Clinical Significance: Sof-Lex discs, showed superior results for surface finishing of glass ionomer cement, while the diamond bur reveled the poorest results.

Glass ionomer cement use as a restorative material has increased due to its adhesion to the dental structure, ^1,2 thermal-linear expansion coefficient being similar to that of the tooth³, biocompatibility³, and fluoride release in the oral cavity^4,7 which provides an anticariogenic effect^8-10.

Although they are highly sensitive to restorative techniques¹¹, especially regarding surface finishing these cements have shown good clinical performance^11,13,14.

An adequate anatomic contour of the restoration is hardly achieved only by using the Mylar strip¹⁵. Consequently, finishing and polishing become essential since they reduce dental plaque adhesion¹⁶ and secondary caries risk¹⁷, thus, maintaining gingival health¹⁸ and increasing the restoration clinical longevity^19,20. They also allow for less color alteration^21,22 which is an important esthetic consideration.

This study evaluated a glass ionomer cement surface roughness and staining, after subjected to different surface finishing and polishing treatments.

Chem-Fil Ii^a glass ionomer cement, and, as abrasive systems, Sof-Lex^b discs, Enhance^a points, and KG Sorensen (3168 FF^c) points were selected to perform this study. Sixteen 13mm diameter round perforation in their internal surface in order to permit the making of cylindrical specimens with those dimensions, were also used.

Metallic matrices were placed on glass slabs for microscopy. Glass ionomer cement was then mixed at 1:1 powder/liquid ratio, for 20 seconds, according to the manufacturer’s recommendations. To avoid air entrapment, the mixed cement was inserted in the matrix with a Centrix^d syringe. Polyester strips^e were then placed over the initial slabs, and other glass slabs for microscopy ware laid on top of them. A 500-g axial load²³ was placed over them for 10 minutes, counting from the beginning of mixing. After the load had been removed. each set was left in the humidifier, at 37 ± 1^oC, for 24 hours. After that period, specimens were taken from their matrices and covered by two layers of colorless nail polish^{f 24}. After 10 minutes, specimens were randomly divided into four groups: Group 1: Mylar strip, Group 2: Sof-Lex discs; Group 3, Enhance points; Group 4 FF diamond points. Group 1, the control group, did not receive any finishing, while the other groups received surface treatments on just one of their sides.

For each stage of the procedure, 16 specimens were made. Half of them were used in the surface roughness test, and the other half in the staining test.

Roughness evaluation - After the surface finishing treatment, specimens were washed in distilled water for 15 seconds and dried with absorbent paper. Afterwards, they were submitted to a profilometer^g reading, taking into consideration the arithmetic mean between peaks and recesses (Ra)²⁵. Three readings were performed in different traces of each specimen surface, totaling 96 readings.

Staining evaluation - Serra’s²⁴ method was employed to quantify the staining. After receiving the treatments previously described, the specimens were immersed into a 2% methylene blue aqueous solution for 3 minutes²³. They were then washed in distilled water, dried with absorbent paper, and placed into test tubes containing 3 ml of nitric acid^h at 65% for 24 hours. After, the specimens were dissolved by the nitric acid, the test tubes were centrifuged for 3 minutes at 4,000 rpm, and the supernatant was used to determine the absorbance in a spectrophotometer.

Using 0, 2, 4, 6, 8, and 10 m g methylene blue standard solutions, and a cement sample in 1 ml of nitric acid at 65%, a calibration curve was elaborated before quantifying the dye impregnated in the samples.

Through the calibration curve standard solutions, the maximum peak of absorbance was reached in the 585-nm wavelength, which was used for all the specimens readings during the staining test.

Surface roughness - Data obtained by the profilometer reading for each of the studied treatments were submitted to ANOVA for totally random design²⁶. The estimated F value was 20.09 showing significant differences at a 5% level. Tukey’s test was performed to better explain individual comparisons²⁶ (Table 1).

Tukey’s test results showed significant differences between the control and the other groups. There was no significant difference in all other comparisons between two means.

Staining test - After obtaining the standard calibration curve, the correlation coefficient (r), whose value equaled 0.9985, was calculated; therefore, proving the efficiency of the used methodology.

To estimate dye concentration in the samples, a linear regression analysis, shown by the following equation, was performed.

The results obtained from the spectrophotometer direct reading for each absorbance treatment were converted into dye concentration and submitted to ANOVA for a totally random design²⁶. The F value was 14,92, showing statistical differences at a 5% level. Tukey’s test was applied to the individual comparisons²⁶ (Table 2).

Tukey’s test results revealed significant differences between the following groups: control and Enhance points; control and FF diamond point; and Sof-Lex discs and FF diamond point. All other comparisons between two means showed no significant difference.

Correlation between surface roughness and staining - The linear correlation study between the surface roughness test (X) and the staining test (Y) presented the X mean equal to 1.4863, with a standard deviation of 0.6773, and the Y mean equal to 2.9678, with a standard deviation of 0.9651. The regression ANOVA with totally random design³⁴ presented an F equal to 22.08, significant at a 5% level. The regression Student t-test result was equal to 4.6990, significant at a 5% level, demonstrating that staining is linearly dependent on the surface roughness.

Longevity of a glass ionomer cement restorations and the surface obtained after surface finishing are closely related to the material’s setting time, the type of instrument used for finishing, and the amount of moisture contamination or dehydration which occurs during the finishing process²⁰.
 

Table 1. Means (± S.D.) surface roughness of glass ionomer samples subjected to various surface treatments.
 

Treatment	Mean (± S.D.)  m m
FF diamond bur  Enhance  Sof-Lex  Control	2.20 (0.38) c  1.74 (0.38) b c  1.30 (0.44) b  0.71 (0.40) a

dms=0.55 a =0.05
Means with the same letter were not significantly different.
 

Table 2. Means (± S.D.) of dye concentration in glass ionomer samples subjected to various surface treatments.
 

Treatment	Mean (± S.D.)  m m
FF diamond bur  Enhance  Sof-Lex  Control	2.20 (0.38) c  1.74 (0.38) b c  1.30 (0.44) b  0.71 (0.40) a

dms=0.86 a =0.05
Means with the same letter were not significantly different.
 

Due to the fact that restorative glass ionomer cements have a slow setting time, they show a more homogeneous surface²⁹ when the finishing is performed 24 hours after the restoration is completed. On the other hand, the rapid setting time can only be achieved at the expense of color and translucency so that, if a restorative aesthetic cements is to be utilized, to obtain optimum results it is necessary to protect the setting cement against water uptake for some hours after placement. In certain materials, physical properties at 15 minutes may be sufficient to be able to contour and polish the newly placed restorations. However, if disturbed at this point there will be sufficient water uptake to reduce the translucency to unacceptable levels, as well as to lower the physical properties and the attachment to dentin⁷. Thus, for this study, surface finishing was done 24 hours after the beginning of mixing.

The aluminum oxide discs produced the smoothest surface of any finishing techniques for both resin composites^15,18,27 and glass ionomer cements ^20,28,30. However, the anatomical from of the tooth does not allow the discs clinical usage in all its surfaces; so disc-shaped rotary instruments seem appropriate to be used on flat surfaces. In concave areas, different shapes of abrasive instruments are necessary. An extra thin cut diamond point and the Enhance surface finishing points were also included in the study.

All surface finishing procedures were made within 30 seconds under water cooling^29,30 to avoid shrink and case of the cement^31,32.

The surface roughness results of this study showed that the best finishing was obtained when the glass ionomer cement was set in contact with the matrix. This corroborates previous studies²⁰ and strengthens the idea that, although difficult, restorations should be done with a minimum of excess.

Among the abrasive materials used for surface finishing, aluminum oxide discs presented the lead surface roughness mean (1.30 m m). This indicates that the choice for the abrasive instrument which best suits the surface to be polished can be made depending on the clinical situation, and considering the anatomy of the several surfaces of a tooth. The highest mean of surfaces roughness was presented by the FF diamond point (2.20 m m), which showed significant difference from the control and the aluminum oxide discs groups. According to Pearson²⁰, the relatively rough surfaced diamond burs produce a glass ionomer cement surface which is irregular and rather undulating. Some disturbance of the underlying substructure occurs by destruction of the polygel matrix rather than the alumino-silicate glass. This may result in the breakdown of the restoration in the long term. Another disadvantage of the diamond point is that it leaves metallic marks when the restorative material has quartz macro-particles³³.

Some foods contribute to color alteration of the silicate cement and resin composite, and the same may occur to the glass ionomer cement. Paulillo et al²³ showed that fully hardened restorative glass ionomer cements have almost no color alteration, but they did not evaluate the surface finishing effect on the material’s staining.

Most in vitro studies have not quantified the staining of esthetic restorative materials, which is a subjective analysis. The quantitative method of measuring staining, which has been used for microleakage evaluation, was chosen because it allows the quantification of dye penetration into the material. Consequently, in vitro evaluation of the esthetic characteristics of the material is more accurate, and therefore makers it possible to use a more powerful statistical test²³.

The staining test demonstrated that spectrophotometry is an efficient method to measure the amount of dye retained in surface irregularities produced during finishing and polishing. That test results indicate that there was no significant statistical difference between the dye concentration in the control, that is, Mylar strip (1.92 m g/ml) and the Sof-Lex aluminum oxide discs (2.72 m g/ml) groups. This may have occurred due to presence of air on the cement surface^29,30,36. The highest standard deviation in the control group seemed to be due to the presence of microbubbles. Surface finishing done with Sof-Lex discs removed the surface microbubbles, leaving the surface smoother and with a lesser amounts of irregular particles²⁹, consequently, presenting a lower standard deviation. The other materials used on surface finishing, the Enhance points and FF diamond point, probably would also remove those surface microbubbles. However, those materials did not promote the same surface smoothness obtained with Sof-Lex discs, showing a significant difference from the surface obtained with the Mylar strip. Surfaces treated with FF diamond points showed a dye concentration significantly higher (3.97 m g/ml) than those treated with Mylar strip and Sof-Lex discs, proving that FF diamond points produce an undulating and irregular surface²⁰, favoring staining. The same test also showed that there was no statistically significant difference between the Sof-Lex discs and the Enhance points (3.22 m g/ml); however, the Enhance points presented a significant difference in relation to the Mylar strip, probably because it produces a slightly rougher surface than those of Sof-Lex discs.

When surface roughness and staining tests are compared, it was verified that there is a statistical correlation between the two tests. Except for the fact that the means of the Mylar strip and the staining test, it could be noted that the other results were similar. Thus, the surface finishing that produced a rougher and more irregular surface, also presented a higher amount of dye on the surface, demonstrating that both test are linearly dependent. All the materials used to remove the excess material during the finishing did not leave the surface the same as that obtained when glass ionomer cement was set in contact with the Mylar strips, which means that none of the materials used for finishing and polishing provided a final surface with the same initial smoothness level²⁸.

Legends

a. De Trey Division, Dentsply Ltd., Weybridge, UK.

b. 3M do Brazil Ltd., Sumaré, SP, Brazil

c. K.G. Sorensen Industry e Commerce Ltd., São Paulo, Brazil.

d. DFL Industry e Commerce, Rio de Janeiro, Brazil.

e. Odahcam, Herpe Produtos Dentários Ltd., Brazil.

f. Colorama, Ceil Coml. Exp. Ind. Ltda., São Paulo, Brazil.

g. Prasis Rug 03, Aro S/A, Argentina.

h. Merck S/A Industrias Químicas, Rio de Janeiro, Brazil.

i. Beckman DU-65, Fullerton, CA, USA.

Acknowledgments: We would like to thank Dr. Lúcio B. Kroll from the Universidade Estadual Paulista - UNESP (Botucatu Campus) for his statistical analysis, and his guidance on statistics.

Drs. Paulillo, Lovadino, and Serra are Professors, School of Dentistry, Campinas State University, Piracicaba, São Paulo, Brazil; Dr. Coradazzi is Professor, School of Dentistry, São Paulo State University, Bauru, São Paulo, Brazil.