Dent Mater 8:10-15, January, 1992
Chemical characterizationof the dentin/adhesiveinterface by FourierTransformInfraredPhotoacousticSpectroscopy P. Spencer ~, T.J. Byerley 2, J.D. Eick 3, J.D. Witt 2 Departments of Pediatric Dentistry, 3Oral Biology University of Missouri-Kansas City, School of Dentistry, Kansas City, MO USA 2Midwest Research Institute, Kansas City, MO USA
Abstract. Irreversible bonding of composite materials to tooth structure depends on chemical as well as mechanical adhesion. The proposed bonding mechanism for several commercial dental adhesives is chemical adhesion to the dentin surface. The purpose of this in vitro investigation was to characterize the chemical nature of the surface interaction between dentin and two commercial adhesives by use of Fourier transform infrared photoacoustic spectroscopy (FTIR/PAS). The occlusal thirds of the crowns of freshly extracted, non-carious, unerupted human molars were sectioned perpendicular to the long axis. Dentin disks, 6 mm x 2 mm, were prepared from these sectioned teeth. The exposed dentin surface was treated with either Scotchbond 2, a BIS-GMA resin, or Dentin-Adhesit, a polyurethane resin. All spectra were recorded from 4000 to 400 cm -1 by use of an Analect RFX-65 FTIR spectrometer equipped with an MTEC Photoacoustics Model 200 photoacoustic cell. An initial spectrum of the dentin surface was collected. This surface was primed according to manufacturer's instructions and spectra recorded of the primed surface plus one to three layers of adhesive. By comparison of these spectra, it was possible for us to record changes in the phosphate and amide I and II bands due to surface interactions between the adhesive and the dentin. Although early results do not indicate covalent bonding between the dentin and these adhesives, this technique presents several advantages for spectroscopic evaluation of the dentin/adhesive interface.
A critical factor in the clinical success of resin composites is the adhesive bond formed at the restorative material/tooth surface interface. Buonocore's acid-etch technique, introduced in 1955, promoted mechanical bonding between the composite restoration and the enamel surface (Buonocore, 1963). Indeed, this technique is recognized today as a fundamental tool in promoting the clinical viability of dental composites. Further improvement in this restoration, however, depends on the establishment of an adhesive bond between the dentin surface and the resin composite. Establishment of such a bond has been a formidable problem. Chemical bonding to the dentin via the collagen matrix or the calcium ions is a challenging problem that is further complicatedby the presence ofthe smear layer (Pashley, 1990). Even chemical and physical analysis of the dentin/ resin bond is compromised by the complex, heterogeneous structure of dentin. Consequently, much of our understanding about this bond has been derived from either morphologic characterization (Roulet and Blunck, 1989; Roulet et al., 1989) or bond strength measurements (Pashley, 1990). Little
10 Spencer et aL/FTIR photoacoustic spectroscopy of dentin~adhesive
information (Misra and Johnston, 1987; Misra, 1989) has been presented on the chemistry of the dentin/resin bond. Infrared spectroscopy is a popular tool for characterization of the chemical reactivity of dental materials (Ferracane and Greener, 1984; Culler et al., 1986; Nicholson et al., 1988; Rueggeberg and Craig, 1988). Several authors (Vankerckhoven et al., 1982; Ruyter and Oysaed, 1987) have used multiple internal reflection infrared spectroscopy to monitor the extent of polymerization conversion in dental composites. Because of poor optical contact between the internal reflection element and rigid, bulk biological samples, it is very difficult for this technique to be used to record spectra from a dentin surface. Thus, this technique is not readily applicable to evaluation of the chemical structure at the dentin/resin interface. Another technique, transmission infrared spectroscopy, requires preparatory procedures that may interfere with evaluation of the chemistry at the dentin/adhesive interface (Spencer et al., 1990). Photoacoustic spectroscopy is a unique sampling method in infrared spectroscopy. In photoacoustic spectroscopy, the heat generated by the absorption of light at the surface of a solid sample is measured. For example, a solid biological sample is placed in a sealed chamber containing an infrared transparent coupling gas, such as helium. When the sample is irradiated, any absorbed light creates heat at the sample surface and a resultant temperature and pressure increase in the surrounding gas. Because the input radiation is modulated at a frequency in the audio range by the interferometer, the pressure fluctuations occurring at this modulation frequencycan be detected by the sensitive microphone in the photoacoustic cell. Because the magnitude of the acoustic signal is proportional to the amount of heat emanating from the solid surface, there is a direct correlation between the strength of the acoustic signal and the light absorbed by the solid sample (Rosencwaig, 1975; Adams and Kirkbright, 1977; Mandelis and Royce, 1980). The purpose ofthis in vitro investigation was to characterize the chemistry of the dentin/adhesive interface by use of FTIR photoacoustic spectroscopy. This report represents, for the first time, the application ofinfrared photoacoustic spectroscopy to the study of solid biological samples and their interaction with adhesive materials. MATERIALS AND METHODS Extracted non-carious, unerupted human molars, which were stored upon extraction at 3°C in buffered saline, were randomly selected for treatment with either Scotchbond 2, a BIS-GMA resin, or Dentin-Adhesit, a polyurethane resin (Table 1). The occlusal one-third of each crown was sectioned perpendicular to
[ ~ l , l ~ l , l 3000
Fig. 1. FTIR photoacoustic spectrum of dentin surface.
the long axis of the tooth by means of a water-cooled low-speed diamond saw (Isomet, Buehler Ltd., Lake Bluff, IL). A second cut was made parallel to and 2 mm below this surface. Dentin disks, 6 mm in diameter and 2 mm thick, were made from these slices by means ofa carbide bur in a high-speed dental handpiece with water spray. The spectra were collected from the diamond cut surface furthest from the pulp. These samples were stored in a desiccator at 3°C for 10 days prior to examination in the photoacoustic cell. FTIR/PAS spectra were recorded by use of an Analect RFX 65 Fourier transform infrared spectrometer, equipped with an MTEC Photoacoustic Model 200 photoacoustic cell. All spectra were collected at a resolution of 8 cm-~,with a scan speed of 0.5 cm/sec. An improved signal-to-noise ratio was obtained by collection of 1000 scans for each sample. The sample singlebeam spectra were ratioed against a carbon black reference supplied by MTEC Photoacoustics. The photoacoustic detector is responsive over the spectral range of 400-4000 cm-L A spectrum was first recorded from the dentin surface. This surface was then primed according to manufacturer's instructions and a spectrum recorded of the primed surface. A layer of dentin adhesive, either Scotchbond 2 or Dentin-Adhesit, was applied and dried to a thin layer by means of an air syringe. The adhesive was then polymerized by exposure to a visible-light source for two min. The thickness of this initial layer of adhesive was subsequently determined by direct measurement on an SEM photomicrograph. The thickness was slightly greater than 1 ~m~. A spectrum was recorded of the dentin surface plus this initial layer of adhesive. Successive spectra were recorded as more adhesive was added. Samples were polymerized with a visible-light source for two min following each application of the adhesive. This sequence was continued until the spectrum showed contributions from only the adhesive. Samples of each adhesive were made by polymerization of layers of the material on a AgC1disk. Spectra of these standard samples were subtracted from composite adhesive/dentin spectra so that the interaction between the adhesive and the
Figs. 2a.d. FTIR photoacoustic spectra of Scotchbond 2 on dentin substrate. (a) primed dentin surface; (b) initial layer of Scotchbond 2 applied to primed surface; (c) second layer of Scotchbond 2; and (d) third and final layer of Scotchbond 2.
substrate could be evaluated. Similarly, the respective dentin spectra were subtracted from the composite adhesive/dentin spectra. RESULTS
The FTIR/PAS spectrum of a dentin sample is shown in Fig. 1. The observed spectrum reveals vibrations of both the organic and inorganic components. The major protein contributions are indicated by the absorption bands at 1650 cm-1 (Amide I, C=O), 1550 cm-1(Amide II, N-H and C-N), and 1240 cm-1(Amide III, C-N and N-H). Contributions from the mineral phase are the orthophosphate bands at 1030-1160 cm 1 and 600-560 cm-1 and the carbonate band at 870 cm-L The carbonate band at 1450-1425 cm-1is overlapped with contributions from the Amide II band. A broad O-H absorption band is recorded between 3500 and 3300 cm-L The FTIR/PAS spectra of Scotchbond 2 on the dentin substrate are shown in Figs. 2a-d. Fig. 2a is a spectrum of the dentin surface following application ofthe Scotchbond 2 primer. The C-H absorption band at 2950-3000 cm-1is enhanced in this
TABLE 1: MATERIALSUSED Product and Composition
Scotchbond 2- BIS-GMA Resin, Primer:maleicacid and hydroxyethylmethacrylate
3M Dental Products Division
Dentin-Adhesit - PolyurethaneResin, Primer:5% sodium hypochlorite
Dental Materials/January 1992 11
. 12~0 (~
Wave, n u m b e r c m 4
Figs. 4a-c. FTIR/PASspectra of Scotchbond 2 on dentin substrate. (a) Composite of Scotchbond 2/dentin; (b) dentin substrate; and (c) subtraction spectrum, a-b.
, J I , I ,I I , I , 3000 2000 Wavenumber
Figs. 3a-d. FTIR photoacoustic spectra of Dentin-Adhesit on dentin substrate. (a) primed dentin surface; (b) initial layer of Dentin-Adhesit applied to primed surface; (c) second layer of Dentin-Adhesit; and (d) third and final layer of Dentin-Adhesit.
spectrum. In Fig. 2b, Scotchbond 2 was applied and reduced to a very thin layer by means of compressed air. The major contributions from the dentin adhesive are the C-H absorption bands at 2950-3000 cm~ and the aryl group at 1520 cm1. Fig. 2c shows the spectrum recorded when two layers of Scotchbond 2 were applied. Tentative assignment of the absorption bands, which are characteristic of the adhesive, is given in Table 2 TABLE 2: TENTATIVEBANDASSIGNMENTFOR BIS-GMARESIN
3. C=O ester
8. C-O-C6H4aliphaticaryl ether
9. C-O-Caliphaticetherfrom TEGDMA
The absorptionband at 2400 is attributableto atmosphericCO2.
12 Spencer et aL/FTIR photoacoustic spectroscopy of dentin/adhesive
(Bellamy, 1958; Williams and Braden, 1981). The spectrum recorded when a third and final coat ofScotchbond 2 was added to this sample is shown in Fig. 2d. This spectrum shows no change in absorption band position from the previous spectrum. The FTIR/PAS spectra of Dentin-Adhesit on a dentin substrate are shown in Figs. 3a-d. Fig. 3a is the spectrum of the dentin surface following application of the dentin conditioner. The phosphate regions at 1030-1160 cm1 and 600-560 cm-1 show broadening, suggestive of possible destruction of some dentin material. Figs. 3b-d show successive spectra of the dentin substrate with layers of Dentin-Adhesit applied. Fig. 3b shows contributions from the adhesive at 2225 cm-1 (isocyanate) and at 1220 cm-1(ester C-O). Bands attributable to the protein fraction of the dentin substrate are recorded at the following positions: 1650 cm-1and 1550 cm-1. With the addition of a second layer of Dentin-Adhesit, additional absorption bands attributable to the adhesive are noted. As shown in Fig. 3c, a C-H band is noted at 2950-3000 cm-1,the isocyanate band at 2225 cm1, and an ester band (C=O) at 1730 cm-L The spectrum recorded following the application of a third coat of Dentin-Adhesit is shown in Fig. 3d. No further changes were noted in this spectrum with the application of additional layers of adhesive. Tentative assignment of the absorption bands, which are characteristic of this adhesive, is given in Table 3 (Bellamy, 1958; Williams and Braden, 1981). Fig. 4 shows the FTIR/PAS spectrum of the initial layer of Scotchbond 2 on a dentin substrate. The upper spectrum is the composite of dentin/adhesive, the middle spectrum is the dentin substrate, and the bottom spectrum is the composite spectrum after subtraction of the dentin. Fig. 5 shows the FTIR/PAS spectrum of three layers of Scotchbond 2 on a dentin substrate. The upper spectrum is the composite of dentin/adhesive, the middle spectrum is the adhesive, and the bottom spectrum is the composite spectrum after subtraction of the adhesive. The straight line, which is computed following subtraction of these spectra, suggests that there is virtually no shift in absorption band position between these spectra. The FTIR/PAS spectrum of the initial layer of DentinAdhesit on a dentin substrate is shown in Fig. 6. The upper
Figs. 6a-c. FTIR/PAS spectra of Dentin-Adhesiton dentin substrate. (a) Compositeof Dentin-Adhesit/dentin;(b) dentin substrate; and (c) subtraction spectrum, a-b.
c m "t
Figs. 5a-c. FTIR/PAS spectra of Scotchbond 2 on dentin substrate. (a) Composite of Scotchbond 2/dentin; (b) Scotchbond 2; and (c) subtraction spectrum, a-b,
spectrum is the composite of dentin/adhesive, the middle spectrum is the dentin substrate, and the bottom spectrum is the composite spectrum after subtraction of the dentin. Fig. 7 shows the FTIPJPAS spectrum of four layers of Dentin-Adhesit on a dentin substrate. The upper spectrum is the composite of dentin/adhesive, the middle spectrum is the adhesive, and the bottom spectrum is the composite spectrum after subtraction of the adhesive. The straight line, which is computed following subtraction suggests that there is virtually no shift in absorption band position. DISCUSSION Infrared spectroscopy has previously been used to study the chemical nature of adhesive resins (Ruyter and Gyoroski, 1976; Williams and Braden, 1981). The present study describes for the first time how the chemical interaction between the dentin surface and an adhesive resin was studied by means of FTIR photoacoustic spectroscopy. A principal advantage of this technique is that the samples may be studied without any additional handling or preparation that might degrade the interface. However, as with most techniques, this procedure has inherent limitations. A fundamental problem is spectral interference from water. Water absorbs strongly in many locations in the infrared spectral region, including the C-H stretching region (3100-2800 cm-1) and part of the biological fingerprint region (1800-1300 cm-1) (Bohlke et al., 1989). In this study, it was necessary to desiccate the dentin samples so that spectra with an adequate signal/noise ratio could be collected. Although it would be preferable to observe
wet dentin samples, biological specimens must frequently be dehydrated so that their morphologic features as well as molecular structure can be observed. It is interesting that, in spite of this limitation, these data corroborate previous evaluations of the chemistry of the dentin/resin bond (Eliades et al., 1990; Misra and Johnston, 1987; Misra, 1989). It has been suggested that dentin adhesives adhere chemically to the dentin surface via bonding to the organic or inorganic phase. Such bonding could be demonstrated spectroscopically by a shift in the absorption bands attributable to the dentin substrate (Urban and Salazar-Rojas, 1990). For example, if a dentin bonding agent were chemically bound to the organic phase, a shift in these absorption bands should be recorded in the composite spectrum of the dentin plus adhesive. In contrast, if the adhesive does not interact with the functional groups of the dentin surface, the composite spectrum would represent the adhesive overlaid on a dentin spectrum with no change in absorption band position. Following spectral subtraction, it is apparent, in this study, that the spectrum of the dentin plus adhesive does not show a TABLE3: TENTATIVEBANDASSIGNMENTFOR POLYURETHANERESIN Band
3. -N=C=O (isocyanate)
4. C=O (ester)
O 5. R-N-C-O-R (urethane)
6. C-O (ester)
7. C-O-C (ether)
Dental Materials~January 1992 13
lS3O ]P~. /
tive, samples can be used for multiple analyses. Spectra can be obtained from samples that are completely opaque to transmitted light, thus eliminating any structural defects that might be produced by either thin-sectioning or grinding of the material (Mendelsohn et al., 1989; Renugopalakrishnan et al., 1989). Because this technique analyzes the molecular structure a few micrometers into the surface of a dentin sample, it allows for evaluation of chemical interactions near the interface between the dentin and the adhesive. Finally, although these early results suggest no covalent bonding between these commercial adhesives and the dentin substrate, further evaluation is needed. Other techniques which do not exhibit significant spectral interference from water, e.g., Raman spectroscopy, should be thoroughly explored.
The authors are indebted to Ms. Eileen Roach and Ms. Laura Marshall for their technical assistance. This investigation was supported in part by USPHS Research Grants K11 DE-00260-02 and DE-08223-03 from the National Institute of Dental Research, Bethesda, MD 20892.
ReceivedMarch 28, 1991/AcceptedAugust 26, 1991 Address correspondenceand reprint requests to: P. Spencer University of Missouri-KansasCity Schoolof Dentistry 650 E. 25th Street Kansas City MO 64108 USA m Wavenumber
Figs. 7a-c. FTIR/PASspectra of Dentin-Adhesit on dentin substrate. (a) Composite of Dentin-Adhesit/dentin; (b) Dentin-Adhesit; and (c) subtraction spectrum, a-b.
shift in the absorption bands attributable to dentin with either the BIS-GMA or the polyurethane resin. These early results suggest that the commercial adhesives used in this study are not covalently bonded to the dentin substrate. Eliades et al. (1990), using electron spectroscopy for chemical analysis (ESCA), showed similar results between dentin and Scotchbond 2. That is, in their comparison of Gluma, Scotchbond 2, and Tenure, they did not report substantial binding energy shifts, which could indicate primary bonding, with any of these commercial adhesives. In the absence of chemical bonding, adhesion at the interface must be the result of other, potentially weaker, forces. For example, because an intimate contact is established between the conditioned dentin surface and the adhesive resin, secondary forces, such as van der Waals forces, could contribute to the adhesion. Mechanical retention via resin penetration into the microscopic porosities in the subsurface intertubular dentin may also contribute to the bond (Pashley, 1990). In summary, a fundamental aspect of our efforts to develop a dentin adhesive that bonds chemically is the identification of available bonding sites on the dentin surface and the measurement of chemical interactions at these sites. In this study, FTIR photoacoustic spectroscopy was used to determine the surface functionality of dentin and to measure the chemical processes at these sites. This technique offers several advantages for the study of adhesive interactions with the dentin substrate. For example, because the technique is non-destruc-
14 Spencer et aL/FTIR photoacoustic spectroscopy of dentin~adhesive
REFERENCES Adams MJ, Kirkbright GF (1977). Analytical optoacoustic spectroscopy. Part III. The optoacoustic effect and thermal diffusivity. Analyst 102:281-292. Bellamy LJ (1958). The infrared spectra of complex molecules. London: Methuen & Co., Ltd., 75-87. Bohlke AP, Lin-Vien RM, Hammaker RM, Fateley WG (1989). Hadamard transform spectrometry: application to biological systems, a review. In: Theophanides T, editor. Spectroscopy of inorganic bioactivators. Theory and applications--chemistry, physics, biology, and medicine. Norwall, MA: Kluwer Academic Publishers, 159-189. Buonocore MG (1963). Principles of adhesive retention and adhesive restorative materials. J A m DentAssoc 67:72-76. Culler SR, Ishida H, Koenig JL (1986). FT-IR characterization of the reaction at the silane/matrix resin: interphase of composite resins. J Colloid Polym Sci 109:1-10. Eliades G, Palaghias G, Vougiouklakis G (1990). Surface reactions of adhesives on dentin. Dent Mater 6:208-216. Ferracane JL, Greener EH (1984). Fourier transform infrared analysis of degree of polymerization in unfilled resins-methods comparison. JDent Res 63:1093-1095. Mandelis A, Royce BSH (1980). Relaxation time measurements in frequency and time-domain photoacoustic spectroscopyof condensed phases. J Opt Soc Am 70:474-480. Mendelsohn R, Hassankhani A, DiCarlo E, Boskey A (1989). FT-IR microscopyofendochondral ossification at 2 0 spatial resolution. Calcif Tissue Int 44:20-24. Misra DN (1989). Adsorption of 4-methacryloxyethyl trimellitate anhydride (4-META) on hydroxyapatite and its role in composite bonding. JDent Res 68:42-27. Misra DN, Johnston AD (1987). Adsorption of N-phenylglycine
on hydroxyapatite: role in the bonding procedure of a restorative resin to dentin. J Biomed Mater Res 21:13291339. Nicholson JW, Brookman PJ, Lacy OM, Wilson AD (1988). Fourier transform infrared spectroscopicstudy of the role of tartaric acid in glass-ionomer dental cements. J Dent Res 67:1451-1454. Pashley DH (1990). Interaction of dental materials with dentin. Proceedings of enamel-dentin-pulp-bone-periodontal tissue interactions with dental materials. Trans Acad Dent Mater 3:55-73. Renugopalakrishnan V, Chandrakanaan G, Moore S, Hutson TB, Berney CV, Bhatnagan RS (1989). Bound water in collagen. Evidence from Fourier transform infrared and Fourier transform infrared photoacoustic spectroscopic study. Macromolecules 22:4121-4124. Rosencwaig A (1975). Photoacoustic spectroscopy. A new tool of investigation of solids. Anal Chem 47:592A-604A. Roulet JF, Blunck U (1989). Effectiveness of dentin bonding agents. Scanning Electron Microsc 3:1013-1022. Roulet JF, Reich T, Blunck U, Noack M (1989). Quantitative margin analysis in the SEM: A powerful tool for the
characterization ofthe behavior ofdental restorations. Scanning Electron Microsc 3:147-159. Rueggeberg FA, Craig RG (1988). Correlation of parameters used to estimate monomer conversion in a light-cured composite. J Dent Res 67:932-937. Ruyter IE, GyoroskiPP (1976). An infrared spectroscopicstudy of sealants. Scand J Dent Res 84:396-400. Ruyter IE, ~ysaed H (1987). Composites for use in posterior teeth: composition and conversion. J Biomed Mater Res 21:11-23. Spencer P, Eick JD, Byerley TJ (1990). FTIR microspectroscopy of the dentin/adhesive interface (abstract). J Dent Res 69:116. Urban MW, Salazar-Rojas EM (1990). Probingorganic-inorganic interactions and curing processesin coatingsby photoacoustic Fourier transform infrared spectroscopy.JPolym Sci 28 (Part A):1593-1613. Vankerckhoven H, Lambrechts P, van Beylen M, Davidson CL, van Herle G (1982). Unreacted methacrylate groups on the surfaces of composite resins. J Dent Res 61:791-795. Williams B, Braden M (1981). Characteristics of fissure sealants. J Dent Res 60:990-994.
Dental Materials~January 1992 15