Dental Materials (2005) 21, 232241

www.intl.elsevierhealth.com/journals/dema

Mechanical stability of resin dentin bond components

Marcela Rocha de Oliveira Carrilhoa, Franklin R. Tayb, David H. Pashleyc, derhaned,e, Ricardo Marins Carvalhof,* Leo Tja
Department of Restorative Dentistry, Dental Materials, Piracicaba School of Dentistry, University of Campinas, Piracicaba, SP, Brazil b Department of Conservative Dentistry, School of Dentistry, Prince Philip Hospital, University of Hong Kong, Hong Kong, China c Department of Oral Biology and Maxillofacial Pathology, School of Dentistry, Medical College of Georgia, Augusta, GA, USA d Institute of Dentistry, University of Helsinki, Helsinki, Finland e Department of Oral and Maxillofacial Diseases, Helsinki University Central Hospital, Helsinki, Finland f Department of Operative Dentistry, Endodontics and Dental Materials, Bauru School of Dentistry, o Paulo, Bauru, SP, Brazil University of Sa
KEYWORDS
Resin composite; Adhesives; Collagen brils; Dentin; Mechanical properties; Durability
a

Summary Objectives: To evaluate the effects of long-term storage on the mechanical properties of the components of resin dentin bonds, that is, resin composite, adhesive system, demineralized and mineralized dentin. Methods: Specimens of resin composite (Z250) and adhesive systems (Single BondSB; One-Step-OS and Clearl Liner Bond 2V-CL) were cast in molds. Dentin specimens were prepared from dentin discs obtained from the crowns of extracted human molars. Specimens of demineralized dentin were obtained by immersion of dentin discs for 6 days in 0.5 mol/l EDTA (pH 7.0). Both dentin and resin-based substrates were shaped to hourglass or I-beam specimens that were used to determine the true stress (TS) or apparent modulus of elasticity E ; respectively. Control specimens were subjected to tensile testing at 0.6 mm/min after 24 h of immersion in distilled water. Experimental specimens were stored at 37 8C in either distilled water or mineral oil and tested after 12 months. The data of each group were individually analyzed by ANOVA and Tukeys test. Results: Both TS and E of the resin-based materials decreased signicantly after 12 months of storage in water p , 0:05; except the TS of SB p . 0:05: No changes were observed for specimens of mineralized dentin, regardless of storage condition p . 0:05: Storage of demineralized dentin in water did not cause any signicant effect in either TS or E p . 0:05; however, signicant reductions of TS and E of demineralized dentin occurred after storage in oil for 1 year p , 0:05: Signicance: Storage time and medium may be deleterious to the mechanical properties of the resin dentin bond components, which ultimately could compromise the durability of resin dentin bonds. Q 2004 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

*Corresponding author. Address: Al. Ota vio Pinheiro Brisolla 9-75, Bauru, SP, Faculdade de Odontologia de Bauru, Universidade de Sao Paulo, Dent stica, Bauru, CEP 17012-901, Brazil. Tel.: 55-14-235-8321; fax: 55-14-224-1388. E-mail address: ricfob@fob.usp.br 0109-5641/$ - see front matter Q 2004 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dental.2004.06.001

Durability of resin dentin bond components

233

Introduction
Recent studies revealed that the resin dentin bonds created by current hydrophilic adhesive systems can severely degrade over time [1 3]. Although several factors may account for such deterioration, there has always been a concern regarding how well resin monomers are able to inltrate into demineralized dentin, and how this monomer inltration might create a stable hybrid layer. The reasons why incomplete inltration can occur are numerous, but the end result is a porous hybrid layer [4] highly susceptible to degradation by oral uids [1,5,6]. Despite our inability to diagnose the early signs of impending bond degradation [7], contemporary methods in dental materials investigation have been useful to characterize several mechanisms responsible for detecting either mechanical or morphological degradation of resin dentin bonds [3,8,9]. Since the mechanical properties of each resin dentin bond component are expected to play signicant roles in the ultimate bond strength values [10 14], the durability of resin dentin bonds depends upon the stability of their components over time. Morphological in vitro studies indicated that both resin and collagen matrices may degrade upon storage [6,9]. Catastrophic failure of resin dentin bonds may initiate in one specic component of the interface. The identication of which component is more likely to be responsible for the overall reduction of the bond strength is impossible when the components are evaluated in the complex bonded condition. By testing them separately, the authors expect to be able to determine which is least stable during various storage conditions. Thus, the objective of this study was to investigate the effects of prolonged aging in distilled water or mineral oil on the true stress (TS) and apparent modulus of elasticity E of resin dentin bond substrates (i.e. resin composite, adhesive system, demineralized and mineralized dentin). Since the materials/substrates tested possess viscoelastic behavior and may undergo deformation upon loading, the ultimate cohesive strength at failure is more properly named true stress rather than ultimate tensile strength. Similarly, apparent E was adopted because displacement during testing was obtained from the crosshead of the machine rather than from extensometers. The null hypothesis tested is that there will be no signicant changes in the mechanical properties of these components, regardless of storage condition.

Material and methods

The institutional review board of the University of Sa o Paulo approved this study under protocol number 147/00. Fifty-three non-carious human third molars were stored in water containing 0.2% sodium azide at 4 8C and used within 6 months of extraction. Resin composite and adhesive systems that were used are presented in Table 1. The procedures used to prepare each substrate are described below.

Specimens preparation
Resin composite Polyvinylsiloxane molds with hourglass or I-beam shaped impressions were prepared (Fig. 1). Resin composite Z250 (3M ESPE, St Paul, MN, USA) approximately 1 mm thick was inserted into each mold, lling it completely. A glass cover slip was placed on the resin composite and the surface was light-cured with three individual 40 s exposures using a light source of 600 mW/cm2, covering the entire surface of the resin composite. After polymerization, cured hourglass or I-beam shaped specimens were removed from the molds and handnished with 600-grit SiC paper. Thirty hourglass and I-beam shaped specimens were prepared and used to determine the true stress (TS) and apparent modulus of elasticity E ; respectively (Fig. 2). Both specimens had a cross-sectional area of 0.7 mm2 (0.1), mean (standard deviation).
Table 1 Resinous materials tested. Material Z250 Single Bond (SB) Composition UDMA, Bis-EMA, TEGDMA, inorganic ller Adhesive: HEMA, Bis-GMA, polyalkenoic acid copolymer, dimethacrylates, ethanol, water and CQ Adhesive: BPDM, Bis-GMA, HEMA, acetone and CQ Primer A: MDP, Hydrophilic dimethacrylates, CQ. Primer B: HEMA, N,N-diethanol toluidine, water Adhesive: MDP, Bis-GMA, HEMA, hydrophobic dimethacrylates, N,N-diethanol toluidine,CQ and colloidal silica

One Step (OS) Clearl Liner Bond 2V(CL)

Bis-GMA: bisphenol A-glycidyl methacrylate; Bis-EMA: bisphenol ethoxy methacrylate; BPDM: biphenyl dimethacrylate; CQ: dil-camphorquinone; HEMA: 2-hydroxylethyl methacrylate; MDP: 10-methacryloyloxydecyl dihydrogen phosphate, TEGDMA: triethylene-glicol-dimethacrylate, UDMA: urethane dimethacrylate.

234

M.R.O. Carrilho et al.

Figure 3 Schematic of how the adhesive systems were directly dispensed into the mold to produce the cured slabs for posterior trimming into hour-glass and I beam shapes. Figure 1 Polyvynil siloxane molds used to prepare resin composite specimens for TS (a) and apparent modulus of elasticity (b). Z 250 resin composite was directly inserted into the molds and light-cured.

Adhesive systems A pilot study showed that adhesive specimens prepared from polyvinylsiloxane molds underwent severe deformation within a few minutes after

Figure 2 Representative illustration of the hour-glass shaped specimens used for testing TS and the I beam shaped specimens used for the E testing.

polymerization. To overcome that problem, acrylic rods were glued to a glass slide to create a 15 15 0.8 mm mold to prepare slabs of the adhesives (Fig. 3), as previously described [15]. The adhesive systems Single Bond (SB-3M ESPE, St Paul, MN, USA) and One-Step (OS-Bisco, Schaumburg, IL, USA) were directly dispensed to ll the mold completely, left undisturbed for 60 s, and gently blown with oil- and water-free compressed air for an additional 60 s to allow proper solvent evaporation. Preliminary testing showed that no further weight loss of the adhesives occurred with prolonged time allowed for evaporation (not shown). A glass cover slip was placed on the adhesive and the surface was light-cured with four individual 80 s exposures to a light source of 600 mW/cm, covering the entire surface of adhesive in the matrix. Equal amounts of primers A and B (seven drops each) of Clearl Liner Bond 2V (CL-Kuraray Med. Inc., Osaka, Japan) were mixed in a well and immediately dispensed into the mold. The mixture was left undisturbed for 60 s and gently air-blown for an additional 60 s. Then, seven drops of the Bond liquid A was added and mixed with the primers. A glass cover slip was placed on the surface followed by the same light-curing protocol described above.

Durability of resin dentin bond components

235

After polymerization, cured slabs measuring 15 15 0.8 mm were obtained from the three adhesives. Each one was sectioned into two halves and each half was trimmed with a ne diamond bur to either an hourglass or an I-beam shape. At least 33 slabs of each adhesive were prepared from the matrix, which resulted in 32 34 specimens for both TS and E : The cross-sectional area of the specimens for both tests was 0.8 mm2 (^ 0.05). Mineralized dentin Thirty dentin discs (0.8 mm thick) were obtained from the mid-crown of 30 extracted human third molars by means of a diamond saw (Labcut 1010, Extec Inc., Eneld, CT, USA). The teeth were stored in saline with 0.2% sodium azide and were used within 3 months after extraction. One hourglass and one I-beam shaped specimens were cut from each dentin disc (Fig. 4) with a ne diamond bur operated in a high-speed handpiece with copious air-water spray irrigation. This resulted in 30 specimens prepared for both TS and E : The crosssectional area of the specimens for both tests was 0.65 mm2 (^ 0.01). Demineralized dentin Twenty-three teeth were used to obtain discs of mid-coronal dentin. Two hourglass shaped specimens were cut from 15 teeth in a similar way as described above n 30: Both ends of these specimens were coated with two layers of nail varnish, leaving a central area of 1.5 mm exposed for demineralization. Three to four dentin sticks, measuring approximately, 0.8 0.8 8.0 mm were obtained from the remaining eight discs, by simply transversally slicing the discs as described previously [16]. Both ends of the dentin sticks were acid-etched (35% phosphoric acid, 3M ESPE), rinsed,

Figure 5 Preparation of the dentin beam specimens for demineralization in EDTA and posterior testing for calculating E : The dentin beams were placed in the center of the mold (a) and the ends covered with resin composite (b).

blot-dried and coated with two layers of Single Bond adhesive system. Each stick was placed into a polyvinylsiloxane mold with an I-beam shaped impression and the ends were covered with Z250 resin composite, leaving an intermediate area of 4.0 mm in length (gauge length) exposed for demineralization (Fig. 5). Thirty specimens were prepared. Both the TS (hourglass) and E (I-beam) specimens were demineralized by immersion in 0.5 mol/l EDTA (pH 7.0) for 6 days at 25 8C [17] followed by extensive rinsing in water (ca. 24 h). The cross-sectional area of the specimens for TS and E tests was 0.65 mm2 (^ 0.03).

Storage conditions
Specimens of each substrate/material were divided into three groups and randomly assigned to be stored according to the following conditions: 24 h in distilled water (control) and 1 year either in distilled water (1yr W) or in mineral mineral oil

Figure 4 Schematic of how the dentin disc was obtained from the molar teeth and the specimens trimmed out of the discs.

236

M.R.O. Carrilho et al.

(1yr O), without the addition of preservatives or antimicrobial agents. Storage in mineral oil was selected to avoid the effects of water storage. All specimens were stored in hermetically sealed vials at 37 8C.

comparisons were performed using Tukeys test. Statistical signicance was preset at a 0:05:

Results
Results are summarized in Tables 2 4. Storage in water for 1 year did not cause any signicant effect in either TS or E of mineralized and demineralized dentin specimens p . 0:05: Conversely, storage in water for 1 year caused signicant reductions in both TS and E of all resin-based substrates (resin composite and adhesives) p , 0:05; except for TS of CL adhesive that did not signicantly change over time p . 0:05: Reductions in the mechanical properties of the self-etch adhesive after 1 year of storage in water were lower (CL, 29 and 14%, respectively, for E and TS, Table 4). The mechanical properties of both total-etch adhesive resins (SB and OS) were signicantly higher at 24 h than the self-etch system (CL) p , 0:05: The same trend was observed after 1 year of water storage, except that TS of OS was not signicantly different from that of CL p . 0:05: Storage in mineral oil for 1 year did not signicantly alter the TS and E of resin composite (Z250) and mineralized dentin substrates, but increased the TS of all adhesivesubstrates (SB, OS and CL) and the E of OS and CL signicantly after 1 year of storage in mineral oil p , 0:05: Storage in oil for 1 year caused
Table 2 Changes in the true stress (TS) of each substrate as a function of storage condition. Substrate Z250 Storage condition TS (SD) n 79.7 (23.5) [10]* 55.3 (14.2) [10] 70.4 (22.7) [10]* 22.3 (3.3) [13] c,d 12.6 (1.9) [10] e 44.3 (6.6) [10] a 18.6 (4.4) [12] d 6.8 (1.8) [10] f 34.6 (8.9) [10] b 8.4 (1.5) [13]*f 7.2 (1.2) [10]*f 25.4 (5.0) [10] c 82.7 (26.2) [10]* 82.7 (15.1) [10]* 81.3 (16.8) [10]* 13.3 (3.6) [10]* 12.0 (3.7) [10]* 2.4 (0.9) [10]

Control specimens testing

The TS and E of each control group were determined after the specimens had been stored for 24 h in distilled water. The specimens were individually attached to the grips of a Vitrodyne V-1000 (Chatillon Bros., Greensboro, NC, USA) testing machine and tested in tension at 0.6 mm/min. Specimens for TS (hourglass) were glued to the testing jig with cyanoacrylate cement (Zapit, Dental Ventures of America, Corona, CA, USA), while E (I-beam) specimens were clamped with special grips that permitted the maintenance of their 4 mm gauge length. The EDTA-demineralized dentin specimens were tested while immersed in distilled water. The load-displacement data were collected in a computer connected to the testing machine. After failure, the specimens were removed from the grips and their cross-sectional area measured with a digital caliper to the nearest 0.01 mm. The cross-sectional area was used to calculate the TS as a function of the maximum load at fracture, and the load displacement data were converted to stress strain curves. From these curves, the moduli were calculated at the steepest part of the curve. All values were expressed in MPa. Because displacement was obtained from movement of the cross-head instead of an extensometer, the moduli of elasticity have been designated as apparent moduli of elasticity.

Experimental specimens testing

The remaining specimens stored for 1 year were removed from the vials, washed in water for 1 h and tested under the same protocol described above for control specimens.

Statistical analysis
The statistical analysis was not designed to compare different substrates. Thus, the data from resin composite, mineralized dentin and demineralized dentin substrates were individually analyzed by one-way ANOVA, with the storage condition (24 h, 1yr W and 1yr O) as the main factor. Data from the adhesive systems were analyzed by two-way ANOVA, with storage condition (24 h, 1yr W and 1yr O) and adhesive systems (SB, OS and CL) as the two factors in the analysis. All post hoc multiple

Control (24 h) 1 yr W 1 yr O SB Control (24 h) 1 yr W 1 yr O OS Control (24 h) 1 yr W 1 yr O CL Control (24 h) 1 yr W 1 yr O Mineralized dentin Control (24 h) 1 yr W 1 yr O Demineralized dentin Control (24 h) 1 yr W 1 yr O

Values are in MPa (SD) n: (*) Indicates no statistical difference among values within each substrate group p . 0:05: Same letter indicates no statistical difference among adhesive system substrates p . 0:05:

Durability of resin dentin bond components

237

Table 3 Changes in the apparent modulus of elasticity E of each resin dentin component as a function of storage condition. Substrate Z250 Storage condition E (SD) n 6532 (2136) [10]* 4662 (1067) [10] 7701 (2199) [10]* 614 (152) [12]*a,b 398 (47) [10] c 579 (39) [10]*b 423 (47) [10] c 227 (44) [10] d 532 (84) [10] b 139 (43) [12] e 98 (20) [10] f 698 (93) [10] a 7902 (2503) [10]* 9046 (1728) [10]* 8864 (1987) [10]* 60 (17) [10]* 69 (36) [10]* 31 (7.2) [10]

Control (24 h) 1 yr W 1 yr O SB Control (24 h) 1 yr W 1 yr O OS Control (24 h) 1 yr W 1 yr O CL Control (24 h) 1 yr W 1 yr O Mineralized dentin Control (24 h) 1 yr W 1 yr O Demineralized dentin Control (24 h) 1 yr W 1 yr O

Values are in MPa (SD) n: (*) Indicates no statistical difference among values within each substrate group p . 0:05: Same letter indicates no statistical difference among adhesive system substrates p . 0:05:

signicant reductions in both TS and E of demineralized dentin specimens p , 0:05:

Discussion
The results of this study showed that the mechanical properties of some of the components of the resin dentin bonds were signicantly affected by storage in water or oil for 1 year. The null hypothesis tested must, therefore, be partially rejected. The reduction in the mechanical properties of the adhesive systems aged in distilled water for 1 year extends and conrms our previous observations that demonstrated signicant decreases in both TS and E of the same adhesives stored in distilled water for 3 and 6 months [15]. Cumulative reductions in TS for SB and OS after 1-year storage in water reached values of 43 and 63%
Table 4 Reduction (%) of the mechanical properties of resin-based substrates stored in water for 1 year, as a function of control condition (24 h). Substrate Z250 SB OS CL TS (%) 30 43 63 14 E (%) 28 35 46 29

of the 24 h controls, while reductions in E were in the range of 35 and 46%, respectively (Table 4). The mechanical properties (ca. TS and E ) of CL did not change signicantly when the specimens were stored in water for 3 6 months [15]. However, the present data demonstrate that while no signicant changes were observed for TS, a signicant decrease in the E of CL was observed after 1-year storage. The mechanical properties of CL were always signicantly lower than the other two adhesive systems tested. The mechanical properties of CL at 24 h were particularly low, and that may have accounted for the lack of differences between the control (24 h) and 3 and 6 months testing periods [15]. However, the prolonged immersion time in water may have allowed for further deterioration of the polymer structure, thereby resulting in further reductions in its properties. It is speculated that between 24 h and 6 months, plasticization of the polymer due to immediate water sorption that occurred within the initial 24 h of water immersion, was mainly responsible for the unaltered properties up to 6 months of storage. However, during the following 6 months of water storage (up to 1 year), hydrolytic degradation may have taken place, signicantly compromising the structure of the polymer. Regardless of the different behavior among the adhesives, the present data conrm the cumulative, deleterious effects of water to adhesive polymers [8]. Likewise, both mechanical properties of resin composite specimens (Z250) were also signicantly reduced after 1 year of storage in water p , 0:05: Such changes represented a decrease of 28% in E and 30% in TS (Table 4). Previous studies have shown that reduction in mechanical properties of resin composites aged in water may occur within 2 6 months [18 20]. Although the authors have not determined the mechanical properties of the resin composite in those periods, the observed reduction of its properties after 1 year is probably a result of a continuous action of water on the structure of the material. The mechanism of water transport and its effects on the mechanical properties of polymers are dependent on several factors [21,22]. Composition and monomer ratio varies according to the specic applications and manufacturers goals [23], and such variability will dene the chemical stability/degradability of resins in a specic environment [8]. It was demonstrated that increasing the ratio of TEGDMA to Bis-GMA in a polymethacrylate blend caused an increase in water uptake [24,25]. Similarly, UDMAbased resin was reported to be highly susceptible to water sorption [26]. Perhaps, the presence of TEGDMA and UDMA in Z250 resin composite contributed to the acceleration of water sorption

238

M.R.O. Carrilho et al.

and favored balance between water sorption and leaching of resin components. However, during a prolonged soaking this balance can be often upset [21] or cannot actually be achieved for years [27]. Besides the hydrophilic nature of different monomers constituting the resin matrix [28], sensitivity of resin-based materials to water effects also depends on the degree of polymer cross-linking [21], degree of monomer conversion [29], presence of llers, volume fraction of intrinsic nanometersized pores [21] and additional environmental aspects, such as the presence of enzymes and changes in pH [8,30,31]. Such factors can also inuence the rate of water sorption [31 35], and, therefore, can regulate the water effect. Accordingly, in a time-dependent manner, water molecules diffuse into the resin and while some of these molecules occupy polymer free volume, others disrupt interchain hydrogen bonds, thus altering the mechanical properties of the materials [27]. In contrast, the maintenance or slight increase in E and TS of all adhesives and resin composite specimens aged in mineral oil support the degrading effects of water storage. The mechanical properties of hydrophilic polymers may be affected by immersion in water immediately after polymerization due to a plasticization effect [36]. Conversely, if polymers are cured and not exposed to water, their mechanical properties do not decrease and even increase over time [15,36]. The results obtained with resin-based specimens stored in oil (Table 4) are probably due to the absence of water in the medium. The higher values observed after 1 year of immersion in oil may not be a result of improved mechanical properties in this medium, rather, the control values are lower because the specimens may have suffered the immediate effects of water immersion after curing. However, we cannot rule out the possibility that long-term immersion in oil could have dehydrated the specimens. This would also explain the higher values observed after long-term immersion in oil [15]. Water storage did not signicantly alter the mechanical properties of EDTA-demineralized dentin specimens aged for 1 year. These results are in agreement with another study, wherein no signicant changes in either TS or E of EDTA-demineralized dentin beams were observed after 48 months of storage in buffered saline [17]. Conversely, analysis of fractured interfaces of long-term resin dentin bond strength studies have provided evidences that collagen brils may degrade over time [3,37,38]. Phosphoric acid-exposed collagen brils unprotected by resin have shown to be morphologically altered after 500 days of water storage [9]. Caries-related literature has suggested

that host-derived proteolytic enzymes that are present in saliva and released from demineralized dentin extracellular matrix can play an important role in collagen degradation in caries lesions [39,40]. These endopeptidases, called matrix metalloproteinases (MMPs), are able to degrade most extracellular matrix proteins, including different collagens in native and denatured forms [41]. The specimens in this study were completely demineralized in EDTA and thoroughly rinsed with water before being tested and stored. Even though EDTA demineralization removes most dentin MMPs, matrix bound MMPs remain in the demineralized dentin [42]. However, mineral cations, especially Zn2 and Ca2 are required to activate MMPs. It is likely that the MMPs remaining bound to collagen were not activated due to the absence of metal ions in the distilled water. It was noticed that EDTA-demineralized specimens exhibited signicant reductions in both TS and E after 1 year of storage in mineral oil. Upon visual inspection, the specimens were seen to be twisted, shrunken and clearly dehydrated (ca. stiff) after 1 year immersed in oil. When they were re-hydrated by re-immersion in water just before being tested, they immediately became less twisted and softer than specimens that were stored in water for the same period. The softness and weakness of the specimens were then conrmed by the dramatic loss of their mechanical properties. Any residual calcium ions that were slowly released from the mineralized ends of the specimens bonded with resin composite may have been diluted by the large volume of water to levels that were so low that the MMPs remained inactive. Conversely, we may speculate that when the specimens were incubated in oil, the water that remained in the matrix was conned in situ by the oil. Any calcium released from the mineralized ends of the specimens remained in this residual water and could have activated the MMPs, allowing them to degrade the matrix over time. However, we believe that is not the case. The dehydrated appearance of the specimens when removed from the oil indicated that remaining water diffused out from the matrix, thus preventing the diffusion of metal ions from the demineralized ends throughout the specimen. Additionally, the specimens in this condition invariably ruptured in the middle of the 4 mm exposed area. If the above rationale was correct, failures were more likely to occur near the mineralized ends of the specimens. While the possibility that the oil itself may have somehow damaged the collagen matrix over 1 year storage cannot be ruled out, to the extent of the authors knowledge, there is no information available on the effects of long-term

Durability of resin dentin bond components

239

dehydration on the mechanical properties of EDTAdemineralized dentin matrix. Both aspects warrant further investigation. No changes in TS or E were observed for mineralized dentin specimens aged either in water or mineral oil during 1 year. It has been reported that the ultimate tensile strength (UTS, more properly named true stress (TS, in this study) is in the range of 70 100 MPa, depending on the tubule orientation, specimen shape and size [10,43 45]. Values of TS for mineralized dentin in this study t well within those values for specimens of similar shape and size. Variations in the UTS or TS values of mineralized dentin specimens are likely to be regulated by aw distribution in the specimen [46]. As there are no reports indicating that mineralized dentin mechanically degrades over time, therefore, our expectations of no changes after storage were conrmed. Conversely, the lack of changes in the E deserves further discussion. Recent reports indicate that the modulus of elasticity of mineralized dentin may signicantly decrease after short-term (ca. 5 15 days) storage in water [46,47]. Those studies used a modied atomic force microscope indentation technique that is capable of determining the mechanical properties of the near-surface layer of the specimens. In that case, storage in water caused partial demineralization of the surface layer, and that was detected by the surface sensitive method. It is likely that specimens stored in water in this study also suffered the surface demineralization effect of water, however, that was probably insignicant enough to cause an overall change in the bulk specimen. The authors observed values of E of mineralized dentin are below the claimed true value of modulus of elasticity for mineralized dentin (ca. 19 29 GPa) [46]. Smaller values of modulus of elasticity may be determined by a strain-rate-dependent viscoelastic response, nonuniform stress distribution that occurs in small specimens or aws introduced during specimen preparation [46]. The specimens in the current study were small in size and a low strain rate was used. Moreover, since an extensometer was not used, displacement was directly measured from the cross-head movement of the testing machine, therefore, these values should be regarded as apparent modulus of elasticity. Although the absolute values in the present study differ from literature values, the maintenance of E values over time in oil are remarkable. It would be desirable to apply pretreatments or use modied materials that would maintain

the stability of resin dentin bond components in vivo. This would prevent replacements of restorations due to degradation over time. Current bonding techniques and materials have not yet proved to be stable over time in clinical service. Contemporary resin components (adhesives and resin composite) are intrinsically susceptible to water sorption. Ideally, bonding procedures should be done in the absence of water, employing hydrophobic resins that are expected to be more stable over the long-term. At the same time, these bonding procedures should be capable of inactivating matrix derived proteolytic enzymes. Low concentration of chlorhexidine (0.03%), a potent antimicrobial agent, has been shown to hinder the activity of several MMPs [48], but its effects on resin dentin bonds formation and durability needs to be investigated. For instance, Tetracycline hydrochloride and EDTA, both used in root surface conditioning in periodontal surgery [49], and both known to inhibit MMPs [50] might provide means to control the activity of dentin-bound MMPs if used as conditioning agents in resin dentin bonds. Future studies should focus on long-term experiments that can validate whether the durability of resin dentin bonds can be improved by using either hydrophobic monomers and/or protease-inhibitors.

Acknowledgements
The authors wish to thank Mr Eduardo Rocha Pansica for illustrations. This study was supported, in part, by grants: 99/10043-0, 02/06682-1 and 01/07250-5 from FAPESP, 300481/95-0 and 474226/03-4 from CNPq, DE014911 from NIDCR, and 104337 from the Academy of Finland.

References
[1] Sano H, Yoshikawa T, Pereira PN, Kanemura N, Morigami M, Tagami J, Pashley DH. Long-term durability of dentin bonds made with a self-etching primer, in vivo. J Dent Res 1999; 78:90611. [2] Armstrong SR, Keller JC, Boyer DB. The inuence of water storage and C-factor on the resindentin composite microtensile bond strength and debond pathway utilizing a lled and unlled adhesive resin. Dent Mater 2001;17: 26876. [3] De Munck J, Van Meerbeek B, Yoshida Y, Inoue S, Vargas M, Suzuki K, Lambrechts P, Vanherle G. Four-year water degradation of total-etch adhesives bonded to dentin. J Dent Res 2003;82:13640. [4] Wang Y, Spencer P. Hybridization efciency of the adhesive/dentin interface with wet bonding. J Dent Res 2003;82:1415.

240

M.R.O. Carrilho et al.

[5] Hashimoto M, Ohno H, Kaga M, Endo K, Sano H, Oguchi H. In vivo degradation of resindentin bonds in humans over 1 to 3 years. J Dent Res 2000;79:138591. [6] Tay FR, Pashley DH, Yoshiyama M. Two modes of nanoleakage expression in single-step adhesives. J Dent Res 2002;81: 4726. [7] Tay FR, Pashley DH. Dental adhesives of the future. J Adhes Dent 2002;4:91103. [8] Santerre JP, Shajii L, Leung BW. Relation of dental composite formulations to their degradation and release of hydrolyzed polymeric-resin-derived products. Crit Rev Oral Biol Med 2001;12:13651. [9] Hashimoto M, Tay FR, Ohno H, Sano H, Kaga M, Yiu C, Kubota M, Oguchi H. SEM and TEM analysis of water degradation of human dentinal collagen. J Biomed Mater Res Part B: Appl Biomater 2003;66B:28798. [10] Sano H, Ciucchi B, Mattews WG, Pashley DH. Tensile properties of mineralized and demineralized human and bovine dentin. J Dent Res 1994;73:120511. [11] Sano H, Takatsu T, Ciucchi B, Russell CM, Pashley DH. Tensile properties of resin-inltrated demineralized human dentin. J Dent Res 1995;74:1093102. [12] Pashley DH, Ciucchi B, Sano H, Yoshiyama M, Carvalho RM, Russell CM. Bond strength versus dentine structure: modeling approach. Arch Oral Biol 1995;40:110918. [13] Hasegawa T, Itoh K, Koike T, Yukitani W, Hisamitsu H, Wakumoto S, Fujishima A. Effect of mechanical properties of resin composite on the efcacy of the dentin bonding systems. Oper Dent 1999;24:32330. [14] Takahashi A, Sato Y, Uno S, Pereira PN, Sano H. Effects of mechanical properties of adhesives on bond strength to dentin. Dent Mater 2002;18:2638. [15] Carrilho MRO, Carvalho RM, Tay FR, Pashley DH. Effects of storage media on mechanical properties of adhesive systems. Am J Dent 2004;17:1048. [16] Carvalho RM, Yoshiyama M, Pashley EL, Pashley DH. In vitro study of the dimensional changes in dentine after demineralization. Arch Oral Biol 1996;41:36977. [17] Carvalho RM, Tay FR, Sano H, Yoshiyama M, Pashley DH. Long-term mechanical properties of EDTA-demineralized dentin matrix. J Adhes Dent 2000;2:1939. [18] Ferracane JL, Hopkin JK, Condon JR. Properties of heattreated composites after aging in water. Dent Mater 1995; 11:3548. [19] Ferracane JL, Berge HX, Condon JR. In vitro aging of dental composites in watereffect of degree of conversion, ller volume, and ller/matrix coupling. J Biomed Mater Res 1998;42:46572. [20] Drummond JL, Botsis J, Zhao D, Samyn D. Fracture properties of aged post-processed dental composites. Eur J Oral Sci 1998;106:6616. [21] Soles CL, Yee AF. A discussion of the molecular mechanisms of moisture transport in epoxy resins. J Polym Sci: Part B: Polym Phys 2000;38:792802. [22] Van Landingham MR, Eduljee RF, Gillespie Jr JW. Moisture diffusion in epoxy systems. J Appl Polym Sci 1999;71: 78798. [23] Ruyter IE, ysaed H. Composites for use in posterior teeth: composition and conversion. J Biomed Mater Res 1987;21: 1123. [24] Kalachandra S, Turner DT. Water sorption of polymethacrylate networks: bis-GMA/TEGDM copolymers. J Biomed Mater Res 1987;21:32938. [25] Beatty MW, Swartz ML, Moore BK, Phillips RW, Roberts TA. Effect of crosslinking agent content, monomer functionally, and repeat unit chemistry on properties of unlled resins. J Biomed Mater Res 1993;27:40313.

[26] Venz S, Dickens B. NIR-spectroscopic investigation of water sorption characteristics of dental resin composites. J Biomed Mater Res 1991;24:123148. [27] Adamson MJ. Thermal expansion and swelling of cured epoxy resin used in graphite/epoxy composite materials. J Mater Sci 1980;173645. [28] Tanaka J, Ishikawa K, Yatani H, Yamashita A, Suzuki K. Correlation of dentin bond durability with water absorption of bonding layer. Dent Mater J 1999;18:1118. [29] Ferracane JL. Elution of leachable components from composites. J Oral Rehabil 1994;21:44152. [30] Yourtee DM, Smith RE, Russo KA, Burmaster S, Cannon JM, Eick JD, Kostoryz EL. The stability of methacrylate biomaterials when enzyme challenged: kinetic and systematic evaluation. J Biomed Mater Res 2001;57: 52231. rtengren U, Andersson F, Elgh U, Terselius B, Karlsson S. [31] O Inuence of pH and storage time on the sorption and solubility behaviour of three composite resin materials. J Dent 2001;29:3541. [32] Braden M, Clarke RL. Water absorption characteristics of dental microne composite lling materials. I. Proprietary materials. Biomaterials 1984;5:37296. [33] So derholm JK. Water sorption in a bis(GMA)/TEGDMA resin. J Biomed Mater Res 1984;18:2719. [34] Oysaed H, Ruyter IE. Water sorption and ller characteristics of composites for use in posterior teeth. J Dent Res 1986;65:13158. rtengren U, Wellendorf H, Karlsoon S, Ruyter IE. Water [35] O sorption and solubility of dental composites and identication of monomers released in an aqueous environment. J Oral Rehabil 2001;28:110615. [36] Paul SJ, Leach M, Rueggeberg FA, Pashley DH. Effect of water content on the physical properties of model dentine primer and bonding resins. J Dent 1999;27: 20914. [37] Burrow MF, Satoh M, Tagami J. Dentin bond durability after three years using a dentin bonding agent with and without priming. Dent Mater 1996;12:3027. [38] Kato G, Nakabayashi N. The durability of adhesion to phosphoric acid etched, wet dentin substrates. Dent Mater 1998;14:34752. [39] Tja derhane L, Larjava H, Sorsa T, Uitto VJ, Larmas M, Salo T. The activation and function of host matrix metalloproteinases in dentin matrix metalloproteinases in dentin matrix breakdown in caries lesions. J Dent Res 1998;77: 16229. [40] van Strijp AJ, Jansen DC, DeGroot J, ten Cate JM, Everts V. Host-derived proteinase and degradation of dentin collagen in situ. Caries Res 2003;37:5865. [41] Tja derhane L, Palossari H, Sulkala M, Wahlgren J, Salo T. The expression of matrix metalloproteinases (MMPs) in human odontoblasts. In: Proceedings of the International Conference on Dentin/Pulp Complex; 2002. p. 4551. [42] Martin-De Las Heras S, Valenzuela A, Overall CM. The metalloproteinase gelatinase A in human dentine. Arch Oral Biol 2000;75765. [43] Bowen RL, Rodriguez MM. Tensile strength and modulus of elasticity of tooth structure and several restorative materials. J Am Dent Assoc 1962;64:37887. [44] Carvalho RM, Fernandes CA, Villanueva R, Wang L, Pashley DH. Tensile strength of human dentin as a function of tubule orientation and density. J Adhes Dent 2001;3: 30914. [45] Inoue S, Pereira PN, Kawamoto C, Nakajima M, Koshiro K, Tagami J, Carvalho RM, Pashley DH, Sano H. Effect

Durability of resin dentin bond components

241

of depth and tubule direction on ultimate tensile strength of human coronal dentin. Dent Mater J 2003; 22:3947. [46] Kinney JH, Marshall SJ, Marshall GW. The mechanical properties of human dentin: a critical review and reevaluation of the dental literature. Crit Rev Oral Biol Med 2003;14:1329. [47] Habelitz S, Marshall GW, Balooch M, Marshall S. Nanoindentation and storage of teeth. J Biomech 2002; 35:9958.

[48] Gendron R, Greiner D, Sorsa T, Mayrand D. Inhibition of the activities of matrix metalloproteinases 2,8 and 9 by chlorhexidine. Clin Diagn Lab Immunol 1999;6: 4379. [49] Karring T. Regenerative periodontal therapy. Review. J Int Acad Periodontol 2000;2:1019. [50] Golub LM, Lee HM, Ryan ME, Giannobile WV, Payne J, Sorsa T. Tetracyclines inhibit connective tissue breakdown by multiple non-antimicrobial mechanisms. Adv Dent Res 1998;12:1226.