Biocompatibility and bond degradation of poly-acrylic acid coated copper iodide-adhesives

Abstract

Objective.

To investigate the effect of poly-acrylic acid (PAA) copper iodide (CuI) adhesives onbond degradation, tensile strength, and biocompatibility.

Methods.

PAA-CuI particles were incorporated into Optibond XTR, Optibond Solo and XP Bondin 0.1 and 0.5 mg/ml. Clearfil SE Protect, an MDPB-containing adhesive, was used as control.The adhesives were applied to human dentin, polymerized and restored with compositein 2 mm-increments. Resin–tensilebond strength after 24 h, 6 months and 1 year. Hourglass specimens (10 × 2 × 1 mm) wereevaluated for ultimate tensile strength (UTS). Cell metabolic function of human gingivalfibroblast cells exposed to adhesive discs (8 × 1 mm) was assessed with MTT assay. Copperrelease from adhesive discs (5 × 1 mm) was evaluated with UV–vis spectrophotometer afterimmersion in 0.9% NaCl for 1, 3, 5, 7, 10, 14, 21 and 30 days. SEM, EDX and XRF were conductedfor microstructure characterization.

Results.

XTR and Solo did not show degradation when modified with PAA-CuI regardless ofthe concentration. The UTS for adhesives containing PAA-CuI remained unaltered relative tothe controls. The percent viable cells were reduced for Solo 0.5 mg/ml and XP 0.1 or 0.5 mg/mlPAA-CuI. XP demonstrated the highest ion release. For all groups, the highest release wasobserved at days 1 and 14.

Significance.

PAA-CuI particles prevented the bond degradation of XTR and Solo after 1 yearwithout an effect on the UTS for any adhesive. Cell viability was affected for some adhesives.A similar pattern of copper release was demonstrated for all adhesives.

1. Introduction

Despite significant progress to improve the survival of com-posite restorations, their long-term durability remains aconcern. Bacterial proliferation due to marginal breakdown,and hydrolysis of both resin-based polymers by salivaryesterases and demineralized collagen by endogenous pro-teases have been extensively investigated in the last fewdecades [1]. Agents with anti-bacterial and anti-proteolyticproperties such as benzalkonium chloride [2], glutaraldehyde[3], chlorhexidine digluconate [4], and methacryloyloxydode-cyl pyridinium bromide (MDPB) [5] have been proposed.

Nanoparticles with antibacterial properties, such as silver[6] and copper [7], have received considerable attention whenincorporated into adhesive materials [8] since bacteria are lesslikely to acquire resistance against metal nano-particles thanto other conventional narrow-target antibiotics [9]. This is pre-sumably because metals are known to act on a broad rangeof microbial targets, and several mutations would have tooccur for microorganisms to resist their antibacterial activ-ity [9]. Recently, adhesive materials containing poly-acrylicacid coated copper iodide (PAA-CuI) particles have demonstrated strong long-term antibacterial properties and reducedcollagen degradation [10,11]. We speculate that, in additionto their strong antibacterial properties, PAA-CuI-adhesives may help delay bond degradation. The biocompatibility of PAA-CuI containing materials, however, is still of concern since biodegradation of dental materials with the consequentbyproduct release into the oral environment, might then ini-tiate local and systemic biological responses. To date, the amount of copper released into the oral environment remains unknown.

Therefore, the purpose of this study was to evaluate adhesives containing PAA-CuI particles for their ability to delaybond degradation. Their effect on the adhesives’ mechanicalproperties, cytotoxicity and copper release was also investi-gated. Scanning electron microscopy (SEM), X-ray diffractionanalysis (EDX) and X-ray fluorescence (XRF) were conductedfor microstructure characterization. Specific aims includedevaluation of the PAA-CuI adhesives for bond strength after24 h, 6 months and 1 year of storage, ultimate tensile strength(UTS), cell viability with MTT assay and copper release.

2. Materials and methods

Three commercially available adhesives, XP Bond (XP,Dentsply, York, PA, USA), Optibond Solo Plus (Solo, Kerr,Orange, CA, USA) and Optibond XTR (XTR, Kerr), were mod-ified with 0.1 and 0.5 mg/ml PAA-CuI particles. Clearfil SEProtect (Protect, Kuraray America, New York, NY, USA) contain-ing MDPB, a known antibacterial agent, was used as positivecontrol.

2.1. Preparation of the PAA-CuI adhesive resins

Both synthesis of PAA-CuI particles and generation of PAA-CuI adhesives was conducted according to Sabatini et al. [11].Briefly, 10 mg of PAA-CuI powder was admixed with 1 ml of one of three adhesives, XP, Solo and XTR, to yield a concentrated solution (10 mg/ml). To ensure uniform dissolution of the particles, a probe tip sonicator (Sonic Dismebrator 100,Fisher Scientific, Waltham, MA, USA) was used for 15 s under dark conditions in an iced-water bath. Immediately after, 10or 50 μl of the concentrated solution was micro-pipetted intoamber vials containing either 990 or 950 μl, respectively of the appropriate stock adhesive to yield a final working tration of 0.1 or 0.5 mg/ml of PAA-CuI adhesive. The following study groups were evaluated: 1) XP; 2) XP-0.1 CuI; 3) XP-0.5 CuI;4) Solo; 5) Solo-0.1 CuI; 6) Solo-0.5 CuI; 7) XTR; 8) XTR-0.1 CuI;9) XTR-0.5 CuI and 10) Protect. All procedures were performedat controlled room temperature (23 ± 2°C) and humidity conditions.

2.2. Micro-tensile bond strength (μTBS)

Thirty recently extracted, healthy human molars, obtainedunder a protocol approved by the State University of NewYork’s institutional review board (IRB ID No. 00000133), wereused to obtain dentin substrate. The teeth were equally andrandomly assigned to ten groups. A flat, transversely cutsurface of occlusal superficial dentin was obtained using awater-cooled lab trimmer (Whip Mix, Louisville, KY, USA),and a standardized smear layer created with 320-, 400- and600-grit silicon carbide abrasive paper (SiC paper, Buehler,Lake Bluff, IL, USA). All materials were applied and poly-merized following manufacturer’s recommendations (Table 1) with LED light-curing unit (VALO, Ultradent, South Jordan, UT,USA) with a power density of 1400 mW/cm2. Resin compos-ite (Filtek Z100, 3M ESPE, Saint Paul, MN, USA) was applied in increments less than 2 mm to the bonded surface and poly-merized for 40 s. After distilled water (DW) storage at 37°C for24 h, sixty resin–dentin beams (0.9 ± 0.1 mm²) per group were obtained according to the non-trimming technique [12], andincubated in DW for μTBS evaluation after 24 h, 6 months or1 year (n = 20). Individual beams were mounted on a stabiliz-ing jig with cyanoacrylate (Zapit, Dental Ventures of America,Corona, CA, USA) and stressed to failure with a universal test-ing machine (Bisco, Schaumburg, IL, USA) at a cross-headspeed of 1 mm/min. The load required to fracture the speci-men was expressed in megapascals (MPa).

Table 1 –

Study materials, including composition and application protocol, as per manufacturer’s recommendations.

Material	Composition	Application protocol
Ultra-etch (Ultradent) (Lot # B8RRK)	35% Phosphoric acid	• Rinse and dry surface • Apply etchant (20 s) • Rinse thoroughly, dry and apply adhesive per manufacturer’s instructions
Z100 (3M ESPE) (Lot #5904A4)	Matrix: Bis-GMA and TEGDMA Filler: zirconia and silica—84.5%–0.6 μm	• Place and light cure in 2 mm increments (40 s) with halogen or LED light with minimum intensity of 400mW/cm2
Optibond XTR (Kerr Corp.)	Primer: HEMA (30–50), acetone (25–35), ethyl alcohol (4–15)	• Apply SE primer with active scrubbing (20 s)
Two-bottle system Self-etch (Lot # 000045)	Bond: ethyl alcohol (20–30), alkyl dimethacrylate resins (47–68), barium aluminoborosilicate glass (5–15), silicon dioxide (3–10), sodium hexafluorosilicate (0.5–3)	• Air thin (5 s) • Apply Bond with active scrubbing (15 s) • Air thin (5 s) • Polymerize (10 s)
Optibond Solo Plus (Kerr Corp.) One-bottle system Etch-and-rinse (Lot #5475754)	Bis-GMA, HEMA, GPDM, ethanol, barium aluminum borosilicate glass, fumed silica (silicon dioxide), sodium hexafluoro-silicate, photoinitiator (CQ)	• Etch with 35% H3PO4 (15 s) • Rinse & lightly dry • Apply adhesive with active scrubbing (15s) • Air thin (3 s) • Polymerize (20 s)
XP Bond (Dentsply) Two-bottle system Etch-and-rinse (Lot # 141027)	TCB resin, PENTA, UDMA, TEGDMA, HEMA, butylated benzenediol (stabilizer), ethyl-4-dimethylaminobenzoate, CQ, functionalized amorphous silica, t-butanol	• Etch with 35% H3PO4 (15 s) • Rinse (15s) & blot dry • Apply adhesive • Leave undisturbed (20s) • Air thin (5 s) • Polymerize (10 s)
Clearfil SE Protect (Kuraray Medical Inc.) Two-bottle system self-etch (Lot #AE0027)	Primer: MDPB, MDP, HEMA, hydrophilic dimethacrylate, photo-initiator, water Bond: MDP, HEMA, bis-GMA, hydrophobic dimethacrylate, photo-initiators, silanated colloidal silica, surface-treated NaF	• Apply SE primer. Leave in place (20 s) • Evaporate excess solvent with mild air stream • Apply bond with microbrush • Gently air thin leaving uniform layer • Polymerize (10 s)

CQ, camphorquinone; HEMA, 2-hydroxyethylmethacrylate; MDP, 10-methacryloyloxydecyl dihydrogen phosphate; MDPB, 12- methacryloyloxydodecylpyridinium bromide; PENTA, phosphoric acid modified acrylate resin; TEGDMA, triethyleneglycol dimethacrylate; TCB resin, carboxylic acid modified dimethacrylate; UDMA, urethane dimethacrylate; Bis-GMA, bisphenol A diglycidyl methacrylate; GPDM, glycerol phosphate dimethacrylate; TCB, butan-1,2,3,4-tetracarboxylic acid di-2-hydroxyethylmethacrylate ester.

2.3. Ultimate tensile strength (UTS)

Hourglass-shaped specimens (10 × 2 × 1 mm) were fabricatedby placing the mold over a polyester strip, which was placedon top of a microscope glass slab. The adhesives were inserted carefully into the mold with care not to incorporate voids. Apolyester strip and a microscope glass slab were placed on topof the mold, the specimens polymerized with LED unit (VALO)for the designated time, and polished with SiC abrasive papers(320–2400). Ten specimens per group (n = 10) were incubated inDW at 37°C for 24 h, and then mounted and stressed to failureas described above.

2.4. Cell viability

Human gingival fibroblasts (HGFs) were isolated from dis-carded healthy gingival tissue collected from subjects undergoing surgical procedures at the University at Buffaloclinics, using an explant outgrowth method fresh tissue undera current IRB protocol at the University at Buffalo (IRB proto-col 663292–1). Tissue pieces were collected immediately andwashed in minimum essential Media (MEM) containing 10×antibiotics, followed by washing three times in media con-taining 1× antibiotics. Tissue pieces were minced into smallpieces and allowed to attach to wells of a 6-well plate for30 min, followed by the addition of 1 ml of fresh MEM + 10%FBS with antibiotics. Explants were incubated at 37°C, 5%CO2until migration of cells from the tissue explant couldbe observed, with replacement of fresh media every 3 days.Once 60% confluence was achieved, cells were detached using0.25% trypsin/0.05% EDTA and reseeded into larger flasks.For routine culture, HGFs were grown to 80% confluencein MEM + 10% FBS, and passaged using trypsin/EDTA. The3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide(MTT) cell viability assay (Sigma–Aldrich, St. Louis, MO, USA)was used to assess the cell metabolic function based onmitochondrial dehydrogenase activity. Three disc specimens(8 × 1 mm) per group (n = 3) were fabricated by micro-pipetting the adhesive into a circumferential polytetrafluoroethylene(PTFE) mold against a polyester strip, which was placed ona microscope glass slab and polymerizing from top and bot-tom with LED unit (VALO) for the designated time. The discswere polished with SiC abrasive papers and sterilized by glowdischarge sterilization for 15 min. Discs corresponding to theten groups were incubated in 24-well plates in 1 ml of gin-gival fibroblast growth medium (alpha-MEM + 10% FBS + 1%antimycotic; pH 7.4; Gibco, Life Technologies, Grand Island,NY) for 24 h. This incubation time was selected since themost toxic effects are normally observed during the first 24 has the resin is still undergoing polymerization [13]. An addi-tional group, containing gingival fibroblast medium only andno disc, was used as control, whose survival rates were set torepresent 100% viability. For the viability assay, HGFs (20,000cells/well) were grown in 96-well plates until 80% confluent,after which the culture media was removed, replaced with200 μl of filtered (0.2 μm filter, EMD Millipore, Billerica, MA, USA) disc-conditioned media and further incubated 48 h. TheMTT assay was performed by aspirating the disc test media,followed by incubation of the HGFs with 200 μl MEM (no phenol red) and 13 μl MTT assay reagent for an additional 4 h at 37°Cin 5% CO2, then replacement with 200 μl of DMSO at the end ofthe incubation period and measurement of the absorbance at 540 nm (Flexstation 3, Molecular Devices, Sunnyvale, CA, USA).

2.5. Copper release

Three disc specimens (5 × 1 mm) per group (n = 3) were fabri-cated as described above, and sterilized with 95 v/v% ethanolfor 5 min. Because even small amounts of contaminationcan lead to large errors when determining trace elementalconcentrations, the vials used for specimen’s storage wereacid-cleaned with an aqueous solution of 0.3 mol/l HNO3for12 h [14]. Specimens were individually stored in a cylindri-cal screw-top polyethylene test tube containing 5 ml of 0.9%sodium chloride (NaCl) buffered to pH 7.0 using 50 mmol/l ofHEPES [15]. The test tubes were incubated in an oscillating trayat 37°C for 30 cycles/min. A nylon monofilament (Fishing line,VWR, Radnor, PA, USA), which passed through a small holein the center of the specimen, was used to allow specimenretrievability without contaminating the solution. The ratio ofspecimen surface area to volume of solution was 0.1 cm²/ml, less than the range of 0.5–6.0 cm²/ml recommended for biolog-ical studies of medical devices by ISO Standard 10993 [16]. This,however, was deemed appropriate since no biological stud-ies were performed. Test tubes containing storage solutiononly and no disc were incubated under the same conditionsas control. To determine the maximum absorption wave-length of copper, a standard solution was first generated byadding 100 μg/ml of copper (Sigma–Aldrich) to 3% nitric acid(EMD Millipore,) in a 1 cm quartz cell. A wavelength scan wasperformed with UV–vis spectrophotometer (ND-1000, Nano-drop. Wilmington, NC, USA) between 250–700 nm to detectchanges in the molecular structure of copper as determinedby a change in the UV–vis spectrum. Copper standard solu-tions (0.03125, 0.0625, 0.125, 0.25, 0.5, 1 μg/ml) in 3% nitric acidwere prepared, and a calibration curve created by plotting theabsorbance values against the concentration of copper. Cop-per release from the adhesive specimens into the media wasmeasured with UV–vis spectrophotometer after immersion for1, 3, 5, 7, 10, 14, 21 and 30 days. At each of the testing peri-ods, the discs were retrieved from the test tube and placedin a fresh NaCl solution to incubate for the next time period.The regression equation derived from the calibration curvewas used to calculate copper concentration (μg/ml) in theunknown samples based on their absorbance values.

2.6. Scanning electron microscopy (SEM), energydispersive X-ray microanalysis (EDX) and X-rayfluorescence (XRF)

A disc (5 × 1 mm) per group was rinsed with DW, air dried,and placed on aluminum stubs with conductive tape, carboncoated and observed under SEM (SU70, Hitachi, Tokyo, Japan)using backscattered electron signal. Energy dispersive X-raymicroanalysis was used for elemental composition analysis.An area of 20 × 15 μm was selected for 2D analysis, whichincluded both resin matrix and filler particles. Another discof the same dimensions was used to quantify trace cop-per amounts in each adhesive using XRF (Innov-X Alpha 2portable X-ray Fluorescence) with an analysis area of 1 mmand expressed in parts per million (ppm).

2.7. Statistical analyses

A two-way analysis of variance (ANOVA) followed by Tukey’stest were used to analyze the effect of ‘treatment’, ‘time’ andtheir interaction on μTBS. A Kruskal Wallis ANOVA on ranksfollowed by Tukey’s test were used to identify differences inmean UTS values among groups. The effect of ‘treatment’ oncell viability was evaluated by one-way ANOVA and Tukey’stest. Friedman’s two-way ANOVA on ranks followed by Tukey’stest were used to evaluate the effect of “treatment”, “time”and their interaction on copper release. A significance level ofp < 0.05 was used for all tests. All the analyses were performedwith statistical software (SigmaStat version 3.5, San Jose, CA,USA).

3. Results

3.1. Micro-tensile bond strength

A significant effect of ‘treatment’ and ‘time’ (p < 0.001), but noeffect of their interaction (p = 0.063) was demonstrated. No sig-nificant variations were observed after 6 months for any of thegroups relative to their 24 h values. Except for Protect, all othercontrol adhesives demonstrated significant bond degradationafter 1 year (p = 0.027, p < 0.001 and p = 0.027 for XTR, Solo andXP, respectively) (Fig. 1). Neither XTR or Solo with PAA-CuIdegraded after 1 year regardless of the concentration, whereasXP Bond did (p < 0.001 and p = 0.003 for 0.1 or 0.5 mg/ml PAA-CuI respectively). Comparisons among groups means at eachtest period individually revealed a trend. While Solo demon-strated no significant variations, relative to their counterpartsmodified with PAA-CuI, XTR demonstrated lower values andXP higher values when modified with copper relative to theircorresponding controls. No differences were shown betweenany of the PAA-CuI-modified adhesives and Protect at anyof the test periods. The only exception was XP 0.1 mg/mlPAA-CuI, which demonstrated significantly lower μTBS thanProtect when evaluated after 1 year (p < 0.001).

Fig. 1 –

Mean micro-tensile bond strengths of PAA-CuI particles containing adhesives and corresponding control groups at24 h, 6 months and 1 year. Bars represent the mean values; brackets indicate the SD values. N = 20. Groups identified bydifferent letters are significantly different (Tukey’s test; p < 0.001). Upper case denotes differences among test periods foreach treatment group. Lower case denotes differences among treatment groups for each test period.

3.2. Ultimate tensile strength

A significant effect of the “treatment” was demonstrated(p < 0.001). While no differences between adhesives modifiedwith PAA-CuI and their corresponding controls were evi-denced, XP as a group, exhibited significantly lower valuesthan the other groups, including Protect (Fig. 2).

Fig. 2 –

Mean ultimate tensile bond strengths (UTS) of PAA-CuI particles containing adhesives and corresponding controlgroups. Bars represent the mean values; brackets indicate the SD values. N = 10. Groups identified by different letters aresignificantly different (Kruskal Wallis; p < 0.001).

3.3. Cell viability

A significant effect of the “treatment” was observed (p = 0.001).While the percent viable cells for XTR 0.1 and 0.5 mg/ml andSolo 0.1 mg/ml PAA-CuI remained unaltered relative to theircorresponding controls without copper, Solo 0.5 mg/ml andXP 0.1 and 0.5 mg/ml PAA-CuI showed significantly lower val-ues than their corresponding controls (p = 0.008 p = 0.005 and p < 0.001, respectively) (Fig. 3). XP as a group revealed signifi-cantly lower cell viability values than all other groups (p < 0.05).

Fig. 3 –

MTT assay results of cytotoxicity against human gingival fibroblast cells obtained from polymerized resin discs.Adhesives with PAA-CuI particles were compared with control adhesive groups and untreated cells (Control), whosesurvival rates were set to represent 100% viability. Results are expressed in percentage of viable cells. Bars represent themean values; brackets indicate the SD values. N = 3. Groups identified by different letters are significantly different (Tukey’stest; p < 0.05).

3.4. Copper release

A significant effect of the “treatment”, “time” and their inter-action was demonstrated (p < 0.001). XP 0.5 and 0.1 mg/mlPAA-CuI had the highest copper release, with values whichwere significantly higher than all other adhesives (p < 0.001). XTR 0.1 and 0.5 mg/ml PAA-CuI followed, with valueswhich were significantly higher than Solo 0.1 mg/ml PAA-CuI(p < 0.001 and p = 0.008, respectively) but no different from Solo0.5 mg/ml PAA-CuI. The control adhesives demonstrated sig-nificantly lower copper release values (0.248, 0.235, 0.224 and0.198 μg/ml for XTR, Solo, Protect and XP, respectively) thanPAA-CuI adhesives (p < 0.001). In general, copper release washigh at day 1 (0.750 μg/ml) followed by a decrease at days3 (0.534 μg/ml) and 5 (0.382 μg/ml), a subsequent increase at days 7 (0.476 μg/ml) and 14 (0.822 μg/ml) and plateauing atdays 21 (0.686 μg/ml) and 30 (0.750 μg/ml). Overall, release val-ues at days 1 and 14 were significantly higher than all otherperiods (p < 0.001). The release value at day 30 was no dif-ferent from days 1, 14 and 21, but was significantly different from days 3, 5 and 7 (p < 0.001). The lowest release values wereobserved at day 5 with a significantly lower value than allother periods (p < 0.001). Fig. 4 summarizes the copper release values in μg/ml for each adhesive at the different incubationperiods, as well as the multiple comparisons among groups.In addition, the mean copper release per day for the differ-ent copper-modified adhesives at days 1, 3, 5, 7, 14, 21 and 30 were calculated as follows: 1.27, 0.38, 0.26, 0.30, 0.15, 0.13and 0.12 μg/ml for XTR 0.1 mg/ml PAA-CuI; 0.82, 0.20, 0.26,0.22, 0.20, 0.16 and 0.13 μg/ml for XTR 0.5 mg/ml PAA-CuI; 0.94,0.18, 0.21, 0.24, 0.15, 0.10 and 0.10 μg/ml for Solo 0.1 mg/mlPAA-CuI; 0.72, 0.34, 0.25, 0.31, 0.16, 0.11 and 0.10 μg/ml forSolo 0.5 mg/ml PAA-CuI; 1.11, 0.58, 0.38, 0.33, 0.17, 0.16 and0.13 μg/ml for XP 0.1 mg/ml PAA-CuI; 1.27, 0.69, 0.34, 0.65, 0.20,0.19 and 0.14 μg/ml for XP 0.5 mg/ml PAA-CuI.

Fig. 4 –

Copper release values (μg/ml) for the different adhesives modified with PAA-CuI particles and corresponding controlgroups at the different test periods. N = 3. For each adhesive-concentration group, different superscript letters indicatesignificant differences among test periods (days-horizontal) (Tukey’s test, p < 0.05).

3.5. Scanning electron microscopy (SEM), energydispersive X-ray microanalysis (EDX) and X-rayfluorescence (XRF)

The SEM micrographs (Fig. 5) reveal the crystallinity and sizepolydispersity of the PAA-CuI particles. The particle sizesranged from 1–3 μm for the individual particles to 150–250 μmfor the CuI agglomerates. The EDX spectrum of the PAA-CuIparticles shows a ratio of Cu (K-line) to I (L-line) of 40:60,confirming its composition as copper iodide. Except for Solo0.1 mg/ml PAA-CuI, the presence of copper was confirmed withSEM and EDX in all other adhesives (Fig. 6). The EDX spec-trum verified the inorganic filler content as silicon dioxide forOptibond XTR, barium aluminoborosilicate and silicon diox-ide for Solo, and primarily silica for XP. XRF demonstrated thepresence of copper in all PAA-CuI containing adhesives, whichwas proportional to the concentration of PAA-CuI in the adhe-sive (Table 2). XRF values were 167 and 597 for Solo 0.1 and0.5 mg/ml, respectively; 150 and 516 for XP 0.1 and 0.5 mg/ml,respectively; and 81 and 555 for XTR 0.1 and 0.5 mg/ml, respec-tively. No copper was detected for any of the control groups.The values for control groups, Solo, XP, XTR and Protect, 67, 45,49 and 38, respectively, represent the threshold above whichthe element would be detected by the software for each group.

Fig. 5 –

SEM micrograph and EDX spectrum of PAA-CuI particles; magnification 250× (A), 3500× (B) and 10,000 (C). Elementsidentified by energy dispersive X-ray spectroscopy microanalysis for the PAA-CuI particles (D).

Fig. 6 –

SEM micrograph and EDX spectrum of PAA-CuI adhesives and corresponding control groups; magnification 1500×. Optibond XTR (A. Control, B. 0.1 mg/ml, C. 0.5 mg/ml), Optibond Solo (D. Control, E. 0.1 mg/ml, F. 0.5 mg/ml), and XP Bond (G.Control, H. 0.1 mg/ml, I. 0.5 mg/ml).

Table 2 –

Summary of all results including micro-tensile bond strength, ultimate tensile strength, ion release, cell viability and X-ray fluorescence.

Group	6-month bond degradation (%)	1-year bond degradation (%)	UTS (MPa)	Cell viability (%)	Overall ion release (μg/ml)	XRF (PPM)
XTR	3	−25	21.9	92.7	0.248	ND < 49
XTR 0.1	12	−19	18.7	94.3	0.871	81 (±20)
XTR 0.5	−12	−19	23.2	96.7	0.827	555 (±39)
Solo	−3	−66	16.2	93.9	0.235	ND < 67
Solo 0.1	−1	−11	16.3	90.9	0.682	167 (±27)
Solo 0.5	0	−28	17.7	85.1	0.745	597 (±35)
XP	−11	−63	9	71.4	0.198	ND < 45
XP 0.1	−17	−112	9.1	62.2	1.015	150 (±22)
XP 0.5	−22	−45	9.5	56.9	1.238	516 (±45)
Protect	−3	−10	18.3	99.2	0.224	ND < 38

4. Discussion

This study provides original information on the effects ofPAA-CuI modified adhesives on bond degradation, tensilestrength, and biocompatibility as per evaluation of the cyto-toxicity and ion release. No degradation was observed forany adhesive after 6 months. Six months has previously beenreported as insufficient time to detect the effects of hydrolyticdegradation [17]. After 1 year, however, all the control groups,except Protect, showed significant degradation with self-etchadhesives degrading less [XTR (25%) and Protect (10%)] thanetch-and-rinse [Solo (66%) and XP (63%)]. The dual adhesivemechanism of self-etch adhesives, that is the chemical adhe-sion to hydroxyapatite in addition to the micro-mechanicalinterlocking, has been reported to make these adhesives sta-ble even in aqueous environments [18], perhaps contributingto explain our findings. These interfaces may then better beable to withstand the hydrolytic breakdown of its components,with the consequent improvement in bond durability [19].Only XP showed signs of degradation after 1 year regardlessof the concentration of copper. The lower performance of XPrelative to other adhesives has been previously documented[20], and it may be the result of its lower filler content, whichwas confirmed by our SEM/EDX observations. Conversely, XTRand Solo demonstrated stable bonds after 1 year irrespective ofthe concentration of PAA-CuI (Table 2). This may be the resultof a myriad of factors. Our group has recently found that cop-per can reduce both degradation of collagen [10] and esteraseactivity (unpublished observations). In addition, copper is alsoknown to be a catalyst for HEMA radical polymerization [21].We speculate that an increased degree of monomer conversionmay have rendered some of the copper-containing adhesivesmore resistant to hydrolytic degradation and thus less suscep-tible to bond degradation.

Incorporation of PAA-CuI particles into the adhesives didnot affect their UTS. Similarly, XP exhibited significantly lowerUTS values than all other adhesives regardless of the presenceof copper. A recent study by Stanislawczuk et al. also reportedlow UTS values for XP (7.2 MPa) relative to other adhesives [22].

Incorporation of PAA-CuI to XTR did not seem to affect cellviability regardless of the concentration indicating that cop-per was not cytotoxic to gingival fibroblasts. Protect did notdemonstrate reduced cell viability either. Higher monomerconcentrations are known to lead to lower cell survival rates[23]. We speculate that the higher survival rates observed forspecimens of self-etch adhesives, XTR and Protect, may beattributed, at least partially, to an overall reduced monomerconcentration since these were fabricated with a solution ofadhesive only, as opposed to specimens of etch-and-rinseadhesives, Solo and XP, which were fabricated with a solutionof primer and adhesive. The cell viability of Solo was signifi-cantly decreased with addition of 0.5 mg/ml, but not 0.1 mg/mlPAA-CuI. As a group, XP demonstrated the lowest cell sur-vival rates, with cell viability values which were lower thanthe main control representing 100% viability. Other authorshave previously reported the low survival rates of XP Bond,with cell viability values of 32% [24], 30% [25] and even 10% [26]. This has been attributed to the higher levels of UDMA,TEGDMA, HEMA and camphorquinone present in the mixturerelative to other adhesives. Both, monomers [27–29] and cam-phorquinone [30–33] in the matrix of contemporary adhesives,are known to be cytotoxic.

Differences in copper release values were also observed.Overall, XP with 0.1 and 0.5 mg/ml demonstrated the highestrelease, and Solo 0.1 mg/ml the lowest release. Both filler parti-cle size and volume are known to play a role on ion release [34].Our SEM/EDX observations indicate that the filler content inXP is minimal relative to XTR and Solo. The increased matrixvolume in XP may have led to a greater ion release, relative tothe other adhesives. Ion release from resin-based materials isalso known to be dependent on pH [15] and hydrophilicity ofthe resin matrix [35]. We further speculate that the higher lev-els of hydrophilic monomers (e.g. HEMA) in the matrix of XP,compared to the other adhesives, may have also contributedto the observed increased rate of copper release. The highestrelease values shown for XP, relative to other groups, may thushelp explain the highest cytotoxicity effect observed for thisgroup. The control groups, interestingly, showed some cop-per release. This could be attributed either to the presence oftrace amounts of copper unreported by the manufacturer, ordifferent elements reactive at the same wavelength used todetect copper (245 nm). Our XRF observations demonstratingthe presence of minimal quantities of copper in the controladhesives, confirmed the former speculation. Analysis of thecopper release overtime demonstrated a consistent pattern forall adhesives. Days 1 and 14 demonstrated the highest releaseand days 5 and 7 the lowest release. An initial high releaseat day 1 was followed by an immediate gradual decrease atdays 3 and 5, another increase at days 7 and 14, finally reach-ing a plateau at days 21 and 30. The initial surge is likelyderived from the additional copper released from the top andbottom surfaces of the specimens. A correlation between theexposed surface of the sample and the release of specificcomponents has been previously reported [36]. The gradual decrease in release values at days 3 and 5 may be a moreaccurate representation of the true release pattern once thesuperficial copper has been exposed and released. A decreasein ion release values after an initial surge has been previ-ously reported in previous studies [34,37]. The second surgemay be the result of the exposure and release of additionalcopper particles as the resin matrix began to undergo swellingand hydrolysis during the first 14 days of incubation. Earlyeffects of water absorption have been previously reported [38].Water aging is known to increase filler particle pull out on theexposed surface, possibly due to a breakdown of the silanebonds between the resin and filler particles [39]. Moreover, since there is no silane coating present to promote adhesionof the copper particles to the matrix, these are expected to bemuch more susceptible to dislodgement.

Because element release plays a critical role in the mate-rial biocompatibility, it was critical to investigate, for the firsttime, the copper release pattern of PAA-CuI-containing mate-rials. Our in vitro model, though it hardly resembles a clinicalsituation characterized by a rapid and continuous flow andremoval of saliva, provides a simplified test set-up to allowfor an initial estimation of the rate and quantity of cop-per expected to be released from resin matrices without theundue influence of mechanical and enzymatic factors. Metalsare known to be more vulnerable to electrochemical degra-dation than other materials [40]. Copper modified adhesiveshave recently demonstrated excellent long-term antibacte-rial properties [11]. Ideally, antibacterial particles should befixed within the restorative material to prevent undesirableside effects to the host tissues as well as the potential lossof its antibacterial properties [41]. Overall, the copper releasevalues over the 30-day period were minimal ranging from0.198 to 1.238 μg/ml. The allowed daily dietary intake of copperhas been reported to be 11 μg/kg [42,43]. Ion release stud-ies normally present data cumulatively or additively for eachincubation period. In this study, the data is presented addi-tively for easier visualization of the amount of copper releasedat each interval. However, calculations of the mean releaseper day at each incubation period revealed that even the high-est release value per day (1.27 μg/ml) was still well below thereported toxicity threshold. This indicates thus that a sys-temic toxic effect from copper containing adhesives is highlyunlikely.

The SEM/EDX micrographs demonstrated the presence ofcopper particles in all adhesives except in Solo 0.1 mg/ml.We speculate that copper particles being relatively dense(8.96 g/cm³) compared to other elements in the periodic table, may tend to agglomerate toward the center of the specimen,especially when added in low concentrations, explaining whythese analytical techniques were unable to detect copper par-ticles in Solo 0.1 mg/ml. Because SEM provides only a visualanalysis of the surface and EDX an analysis to a depth of 2 μm,it is unlikely that the particles would remain suspended on thesurface and available for detection at all times. Moreover, EDXhas a detection level of 1%, which is higher than the concen-tration of copper in our experimental adhesives. The presenceof copper particles was confirmed by XRF in all groups, including Solo 0.1 mg/ml, validating thus our assumption con-cerning the particles density.

The present study provides evidence in support of thebeneficial role of PAA-CuI particles as an additive to dentaladhesives to help improve the bond stability of resin-dentinbonds without an adverse effect to the adhesive’s UTS, andwith the release of only minimal levels of copper that remainunder the accepted toxicity level. Additional studies using dif-ferent cytotoxicity models, adhesives of different compositionand concentrations of the additive need to be undertaken todetermine the optimal balance between antibacterial proper-ties without an adverse effect on the material’s bond strength,mechanical properties and toxicity to the host. A pH of 7 wasused in the present study for the incubation solutions as astarting point. However, ion release from dental materials hasshown to be highest at pH 4, decreasing at pH 5, and becomingminimal at pH 7 [44,45]. Because adhesives may be able to actas “smart” materials increasing copper release at cariogeniclow pH ranges (pH 5.5–6.5) during a biofilm acid challenge,future studies should also investigate ion release patterns atlower pH ranges to evaluate whether they remain under thereported toxicity thresholds.

5. Conclusion

Within the limitations of this study, it can be concludedthat incorporation of PAA-CuI particles was able to preventbond degradation after 1 year for XTR, Solo, but not XP. Nobond degradation was observed for any of the adhesives after6 months. The UTS remained unaffected for all adhesivesregardless of the concentration of PAA-CuI particles. Cell via-bility was not affected with the incorporation of 0.1 and0.5 mg/ml PAA-CuI to XTR or 0.1 to Solo, but it was reducedwhen incorporating 0.5 mg/ml to Solo or 0.1 and 0.5 mg/mlPAA-CuI to XP. XP demonstrated the highest copper release ofall adhesives. A similar pattern of copper release was observedfor all adhesives with a high release at day 1, followed by agradual decrease at days 3 and 5, another increase at days 7and 14 and release values plateauing at days 21 and 30.