A Confocal Microscopic Study of Some Factors Affecting the Adaptation of a Light-cured Glass Jonomer to Tooth Tissue T.F. WATSON The Department of Conservative Dentistry, United Medical and Dental Schools, Guys Hospital, London Bridge, SE] 9RT*; and the Department of Anatomy and Developmental Biology, University College London, Gower Street, London, WCJE 6BT, England (*address for correspondence)
Vitrabond consists of a conventional glass ionomer, in conjunction with a light-curing resin and hydroxy-ethylmethacrylate. This study, which used a tandem scanning reflected light microscope for confocal imaging, looked at factors affecting the adaptation of this material to tooth tissue. Wedgeshaped cervical cavities were cut and restored in three ways: (1) Vitrabond was applied as a thin sub-base and either extended onto the enamel margin or kept clear of it. P50 resin composite was then placed, following phosphoric-acid-etching of the enamel margins. (2) The dentin surfaces were conditioned with Scotchprep (maleic acid), then with the Vitrabond, the enamel was etched, and the Scotchbond 2 adhesive applied prior to addition of the P50. (3) Vitrabond was applied alone in bulk, with and without Scotchprep acid-dentin conditioning with a 1:1 (normal) and 3:1 powder:liquid ratio (P:L). Adaptation of the Vitrabond was excellent when maleic acid was used for conditioning of the dentin. When the Vitrabond was used with P50 but extended onto the enamel, the enamel margin occasionally failed. Enamel invariably fractured when the Vitrabond was used alone in bulk. An increase in the P:L ratio decreased contraction gaps when the dentin was not conditioned, but Vitrabond failed cohesively when the dentin was conditioned. The Vitrabond was very susceptible to shrinkage on dehydration. This study suggests that Vitrabond should only be applied to dentin in thin layers, should not be extended onto enamel margins, and should not be allowed to dehydrate. Maleic acid conditioning of the dentin improved adaptation. J Dent Res 69(8):1531-1538, August, 1990
Introduction. Light-cured glass ionomers are designed for use in the composite/glass-ionomer sandwich technique, as first suggested for chemically-set glass ionomers (McLean and Wilson, 1977), where they should act as a protective covering for the dentin. The aim of this study was to determine how the light-cured glass ionomer Vitrabond (3M Co.) is affected by, and affects, the tooth/restoration complex. Failure of this material at the cervical margin of a cavity could shorten the life expectancy of a restoration. At least three manufacturers have introduced materials that incorporate the active components of a glass ionomer (e.g., acid-leachable glass, polyalkenoic acid) with a photo-curing resin system. To date (1989), no patent information is available on these materials. One of the light-cured glass ionomers (Vitrabond, 3M PLC) contains a light-sensitive glass, an ionic polymer (ionomer), and water. This system uses a polycarboxylic acid with some pendent methacryloxy groups. In Received for publication November 20, 1989 Accepted for publication February 5, 1990 This investigation was supported by Science and Engineering Research Council Grant GR/D/68504 at UCL and by the Special Trustees of University College and Middlesex School of Medicine.
addition to the polycarboxylic acid, the liquid portion contains some hydroxyethyl-methacrylate (HEMA) and a photo-active agent (manufacturer's data). This would suggest that Vitrabond may have some mechanisms of attachment to tooth tissue in common with some of the "third-generation" adhesives, such as Scotchbond 2. According to information supplied by the manufacturers, when the powder and liquid components are mixed together, the conventional ionomer reaction begins immediately. The polycarboxylic acid ionizes in water in the presence of the glass, so that a negatively-charged carboxylate polymer and protons are formed. The protons then attack the glass, causing it to release positive ions, e.g., Ca2+, Al3+, AlF2+, etc. These positive ions complex with the negatively-charged polycarboxylate polymer to form an ionically-cross-linked network. Some of the positive ions released from the glass are complex in structure and have fluorine present in them, e.g., A1F2+, A1F2+. The polycarboxylate ions replace the fluorine ions of these complex ions, resulting in the very important fluoride-releasing benefit of glass ionomers. Water serves as the medium through which any transportation of ions occurs. Thus far, the material is identical to any glass-ionomer cement. The light-curing capability of this material arises from the presence of the pendant methacryloxy groups of the polycarboxylate. These will react with the available HEMA and set with a photo-initiated reaction, giving a hard cement on exposure to a blue curing light. Even after light-curing, the glassionomer reaction will continue to completion. The Vitrabond liquid does not undergo curing if it alone is exposed to light; it requires the presence of the photo-sensitive glass to trigger the setting reaction. The material has been described as an "inter-connecting polymer network" (B.E. Causton, personal communication), reflecting the fact that the two polymerization systems are separate from one another, but not mutually exclusive. The 3M Company claims high tensile and compressive strengths for Vitrabond once it is cured, as well as good adhesive properties. The company also claims that it leaches fluoride in laboratory studies and so is acting as a true glass ionomer. The material will set within 12 min even if it is not light-cured. The chemically-cured glass ionomers show minimal contraction on setting; it is hard for one to imagine that this will be the case with a system that contains a resin. A confocal optical microscope was used for examination of factors affecting the adaptation of Vitrabond to tooth tissue, especially dentin conditioning agents, polymerization shrinkage, and the susceptibility of the material to dehydration. Confocal imaging enabled subsurface features of the tooth/restoration interface and microscopic failure of the tissues, materials, or bond to be examined with the restoration in situ.
Materials and methods. The penetration of bonding agents that are applied to dentin may be influenced by the contents of the pulp and dentin tubules (Watson and Boyde, 1987). For this reason, the practice 1531
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
1532
WA TSON
TABLE 1 VITRABOND PLACED WITHOUT SCOTCHPREP DENTIN CONDITIONING Tooth Scotchprep Vitrabond H3PO4 Scotchbond 2 P50 (Figures) 1 *Fl *Rh * * 2 *Rh *Fl * * * * * 3 *Fl * * * 4 *Rh * * * 5 *Rh 6 (1, 2) * * *Fl * * * * 7 (7) 8 * * ** * * * 9 (3, 4) *FL 10 * *Fl * * 11 *Fl * * * 12 * * *Fl * 13 * *Fl * * Schedule: -, not present; *, present; Fl, fluorescein; and Rh, rhodamine. Figures corresponding to the specimens are shown in parentheses.
of using only fresh teeth taken directly from the extraction clinic was adopted, all operative procedures being completed within 30 min to one h after extraction. Thirty-one third molars were carefully checked by examination with a loupe for any damage resulting from their removal. For reproduction of restorations approximating those in normal clinical practice, cervical cavities 4 mm wide, 3 mm high, and 1-1.5 mm deep, with a cavosurface angle of 900, were cut with a type-012 diamond bur running wet in an ultra-high-speed handpiece. Cavities were placed mesially and distally and were extended below the enamel-cementum and enamel-dentin junctions (i. e., finished onto dentin, thus imitating a large approximal cavity or an abrasion cavity). The teeth were restored in three ways: (1) Vitrabond was applied as a thin sub-base and either extended halfway onto the enamel margin in the mesial cavity, or taken only to the enamel-dentin junction in the distal cavity and then light-cured for 20 s. The exposed enamel margins were acid-etched with 37% phosphoric acid for 15 s, then washed with water for 20 s. Scotchbond 2 adhesive (3M Co.; HEMA, Bis-GMA, and photo-initiator) was then applied as a thin layer, and the cavity was restored with P50 resin composite (3M Co.), placed, and light-cured in two layers (see Table 1). (2) The dentin surfaces were conditioned with Scotchprep (3M Co.; maleic acid, HEMA, and water) for 60 s, dried, and then the Vitrabond was placed, enamel was etched, and the Scotchbond 2 applied prior to the addition of the P50, as above (see Table 2). (3) Vitrabond was applied alone in bulk, with and without Scotchprep acid-conditioning of the tooth surface, but without phosphoric acid etching of the enamel. The light-cured glass TABLE 3 VITRABOND PLACED WITH A NORMAL AND 3/1 POWDER/ LIQUID RATIO Tooth Scotchprep Vitrabond H3P04 Scotchbond 2 P50
26 27 28 29 30 31
(Figures) (16) (10-12) (13-15)
* * * * * * (18) * * (17) Schedule: -, not present; *, present; Fl, fluorescein; and Rh, rhodamine. Figures corresponding to the specimens are shown in parentheses.
J Dent Res August 1990 TABLE 2 VITRABOND PLACED WITH SCOTCHPREP DENTIN CONDITIONING Tooth Scotchprep Vitrabond HPO4 Scotchbond 2
P50
(Figures) * * * * (6) * 15 * 16 *Rh * * * * 17 (7, 8, 9) *Rh *Rh *Fl * * 18 *Rh *Fl *Fl * * 19 *Fl *Fl * * * 20 *Rh * * *Rh * 21 * * * *Fl * 22 * * *Fl * * 23 * *Fl * 24 * * *Fl * * 25 * * *Fl * * Schedule: -, not present; *, present; Fl, fluorescein; and Rh, rhodamine. Figures corresponding to the specimens are shown in parentheses.
14
ionomer was mixed with a 1:1 (normal) powder:liquid ratio (P/L) in the mesial cavity, and a 3:1 P/L in the distal cavity. The restoration was light-cured for 30 s (see Table 3). Samples were maintained in a damp environment for 24 h and then sectioned, once, with a slow-speed diamond saw under water. The tooth/restoration interface was then examined immediately (i.e., before any drying could take place) with a tandem-scanning reflected-light microscope or TSM (Tracor Northern, 2551 W. Beltline Highway, Middleton, WI 53562; Watson and Boyde, 1987). The fit of the restorations was examined with an X40/0.8NA dry objective, and the sub-surface adaptation of the restoration observed initially with an X50/1.ONA water-immersion objective, and then an X40/1.3NA oil-immersion objective, all in conjunction with an X10 eyepiece. Specimens were allowed to dehydrate under the microscope for up to ten min, and the movement of the materials was recorded photographically during this time. Fluorescent markers (fluorescein and rhodamine B; see Watson, 1989a, b) were incorporated into the various components of the system, and their distribution within the tooth and restoration was noted with use of appropriate fluorescence excitation and barrier filters.
Results. Optical contrast between the glass- and the light-cured matrix was sufficient for simple reflection imaging in the TSM. Incorporation of fluorophore into the Vitrabond helped to improve this contrast, giving a better image of particle distribution, but it was not essential for routine applications. When the Vitrabond was extended onto the enamel margins and the cavity was restored with P50, it was noticed that the enamel was subject to cracking beneath the glass ionomer. This could have been caused by polymerization contraction of the adhesive and composite (Figs. 1 and 2), or perhaps by the light-cured glass ionomer (Figs. 3 and 4). Gap formation along part of the margin was frequently present between the lining material and dentin (Fig. 5), especially toward the cervical margin. Adaptation of the glass ionomer was greatly improved by the use of Scotchprep. This could be seen when a dry objective was used for the cut surface of the section (Fig. 6) and when an immersion objective was used (Fig. 7). The distribution of the components could be verified by the use of fluorescent labels. Fig. 7 shows a reflection image of dentin tubules with the glass ionomer labeled with fluorescein and the Scotchbond 2 labeled with rhodamine B. In Fig. 8, a
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
Vol. 69 No. 8
MORPHOLOGY OF TOOTH!1VITRABOND INTERFACE
1533
Fig. 1-Enamel margin bordering the junction between P50, Scotchbond 2 resin (SB), and a small extension of Vitrabond (bottom, right-hand corner). Scotchbond 2 labeled with fluorescein. X40/1.3NA oil. Bar = 100 Pm.
Fig. 3-Field similar to Figs. 1 and 2. The Vitrabond (V) is extended onto enamel (E) beneath the Scotchbond 2 and P50. Cracking is visible in the enamel (top, left-hand corner). The glass ionomer has been etched and the roughened surface penetrated by the fluorescein-labeled Scotchbond 2. 40/1.3 oil. Bar = 100 ,um.
Fig. 2-Same field as Fig. 1. Fluorescence illumination highlights the cracks formed in the enamel between the Scotchbond 2 and Vitrabond (arrows). 450-nm excitation filter, 520-nm barrier filter. 40/1.3 oil. Bar = 100lm.
Fig. 4-Same field as Fig. 3. Fluorescence image showing the adaptation of the Scotchbond 2 resin and Vitrabond. Cracking of the enamel is just discernible (arrow). 450-nm excitation, 520-nm barrier filter. 40/ 1.3 oil. Bar = 100 Atm.
green band-pass excitation filter (546 + 10 nm) and a red long-pass filter (600 nm) show fluorescence from only the rhodamine, highlighting the contribution of the Scotchprep and Scotchbond 2 resin to the adhesive system. The fluorescence signal from the labeled glass ionomer was often reduced at the interface with the dentin, and there was occasional evidence for some movement of fluorophore into the dentin (Fig. 9). The stained band within the dentin, which is characteristically seen with conventionally cured glass ionomers, was not present (Watson, 1989a; Boyde et al., 1990). In the same sample in which the Vitrabond was labeled with fluorescein, a blue excitation filter (450 nm) and a yellow pass filter (520 nm) allowed the distribution of leached products to be seen alone
(Fig. 9). The Scotchbond 2 adhesive was frequently seen to have mixed with the accidentally-etched glass ionomer, embedding the irregular glass particles in fluorescent-labeled resin at the interface, with the overlying P50 composite restoration (Figs. 3, 4, 7, 8, and 9). Bulk placement of the Vitrabond in a cavity increased the likelihood of substantial polymerization contraction. When the section surface was allowed to dehydrate under the microscope, but without the sample continuously illuminated, cracking at the interface became particularly apparent. When Scotchprep was used on dentin, then fracturing of the enamel was common, with the maintenance of a bond to dentin. The appearance of the section surface can be seen in Fig. 10 im-
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
1534
WA TSON
Fig. 5-Dentin (D) near the cervical margin of a cavity lined with Vitrabond (V). The stage at which a gap formed during the placement of the restoration cannot be ascertained. 40/1.3 oil. Bar = 100 pLm.
mediately following sectioning, with immediate fracture of the enamel close to the EDJ. After ten min of dehydration, the crack had opened (Fig. 11). When examined with an immersion objective, cohesive failure of the enamel was apparent (Fig. 12), with prisms straddling the void between the glass ionomer and the tooth. A reduction in polymerization shrinkage was observed when the powder:liquid ratio was increased to 3:1. In Fig. 13 there was no fracture visible at the surface after sectioning, but after ten min of dehydration, a crack had opened (Fig. 14) and enamel prisms had failed cohesively, remaining attached to the glass ionomer (Fig. 15). The width of the gap between restoration and tooth was halved when the powder:liquid ratio in these test cavities was increased, but the incidence of cohesive failure within the glass ionomer was increased. Even so, where the prism course was parallel with the cavity margin, the contraction of the glass ionomer frequently caused interprismatic failure (Fig. 16). When a 3:1 P/L ratio was used with Scotchprep dentin conditioning, a cross-over fracture was commonly observed at the EDJ (Fig. 17). Cohesive failure within the enamel gave way to cohesive failure within the glass ionomer at the EDJ, which spread along the interface with the dentin (Fig. 18).
Discussion. The good optical contrast between the glass and its matrix shown by this light-cured glass ionomer may be attributable to the presence of the resin within the system or, alternatively, the photosensitivity of the glass itself. The optical properties of this material are significantly different from those of a conventional glass ionomer (Watson, 1989a). Evidence for the movement of fluids into and out of the glass ionomer was not as apparent from the present study with fluorescent-labeled Vitrabond (q. v. chemically-cured glass-ionomer cements; Boyde et al., 1990; Watson, 1989a), but this may indicate non-detection rather than absence. The manufacturer claims considerable leaching of fluoride from Vitrabond. The glass-ionomer/resin-composite sandwich technique sets up many complex competing stresses within both the tooth and the restoration (Garcia-Godoy et al., 1988). The failures seen
J Dent Res August 1990
Fig. 6-Surface reflection image of a sectioned restoration where Scotchprep was placed before the Vitrabond (V). No gap can be seen next to the dentin (D). Scotchprep not visible in a surface reflection image. 40/ 0.70 dry. Bar = 100 pum.
Fig. 7-Figs. 7, 8, and 9 are of the same field, showing the interfaces between P50, Scotchbond (SB), Vitrabond (V), Scotchprep (SP), and dentin (D). The Scotchprep and Scotchbond are labeled with rhodamine B, and the Vitrabond is labeled with fluorescein. Fig. 7 shows a reflection image: dentin interface with Vitrabond, Scotchbond 2 resin, and P50. Dentin surface conditioned with Scotchprep. Good adaptation of restoration. 40/1.3 oil. Bar = 50 pum. Fig. 8 shows fluorescence from the rhodamine, and Fig. 9 shows fluorescence from the fluorescein.
close to the EDJ in the present study could be expected because of natural flaws such as tufts and lamellae. The presence of the overlying composite also introduces too many variables into the system, because of the polymerization stresses that this will apply to the roughened, etched surface of a glass ionomer (Lutz et al., 1986). The manufacturer suggests that the glass ionomer should not be etched, but the present study has shown that this is virtually impossible to avoid, even in conditions with good access in a laboratory. With the chemically-cured glass-ionomer cements, the slow set of these materials favors an approach whereby the glass
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
Vol. 69 No. 8
MORPHOLOGY OF TOOTH!1VITRABOND INTERFACE
Fig. 8-Same field as Fig. 7. Scotchprep (SP) and Scotchbond 2 resin (SB) labeled with rhodamine B. Scotchprep has penetrated dentin tubules, and the air-inhibited layer has mixed with the resin composite. 546-nm 600-nm barrier filter. 40/1.3 oil. Barr= 50 m . m anditheatir-inhibited larer hasfmiter.40/1.3hoi excitation,
Fig. 9-Same field as Figs. 7 and 8. Vitrabond (V) labeled with fluorescein. Some movement of the fluorophore into the dentin tubules is visible. 40/1.3 oil. Bar = 50 pLrm.
ionomer is placed at one visit and then cut back and veneered with the composite when the material is mature, a few days later. In this way, the bond between the glass ionomer and tooth tissue is at its maximum, and the bulk of resin composite may be reduced, with a consequent reduction in polymerization stresses. Light-cured glass ionomers have been developed for production of a material that achieves an instantaneous set, thus allowing the restorative process outlined above to be undertaken in one visit. An important difference that has been highlighted by this study is that this light-cured glass ionomer should not be placed in thick layers. Unless the material is built up in thin increments, considerable contraction stresses produced by composite restorations will still be present in large cavities. The major problems associated with this type of glass
1535
Fig. 10-Horizontal section of a tooth filled with Vitrabond (V). Subsurface structure is not visible, but the difference in reflectivity between the glass ionomer and enamel (E) and dentin (D) enables a crack to be seen within the enamel. Some of the enamel has remained attached to the Vitrabond. 40/0.70 dry. Bar = 50
purm.
Fig. 11-Same field as Fig. 10, but the sample had been allowed to dry out for ten min. Notice that the crack size has increased. 40/0.70 dry. Bar = 50 pm.
ionomer only became apparent when the cement/tooth interface was studied in isolation. While the use of Vitrabond in bulk is not recommended by the manufacturers, the placement of this material in an exaggerated clinical condition highlighted some of the problems with this glass-ionomer cement. When vitrabond was placed without any form of cavity conditioning, the strength of bond to enamel was clearly superior to that with dentin, as evidenced by the frequent presence of a gap between the Vitrabond and the cervical margin. Prior treatment of the cavity with Scotchprep produced highly reliable adaptation between Vitrabond and tooth tissue. This was especially noticeable at the dentin interface, but, unfortunately, the light-cured glass ionomer was then capable of causing frac-
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
1536
WA TSON
J Dent Res August 1990
Fig. 12-Same sample as Figs. 10 and 11, but slightly further toward the natural surface of the tooth. Enamel can be seen attached to the Vitrabond, with the contraction gap straddled by dislodged enamel prisms (arrowed). 40/1.3 oil. Bar = 50 pum.
Fig. 14-Same field as Fig. 13 after a ten-minute exposure to air. Previously-visible enamel cracks have now opened into gaps. The Vitrabond is still, apparently, attached to the dentin. 40/0.70 dry. Bar = 50 pm.
Fig. 13-A cavity filled with Vitrabond (V) with a 3:1 P/L ratio, with no prior Scotchprep conditioning of dentin (D). View of horizontal section surface immediately after being sectioned. Small cracks can be discerned within the enamel (right-hand side, arrowed), subjacent to the Vitrabond, especially if this field is compared with Fig. 14. 40/0.70 dry. Bar = 50
Fig. 15-Similar field, same sample as Figs. 13 and 14. Enamel prisms can be seen attached to the glass ionomer (arrow), but this material is showing signs of cohesive failure close to the dentin. 40/1.3 oil. Bar = 50 Wm.
pm. ture of enamel margins of the tooth, especially where the prism course was
parallel with the cavity margin. Enamel prisms
are
most easily separated in this direction (Boyde, 1989; Watson,
1990), and it would appear that the bond that is made between the glass ionomer and enamel is stronger than the cohesive strength of enamel in this orientation. Subsequent dehydration greatly increased the size of these cracks. Increasing the filler content is a classical method for the decrease of polymerization contraction. In the experiments reported here (with a substantial increase in filler loading), the polymerization contraction, as illustrated by gap formation in
the non-conditioned cavities, was reduced. It was evident that the Vitrabond had a much shorter working time and behaved more like a conventional glass-ionomer cement. The size of the gaps would be unlikely to allow the broken bonds to be re-established, a mechanism that has been said to occur in conventional glass-ionomer cements (Wilson and McLean, 1988). When combined with Scotchprep, the adaptation was again excellent, with crack formation becoming apparent only on dehydration. The Vitrabond showed cohesive failure, especially associated with the dentin interface. This may be a result of weakening of the material, its tensile strength normally being partly derived from the resin component, with fluid movement from the dentin tubules causing a localized
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
Vol. 69 No. 8
MORPHOLOGY OF TOOTH!VITRABOND INTERFACE
Fig. 16-The outer enamel margin of a cavity filled with Vitrabond (V) after longitudinal sectioning and dehydration for ten min. No Scotchprep conditioning. Notice that the enamel (E, arrowed) and glass ionomer have both failed cohesively in different parts of the adhesive/tooth complex40/1.3 oil. Bar = 50 pum.
Fig. 17-A cavity filled with Vitrabond with a 3:1 P/L, following Scotchprep dentin conditioning and dehydration. The fracture line has crossed from the enamel into the restoration at the EDJ, with glass ionomer well-adapted and probably adhering to the dentin. This good adaptation to the dentin was determined by prior examination of the specimen surface. 40/1.3 oil. Bar = 50 pum.
1537
Fig. 18-A cavity treated similarly to the sample in Fig. 17. Cohesive failure of the Vitrabond (V) can be seen running parallel with the dentin (D). Prior examination with a dry objective showed that the Vitrabond was well-adapted to the dentin. The altered appearance at the dentin interface is probably attributable to the Scotchprep conditioning agent. 40/ 1.3 oil. Bar = 50 pm.
(1) When the material is extended onto enamel margins, there is considerable risk of fracture of the enamel due to the good adhesive properties of the glass ionomer and its tendency to polymerization shrinkage, especially when used in conjunction with a resin composite. (2) The adaptation of the material to dentin can be greatly improved by the use of Scotchprep (maleic acid, HEMA, and water) as a conditioning agent. Failure to use a conditioning agent will lead to gap formation, especially if the glass ionomer is subjected to contraction stresses from an overlying composite restoration. (3) An increase in the powder/liquid ratio decreases the polymerization contraction of the light-cured glass ionomer, but also decreases its strength. (4) The material is very susceptible to dehydration, exhibiting considerable shrinkage. It should be protected against dehydration in the mouth during lengthy operative procedures. The TSM has enabled a study of the interface between a light-cured glass ionomer and tooth tissue to be undertaken with minimal disruption due to specimen preparation. Controlled dehydration of the specimen under the microscope indicated some of the stresses that can be placed on the tooth/ restoration complex during extended operative procedures. Fluorescent labeling facilitated visualization of separate components in this complex.
Acknowledgments. decrease in strength of the glass-ionomer component close to the dentin. One is tempted to speculate that the use of Scotchprep in conjunction with Vitrabond is producing an adhesive system similar to that of Scotchbond 2 (Watson, 1989b), but with the added advantages of a glass-ionomer cement. The improved adaptation of Vitrabond could, however, be considered a disadvantage because of the polymerization and dehydration shrinkage that this material exhibits. The results of this study suggest the following conclusions or recommendations:
I thank Alan Boyde and Sheila Jones for discussions relating to this work. Maureen Arora and Roy Radcliffe provided expert technical assistance. REFERENCES BOYDE, A. (1989): Enamel. In: Handbook of Microscopic Anatomy, Vol. 6, A. Oksche and L. Vollrath, Eds., Berlin: SpringerVerlag, pp. 309-473. BOYDE, A.; JONES, S.J.; TAYLOR, L.; WOLFE, L.; and WAT-
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
1538
J Dent Res August 1990
WA TSON
SON, T.F. (1990): Fluorescence in the Tandem Scanning Micro-
J Microsc 157:39-49. GARCIA-GODOY, F.; DRAHEIM, R.N.; TITUS, H.W.; and CHIESA, D. (1988): Microleakage of Composite Restorations with Etched and Non-etched Glass Ionomer Bases, Am J Dent 1:159162. LUTZ, F.; KREJCI, I.; and OLDENBURG, T.R. (1986): Elimination of Polymerization Stresses at the Margins of Posterior Composite Resin Restorations: a New Restorative Technique, Quint Int 17:777784. McLEAN, J.W. and WILSON, A.D. (1977): The Clinical Development of Glass-lonomer Cement. II. Some Clinical Applications, Aust Dent J 22:120-127. WATSON, T.F. (1989a): Fluorescence Confocal Microscopy for the scope,
Rapid Evaluation of Tooth/Restoration Interfaces. In: Proceedings of the 1st International Conference on Confocal Microscopy, Amsterdam. WATSON, T.F. (1989b): A Confocal Optical Microscope Study of the Tooth/Restoration Interface Using Scotchbond 2 Dentin Adhesive, J Dent Res 68:1124-1131. WATSON, T.F. (1990): The Application of Real-time Confocal Microscopy to the Study of High Speed Dental Bur/Tooth Cutting Interactions, J Microsc 157:51-60. WATSON, T.F. and BOYDE, A. (1987): Tandem Scanning Reflected Light Microscopy: Applications in Clinical Dental Research, Scanning Microsc 1:1971-1981. WILSON, A.D. and McLEAN, J.W. (1988): Glass lonomer Cement, Chicago, IL: Quintessence.
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
Vitrabond consists of a conventional glass ionomer, in conjunction with a light-curing resin and hydroxy-ethylmethacrylate. This study, which used a tandem scanning reflected light microscope for confocal imaging, looked at factors affecting the adaptation of this material to tooth tissue. Wedgeshaped cervical cavities were cut and restored in three ways: (1) Vitrabond was applied as a thin sub-base and either extended onto the enamel margin or kept clear of it. P50 resin composite was then placed, following phosphoric-acid-etching of the enamel margins. (2) The dentin surfaces were conditioned with Scotchprep (maleic acid), then with the Vitrabond, the enamel was etched, and the Scotchbond 2 adhesive applied prior to addition of the P50. (3) Vitrabond was applied alone in bulk, with and without Scotchprep acid-dentin conditioning with a 1:1 (normal) and 3:1 powder:liquid ratio (P:L). Adaptation of the Vitrabond was excellent when maleic acid was used for conditioning of the dentin. When the Vitrabond was used with P50 but extended onto the enamel, the enamel margin occasionally failed. Enamel invariably fractured when the Vitrabond was used alone in bulk. An increase in the P:L ratio decreased contraction gaps when the dentin was not conditioned, but Vitrabond failed cohesively when the dentin was conditioned. The Vitrabond was very susceptible to shrinkage on dehydration. This study suggests that Vitrabond should only be applied to dentin in thin layers, should not be extended onto enamel margins, and should not be allowed to dehydrate. Maleic acid conditioning of the dentin improved adaptation. J Dent Res 69(8):1531-1538, August, 1990
Introduction. Light-cured glass ionomers are designed for use in the composite/glass-ionomer sandwich technique, as first suggested for chemically-set glass ionomers (McLean and Wilson, 1977), where they should act as a protective covering for the dentin. The aim of this study was to determine how the light-cured glass ionomer Vitrabond (3M Co.) is affected by, and affects, the tooth/restoration complex. Failure of this material at the cervical margin of a cavity could shorten the life expectancy of a restoration. At least three manufacturers have introduced materials that incorporate the active components of a glass ionomer (e.g., acid-leachable glass, polyalkenoic acid) with a photo-curing resin system. To date (1989), no patent information is available on these materials. One of the light-cured glass ionomers (Vitrabond, 3M PLC) contains a light-sensitive glass, an ionic polymer (ionomer), and water. This system uses a polycarboxylic acid with some pendent methacryloxy groups. In Received for publication November 20, 1989 Accepted for publication February 5, 1990 This investigation was supported by Science and Engineering Research Council Grant GR/D/68504 at UCL and by the Special Trustees of University College and Middlesex School of Medicine.
addition to the polycarboxylic acid, the liquid portion contains some hydroxyethyl-methacrylate (HEMA) and a photo-active agent (manufacturer's data). This would suggest that Vitrabond may have some mechanisms of attachment to tooth tissue in common with some of the "third-generation" adhesives, such as Scotchbond 2. According to information supplied by the manufacturers, when the powder and liquid components are mixed together, the conventional ionomer reaction begins immediately. The polycarboxylic acid ionizes in water in the presence of the glass, so that a negatively-charged carboxylate polymer and protons are formed. The protons then attack the glass, causing it to release positive ions, e.g., Ca2+, Al3+, AlF2+, etc. These positive ions complex with the negatively-charged polycarboxylate polymer to form an ionically-cross-linked network. Some of the positive ions released from the glass are complex in structure and have fluorine present in them, e.g., A1F2+, A1F2+. The polycarboxylate ions replace the fluorine ions of these complex ions, resulting in the very important fluoride-releasing benefit of glass ionomers. Water serves as the medium through which any transportation of ions occurs. Thus far, the material is identical to any glass-ionomer cement. The light-curing capability of this material arises from the presence of the pendant methacryloxy groups of the polycarboxylate. These will react with the available HEMA and set with a photo-initiated reaction, giving a hard cement on exposure to a blue curing light. Even after light-curing, the glassionomer reaction will continue to completion. The Vitrabond liquid does not undergo curing if it alone is exposed to light; it requires the presence of the photo-sensitive glass to trigger the setting reaction. The material has been described as an "inter-connecting polymer network" (B.E. Causton, personal communication), reflecting the fact that the two polymerization systems are separate from one another, but not mutually exclusive. The 3M Company claims high tensile and compressive strengths for Vitrabond once it is cured, as well as good adhesive properties. The company also claims that it leaches fluoride in laboratory studies and so is acting as a true glass ionomer. The material will set within 12 min even if it is not light-cured. The chemically-cured glass ionomers show minimal contraction on setting; it is hard for one to imagine that this will be the case with a system that contains a resin. A confocal optical microscope was used for examination of factors affecting the adaptation of Vitrabond to tooth tissue, especially dentin conditioning agents, polymerization shrinkage, and the susceptibility of the material to dehydration. Confocal imaging enabled subsurface features of the tooth/restoration interface and microscopic failure of the tissues, materials, or bond to be examined with the restoration in situ.
Materials and methods. The penetration of bonding agents that are applied to dentin may be influenced by the contents of the pulp and dentin tubules (Watson and Boyde, 1987). For this reason, the practice 1531
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
1532
WA TSON
TABLE 1 VITRABOND PLACED WITHOUT SCOTCHPREP DENTIN CONDITIONING Tooth Scotchprep Vitrabond H3PO4 Scotchbond 2 P50 (Figures) 1 *Fl *Rh * * 2 *Rh *Fl * * * * * 3 *Fl * * * 4 *Rh * * * 5 *Rh 6 (1, 2) * * *Fl * * * * 7 (7) 8 * * ** * * * 9 (3, 4) *FL 10 * *Fl * * 11 *Fl * * * 12 * * *Fl * 13 * *Fl * * Schedule: -, not present; *, present; Fl, fluorescein; and Rh, rhodamine. Figures corresponding to the specimens are shown in parentheses.
of using only fresh teeth taken directly from the extraction clinic was adopted, all operative procedures being completed within 30 min to one h after extraction. Thirty-one third molars were carefully checked by examination with a loupe for any damage resulting from their removal. For reproduction of restorations approximating those in normal clinical practice, cervical cavities 4 mm wide, 3 mm high, and 1-1.5 mm deep, with a cavosurface angle of 900, were cut with a type-012 diamond bur running wet in an ultra-high-speed handpiece. Cavities were placed mesially and distally and were extended below the enamel-cementum and enamel-dentin junctions (i. e., finished onto dentin, thus imitating a large approximal cavity or an abrasion cavity). The teeth were restored in three ways: (1) Vitrabond was applied as a thin sub-base and either extended halfway onto the enamel margin in the mesial cavity, or taken only to the enamel-dentin junction in the distal cavity and then light-cured for 20 s. The exposed enamel margins were acid-etched with 37% phosphoric acid for 15 s, then washed with water for 20 s. Scotchbond 2 adhesive (3M Co.; HEMA, Bis-GMA, and photo-initiator) was then applied as a thin layer, and the cavity was restored with P50 resin composite (3M Co.), placed, and light-cured in two layers (see Table 1). (2) The dentin surfaces were conditioned with Scotchprep (3M Co.; maleic acid, HEMA, and water) for 60 s, dried, and then the Vitrabond was placed, enamel was etched, and the Scotchbond 2 applied prior to the addition of the P50, as above (see Table 2). (3) Vitrabond was applied alone in bulk, with and without Scotchprep acid-conditioning of the tooth surface, but without phosphoric acid etching of the enamel. The light-cured glass TABLE 3 VITRABOND PLACED WITH A NORMAL AND 3/1 POWDER/ LIQUID RATIO Tooth Scotchprep Vitrabond H3P04 Scotchbond 2 P50
26 27 28 29 30 31
(Figures) (16) (10-12) (13-15)
* * * * * * (18) * * (17) Schedule: -, not present; *, present; Fl, fluorescein; and Rh, rhodamine. Figures corresponding to the specimens are shown in parentheses.
J Dent Res August 1990 TABLE 2 VITRABOND PLACED WITH SCOTCHPREP DENTIN CONDITIONING Tooth Scotchprep Vitrabond HPO4 Scotchbond 2
P50
(Figures) * * * * (6) * 15 * 16 *Rh * * * * 17 (7, 8, 9) *Rh *Rh *Fl * * 18 *Rh *Fl *Fl * * 19 *Fl *Fl * * * 20 *Rh * * *Rh * 21 * * * *Fl * 22 * * *Fl * * 23 * *Fl * 24 * * *Fl * * 25 * * *Fl * * Schedule: -, not present; *, present; Fl, fluorescein; and Rh, rhodamine. Figures corresponding to the specimens are shown in parentheses.
14
ionomer was mixed with a 1:1 (normal) powder:liquid ratio (P/L) in the mesial cavity, and a 3:1 P/L in the distal cavity. The restoration was light-cured for 30 s (see Table 3). Samples were maintained in a damp environment for 24 h and then sectioned, once, with a slow-speed diamond saw under water. The tooth/restoration interface was then examined immediately (i.e., before any drying could take place) with a tandem-scanning reflected-light microscope or TSM (Tracor Northern, 2551 W. Beltline Highway, Middleton, WI 53562; Watson and Boyde, 1987). The fit of the restorations was examined with an X40/0.8NA dry objective, and the sub-surface adaptation of the restoration observed initially with an X50/1.ONA water-immersion objective, and then an X40/1.3NA oil-immersion objective, all in conjunction with an X10 eyepiece. Specimens were allowed to dehydrate under the microscope for up to ten min, and the movement of the materials was recorded photographically during this time. Fluorescent markers (fluorescein and rhodamine B; see Watson, 1989a, b) were incorporated into the various components of the system, and their distribution within the tooth and restoration was noted with use of appropriate fluorescence excitation and barrier filters.
Results. Optical contrast between the glass- and the light-cured matrix was sufficient for simple reflection imaging in the TSM. Incorporation of fluorophore into the Vitrabond helped to improve this contrast, giving a better image of particle distribution, but it was not essential for routine applications. When the Vitrabond was extended onto the enamel margins and the cavity was restored with P50, it was noticed that the enamel was subject to cracking beneath the glass ionomer. This could have been caused by polymerization contraction of the adhesive and composite (Figs. 1 and 2), or perhaps by the light-cured glass ionomer (Figs. 3 and 4). Gap formation along part of the margin was frequently present between the lining material and dentin (Fig. 5), especially toward the cervical margin. Adaptation of the glass ionomer was greatly improved by the use of Scotchprep. This could be seen when a dry objective was used for the cut surface of the section (Fig. 6) and when an immersion objective was used (Fig. 7). The distribution of the components could be verified by the use of fluorescent labels. Fig. 7 shows a reflection image of dentin tubules with the glass ionomer labeled with fluorescein and the Scotchbond 2 labeled with rhodamine B. In Fig. 8, a
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
Vol. 69 No. 8
MORPHOLOGY OF TOOTH!1VITRABOND INTERFACE
1533
Fig. 1-Enamel margin bordering the junction between P50, Scotchbond 2 resin (SB), and a small extension of Vitrabond (bottom, right-hand corner). Scotchbond 2 labeled with fluorescein. X40/1.3NA oil. Bar = 100 Pm.
Fig. 3-Field similar to Figs. 1 and 2. The Vitrabond (V) is extended onto enamel (E) beneath the Scotchbond 2 and P50. Cracking is visible in the enamel (top, left-hand corner). The glass ionomer has been etched and the roughened surface penetrated by the fluorescein-labeled Scotchbond 2. 40/1.3 oil. Bar = 100 ,um.
Fig. 2-Same field as Fig. 1. Fluorescence illumination highlights the cracks formed in the enamel between the Scotchbond 2 and Vitrabond (arrows). 450-nm excitation filter, 520-nm barrier filter. 40/1.3 oil. Bar = 100lm.
Fig. 4-Same field as Fig. 3. Fluorescence image showing the adaptation of the Scotchbond 2 resin and Vitrabond. Cracking of the enamel is just discernible (arrow). 450-nm excitation, 520-nm barrier filter. 40/ 1.3 oil. Bar = 100 Atm.
green band-pass excitation filter (546 + 10 nm) and a red long-pass filter (600 nm) show fluorescence from only the rhodamine, highlighting the contribution of the Scotchprep and Scotchbond 2 resin to the adhesive system. The fluorescence signal from the labeled glass ionomer was often reduced at the interface with the dentin, and there was occasional evidence for some movement of fluorophore into the dentin (Fig. 9). The stained band within the dentin, which is characteristically seen with conventionally cured glass ionomers, was not present (Watson, 1989a; Boyde et al., 1990). In the same sample in which the Vitrabond was labeled with fluorescein, a blue excitation filter (450 nm) and a yellow pass filter (520 nm) allowed the distribution of leached products to be seen alone
(Fig. 9). The Scotchbond 2 adhesive was frequently seen to have mixed with the accidentally-etched glass ionomer, embedding the irregular glass particles in fluorescent-labeled resin at the interface, with the overlying P50 composite restoration (Figs. 3, 4, 7, 8, and 9). Bulk placement of the Vitrabond in a cavity increased the likelihood of substantial polymerization contraction. When the section surface was allowed to dehydrate under the microscope, but without the sample continuously illuminated, cracking at the interface became particularly apparent. When Scotchprep was used on dentin, then fracturing of the enamel was common, with the maintenance of a bond to dentin. The appearance of the section surface can be seen in Fig. 10 im-
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
1534
WA TSON
Fig. 5-Dentin (D) near the cervical margin of a cavity lined with Vitrabond (V). The stage at which a gap formed during the placement of the restoration cannot be ascertained. 40/1.3 oil. Bar = 100 pLm.
mediately following sectioning, with immediate fracture of the enamel close to the EDJ. After ten min of dehydration, the crack had opened (Fig. 11). When examined with an immersion objective, cohesive failure of the enamel was apparent (Fig. 12), with prisms straddling the void between the glass ionomer and the tooth. A reduction in polymerization shrinkage was observed when the powder:liquid ratio was increased to 3:1. In Fig. 13 there was no fracture visible at the surface after sectioning, but after ten min of dehydration, a crack had opened (Fig. 14) and enamel prisms had failed cohesively, remaining attached to the glass ionomer (Fig. 15). The width of the gap between restoration and tooth was halved when the powder:liquid ratio in these test cavities was increased, but the incidence of cohesive failure within the glass ionomer was increased. Even so, where the prism course was parallel with the cavity margin, the contraction of the glass ionomer frequently caused interprismatic failure (Fig. 16). When a 3:1 P/L ratio was used with Scotchprep dentin conditioning, a cross-over fracture was commonly observed at the EDJ (Fig. 17). Cohesive failure within the enamel gave way to cohesive failure within the glass ionomer at the EDJ, which spread along the interface with the dentin (Fig. 18).
Discussion. The good optical contrast between the glass and its matrix shown by this light-cured glass ionomer may be attributable to the presence of the resin within the system or, alternatively, the photosensitivity of the glass itself. The optical properties of this material are significantly different from those of a conventional glass ionomer (Watson, 1989a). Evidence for the movement of fluids into and out of the glass ionomer was not as apparent from the present study with fluorescent-labeled Vitrabond (q. v. chemically-cured glass-ionomer cements; Boyde et al., 1990; Watson, 1989a), but this may indicate non-detection rather than absence. The manufacturer claims considerable leaching of fluoride from Vitrabond. The glass-ionomer/resin-composite sandwich technique sets up many complex competing stresses within both the tooth and the restoration (Garcia-Godoy et al., 1988). The failures seen
J Dent Res August 1990
Fig. 6-Surface reflection image of a sectioned restoration where Scotchprep was placed before the Vitrabond (V). No gap can be seen next to the dentin (D). Scotchprep not visible in a surface reflection image. 40/ 0.70 dry. Bar = 100 pum.
Fig. 7-Figs. 7, 8, and 9 are of the same field, showing the interfaces between P50, Scotchbond (SB), Vitrabond (V), Scotchprep (SP), and dentin (D). The Scotchprep and Scotchbond are labeled with rhodamine B, and the Vitrabond is labeled with fluorescein. Fig. 7 shows a reflection image: dentin interface with Vitrabond, Scotchbond 2 resin, and P50. Dentin surface conditioned with Scotchprep. Good adaptation of restoration. 40/1.3 oil. Bar = 50 pum. Fig. 8 shows fluorescence from the rhodamine, and Fig. 9 shows fluorescence from the fluorescein.
close to the EDJ in the present study could be expected because of natural flaws such as tufts and lamellae. The presence of the overlying composite also introduces too many variables into the system, because of the polymerization stresses that this will apply to the roughened, etched surface of a glass ionomer (Lutz et al., 1986). The manufacturer suggests that the glass ionomer should not be etched, but the present study has shown that this is virtually impossible to avoid, even in conditions with good access in a laboratory. With the chemically-cured glass-ionomer cements, the slow set of these materials favors an approach whereby the glass
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
Vol. 69 No. 8
MORPHOLOGY OF TOOTH!1VITRABOND INTERFACE
Fig. 8-Same field as Fig. 7. Scotchprep (SP) and Scotchbond 2 resin (SB) labeled with rhodamine B. Scotchprep has penetrated dentin tubules, and the air-inhibited layer has mixed with the resin composite. 546-nm 600-nm barrier filter. 40/1.3 oil. Barr= 50 m . m anditheatir-inhibited larer hasfmiter.40/1.3hoi excitation,
Fig. 9-Same field as Figs. 7 and 8. Vitrabond (V) labeled with fluorescein. Some movement of the fluorophore into the dentin tubules is visible. 40/1.3 oil. Bar = 50 pLrm.
ionomer is placed at one visit and then cut back and veneered with the composite when the material is mature, a few days later. In this way, the bond between the glass ionomer and tooth tissue is at its maximum, and the bulk of resin composite may be reduced, with a consequent reduction in polymerization stresses. Light-cured glass ionomers have been developed for production of a material that achieves an instantaneous set, thus allowing the restorative process outlined above to be undertaken in one visit. An important difference that has been highlighted by this study is that this light-cured glass ionomer should not be placed in thick layers. Unless the material is built up in thin increments, considerable contraction stresses produced by composite restorations will still be present in large cavities. The major problems associated with this type of glass
1535
Fig. 10-Horizontal section of a tooth filled with Vitrabond (V). Subsurface structure is not visible, but the difference in reflectivity between the glass ionomer and enamel (E) and dentin (D) enables a crack to be seen within the enamel. Some of the enamel has remained attached to the Vitrabond. 40/0.70 dry. Bar = 50
purm.
Fig. 11-Same field as Fig. 10, but the sample had been allowed to dry out for ten min. Notice that the crack size has increased. 40/0.70 dry. Bar = 50 pm.
ionomer only became apparent when the cement/tooth interface was studied in isolation. While the use of Vitrabond in bulk is not recommended by the manufacturers, the placement of this material in an exaggerated clinical condition highlighted some of the problems with this glass-ionomer cement. When vitrabond was placed without any form of cavity conditioning, the strength of bond to enamel was clearly superior to that with dentin, as evidenced by the frequent presence of a gap between the Vitrabond and the cervical margin. Prior treatment of the cavity with Scotchprep produced highly reliable adaptation between Vitrabond and tooth tissue. This was especially noticeable at the dentin interface, but, unfortunately, the light-cured glass ionomer was then capable of causing frac-
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
1536
WA TSON
J Dent Res August 1990
Fig. 12-Same sample as Figs. 10 and 11, but slightly further toward the natural surface of the tooth. Enamel can be seen attached to the Vitrabond, with the contraction gap straddled by dislodged enamel prisms (arrowed). 40/1.3 oil. Bar = 50 pum.
Fig. 14-Same field as Fig. 13 after a ten-minute exposure to air. Previously-visible enamel cracks have now opened into gaps. The Vitrabond is still, apparently, attached to the dentin. 40/0.70 dry. Bar = 50 pm.
Fig. 13-A cavity filled with Vitrabond (V) with a 3:1 P/L ratio, with no prior Scotchprep conditioning of dentin (D). View of horizontal section surface immediately after being sectioned. Small cracks can be discerned within the enamel (right-hand side, arrowed), subjacent to the Vitrabond, especially if this field is compared with Fig. 14. 40/0.70 dry. Bar = 50
Fig. 15-Similar field, same sample as Figs. 13 and 14. Enamel prisms can be seen attached to the glass ionomer (arrow), but this material is showing signs of cohesive failure close to the dentin. 40/1.3 oil. Bar = 50 Wm.
pm. ture of enamel margins of the tooth, especially where the prism course was
parallel with the cavity margin. Enamel prisms
are
most easily separated in this direction (Boyde, 1989; Watson,
1990), and it would appear that the bond that is made between the glass ionomer and enamel is stronger than the cohesive strength of enamel in this orientation. Subsequent dehydration greatly increased the size of these cracks. Increasing the filler content is a classical method for the decrease of polymerization contraction. In the experiments reported here (with a substantial increase in filler loading), the polymerization contraction, as illustrated by gap formation in
the non-conditioned cavities, was reduced. It was evident that the Vitrabond had a much shorter working time and behaved more like a conventional glass-ionomer cement. The size of the gaps would be unlikely to allow the broken bonds to be re-established, a mechanism that has been said to occur in conventional glass-ionomer cements (Wilson and McLean, 1988). When combined with Scotchprep, the adaptation was again excellent, with crack formation becoming apparent only on dehydration. The Vitrabond showed cohesive failure, especially associated with the dentin interface. This may be a result of weakening of the material, its tensile strength normally being partly derived from the resin component, with fluid movement from the dentin tubules causing a localized
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
Vol. 69 No. 8
MORPHOLOGY OF TOOTH!VITRABOND INTERFACE
Fig. 16-The outer enamel margin of a cavity filled with Vitrabond (V) after longitudinal sectioning and dehydration for ten min. No Scotchprep conditioning. Notice that the enamel (E, arrowed) and glass ionomer have both failed cohesively in different parts of the adhesive/tooth complex40/1.3 oil. Bar = 50 pum.
Fig. 17-A cavity filled with Vitrabond with a 3:1 P/L, following Scotchprep dentin conditioning and dehydration. The fracture line has crossed from the enamel into the restoration at the EDJ, with glass ionomer well-adapted and probably adhering to the dentin. This good adaptation to the dentin was determined by prior examination of the specimen surface. 40/1.3 oil. Bar = 50 pum.
1537
Fig. 18-A cavity treated similarly to the sample in Fig. 17. Cohesive failure of the Vitrabond (V) can be seen running parallel with the dentin (D). Prior examination with a dry objective showed that the Vitrabond was well-adapted to the dentin. The altered appearance at the dentin interface is probably attributable to the Scotchprep conditioning agent. 40/ 1.3 oil. Bar = 50 pm.
(1) When the material is extended onto enamel margins, there is considerable risk of fracture of the enamel due to the good adhesive properties of the glass ionomer and its tendency to polymerization shrinkage, especially when used in conjunction with a resin composite. (2) The adaptation of the material to dentin can be greatly improved by the use of Scotchprep (maleic acid, HEMA, and water) as a conditioning agent. Failure to use a conditioning agent will lead to gap formation, especially if the glass ionomer is subjected to contraction stresses from an overlying composite restoration. (3) An increase in the powder/liquid ratio decreases the polymerization contraction of the light-cured glass ionomer, but also decreases its strength. (4) The material is very susceptible to dehydration, exhibiting considerable shrinkage. It should be protected against dehydration in the mouth during lengthy operative procedures. The TSM has enabled a study of the interface between a light-cured glass ionomer and tooth tissue to be undertaken with minimal disruption due to specimen preparation. Controlled dehydration of the specimen under the microscope indicated some of the stresses that can be placed on the tooth/ restoration complex during extended operative procedures. Fluorescent labeling facilitated visualization of separate components in this complex.
Acknowledgments. decrease in strength of the glass-ionomer component close to the dentin. One is tempted to speculate that the use of Scotchprep in conjunction with Vitrabond is producing an adhesive system similar to that of Scotchbond 2 (Watson, 1989b), but with the added advantages of a glass-ionomer cement. The improved adaptation of Vitrabond could, however, be considered a disadvantage because of the polymerization and dehydration shrinkage that this material exhibits. The results of this study suggest the following conclusions or recommendations:
I thank Alan Boyde and Sheila Jones for discussions relating to this work. Maureen Arora and Roy Radcliffe provided expert technical assistance. REFERENCES BOYDE, A. (1989): Enamel. In: Handbook of Microscopic Anatomy, Vol. 6, A. Oksche and L. Vollrath, Eds., Berlin: SpringerVerlag, pp. 309-473. BOYDE, A.; JONES, S.J.; TAYLOR, L.; WOLFE, L.; and WAT-
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
1538
J Dent Res August 1990
WA TSON
SON, T.F. (1990): Fluorescence in the Tandem Scanning Micro-
J Microsc 157:39-49. GARCIA-GODOY, F.; DRAHEIM, R.N.; TITUS, H.W.; and CHIESA, D. (1988): Microleakage of Composite Restorations with Etched and Non-etched Glass Ionomer Bases, Am J Dent 1:159162. LUTZ, F.; KREJCI, I.; and OLDENBURG, T.R. (1986): Elimination of Polymerization Stresses at the Margins of Posterior Composite Resin Restorations: a New Restorative Technique, Quint Int 17:777784. McLEAN, J.W. and WILSON, A.D. (1977): The Clinical Development of Glass-lonomer Cement. II. Some Clinical Applications, Aust Dent J 22:120-127. WATSON, T.F. (1989a): Fluorescence Confocal Microscopy for the scope,
Rapid Evaluation of Tooth/Restoration Interfaces. In: Proceedings of the 1st International Conference on Confocal Microscopy, Amsterdam. WATSON, T.F. (1989b): A Confocal Optical Microscope Study of the Tooth/Restoration Interface Using Scotchbond 2 Dentin Adhesive, J Dent Res 68:1124-1131. WATSON, T.F. (1990): The Application of Real-time Confocal Microscopy to the Study of High Speed Dental Bur/Tooth Cutting Interactions, J Microsc 157:51-60. WATSON, T.F. and BOYDE, A. (1987): Tandem Scanning Reflected Light Microscopy: Applications in Clinical Dental Research, Scanning Microsc 1:1971-1981. WILSON, A.D. and McLEAN, J.W. (1988): Glass lonomer Cement, Chicago, IL: Quintessence.
Downloaded from jdr.sagepub.com at NORTH CAROLINA STATE UNIV on March 17, 2015 For personal use only. No other uses without permission.
