Influence of Polishing Systems on Surface Roughness of Composite Resins: Polishability of Composite Resins
The objective of this in vitro study was to compare, with a threshold value of 200 nm, the surface roughness obtained when using 12 different polishing systems on four different composite resins (microfill, nanofill, and two nanohybrids). A total of 384 convex specimens were made using Durafill VS, Filtek Supreme Ultra, Grandio SO, and Venus Pearl. After sandblasting and finishing with a medium-grit finishing disc, initial surface roughness was measured using a surface roughness tester. Specimens were polished using 12 different polishing systems: Astropol, HiLuster Plus, D♦Fine, Diacomp, ET Illustra, Sof-Lex Wheels, Sof-Lex XT discs, Super-Snap, Enhance/Pogo, Optrapol, OneGloss and ComposiPro Brush (n=8). The final surface roughness was measured, and data were analyzed using two-way analysis of variance. Pairwise comparisons were made using protected Fisher least significant difference. There were statistical differences in the final surface roughness between polishing systems and between composite resins (p<0.05). The highest surface roughness was observed for all composite resins polished with OneGloss and ComposiPro Brush. Enhance/Pogo and Sof-Lex Wheels produced a mean surface roughness greater than the 200-nm threshold on Filtek Supreme Ultra, Grandio SO, and Venus Pearl. Data showed that there was an interaction between the composite resins and the polishing systems. A single polishing system does not perform equally with all composite resins. Except for Optrapol, multi-step polishing systems performed generally better than one-step systems. Excluding Enhance/Pogo, diamond-impregnated polishers led to lower surface roughness. Durafill VS, a microfill composite resin, may be polished more predictably with different polishers.SUMMARY
Objectives:
Methods and Materials:
Results:
Conclusions:
INTRODUCTION
Composite resin is extensively used as a dental restorative material, as it is very conservative and has high esthetic potential. Although composite resin is used frequently, it remains a challenge to identify appropriate polishing systems to obtain high surface gloss. A smooth surface is important to prevent discoloration and plaque accumulation, which can increase caries risk and gingival inflammation.1,2 A surface roughness value of 200 nm has been established as the threshold under which bacterial adhesion could be prevented.3 Long-term success and the esthetics of composite resin restorations may be improved through proper polishing, which prevents marginal staining and discoloration.4-8 Moreover, proper polishing may preserve high surface quality and gloss over time.9
The surface quality of composite resin is influenced by several factors, including filler particle size, filler loading and resin content, type of filler, and particle morphology.10-13 Polishing success is reported to be increased when smaller particles are included in composite resin materials.14 Microfilled composite resins are known to obtain the highest gloss and surface quality because of their small particles and high resin content. However, microfilled composite resins have lower mechanical properties than universal composite resins, such as nanohybrid and nanofill composite resins.12
Several systems are available to finish and polish composite resin materials. These systems require one or multiple steps, and they differ greatly in their composition, presentation, type and hardness of abrasive particles. These differences significantly influence the surface gloss and roughness of composite resin materials.8,9,15-19 Considering that simplified systems are less time-consuming, it is important for dental practitioners to know what systems offer adequate surface quality to improve both esthetics and longevity of composite resin restorations.
Several studies9,16-25 have assessed different finishing and polishing systems using various types of composite resins. However, many of these studies limit the number of finishing/polishing systems evaluated and use discs of composite resin presenting flat surfaces. To the knowledge of the authors, there is currently no study comparing the surface roughness obtained when using several different polishing systems on convex composite resin surfaces, as is normally the case in clinical situations.
The purpose of this study was to compare the surface roughness obtained when using 12 different polishing systems commonly used in private dental practice on four different composite resin specimens with a convex surface. The main null hypothesis was that there is no difference in surface roughness between the different polishing systems tested for each composite resin. Secondary null hypotheses were that there are no differences in the surface roughness between the four composite resins tested for each polishing system and that there is no interaction among the two variables: polishing system and composite resin.
METHODS AND MATERIALS
Impressions of a VITA shade tab (VITA North America, Yorba Linda, CA, USA) were made using polyvinyl siloxane putty material (Extrude XP, Kerr Corporation, Orange, CA, USA) Specimens of each composite resin were made by placing composite resin in one increment into the mold (Figure 1). A total of 384 specimens were fabricated: 96 specimens were made from a nanofill composite resin (Filtek Supreme Ultra), 96 from a nanohybrid composite resin (GrandioSO), 96 using a second nanohybrid composite resin (Venus Pearl), and 96 from a microfilled composite resin (Durafill VS). Composite resin specifications are listed in Table 1.



Citation: Operative Dentistry 44, 3; 10.2341/17-140-L

To improve handling during finishing and polishing procedures, a lubricated flat-end screw was inserted in the back of the specimen in its cervical portion to a depth of approximately 1.5 mm (Figure 2) before polymerization. The composite resin was subsequently light-cured for 40 seconds (Optilux 501, Demetron, Kerr, Danbury, CT, USA). After the first polymerization cycle, the screw was removed, and the specimens were light-cured for an additional 40 seconds. The specimen was then removed from the putty material, and the unexposed surface was light-cured for an additional 40 seconds. The intensity of the curing light was verified periodically (after curing five specimens) using the radiometer on the unit to ensure that at least 600 mW/cm2 was delivered to the material.
To ensure uniform initial roughness, the composite resin surface was first sandblasted with 50-μm aluminum oxide particles until the surface layer appeared uniformly rough (Microetcher II, Danville, San Ramon, CA, USA). Specimens were then cleaned in 70% ethanol in an ultrasonic bath (Ultrasonic 08849-00, Cole-Parmer, Vernon Hills, IL, USA) for 2 minutes, rinsed, and dried. Finishing was simulated using a medium-grit Sof-Lex XT disc with an electric handpiece (ForZaElm, Brasseler, Savannah, GA, USA). Next, specimens were rinsed with combined air and water spray and air-dried to remove excess moisture. To further remove surface debris, impressions using a low-viscosity polyvinyl siloxane (Aquasil XLV Ultra, fast set, Dentsply Caulk, Milford, DE, USA) were taken and allowed to set for 5 minutes. These impressions were discarded.
The initial surface roughness (Ra) of each specimen was then measured with a surface roughness tester (Surftest 402, Mitutoyo, Kanagawa, Japan) using a tracing length of 3 mm and a cutoff λC of 0.25 mm. Three measurements were taken of each specimen by rotating the specimen 60°, and the average was calculated for statistical analysis (Figure 3).



Citation: Operative Dentistry 44, 3; 10.2341/17-140-L
Specimens of each composite resin type were then randomly and equally divided into 12 groups according to the finishing and polishing system used, as listed in Table 2 (n=8).

Specimens were polished by a single operator according to the polisher manufacturer's instructions regarding the speed, pressure, and need for water during the procedure (Table 3). Specimens were thoroughly rinsed with water between each polishing step. An electric handpiece was used to standardize the polishing speed, and a chronometer (Traceable timer, Control Company, Webster, TX, USA) was used to control the polishing time. The operator rehearsed and tested the protocol until the highest gloss was achieved for each polisher using extra specimens of Filtek Supreme Ultra that were discarded.

To minimize the variable of operator improvement throughout the experiment, a list of specimens placed in random group order was established using a random-sequence generator (Random.org, Dr. Mads Haahr, School of Computer Science and Statistics, Trinity College, Dublin, Ireland). The goal was to randomly polish one specimen in each group before moving forward to the second specimen.
Surface roughness was then measured with the same surface roughness tester (Surftest 402, Mitutoyo) using the same protocol as for the initial surface roughness measurements.
Statistical analysis was conducted using SAS for Windows (version 9.4, 2015, SAS Institute Inc, Cary, NC, USA). A two-way analysis of variance (ANOVA) model was used to study the effect of polishing system and composite resin on surface roughness. The model was adjusted for the roughness measurements before and after polishing. Pairwise comparisons were made using protected Fisher least significant difference. A p-value of less than 0.05 was used as a criterion for statistical significance. Residual analysis was performed to verify the normality and the homogeneity of the variance assumptions.
RESULTS
Descriptive statistics for initial surface roughness and final surface roughness measurements are presented in Figures 4 and 5. Figure 4 shows that the initial surface roughness is greater than the 200-nm threshold for all composite resins tested and that Grandio SO and Venus Pearl have an overall lower surface roughness value. Figure 5 depicts that the highest final surface roughness for each composite resin was observed with OneGloss and ComposiPro Brush. Enhance/Pogo and Sof-Lex Wheels produce a surface roughness below the 200-nm threshold on Durafill VS, whereas the surface roughness was above the threshold on Filtek Supreme Ultra, Grandio SO, and Venus Pearl.



Citation: Operative Dentistry 44, 3; 10.2341/17-140-L
Results of the two-way ANOVA are presented in Table 4. For initial surface roughness, only the main effect of composite resins was significant (p<0.0001), meaning that the mean initial surface roughness measurements between the composite resins tested were statistically different. Furthermore, differences observed between composite resins were the same for all polishing systems, since the interaction term was not significant (p=0.34). Pairwise comparisons of composite resins are presented in Table 5, which shows that Filtek Supreme Ultra had the highest initial surface roughness and Grandio SO the lowest.


Regarding the final surface roughness, results of the two-way ANOVA revealed that the interaction between composite resins and polishing systems was statistically significant (p<0.0001). This indicated that the final surface roughness obtained with different polishing systems was not the same for each composite resin tested. Table 6 illustrates the pairwise comparisons of different polishing systems for each composite resin as well as the pairwise comparisons of different composite resins for each polishing system. It may be concluded from this table that the polishing systems leaving the smoothest surfaces for each composite resin are not the same. Durafill VS, D♦Fine, Optrapol, and HiLuster Plus produced the lowest surface roughness. However, except for OneGloss and Composipro Brush, most systems reached a mean surface roughness below the 200-nm threshold, although some standard deviations were slightly higher. For Filtek Supreme Ultra, Super-Snap achieved the smoothest surface, but Astropol, HiLuster Plus, D♦Fine, Diacomp, and OptraPol also obtained mean surface roughness value less than 200 nm. For Grandio SO and Venus Pearl, the lowest surface roughness was obtained when using Super-Snap, OptraPol, and Astropol adding ET Illustra for Grandio SO and Sof-Lex discs for Venus Pearl.

DISCUSSION
Composite resin surface quality is important to optimize esthetics and longevity of restorations. The best surface quality with the lowest surface roughness has often been obtained with a composite resin cured against a Mylar strip.17-19,26,30,31 However, this surface has a resin-rich layer and presents a lower hardness. To prevent wear and discoloration, it is suggested to finish and polish this surface.32 This study evaluated the effect of different commercially available polishing systems on the surface roughness of composite resins.
The results of the present study suggest that a single polishing system does not produce the same surface quality for all composite resins. This may not be entirely attributed to the quality of the polishers but to the interaction between the polisher and the composite resin. This is in accordance with the findings of previous studies.9,16,19 Several factors have been proposed to affect the polishability of composite resin, including the polishing system, the composite resin used, and variables associated with the operator.
Traditionally, ideal polishing protocols have been explained as a selective wear process using a sequence of abrasive particles from coarse grit gradually decreasing toward fine grit.5,15 Currently, a variety of polishing systems are commercially available. Some systems require multiple steps, whereas others are simplified and require only one grit used with gradually decreasing pressure. The hardness of the backing or rubber media into which the abrasive particles are embedded influences the surface quality.15 The hardness of the abrasive particles vary and may be classified as follows according to the Mohs's hardness scale: diamond > silicon carbide > tungsten carbide > aluminum oxide > zirconium silicate.15 The hardness and the size of abrasive particles are very important. First, the abrasive particles must be harder than the filler particles present in the composite resin to avoid abrading only the resin matrix and leaving the filler particles protruding. Second, the abrasive particles must be small to prevent scratches on the composite resin. Multi-step systems use smaller particles for each step to remove scratches from the previous polisher until a highly shined surface is obtained. For one-step systems, the grit size is important because it may leave scratches on the composite resin. Some studies reported that multi-step polishers perform better than one-step polishers.9,16 Indeed, in the literature, it may be found that Sof-Lex discs produced higher gloss along with Astropol,8,9,16 whereas brushes produced high surface roughness.16 On the contrary, some studies reported that the Pogo system, used as a one-step or a two-step system, showed the highest gloss, whereas aluminum oxide discs produced a poorer surface finish.16-19 In the present study, data showed that following final polishing, most polishing systems obtained a clinically adequate surface roughness of less than 200 nm. However, OneGloss and ComposiPro Brush, two simplified one-step polishing systems, were unable to reach an acceptable surface roughness and left roughness significantly above the threshold for all the composite resins tested. The final surface roughness obtained with OneGloss was even higher than before polishing procedures. Of the one-step systems, only Optrapol had a surface of less than 200 nm. Therefore, multi-step systems generally provided a better surface finish, a finding that is in partial agreement with the current literature. Although the Enhance/Pogo system showed good results in a previous study, the present data revealed that it left surface roughness above the 200-nm threshold on Filtek Supreme Ultra, Grandio SO, and Venus Pearl. Therefore, except for the Enhance/Pogo system, polishing systems containing diamond particles produced, in general, a superior surface finish, which is consistent with previous studies.15,16 Super-Snap, which contains silicon carbide and aluminum oxide particles, also left an excellent surface roughness, especially with Filtek Supreme Ultra. The Sof-Lex Wheels left a surface roughness greater than 200 nm on Filtek Supreme Ultra, Grandio SO, and Venus Pearl. The Sof-Lex Wheels used in the present study contained only aluminum oxide particles. Since the study was performed, the manufacturer modified the composition of the final polisher by replacing aluminum oxide with diamond particles. This would have probably altered the efficacy of this system.
Factors related to the composite resin also influence surface quality. These include the resin matrix content and formulation, the filler particle characteristics (type, hardness compared with the abrasiveness of the polishers, size, and shape), the composite resin filler load, the quality of the silane coupling agent, and the degree of conversion after light curing.33-36 The matrix and filler particles have different hardness, which may influence the polishability of the composite resins. Insufficient abrasiveness of the polisher particles compared with the composite resin fillers will mostly abrade the matrix, leaving the filler particles in protrusion. In addition, insufficiently bonded fillers may debond and dislodge, leaving a dull surface. Therefore, the results suggest that the combination of composite resin and polisher has an influence on the result, with some polishers leaving an excellent finish on some composite resins but a less optimal finish on others. It is well known that smaller particles reduce the surface roughness after polishing procedures.34 It has also been reported that spherical particles allow for a better light reflection than irregular particles.36 It has been suggested that composite resin should be polished with the polishing system of the same manufacturer,37 an assertion that could not be confirmed in the present study. The result of the present study reveals that, although not statistically different for all polishing systems used, Durafill VS, a microfill composite resin containing small particles and high resin content, showed less variability among the polishing systems. Filtek Supreme Ultra, a nanofilled composite resin, was reported by the manufacturer to have a unique formulation, in which nanosized particles were individually silanized38 and agglomerations or nanoclusters that seem to resist particle loss during the polishing procedure, leaving a more uniform surface with less roughness.33 The results of this study generally confirm this, except when using some polishers (Enhance/Pogo and Sof-Lex Wheels).
A high surface roughness has been found with Grandio in some studies.8,9 However, compared with the other composite resins tested in the present study, this high surface roughness was not observed. Although the mean surface roughness was close to the threshold for some polishers (Sof-Lex discs and Diacomp) or above the threshold for some others (Enhance/Pogo, Sof-Lex Wheels, OneGloss, and ComposiPro Brush), many polishing systems left a mean surface roughness less than 200 nm. This may be due to the recent improvement in the Grandio SO formulation to obtain a better surface finish. Venus Pearl also seemed to adequately polish, with most polishers tested except with Sof-Lex Wheels and Enhance/Pogo.
An interaction was found in the present study, meaning that the surface roughness depends on the combination of polishing system used and composite resin. This is in accordance with the results of previous studies comparing polishing systems and composite resins.18,26
Possible bias may be attributed to operator variables such as the polishing time, the speed of the handpiece, the pressure applied to the composite resin, the hand skill improvement, and the experience of the operator. Heintze and others reported that surface roughness and gloss are time dependant, with the greatest improvement after five seconds and continued improvement for up to 30 seconds for each of the steps of the Astropol system (three-step polishing system).39 In the present study, the operator, in accordance with the manufacturer's instructions, established the best polishing protocol (pressure, speed, and time) for all polishers to obtain the highest surface quality possible. These specimens were not included in the study. In the present study, specimens were polished for a minimum of 30 seconds for each step controlled with a chronometer (Table 3). Establishing this protocol also allowed the operator to rehearse prior to the study to obtain a standardized pressure while polishing the composite resin specimens. The speed of the handpiece was standardized using an electric handpiece. All steps were performed by the same rehearsed operator, who was a trained second-year dental student. According to Zimmerli and others, the age and experience of the operator do not seem to influence the surface quality of the composite resin after polishing procedures.40 However, to control for the variability of improvement in hand skills throughout the study, specimens were polished in random order.
Previous studies evaluating the efficacy of polishing protocols in terms of surface roughness and gloss are mostly performed on discs of composite resin.8,9,18,26,29,41,42 However, in clinical situations, composite resin is often placed in convex morphology, which may influence the result of polishing procedures, and this is the reason why convex specimens were used in the present study. Although disks allow gloss measurements, only surface roughness was measured in the present study because it was impossible to obtain reproducible gloss values due to specimens' convexity.
To establish an initial roughness for all the composite resins tested, specimens were first roughened using the same sequence of sandblasting (50 μm aluminum oxide particle) followed by a medium-grit Sof-Lex XT disk indicated for excess composite resin removal and recontouring before polishing. Although the same procedure was applied to all specimens, the initial roughness varied among the four composite resins tested. However, these differences did not seem to influence the final surface roughness results.
Many different factors may affect polish retention over time. It has been shown that the surface may be altered by bacterial biodegradation.43 Surface quality may also be influenced by alcohol and acidic solution exposure.44-49
The results of the present study should be interpreted with caution and may not apply to other composite resins, polishing instruments, or polishing techniques. Clinical studies evaluating the effect of polishing quality on the longevity of composite resin restorations would be relevant.
CONCLUSION
The null hypotheses were rejected, since surface roughness was influenced by the polishing system and the composite resin tested. In addition, there was an interaction between the polishing systems and the composite resins. Indeed, a given polishing system does not perform equally with all composite resins. The results of the present study suggest that, except for Optrapol, multi-step polishing systems performed generally better than one-step systems. Moreover, excluding Enhance/Pogo, diamond-impregnated polishers allowed for lower surface roughness. In addition, Durafill VS, a microfill composite resin, may be polished more predictably with different polishers.

Impression of a Vita Shade guide in which composite resin was inserted in one increment.
Figure 2 A flat-end screw was inserted in the back of the specimen in its cervical portion to a depth of approximately 1.5 mm for handling purposes.

Three surface roughness measurements were taken by rotating the specimen at an angle of 60°.

Initial surface roughness measurements after sandblasting procedures and medium grit Sof-Lex XT disc.
Figure 5 Comparison of composite resin final surface roughness measurements after polishing with different systems.
Contributor Notes