alloy’s bond strength potential. When examining the Ceramco porcelain results (N-C, bond strength, 8,000 p.s.i.; N-C2, 17,000 p.s.i.; and N-C&, 14,000 p.s.i., Table II), it becomes apparent that individual nonprecious alloys may produce bond strengths smaller than or in excess of those of traditional goldbased alloys. The practical significance of this study suggests that at this time, the dentist who plans to use nonprecious ceramic-metal restorations needs to be cautious in the selection of a ceramic-alloy system. He should not allow his patients to become the testing environment for the injudicious selection of nonprecious alloys. Further, objective research on nonprecious alloys must be continued and broadened in order that progress and improvement are not fraught with trial and needless failure. SUMMARY
One-hundred cylindrical pull-through ceramic-metal test specimens were subjected to shear loading forces. Two porcelain systems were tested in association with five ceramic alloys-one precious, one semiprecious, and three nonprecious alloys. 1. There was no significant difference in bond strength between Ceramco and Vita porcelains. 2. The various alloys tested demonstrated significantly different bond strengths. 3. Differential statistical tests suggested that nonprecious alloys performed better with Ceramco porcelain than with Vita porcelain. 4. The bond strength of Ceramco porcelain fused to nonprecious alloy N-C2
was significantly greater than that of Ceramic and traditional gold-based alloys. 5. Nonprecious alloy N-C, produced significantly less bond strength than the traditional gold-based alloys. 6. Semiprecious alloy W produced high bond strengths with both Ceramco and Vita porcelains. 7. The cylindrical pull-through test was a reliable, reproducible method of testing ceramic-metal bond shear strength. 8. The surface roughness appears to be the one common factor in nonprecious alloys that relates to large differences in bond strength. Additional tests must be made to verify this hypothesis. The authors thank Kathleen Keenan, Ph.D., Associate Professor, Department of Human Oral Genetics, University of Minnesota, School of Dentistry, for her assistance in the statistical interpretation of the data; William Gerberich, Ph.D., Associate Professor, Department University of Minnesota; and all manuof Chemical Engineering and Materials Sciences, facturers and representatives who contributed their time and materials for the execution of this investigation-The J. M. Ney Company, Nobillium Company, Permabond Company, and Unitek Corp. and
References 1. Moffa, J., Lugassy, A. A., Guckes, A. D., and Gettleman, L.: An Evaluation of Nonprecious Alloys for Use With Porcelain Veneers. Part I. Physical Properties, J. PROSTNET. DENT. 30: 424-431, 1973. 2. Ney Company: Physical Properties Chart, Bloomfield, Conn. 3. Nobillium Company: Physical Properties Chart, Chicago, Ill. 4. Permabond Company: Physical Properties Chart, Walnut Creek, Calif. 5. Unitek Corp.: Interoffice Memorandum, Monrovia, Calif., Oct. 16, 1974. 6. McClean, J. W.: The Art and Science of Dental Ceramics, Louisiana State University, School of Dentistry, Continuation Education Program, New Orleans, La. 7. Shell, J. S., and Nielson, J. P.: Study of the Bond Between Gold Alloys and Porcelain, Bull. South. Calif. State Dent. Lab. Sot., Aug.-Sept., 1963, pp. l-15. 8. Shell, J. S., and Nielson, J. P.: Study of the Bond Between Gold Alloys and Porcelain, J. Dent. Res. 41: 1424-1437, 1962. 9. Anthony, D. H., and Burnett, D. L.: Shear Test for Measuring Bonding in Cast GoldAlloy-Porcelain Systems, J. Dent. Res. 49: 27-33, 1970. of Attachment Forces in Porcelain Gold 10. Vickery, R. C., and Bandinelli, L. A. : Nature Systems, J. Dent. Res. 47: 683-689, 1968. 11. Borom, M. P., and Pask, J.: Role of Adherence Oxides in the Development of Chemical Bonding at Glass Metal Interfaces, J. Am. Ceram. Sot. 49: 1-6, 1966. 12. Custer, F., and Coyle, T.: Techniques Influencing Strength of the Porcelain Fused to Metal Restorations, Part I, N. Y. Dent. J. 39: 118-123, 1969. 13. Moore, D. G., Pitts, J. C., Richmond, J. C., and Harrison, W. N.: Galvanic Corrosion Theory for Adherence of Porcelain Enamel Ground Coats to Steel, J. Am. Ceram. Sot. 37: 1-6, 1954. 14. Rosenberg, J. E.: Outline of Interesting Enameling Problems, Ceram. Abst. 6: 330, 1937. 15. Staley, H. J.: Electrolytic Reactions in Vitreous Enamels and Their Relation to Adherence of Enamels to Steel, J. Am. Ceram. Sot. 17: 163-167, 1934. 16. Lavine, M. H., and Custer, R.: Variables Affecting the Strength of Bond Between Porcelain and Gold, J. Dent. Res. 45: 32-36, 1966. 17. King, B. W., Tripp, H. P., and Duckworth,, W. H.: Nature of Adherence of Porcelain Enamels to Metals, J. Am. Ceram. Sot. 42: 504-525, 1959. 18. Alner, D. J.: Aspects of Adhesion, London, 1965, University of London Press, pp. 1 l-45.
Volume Number 19. 20. 21. 22. 23. 24. 25. 26.
Berger, WH.: Structural Adhesives, Allentown, Pa., 1958, Structural Adhesives Associates, pp. 1-6. J. F. Jelenko & Company: Questions and Answers About Jel-Span, New Rochelle, N. Y. Kelly, M., and Asgar, K.: Tensile Strength Determination of the Interface Between Porcelain Fused to Gold, J. Biomed. Mater. Res. 3: 403-406, 1969. Parker, R. S., and Taylor, P.: Adhesion and Adhesives, Oxford, 1966, Pergamon Press, Inc. Ryge, G.: Current American Research on Porcelain Fused to Metal Restorations, Int. Dent. J. 15: 385-392, 1965. Gilman, J. J.: Fracture, New York, 1959, John Wiley & Sons, Inc. Nally, J. N.: Chemical-Physical Analysis and Mechanical Tests of the Ceramo-Metallic Complex, Int. Dent. J. 18: 309-325, 1968. Moffa, J. P., Guckes, A. D., Okawa, M. T., and Lilly, G. E.: An Evaluation of Nonprecious Alloys for Use With Porcelain Veneers. Part II. Industrial Safety and Biocompatibility, J. PROSTHET. DENT. 30: 424-431, 1973. Kuhlmann, W. H.: The Effects of Crown Design Upon Fracture Resistance of Fused Porcelain Veneers, Unpublished Masters Thesis, University of Minnesota, June, 1972. Warpeha, W.: Design and Technique Variables Affecting Published Masters Thesis, University of Minnesota, 1973. DR. LUBOVICH 6008 BETHANY BLVD. NASHVILLE, TENN. 37221 DR. GOODKIND 4123 SUNSET MIN’NEAPOLIS,
cast technique Todo,
A study of the variability of setting a fully adjustable articulator to a pantographic tracing R. Bruce
A clinical John
of the maxillectomy
Selective resection of the circumoral mandibular denture stability
resin and Joseph
of fractured D.M.D.