Influence of calcium concentration on the dentin wettability by an adhesive M. Panighi* and C. G'Sell Laboratoire de Mktallurgie Physique & Sciences des Matkriaux, URA C.N.R.S. DO 155, Ecole des Mines, Parc de Saurupt, 54042 Nancy CMex, France Plane dentin surfaces were abraded perpendicular to the radicular axis of sound human molars. They were cleaned to reveal the tubules, and the morphological features of each surface were studied microscopically. For a first series of teeth, the Vickers microhardness of the dentin surfaces was measured and the calcium and phosphorus composition was deter-
mined by electron microprobe analysis. For a second series, the microhardness a n d wettability of t h e surface by t h e Scotchbond adhesive were compared. Positive correlations were found between the following parameters: degree of minerality, dentin compactness, hardness, and spreading capability of the adhesive.
Although the chemical stability and mechanical strength of the so-called "posterior resin composites" are gradually improving, their use in approxima1 cavities frequently results in clinical failures.' These outcomes are often associated with an unsuccessful seal between the composite material and the dental tissues.' Since shortcomings in operative techniques cannot be always suspected, it is therefore evident that the adhesion ability is limited by illdefined parameters regarding the filling materials or the tooth itself. It has been shown several time^^-^ that the most serious factor for the resin composite is the volumetric shrinkage that occurs during the setting of the resin and the adhesive, giving rise to internal shear stresses at the adhesion interface. To compensate for this effect, some authors recommended increasing the contact area by beveling the enamel on the periphery of the cavity8r9 since it is well known that the adhesion is more efficient on enamel than on dentin. However, this procedure is not sufficient'"," because it does not prevent the formation of dangerous flaws at the resin-dentin interface.12 In addition to the efforts to develop expansible resins, it is important to understand better the elementary mechanisms that control the adhesion of a resin composite to dentin surfaces since the strength of the adhesion appears to be dependent upon the dentin structure and preparati~n.'~ Since the first step of the interaction between an adhesive and a substrate is the wetting of the sub~trate,'~ the aim of this investigation was to study this "To whom correspondence should be addressed Journal of Biomedical Materials Research, VoI. 26, 1081-1089 (1992) 0 1992 John Wiley & Sons, Inc. CCC 0021-9304/92/081081-09$4.00
PANIGHI A N D G'SELL
wetting mechanism in some detail by relating the spreading capability of a given adhesive on various dentin surfaces to their chemical and microstructural features.
Sample preparation Sound human extracted molars were soaked for 24 h in a 0.5% sodium hypochlorite solution. After brushing and rinsing, they were kept in the dark at 4°C in a normal saline solution. For each tooth, a plane dentin surface was prepared by means of a heatless wheel, perpendicular to the radicular axis. To make the operation easier, the teeth were embedded with a special resin (Mecaprex, Presi, Poisat, Grenoble, France) in a copper band #20 (Produits Dentaires, Vevey, Switzerland), the plane surface being oriented perpendicular to the cylindrical band. The final abrasion was carried out using a wet The re320-grit abrasive paper in a specially designed guiding sultant smear layer was cleaned with two drops of a 50 wt % citric acid solution for 30 s, and then the tooth surface was thoroughly rinsed with water for the same time.
Morphological analysis Each etched surface was observed by reflection optical microscopy to determine the density and average diameter of the tubules for several selected areas of 10 X 10 ,urn2. The mean surface fraction between the tubules was then determined as the intertubular "real dentin surface" (RDS).
Microhardness measurements Each dentin surface was tested with a Vickers microindenter (Durimet, Leitz, Wetzlar, Germany) loaded with a 50-g weight. The diagonals of 10 indentations were carefully measured for each tooth in a definite zone. The mean Vickers number, H,, obtained from the measurements was then taken as the hardness of the surface.
Electron microprobe analysis For a first series of samples, the hardness measurements were correlated with a quantitative analysis of the calcium and phosphorus content of the dentin surface. The teeth were dehydrated under vacuum for 40 h at 40°C, coated with a 5-nm gold layer, and mounted in a Camebax Electron Microprobe analyzer (Cameca, Paris, France). The voltage was set at 10 kV, about
INFLUENCE OF CALCIUM CONCENTRATION O N DENTIN WETTABILITY
three times more than the x-ray emission threshold of the elements considered. The electron beam current was set at 25 2 3 nA, a value that optimizes the sensitivity/contamination ratio. The electron beam was focused at randomly distributed spots in the zone where microindentations had been made. At least 10 analyses were performed for each surface so that the experimental errors were reduced to below l % . I 7 The emission intensities, I, of the Ca and P peaks were systematically measured for the sample and for a synthetic hydroxyapatite standard [Ca5(P04)3,F, C1, OH]. The weight-based percentages [Ca] and [PI of the samples in calcium and phosphorus were then obtained with the following relation:
where BN represents the background noise recorded with the sample and the standard and CF is the appropriate correction factor, accounting for the major other elements present at the dentin surface but not specifically analyzed in this series. The CF determination is calculated on the fact that if all the surface elements were measured the total sum would yield loo%." Wett ability measurements
A second series of specimens was used for the microhardness and wetting experiments because it was impossible, after the preceding analyses, to restore the original degree of the teeth hydration with any confidence. Conversely, the presence of the adhesive would have polluted the microprobe camera. The wettability measurements were performed with a 0.6-pL chemosetting adhesive drop (Scotchbond, 3M Co., St. Paul, MN) carefullly deposited on each dentin surface with a micrometric syringe (Terumo & Co., Tokyo, Japan). Fifteen seconds after its deposition, the drop meniscus was photographed with a specially adapted binocular. The contact angle 8 between the meniscus and substrate was measured at X30 enlargement. The cosine of this angle measures in a direct way the wettability of the dentin surface by the spreading adhesive .I4 RESULTS
A complete set of data concerning the tubule morphology and the microhardness of the first series of dentin surfaces is presented in Table I. The intertubular RDS fraction varies from 61437% and is strongly correlated with the microhardness of the dentin, as illustrated graphically in Figure 1. The basic interpretation of this result is to consider that the presence of tubules weakens the material stressed by the Vickers indenter and makes easy its deformation. Extrapolating the data leads to a microhardness of about 100 H , for a hypothetical zero-tubule dentin, in good agreement with the corresponding value of 98 Knoop Hardness (about 94 H,) obtained by Pashley et a1.I'
PANIGHI AND G’SELL
TABLE I Microhardness, Tubule Density, Mean Tubule Diameter, and Intertubular RDS Fraction of the First Series of Dentin Surfaces Tooth Number
D2 D3 D5 D6 D7 Dl1 D13 D15 D19 D35
71.4 70.9 80.9 75.8 70.2 66.4 68.2 80.6 57.5 44.5
t 4.5 2 3.0 ? 5.0 i 4.6 5 2.3 2 1.3 t 2.7 t 4.6 f 5.4 5 3.1
Tubule Density (mrn-’)
29,100 22,100 19,500 29,200 24,900 22,200 25,600 22,000 21,500 25,000
3.1 3.7 3.3 2.8 3.4 3.8 3.6 2.8 4.1 4.5
78 76 83 81 72 75 74 87 72 61
= 75 Z
40 60 80 100 R . D . S . (%I Figure 1. Correlation between the RDS fraction and the microhardness of 0
The electron microprobe analyses for the same series of specimens led to the [Ca] and [PI weight percentages listed in Table 11. On the average, the composition is made of 25.4 wt % ’ calcium and 12.2 wt 7% phosphorus. The mean stoichiometric ratio is then equal to [Ca]/[P] = 2.08. Furthermore, it is evident from the data of Table I1 that [Ca]/[P] is nearly constant from one tooth to another, while individual values of [Ca] and [PI respectively are subjected to considerable variations. Although the cleansing of the samples with an etchant during 30 s removes the abrasion debris, opens the tubules, and could alter the chemical composition of the dentin to some extent, the acidic dissolution process leaves the [Ca]/[P] ratio in a stoichiometric way. Moreover, the observed data closely approximate those observed in another work,19 where the calculated [Ca]/[P] ratio equals 2.09 for human dentin
INFLUENCE OF CALCIUM CONCENTRATION ON DENTIN WETTABILITY
TABLE I1 Ca and P Composition of the First Series of Dentin Specimens Whose Morphological Features are Listed in Table I Tooth Number
D2 D3 D5 D6 D7 D11 D13 D15 D19 D35
71.4 70.9 80.9 75.8 70.2 66.4 68.2 80.6 57.5 44.5
4.5 3.0 5.0 4.6 2.3 t 1.3 ? 2.7 ? 4.6 ? 5.4 ? 3.1 ? ? ? ? ?
27.0 26.9 28.2 27.5 25.9 24.1 26.5 28.1 21.1 18.5
13.5 12.9 13.7 13.6 12.2 11.8 12.5 14.0 9.2 9.0
2.00 2.09 2.06 2.02 2.12 2.04 2.12 2.01 2.30 2.06
in viva It is also interesting to note that, since the beam of the electron microprobe penetrates the dentin to about 1 pm deep,'7 the data reported in Table I1 characterize actually the superficial mineral composition of the samples after etching and water rinsing. Therefore, it is now possible, from these data, to test the influence of the calcium content on the dentin microhardness. It is clear from Figure 2 that H, increases with [Ca] and that a simple linear correlation holds for the investigated data range with H, = 2.65 [Ca]. For the second series of dentin specimens, the values of the hardness and the adhesive contact angle are reported in Table 111. The large variations of the angle (16.7' < 8 < 54.9') are represented by the cos B vs. H,plot of Figure 3, and show that the wettability of the dentin is quite dependent on its structural state. A positive linear correlation also holds in this case, with cos 8 = 0.0134 H,.
s 2 5 -
( w t Yo)
Figure 2. Influence of Ca content on the microhardness of dentin surfaces.
PANIGHI AND G'SELL
TABLE I11 Hardness and Contact Angle of the Adhesive for the Second Series of Dentin Samples Tooth Number
PI27 D134 P135 D149 P157 D168 D170 0173 D184 D196 D354 P355
53.5 54.7 67.5 69.1 45.3 59.5 59.6 59.0 68.0 71.0 73.4 57.5
36.5 36.0 19.4 31.5 54.9 39.6 35.1 36.9 24.8 19.8 16.7 36.0
Figure 3. Correlation between the microhardness of dentin surfaces and their wettability by the adhesive (cosine of the contact angle).
Comparing lastly the results obtained from the two series of samples, it is now clear that the dentin surfaces with the highest degree of minerality exhibit the largest microhardness and are also the ones best wetted by the Scotchbond adhesive.
Although the chemical composition of dental tissues has an essential influence on their physical properties, only a small number of quantitative studies have included elemental analyses. Among them, the work of J e n k i n ~ ' ~ is of particular interest since it is based on absolute analytical techniques. For
INFLUENCE OF CALCIUM CONCENTRATION ON DENTIN WETTABILITY
human dentin, Jenkins found that the content of calcium and phosphorus in fully dried material are 27 and 1376, respectively. Correcting these data for hydrated tissue, he deduced the corresponding data for the in vim conditions: 24 and 11.5%.The mean values obtained in the present study (25.4 and 12.2%) fall just between those for dry and in vivo states. In fact, the dehydration for the microprobe analysis was performed only under a primary vacuum. During this treatment, it is assumed that the teeth lost a major fraction of their free moisture but retained the whole of the bound water molec~les.'~ In this investigation, microhardness has been shown to be a structural indicator for dentin composition. It is easily measurable and causes negligible damage to the dentin surface. The observed positive correlation between hardness and tooth minerality can be deduced from two parameters. The first is the direct influence of the mineral content on the hardness of the intertubular material. It should be remarked that the electronic beam of the microprobe analyzer (1 pm diameter) was always located between the tubular openings so that the [Ca] and [PI values characterize the intertubular material only. It is then likely that the higher mineralized dentin surface (i.e., with more hydroxyapatite and less collagen) is intrinsically more resistant to the pyramidal Vickers indenter. The second factor corresponds to the indirect effect of the dentin composition on the hardness via the RDS (Tables I and 11). It appears that the surface fraction of the tubules is smaller in specimens with higher minerality, so the weakening effect is reduced. The two factors have thus a cooperative role in the overall increase of H,vs. [Ca]. It is lastly noteworthy that the results discussed in this section are related to those of Fusayama,2" who observed that carious dentin, which is severely depleted in mineral elements, is one third to one fourth less hard than sound dentin, showing that microhardness is very sensitive to tooth composition. We can now examine the significance of the contact angle 8. This parameter appears to be an indicator of the potential of the adhesive drop to spread on a dentin surface. It has been shown that the wettability factor cos 13 increases with the surface roughness of the substrate." In the present case, the spreading ability seems to depend on yet another factor. Since the same grit abrasive paper has been utilized for all the teeth, the roughness Wenzel's factor (i.e., ratio of the developed surface to the apparent surface) is approximately constant. The open tubules correspond to local accidents of the surface and therefore contribute to the roughness at a microscopic scale. However, Tables I and I11 indicate that the highest wettability is observed on the dentin samples with the least tubular surface. Consequently, the above observations suggest that the dominant parameter in the wettability is the intrinsic surface free energy of the intertubular dentin, which in turn depends on the Ca and P concentrations. It is important to note at this point that the ability of the adhesive to wet the substrate indicates its tendency to form bonds with the dentin. Furthermore, following the general conclusions of previous studies (Wu"), it appears that a high wettability of the dentin by the adhesive is a necessary, although not sufficient, condition to ensure a strong adhesion of the resin composite.
PANIGHI AND G’SELL
From a clinical standpoint, the correlation between the adhesive spreading factor and the dentin minerality has interesting consequences, e.g., the poor adhesion that should be expected on the dentin of teeth with very low mineral content. This suggestion is supported by the study of Eidelman et a1.,22 who showed that nearly 75% of class-2 composite fillings placed in deciduous teeth show a distinct crevice at the cervical margin of the cavity. As the polymerization shrinkage is a constant for a given composite, it is logical to assign this extensive cracking to the lower mineralization of the deciduous teeth. Beyond this particular conclusion, the inf hence of the tooth minerality has to be investigated with regard to the adhesion strength of resin composites to the hard tooth tissues. The authors are indebted to the CNRS (Centre National de la Recherche Scientifique, PIRMAT) for supporting this work through a research grant.
References 1. B. Torstenson, M. Brannstrom, and B. Mattsson, ’A new method for 2. 3. 4.
5. 6. 7.
9. 10. 11. 12. 13. 14.
sealing composite resin contraction gaps in lined cavities,” J. Dent. Res., 64, 450-453 (1985). J. R. Swift Jr., “Pulpal effects of composite resin restorations,” Operative Dent., 14, 20-27 (1989). S. Bandyopadhyay, ‘A study of the volumetric setting shrinkage of some dental materials,” J. Biomed. Mater. Res., 16, 135-144 (1982). E. K. Hansen, ”Contraction pattern of composite resins in dentin cavities,’’ Scand. J. Dent. Res., 90, 480-483 (1982). M. Goldman, ”Polymerization shrinkage of resin-based restorative materials,” Aust. Dent. J., 28, 156-161 (1983). T. Hirasawa, S. Hirano, S. Hiribayashi, I. Harashima, and M. Aizawa, ”Initial dimensional change of composites in dry and wet conditions,” J. Dent. Res., 61, 28-31 (1983). A. J. Feilzer, A. J. De Gee, and C. L. Davidson, “Curing contraction of composites and glass-ionomer cements,” J. Prosthet. Dent., 59, 297-300 (1988). D. H. Retief, E. Woods, and H. C. Jamison, ”Effect of cavosurface treatment on marginal leakage in class V composite resin restorations,” I. Prostket. Dent., 47, 496-501 (1982). E. K. Hansen, “Effect of Scotchbond dependant of cavity cleaning, cavity diameter and cavosurface angle,” Scand. J. Dent. Xes., 92, 141-147 (1984). G. Vanherle, M. Verschueren, P. Lambrechts, and M. Braem, ”Clinical investigation of dental adhesive systems. Part I: An in-vivo study,” J. Prostket. Dent., 55, 157-163 (1986). M. Braem, P. Lambrechts, and G. Vanherle, ”Clinical investigation of dental adhesive systems. Part 11: A scanning electron microscopy study,” J. Prosthet. Dent., 55, 551-560 (1986). 0. Zidan, 0. Gomez-Marin, and T. Tsuchiya, ‘A comparative study of the effects of dentinal bonding agents and application techniques on marginal gaps in class V cavities,” 1. Dent. Res., 66,716-721 (1987). Council on Dental Materials, Instruments, and Equipment, “Dentin bonding system: An update,” 1.A m . Dent. Assoc., 114, 91-95 (1987). S. Wu, ”Polymer interface and adhesion,” Marcel Dekker, New York, 1982.
INFLUENCE OF CALCIUM CONCENTRATION ON DENTIN WETTABILITY 15.
19. 20. 21. 22.
M. Panighi and C. G’Sell, ”Corrglation de I’adh6sion d u n e r6sine composite aux tissus am6lo-dentinaires avec leur microduret6,” Proceedings of the 3rd Annual Conference, C.EB.D., Paris, April 18, 1986, p. 71. M. Panighi and C. G’Sell, “Etude des mgcanismes d’adhgsion de r6sines composites 2 l’6mail et 21 la dentine,” in Progu2s Ricents des Biornatt!riaux, R. Bgnichoux and J. Lacoste (eds.), Masson, Paris, 1988, pp. 153-178. J. Ruste, ’knalyse quantitative au microanalyseur 21 sonde glectronique de Castaing,” ThPse de Docteur-Inghnieur, Universitg de Nancy I, Nancy, France, 1976. D.H. Pashley, A. Okabe, and P. Parham, “The relationship between dentin microhardness and tubule density,” Endod. Dent. Traumatol., 1, 176-179 (1985). G. N. Jenkins, “Chemical composition of teeth,” in The Physiology und Biochemistry of the Mouth, 4th ed., Blackwell Scientific Publications, Oxford, 1978, pp. 57-59. T. Fusayama, ”Two layers of carious dentin,” in New Concepts in Operative Dentistry, Quintessence Publishing, Chicago, IL, 1980, pp. 18-19. R. N. Wenzel, “Surface roughness and contact angles,” 1. Phys. Chem., 53, 1466 (1949). E. Eidelman, A. Fuks, and A. Chosack, ’kclinical, radiographic, and SEM evaluation of class 2 composite restorations in primary teeth,” Operative Dent., 14, 58-63 (1989).
Received May 10, 1991 Accepted December 20, 1991