Effect of digerent metal ceramic opaque and dentin porcelain Bruce J. Crispin, D.D.S., M.S.,* Robert and Harold Globe***
alloys
on the color
of
R. Seghi, D.D.S., M.S.,**
University of California, School of Dentistry, Los Angeles, Calif. Certain metal ceramic alloys alter the final color of bonded porcelain. Contradicting results, inconsistent test designs, and incomplete alloy selections have led to some confusion. In this study, differential calorimetric analysis was done at the opaque and dentin porcelain stages with five major types of metal ceramic alloys. The color stability of the porcelain on the high-noble metal alloys was found to be excellent. The palladium-silver and nickel-chromium alloys resulted in significant color changes of the dentin porcelain only. The greatest changes in color were found with the palIadiu:m-silver alloy, which resulted in a higher yellow-green saturation. The nickel-chromium alloy also produced a color change, although not as severe, resulting in a porcelain shade with a reduced Value or lightness. (J PROSTHET DENT 1991;65:361-6.)
E
xcellence in esthetic dentistry requires an understanding of the variables in the metal ceramic processthat affect the final perceived color.l-s A number of alloys are available for useasmetal ceramic substructures and some controversy exists asto their effect on the final shadeof the porcelain. Barghi and Richardson2studied the effect of repeated firing on thle resulting color of porcelain applied to various high-gold-containing ceramic alloys and found no difference visually. In 1982, Barghi3 studied the effect of repeated firing on the perceived color of four porcelains baked five high-gold ceramicalloys and one nickel-chrome alloy. He concludedthat repeated firings did not noticeably affect the visually perceivable shadeof the porcelains for the first five glaze cycles. In a preliminary visual study by Crispin et al.,8 actual crowns were made with nickel-chromium and gold-palladium substructuresand comparedvisually for differences in perceived value. One porcelain shadewasused (Vita VMK-68 shadeAl, Vident, Baldwin Park, Calif.) and :dgnificant differences were noted. Brewer et a1.6used spectrophotometric analysis techniques to compare the influence of three different alloy substrateson the resulting color of metal ceramic restorations with one sh.adeof porcelain. Reflectance measurements were made at each of seven fabrication steps to evaluate color changes.The results of this study indicated that although little calorimetric changeoccurred with the
Supportedin part b’yUniversity of Californiaat LosAngelesAcademicSenateGrant No. 4090. *AssociateProfessor,Departmentof Fixed Prosthodontics. **AssistantProfessor,Departmentof OperativeDentistry. ***ResearchAssociate. 10/1/20619
THE
JOURNAL
OF PROSTHETIC
DENTISTRY
three alloys through the opaque porcelain addition steps, significant color changesoccurred after the dentin porcelain addition steps.The resulting calorimetric values of the porcelain baked on the palladium-silver alloy differed significantly from those of the porcelain baked on both the high-gold and the nickel-chromium alloys, which were very similar. In 1987,Jacobset al7 used spectrophotometric and visual assessments of Hue, Value, and Chroma to study the effects of several variables on the resulting color of metal ceramic restorations. In this study, three alloys (gold-platinum-palladium, nickel-chromium, and high-palladium), three porcelain shades(Vita VMK-68, Bl, A3, C4, Vident) and three dentin porcelain thicknesses(0.5,1, and 1.5mm) wereusedasvariables. The spectrophotometric assessment indicated that for certain shades,the nickel-chromium and high-palladium alloys resulted in significantly different Hue valuesthan the gold-platinum-palladium alloy group. Visual assessments of thesedifferences,however, indicated that the nickel-chromium alloy group was significantly different than both the high-palladium and the gold-platinum-palladium alloy groups.All detected differenceswere most obviouswhen thinner layers of dentin porcelain were used. Although it appearsthat various alloys can affect the resulting color of dental porcelain shades,somedisagreement continuesamongresearchersasto which alloys producethe most clinically significant changes.Differencesin porcelain thicknessesand the techniques of analysisusedin the various studiesmake it difficult to compare results and draw clinically significant conclusionsfrom the data. This study (1) evaluated and comparedthe effects of five commonly used alloy substructures on the resulting shade of metal ceramic restorations and (2) establisheda clinically relevant test model that can be used for future testing.
351
CRISPIN,
SEGHI,AND
GLOBE
Fig. 1. Metal disks in as-cast condition show flat porcelain-bearing surfaces and reinforcing supports on opposite side. Fig. 2. Dial calipers ptaced on reference point for measurement of porcelain thickness. Fig. 3. Metal test disks with opaque porcelain only. Fig. 4. Minolta CR100 Chroma Meter instrument.
MATERIAL AND METHODS Metal substructures Five metal ceramic alloys representing the most common alloy types used for metal ceramic restorations were tested (Table I). Ten disks were made from each alloy. Disks approximately 11 mm in diameter were cast by use of a lost wax technique and following the manufacturer’s instructions (Fig. 1). All of the alloys were cast by u8e of a Harris oxygen-gas torch with the exception of the nickel-chromium alloy, which wa8 cast in an induction casting machine (ECM III, Howmedica Inc., Chicago, Ill.). The disks were designed with reinforcing supports on the side not bearing porcelain to aid in handling, to minimize warping, and to allow positioning in a special device used to facilitate application of dentin porcelain. The flat porcelain-bearing surface on each disk was smoothed with mounted stones (Shofu, Menlo Park, Calif.), air-abraded (Mark III, Belle de St. Claire, Chatsworth, Calif.) with 25 pm alumina oxide, and cleaned with steam
352
prior to oxidation. Each metal was handled according to the manufacturer’8 instructions before the addition of opaque porcelain (Table I). Four reference point8 were placed on each disk on the side not bearing porcelain to aid in the assessment of the porcelain thickness (Fig. 2). Measurements were made to the nearest 0.05 mm. The thickness of the test disks was measured and recorded at each of the four recording point8 before the addition of porcelain. This measurement wa8 ueed to determine the thickness of the subsequent porcelain additions.
Porcelain
application
Vita VMK-66 paint-on 88 opaque shade Al (No. 501, Vident) porcelain was applied to each of the test disks with a brush-on application technique. Three layers of opaque were applied. The first layer wa8 applied as a thin slurry, the second a8 a normal paint-on layer, followed by a final correction layer to achieve uniform coverage and the
MARCH
1991
VOLUME
65
NUMBER
3
EFFECT
OF ALLOYS
Table
I. Alloy product information
ID
ON PORCELAIN
Alloy
type
COLOR
Product
name
Composition
(%)
Hi-Au
High-gold
Williams Y-l
84 Au,7 Pt,6 Pd
Au-Pd
Gold-palladium
Galaxy
51.5 Au,31.9 Pd,7.5 In
Oxidation
Manufacturer
Williams Co. Inc., Buffalo, N.Y. APM Sterngold, Huntington Beach, Calif.
Ni-Cr
Nickel-chromium
Rexillium III
78 Ni,14 Cr, 6 MO, 2 Be
Jeneric Ind. Inc, Wallingford, Conn.
Pd-Ag
Palladiumsilver
Williams W-l
Williams Co. Inc.
Hi-Pd
High-palladium
Option
53.5 Pd, 37.5 Ag, 8.5 Sn 79 Pd,2 Au
Table
Air fire at 1850’ F for 5 min
Air fire 1200’-1900’ F hold 1 min Place in hydrofluoric acid 2 min l Sandblast with aluminous oxide (75 - 80 psi) l Clean ultrasonically in alcohol for 5 min l Rinse in distilled HxO l Sandblast (80 psi) aluminous oxide l Clean ultrasonically in distilled Hz0 0 Fire 1200’ - 1825O F under vacuum l No coating agents, slow cool porcelain l Vacuum fire to 1850’ F, hold 5 min l No coating agents l Air fire to 1900“ F and hold 5 min l Sandblast with aluminous oxide l l
II. Summary of porcelain firing cycles 1st Opaque (slurry bake) 2nd Opaque bake 3rd Opaque (correction bake) 1st Body bake 2nd Body bake Final :glaxe
desired range of thickness. The average opaque thickness for the five alloys ranged from 0.26 to 0.29 mm. All layers were fired from 6510’C to 950” C under vacuum in an oven (Ultra-Mat II, Unitec, Monrovia, Calif.). No porcelain grinding was necessaryfor these steps (Fig. 3). Five of the 10 disks in each alloy group received Vita VMK-68 shadeAl dentin porcelain (No. 540, Vident). A cylindrical
holding
device was used to hold the disks and
contain the porcelain powders. The disks were overbuilt and ground down to their final measurementsby use of Busch Silent stones(Pfingst Inc., South Plainfield, N.J.1. The first two dentin applications were fired from 650’ C to 910” C under vacuum. The final glazing cycle wasfired from 880’ C to 920° C with no vacuum and a l-minute hold time at 920° C. The averagethicknessof the dentin porcelain for each of the five alloys ranged from 0.74 to 0.79 mm. The remaining five opaque-only test disks in each alloy group were subjectedto the sameadditional thermal cyclesasthe dentin-covered disks. Table II summarizesthe thermal cycles of the test disks.
Reference
porcelain
disks
To facilitate calorimetric data analysis and interpretation, colorimetrio reference standardswere madefrom the porcelain powdersby use of the techniques describedpreviously by Seghi et aI9 An opaque disk approximately 14
TEE
J.M. Ney Co., Bloomfield, Conn.
l
instructions
JOURNAL
OF PROSTHETIC
DENTISTRY
650” 650” 650° 650° 650’ 880”
- 980’ - 980” - 980” - 910a - 910’ - 920”
C C C C C C
(1202’ (1202’ (1202’ (1202’ (1202“ (1616’
-
1796” 1796’ 1796O 1670° 1670’ 1688’
F) F) F) F) F) F)
under under under under under 1-min
vacuum vacuum vacuum vacuum vacuum hold; no vat
mm in diameter and 5 mm thick was used as the opaque reference.The layered referencesamplewasestablishedby optically connecting a 0.75mm thick dentin porcelain disk to the correspondingopaque disk with 1.5 index of refraction fluid (R. P. Cargille Laboratories Inc., Cedar Grove, N.J.). The reference standards were consideredto represent the ideal color of the porcelains.
Color measuring
procedures
A Minolta CR 100ChromaMeter (Minolta Inc., Ramsey, N.J.) instrument wasusedfor the calorimetric analysisof the samples
(Fig.
4). This
instrument
provides
color-
difference data on opaque and translucent dental porcelains that correlate well with referencespectrophotometric data.1° The
instrument
was calibrated
according
to the
manufacturer’s instructions by use of the supplied white working tile. Each samplewasmeasuredduring three separate recording sessions.Each sessionconsistedof the averageof three measurements. The instrument calibration was verified after five sample measurementsand recalibrated
if necessary.
The reference porcelain disks were measured during eachrecording session.Calorimetric recordingswere made on all of the disks after the third addition of opaque porcelain. Additional measurementswere madeon the opaque and layered test disks after the final glazing cycle. All mea-
353
CRISPIN,
EFFECTS
OF SUBSTRUCTURE
+.----a-*
ON PORCELAIN
COLOR
I t------1 -------I II
Alloy
Substructure
Type
Fig. 6. A, Summary of mean and standard deviation of AE values obtained for each alloy group. B, Final test disks after addition of dentin porcelain.
surements were recorded in CIELAB coordinates relative to standard light source D65 and the 1931 standard observer functions (20)i1 and were transferred to a personal computer (Macintosh II, Apple Computer, Inc., Cupertino, Calif.) for analysis. The CIELAB color difference formulall was used to calculate the color differences (AE) in the evaluation. Color differences between the simulated metal ceramic crown samples and their corresponding porcelain reference standards were calculated from the measured data. The AE values were statistically evaluated by analysis of variance and the Tukey test. Differences between means at p < 0.05 were considered statistically significant.
RESULTS Fig. 5 summarizes the effect of the alloy substructure on the resulting color of the porcelain at the opaque and dentin-layered stages of metal ceramic crown fabrication. The means and standard deviations of the resulting calculated AE values for each alloy group are reported. The AE determined between the opaque-covered alloy disks subject to three firing cycles and the reference opaque disk are the values designated as AEos and those subject to six firing 354
SEGHI,
AND
GLOBE
cycles are designated as AEoe. The mean AE determined for the opaque and dentin layered alloy disks relative to the corresponding layered reference disk produced the values designated as AEns. The AE values obtained from the samples made with the high-gold-containing alloy were used as the control group for each fabrication step assessed. Multiple comparisons were made relative to the control values and significant differences (*) are indicated. No statistically significant differences were found between the control group and the experimental groups after the opaque-only disks were subjected to either three or six thermal cycles. The addition of the dentin porcelain layer onto the opaque-covered disks did show significantly greater AE values for the palladium-silver and nickelchromium groups when compared with the control group. The palladium-silver alloy group resulted in a mean AE of approximately five units greater than the control group whereas the nickel-chromium group produced a relative difference of just more than one unit greater than the control group. Within each alloy group, the opaque-only samples that were subjected to six firing cycles resulted in greater AE values than similar samples subjected to fewer cycles. However, when similarly fired opaque samples were compared with the corresponding high-gold-containing control group, no significant differences were found. Fig. 6 is a graphic representation of the measured CIELAB data obtained on the final dentin porcelain layered samples. Each point represents the mean measurements of the five samples within each alloy group. Graphic representations of the data are essential for the visual interpretation of directional information related to the color changes occurring between the groups. The circles represent areas in the color space that are within one color-difference unit away from the control group. The points that lie outside this area are considered visually significantly different than the control group. Fig. 6, A represents a planar projection of the a*b* axis of the color space and allows for the assessment of the chromatic differences that exist between the samples. Only the palladium-silver group was found to have mean chromatic changes greater than one unit from the control group. The plots indicate that the porcelain backed by this metal resulted in a shade that was much more saturated with yellow or yellow-green than the control group. Fig. 6, B represents a planar projection of the L*b* axis and allows for graphic interpretation of achromatic changes in color. Changes occurring along the L* axis only are considered in the assessment of achromatic or lightness changes. Both the palladium-silver and nickel-chromium alloys show a decrease in lightness from the control group that was greater than one color difference unit.
DISCUSSION The high-gold-containing alloys have been considered for many years as the standard of the industry with respect to color-fidelity production of metal ceramic restorations.
MARCH
1991
VOLUME
65
NUMBER
3
EFFECT
OF ALLOYS
ON PORCELAIN
COLOR
The relatively low measured values for AE between the reference porcelain disks and the samples made with an alloy of this type would support this consideration. The calorimetric assessment of the gold-palladium and highpalladium alloy groups in this study has shown that these noble-metal alloys Iperformed as well as the high-gold control group. This finding supports previous visual and spectrophotometric evaluations that found no significant difference between porcelain color produced from samples backed by a high-gold type alloy and a high-palladium type alloy.7 This analysis has shown the palladium-silver alloy system to produce the most significant AE relative to the other alloys tested. The data are in agreement with previous studies that report.ed this effect through changes in measured reflectance spectra and tristimulus values6 The relative difference in the Al3 values obtained between the control group and the palladium-silver group was more than 5 AE units and would be most likely to be perceived by even the most indiscriminate eye. Although the numeric data are extremely useful in assessing the overall significance of the perceived color problem, they do not provide useful directional information. Graphic analysis of the measured CIELAB values can provide this additional visually related information, which may be applied to the assessment or management of color-related problems. The palladium-silver alloys have been reported by dentists and technicians to cause a yellow or yellow-green discoloration of the porcelain. l2 Graphic analysis of the colorimetric data obtained in this study verifies these observations. The yellow c:oloration is assumed to be caused by the absorption of silver into the body porcelain, according to work reported by :Doremus, l3 who showed that at elevated temperatures in glass, silver atoms can move diffusely and cluster into colloidal agglomerates that can import a yellow color to the glass. The exact mechanism by which the silver ions are transported from the alloy to the porcelain is not yet well known, but mechanisms have been proposed.14 New porcelain products that are claimed to minimize this effect have been developed. Evaluations of these materials and techniques are under way. The nickel-chromium alloy systems are among the most popular base metal alloys in use today. The AE data indicate that these alloy types can result in significantly greater color changes than the high-gold control group. Although other visual and instrumental studies have reported the existence of such color changes resulting from the use of nickel-chromium alloys, conclusions based on the clinical significance of these differences have varied.6-8 This variability is most likely a result of the relatively small color changes these alloys produce. Our results have shown the relative difference between the nickel-chromium alloy and the control group to be approximately one unit for the Al shade evaluated. Although this magnitude of difference is in the range of detectability by a trained dental observer under ideal viewing conditions,15 it may be well within an
THE
JOURNAL
OF PROSTHETIC
DENTISTRY
21 20 19 16 17 16 15 14 13 12 11
Red (+a’)
uA
RELATIVE
CHROMATIC
SHIFTS
79
79 77 76 75 74
, I
73
I
I
I
_
72 -. 71 I
I
I
70"""' 70
0B’
69 I 69,, 11
1 , 12
RELATIVE
, 13
(,( 14
, 15
, 16
ACHROMATIC
, 17
, ,fl 16
Yellow Ih*i
19
'"'
SHIFTS
Fig. 6. Mean measured color data for each alloy group in graphic form. Circle represents area within one AE unit of control group. Groups found within this area are not considered visually significant. A, a*b axis; B, L*b* axis.
acceptable range for normal viewing conditions, Further work is needed to evaluate the acceptability limits that can be tolerated. Graphic analysis indicated that the chromatic changes resulting from the nickel-chromium alloy were minimal and that the measured color differences recorded were primarily a result of a decrease in the L* value or lightness of the resulting shade. This analysis is in agreement with other studies that have reported the resulting color changes in terms of Munsell Value changes and
CRISPIN,
decreased luminance reflectance factor, both of which correlate closely with the perceived lightness of an object. Nickel oxide is a colorant that has been shown to produce a neutral grey color in sodium silicate g1asses14 and is most likely responsible for the color changes caused by these alloys.
CLINICAL
IMPLICATIONS
The results of this study suggest that the type of alloy substructure used in the fabrication of metal ceramic restorations can significantly affect the resulting shade. The selection of shade or porcelain material may need to be modified to accommodate these color changes, particularly when palladium-silver and nickel-chromium based alloys are used.
CONCLUSIONS Five different alloys used as porcelain substructures were analyxed to determine the effect they have on the resulting color of opaque and dentin porcelain. Differential colorimetry was used to quantify these effects. The findings were as follows. 1. The high-noble metal-containing alloys showed the best performance with respect to color production with no significant differences found among the three alloy groups. 2. The palladium-silver and nickel-chromium alloy groups caused significantly greater color changes relative to the high-gold-containing control at the body porcelain stage. 3. No significant color changes were noted with alloys for the opaque-only groups. 4. The results of this calorimetric evaluation technique were found to be in good agreement with results obtained by other investigators using both the visual and the spectrophotometric techniques.
356
SEGHI,
AND
GLOBE
REFERENCES 1. Barghi
N, Goldberg J. Porcelain shade stability after repeated firings. J PROSTHS~ DENT 1977;37:173-5. 2. Barghi N, Richardson JT. A study of various factors influencing shade of bond porcelain. J PROSTHET DENT 1978;39:282-4. 3. Barghi N. Color and glaze: effects of repeated firings. J PROSTHET DENT 1982;47:393-5.
4. Jorgenson MW, Goodkind RJ. Spectrophotometric study of five porcelain shades relative to the dimensions of color, porcelain thickness, and repeated firings. J PROWHET DENT 1979;42:96-105. 5. Obregon A, Goodkind RJ, Schwabacher WB. Effects of opaque and porcelain texture on the color of ceramometal restorations. J PROSTHET DENT
1981;46:330-40.
6. Brewer JD, Akers CK, Garlapo DA, Sorensen SE. Spectrometric analysis of the influence of metal substrates on the color of metal-ceramic restorations. J Dent Res 1985;64:74-7. 7. Jacobs SH, Goodacre CJ, Moore BK, Dykema RW. Effect of porcelain thickness and type of metal-ceramic aboy on color. J PROSTHET DENT 1987;57:138-45.
8. Crispin BJ, Okamoto SK, and Globe H. Porcelain crown substructures effect on visually perceivable Value. J Prosthet Dent 1989 (in-press). 9. Seghi RR, Johnston WM, O’Brien WJ. Spectrophotometric analysis of color differences between porcelain systems. J PROSTHET DENT 1986$&35-40.
10. Seghi RR, Johnston WM, O’Brien WJ. Performance assessment of colorimetric devices on dental porcelains. J Dent Res 198%68:1755-g. 11. CIE Publication No. 15.2 Calorimetry. Central Bureau of the CIE, Vienna, 1986. 12. Tuccillo JJ, Cascone PJ. The evolution of porcelain-fused-to metal alloy systems in dental ceramics. In McLean JW, ed. Proceedings of the first international symposium on ceramics. Chicago: Quintessence Pub1 Co Inc, 1983;347-70. 13. Doremus RH. Optical properties of small silver particles. J Chem Phys 1965;42:414-7. 14. Fanderlik I. Glass science and technology 5: optical properties of glass. New York: Elsevier Science Pub1 Co, Inc, 1983;320. 15. Segbi RR, Hewlett ER, Kim J. Visual assessment of calorimetric analysis on translucent dental porcelain. J Dent Res 1989;68:1766-4. Reprint
requests to:
DR. BRUCE J. CRISPIN SCHOOL OF DENTISTRY UNIVERSITY OF CALIFORNIA 10833 LE CONTE AVE. Los ANGELES, CA 90024
MARCH
lSS1
VOLUME
65
NUMBER
S
alloys
on the color
of
R. Seghi, D.D.S., M.S.,**
University of California, School of Dentistry, Los Angeles, Calif. Certain metal ceramic alloys alter the final color of bonded porcelain. Contradicting results, inconsistent test designs, and incomplete alloy selections have led to some confusion. In this study, differential calorimetric analysis was done at the opaque and dentin porcelain stages with five major types of metal ceramic alloys. The color stability of the porcelain on the high-noble metal alloys was found to be excellent. The palladium-silver and nickel-chromium alloys resulted in significant color changes of the dentin porcelain only. The greatest changes in color were found with the palIadiu:m-silver alloy, which resulted in a higher yellow-green saturation. The nickel-chromium alloy also produced a color change, although not as severe, resulting in a porcelain shade with a reduced Value or lightness. (J PROSTHET DENT 1991;65:361-6.)
E
xcellence in esthetic dentistry requires an understanding of the variables in the metal ceramic processthat affect the final perceived color.l-s A number of alloys are available for useasmetal ceramic substructures and some controversy exists asto their effect on the final shadeof the porcelain. Barghi and Richardson2studied the effect of repeated firing on thle resulting color of porcelain applied to various high-gold-containing ceramic alloys and found no difference visually. In 1982, Barghi3 studied the effect of repeated firing on the perceived color of four porcelains baked five high-gold ceramicalloys and one nickel-chrome alloy. He concludedthat repeated firings did not noticeably affect the visually perceivable shadeof the porcelains for the first five glaze cycles. In a preliminary visual study by Crispin et al.,8 actual crowns were made with nickel-chromium and gold-palladium substructuresand comparedvisually for differences in perceived value. One porcelain shadewasused (Vita VMK-68 shadeAl, Vident, Baldwin Park, Calif.) and :dgnificant differences were noted. Brewer et a1.6used spectrophotometric analysis techniques to compare the influence of three different alloy substrateson the resulting color of metal ceramic restorations with one sh.adeof porcelain. Reflectance measurements were made at each of seven fabrication steps to evaluate color changes.The results of this study indicated that although little calorimetric changeoccurred with the
Supportedin part b’yUniversity of Californiaat LosAngelesAcademicSenateGrant No. 4090. *AssociateProfessor,Departmentof Fixed Prosthodontics. **AssistantProfessor,Departmentof OperativeDentistry. ***ResearchAssociate. 10/1/20619
THE
JOURNAL
OF PROSTHETIC
DENTISTRY
three alloys through the opaque porcelain addition steps, significant color changesoccurred after the dentin porcelain addition steps.The resulting calorimetric values of the porcelain baked on the palladium-silver alloy differed significantly from those of the porcelain baked on both the high-gold and the nickel-chromium alloys, which were very similar. In 1987,Jacobset al7 used spectrophotometric and visual assessments of Hue, Value, and Chroma to study the effects of several variables on the resulting color of metal ceramic restorations. In this study, three alloys (gold-platinum-palladium, nickel-chromium, and high-palladium), three porcelain shades(Vita VMK-68, Bl, A3, C4, Vident) and three dentin porcelain thicknesses(0.5,1, and 1.5mm) wereusedasvariables. The spectrophotometric assessment indicated that for certain shades,the nickel-chromium and high-palladium alloys resulted in significantly different Hue valuesthan the gold-platinum-palladium alloy group. Visual assessments of thesedifferences,however, indicated that the nickel-chromium alloy group was significantly different than both the high-palladium and the gold-platinum-palladium alloy groups.All detected differenceswere most obviouswhen thinner layers of dentin porcelain were used. Although it appearsthat various alloys can affect the resulting color of dental porcelain shades,somedisagreement continuesamongresearchersasto which alloys producethe most clinically significant changes.Differencesin porcelain thicknessesand the techniques of analysisusedin the various studiesmake it difficult to compare results and draw clinically significant conclusionsfrom the data. This study (1) evaluated and comparedthe effects of five commonly used alloy substructures on the resulting shade of metal ceramic restorations and (2) establisheda clinically relevant test model that can be used for future testing.
351
CRISPIN,
SEGHI,AND
GLOBE
Fig. 1. Metal disks in as-cast condition show flat porcelain-bearing surfaces and reinforcing supports on opposite side. Fig. 2. Dial calipers ptaced on reference point for measurement of porcelain thickness. Fig. 3. Metal test disks with opaque porcelain only. Fig. 4. Minolta CR100 Chroma Meter instrument.
MATERIAL AND METHODS Metal substructures Five metal ceramic alloys representing the most common alloy types used for metal ceramic restorations were tested (Table I). Ten disks were made from each alloy. Disks approximately 11 mm in diameter were cast by use of a lost wax technique and following the manufacturer’s instructions (Fig. 1). All of the alloys were cast by u8e of a Harris oxygen-gas torch with the exception of the nickel-chromium alloy, which wa8 cast in an induction casting machine (ECM III, Howmedica Inc., Chicago, Ill.). The disks were designed with reinforcing supports on the side not bearing porcelain to aid in handling, to minimize warping, and to allow positioning in a special device used to facilitate application of dentin porcelain. The flat porcelain-bearing surface on each disk was smoothed with mounted stones (Shofu, Menlo Park, Calif.), air-abraded (Mark III, Belle de St. Claire, Chatsworth, Calif.) with 25 pm alumina oxide, and cleaned with steam
352
prior to oxidation. Each metal was handled according to the manufacturer’8 instructions before the addition of opaque porcelain (Table I). Four reference point8 were placed on each disk on the side not bearing porcelain to aid in the assessment of the porcelain thickness (Fig. 2). Measurements were made to the nearest 0.05 mm. The thickness of the test disks was measured and recorded at each of the four recording point8 before the addition of porcelain. This measurement wa8 ueed to determine the thickness of the subsequent porcelain additions.
Porcelain
application
Vita VMK-66 paint-on 88 opaque shade Al (No. 501, Vident) porcelain was applied to each of the test disks with a brush-on application technique. Three layers of opaque were applied. The first layer wa8 applied as a thin slurry, the second a8 a normal paint-on layer, followed by a final correction layer to achieve uniform coverage and the
MARCH
1991
VOLUME
65
NUMBER
3
EFFECT
OF ALLOYS
Table
I. Alloy product information
ID
ON PORCELAIN
Alloy
type
COLOR
Product
name
Composition
(%)
Hi-Au
High-gold
Williams Y-l
84 Au,7 Pt,6 Pd
Au-Pd
Gold-palladium
Galaxy
51.5 Au,31.9 Pd,7.5 In
Oxidation
Manufacturer
Williams Co. Inc., Buffalo, N.Y. APM Sterngold, Huntington Beach, Calif.
Ni-Cr
Nickel-chromium
Rexillium III
78 Ni,14 Cr, 6 MO, 2 Be
Jeneric Ind. Inc, Wallingford, Conn.
Pd-Ag
Palladiumsilver
Williams W-l
Williams Co. Inc.
Hi-Pd
High-palladium
Option
53.5 Pd, 37.5 Ag, 8.5 Sn 79 Pd,2 Au
Table
Air fire at 1850’ F for 5 min
Air fire 1200’-1900’ F hold 1 min Place in hydrofluoric acid 2 min l Sandblast with aluminous oxide (75 - 80 psi) l Clean ultrasonically in alcohol for 5 min l Rinse in distilled HxO l Sandblast (80 psi) aluminous oxide l Clean ultrasonically in distilled Hz0 0 Fire 1200’ - 1825O F under vacuum l No coating agents, slow cool porcelain l Vacuum fire to 1850’ F, hold 5 min l No coating agents l Air fire to 1900“ F and hold 5 min l Sandblast with aluminous oxide l l
II. Summary of porcelain firing cycles 1st Opaque (slurry bake) 2nd Opaque bake 3rd Opaque (correction bake) 1st Body bake 2nd Body bake Final :glaxe
desired range of thickness. The average opaque thickness for the five alloys ranged from 0.26 to 0.29 mm. All layers were fired from 6510’C to 950” C under vacuum in an oven (Ultra-Mat II, Unitec, Monrovia, Calif.). No porcelain grinding was necessaryfor these steps (Fig. 3). Five of the 10 disks in each alloy group received Vita VMK-68 shadeAl dentin porcelain (No. 540, Vident). A cylindrical
holding
device was used to hold the disks and
contain the porcelain powders. The disks were overbuilt and ground down to their final measurementsby use of Busch Silent stones(Pfingst Inc., South Plainfield, N.J.1. The first two dentin applications were fired from 650’ C to 910” C under vacuum. The final glazing cycle wasfired from 880’ C to 920° C with no vacuum and a l-minute hold time at 920° C. The averagethicknessof the dentin porcelain for each of the five alloys ranged from 0.74 to 0.79 mm. The remaining five opaque-only test disks in each alloy group were subjectedto the sameadditional thermal cyclesasthe dentin-covered disks. Table II summarizesthe thermal cycles of the test disks.
Reference
porcelain
disks
To facilitate calorimetric data analysis and interpretation, colorimetrio reference standardswere madefrom the porcelain powdersby use of the techniques describedpreviously by Seghi et aI9 An opaque disk approximately 14
TEE
J.M. Ney Co., Bloomfield, Conn.
l
instructions
JOURNAL
OF PROSTHETIC
DENTISTRY
650” 650” 650° 650° 650’ 880”
- 980’ - 980” - 980” - 910a - 910’ - 920”
C C C C C C
(1202’ (1202’ (1202’ (1202’ (1202“ (1616’
-
1796” 1796’ 1796O 1670° 1670’ 1688’
F) F) F) F) F) F)
under under under under under 1-min
vacuum vacuum vacuum vacuum vacuum hold; no vat
mm in diameter and 5 mm thick was used as the opaque reference.The layered referencesamplewasestablishedby optically connecting a 0.75mm thick dentin porcelain disk to the correspondingopaque disk with 1.5 index of refraction fluid (R. P. Cargille Laboratories Inc., Cedar Grove, N.J.). The reference standards were consideredto represent the ideal color of the porcelains.
Color measuring
procedures
A Minolta CR 100ChromaMeter (Minolta Inc., Ramsey, N.J.) instrument wasusedfor the calorimetric analysisof the samples
(Fig.
4). This
instrument
provides
color-
difference data on opaque and translucent dental porcelains that correlate well with referencespectrophotometric data.1° The
instrument
was calibrated
according
to the
manufacturer’s instructions by use of the supplied white working tile. Each samplewasmeasuredduring three separate recording sessions.Each sessionconsistedof the averageof three measurements. The instrument calibration was verified after five sample measurementsand recalibrated
if necessary.
The reference porcelain disks were measured during eachrecording session.Calorimetric recordingswere made on all of the disks after the third addition of opaque porcelain. Additional measurementswere madeon the opaque and layered test disks after the final glazing cycle. All mea-
353
CRISPIN,
EFFECTS
OF SUBSTRUCTURE
+.----a-*
ON PORCELAIN
COLOR
I t------1 -------I II
Alloy
Substructure
Type
Fig. 6. A, Summary of mean and standard deviation of AE values obtained for each alloy group. B, Final test disks after addition of dentin porcelain.
surements were recorded in CIELAB coordinates relative to standard light source D65 and the 1931 standard observer functions (20)i1 and were transferred to a personal computer (Macintosh II, Apple Computer, Inc., Cupertino, Calif.) for analysis. The CIELAB color difference formulall was used to calculate the color differences (AE) in the evaluation. Color differences between the simulated metal ceramic crown samples and their corresponding porcelain reference standards were calculated from the measured data. The AE values were statistically evaluated by analysis of variance and the Tukey test. Differences between means at p < 0.05 were considered statistically significant.
RESULTS Fig. 5 summarizes the effect of the alloy substructure on the resulting color of the porcelain at the opaque and dentin-layered stages of metal ceramic crown fabrication. The means and standard deviations of the resulting calculated AE values for each alloy group are reported. The AE determined between the opaque-covered alloy disks subject to three firing cycles and the reference opaque disk are the values designated as AEos and those subject to six firing 354
SEGHI,
AND
GLOBE
cycles are designated as AEoe. The mean AE determined for the opaque and dentin layered alloy disks relative to the corresponding layered reference disk produced the values designated as AEns. The AE values obtained from the samples made with the high-gold-containing alloy were used as the control group for each fabrication step assessed. Multiple comparisons were made relative to the control values and significant differences (*) are indicated. No statistically significant differences were found between the control group and the experimental groups after the opaque-only disks were subjected to either three or six thermal cycles. The addition of the dentin porcelain layer onto the opaque-covered disks did show significantly greater AE values for the palladium-silver and nickelchromium groups when compared with the control group. The palladium-silver alloy group resulted in a mean AE of approximately five units greater than the control group whereas the nickel-chromium group produced a relative difference of just more than one unit greater than the control group. Within each alloy group, the opaque-only samples that were subjected to six firing cycles resulted in greater AE values than similar samples subjected to fewer cycles. However, when similarly fired opaque samples were compared with the corresponding high-gold-containing control group, no significant differences were found. Fig. 6 is a graphic representation of the measured CIELAB data obtained on the final dentin porcelain layered samples. Each point represents the mean measurements of the five samples within each alloy group. Graphic representations of the data are essential for the visual interpretation of directional information related to the color changes occurring between the groups. The circles represent areas in the color space that are within one color-difference unit away from the control group. The points that lie outside this area are considered visually significantly different than the control group. Fig. 6, A represents a planar projection of the a*b* axis of the color space and allows for the assessment of the chromatic differences that exist between the samples. Only the palladium-silver group was found to have mean chromatic changes greater than one unit from the control group. The plots indicate that the porcelain backed by this metal resulted in a shade that was much more saturated with yellow or yellow-green than the control group. Fig. 6, B represents a planar projection of the L*b* axis and allows for graphic interpretation of achromatic changes in color. Changes occurring along the L* axis only are considered in the assessment of achromatic or lightness changes. Both the palladium-silver and nickel-chromium alloys show a decrease in lightness from the control group that was greater than one color difference unit.
DISCUSSION The high-gold-containing alloys have been considered for many years as the standard of the industry with respect to color-fidelity production of metal ceramic restorations.
MARCH
1991
VOLUME
65
NUMBER
3
EFFECT
OF ALLOYS
ON PORCELAIN
COLOR
The relatively low measured values for AE between the reference porcelain disks and the samples made with an alloy of this type would support this consideration. The calorimetric assessment of the gold-palladium and highpalladium alloy groups in this study has shown that these noble-metal alloys Iperformed as well as the high-gold control group. This finding supports previous visual and spectrophotometric evaluations that found no significant difference between porcelain color produced from samples backed by a high-gold type alloy and a high-palladium type alloy.7 This analysis has shown the palladium-silver alloy system to produce the most significant AE relative to the other alloys tested. The data are in agreement with previous studies that report.ed this effect through changes in measured reflectance spectra and tristimulus values6 The relative difference in the Al3 values obtained between the control group and the palladium-silver group was more than 5 AE units and would be most likely to be perceived by even the most indiscriminate eye. Although the numeric data are extremely useful in assessing the overall significance of the perceived color problem, they do not provide useful directional information. Graphic analysis of the measured CIELAB values can provide this additional visually related information, which may be applied to the assessment or management of color-related problems. The palladium-silver alloys have been reported by dentists and technicians to cause a yellow or yellow-green discoloration of the porcelain. l2 Graphic analysis of the colorimetric data obtained in this study verifies these observations. The yellow c:oloration is assumed to be caused by the absorption of silver into the body porcelain, according to work reported by :Doremus, l3 who showed that at elevated temperatures in glass, silver atoms can move diffusely and cluster into colloidal agglomerates that can import a yellow color to the glass. The exact mechanism by which the silver ions are transported from the alloy to the porcelain is not yet well known, but mechanisms have been proposed.14 New porcelain products that are claimed to minimize this effect have been developed. Evaluations of these materials and techniques are under way. The nickel-chromium alloy systems are among the most popular base metal alloys in use today. The AE data indicate that these alloy types can result in significantly greater color changes than the high-gold control group. Although other visual and instrumental studies have reported the existence of such color changes resulting from the use of nickel-chromium alloys, conclusions based on the clinical significance of these differences have varied.6-8 This variability is most likely a result of the relatively small color changes these alloys produce. Our results have shown the relative difference between the nickel-chromium alloy and the control group to be approximately one unit for the Al shade evaluated. Although this magnitude of difference is in the range of detectability by a trained dental observer under ideal viewing conditions,15 it may be well within an
THE
JOURNAL
OF PROSTHETIC
DENTISTRY
21 20 19 16 17 16 15 14 13 12 11
Red (+a’)
uA
RELATIVE
CHROMATIC
SHIFTS
79
79 77 76 75 74
, I
73
I
I
I
_
72 -. 71 I
I
I
70"""' 70
0B’
69 I 69,, 11
1 , 12
RELATIVE
, 13
(,( 14
, 15
, 16
ACHROMATIC
, 17
, ,fl 16
Yellow Ih*i
19
'"'
SHIFTS
Fig. 6. Mean measured color data for each alloy group in graphic form. Circle represents area within one AE unit of control group. Groups found within this area are not considered visually significant. A, a*b axis; B, L*b* axis.
acceptable range for normal viewing conditions, Further work is needed to evaluate the acceptability limits that can be tolerated. Graphic analysis indicated that the chromatic changes resulting from the nickel-chromium alloy were minimal and that the measured color differences recorded were primarily a result of a decrease in the L* value or lightness of the resulting shade. This analysis is in agreement with other studies that have reported the resulting color changes in terms of Munsell Value changes and
CRISPIN,
decreased luminance reflectance factor, both of which correlate closely with the perceived lightness of an object. Nickel oxide is a colorant that has been shown to produce a neutral grey color in sodium silicate g1asses14 and is most likely responsible for the color changes caused by these alloys.
CLINICAL
IMPLICATIONS
The results of this study suggest that the type of alloy substructure used in the fabrication of metal ceramic restorations can significantly affect the resulting shade. The selection of shade or porcelain material may need to be modified to accommodate these color changes, particularly when palladium-silver and nickel-chromium based alloys are used.
CONCLUSIONS Five different alloys used as porcelain substructures were analyxed to determine the effect they have on the resulting color of opaque and dentin porcelain. Differential colorimetry was used to quantify these effects. The findings were as follows. 1. The high-noble metal-containing alloys showed the best performance with respect to color production with no significant differences found among the three alloy groups. 2. The palladium-silver and nickel-chromium alloy groups caused significantly greater color changes relative to the high-gold-containing control at the body porcelain stage. 3. No significant color changes were noted with alloys for the opaque-only groups. 4. The results of this calorimetric evaluation technique were found to be in good agreement with results obtained by other investigators using both the visual and the spectrophotometric techniques.
356
SEGHI,
AND
GLOBE
REFERENCES 1. Barghi
N, Goldberg J. Porcelain shade stability after repeated firings. J PROSTHS~ DENT 1977;37:173-5. 2. Barghi N, Richardson JT. A study of various factors influencing shade of bond porcelain. J PROSTHET DENT 1978;39:282-4. 3. Barghi N. Color and glaze: effects of repeated firings. J PROSTHET DENT 1982;47:393-5.
4. Jorgenson MW, Goodkind RJ. Spectrophotometric study of five porcelain shades relative to the dimensions of color, porcelain thickness, and repeated firings. J PROWHET DENT 1979;42:96-105. 5. Obregon A, Goodkind RJ, Schwabacher WB. Effects of opaque and porcelain texture on the color of ceramometal restorations. J PROSTHET DENT
1981;46:330-40.
6. Brewer JD, Akers CK, Garlapo DA, Sorensen SE. Spectrometric analysis of the influence of metal substrates on the color of metal-ceramic restorations. J Dent Res 1985;64:74-7. 7. Jacobs SH, Goodacre CJ, Moore BK, Dykema RW. Effect of porcelain thickness and type of metal-ceramic aboy on color. J PROSTHET DENT 1987;57:138-45.
8. Crispin BJ, Okamoto SK, and Globe H. Porcelain crown substructures effect on visually perceivable Value. J Prosthet Dent 1989 (in-press). 9. Seghi RR, Johnston WM, O’Brien WJ. Spectrophotometric analysis of color differences between porcelain systems. J PROSTHET DENT 1986$&35-40.
10. Seghi RR, Johnston WM, O’Brien WJ. Performance assessment of colorimetric devices on dental porcelains. J Dent Res 198%68:1755-g. 11. CIE Publication No. 15.2 Calorimetry. Central Bureau of the CIE, Vienna, 1986. 12. Tuccillo JJ, Cascone PJ. The evolution of porcelain-fused-to metal alloy systems in dental ceramics. In McLean JW, ed. Proceedings of the first international symposium on ceramics. Chicago: Quintessence Pub1 Co Inc, 1983;347-70. 13. Doremus RH. Optical properties of small silver particles. J Chem Phys 1965;42:414-7. 14. Fanderlik I. Glass science and technology 5: optical properties of glass. New York: Elsevier Science Pub1 Co, Inc, 1983;320. 15. Segbi RR, Hewlett ER, Kim J. Visual assessment of calorimetric analysis on translucent dental porcelain. J Dent Res 1989;68:1766-4. Reprint
requests to:
DR. BRUCE J. CRISPIN SCHOOL OF DENTISTRY UNIVERSITY OF CALIFORNIA 10833 LE CONTE AVE. Los ANGELES, CA 90024
MARCH
lSS1
VOLUME
65
NUMBER
S
