Effect of different light curing methods on mechanical and physical properties of resin-cements polymerized through ceramic discs.

OBJECTIVE: The aim of this study was to compare the polimerization ability of three different light-curing units (quartz tungsten halogen, light-emitting diodes and plasma arc) and their exposure modes (high-intensity and soft-start) by determination of microhardness, water sorption and solubility, and diametral tensile strength of 5 dual-curing resin cements.
MATERIAL AND METHODS: A total of 720 disc-shaped samples (1 mm height and 5 mm diameter) were prepared from different dual-curing resin cements (Duolink, Nexus, Bifix-QM, Panavia F and RelyX Unicem). Photoactivation was performed by using quartz tungsten halogen (high-power and soft-up modes), light-emitting diode (standard and exponential modes) and plasma arc (normal and ramp-curing modes) curing units through ceramic discs. Then the samples (n=8/per group) were stored dry in the dark at 37°C for 24 h. The Vickers hardness test was performed on the resin cement layer with a microhardness tester (Shimadzu HMV). For sorption and solubility tests; the samples were stored in a desiccator at 37°C and weighed to a constant mass. The samples were weighed both before and after being immersed in deionized water for different periods of time (24 h and 7 days) and being desiccated. The diametral tensile strength of the samples was tested in a universal testing machine at a crosshead speed of 0.5 mm/min. Data were analyzed statistically by nonparametric Kruskal Wallis and Mann-Whitney U tests at 5% significance level.
RESULTS: Resin cement and light-curing unit had significant effects (p<0.05) on microhardness, diametral tensile strength, water solubility and sorption. However, no significant differences (p>0.05) were obtained with different modes of LCUs.
CONCLUSION: The study indicates that polymerization of resin cements with different light-curing units may result in various polymer structures, and consequently different mechanical and physical properties.

INTRODUCTION

High-technology processes have led to the development of glass ceramics that are increasingly attractive in restorative dentistry because of their excellent mechanical properties, aesthetics and etchability[6,27]. Long-term survival of adhesive porcelain restorations depends on the success of a reliable bond between porcelain, the luting agents and the dental substrates. Because of their brittle nature, all-ceramic restorations rely on adequate bonding[7]. For luting ceramic restorations, the use of resin cements has increased considerably in the last years[15,22]. In most clinical cases, dual-curing resin cement is used when bonding ceramic to enamel and dentin[17]. Dual polymerization, the combination of light and chemical polymerization, provides a better conversion of monomers. This is important because inadequate polymerization is usually associated with poor mechanical and biological properties of the resin cements[19]. In addition, the polymerization of resin cement might also be affected by the characteristics of the luting material, such as chemical composition, filler particle size and shade[27]. Moreover, adequate polymerization of resin cements depends not only on resin cement but also on the light-curing unit (LCU) intensity, wavelength of the visible light and polymerization time[19].

Until recently, conventional quartz tungsten halogen (QTH) LCUs were widely used to polymerize resin cements[16]. These LCUs are susceptible to intensity output degradation with time as a result of the age of the bulb and its reflector, blistering and cracking of the filter, damage to the fiber-optic tips, because of repeated sterilization or heat generation[14]. To overcome these problems, boosted versions of QTH or high intensity LCUs such as plasma arc (PAC) and light-emitting diode (LED) LCUs that possess higher light intensity and shorter polymerization cycles than conventional LCUs have been developed[14,17]. In addition to increasing light intensity, various irradiation protocols, such as ramped and stepped intensity LCUs are marketed for their ability to "soft-start" polymerization[7]. These LCUs first use a lower intensity light followed by high-intensity light. This slower polymerization process allows the material to flow in the pregel stage, which provides some stress relief from polymerization contraction at the resin/dentin interface and reduces marginal gaps in these joints[4]. As a result, manufacturers claim that polymerization could be completed with higher intensity lights, which result in higher degree of monomer conversion, and is associated with improvements in the mechanical properties and of resin-based materials[13,20]. In this regard, modified light polymerization protocols can lead to the resulting polymer having different structures and properties, such as hardness, diametral tensile strength (DTS), and water solubility and sorption[1,11,20].

The interaction of dental resin-based materials with the aqueous oral environment may result in deterioration of their mechanical properties, dissolving and leaching of some of the components, such as unreacted monomers or fillers out of the samples leading to degradation and erosion[12,26]. In other words, sorption and solubility may act as precursors to a variety of chemical and physical processes that can lead to deleterious effects on the structure of the polymeric network, which can affect the suitability of the clinical applications of these materials[12,30]. Furthermore, surface hardness is also one of the most important physical characteristics of dental materials, and is defined as the resistance of a material to indentation or penetration[23]. Hardness testing is commonly used as a simple and reliable method to indicate the degree of conversion of resin-based cements[9]. Moreover, a basic mechanical property of resin-based materials, usually represented by manufacturers, is its strength[5]. An alternative method to test the strength of brittle resin-based materials, in which the ultimate tensile strength of a brittle material is determined through compressive testing, has become popular because of its relative simplicity and reproducibility of results[8,10]. The method is described in the literature as the diametral compression test for tension[10].

Accordingly, the following null hypotheses were tested: 1. The microhardness, water sorption and solubility, and DTS show dependence on the type of resin cements cured with different curing modes of high power LCUs; 2. High-power PAC LCUs might be considered for optimal curing efficiency of resin cements.

MATERIAL AND METHODS

Sample preparation

IPS Empress 2 (Ivoclar-Vivadent; shade A1, Schaan, Liechtenstein) cylindrical ceramic sample (5 mm in diameter and 2 mm in height) was fabricated by the lost wax technique and the ingot was injected to an EP 600 furnace (Ivoclar-Vivadent). The ceramic disc was ground with 220, 360, and 600 grit [Federation of European Producers of Abrasives (FEPA)] and air-abraded with 50 µm Al2O3 particles (Korox; Bego, Bremen, Germany) for 14 s from a distance of approximately 10 µm at 400 kPa with a sandblasting device (Ar-Ge Dental, Denizli, Turkey). Then the disc was cleaned in distilled water for 10 min in an ultrasonic bath (Healthsonics; Livermore, CA, USA) to ensure a contaminant-free ceramic surface.

The tested resin cements (Duolink, Bifix-QM, Nexus, RelyX Unicem and Panavia F), their manufacturers and compositions are shown in Figure 1. Seven hundred and twenty cylindrical discs (5 mm diameter and 1 mm high) were prepared by the same operator according to their manufacturers' directions. The methodology used in this study was based on the ADA Specification no. 27 and ISO standard 4049:2000[2,16].

Figure 1

Test materials and their composition according to the manufacturers

Trade Name	Chemical composition* (filler content by weight*)	Lot number	Manufacturer
Duolink	Base: Bis-GMA, TEGDMA, glass filler, UDMA	500009783	Bisco Inc., Schaumburg, U.S.A
	Catalyst: Bis-GMA, TEGDMA, glass filler (67% wt)
Bifix-QM	Bis-GMA, benzoylperoxide, amines, barium-aluminiumboro-silicate glass (71-73% wt)	640422	Voco GmbH, Cuxhaven, Germany
Nexus	SP-345, EBPADM, Fumed silica, UDMA, TEGDMA, Silane, SrSiF6, BPO, BisGMA, UV-9, ZnSiF6, BHT (68% wt)	449569	Kerr Co., Orange, CA
RelyX Unicem	Powder: glass powder, initiator, silica, substituted pyrimidine, calcium hydroxide, peroxy compound and pigment	290546	3M Espe AG,St. Paul, MN, USA
	Liquid: methacrylated phosphoric ester, dimethacrylate, acetate, stabilizer and initiator (72% wt)
Panavia F	Paste A: BPEDMA, MDP, DMA	Paste A: 00304A	Kuraray Medical Inc., Okayama, Japan
	Paste B: barium, boron, silicium glass and NaF(73% wt.)	Paste B: 00052A

Information provided by manufacturers.

Bis-GMA: Bis-phenol-A diglycidylmethacrylate; TEGDMA: Triethyleneglycol dimethacrylate; UDMA: Urethane dimetacrylate; SP-345: barium aluminoborosilicate; EBPADM: ethoxylated Bis-phenol-A-dimethacrylate; BPO: benzoyl peroxide; BHT: Butylated hydroxy toluene; HDDMA: Hexanediol dimethacrylate; SrSiF6: strontium hexafluorsilicate; UV-9: 2-hydroxy-4-methoxybenzophene; ZnSiF6: zinc hexafluorosilicate; BHT: 2,6-di-tert-butyl-4-methylphenol; BPEDMA: bisphenol-A polyethoxydimethacrylate; MDP: 10-methacryloyloxy decyl dihydrogenphosphate; DMA: aliphatic dimethacrlyate; NaF: sodium fluoride.

Figure 2 shows the experimental set-up of the study. The pastes A and B of the Panavia F hand-mixing cement was mixed in a 1:1 ratio on a mixing pad for 10 s. The other self-mixing cements (Duolink, Bifix-QM, Nexus and RelyX Unicem) were mixed by activating the syringes and triturating for 5 s. After placing the resin cements in a circular polytetrafluoroethylene mold, a transparent polyethylene film was placed over the filled mold and the ceramic disc was positioned over the resin cement.

Figure 2

Schematic illustration of sample preparation and the experimental design

Light activation was performed through the ceramic disc using three high-power LCUs: a quartz tungsten halogen (QTH, Blue Swan Digital; Dentanet, Turkey) (high-power and soft-up modes), a light-emitting diode (LED, Elipar Freelight 2; 3M Espe, St. Paul, MN, USA) (standard and exponential modes) and a plasma arc LCU (PAC, PlasmaStar, SP-2000; Monitex Industrial Co. Ltd., Taiwan) (normal and ramp-curing modes). While high-intensity mode corresponds to high power, standard and normal modes, soft-start mode corresponds to soft-up, exponential and ramp-curing modes. The characteristics of the LCUs are shown in Figure 3. After light polymerization, the samples were stored dry in dark at 37ºC for 24 h. All of the mechanical tests were performed by the same operator.

Figure 3

Light-curing units and modes

Light curing unit	Type of light source and diameter of tip	Curing Modes	Profiles (exposure period)	Output* (mW/cm2)
QTH	8-11.5 mm	Soft-up mode	The polymerization cycle of boost automatically from soft- start to high output (20 s)	1000
		High-power Mode	Unit constantly generates the full polymerizing energy (20 s)
PAC	8 mm	Normal mode	Provides full light intensity for the entire exposure period (10 s)	2250±50
		Ramp-curing Mode	Light exposure time 6 s total, intensity as follows: start 2 s at 50%, 2 s at 50%, 2 s at 100%
LED	8 mm	Standard mode	Provides full light intensity for the entire exposure period (20 s)	1200
		Exponential Mode	Provides light increasing to full intensity over the course of 5 s (20 s)

QTH=Quartz tungsten halogen - PAC=plasma arc - LED=light-emitting diode

Water solubility and sorption

For sorption and solubility measurements, 48 samples were prepared for each of the 5 test materials (n=48/per cement group). The thickness of the samples was measured using a digital caliper (Liaoning MEC Group; Dalian, China). All samples were placed in a desiccator containing freshly dried silica gel at 37ºC 24 h and weighed to a precision of 0.0001 g using a calibrated electronic balance (AX205 DeltaRang; Mettler Toledo, Giessen, Germany). This drying cycle was repeated until a constant mass (m1) for each disc was attained.

The samples were stored in distilled water at 37ºC for 24 h and 7 days. Then they were weighed after being carefully wiped with an absorbent paper. When constant weight was obtained, these values were recorded as m2. After these weighing procedures, the samples were returned to the desiccator at 37ºC until a constant mass was achieved (m3). The volume (V) of each sample was calculated in mm3.

The values for water sorption (Wsp) and solubility (Wsl), in µg/mm3, were calculated using the following equations[26]:

DTS

For DTS test measurements 240 samples were prepared (n=8/per group). The DTS of the samples (n=8/per group) was investigated through a diametral compression test. A compressive load was applied on the diametral surface of the samples to obtain DTS at a crosshead speed of 0.5 mm/min in a testing machine (Lloyd LRX; Lloyd Instruments, Fareham Hants, UK) until failure. After each compressive test, the fracture load (F) in Newton (N), was recorded and the DTS (MPa) was calculated from the equation[9]:

where P is the load at fracture, d is the diameter and t is the thickness.

Vicker's hardness

A total of 240 samples were polished under wet conditions with 220, 360, and 600 grit (FEPA) and placed on the platform of the tester with the surface being tested facing the diamond indenter (n=8/per group). The Vicker's microhardness test was performed on the cement layer with a microhardness tester (Shimadzu HMV; Shimadzu Corporation, Tokyo, Japan) with 200 g of load application for 15 seconds. Three indentations taken for each sample were not closer than 1 mm to the margin and were averaged to determine the hardness value for each sample.

Statistical analysis

Statistical analysis was performed using the Statistical Package for Social Sciences (SPSS) 11.5 software (SPSS Inc; Chicago, IL, USA). Whether the continuous variables were normally distributed or not was determined by using Shapiro Wilk test. Data were expressed as median (25th-75th) percentiles. The differences among resin cement and LCU groups were evaluated by using Kruskal Wallis test, and polymerization modes were compared by Mann Whitney U test. If the p value obtained from Kruskal Wallis test was statistically significant, multiple-comparison tests were used to determine which group differed from the others. Whether the differences between 24 h and 7 day measurements regarding solubility and sorption were statistically significant or not, was evaluated using Wilcoxon Sign Rank test (p<0.05). All possible subgroup analyses with Bonferroni adjustment were applied to control Type I error. A non-parametric Spearman's correlation analysis and Spearman's coefficient (rho) was used to assess the correlation between the variables.

RESULTS

The water solubility and sorption median values and (25th - 75th) percentiles are represented in Tables 3 and 4. Comparing the median values of solubility and sorption, Kruskal Wallis test revealed statistically significant differences among resin cement and LCU groups (p < 0.05). The effect of the storage time factor (24h and 7 days) on solubility and sorption behaviors was shown to be significant for the resin cements and LCUs (p < 0.05). Additionally, there was no significant difference between water solubility and sorption values of high-intensity and soft-start modes of the same LCU either after 24 h (water solubility: p = 0.635; sorption: p = 0.092) or 7 days (water solubility: p = 0.145; sorption: p = 0.189).

Table 3

Median values and (25th-75th) percentiles for diametral tensile strength (MPa) of the resin cements

Resin cement	LCU	Polymerization modes
		High intensity	Soft-start
Duolink	QTH	58.26 (55.08-59.71)	61.26 (54.4-65.26)
	LED	51.80 (50.07-57.76)	63.86 (52.40-68.44)
	PAC	57.64 (50.06-62.29)	43.58 (38.71-46.16)
Bifix-QM	QTH	45.36 (37.19-50.93)	51.21 (44.21-57.59)
	LED	56.07 (51.57-63.2)	55.86 (46.18-61.33)
	PAC	46.94 (44.45-51.39)	50.74 (46.9-58.72)
Nexus	QTH	54.57 (44.96-58.09)	54.53 (48.66-58.06)
	LED	57.41 (44.99-61.02)	63.82 (54.39-66.83)
	PAC	49.65 (42.63-55.33)	31.17 (26.92-36.14)
RelyX Unicem	QTH	41.06 (38.84-45.28)	32.87 (28.86-34.78)
	LED	43.57 (38.37-47.82)	35.57 (32.89-38.05)
	PAC	34.84 (32.74-42.48)	27.19 (24-30.24)
Panavia F	QTH	36.23 (34.19-39.31)	34.01 (30.53-37.1)
	LED	44.78 (41.72-46.15)	39.06 (34.65-41.94)
	PAC	24.28 (20.20-29.82)	27.55 (25.17-29.36)

LCU=light curing unit. - QTH=quartz tungsten halogen - LED=light-emitting diode - PAC=plasma arc

Table 4

Median values and (25th-75th) percentiles for Vicker's hardness number (VHN) of the resin cements

Resin cement	LCU	Polymerization modes
		High intensity	Soft-start
Duolink	QTH	42.57 (40.28-43.27)	44.03 (43.18-45.46)
	LED	44.63 (43.68-45.66)	44.13 (42.97-45.57)
	PAC	36.65 (35.73-37.75)	29.18 (28.23-30.71)
Bifix-QM	QTH	52.38 (51.32-53.18)	54.42 (54.24-57.06)
	LED	54.50 (53.03-56.63)	55.25 (52.97-56.63)
	PAC	43.90 (42.51-44.38)	33.05 (31.85-34.93)
Nexus	QTH	34.55 (33.32-36.38)	31.53 (30.77-35.43)
	LED	40.63 (39.68-41.13)	39.43 (37.59-40.84)
	PAC	28.35 (27.13-29.95)	20.15 (19.03-22.28)
RelyX Unicem	QTH	47.92 (47.32-49.98)	52.60 (50.82-53.48)
	LED	52.17 (51.16-53.83)	50.42 (48.88-51.87)
	PAC	37.90 (37.16-38.52)	21.15 (19.28-23.07)
Panavia F	QTH	32.13 (31.84-34.63)	31.25 (29.58-33.57)
	LED	24.92 (24.13-27.63)	27.57 (25.65-28.91)
	PAC	15.65 (15.2-18.05)	9.76 (8.81-11.02)

LCU=light curing unit. - QTH=quartz tungsten halogen - LED=light-emitting diode - PAC=plasma arc

Regarding water solubility, resin cements light polymerized with PAC showed the highest water solubility values except Bifix-QM. For Bifix-QM, the highest water solubility values after 24 h and 7 days storage in water were seen when this cement was polymerized with high-intensity mode of QTH LCU (24 h: QTH (9.91), LED (5.10), PAC (7.64); 7 days: QTH (7.88), LED (2.85), PAC (5.88)) (Table 3).

A ranking for water solubility of resin cements was obtained as follows: Duolink < Bifix-QM < Nexus < RelyX Unicem < Panavia F. Moreover, according to the water solubility results obtained from resin cements polymerized with different LCUs after 24 h and 7 days, LCUs were ranked as follows: LED < QTH < PAC (p < 0.05).

While the highest solubility value was observed after 7 days storage in distilled water with Panavia F polymerized with soft-start mode of PAC (152.14 µg/mm3), the lowest was observed after 7 days storage in distilled water with Duolink polymerized with high intensity mode of QTH (1.27 µg/mm3) (Table 3). All the tested resin cements, except Nexus, Bifix-QM and RelyX Unicem polymerized with soft-start mode of PAC, and Panavia F had a decrease in water solubility values with respect to water storage duration. According to the median values of the resin cements, the highest percentage of weight change in solubility values from 24 h to 7 days was seen in Panavia F (43.44 %, p < 0.05).

Regarding water sorption, sorption values measured after 24 h and 7 days were ranked as follows: Bifix-QM < Nexus < Duolink < RelyX Unicem < Panavia F. Furthermore, LCU ranking after 24 h can be shown as follows: LED < QTH < PAC. However, after 7 days the sorption values of the resin cements polymerized with LED presented statistically lower sorption values than PAC and higher values than QTH (p < 0.05).

The highest sorption value was obtained after 7 days storage in distilled water with Panavia F polymerized with soft-start mode of PAC (106.24 µg/mm3), the lowest was obtained after 7 days storage in distilled water with Bifix-QM polymerized with soft-start mode of QTH (4.59 µg/mm3) (Table 4). All the tested resin cements except Nexus polymerized with soft-start mode of QTH, high intensity modes of LED and PAC; Duolink polymerized with high intensity modes of QTH and LED, RelyX polymerized with soft-start mode of QTH and Bifix-QM polymerized with soft-start and high intensity modes of QTH had an increase in water sorption values with respect to water storage duration (Table 4). Similar to the solubility test, according to the median values of the resin cements, Panavia F showed the highest percentage of weight change in sorption values from 24 h to 7 days (26.54 %, p < 0.05).

Diametral tensile strength

Results of the DTS test are presented in Table 5. Resin cement and LCU had significant effects on DTS (p < 0.05). However, no significant differences were detected between the polymerization modes of the same LCU (p = 0.086).

The median DTS values of Duolink, Bifix-QM and Nexus were higher than that of RelyX Unicem and Panavia F and the ranking was obtained as follows: Panavia F < RelyX Unicem < Nexus < Bifix-QM < Duolink. Moreover, resin cements polymerized with LED LCU exhibited higher DTS values than QTH and PAC LCUs. Additionally, in all groups except Bifix-QM and Duolink polymerized with high intensity mode of PAC, DTS values decreased with the use of PAC LCU (Table 5). Duolink polymerized with soft-start mode of LED LCU had the highest median DTS value (63.86 MPa), whereas Panavia F polymerized with high intensity mode of PAC had the lowest (24.28 MPa).

Vicker's microhardness

The median VHN values and (25th - 75th) percentiles for samples are shown in Table 6. Although the data showed a significant difference for the resin cement groups and LCU groups (p < 0.05), the differences were not significant for the polymerization mode of the same LCU (p = 0.091).

Comparisons for VHN median values among resin cements revealed that Panavia F < Nexus < Duolink < RelyX Unicem < Bifix-QM. Bifix-QM polymerized with soft-start mode of LED yielded a highest median VHN value (55.25) whereas Panavia F polymerized with soft-start mode of PAC LCU presented the lowest value overall (9.76). Furthermore, QTH LCU presented higher VHN values than PAC LCU and lower values than LED LCU.

DISCUSSION

In this study, IPS Empresss 2 was selected as a representative ceramic material, because it has been widely used for single unit restorations as well as for three-unit fixed dental prosthesis of the anterior region extending to second premolar[6,17]. The adhesion of ceramic to dental structure with resin luting materials increases the fracture resistance of the restoration and tooth itself; it also minimizes microleakage, which may be the determining factor in the success or the failure of the treatment[6].

It is known that inadequate polymerization of resin cements might be a problem, especially under ceramic restorations[17]. Therefore, to reach maximum physical properties of resin cements; the conversion rate should be as high as possible[18]. In the present study, to simulate clinical conditions, the resin cement samples were irradiated from the top of the ceramic discs using LCUs, where the end of the light guide was in contact with the discs.

Investigations about the performance of new LCUs and polymerization modes by evaluating the physical and mechanical properties of resin-based materials are becoming increasingly available[17,27]. Under the experimental conditions of this study, the results indicated that three LCUs (QTH, LED and PAC) with different polymerization modes may point out different properties of resin cements and heterogeneity in the results.

In the literature, PAC LCUs are often discussed as an alternative to high-power LED and QTH LCUs[6,16]. However, this study indicated poor mechanical and physical properties for resin cements, polymerized with PAC LCU. The results showed that light curing with PAC had negative effects on water solubility, sorption, DTS and VHN of almost all of the resin cements, especially Panavia F, RelyX Unicem and Nexus. The special spectrum of this lamp, which has very high light intensities (2250±50 mW/cm[2] )) at certain wavelengths, might have caused this poor outcome. This might be related with the sudden polymerization shrinkage that follows curing with the PAC LCU. Additionally, high light intensity may lead the rapid development of the polymeric network in shorter molecular chains with low molecular weight and less cross-linking[9,28]. Moreover, the fact that shorter polymerization period of PAC is shorter than other LCUs might have resulted in an insufficient degree of polymerization. Besides, Uctasli, et al.[31] (2005) confirmed that PAC LCUs are so concentrated in narrow peaks (440-500 nm) that the initiator may not be decomposed in the given time.

A previous study reported that different polymerization modes can lead to resulting polymers having different structures[1]. However, in the present study, high-intensity and soft-start techniques exhibited similar values under almost all conditions, except Panavia F and RelyX Unicem in which soft-start mode of PAC LCU showed significantly higher solubility and sorption values than high-intensity mode (Tables 1 and 2). A previous study[19] reported that the two-step light-curing approach (soft-start technique) using different intensity distributions during polymerization, does not affect surface hardness. Similarly, Mehl, et al.[21] (1997) found that initial polymerization using decreased light intensity followed by high intensity has no effect on microhardness. For the DTS test performed in this study, there were no statistical differences between the polymerization modes. It could be possible that the DTS test was not sensitive to different polymerization modes, because it leads to fracture of the samples[1].

Table 1

Median values and (25th-75th) percentiles for solubility (μg/mm3) of the resin cements tested after 24 h and 7 days

Resin cement	SOLUBILITY (μg/mm3)
	Polymerization modes
		24 h		7 days
	LCU	High intensity	Soft-start	High intensity	Soft-start
Duolink	QTH	5.09 (5.09-7.77)	9.14 (5.73-10.19)	1.27 (0.64-2.34)	2.26 (1.40-2.99)
	LED	5.09 (5.09-7.01)	2.73 (2.28-2.96)	1.95 (1.78-2.47)	2.55 (1.27-4.46)
	PAC	10.47 (9.71-11.99)	10.19 (6.37-10.19)	2.33 (1.54-2.80)	4.08 (1.40-7.01)
Bifix-QM	QTH	9.91 (8.39-10.19)	5.1 (5.1-5.1)	7.88 (6.30-8.53)	1.66 (1.15-2.43)
	LED	5.1 (5.1-5.91)	5.1 (5.1-5.69)	2.85 (2.55-3.84)	4.33 (1.27-4.97)
	PAC	7.64 (5.1-10.19)	10.19 (8.88-10.19)	5.87 (5.22-8.09)	11.47 (10.19-12.93)
Nexus	QTH	13.32 (10.67-14.3)	10.19 (10.19-14.65)	9.91 (8.62-10.47)	8.66 (4.97-9.17)
	LED	5.86 (5.1-9.55)	10.19 (5.1-10.67)	1.51 (1.26-2.01)	8.13 (5.67-9.21)
	PAC	18.59 (15.75-20.38)	25.47 (23.75-26.31)	17.99 (14.39-20.25)	63.69 (59.76-70.96)
RelyX Unicem	QTH	27.47 (25.48-30.38)	35.67 (35.67-40.29)	23.65 (21.78-28.30)	32.33 (28.28-35.59)
	LED	20.38 (15.29-20.86)	17.95 (15.29-20.38)	10,96 (8.74-13.04)	10.5 (9.17-12.98)
	PAC	43.07 (30.57-45.86)	64 (61.61-67.83)	39.03 (35.8-46.21)	79.59 (69.38-87.91)
Panavia F	QTH	33.12 (31.84-34.63)	31.25 (29.58-33.57)	44.78 (43.26-47.64)	53.05 (48.49)
	LED	29.55 (24.20-40.25)	20.64 (19.11-28.66)	32.22 (30.19-35.33)	31.81 (29.01-34.03)
	PAC	31.79 (21.91-36.93)	77.01 (72.3-85.75)	73.93 (73.93-76.57)	152.14 (127.52-158.56)

LCU=light curing unit. - QTH=quartz tungsten halogen - LED=light-emitting diode - PAC=plasma arc

The values written in italics are higher than the required values reported in ADA Specification No. 27 (7.5 μg/mm3)

Table 2

Median values and (25th-75th) percentiles for sorption (μg/mm3) of the resin cements tested after 24 h and 7 days

Resin cement	SORPTION (μg/mm3)
	Polymerization modes
		24 h		7 days
	LCU	High intensity	Soft-start	High intensity	Soft-start
Duolink	QTH	10.61 (8.92-12.61)	10.19 (10.19-10.19)	7.13 (6.75-7.64)	13.04 (10.23-15.54)
	LED	10.19 (6.37-10.19)	10.45 (10.19-10.7)	9.68 (8.79-10.57)	20.89 (15.8-21.91)
	PAC	10.19 (10.19-10.19)	10.19 (10.19-15.29)	11.46 (7.26-15.8)	10.70 (9.81-14.01)
Bifix-QM	QTH	5.10 (5.1-6.27)	5.10 (5.1-10.19)	6.15 (5.1-7.53)	4.59 (4.08-6.5)
	LED	6.96 (5.1-7.51)	10.19 (10.19-10.19)	6.62 (5.22-8.97)	7.13 (5.6-8.03)
	PAC	5.10 (5.1-6.27)	7.64 (5.1-10.19)	6.88 (6.62-7.13)	9.17 (7.77-10.06)
Nexus	QTH	10.19 (10.19-14.01)	7.64 (5.1-10.19)	9.17 (8.79-10.38)	8.15 (7.64-8.66)
	LED	10.19 (5.1-10.19)	5.10 (5.1-10.19)	8.41 (5.1-9.68)	8.92 (7.26-9.55)
	PAC	10.75 (10.19-14.97)	24.01 (16.56-25.65)	11.97 (11.21-17.58)	16.31 (14.27-17.70)
RelyX Unicem	QTH	20.38 (16.56-25.48)	30.57 (30.57-34.4)	35.16 (33.63-38.6)	26.24 (22.04-30.96)
	LED	17.44 (15.29-20.38)	17.83 (15.29-20.38)	26.75 (24.97-28.41)	16.56 (15.29-19.24)
	PAC	30.57 (25.48-34.4)	50.96 (45.86-59.87)	36.18 (31.72-40)	51.21 (48.15-55.8)
Panavia F	QTH	36.69 (30.06-44.08)	36.18 (30.64-36.6)	36.18 (32.36-45.98)	53.50 (46.5-56.56)
	LED	29.55 (24.20-40.25)	34.14 (30.45-38.47)	32.22 (30.19-35.33)	40.25 (35.8-44.08)
	PAC	37.2 (35.10-38.06)	64.33 (57.32-71.34)	73.93 (73.93-76.57)	106.24 (97.83-108.54)

LCU=light curing unit. - QTH=quartz tungsten halogen - LED=light-emitting diode - PAC=plasma arc

The values written in italics higher than the required values reported in ADA Specification No.27 (40 μg/mm3)

In the present investigation, the tested materials exhibited considerable differences with respect to water solubility, sorption, DTS and VHN values (Tables 1-4). The samples polymerized with PAC presented significantly more water sorption and solubility than those polymerized with LED and QTH. A hypothesis for the high solubility and sorption values for the PAC LCU could be that, a reduced polymerization time, 6 s or 10 s, is not sufficient for effective polymerization in the deepest layers of the sample, resulting in shorter chain length, thus making it more sensitive in water.

A previous study by Fleming, et al.[14] (2007) compared the water solubility, sorption and VHN of different resin composites. In that study, the authors indicated that the increased filler content would result in reduced water sorption and solubility values. However, in the current study, Panavia F having high filler content (73% by weight) had the lowest values in almost all tests (highest values in water sorption and solubility tests). The reason for this could be that the use of hand-mixing Panavia F, unlike the other self-mixing cements tested in this study, might have resulted in mixing errors associated with different polymerization initiators and reactions. Self-mixing may help discard susceptibility to operator-induced variability[24]. Thus, unfavorable results of Panavia F could be related with the air inclusions during the mechanical mixing or consistency of the material, which might more readily accept air inclusion during the hand-mixing process[25]. Additionally, the poor results of RelyX Unicem might be due to the relatively high viscosity of the material and the limited penetration/interaction time. Higher viscosity results in lower degree of conversion increasing the mobility of molecules[7]. As the physical and mechanical properties of resin based materials are strongly influenced by the degree of conversion, Rely X Unicem showed inferior properties when compared with the other resin cements[29].

According to the ADA Specification No. 27[2], water solubility of all materials shall be less than 7.5 µg/mm3 and the sorption of resin-based materials shall be less than 40 µg/mm3. The median solubility values of the resin cements tested in this study were between 1.27-152.14 µg/mm3 and most of them were above the required values (Table 1). In addition, the sorption values of the resin cements tested in this study were between 4.59-106.24 µg/mm3 (Table 2). The sorption values except some of the values of Panavia F and RelyX Unicem indicated in Table 2, met the sorption standard's limits. Additionally, a strong positive relationship was observed between sorption and solubility values. Therefore, it is clear that the greater the amount of water absorbed, the greater the amount of components that could leach out from resin cements.

It has been shown that hardness is useful in determining the development of the mechanical properties of composites during their polymerization reaction, and that there is a direct correlation between degree of conversion and hardness development during polymerization, as a consequence of the increase in stiffness and strength of the material[27]. However, unlike degree of conversion, other important characteristics of the polymer network might affect hardness, such as the chemical structure of the monomers involved and the type and density of cross-linking. Therefore, indentation testing may give a more accurate characterization of the polymerized material than degree of conversion analysis[27]. To illustrate, a previous study by Reges, et al.[27] (2008) investigated the Knoop hardness of different shades of dual-polymerized resin cement light-activated through a ceramic restoration or not and found that hardness analysis might be more accurate than degree of conversion assessment for evaluating polymeric network structures.

On these grounds, in the current study, the hardness testing was used to evaluate the resin cement's polymerization. Furthermore, when comparing the effect of LCUs on microhardness, significant differences were observed between PAC and the other two LCUs (QTH and LED) (p<0.05). This finding was in accordance with a previous study that measured the hardness of luting composites and demonstrated differences between LCUs (QTH and PAC)[18]. Further on the effect of activation modes on resin cements, an investigation by Fonseca, et al.[15] (2005) related the reduced hardness of dual-curing resin cements with the partial absorption of light by the restoring esthetic materials.

As for the comparison of microhardness values of the resin cements, Duolink and Nexus showed lower microhardness values than Bifix-QM and RelyX Unicem (Table 4). This could be partially attributed to the differences in filler load, filler type, resin matrix, and formulation. The amounts of fillers used in this study were 67% and 68% by weight for Duolink and Nexus respectively, which were lower than Bifix-QM and RelyX Unicem (71-73% and 72%, respectively) (Figure 1). Moreover, Duolink and Nexus, consisting TEGDMA, presented lower values for hardness, while resin cement with Bis-GMA (Bifix-QM) showed higher values. A recent study by Moraes, et al.[22] (2010) confirmed that hardness is lower for TEGDMA-rich resin cements.

DTS is an acceptable and common test for resin based materials[8]. In a previous study, the DTS values of RelyX Unicem and Panavia F polymerized with QTH for 60 sec, were 44 MPa and 51.6 MPa, respectively[15]. In the present study, the DTS values of RelyX Unicem and Panavia F polymerized with QTH for 20 sec, were 41.06 MPa and 36.23 MPa, respectively (Table 3). These differences in DTS values might be attributed to the variation in polymerization periods.

The diametral tensile strength test might reveal different values for apparently similar materials. However, this variation could be explained by the difference between polymeric matrix, size of fillers and bond between fillers and matrix[3]. The higher the amount of TEGDMA, the more polar the organic phase, and therefore the better the interaction with the inorganic fillers, increasing the DTS[22,29]. The flexibility and low-molecular weight of TEGDMA, in addition to being responsible for its increased conversion potential, give this monomer great cross-linking capacity. Thus, high amounts of TEGDMA may enhance the 3D microstructure of the network by increased cross-linking reactions, improving the mechanical strength[22]. Additionally, Fonseca, et al.[15] (2005) evaluated the DTS of various dual-curing resin cements and indicated that the replacement of Bis-GMA or TEGDMA by UDMA results in an increase in DTS. This result is consistent with the present study, whereby Duolink and Nexus were the compounds comprising Bis-GMA, TEGDMA and UDMA. Therefore, these cements presented higher DTS values than RelyX Unicem and Panavia F.

In the context of this study, a strong negative correlation was seen between DTS and solubility after 24 h and 7 days (rho=-0.738 and rho =-0.756, respectively, p<0.001), while a weak positive correlation between DTS and VHN was observed (rho=0.385, p<0.001). Similarly, Medeiros, et al.[20] (2007) investigated the VHN and DTS of resin composites and observed a positive correlation between these two tests. Moreover, a strong negative correlation was found between DTS and sorption after 24 h and 7 days (rho=-0.695 and rho=-0.660, respectively, p<0.001). Additionally, VHN was found to be strongly and negatively correlated with solubility (24 h: rho=-0.534 and 7 days: rho=-0.589, p<0.001) and sorption tests (24 h: rho=-0.526 and 7 days: rho=-0.539, p<0.001).

The present results can be compared with those obtained by previous authors only to a limited extent, as the in vitro conditions mean that a large number of different variables may affect the results of the study; that is these are not reproducible. It is worthwhile to point out that in vitro studies are limited in their attempt to simulate clinical conditions. Despite their limitations, in vitro studies are simple, repeatable and inexpensive to perform. In the present study, the first limitation was the use of water for sorption and solubility tests instead of saliva that was encountered intraorally. Another limitation concerns the fact that aging process could be useful to understand clinical behavior of these resin cements. The data obtained in this in vitro study should be supported by the results of clinical investigations. Further research is necessary to evaluate the effects of these LCUs on other properties of resin cements.

CONCLUSIONS

Within the limitations of this in vitro study, the following conclusions could be drawn:

PAC units may require considerably longer exposure times than the manufacturers' recommendations to adequately polymerize resin cements. Thereby, this may lead to better physical and mechanical properties;

High-power LED LCUs might be considered as effective as or more effective than QTH LCU for polymerization of the resin-based materials;

The results of this study do not justify the polymerization of hand-mixing Panavia F with PAC LCU;

A positive and linear correlation could be observed between DTS and VHN. Moreover, with the increase in water solubility and sorption, a decrease in DTS and VHN values could be seen.