Dent Mater 8:125-130, March, 1992
Effect of HEMA on bonding to dentin N. Nakabayashi, K. Takarada Division of Organic Materials, Institute for Medical and Dental Engineering, Tokyo Medical and Dental University, Tokyo, Japan
Abstract. The present study investigated the effectiveness of treating dentin with 2-hydroxyethyl methacrylate (HEMA) prior to application of an adhesive resin. The adhesive resin was 5% 4methacryloxyethyl trimellitate anhydride (4-META) in methyl methacrylate (MMA) combined with poly-MMA powder. Polymerization of this resin was initiated by tri-n-butyl borane (TBB). Bovinedentin samples were ground with 600-grit Carbimet paper discs, and demineralized with either an aqueous solution of 10% citric acid/3% ferric chloride (10-3) or an aqueous solution of 10% citric acid (10-0). Improved bond strengths were achieved with HEMA treatment of bovine dentin samples, and improvement of bond strengths was dependent upon the time period of HEMA application. Scanning electron microscope (SEM) examination revealed the formation of a transitional zone of resin-reinforced dentin, termed the "hybrid" layer, in those specimens receiving 103 pre-treatment. The adhesive resin impregnated the exposed collagen bundles with which it entangled to create the "hybrid", essential in the attainment of high tensile bond strengths. Specimens pre-treated with 10-0 did not readily form "hybrid" layers. However, if HEMA application followed the 10-0 pretreatment, "hybrids"were demonstrated on SEM, and bondstrength increased to 13 MPa. The ferric ions in the 10-3 effectively improved the diffusivity of dentinal substrates, asdid HEMA. This study indicates that HEMA applied to denti nal substrates enhances monomer diffusion and entanglement with dentinal components, and facilitates the formation of a "hybrid" layer. Methacrylates that contain both hydrophobic and hydrophilic groups on the ester molecule have been developed. These groups improve the adhesive strength of resins to teeth by promotinginterpenetration, impregnation, and entanglement ofthe methacrylate-based monomers into dentinal substrates and their polymerization therein (Nakabayashi, 1982; 1984; Nakabayashiet al., 1982; Nikaido, 1989; Inokoshiet al., 1990; Wang andNakabayashi, 1991). Examples ofthese methacrylates include 4-methacryloxyethyltrimellitate anhydride (4META; Takeyama et al., 1978), 2-(methacryloxy)ethylphenyl hydrogen phosphate (phenyl-P; Yamauchi, 1986), and derivatives ofphenyl-P(Nakabayashi and Kanda, 1988). An adhesive system, 5% 4-methacryloxyethyl trimellitate anhydride (4META) in methyl methacrylate (MMA) with poly-MMA powder initiated by partially oxidized tri-n-butyl borane (TBB; Nakabayashi et al., 1978) and designated as 4-META/MMATBB resin, has been widely studied as a bonding system for tooth substrates (Nakabayashi, 1984), metals (Takeyama et al., 1978; Matsumura and Nakabayashi, 1988; Tanaka etal., 1988) and porcelains (Maeda et al., 1987). The bonding mechanism of 4-META/MMA-TBB resin has been reported (Nakabayashi, 1984; Kato et al., 1986), a n d recently the mechanism for phenyl-P/MMA-TBB has also been studied and reported (Wang and Nakabayashi, 1991). This
same bonding mechanism might be applicable to the adhesion of photo-cure bonding agents to EDTA-pre-treated dentin (Nikaido, 1989; Wanget al., 1991). These investigations have shown that the bond strength of adhesive resins to dentinal substrates can be altered by the rate ofmonomer diffusion into demineralized dentin after smear layer removal. The diffusion rate of monomer is a function of both the diffusivity of the dentinal substrate and the diffusibility ofthe monomer itself. Bond strengths were lower if dentin was pre-treated with either phosphoric acid or citric acid alone, compared with bond strengths attained if dentin was pre-treated with 10-3 (an aqueous mixture of 10% citric acid and 3% ferric chloride). SEM examinations ofcross-sectional samples confirmedthat monomers did not diffuse into dentin substrates pre-treated with the 'pure' acids, either phosphoricor citric (Mizunuma, 1986;Wang andNakabayashi, 1991). In 1990, Suzukietal. reported that 2-hydroxyethyl methacrylate (HEMA) applied to phosphoric acid pre-treated dentin made the dentinal substrate more penetrable. The report suggested that HEMAimproveddentin diffusivity. Several commercially available adhesive agents that claim bonding to dentin contain HEMA. According to a 1984 studybyMunksgaard andAsmussen, the bondingmechanism for GLUMA might be a chemical reaction of HEMA with glutaraldehyde and collagen. In a 1985 study by Itoh et al., glutaraldehyde was eliminated from the GLUMA system, and that material was combinedwith Clearfil New Bond (Kuraray, Japan), which contained methacryloyloxydecamethylenephosphoric acid. In the present study, the influence of HEMA on dentin bondingbecame more obvious during SEM and TEM examinations oftest specimens before and after its application. HEMA improved the diffusivity ofdentin substrates, and higher bond strengths were obtained.
MATERIALSAND METHODS Bonding of a PMMA Rod to Extracted Bovine Dentin with 4-META/MMA-TBB Resin. Extracted bovine teeth samples were frozen so that freshness would be maintained during storage. The specimens were ground under a stream of water with Carbimet paper discs (Buehler, Lake Bluff, IL, USA) sequentially from 180- to 600-grit. These surfaces were then treated with 40 pL of either the 10-3 solution (10% citric acid/ 3% ferric chloride) or the 10-0 solution (10% citric acid/0% ferric chloride). Dentin pre-treatment times with these solutionswere 10 s, followedby a 15 s water rinse. The samples were then blown dry. The pre-treated specimens were divided into two groups. One group received immediate application of the adhesive resin. The second received a 40 pL application of 30% HEMA (Mitsubishi Rayon Co., Tokyo, Japan)before adhesive resin was applied. This second group was further divided into subgroups that received HEMA application for different Dental Materials~March 1992 125
~riods. All test groups had a piece of tape with a Ldiameter hole firmly attached to dentin surfaces. cved to standardize the dentin surface area available ding. The adhesive resin used in this study was a e of 0.1 g of 5% 4-methacryloxyethyl trimellitate ide (Sun Medical Co., Kyoto, Japan)/methyl methacMitsubishi Rayon Co.)/tri-n-butyl borane (Sun MediI and 0.1 g of ground PMMA powder (Sun Medical ?his adhesive resin, designated as 4-META/MMAsin, was applied (by means of a brush-on technique) ;.4 mm diameter of exposed, pre-treated dentin test ens. Finally, a 6 mm diameter PMMA rod (Mitsubishi Co.) was perpendicularly bonded to the adhesive 1 the flat dentin specimen surfaces. This PMMA rod as a handle (Tanaka et al., 1988). The entire assemleft standing at room temperature for 30 min before laced into 37°C water for 24 h storage prior to being ~ested. The tensile bond strengths of test specimens masured with a universal testing machine (Auto)SS 500, Shimadzu Co., Kyoto, Japan) at a crossreed of 2.0 mm/min. Adhesion of the PMMA rod to dentin after HEMA-only application was also tried. vere four specimens in each test grouping. Data were ~'d to Student's t test at the 5% level to determine ~ y significant differences. ~ning Electron Microscope (SEM) Observations. Speci'ere cut so that cross-sectional samples of the dentin/ e resin interface could be obtained. The specimens were cl to a size of 3 mm wide x 3 mm long x 2 mm high. ag specimen treatment (described below), they were air. room temperature, then dried further in a vacuum at ~r 24 h. Surfaces were then gold-sputtered for SEM ation with a CSM-500 (Comtec, Tokyo, Japan). imen Treatment Prior to Being Dried. Each sample )und with 600-grit emery paper and polished with alumina under a stream of water. The polished ~ns were immersed in 6 mol/L HC1 for 30 s for ~ion of the mineral component of the dentin and then ed by SEM. This was followed by immersion in 1% for either 10 rain, 30 min, or 60 min for removal of ]ous material and low resin content components, hich the specimens were re-examined under SEM. ~smission Electron Microscope (TEM) Observations. dn sections of the dentin/adhesive resin interface epared directly from bonded specimens without epoxy ting. TEM examinations were carried out with a I H600 (Hitachi, Hitachi, Japan) before and after 5cation with 0.001 mol/L HC1 for 4 s to remove the gapatite.
ross-sectional SEM view of a specimen of 4-MET/VMMA-TBB resin ground bovine dentin that had been pre-treated with 10-3 solution a=left): Specimen after immersion in 6 mol/L HCI for 30 s; (b=right): cimen after a 10 min immersion in 1% NaOCI. R = resin; H = "hybrid" = demineralized dentin; and T = resin tag.
abayashi & Takarada/Effect of HEMA on bonding to dentin
TABLE:EFFECTOF HEMAAS A SECONDTREATMENTON BONDSTRENGTH* Treatmentlength (min)
13.3±1.9 4.9±0.5 18.7±0.9
There are significant differences in the bond strengthswith HEMA2nd treatment (p < 0.05), except as indicated by the vertical line (p>0.05).
Effects of HEMA on Adhesion to Dentin. The effects of HEMA application directly and following dentin pre-treatment with either the 10-3 or 10-0 solution, and the application time periods ofHEMA to test samples, are summarized in the Table. HEMA significantly improved the tensile bond strength of 4META/MMA-TBB resin to ground and pre-treated bovine dentin (p < 0.05). The enhanced bond strengths were strongly dependent upon the length of time of HEMA application. Scanning Electron Microscope Observations. Fig. la is a cross-sectional view ofthe dentin/adhesive resin interface after immersion in 6 mol/L HC1 for 30 s, as seen in an SEM examination. Dentin has been pre-treated with the 10-3 solution. The "hybrid" layer (H) is visible in the middle ofthis view; to the right is cured resin (R); and to the left is partially demineralized dentin (DD). The DD is at a lower level than H and R. This specimen was subjected to 1% NaOC1 for 10 min, and its altered appearance is seen in Fig. lb. Collagen-rich materials were depleted by NaOC1 immersion, and resin tags (T) that were embedded in dentin become more apparent, as did the "hybrid" layer (H). Note that the "hybrid" layer (H) appears narrower in Fig. lb, after NaOC1 immersion, than in Fig. la. The NaOClimmersion has depleted that portion ofthe"hybrid" layer that was not fully impregnated and entangled with in situ-polymerized resin--in other words, that portion of"hybrid" with low resin concentrations. Fig. 2a is a specimen prepared in the same way as that pictured in Fig. la, with the exception that dentin pre-treatment was accomplished with the 10-0 rather than the 10-3 solution. A thin "hybrid-like" layer (H') is visible in the middle of this view. A gap between H' and cured resin (R) is visible.
Fig. 2. Cross-sectional SEM view of a specimen of 4-METAJMMA-TBB resin bonded to ground bovine dentin that had been pre-treated with 10-0 for 10 s. (a=left): Specimen subjected to immersion in 6 mol/L HCI for 30 s. Note partial detachment at interface, H cannot be seen. Thin H' is visible. (b=right): Same specimen after an additional 10 min immersion in 1% NaOCI. Note that H' has decomposed completely and has disappeared, H' = "hybrid-like" layer; other codes are the same as in Fig, 1.
Fig. 3. (a=left): Another perspective of specimen seen in Fig. lb; (b=right): Another perspective of the specimen seen in Fig. 2b, Compare tag "necks" between these specimens. Note H' is difficult to identify. Codes are the same as in Fig. 1.
Fig. 4. Cross-sectional SEM view of a specimen of 4-META/MMA-TBB resin bonded to ground bovine dentin that had been pre-treated with 10-0 for 10 s. Specimen, in (a=left), was treated to a 10 min application of 30% HEMA; then immersed in 6 mol/L HCI for 30 s, and 10 rain of immersion in 1% NaOCI. Note this "hybrid" layer (H) at the interface. Specimen, in (b=right) received the same treatment except the 30% HEMA application was for 60 min. Note wider "hybrid" layer (H) than is visible in (a). Codes are the same as in Fig. 1.
Fig. 2b depicts the same sample after immersion in 1% NaOC1 for 10 min to decompose collagenous materials. In addition to collagenous material, the 1% NaOC1 immersion also decomposed the "hybrid-like" layer (H'). Compare this with the "hybrid" (H) that remained after NaOC1 treatment (depicted in Fig. lb). In Fig. 2b, this low resin-content zone has completely disappeared. Resin tags (T) are continuous with cured resin (R), but their "necks" [that portion immediately adjacent to cured resin (R)] are narrower than those visible in Fig. lb. Figs. 3a and 3b are different perspectives of samples depicted in Figs. lb and 2b. Fig. 4 shows specimens prepared in the same manner as that seen in Fig. 3a, except that HEMA was applied to 10-0 pretreated dentin for 10 rain (Fig. 4a) and 60 rain (Fig. 4b). Note that HEMA application has encouraged"hybrid"formation in both specimens, but they appear narrower than the "hybrid" apparent in the sample depicted in Fig. 3a. "Necks" of resin tags are wider in Figs. 3a, 4a, and 4b than in Fig. 3b, even though the same 4-META/MMA-TBBresin in the same amounts were used in each instance. Fig. 5 depicts SEM views of samples prepared in the same manner as those seen in Figs. 3 and 4, except that 1% NaOC1 immersion was extended to 60 min. Note that the longer NaOC1 immersion did not alter the appearance ofthe samples shown in Figs. 5a and 5b compared to the samples seen in Fig.3. On the other hand, the specimens in Fig. 4 have been altered by the extended NaOC1 immersion, as seen in Figs. 5c and 5d. The "hybrid" seen in Fig. 4a has completely disappeared (see Fig. 5c), and the thin "hybrid" visible in Fig. 4b has been partially decomposed. A crack replaces it at the adhesive resin/dentin interface (see Fig. 5d). Transmission Electron Microscope Observations. TEM views of ultrathin cross-sections of the interface between
Fig. 5. (a=top left): Cross-sectional SEM view of a specimen of 4-META/MMATBB resin bonded to ground bovine dentin that had been pre-treated with 103 for 10 s. Specimen subjected to immersion in 6 mol/L HCI for 30 s, and 60 min of immersion in 1% NaOCl. (b=top right): Specimen received the same preparation except pre-treated with 10-0 for 10 s. Compare "necks" of resin tags between specimen. (c=bottom left): Specimen received the same treatment as sample in (b): pre-treated with 10-0 for 10 s but this was followed by 30% HEMA application for 10 min. Note that thin "hybrid" layer (H) is visible at the interface, and that "necks" of resin tags have narrowed compared with those seen in Fig. 4a; comparable with those shown in Figs. 3a and 5b. (d=bottom right): Specimen received the same treatment as sample shown in (c), but after pre-treetment with 10-0 for 10 s, the 30% HEMA was applied for 60 min. Note that the "hybrid" layer (H) has detached from cured resin, and tag "necks" appear narrower than those seen in Fig. 4b. Codes are the same as in Fig. 1
4-META/MMA-TBB resin (R) and bovine dentin (D) that has been demineralized with 10-0 solution for 10 s then treated with 30% HEMA for 60 min are shown in Fig. 6. Fig. 6a depicts a sample prepared without epoxy embedding. This was not possible when a specimen had not had HEMA applied to the 10-0 pre-treated dentin. The assumption is that without HEMA application, a true "hybrid" does not form when dentin is pre-treated with 10-0 solution, so the specimen could not maintain its integrity during the slicing procedures that created the ultra-thin section. In the latter instance, epoxy embedding is necessary for TEM specimen fabrication. However, in Fig. 6a (sample made without embedding), the dentin has been reinforced by the adhesive resin, 4-META/MMA-TBB resin. The transparent zone on the TEM is the "hybrid" layer (H) formed when demineralized dentin has been impregnated and entangled with polymerized resin. This "hybrid" unites the resin with dentin. Thickness of this "hybrid" is about one micron. Fig. 6b depicts the sample after immersion in 0.001 mol/L HC1 for 4 s. The acid immersion demineralized hydroxyapatite that has not been completely encapsulated by resin and incorporated into the "hybrid" layer ofresin-reinforced dentin. The black line apparent between the "hybrid" layer (H) and partially demineralized dentin (DD)is hydroxyapatite remainingin the acid-immersed specimen. Fig. 6c is a higher magnification (50,000x) of the sample shown in Fig. 6b. Fig. 7 depicts specimens prepared in the same manner as those shown in Fig. 6, except that the dentin was pre-treated with 10-3 solution prior to 30% HEMA application for 60 s. Note that the black line created by the much shorter (60 s) HEMA application is seen more clearly in Fig. 7b. The "hybrid" layer is about 3 ~m wide. Dental Materials~March 1992 127
7. 6. Ultra-thin, cross-sectional TEM views of a specimen of 4-META/MMA-TBB resin bonded to ground bovine dentin; (a=left): Specimen had been pre-treated th 10-0 for 10 s followed by 30% HEMA application for 60 min, Specimen was sectioned directly without epoxy-embedding for TEM examination. (b=center): me specimen after immersion in 0.001 mol/L HCI for 4 s to remove the hydroxyapatite, (if possible). Note black line of resin-encapsulated hydroxyapatite crystals it has resisted dilute acid demineralization. (c=right): Higher magnification of view in (b). Codes are the same as in Fig. 1.
DISCUSSION ;veral concepts of the bonding mechanism of adhesive sins to dentin have been proposed. One is bonding via tag cmation in the dentinal tubules ofetched dentin (Nordenvall [d BriinnstrSm, 1980). A second is the formation ofprecipites on pre-treated dentinal substrates to which an adheve resin may be chemically or mechanically bonded (Bowen al., 1982). A third is chemical union to either inorganic ~bar and Farley, 1974) and/or organic components of the Lbstrate (Munksgaard and Asmussen, 1984). A fourth ncept is proposed by these authors--that is, diffusion of onomers into the subsurfaces of pre-treated dentinal subrates and their polymerization therein, to create a "hybrid" yer of resin-reinforced dentin. The "hybrid" is an acidsistant intermixture of adhesive resin with components of ~ntin at the molecular level. This newly formed material or termixture (the "hybrid" layer) unites underlying dentin Lth cured resin, which overlies the resin-reinforced dentin [akabayashi, 1982; Wang and Nakabayashi, 1991).
I. 7. Ultra-thin, cross-sectional TEM views of a specimen of 4-META/MMA;B resin bonded to ground bovine dentin that had been pre-treated with 10or 10 s followed by 30% HEMA application for 60 s. Specimen in (a=left) was ctioned directly without epoxy-embedding for TEM examination. Note that ybrid" layer (H) is wider than that seen in Fig. 6a; (b=right): same specimen er immersion in 0.001 mol/L HCI for 4 s to remove the hydroxyapatite, (if ssible). Note black line between "hybrid" layer (H) and demineralized dentin :)). It is resin-encapsulated hydroxyapatite crystals that have resisted dilute id demineralization. Codes are the same as in Fig. 1.
B Nakabayashi & Takarada/Effect of HEMA on bonding to dentin
There appear to be two ways to promote and enhance adhesion to dentin. The first is to improve monomer impregnation into the substrate, and the second is to increase the diffusivityor penetrability ofthe dentinal substrate itself. The accomplishmentofeither or both ofthese goals would serve to enhance the formation of a "hybrid" layer created by the impregnation, penetration, and subsequent entanglement of adhesive resin and demineralized dentinal components. In the present study, the authors have found that HEMA enhances "hybrid" layer formation by increasing the penetrability and diffusivity of the dentin substrates. As can be seen by the data in the Table, HEMA application to dentin substrates that have been ground and pre-treated with either the 10-3 or 10-0 solution significantly improves the bond strength of 4-META/MMA-TBB resin to that dentin. Further, the data indicate that the improvement by HEMAis also dependent upon time-length ofthe application. The bond strength of adhesive resin to 10-0 pre-treated dentin (without HEMAapplication) was 5 MPa, but this strength was improved by HEMA application and enhanced even more as the length of time of that applicationincreased. Abond strength ofl3 MPa was attained on 10-0 pre-treated specimens that were subjected to 60 min HEMA application. The data strongly suggest that 2-hydroxyethyl methacrylate applied to ground and pretreated dentin improved its diffusivity. The same effect, enhancementofbond strength, was also attained with 10-3 pretreated dentin samples, but the improvement occurred with shorter HEMA-application times. This is due to the fact that the ferric ions in the 10-3 solution increase the ability of the monomer to impregnate demineralized dentin, and, when coupledwith HEMA's capacity for improvingdentin diffusivity, even higher strengths were achieved with shorter HEMAapplication times, attaining 18.7 MPa with a 60 s application. SEM and TEM examinations of samples clearly supported the concept of improved dentin diffusivity by HEMA application. Observations confirmed greater monomer infusion and impregnation into demineralized dentinal substrates which had higher diffusivityas a result ofapplication ofHEMAto the pre-treated surfaces. Fig. 1 suggests a relatively wide (5 pro) "hybrid" layer (H) with good resistance to decomposition by NaOC1. On the
other hand, Fig. 2 demonstrates a rather narrow zone (H'), with an appearance relatively similar to that of H, that readily decomposed in NaOC1. This has been termed the "hybrid-like" layer because ofits similar appearance on microscopic examination. It easily decomposed in NaOC1, because this narrow "hybrid-like" layer was mainly collagen with little polymerizedresin-entanglement to protect it (collagen)from decomposition by NaOC1immersion. The explanationforthis phenomenonisthat 10-0pre-treated dentin (Fig. 2b) had less diffusivity than did 10-3 pre-treated samples, even though both pre-treatments achieved the same depths of demineralization (5 ~m). However, 10-0 solution (10% citric acid/0% ferric chloride) denatured dentinal peptides in the demineralized zone (Mizunuma, 1986). [Ferric ions in the 10-3 solution (10% citric acid/3% ferric chloride) protect the peptides from denaturing.] The denatured protein molecules of 10-0 pretreated specimens shrank or collapsed during the drying process. Thus, diffusivity or penetrability of these collapsed peptides was low,even though original demineralizationdepth was the same as that achieved with the 10-3 solution. This low diffusivity made it difficult for monomers to impregnate, entangle, polymerize, and resin-reinforce the pre-treated dentin. The lack ofresin-reinforcement ofthese specimens made it impossible for ultrathin samples to be directly fashioned for TEM examination, and epoxy embedding was necessary, but it was possible for ultrathin specimens to be made without epoxy-embeddingin10-3 pre-treated specimens, because they were resin-reinforced (Takarada et al., 1990). The phenomenon can be clearly seen in Figs. 3 through 5, which demonstrate the comparable resistance of"hybrid" and "hybrid-like" layers to NaOC1 decomposition. The "necks" of resin tags continuous with the "hybrid" layer are wide in Fig. 3a. This is due to the fact that these 10-3 demineralized specimens had good diffusivity. Resin tag "necks" seen in 10-0 pre-treated specimens (Fig. 3b) are narrower and continuous with cured resin. These specimens had lower diffusivity and poor"hybrid" layer formation. When HEMA was applied to the 10-0 demineralized dentin sample, the picture changed to that depicted in Fig. 4. Although thinner than that seen in Fig. 3a, a definite "hybrid" layer is apparent with wider "necked" resin tags continuous with that "hybrid". The 10-0 demineralized dentin obviously had enhanced diffusivity after HEMA application, and monomer penetration and polymerization to create the "hybrid" was improved. The phenomenon has been studied and reported by Sugizaki (1991). When NaOClimmersion time was increased from 10 rain to 60 min, samples shown in Fig. 4 (10-0 pre-treated and HEMA) took on the appearance of those shown in Figs. 5c and 5d. However, these increased NaOC1 immersion times had little effect on the 10-3 and 10-0 pre-treated specimens shown in Fig. 3. Rather, these took on the appearance of Figs. 5a and 5b. These microscopic examinations confirm the improved diffusivityof samples receiving HEMA application and that the degree ofimprovement, as indicated by higher bond strengths, was time-dependent. "Hybrid" or "hybrid-like" layers with lower resin concentrations decomposed more readily in NaOC1 immersion. Figs. 6a and 7a are TEM views of specimens that were sliced ultrathin without epoxy-embedding,avery unusual procedure. The 4-META/MMA-TBB resin-reinforcementis akin to embedding samples in epoxy for TEM preparation (Watanabe and Nakabayashi, 1987). If monomers are not capable of impregnating demineralized dentin to polymerize therein, and reinforce the specimens, then direct preparation ofultrathin
TEM sections without epoxy-embedding is not possible. Figs. 6b and 6c dearly reveal that HEMA application improves the diffusivity and penetrability of the substrates, and that monomers have impregnated and polymerized deep within the dentin (D in Fig. 6a). Hydroxyapatite crystals within dentin are dissolved by HC1 immersion if they are not shielded from the acid attack. The black line in HCl-immersed samples (shown in Figs. 6b and 6c between H and DD) is undissolved hydroxyapatite. These hydroxyapatitecrystals resisted dissolution by HClbecause they were encapsulated by polymerized resin that had impregnated and surrounded them along with collagen. This demonstrates the acid-resistance of the "hybrid" layer. Fig. 7b depicts a 10-3 pre-treated sample that received HEMA application. A"hybrid" layer with hydroxyapatite at its bottom (toward the dentin) is visible. This black line of resinencapsulated hydroxyapatite was not seen ifa specimen was not subjected to HEMA application (Takarada et al., 1990). If one compares the width of"hybrid"(H) in Fig. 7a with that in Fig. 6a, a difference can be observed. In Fig. 7a, the"hybrid" is 3 pm, and in Fig. 6a, 1 ~m. The ferric chloride in the 10-3 pretreatment (sample in Fig. 7a) prevented denaturation of peptides and maintained their penetrability (Mizunuma, 1986). Such was not the case in the 10-0 pre-treated specimen (Fig. 6a), and good diffusivity could not be fully recovered even after prolonged, 60 rain, HEMA application. The depth of dentin demineralization following 10-3 pre-treatment for 10 s was 5 pm (see Fig. la). The same demineralization depth of 103 pre-treated dentin has been previously reported (Kato et al., 1986; Takarada et al., 1990; Wang and Nakabayashi, 1991). However, following HEMA application of 10-3 pretreated dentin, the zone of demineralization was reduced to 3 pm (see Fig. 7a). This reduction in depth of the demineralized dentinal substrate did not have an adverse effect on attained bond strengths. It is possible that bond strengths can be improved by enhanced polymerization of resin in the presence of ferric ions absorbed onto dental surfaces (Bowen, 1980; Akimoto et al., 1990), but bond strength levels may also be increased by HEMA application in the absence offerricions. There is no significant difference in bond strengths to dentin on 10-3 pre-treated samples (13.3 MPa) or 10-0 pre-treated samples that had HEMA applied for 60 min (13.3 MPa) (p > 0.05). It should be emphasized here that impregnation of monomers into demineralized dentin and their polymerization therein to form a "hybrid" layer are essential ingredients in attaining high bond strengths. Further evidence of the effects of ferric ions in improving dentin diffusivity and increasing bond strength is demonstrated by the influence of HEMA application to 10-3 pretreated specimens. Higher bond strengths were achieved in these samples with much shorter HEMA-application times than were possible with 10-0 pre-treated specimens. Therefore, the rate of monomer diffusion into HEMA-treated demineralized dentin samples was faster in the former (10-3) than the latter (10-0). HEMA and monomer diffused into dentin substrates with ferric ions absorbed onto their surfaces (10-3 pre-treated) at a much more rapid rate than into non-ferric-ionabsorbed dentinal substrates (10-0 pre-treated) to attain high bond strength.
CONCLUSIONS In order for high bend strengths to be attained, it was essential for dentinal substrates to have good diffusivity and for monomers to Dental Materials~March 1992 129
have good penetrability into the substrate. Impregnated monomers must be well-polymerized in situ. If these conditions are achieved, collagen in demineralized dentin could entangle with impregnating co-polymer chains. HEMA was found to be effective in improving the diffusivity of demineralized dentin whose peptides were denatured, causing the protein molecules to be less penetrable and demonstrated on microscopic views as narrowed zones of demineralization, during smear layer removal by acidic pre-treatment with 10-0 solution (10% citric acid/0% ferric chloride). HEMA was also shown to enhance the rate of monomer diffusion into ground dentinal substrates that had smear layers removed with 10-3 solution (10% citric acid/3% ferric acid). Higher bond strengths were achieved with HEMA application in specimens pre-treated with both solutions. Increased bond strength was also found to be dependent upon the length of time of HEMA application. Finally, monomer penetration and encapsulation of hydroxyapatite crystals rendering the crystals acid-resistant, coupled with impregnation and envelopment by resin of collagenous componentsofdemineralized dentinal substrates was enhanced by HEMA application.
ACKNOWLEDGMENTS This investigation was supported in part by a Grant-in-Aid for Scientific Research (No. 02205034) from The Ministry of Education, Science and Culture, Japan. Received March 18, 1991/Accepted July 12, 1991 Address correspondenceand reprint requests to: N. Nakabayashi Divisionof Organic Materials Institute for Medical and Dental Engineering Tokyo Medical and Dental University Kanda, Tokyo, 101, Japan
REFERENCES Akimoto T, Kadoma Y, Imai Y (1990). Studies on adhesion mechanism of MMA-TBBO resin to dentin. A model experiment using decalcified dentin. J Jpn Dent Mater 9:320-325. Anbar M, Farley EP (1974). Potential use of organic polyphosphonateas adhesives in the restoration ofteeth. J Dent Res 53:879-888. Bowen RL (1980). Adhesive bonding of various materials to hard tooth tissues. XXII. The effects ofa cleanser, mordant, and PolySAC on adhesion between a composite resin and dentin. J Dent Res 59:809-814. Bowen RL, Cobb EN, Rapson JE (1982). Adhesive bonding of various materials to hard tooth tissues: improvement in bond strength to dentin. JDent Res 61:1070-1076. Inokoshi S, Hosoda H, Harnirattisai C, Shimada Y, Tatsumi T (1990). Astudyontheresinimpregnatedlayerofdentin. I. A comparative study on the decalcified and undecalcified sections and the application of argon ion beam etching to disclose the resin impregnated layer of dentin. Jpn J Conserv Dent 33:427-442. Itoh K, Hashimoto K, Wakumoto S (1985). Effect of GLUMA concentration on adhesion to dentin. Jpn JConserv Dent 28:895-901. Kato H, Wakumoto S, Suzuki M (1986). Chemical analysis of interface layer between resin and dentine by Raman microprobe. J Jpn Dent Mater 5:232-237. Maeda M, Mogi M, Miura F, Nakabayashi N (1987). Study 130 Nakabayashi & Takarada/Effect of HEMA on bonding to dentin
on the direct bonding method for porcelain and ceramic brackets. J Jpn Orthod Soc 46:370-379. Matsumura H, Nakabayashi N (1988). Adhesive 4-META/ MMA-TBB opaque resin with poly(methyl methacrylate)coated titanium dioxide. JDentRes 67:29-32. Mizunuma T (1986). Relationship between bond strength of resin to dentin and structural change of dentin collagen during etching. Influence of ferric chloride to structure of the collagen. JJpn DentMater 5:54-64. Munksgaard E C, Asmussen E (1984). Bond strength between dentin and restorative resins mediated by mixtures ofHEMA and glutaraldehyde. JDentRes63:1087-1089. Nakabayashi N (1982). Resin reinforced dentine due to infiltration ofmonomers into the dentine at the adhesive interface. J Jpn Dent Mater 1:78-81. Nakabayashi N (1984). Biocompatibility and promotion ofadhesion to tooth substrates. CRCCriticRevBiocompatibil 1:25-52. Nakabayashi N, Kanda K (1988). Synthesis of phosphoric compounds and their adhesives to bovine teeth. Kobunshi Ronbunshu 45:91-96. Nakabayashi N, Kojima K, Masuhara E (1982). The promotion of adhesion by the infiltration of monomers into tooth substrates. JBiomedMaterRes 16:265-273. Nakabayashi N, Masuhara E, Mochida E, Ohmori I (1978). Development of adhesive pit and fissure sealants using a MMA resin initiated by a tri-n-butyl borane derivative. J Biomed Mater Res 12:149-165. NikaidoT(1989). Formation ofphotocurablebondinghner and adhesion to dentin. Effect ofphotoinitiator, monomer and photoirradiation. J Jpn Dent Mater 8:862-876. Nordenvall K-J, Br~innstrSm M (1980). In vivo resin impregnation ofdentinal tubules. JProsthetDent 44:630-637. Sugizaki J (1991). The effect of the various primers on the dentin adhesion of resin composites. SEM and TEM observation ofthe resin impregnated layer and adhesion promoting effect of the primers. Jpn J Conserv Dent 34:228-265. Suzuki K, Takahashi M, Nakai H (1990). Treatment ofdentin by aqueous solution ofamino acid derivative-HEMA. Adhesive Dent 8:43-51. Takarada K, Kojima M, Ishihara K, Nakabayashi N (1990). Durability ofbondingbetween 4-META/MMA-TBBresin to dentin pretreated with 10-3. The effect of10-3 pretreating period and subsequent glutaraldehyde treatment. J J p n Dent Mater 9:831-840. Takeyama M, Kashibuchi N, Nakabayashi N, Masuhara E (1978). Studieson dental self-curing resins (17). Adhesion of PMMA with bovine enamel or dental alloys. JJpn Dent Appar Mater 19:179-185. Tanaka T, Hirano M, Kawahara M, Matsumura H, Atsuta M (1988). A new ion-coating surface treatment of alloys for dental adhesive resins. JDentRes 67:1376-1380. WangT, NakabayashiN(1991). Effect of2-(methacryloxy)ethyl phenyl hydrogenphosphate on adhesion to dentin. JDent Res 70:59-66. Wang T, Nikaido T, Nakabayashi N (1991). Photocure bonding agent containing phosphoric methacrylate. DentMater 7:59-62. Watanabe A, Nakabayashi N (1987). TEM observation of demineralized dentin surface for adhesion. Comparison of etching agents. Rep Inst Med Dent Eng 21:17-25. Yamauchi J (1986). Study of dental adhesive resin containing phosphoric acid methacrylate monomer. J Jpn Dent Mater 5:144-154.