- Email: [email protected]
The effect of water storage and light exposure on the color and translucency of a hybrid and a microfilled composite
The effect of water storage and light exposure on the color and translucency of a hybrid and a microfilled composite Wolfgang Buchalla, Dr med dent,a Thomas Attin, Prof Dr med dent,b Ralf-Dieter Hilgers, Dr rer nat,c and Elmar Hellwig, Prof Dr med dentd University Clinic for Dentistry, Georg-August University, Göttingen, and Albert-Lugwigs University, Freiberg, Germany Statement of problem. Internal discoloration may occur with the intraoral use of resin-based restoration materials. Water storage and light exposure influence the color properties of microfilled and hybrid composites. Purpose. The purpose of this study was to determine color and translucency changes in a hybrid and a microfilled composite after light exposure with and without water storage. Material and methods. A hybrid (Tetric) and a microfilled composite (Silux Plus) were subjected to artificial daylight with and without water storage. Tristimulus Yxy values were determined colorimetrically against a black or a white background. Differences from baseline were calculated as ∆E*ab for up to 1 month. After 1 month, ∆L*, ∆a*, ∆b*, and the contrast ratio ∆C were calculated. Data were analyzed with 2-way analysis of variance (P<.05). Results. The ∆Eab* increased over time for wet and dry stored specimens of both materials. Wet storage for 1 month resulted in significantly higher ∆Eab* and ∆C but lower ∆b* than dry storage. The ∆a* was significantly higher for Tetric compared with Silux Plus. Both materials showed negative ∆L* values under both storage conditions without significant differences between materials or storage conditions. Conclusion. The results of this in vitro study suggest that resin-based restoration materials undergo measurable changes due to daylight exposure. Increased changes occurred under the influence of water storage. (J Prosthet Dent 2002;87:264-70.)
CLINICAL IMPLICATIONS The microfilled and hybrid composites tested in this in vitro study showed color shifts after 1 month of artificial aging, but to an extent that may be undetectable under normal clinical conditions.
T
he appearance of resin composite restorations changes over time. External stain accumulation, marginal leakage, secondary caries, and internal discoloration can make a restoration visually unacceptable. In addition to secondary caries, discoloration is one of the main reasons for the removal of resin composite fillings.1 Water accumulation, changes in chemical compounds necessary for photopolymerization, photo-oxidation, and other processes have been
This study was supported in part by Ivoclar AG-Vivadent Ets (Schaan, Liechtenstein). aAssociate Professor, Department of Conservative and Preventive Dentistry and Periodontology, University Clinic for Dentistry, Georg-August University. Adjunct Associate Professor, Oral Health Research Institute, School of Dentistry, Indiana University, Indianapolis, Ind. bHead, Department of Operative Dentistry and Periodontology, University Clinic for Dentistry, Georg-August University. cHead, Institute for Medical Statistics, University Clinic, RWTH Aachen, Germany. dHead, Department of Operative Dentistry and Periodontology, University Clinic for Dentistry, Albert-Ludwigs University. 264 THE JOURNAL OF PROSTHETIC DENTISTRY
thought to be responsible for internal color changes.2,3 Numerous tests have been used for artificial aging of restorative materials to investigate color stability in vitro. Hot water4 and artificial light at visible or ultraviolet (UV) ranges after water storage5 have been used for artificial aging. The most common protocol is a combination of artificial light and storage at 100% relative humidity in water or water spray.6, 7 Hybrid composites and microfilled composites are widely used as tooth-colored restorative materials. A weathering procedure that combined heat, UV-light, and moisture resulted in greater initial discoloration in a microfilled composite than in a hybrid composite.8 It has been suggested that the matrix-filler interface plays a major role in the uptake of water by composites9 and that microfilled composites absorb more water at the matrix-filler interface than other composites.10 Therefore, the presence of water during artificial aging may affect the optical properties of microfilled composites more than those of hybrid composites. The purpose of this study was to investigate color and VOLUME 87 NUMBER 3
BUCHALLA ET AL
THE JOURNAL OF PROSTHETIC DENTISTRY
Table I. Composites used in this study Composite
Manufacturer
Filler type
Silux Plus
3M, St Paul, Minn.
Microfilled
Tetric
Vivadent, Schaan, Liechtenstein
Hybrid small particle
Filler content* (% by weight)
Shade
56
U
82.3 ± 0.7
22
Batch no.
19930923 19930924 560584
*Information provided by the manufacturers.
translucency changes in a microfilled and a hybrid composite after artificial aging, with an emphasis on the influence of the presence or absence of water.
MATERIAL AND METHODS Ten specimens each were prepared from a hybrid composite (Tetric, shade 22) and a microfilled composite (Silux Plus, shade U). Manufacturing information is provided in Table I. The shades were chosen after a visual comparison of Tetric samples (shade U, 22, 23) and a Silux Plus sample (shade U) revealed that Tetric 22 was the best match. A plane parallel stainless steel plate was covered with polyethylene foil for preparation of the test specimens. A polytetrafluoroethylene (PTFE) ring (15.5 mm in internal diameter and 7 mm high) was positioned on the polyethylene foil and clamped in place. This mold was slightly overfilled with the respective composite and covered with a second polyethylene foil. Subsequently, another plane parallel stainless steel plate was adapted on top of the polyethylene foil. A hydraulic press was used to induce pressure of up to 15 bar on the steel plates to remove excess material. After removal of the upper steel plate, the mold was placed in a pressure-chamber (Ivomat; Ivoclar, Schaan, Liechtenstein) at 5 bar for 5 minutes to ensure specimen homogeneity with no detectable trapped air bubbles. The mold was placed in a light-chamber (Spectramat; Vivadent, Schaan, Liechtenstein) to polymerize of each of the plan parallel sides of the composite specimen for 5 minutes at 20,000 lux. This procedure has been shown to completely transform the camphoroquinone to colorless decomposition products.11 Due to the adsorption of the blue portion of the spectrum of visible light, champhoroquinone would appear yellow. After polymerization was complete, the polyethylene foils and PTFE ring were removed. A low-speed diamond saw (Isomet Plus, Buehler, Ill.) was used to cut two 1.3-mm–thick slices out of the middle of the composite cylinder. Each of these was designated for 1 of 2 experimental groups to ensure an equal distribution. The 2 surfaces of each specimen were ground wet with 1000-grit silicon carbide abrasive paper. This procedure resulted in plane parallel specimens 15.5 mm in diameter and 1.2 ± 0.03 mm thick. Specimens with a thickness between 1.0 and 1.3 mm have been widely MARCH 2002
Fig. 1. Experimental design.
used in investigations of contrast ratios.5,7,8 The specimens were stored in a dark refrigerator at 5°C when not in use. Ten specimens of each material were divided into 2 groups (Fig. 1). Group I specimens were submerged in distilled water. A single vial stored at room temperature (23°C) was used for each specimen. Group II specimens remained dry in open vials under room conditions of 50% relative humidity and a temperature of 23°C. Both groups were subjected to an artificial light source for up to 1 month (True-Lite; Duro-Test Corp, Midland Park, N.J.). True-Lite is a fluorescent tube that emits visible, ultraviolet, and infrared light between a wavelength of 290 and 770 nm. This corresponds well with daylight at 5500 K light temperature. 265
THE JOURNAL OF PROSTHETIC DENTISTRY
BUCHALLA ET AL
green, and blue filters (x-, y-, and z-values). These tristimulus reflectance values were converted into the CIELAB values L*, a*, and b* by the microcomputer.13 L* is defined as lightness, a* as the red-green component, and b* as the yellow-blue component (Fig. 2). Measurements with black backings were used for the calculation of the L*, a*, and b* values. A black background corresponds to the clinical situation of a class III or IV restoration, where the filling has the dark background of the oral cavity rather than a lightreflecting tooth substance. Translucency was calculated as the contrast-ratio as follows: yb C = ––– yw
Fig. 2. CIELAB color space. Three coordinate axes perpendicular to each other define 3-dimensional space. L* represents lightness between black and white, a* represents chromatic value between green and red, and b* represents chromatic value between blue and yellow. Three coordinates can be used to express every color. Differences between 2 different colors are expressed as ∆E*ab.
The light intensity of True-Lite was 500 lux. The specimen bottles were placed 1.5 m from the light source for exposure. The exposure time was limited to 10 hours per day. The color properties of the specimens were measured against a black and white backing at baseline and after 1, 2, 4, 8, 16, 32, and 48 hours and 1 month. Colorimetric data were generated with a chromameter (CR 300; Minolta Co, Dietikon, Switzerland) connected to a personal computer (PS/2 40 SX; IBM, White Plains, NY) with the software that belonged to the system. Group I specimens were removed from the distilled water and gently dried with cellulose for the measurement procedure. Either black or white ceramic tiles (Vivadent) served as backgrounds. Black and white backings are necessary to calculate opacity.12 The specimens were fixed to the tiles with n-di-butylphthalate to provide an optical contact between the specimen and the ceramic tile. Butylphthalate is colorless and has a refractive index similar to the resin of composites. The chromameter used was a tristimulus colorimeter with an 8-mm wide measuring port and a d/0° measuring geometry (diffuse illumination/0° viewing angle). A tristimulus colorimeter simulates the interpretation of color by the human eye. A xenon flash tube generates a short flash during the measurement. The remitted light is quantitatively registered between 380 to 770 nm wavelength from sensors behind red, 266
(1)
where b represents the black backing and w the white backing. The colorimeter was calibrated with a white standard (CR-A43; Minolta Co) before the first measurement and at regular time intervals. Additionally, the system was calibrated for 100% reflectance with the white ceramic tile and for 0% reflectance with the black ceramic tile. Using the values of L*, a*, b*, and C, differences between the first measurement and the time-dependent measurements were calculated as the color difference values ∆L*, ∆a*, ∆b*, and ∆C. For example, the formula for ∆L* was as follows: ∆L*(t) = L*(t) – L*(0)
(2)
where L*(t) is the L* value measured at time t and L*(0) is the L* value from the first measurement (baseline value, time = 0). To derive an overall color change, ∆E*ab(t) at time t was calculated as follows according to CIE 13: ∆Ε * ab (t) =
[∆L* (t)]2 + [∆a*(t)]2 + [∆b* (t)]2
(3)
Statistical analysis was performed by fitting a 2-way analysis of variance (ANOVA, P<.05) model to the independent observations of ∆E*ab after 1 month based on the 20 specimens. Each of the 2 between factors (material and dryness) had 2 levels. The data were analyzed in further detail by fitting the corresponding 2-way ANOVA models to the data for the single values of L*, a*, b*, and C after 1 month. All computations were performed with a statistical software package (SAS Institute, Cary, NC).
RESULTS Increased changes in the development of the ∆E*ab color differences were found over the experimental period (Fig. 3). The mean ∆E*ab for Tetric and Silux Plus under both wet and dry storage conditions increased over the period of the investigation. The wet-storage specimens showed a tendency to higher VOLUME 87 NUMBER 3
BUCHALLA ET AL
THE JOURNAL OF PROSTHETIC DENTISTRY
Fig. 3. Mean ∆E*ab(t) color differences at each time point t from baseline (time = 0) of investigated materials under wet and dry conditions.
Table II. Means and SD of the materials under different storage conditions at 1 month from baseline Parameter
∆E*ab
Material
Storage
N
Mean*
SD
Tetric Tetric Silux Plus Silux Plus
Wet Dry Wet Dry
5 5 5 5
5.07a 3.17b 4.08a 3.01b
1.327 1.464 0.706 1.057
∆L*
Tetric Tetric Silux Plus Silux Plus
Wet Dry Wet Dry
5 5 5 5
–1.05ab –1.18ab –1.28a –0.88b
0.120 0.168 0.178 0.293
∆a*
Tetric Tetric Silux Plus Silux Plus
Wet Dry Wet Dry
5 5 5 5
1.23a 1.06a 0.30b 0.31b
0.408 0.484 0.143 0.353
∆b*
Tetric Tetric Silux Plus Silux Plus
Wet Dry Wet Dry
5 5 5 5
–4.80a –2.44b –3.84a –2.79b
1.278 1.862 0.753 1.188
∆C
Tetric Tetric Silux Plus Silux Plus
Wet Dry Wet Dry
5 5 5 5
0.015a –0.008b 0.004ac 0.002bc
0.009 0.015 0.008 0.008
N, Number of specimens. *Mean values with the same superscript letter were not significantly different within each parameter (P<.05).
∆E*ab color differences after 4 hours than the dry-storage specimens. Tetric tended to higher ∆E*ab values after 32 hours than Silux Plus under both wet and dry storage conditions. The results after 1 month of artificial aging are presented in Tables II and III. Wet storage resulted in a significantly higher color difference ∆E*ab than dry MARCH 2002
storage (P=.012). Differences between the materials (P=.291 of main effect) and materials under the same storage conditions were not significant (P=.437 of interaction). When the possible reasons for the ∆E*ab were examined, a similarity to ∆b* was found. The means of ∆E*ab between the experimental groups were of the same order as for ∆b*. The assumption that ∆b* 267
THE JOURNAL OF PROSTHETIC DENTISTRY
BUCHALLA ET AL
Table III. ANOVA results at 1 month from baseline Parameter
Source of variation
∆E*ab
Material Storage Material × storage Error
∆L*
df
Mean square
F value
Pr > F
1 1 1 16
1.644 11.00 0.878 1.380
1.19 7.97 0.64
0.291 0.012 0.437
Material Storage Material × storage Error
1 1 1 16
0.005 0.086 0.351 0.040
0.12 2.15 8.78
0.733 0.162 0.009
∆a*
Material Storage Material × storage Error
1 1 1 16
3.503 0.034 0.036 0.136
25.66 0.25 0.26
0.000 0.623 0.614
∆b*
Material Storage Material × storage Error
1 1 1 16
0.465 14.53 2.119 1.770
0.26 8.21 1.20
0.615 0.011 0.290
∆C
Material Storage Material × storage Error
1 1 1 16
0.000 0.001 0.001 0.000
0.04 7.20 4.94
0.835 0.016 0.041
Pr > F, Probability of a larger F value based on the null hypothesis of no effect from source of variation.
has a strong influence on ∆E*ab is supported by the high values for ∆b* compared to ∆a* and ∆L*. According to Equation 3, the parameter with the highest absolute value exerts the greatest influence on ∆E*ab. The opacity was calculated as the contrast ratio (Eqn. 1). Negative values indicate a more translucent appearance after 1 month compared to the beginning of the experiment; such values were observed for Tetric under dry conditions. Additionally, wet-storage Tetric had a significantly higher ∆C as the same material under dry storage conditions (P=.041). Results similar to those for ∆E*ab were obtained for ∆b* and ∆C after 1 month from baseline. Significant differences between wet and dry conditions were observed for ∆b* (P=.011) and ∆C (P=.016). However, the negative values for ∆b* indicated a decrease over time. Negative ∆b* values indicate a less yellow and more blue appearance of the respective materials. No significant differences were found for ∆b* between the materials (P=.615 of main effect, P=.290 of interaction) under the same storage conditions. No significant differences were observed between wet- and dry-storage specimens after 1 month of light exposure with respect to ∆a* for either of the investigated materials. However, regardless of the storage conditions, ∆a* means were significantly higher for Tetric than for Silux Plus (P=.0001). All ∆a* values were positive. Positive ∆a* values after artificial aging indicate a less green and more red material appearance. All mean ∆L* values were negative. Negative values 268
of ∆L* after artificial aging indicate a darker material appearance. Silux Plus demonstrated the largest ∆L* of the investigated groups under wet conditions and the smallest ∆L* under dry conditions. Silux Plus changed more under wet conditions than under dry conditions (P=.009). No significant differences (P=.733 main effect material, P=.162 main effect dry condition) were observed for ∆L* between the other experimental groups.
DISCUSSION Clinical experience demonstrates that tooth-colored direct restoration materials change their optical properties over time and that internal discolorations often necessitate restoration replacement. Different testing procedures have been developed to artificially age restorative materials in vitro.4 Most studies include light exposure in the artificial aging procedure.6-8 UVlamps5 and high power xenon-arc lamps6 both create an intensive light with UV and infrared (IR) and are associated with xenon-specific peaks within the spectrum. Tungsten filament lamps create UV only if the lamps are high powered but have large amounts of IR. Their spectral distribution mimics blackbody radiation more than natural sunlight. The light source used in the present study emits light comparable to natural daylight at 5500 K light temperature, with some amount of UV. Because these experimental conditions caused a more intensive light exposure than found in the clinical situation, they cannot be correlated with a clinical time frame. VOLUME 87 NUMBER 3
BUCHALLA ET AL
In the present study, both of the investigated materials revealed a color shift from yellow to blue (negative ∆b*) and from green to red (positive ∆a*) under both storage conditions. All specimens became darker during the investigation (negative ∆L*). The overall color change was in a range similar to that reported in a previous study8 in which a UV-light weathering technique with high temperature in a humid atmosphere was used. The authors found an ∆E*ab of approximately 5.25 and 5.44 units for a hybrid and a microfilled composite, respectively, with no significant difference between the two. In the present study, an ∆E*ab of 5.1 for the hybrid composite and 4.1 for the microfilled composite was found under wet conditions, with no significant difference between the materials. Significant differences in ∆E*ab were found between wet- and dry-storage specimens. It has been stated that the matrix-filler interface plays a major role in the uptake of water by composites.9 Changes in optical properties within the matrices or the matrixfiller interface therefore could have been responsible for the different values of ∆E*ab. The increase of ∆E*ab was higher for the hybrid composite than the microfilled composite, but there was no significant difference between the materials. If the matrix-fillerinterface plays a major role in water uptake,10 it may be assumed that a hybrid composite that includes fine particle fillers behaves similarly to a microfilled composite. The hybrid composite tested in this study contains ytterbiumtrifluoride to provide a certain amount of fluoride release.14 The fact that this component may be water soluble also may have influenced optical properties. Given the size of the ∆E*ab values, it is questionable whether the statistical differences found in this study would be clinically significant. Reported thresholds for ∆E*ab values becoming visible to the human eye vary widely in the literature and are dependent on the individual observer. It has been claimed that the 50% threshold for perceptibility of ∆E*ab color differences is lower than 1 under perfect standardized laboratory conditions.15,16 However, under clinical conditions in the mouth, ∆E*ab color differences have been reported as relevant only when higher than 3.317 or 3.6.18 The ∆b* showed the highest color shift among ∆L*, ∆a*, and ∆b*. This parameter demonstrated the same behavior as ∆E*ab, with significant differences between wet and dry storage conditions but not between the materials tested. Under the experimental conditions of this study, it may be stated that the color shift from yellow to blue had the greatest influence on the overall color difference for all of the experimental groups. The presence of water during artificial aging with light enhanced this color change. The materials investigated MARCH 2002
THE JOURNAL OF PROSTHETIC DENTISTRY
contain diketones as polymerization initiators. Diketones absorb light of a wavelength between 420 and 450 nm (blue light). After polymerization, less diketone is present and therefore less blue light is absorbed. This was confirmed in a previous study2 in which color shifts of 5 to 10 ∆b* units were found due to polymerization. In the present study, the light-polymerization process was performed at an intense level; the amount of remaining diketones therefore should have been very low. The results suggest that other factors may play an important role in the shift from yellow to blue up to 5 ∆b* units. Given that the presence of water enhanced the blue shift of the investigated materials, it is conceivable that oxidation reactions of unreacted C=C double bonds produce colored peroxide products.3 No correlation between dry and wet storage could be detected regarding ∆a*, but the hybrid composite exhibited a significantly higher ∆a* than the microfilled composite after 1 month. The same ∆a* behavior was reported in a previous investigation of hybrid and microfilled composites.8 Both investigated materials had a darker appearance (lower ∆L*) after 1 month of artificial aging. A significant difference between wet and dry storage was detected only for the microfilled composite. Because microfilled composites consist of a greater proportion of filler-matrix interface than hybrid composites, it can be argued that the hydrolytical instability of the silane bond between fillers and the matrix was responsible for the decrease in lightness due to water storage. It should be borne in mind that a correlation between lightness and opacity may exist. When black backings are used for colorimetry, an increase in opacity may result in an increase in lightness, while a decrease in opacity may result in a decrease in lightness. In a more opaque specimen, more light will be reflected back. In a less opaque specimen, more light will penetrate the material and become absorbed by the black backing. When the results for ∆L* and ∆C were compared, a correlation between lightness and opacity could not be found. Mechanisms other than opacity may have been responsible for changes in lightness. Significant differences were detected for ∆C between wet- and dry-storage specimens, in that wet storage caused significantly greater opacity. While opacity increased for the other experimental groups, it decreased for Tetric under dry storage conditions, resulting in a significant difference between wet and dry specimens. As mentioned above for ∆E*ab, the values for ∆L*, ∆a*, ∆b*, and ∆C were within a range that probably would not be discernable in the clinical situation. It can be assumed that the color stability of both investigated materials was within the clinically acceptable range. 269
THE JOURNAL OF PROSTHETIC DENTISTRY
CONCLUSIONS Within the limitations of this in vitro study, the following conclusions were drawn: 1. Water storage caused a blue shift (negative ∆b*) and a higher ∆E*ab for both the microfilled and hybrid composite tested. 2. Water storage led to greater darkening (negative ∆L*) of the microfilled composite than the hybrid composite tested. 3. The hybrid composite tested demonstrated a larger shift toward red (higher ∆a*) than the microfilled composite tested. We thank Dr Áine Lennon for her correction of the manuscript and Ivoclar AG-Vivadent Ets (Schaan, Liechtenstein) for its support of this study.
REFERENCES 1. Mjör IA, Toffenetti F. Placement and replacement of resin-based composite restorations in Italy. Oper Dent 1992;17:82-5. 2. Seghi RR, Gritz MD, Kim J. Colorimetric changes in composites resulting from visible-light-initiated polymerization. Dent Mater 1990;6:133-7. 3. Ferracane JL, Moser JB, Greener EH. Ultraviolet light-induced yellowing of dental restorative resins. J Prosthet Dent 1985;54:483-7. 4. Asmussen E. An accelerated test for color stability of restorative resins. Acta Odontol Scand 1981;39:329-32. 5. Uchida H, Vaidyanathan J, Viswanadhan T, Vaidyanathan TK. Color stability of dental composites as a function of shade. J Prosthet Dent 1998;79:372-7. 6. Powers JM, Dennison JB, Koran A. Color stability of restorative resins under accelerated aging. J Dent Res 1978;57:964-70. 7. Powers JM, Fan PL, Raptis CN. Color stability of new composite restorative materials under accelerated aging. J Dent Res 1980;59:2071-4. 8. Fruits TJ, Duncanson MG Jr, Miranda FJ. In vitro weathering of selected direct esthetic restorative materials. Quintessence Int 1997;28:409-14. 9. Kalachandra S. Influence of fillers on the water sorption of composites. Dent Mater 1989;5:283-8.
270
BUCHALLA ET AL
10. Kalachandra S, Wilson TW. Water sorption and mechanical properties of light-cured proprietary composite tooth restorative materials. Biomaterials 1992;13:105-9. 11. Domininghaus H. Plastics for engineers. 1st ed. Munich: Hanser; 1992. p. 610. 12. Inokoshi S, Burrow MF, Kataumi M, Yamada T, Takatsu T. Opacity and color changes of tooth-colored restorative materials. Oper Dent 1996;21:73-80. 13. Commission Internationale de L’Eclairage. Colorimetry. CIE Publication No. 15.2. Vienna: Central Bureau of the CIE; 1986. 14. Buchalla W, Attin T, Kalb K, Hellwig E. [Fluoride release and uptake of an experimental composite in vitro and in situ.] Dtsch Zahnärztl Z 1998;53:707-12. German. 15. Douglas RD, Brewer JD. Acceptability of shade differences in metal ceramic crowns. J Prosthet Dent 1998;79:254-60. 16. Kuehni RG, Marcus RT. An experiment in visual scaling of small color differences. Color Res Appl 1979;4:31-9. 17. Ruyter IE, Nilner K, Möller B. Color stability of dental composite resin materials for crown and bridge veneers. Dent Mater 1987;3:246-51. 18. Johnston WM, Kao EC. Assessment of appearance match by visual observation and clinical colorimetry. J Dent Res 1989;68:819-22. Reprint requests to: DR WOLFGANG BUCHALLA GEORG-AUGUST-UNIVERSITÄT GÖTTINGEN ZENTRUM ZAHN-, MUND-, UND KIEFERHEILKUNDE ABTEILUNG FÜR ZAHNERHALTUNG, PRÄVENTIVE ZAHNHEILKUNDE PARODONTOLOGIE ROBERT-KOCH-STR. 40 37075 GÖTTINGEN GERMANY FAX: (49)551-39-2037 E-MAIL: [email protected]
UND
Copyright © 2002 by The Editorial Council of The Journal of Prosthetic Dentistry. 0022-3913/2002/$35.00 + 0. 10/1/121743
doi:10.1067/mpr.2002.121743
VOLUME 87 NUMBER 3
CLINICAL IMPLICATIONS The microfilled and hybrid composites tested in this in vitro study showed color shifts after 1 month of artificial aging, but to an extent that may be undetectable under normal clinical conditions.
T
he appearance of resin composite restorations changes over time. External stain accumulation, marginal leakage, secondary caries, and internal discoloration can make a restoration visually unacceptable. In addition to secondary caries, discoloration is one of the main reasons for the removal of resin composite fillings.1 Water accumulation, changes in chemical compounds necessary for photopolymerization, photo-oxidation, and other processes have been
This study was supported in part by Ivoclar AG-Vivadent Ets (Schaan, Liechtenstein). aAssociate Professor, Department of Conservative and Preventive Dentistry and Periodontology, University Clinic for Dentistry, Georg-August University. Adjunct Associate Professor, Oral Health Research Institute, School of Dentistry, Indiana University, Indianapolis, Ind. bHead, Department of Operative Dentistry and Periodontology, University Clinic for Dentistry, Georg-August University. cHead, Institute for Medical Statistics, University Clinic, RWTH Aachen, Germany. dHead, Department of Operative Dentistry and Periodontology, University Clinic for Dentistry, Albert-Ludwigs University. 264 THE JOURNAL OF PROSTHETIC DENTISTRY
thought to be responsible for internal color changes.2,3 Numerous tests have been used for artificial aging of restorative materials to investigate color stability in vitro. Hot water4 and artificial light at visible or ultraviolet (UV) ranges after water storage5 have been used for artificial aging. The most common protocol is a combination of artificial light and storage at 100% relative humidity in water or water spray.6, 7 Hybrid composites and microfilled composites are widely used as tooth-colored restorative materials. A weathering procedure that combined heat, UV-light, and moisture resulted in greater initial discoloration in a microfilled composite than in a hybrid composite.8 It has been suggested that the matrix-filler interface plays a major role in the uptake of water by composites9 and that microfilled composites absorb more water at the matrix-filler interface than other composites.10 Therefore, the presence of water during artificial aging may affect the optical properties of microfilled composites more than those of hybrid composites. The purpose of this study was to investigate color and VOLUME 87 NUMBER 3
BUCHALLA ET AL
THE JOURNAL OF PROSTHETIC DENTISTRY
Table I. Composites used in this study Composite
Manufacturer
Filler type
Silux Plus
3M, St Paul, Minn.
Microfilled
Tetric
Vivadent, Schaan, Liechtenstein
Hybrid small particle
Filler content* (% by weight)
Shade
56
U
82.3 ± 0.7
22
Batch no.
19930923 19930924 560584
*Information provided by the manufacturers.
translucency changes in a microfilled and a hybrid composite after artificial aging, with an emphasis on the influence of the presence or absence of water.
MATERIAL AND METHODS Ten specimens each were prepared from a hybrid composite (Tetric, shade 22) and a microfilled composite (Silux Plus, shade U). Manufacturing information is provided in Table I. The shades were chosen after a visual comparison of Tetric samples (shade U, 22, 23) and a Silux Plus sample (shade U) revealed that Tetric 22 was the best match. A plane parallel stainless steel plate was covered with polyethylene foil for preparation of the test specimens. A polytetrafluoroethylene (PTFE) ring (15.5 mm in internal diameter and 7 mm high) was positioned on the polyethylene foil and clamped in place. This mold was slightly overfilled with the respective composite and covered with a second polyethylene foil. Subsequently, another plane parallel stainless steel plate was adapted on top of the polyethylene foil. A hydraulic press was used to induce pressure of up to 15 bar on the steel plates to remove excess material. After removal of the upper steel plate, the mold was placed in a pressure-chamber (Ivomat; Ivoclar, Schaan, Liechtenstein) at 5 bar for 5 minutes to ensure specimen homogeneity with no detectable trapped air bubbles. The mold was placed in a light-chamber (Spectramat; Vivadent, Schaan, Liechtenstein) to polymerize of each of the plan parallel sides of the composite specimen for 5 minutes at 20,000 lux. This procedure has been shown to completely transform the camphoroquinone to colorless decomposition products.11 Due to the adsorption of the blue portion of the spectrum of visible light, champhoroquinone would appear yellow. After polymerization was complete, the polyethylene foils and PTFE ring were removed. A low-speed diamond saw (Isomet Plus, Buehler, Ill.) was used to cut two 1.3-mm–thick slices out of the middle of the composite cylinder. Each of these was designated for 1 of 2 experimental groups to ensure an equal distribution. The 2 surfaces of each specimen were ground wet with 1000-grit silicon carbide abrasive paper. This procedure resulted in plane parallel specimens 15.5 mm in diameter and 1.2 ± 0.03 mm thick. Specimens with a thickness between 1.0 and 1.3 mm have been widely MARCH 2002
Fig. 1. Experimental design.
used in investigations of contrast ratios.5,7,8 The specimens were stored in a dark refrigerator at 5°C when not in use. Ten specimens of each material were divided into 2 groups (Fig. 1). Group I specimens were submerged in distilled water. A single vial stored at room temperature (23°C) was used for each specimen. Group II specimens remained dry in open vials under room conditions of 50% relative humidity and a temperature of 23°C. Both groups were subjected to an artificial light source for up to 1 month (True-Lite; Duro-Test Corp, Midland Park, N.J.). True-Lite is a fluorescent tube that emits visible, ultraviolet, and infrared light between a wavelength of 290 and 770 nm. This corresponds well with daylight at 5500 K light temperature. 265
THE JOURNAL OF PROSTHETIC DENTISTRY
BUCHALLA ET AL
green, and blue filters (x-, y-, and z-values). These tristimulus reflectance values were converted into the CIELAB values L*, a*, and b* by the microcomputer.13 L* is defined as lightness, a* as the red-green component, and b* as the yellow-blue component (Fig. 2). Measurements with black backings were used for the calculation of the L*, a*, and b* values. A black background corresponds to the clinical situation of a class III or IV restoration, where the filling has the dark background of the oral cavity rather than a lightreflecting tooth substance. Translucency was calculated as the contrast-ratio as follows: yb C = ––– yw
Fig. 2. CIELAB color space. Three coordinate axes perpendicular to each other define 3-dimensional space. L* represents lightness between black and white, a* represents chromatic value between green and red, and b* represents chromatic value between blue and yellow. Three coordinates can be used to express every color. Differences between 2 different colors are expressed as ∆E*ab.
The light intensity of True-Lite was 500 lux. The specimen bottles were placed 1.5 m from the light source for exposure. The exposure time was limited to 10 hours per day. The color properties of the specimens were measured against a black and white backing at baseline and after 1, 2, 4, 8, 16, 32, and 48 hours and 1 month. Colorimetric data were generated with a chromameter (CR 300; Minolta Co, Dietikon, Switzerland) connected to a personal computer (PS/2 40 SX; IBM, White Plains, NY) with the software that belonged to the system. Group I specimens were removed from the distilled water and gently dried with cellulose for the measurement procedure. Either black or white ceramic tiles (Vivadent) served as backgrounds. Black and white backings are necessary to calculate opacity.12 The specimens were fixed to the tiles with n-di-butylphthalate to provide an optical contact between the specimen and the ceramic tile. Butylphthalate is colorless and has a refractive index similar to the resin of composites. The chromameter used was a tristimulus colorimeter with an 8-mm wide measuring port and a d/0° measuring geometry (diffuse illumination/0° viewing angle). A tristimulus colorimeter simulates the interpretation of color by the human eye. A xenon flash tube generates a short flash during the measurement. The remitted light is quantitatively registered between 380 to 770 nm wavelength from sensors behind red, 266
(1)
where b represents the black backing and w the white backing. The colorimeter was calibrated with a white standard (CR-A43; Minolta Co) before the first measurement and at regular time intervals. Additionally, the system was calibrated for 100% reflectance with the white ceramic tile and for 0% reflectance with the black ceramic tile. Using the values of L*, a*, b*, and C, differences between the first measurement and the time-dependent measurements were calculated as the color difference values ∆L*, ∆a*, ∆b*, and ∆C. For example, the formula for ∆L* was as follows: ∆L*(t) = L*(t) – L*(0)
(2)
where L*(t) is the L* value measured at time t and L*(0) is the L* value from the first measurement (baseline value, time = 0). To derive an overall color change, ∆E*ab(t) at time t was calculated as follows according to CIE 13: ∆Ε * ab (t) =
[∆L* (t)]2 + [∆a*(t)]2 + [∆b* (t)]2
(3)
Statistical analysis was performed by fitting a 2-way analysis of variance (ANOVA, P<.05) model to the independent observations of ∆E*ab after 1 month based on the 20 specimens. Each of the 2 between factors (material and dryness) had 2 levels. The data were analyzed in further detail by fitting the corresponding 2-way ANOVA models to the data for the single values of L*, a*, b*, and C after 1 month. All computations were performed with a statistical software package (SAS Institute, Cary, NC).
RESULTS Increased changes in the development of the ∆E*ab color differences were found over the experimental period (Fig. 3). The mean ∆E*ab for Tetric and Silux Plus under both wet and dry storage conditions increased over the period of the investigation. The wet-storage specimens showed a tendency to higher VOLUME 87 NUMBER 3
BUCHALLA ET AL
THE JOURNAL OF PROSTHETIC DENTISTRY
Fig. 3. Mean ∆E*ab(t) color differences at each time point t from baseline (time = 0) of investigated materials under wet and dry conditions.
Table II. Means and SD of the materials under different storage conditions at 1 month from baseline Parameter
∆E*ab
Material
Storage
N
Mean*
SD
Tetric Tetric Silux Plus Silux Plus
Wet Dry Wet Dry
5 5 5 5
5.07a 3.17b 4.08a 3.01b
1.327 1.464 0.706 1.057
∆L*
Tetric Tetric Silux Plus Silux Plus
Wet Dry Wet Dry
5 5 5 5
–1.05ab –1.18ab –1.28a –0.88b
0.120 0.168 0.178 0.293
∆a*
Tetric Tetric Silux Plus Silux Plus
Wet Dry Wet Dry
5 5 5 5
1.23a 1.06a 0.30b 0.31b
0.408 0.484 0.143 0.353
∆b*
Tetric Tetric Silux Plus Silux Plus
Wet Dry Wet Dry
5 5 5 5
–4.80a –2.44b –3.84a –2.79b
1.278 1.862 0.753 1.188
∆C
Tetric Tetric Silux Plus Silux Plus
Wet Dry Wet Dry
5 5 5 5
0.015a –0.008b 0.004ac 0.002bc
0.009 0.015 0.008 0.008
N, Number of specimens. *Mean values with the same superscript letter were not significantly different within each parameter (P<.05).
∆E*ab color differences after 4 hours than the dry-storage specimens. Tetric tended to higher ∆E*ab values after 32 hours than Silux Plus under both wet and dry storage conditions. The results after 1 month of artificial aging are presented in Tables II and III. Wet storage resulted in a significantly higher color difference ∆E*ab than dry MARCH 2002
storage (P=.012). Differences between the materials (P=.291 of main effect) and materials under the same storage conditions were not significant (P=.437 of interaction). When the possible reasons for the ∆E*ab were examined, a similarity to ∆b* was found. The means of ∆E*ab between the experimental groups were of the same order as for ∆b*. The assumption that ∆b* 267
THE JOURNAL OF PROSTHETIC DENTISTRY
BUCHALLA ET AL
Table III. ANOVA results at 1 month from baseline Parameter
Source of variation
∆E*ab
Material Storage Material × storage Error
∆L*
df
Mean square
F value
Pr > F
1 1 1 16
1.644 11.00 0.878 1.380
1.19 7.97 0.64
0.291 0.012 0.437
Material Storage Material × storage Error
1 1 1 16
0.005 0.086 0.351 0.040
0.12 2.15 8.78
0.733 0.162 0.009
∆a*
Material Storage Material × storage Error
1 1 1 16
3.503 0.034 0.036 0.136
25.66 0.25 0.26
0.000 0.623 0.614
∆b*
Material Storage Material × storage Error
1 1 1 16
0.465 14.53 2.119 1.770
0.26 8.21 1.20
0.615 0.011 0.290
∆C
Material Storage Material × storage Error
1 1 1 16
0.000 0.001 0.001 0.000
0.04 7.20 4.94
0.835 0.016 0.041
Pr > F, Probability of a larger F value based on the null hypothesis of no effect from source of variation.
has a strong influence on ∆E*ab is supported by the high values for ∆b* compared to ∆a* and ∆L*. According to Equation 3, the parameter with the highest absolute value exerts the greatest influence on ∆E*ab. The opacity was calculated as the contrast ratio (Eqn. 1). Negative values indicate a more translucent appearance after 1 month compared to the beginning of the experiment; such values were observed for Tetric under dry conditions. Additionally, wet-storage Tetric had a significantly higher ∆C as the same material under dry storage conditions (P=.041). Results similar to those for ∆E*ab were obtained for ∆b* and ∆C after 1 month from baseline. Significant differences between wet and dry conditions were observed for ∆b* (P=.011) and ∆C (P=.016). However, the negative values for ∆b* indicated a decrease over time. Negative ∆b* values indicate a less yellow and more blue appearance of the respective materials. No significant differences were found for ∆b* between the materials (P=.615 of main effect, P=.290 of interaction) under the same storage conditions. No significant differences were observed between wet- and dry-storage specimens after 1 month of light exposure with respect to ∆a* for either of the investigated materials. However, regardless of the storage conditions, ∆a* means were significantly higher for Tetric than for Silux Plus (P=.0001). All ∆a* values were positive. Positive ∆a* values after artificial aging indicate a less green and more red material appearance. All mean ∆L* values were negative. Negative values 268
of ∆L* after artificial aging indicate a darker material appearance. Silux Plus demonstrated the largest ∆L* of the investigated groups under wet conditions and the smallest ∆L* under dry conditions. Silux Plus changed more under wet conditions than under dry conditions (P=.009). No significant differences (P=.733 main effect material, P=.162 main effect dry condition) were observed for ∆L* between the other experimental groups.
DISCUSSION Clinical experience demonstrates that tooth-colored direct restoration materials change their optical properties over time and that internal discolorations often necessitate restoration replacement. Different testing procedures have been developed to artificially age restorative materials in vitro.4 Most studies include light exposure in the artificial aging procedure.6-8 UVlamps5 and high power xenon-arc lamps6 both create an intensive light with UV and infrared (IR) and are associated with xenon-specific peaks within the spectrum. Tungsten filament lamps create UV only if the lamps are high powered but have large amounts of IR. Their spectral distribution mimics blackbody radiation more than natural sunlight. The light source used in the present study emits light comparable to natural daylight at 5500 K light temperature, with some amount of UV. Because these experimental conditions caused a more intensive light exposure than found in the clinical situation, they cannot be correlated with a clinical time frame. VOLUME 87 NUMBER 3
BUCHALLA ET AL
In the present study, both of the investigated materials revealed a color shift from yellow to blue (negative ∆b*) and from green to red (positive ∆a*) under both storage conditions. All specimens became darker during the investigation (negative ∆L*). The overall color change was in a range similar to that reported in a previous study8 in which a UV-light weathering technique with high temperature in a humid atmosphere was used. The authors found an ∆E*ab of approximately 5.25 and 5.44 units for a hybrid and a microfilled composite, respectively, with no significant difference between the two. In the present study, an ∆E*ab of 5.1 for the hybrid composite and 4.1 for the microfilled composite was found under wet conditions, with no significant difference between the materials. Significant differences in ∆E*ab were found between wet- and dry-storage specimens. It has been stated that the matrix-filler interface plays a major role in the uptake of water by composites.9 Changes in optical properties within the matrices or the matrixfiller interface therefore could have been responsible for the different values of ∆E*ab. The increase of ∆E*ab was higher for the hybrid composite than the microfilled composite, but there was no significant difference between the materials. If the matrix-fillerinterface plays a major role in water uptake,10 it may be assumed that a hybrid composite that includes fine particle fillers behaves similarly to a microfilled composite. The hybrid composite tested in this study contains ytterbiumtrifluoride to provide a certain amount of fluoride release.14 The fact that this component may be water soluble also may have influenced optical properties. Given the size of the ∆E*ab values, it is questionable whether the statistical differences found in this study would be clinically significant. Reported thresholds for ∆E*ab values becoming visible to the human eye vary widely in the literature and are dependent on the individual observer. It has been claimed that the 50% threshold for perceptibility of ∆E*ab color differences is lower than 1 under perfect standardized laboratory conditions.15,16 However, under clinical conditions in the mouth, ∆E*ab color differences have been reported as relevant only when higher than 3.317 or 3.6.18 The ∆b* showed the highest color shift among ∆L*, ∆a*, and ∆b*. This parameter demonstrated the same behavior as ∆E*ab, with significant differences between wet and dry storage conditions but not between the materials tested. Under the experimental conditions of this study, it may be stated that the color shift from yellow to blue had the greatest influence on the overall color difference for all of the experimental groups. The presence of water during artificial aging with light enhanced this color change. The materials investigated MARCH 2002
THE JOURNAL OF PROSTHETIC DENTISTRY
contain diketones as polymerization initiators. Diketones absorb light of a wavelength between 420 and 450 nm (blue light). After polymerization, less diketone is present and therefore less blue light is absorbed. This was confirmed in a previous study2 in which color shifts of 5 to 10 ∆b* units were found due to polymerization. In the present study, the light-polymerization process was performed at an intense level; the amount of remaining diketones therefore should have been very low. The results suggest that other factors may play an important role in the shift from yellow to blue up to 5 ∆b* units. Given that the presence of water enhanced the blue shift of the investigated materials, it is conceivable that oxidation reactions of unreacted C=C double bonds produce colored peroxide products.3 No correlation between dry and wet storage could be detected regarding ∆a*, but the hybrid composite exhibited a significantly higher ∆a* than the microfilled composite after 1 month. The same ∆a* behavior was reported in a previous investigation of hybrid and microfilled composites.8 Both investigated materials had a darker appearance (lower ∆L*) after 1 month of artificial aging. A significant difference between wet and dry storage was detected only for the microfilled composite. Because microfilled composites consist of a greater proportion of filler-matrix interface than hybrid composites, it can be argued that the hydrolytical instability of the silane bond between fillers and the matrix was responsible for the decrease in lightness due to water storage. It should be borne in mind that a correlation between lightness and opacity may exist. When black backings are used for colorimetry, an increase in opacity may result in an increase in lightness, while a decrease in opacity may result in a decrease in lightness. In a more opaque specimen, more light will be reflected back. In a less opaque specimen, more light will penetrate the material and become absorbed by the black backing. When the results for ∆L* and ∆C were compared, a correlation between lightness and opacity could not be found. Mechanisms other than opacity may have been responsible for changes in lightness. Significant differences were detected for ∆C between wet- and dry-storage specimens, in that wet storage caused significantly greater opacity. While opacity increased for the other experimental groups, it decreased for Tetric under dry storage conditions, resulting in a significant difference between wet and dry specimens. As mentioned above for ∆E*ab, the values for ∆L*, ∆a*, ∆b*, and ∆C were within a range that probably would not be discernable in the clinical situation. It can be assumed that the color stability of both investigated materials was within the clinically acceptable range. 269
THE JOURNAL OF PROSTHETIC DENTISTRY
CONCLUSIONS Within the limitations of this in vitro study, the following conclusions were drawn: 1. Water storage caused a blue shift (negative ∆b*) and a higher ∆E*ab for both the microfilled and hybrid composite tested. 2. Water storage led to greater darkening (negative ∆L*) of the microfilled composite than the hybrid composite tested. 3. The hybrid composite tested demonstrated a larger shift toward red (higher ∆a*) than the microfilled composite tested. We thank Dr Áine Lennon for her correction of the manuscript and Ivoclar AG-Vivadent Ets (Schaan, Liechtenstein) for its support of this study.
REFERENCES 1. Mjör IA, Toffenetti F. Placement and replacement of resin-based composite restorations in Italy. Oper Dent 1992;17:82-5. 2. Seghi RR, Gritz MD, Kim J. Colorimetric changes in composites resulting from visible-light-initiated polymerization. Dent Mater 1990;6:133-7. 3. Ferracane JL, Moser JB, Greener EH. Ultraviolet light-induced yellowing of dental restorative resins. J Prosthet Dent 1985;54:483-7. 4. Asmussen E. An accelerated test for color stability of restorative resins. Acta Odontol Scand 1981;39:329-32. 5. Uchida H, Vaidyanathan J, Viswanadhan T, Vaidyanathan TK. Color stability of dental composites as a function of shade. J Prosthet Dent 1998;79:372-7. 6. Powers JM, Dennison JB, Koran A. Color stability of restorative resins under accelerated aging. J Dent Res 1978;57:964-70. 7. Powers JM, Fan PL, Raptis CN. Color stability of new composite restorative materials under accelerated aging. J Dent Res 1980;59:2071-4. 8. Fruits TJ, Duncanson MG Jr, Miranda FJ. In vitro weathering of selected direct esthetic restorative materials. Quintessence Int 1997;28:409-14. 9. Kalachandra S. Influence of fillers on the water sorption of composites. Dent Mater 1989;5:283-8.
270
BUCHALLA ET AL
10. Kalachandra S, Wilson TW. Water sorption and mechanical properties of light-cured proprietary composite tooth restorative materials. Biomaterials 1992;13:105-9. 11. Domininghaus H. Plastics for engineers. 1st ed. Munich: Hanser; 1992. p. 610. 12. Inokoshi S, Burrow MF, Kataumi M, Yamada T, Takatsu T. Opacity and color changes of tooth-colored restorative materials. Oper Dent 1996;21:73-80. 13. Commission Internationale de L’Eclairage. Colorimetry. CIE Publication No. 15.2. Vienna: Central Bureau of the CIE; 1986. 14. Buchalla W, Attin T, Kalb K, Hellwig E. [Fluoride release and uptake of an experimental composite in vitro and in situ.] Dtsch Zahnärztl Z 1998;53:707-12. German. 15. Douglas RD, Brewer JD. Acceptability of shade differences in metal ceramic crowns. J Prosthet Dent 1998;79:254-60. 16. Kuehni RG, Marcus RT. An experiment in visual scaling of small color differences. Color Res Appl 1979;4:31-9. 17. Ruyter IE, Nilner K, Möller B. Color stability of dental composite resin materials for crown and bridge veneers. Dent Mater 1987;3:246-51. 18. Johnston WM, Kao EC. Assessment of appearance match by visual observation and clinical colorimetry. J Dent Res 1989;68:819-22. Reprint requests to: DR WOLFGANG BUCHALLA GEORG-AUGUST-UNIVERSITÄT GÖTTINGEN ZENTRUM ZAHN-, MUND-, UND KIEFERHEILKUNDE ABTEILUNG FÜR ZAHNERHALTUNG, PRÄVENTIVE ZAHNHEILKUNDE PARODONTOLOGIE ROBERT-KOCH-STR. 40 37075 GÖTTINGEN GERMANY FAX: (49)551-39-2037 E-MAIL: [email protected]
UND
Copyright © 2002 by The Editorial Council of The Journal of Prosthetic Dentistry. 0022-3913/2002/$35.00 + 0. 10/1/121743
doi:10.1067/mpr.2002.121743
VOLUME 87 NUMBER 3