Problems
Associated With Resin Curing
Resin polymerization problems — Are
they caused by resin curing lights, resin formulations, or both?
| Contents: | Abstract | Discussion | List of Tables 1-3 |
| Introduction | Conclusions | Figures 1-6 | |
| Materials & Methods | References | Figures 7-12 | |
| Results | Footnotes | Figures 13-18 |
Possible negative effects of rapid, high intensity resin curing have been questioned since argon lasers and plasma arc curing lights were introduced. To address questions, six different resin restorative materials were cured with 14 different resin curing lights representing differences in intensities ranging from 400 to 1900 mW/cm2; delivery modes using constant, ramped, and stepped; cure times ranging from 1 to 40 seconds; and spot sizes of 6.7 to 10.9 mm were tested. The lights included two lasers, five plasma arc lights, and seven halogen lights. Tests were performed to measure shrinkage, modulus, heat generation during curing, and strain during polymerization. Physical changes on the teeth and resins during strain testing were documented photographically and graded. Results of shrinkage, modulus, and strain tests showed consistently that effects associated with lights were not statistically significant, but resin formulation was highly significant. The two microfil resins had least shrinkage and lowest modulus, and hybrid resins had highest shrinkage and modulus. The autocure control resin shrinkage and modulus was as high or higher than the light cured hybrid resins, even though no light was used for its polymerization. The lasers and plasma arc lights produced highest heat increases on the surface (up to 21 degrees Celsius) and within the resin restorations (up tp 14 degrees Celsius), and the halogen lights produced the most heat within the pulp chamber (up to 2 degrees Celsius). Strain within the tooth was least with HeliomolarTM and greatest with Z100TM. Bisfil IITM autocure resin produced highest overall strain. Clinical effects of strain relief were evident as "white lines" at the tooth-resin interface which were visible clinically and cracks in enamel adjacent to the margins which were visible microscopically. Clinical significance of the "white lines" and cracks is unknown. However, they may be contributing to the post-operative sensitivity reported often by clinicians using resin restorative materials. Results of this work implicate resin formulation as the important factor in polymerization problems. Lower light intensity and use of ramped and stepped energy delivery did not provide significant lowering of shrinkage, modulus, or strain within the tooth during polymerization, and did not prevent enamel cracking adjacent to margins and formation of "white line" defects at the margins. Until materials with lower shrinkage and modulus become available, use of low viscosity surface sealants as a final step in resin placement is suggested to seal defects.
Resin restorative materials shrink during polymerization.1-5 This shrinkage causes gaps to form at the resin-tooth interface1-4 where a contiguous sealed junction is desired. It has been proposed that a longer gel phase during the resin’s set might minimize or eliminate the gap formation. To gain a longer gel phase, several investigators have suggested use of multi-mode resin curing lights, where an initial cure with low intensity light is followed by final cure with high intensity light.2,3,6 This technique has been termed "soft polymerization" or "soft start polymerization". A several minute wait between the low and high intensity irradiation phases has also been proposed to try to further extend the gel stage.6 The "soft start" techniques intentionally prolong total resin curing time which, over the years, has become more involved and time consuming as more steps have been added. Today, light curing steps can include curing the adhesive, a liner or base, multiple resin layers, exposure from various directions, and a surface sealant. It has been estimated that clinicians placing about 10 three surface Class II resin restorations per practice day, and practicing about 240 days per year spend about 30 eight-hour days per year holding a light waiting for polymerization to be completed.7
Other investigators have proposed use of high intensity curing lights, such as argon lasers or plasma arc lights, to achieve polymerization quickly.8 Although clinicians have displayed interest in the fast cure methods to speed the light cure process, they have been warned that increased resin brittleness and shrinkage could result due to the rapid, profound cure possible with lights two to five times more intense than curing lights of the past.9 Yet others have reported that high intensity curing lights do not increase polymerization shrinkage1,7 or modulus.7 It has been suggested that the very short cure times used with high intensity lights may mitigate the predicted negative effects.1,7
Since an increasing number of resin restorations are being placed, it is important for clinicians to seek the best possible polymerization methods. The challenge is to optimize resin physical properties while speeding the cure process and minimizing any negative post-operative effects.
Goals of this ongoing study are:
- To test different resin curing light designs with a variety of resin formulations and determine cure times, shrinkage, modulus, and heat generation, in vitro.
- To use the resin curing lights and resin formulations with extracted human teeth and quantitatively measure polymerization strain, and document effects to the teeth and/or resin restorative materials that become evident during the testing.
Materials and Methods
Resin Curing Light Characterization
Table 1 lists 14 resin curing lights tested, and pertinent information about each light. To characterize light output, light intensity was determined with an integrating spherea, research radiometera, and the Curing Radiometer model 100Ô,b handheld radiometer. A spectral curve was generated for each light using a scanning monochromatora. Table 2 lists six resins used in tests described below, and some of their relevant characteristics.
Cure Times
Minimum cure times were determined for various colors of five light cure resins listed in Table 3 by placing resin into a steel mold 2 mm deep x 5 mm diameter, covering it with a mylar strip, and light curing against the mylar. Cure times shorter than the curing light manufacturer’s recommended time were used first, and Barcol hardness was measured, top and bottom, using the Barcol GYZJ 935Ô,c hard material tester. In this study, cure was defined as hardness at a 2 mm depth that was ³90% of the top surface hardness and greater than a Barcol hardness of 80. The previous steps were repeated with additional samples, adjusting cure time up or down until the minimum time was determined that produced the cure defined above. Cure time tests using the above steps were performed in triplicate. Statistical analyses were performed on the measurements.
Effects of exaggerated cure times to resins alone and resins within extracted teeth were also investigated by measuring shrinkage and modulus using the methods described below at various cure times and by photographing test restorations in teeth at 12x after curing at minimal, moderate, and exaggerated times. All five of the light cure resins listed in Table 2 and the following seven resin curing lights from the list in Table 1 were used: AccuCure 3000Ô, Apollo 95 EÔ, BriteSmileÔ, Elipar HighlightÔ, Optilux 401Ô, Optilux 500Ô, and the PAC LightÔ.
Shrinkage
Shrinkage was determined using a mercury dilatometerd. Resin samples were placed on a glass slide and mass was determined. The slide was clamped to the funnel on the dilatometer, and the funnel was filled with mercury. A linear variable displacement transducer was positioned onto the mercury column. The resin was light cured using the light or resin manufacturer’s recommended times, and after a 60 minute wait, the curing light was again activated for the full cure time to provide thermal correction. The sample was removed and sample density was measured with a pycnometer. Percent shrinkage was calculated and statistical analyses were performed on 114 measurements using the 10 resin curing lights listed in Figure 7 and the six resins listed in Table 2.
Modulus
Modulus was determined according to the methods outlined in ISO Specification
#4049:1988(E) entitled "Dentistry — Resin-Based Filling Materials". Three point bending tests were performed on an MTS 810 testing machinee, and statistical analyses were performed on data from 353 tests using the 12 resin curing lights listed in Figure 9 and the six resins listed in Table 2.
Heat Generation
To determine heat generated at the resin-tooth surface, within the resin restoration, and within the pulp chamber, the root canal in one root and the pulp chamber were excavated in a human lower first molar. A type k thermocouple was inserted into the pulp chamber and the chamber was sealed to prevent water leakage through the entry hole. A slot preparation was cut into the distal proximal surface of the tooth (3 mm x 3 mm x 2 mm deep). Z100Ô was placed into the preparation and a second thermocouple was inserted into the center of the restoration at a depth of 1.8 mm. A third thermocouple was attached on the restoration surface at its center. The tooth was placed into a water bath at 37°C and submerged to the CEJ, leaving the enamel crown and restoration exposed to ambient air at 21°C. After the tooth temperatures stabilized, initial temperatures were recorded at the 3 thermocouple locations. The light guide of each of the 12 resin curing lights listed in Figure 12 was then centered 1 mm above the restoration and activated for the times indicated within the parentheses of labels shown in Figure 12. Temperatures were recorded and statistical analyses were performed.
Polymerization Strain Measurement
To monitor development of strain within the tooth as resin polymerization progressed, a strain gauge was mounted just below the gingival margin of a slot preparation in the proximal surface of freshly extracted molars (See Figure 1). Eighty-five freshly extracted human molars were cleaned with flour of pumice slurry on a prophy angle, and one proximal surface was selected for the preparation on each tooth. Some of the teeth were stained in erythrosin B dye, and polished again with flour of pumice slurry to remove residual surface stain and emphasize stain within the pre-existing cracks in the enamel. The teeth were photographed at 12x and 20x on a stereo microscope to document locations of pre-existing cracks. Proximal slot preparations with all margins in enamel were cut using a new #56 six-bladed burf for each tooth. Preparation dimensions were 3 mm from the central groove towards the CEJ, 3 mm wide (buccal-lingual), and 2 mm deep. After preparations were cut, teeth were photographed to document locations of cracks caused by cutting. An EA-06-062AP-120 strain gaugeg was seated about 0.5 mm below the gingival margin, with all parts on enamel, and aligned to measure strain in the buccal-lingual direction. Molten wax was applied to the surface of the gauge to seal it from ingress of water. The preparation was etched for 15 seconds with 32% phosphoric acid and One-StepÔ,h adhesive was applied. Resins were placed into each preparation, with and without layering and with slight overfill. Light cure resins used in this test were HeliomolarÔ, TPH SpectrumÔ, and Z100Ô. Bisfil IIÔ autocure resin was included as a control. Seven lights were included in these tests (AccuCure 3000Ô, Apollo 95 EÔ, BriteSmileÔ, Elipar TriLightÔ, Optilux 401Ô, PAC LightÔ, and VIPÔ). The strain gauge was connected to a data acquisition system set to record data 5 times per second. Three minutes after curing, the tooth was placed into a water bath at 37°C to simulate clinical conditions. The restorations were finished using a new #7406 12-bladed burf and Sof-LexÔ fine and superfine disksi. The tooth was again photographed to identify and document cracks and/or "white lines" that may have formed during resin polymerization. Buccal, lingual, and gingival margins of 79 treated teeth were graded for presence and magnitude of cracks and/or "white lines" by four investigators. Teeth were kept hydrated throughout the entire above process, except during actual manipulation.
Statistical Analyses
Statistical analyses were performed on data from all tests conducted to determine effects of curing light and resin. Analysis of variance (a = 0.05) was used with Tukey’s multiple comparison procedure to determine significance. Logistic regression was used to determine the effect of strain on types of defects produced in the extracted teeth.
Resin Curing Light Characterization
Figures 2 and 3 show spectral curves for lasers and plasma arc lights, and halogen lights, respectively. The curves illustrate the relative output of the laser, plasma arc, and halogen lights. Figures 2 curves show the intensity output of lasers and plasma arc lights is considerably higher than halogen light curves in Figure 3. Figure 3 curves also show the similarity of the output of various brands of halogen lights. The very high intensity output of the lasers and plasma arc lights at wavelengths critical to the chemical initiator enables their rapid cure of current generation resins. The curves also show why lights in the same category designation can perform differently from one another. For example, the BriteSmileÔ and AccuCure 3000Ô are both designated as argon lasers, but their different intensities and wavelengths illustrated in Figures 2 affect clinical performance in areas such as cure time and heat generation.
Cure Times
Table 3 displays the minimum times required to cure 2 mm thick resin samples. Times in the table are ideal times and do not provide any margin of safety to adjust for less than optimal clinical conditions. Times range from as short as one second for the Apollo 95 EÔ plasma arc light used to cure light shades of Z100Ô and TPH SpectrumÔ hybrid resins, to 40 times longer for the medium intensity control halogen light used at 400 mW/cm2 to cure a darker shade microfill resin. In general, regardless of the curing light used, Z100Ô required shortest cure times and HeliomolarÔ required longest cure times, and darker shades of the same resin formulations required longer cure times.
Results of tests where cure times were significantly extended are shown in Figure 4. Samples of three separate teeth restored with Z100Ô and cured with the Apollo 95 EÔ plasma arc light show increasing amounts of damage from "white lines" at the margins to relatively pronounced cracking around margins due to cure times increased from 1 second to 40 seconds. Many clinicians significantly overextend cure times in an effort to maximize resin physical properties, not realizing they can be affecting the surrounding tooth negatively. Figure 5 shows results of exaggerated cure times on shrinkage. The majority of the shrinkage occured within the first few seconds. Extended cure times produced little or no additional shrinkage. Figure 6 shows results of exaggerated cure times on modulus. Modulus increased rapidly during first few seconds of cure, then more slowly afterwards. Extended cure times produced some additional increase in modulus.
Shrinkage
Figures 7 and 8 show results of mercury dilatometer tests using ten curing lights and six resin formulations. Statistical analyses showed lights with different modes of delivery (constant, ramped, and stepped) and different intensities (ranging from 400 to 1900 mW/cm2), different cure times (6 seconds to 40 seconds), and spot sizes (6.7 mm to 10.9 mm) produced statistically similar shrinkage (p = 0.9757). However, shrinkage associated with resin formulation showed significant differences (p = 0.0001). HeliomolarÔ and Silux PlusÔ microfill resins had statistically similar shrinkage, but were statistically different from the other three light cure and one autocure hybrid resins. Overall, these data indicate that the important factor in resin shrinkage is resin formulation, not curing light design.
Modulus
Figures 9 and 10 show results of three-point bending tests for modulus using 12 curing lights with the same six resins listed above for shrinkage. Again, statistical analyses showed no differences among the lights (p = 0.9950), but resin formulations resulted in strong differences
(p = 0.0001). The six resins formed five statistical groupings. The two microfill resins HeliomolarÔ and Silux PlusÔ had lowest modulus, but were statistically different from each other, and from all the other resins tested. Bisfil IIÔ autocure resin had a modulus that was significantly higher than all the light cure resins tested. These data indicate that the important factor in resin modulus (stiffness) is resin formulation, not curing light design.
Heat Generation
Figure 11 shows the location of three thermocouples implanted to record temperatures at various locations in a tooth with a resin restoration ready for curing. Tests were conducted on the resin surface, 1.8 mm within the resin, and within the pulp chamber. Figure 12 shows the location of greatest temperature change caused by curing lights was the resin surface. Bisfil IIÔ autocure resin produced the least heat at the resin surface. BriteSmileÔ argon laser and all five plasma arc lights generated higher temperatures at the resin surface compared to the halogen lights. However, in the pulp, the halogen lights tended to generate higher temperatures compared to the lasers and plasma arc lights, probably due to halogen light cure times, which were three to ten times longer than the lasers and plasma arc lights. Figure 13 shows that increasing cure times can result in an increase of temperature with both a moderate intensity and high intensity light. Negative effects of exaggerated cure times, possibly due to increased heat at the surface, were apparent on extracted teeth (See Figure 4).
Polymerization Strain Measurement
Strain gauges mounted on freshly extracted human molars showed different levels of stress development and strain relief dependent on resin formulation. Figure 14, showing three example strain graphs, indicates that HeliomolarÔ developed substantially less strain more slowly; Z100Ô developed a high level of strain most rapidly; and Bisfil IIÔ autocure resin developed strain slowly over the first 10 minutes of its set, then suddenly developed substantial strain that continued to increase over the next 10 minutes, until it had the most profound strain of the three resins. Both Z100Ô and HeliomolarÔ samples show obvious strain relief after about one minute, however, strain relief in the Bisfil IIÔ sample is not as obvious. Observable signs of strain relief in the extracted teeth were "white lines" at the resin-tooth interface and cracks in the enamel surrounding margins (See Figure 15). All the samples examined had these defects to some degree. Figure 16 illustrates the improbability of detecting the polymerization strain cracks clinically due to their small size and lack of contrast with surrounding tooth structure; however, the "white lines" at the margins are visually apparent.
Overall, 89% of the 76 teeth treated with slot preparations on a proximal surface exhibited cracks correlated directly to resin polymerization. TPH SpectrumÔ and Z100Ô had the highest percentage of enamel cracks, and HeliomolarÔ and Bisfil IIÔ were associated with the lowest percentage of enamel cracks. Of the 89% of teeth with cracks, large cracks, of the size shown in Figure 15 or larger, were observed on 82% of Z100Ô restorations, 67% of TPH SpectrumÔ, 50% of Bisfil IIÔ, and 16% of HeliomolarÔ.
Overall, 61% of the 76 teeth treated exhibited "white lines" at the margins correlated directly to resin polymerization. Frequency of "white lines" was resin dependent, with 92% of HeliomolarÔ restorations, 57% of Z100Ô, 33% of TPH SpectrumÔ, and 25% of Bisfil IIÔ. Figure 17 shows how the "white line" defects can be seen through a translucent resin material. This confirms observations made by others stating that "white lines" are formed during polymerization, and are not caused by finishing procedures.6
Statistical analyses showed no differences among the seven lights related to enamel cracking and "white lines" (p = 0.5740), but resin formulations were highly significant (p = 0.0001).
In order to determine how the physical signs of strain relief on a tooth ("white lines" at margins and cracks in enamel) related to the microstrain values, logistic regression was used. Figure 18 shows that as the maximum microstrain measured by the strain gauge increases, probability for large cracks increases rapidly while probability for "white lines" at the margin decreases rapidly.
To our knowledge, this study includes the largest number of different designs of resin curing lights and different formulations of restorative resins yet investigated. This diversity was sought in order to establish a broader overview of the problems and possible solutions, and to reflect clinical actuality. The study design included lights that provide high intensity with short cure times, moderate intensity with moderate cure times, and stepped and ramped intensities with delayed cure times. It included two different laser designs and several different plasma arc designs. It included light cure resins that represent two different microfill technologies and three different hybrid filler systems, plus a currently sold autocure resin, as a control. The diversity may have contributed to conclusions that differ from previous investigators.1-4, 6 The significant factor in this study was shown repeatedly to be resin formulation, rather than curing light design.
Although several investigators concluded that lower intensity lights improved defects at the resin-tooth interface,1-5 none claimed to solve the "white line" and gap problems. Several stated more than light intensity was involved in the problem, and specifically mentioned resin modulus was also a key factor.1-6 This investigation did not find that lowering the curing light intensity either initially or throughout the cure aided in improvement of the "white line" and enamel cracking defects reported. Differences in cavity design and location and methods of grading the defects may have contributed to the different findings.
Several of the newer resin curing lights tested place many options at the discretion of dentists (intensities, cure time, cure modes) without giving criteria or suggestions on which to base decisions. This requires dentists to make decisions about points still under debate among researchers and curing light manufacturers. This is not a reasonable situation.
Some clinicians and researchers have proposed that resins should bear light sensitivity designations referred to as "resin signatures" which would be used in a manner similar to ASA numbers on photographic film to guide clinicians in selecting curing light settings. This idea has some merit because it places responsibility on resin manufacturers to determine and publicize information on how to obtain best cure of their resin, and it places responsibility on curing light manufacturers to produce lights with energy output in relevant areas of the various resin initiators. It also places responsibility on dentists to test their light’s output, note the "resin signature" value, and set their light as recommended. However, until new resin formulations with low modulus and shrinkage are available, results of this study indicate "resin signatures" will not solve the polymerization defect problems.
Results of this study indicate an urgent need for new resin formulations exhibiting low modulus and low shrinkage in order to overcome defects at the resin-tooth interface during polymerization. HeliomolarÔ showed the best combination of low modulus and low shrinkage in this study. Teeth restored with HeliomolarÔ had few enamel cracks, but all had "white lines" at margins somewhere, indicating that use of HeliomolarÔ helped, but did not solve the problem. Other microfill resins investigated for modulus and shrinkage, such as Silux PlusÔ and Renamel MicrofillÔ,j, had either modulus or shrinkage much higher than HeliomolarÔ and produced enamel cracking in addition to "white lines" at the margins. This study showed clearly that resin formulation problems were the significant factor in formation of polymerization defects, not resin curing light design. This suggests that emphasis for new market ideas should be shifted from curing lights to resin formulations.
This study has not addressed several important points related to the resin polymerization defects described, such as frequency and appearance of the defects in vital teeth and in different types of cavity preparations. The contribution of different types of layering and insulating adhesives also needs attention. These points are under study presently and will be addressed in future publications.
For the present, with no resins or curing lights available that solve the polymerization defect problems, we suggest careful use of surface sealants at the completion of resin finishing to seal the "white line" and enamel crack defects created by polymerization. The surface sealants placed over resin restorations wear off of external surfaces, but are retained at margins and within "V"-shaped defects.10
Regardless of the type of testing performed, resin formulation, rather than resin curing light design and intensity was the significant factor associated with shrinkage, modulus, strain, and "white lines" and cracks at the margins. However, resin curing light design was a factor related to cure time and heat generated at the restoration surface and within the tooth. Higher intensities and longer cure times resulted in greater temperature increases. Lasers and plasma arc lights caused higher temperatures at the surface, and halogen lights caused higher temperatures internally due to longer exposure times needed for polymerization. Seventy-six molars with slot preparations all in enamel exhibited cracks adjacent to margins and "white lines" at the margins in 89% and 61% of the cases, respectively. These defects were photographically correlated directly to resin polymerization. The clinical significance of the cracks and "white lines" is unknown. However, they may be contributing to the post-operative sensitivity often experienced by clinicians placing direct resin restorations. Until materials with lower shrinkage and modulus become available, it may be prudent to use low viscosity surface sealants blown and cured into the "white lines" and cracks to seal the defects.
We would like to thank Todd Wahlin, Jerod Bybee, and Zach Houser for technical assistance in laboratory test procedures.
a International Light, Inc., Newburyport, MA 01950, 978-465-5923
b Demetron/Kerr, Danbury, CT 06810, 203-748-003
c Barber-Coleman Company, Loves Park, IL 61132, 815-637-3480
d National Institute of Standards and Technology, Gaithersburg, MD 20899, 301-975-6806
e MTS Systems Corporation, Minneapolis, MN 55424, 612-937-4000
f Dentsply Midwest, Des Plaines, IL 60018, 847-640-4800
g Measurements Group, Inc., Raleigh, NC 27611, 919-365-3800
h Bisco Dental Products, Schaumburg, IL 60193, 847-534-6000
i 3M Dental Products Division, St. Paul, MN 55144, 651-575-5144
j Cosmedent, Inc., Chicago, IL 60640, 773-989-6844