LANG AND STARR
2. MacCuIlochWT.Advancesindentalceramics.BrDentJ1968;124:361-5. 3. Grossman DC. Processing a dental ceramic by casting methods. Ceramic engineering and science proceeding. Am Ceram Sot 1985;6:1940. 4. Grossman DG. Cast glass ceramics. Dent Clin North Am 1985;29:72537. 5. Pameijer CH, Grossman DG, Adair PJ. Physical properties of a castable ceramic dental restorative material [abstract]. J Dent Res 1980;59:474. 6. Adair PJ, Grossman DG. The castable ceramic crown. Int J Periodont Restor Dent 1984;4:32-46. 7. Cave1WT, Kelsey WT III, Barkmeier WW, Blankenau RJ. A pilot study of the clinical evaluation of castable ceramic inlays and a dual-cure resin cement. Quintessence Int 1988;19:257-62. 8. Grossman DG, Walters HV. The chemical durability of dental ceramics [abstract]. J Dent Res 1984;63:234. 9. Mueninghoff LA, Neal SJ, Ramus DL. Six months’ evaluation of clinical esthetic veneers (Dicer). J Dent Res 1988;67:305. 10. Anonymous. Dicer clinical instructions for Dicer restorations and use of the Dicer light-activated cementation kit. York, Pa: Dentsply International Inc, 198&l-13. 11. Savitt E, Socransky S, Melter A, Malament K, Backman H. Effects of oral microbiota by Dicer glass ceramic. Boston: Forsythe Institute, 1986. 12. Garber DA, Goldstein RE, Feinman RA. Porcelain laminate veneers. 1st ed. Chicago: Quintessence Publishing, 198814-23. 13. Jordan RE. Esthetic composite bonding. 1st ed Revised. Toronto: BC Decker, 1988122-39.
fit of castable
14. Judge SF. Dicer castable ceramic veneers. Dent Lab Prac 1987;12:12-5. 15. Bennett RJ, Baily LF. Bonding to Dicer laminate veneers [abstract]. J Dent Res 1986;65:314. 16. Reid JS. Tooth color modification and porcelain veneers. Quintessence Int 1988;19:477-81. 17. Christensen GJ. Veneering of teeth-state of tbe art. Dent Clin North Am 1985;29:373-90. 18. Covey DA, Oliveira FC Jr, Denehy GE. Selecting an esthetic veneering technique. Quintessence Int 1987;18:247-52. 19. Nathanson D. Etched porcelain restorations for improved esthetics. Part I. Anterior veneers. Compend Contin Educ Dent 1986;7:706-12. 20. Goldstein RE. Diagnostic dilemma: to bond, laminate, or crown? Int J Periodont Restor Dent 1987;5:9-29. 21. Farah JW, Powers JM, eds. All ceramic restorations. Dent Advisor 1989;6:1-3. 22. Woolsey G, Weir D, Dootz E, Morris H. Marginal accuracy of Dicer, Cerestore, porcelain-butt, and cast metal margins [abstract]. J Dent Res 1987;66:284. 23. Jones DW, Sutow EJ, Rizkalla AS, Black D. Opacity and colour of a castable glass-ceramic and cement system [abstract]. J Dent Res 1988;67:118. Reprint requests to: DR. CLIFFORD B. STARR 4062 QUAIL BUSH DRIVE DAYTON, OH 45424
J. Robert Holmes, DDS, MS, MEd,a William D. Sulik, DDS, MS,b Gene A. Holland, DDS, MS,” and Stephen C. Bayne, MS, PhDd Medical University of South Carolina, College of Dental Medicine, Charleston, SC., and University of North Carolina, School of Dentistry, Chapel Hill, N.C. The objective of this study was to measure marginal fit of castable ceramic versus gold crowns. Full veneer gold and ceramic crowns were made on Ivorine dies. Crowns were cemented, embedded, sectioned faciolingually and mesiodistally, and photographed for measurement of absolute marginal discrepancies (cavosurface angle to casting margin) to evaluate fit. Results revealed no statistically significant differences in fit among four locations around the margins of either ceramic or gold crowns. There was no statistically significant difference in the combined absolute marginal discrepancy (fit) between ceramic and gold crowns. The variance of the combined absolute marginal discrepancy (fit) of the ceramic crowns was significantly different than that of the gold crowns (p = 0.01 level). The standard deviation (variability) of the gold crowns was more than twice that of the ceramic crowns, Randomized block ANOVA demonstrated statistically significant differences among individual gold crown specimens, but none among individual ceramic crown specimens. (J PROSTHET DENT 1992;67:594-9.)
Supported in part by NIH-NIDR grant No. 2-S07-RR05333. ‘Assistant Professor, Department of Crown and Bridge Dentistry, Medical University of South Carolina. College of Dental Medicine. %linical Associate Professor (part-time), Department of Prosthodontics, University of North Carolina, School of Dentistry. cProfessor, Department of Prosthodontics, University of North Carolina, School of Dentistry. dAssociate Professor, Department of Operative Dentistry, University of North Carolina, School of Dentistry. 19/l/35864 I
mce the introduction of the porcelain jacket crown by Land in the late 188Os,there have been continual efforts to improve all-ceramic restorations. The esthetic advantages inherent in eliminating the metal substructure of ceramometal crowns is well d0cumented.i Castable ceramic (Dicer, Dentsply Int., York, Pa.) crowns, made by the lost wax technique, are a recent development of the all-ceramic crown concept. These benefits, however, must be evaluated against the possible disadvantage of a poorer “fit.”
MAY 1992 VOLUME67
Fig. 1. Typical faciolingual and embedded specimen.
section through cemented
Clinical acceptability of the Dicer castable ceramic system suggests that the castability and marginal fit of the restorations are excellent. However, when marginal discrepancies have been noted by these investigators, the misfit most often seems to be associated with proximal margins of teeth that exhibit a large degree of rise and fall of the cavosurface angle in an occlusogingival direction because of the asymmetry of the tooth preparation. The objectives of this study were to (1) compare the relative marginal fit of castable ceramic crowns versus type III gold crowns, and (2) compare the marginal fit of castable ceramic crowns at various locations around the margin of an asymmetric full veneer crown preparation. MATERIAL
Ten castable ceramic (Dicer), and 10 type III gold (Firmilay, Pennwalt/Jelenko, Armonk, N.Y.) full veneer crowns were made on identical Ivorine dies (Dentsply Int.). A maxillary central incisor was selected because it typically has more rise and fall (occlusogingivally) in the cementoenamel junction than other teeth, and a typical crown preparation would also reflect this same rise and fall. Although castable ceramic crowns are used on anterior teeth most often (and full veneer metal castings rarely), use of porcelain-metal crowns for comparison would have introduced too many additional variables. Differences in wax thickness of coping and ceramic crown patterns and the potential change in fit at the margin during porcelain application would have jeopardized any meaningful comparison. A posterior tooth, on which metal crowns are more commonly used, was not selected because it would not have allowed testing of differences around the margins of a tooth with as severe a marginal rise and fall. A special waxing mold was used to ensure that all wax patterns were identical. The marginal design was a heavy chamfer for both the ceramic and gold crowns. Two coats of die spacer (Dicer die spacer, Dentsply Int.) were placed to within 1 mm of the margins on all dies. Wax patterns for
Fig. 2. Measurement of rounded margins.
ceramic crowns were invested in phospbate-bonded investment (Dicer castable ceramic investment, Dentsply Int.) and wax patterns for gold crowns were invested in gypsum bonded investment (Super-Span investment, Pennwalt/ Jelenko). Both investments were mixed according to manufacturer’s specifications and were thermally expanded only. Each wax pattern was burned out in a two-step procedure. The first step involved heating to 250“ C with a hold time of 30 minutes. The second step involved heating to 954’ C for ceramic crowns and to 704O C for gold crowns, with a hold time of 30 minutes. The Dicer ceramic crowns were processed according to the manufacturer’s directions. The glass was cast at 1360° C using a motorized centrifugal casting machine (Dicer casting machine, Dentsply Int.) and then bench cooled. The castings were sandblasted with 25 urn aluminum oxide particles at 40 psi, cleaned in an ultrasonic cleaner in distilled water for 5 minutes, and embedded in a special gypsum bonded investment material (Dicer castable ceramic embedment, Dentsply Int.) for ceramming. During ceramming the temperature was raised to 107P C over a period of 90 minutes, held for 6 hours, then cooled to 200’ C over a period of 60 minutes. The eerammed crowns were removed from the embedment, sandblasted as before, and cleaned ultrasonically. Any nodules on the inside of the castings were removed with an abrasive stone. The gold crowns were processed according to manufacturer’s directions. The alloy was melted with a gas and air torch and cast in a broken-arm centrifugal casting machine (Kerr Centrifico casting machine, Sybron/Kerr, Romulus, Mich.). Castings were quenched and the investment removed. The castings were sandblasted with aluminum ox-
Fig. 3. Typical castable ceramic (Dicer) crown margin. Arrows indicate measurement points.
Fig. 4. Typical type III gold (Firmilay) Arrows indicate measurement points.
diamond wafering saw (slow-speed Isomet saw, Buehler Ltd., Lake Bluff, Ill.) in both the faciolingual and the mesiodistal direction at the midpoint of each surface. The sections were polished sequentially with 240,320,400, and 600 grit silicon carbide sandpapers (Carbimet Paper, Buehler Ltd.). An example of a sectioned specimen is shown in Fig. 1. The polished specimens were examined with a light microscope and photographed at X250 magnification for measurement. The measurement points for each specimen were the cavosurface angle for each preparation and the casting margin for each crown. When a margin appeared rounded, the point chosen on that margin was along a line bisecting the angle between the main contours of the die or casting. The determination of that measurement point is illustrated in Fig. 2. A variety of possible choices for the measurement of marginal errors have been documented and defined.2 The linear distance from the cavosurface angle of the preparation to the margin of the restoration is defined as the absolute marginal discrepancy and was chosen for this study. The absolute marginal discrepancy always represents the maximum measurement of misfit at the margin. Measurements of the absolute marginal discrepancy were made at the middle of the facial, lingual, mesial, and distal surfaces for both ceramic and gold crowns. Differences among those reference locations were analyzed for statistical significance at the p = 0.01 level by a randomized block ANOVA. The combined mean of all four locations for each specimen was also calculated. To compare the combined means for the two materials, an unpaired t-test was used. An F-test for variance showed that the Welch approximation was required to determine the degrees of freedom for the t-test.
RESULTS ide, pickled, and cleaned in an ultrasonic cleaner in distilled water for 5 minutes. Any nodules on the internal surfaces of the castings were removed with an abrasive stone. The gold crowns and Dicer crowns were compared in the as-cast state. Finishing, polishing, or burnishing of gold margins was deliberately not done. Application of shading porcelain to the Dicor crowns was also not done. Obviously, those steps might contribute to the final casting fit clinically. However, those steps would have obscured the identification of casting effects on the overall misfit. The as-cast crowns were cemented onto their respective dies after removal of the die spacer. An unfilled resin (Comspan, L. D. Caulk Co., Milford, Del.) was used to lute the crowns to the dies and minimize hydraulic resistance during cementation. After initial crown placement, a constant 25 kg seating pressure was maintained for 10 minutes. The cemented crowns were then embedded in epoxy resin (Araldite 502/Hardener 956, Stuart Pharmaceuticals, Wilmington, Del.). Each crown was sectioned with a 596
The values for the measurements of absolute marginal discrepancy at the midpoints of the facial, lingual, mesial, and distal surfaces for both ceramic and gold crowns are reported in Table I. The means and standard deviations for individual locations and for the combinations of all locations are indicated. Examples of cross sections of the cemented margins for ceramic and gold specimens are presented in Figs. 3 and 4. There was no consistent pattern of overextension or underextension of the margins. As demonstrated in Figs. 3 and 4, approximately 20 % of the specimens showed irregular projections at the casting margins that may have arisen during waxing. No statistically significant difference was found at the p 0.01 level among any of the four individual locations on the crown for either ceramic or gold. No statistically significant difference was found at the p = 0.01 level between the combined means for ceramic and gold crowns. As shown in Table I, larger standard deviations were associated with the gold crown measurements, An F-test was used to compare
Table I. Absolute marginal discrepancies (pm)
Ceramic crowns Gold crowns
51 t 14 43 27 11
54 +- 11
51 + 16 73 F 33
35 + 13 61 k 28
51 & 28
48 k 7$ 57 i 19$
“No statistically significant difference at p = 0.01 level from randomized block ANOVA for comparison of F, L, M, D groups. tNo statistically significant difference at p = 0.01 level from t-test for comparison of combined data from ceramic and gold crowns. IF-test for variance indicates statistically significant difference at p = 0.01 level for combined standard deviations.
the variability of the two materials. The greater variability of the absolute marginal discrepancy for the gold crowns was statistically significant at the p = 0.01 level. DISCUSSIQN Previous studies involving various materials and techniques have resulted in a tremendous range of values for measurement of fit. Measurements have been reported from as small as 3.63 pm to greater than 100 ym, with the midrange 50 to 60 pm. 3,* The values obtained in the present study fall well within the range of values reported by others. Fit of Dicer crowns (48 ,um) determined by this study compared favorably with other studies of all-ceramic restorations, including the mean marginal gap openings of 28 grn reported by the manufacturer for Dic0r.s However, comparisons between the present and previous studies must be made cautiously. The values reported often depend more on study design and measurement method than on the materials tested. Another important consideration of fit measurements is the standard deviation associated with mean values for fit. Investigators do not often emphasize the variability that occurs or the significance of this variability. The midrange of standard deviations reported in the literature is approximately 15 to 20 pm. The standard deviations reported in this study were within this range. An interesting result in the present study was a standard deviation for type III gold crowns approximately twice that of castable ceramic crowns. A similar unexpected result noted in a study by Eden et al.‘j was the nonuniformity of casting size for type III and gold-substitute alloys. The results of the present study seem to support the observation made by Eden et a1.6 Since the standard deviation is a measure of variability, it can be concluded that fit of the castable ceramic crowns was more consistent, that is, the distribution of the dispersion was not as wide for the ceramic crowns. Even though the mean absolute marginal discrepancy between ceramic and gold crowns was not statistically different, a more consistent distribution of fit was achieved with the ceramic crowns. This might be a clinical advantage for ceramic crowns with regard to the film thickness of cement. The marginal fit of gold crowns in the upper end of the range would be more likely to lie outside the limit of acceptability. Those in the lower end of the range might not actually benefit from the better fit because of additional misfit introduced by the film thickness of the cement.
A type III gold alloy was selected for comparison in this study because of general agreement that, if properly made, the fit obtained with this material is clinically acceptable. This does not mean that one should be satisfied with this level of fit, and clearly better techniques and materials should be sought. However, if castable ceramic crowns fit at least as well as type III gold crowns, they would not be contraindicated because of lack of fit, and their esthetic potential could be capitalized on without concern. The attainable limits of marginal fit theoretically should result in zero discrepancy at the cavosurface margin of the preparation. From a practical standpoint, a perfect result is not possible. Christensen7 demonstrated the difficulty in evaluating fit clinically, particularly with regard to gingival margins. He computed a value of 39 Mm as the maximum acceptable marginal opening in visually accessible surfaces of a casting, but concluded that a value for gingival margins would be greater because they were more difficult to assess. Most experienced clinicians would be satisfied with discrepancies in marginal fit of 50 pm or less and probably deem a fit of 100 ym clinically acceptable in some cases. McLean and von Fraunhofe? in a &year clinical study of 1000 restorations concluded that 120 pm was the maximum acceptable marginal opening. Even if discrepancies of less than 10 pm (as obtained by Byrne3) were routinely possible, the film thickness of the cement would not allow castings to be fully seated. Numerous studies have documented the problem of fully seating castings because of the hydraulic effects of cements. Various techniques (vents, flow channels, and die relief) have been advocated to minimize this problem. The ADA Specification No. 8 for zinc phosphate cement recommends a film thickness of 25 ;um, but this degree of seating of castings is probably rarely achieved clinically. Therefore, the practical range for clinical acceptability of fit seemsto be approximately 50 to 100 pm. Choices of measurement location for fit in most studies seem fairly arbitrary. Differences in fit at various locations around the margin are seldom considered. Some studies measure locations other than the midpomt of a surface (for example, line angles), and little mention is ever made of differences at the various locations. One of the results of this study is that differences in fit at various locations around the margin are not significantly different. It was observed clinically before this study that marginal discrepancy with Dicor crowns most often ocmrred in the proxi-
Table II. Calculation of expected marginal error Ceramic cr0wn.s
AT = solidification temp. - room temp = 566- 25 = 541° C cy= coeff. of thermal expansion = 7.2 x [email protected]
/” C W = width of casting at margin = 7 mm Setting expansion =