Hydrophilic dyes as photosensitizers for photopolymerization of dental adhesives

Abstract

Objectives

This study explored hydrophilic dyes as photosensitizers for application in dental adhesives. The goal was to identify dyes that enhance the degree of conversion (DC) of the hydrophilic-rich phase without impairing polymerization of the hydrophobic-rich phase.

Methods

Properties that were investigated included the molar extinction coefficient at 480 nm, relative normalized photon absorption efficiency (PAE), rate of polymerization and degree of conversion (DC). The following hydrophilic dyes: Bromophenol blue sodium salt, Rosebengal sodium salt, Erythrosin B, New Fuchsin and Victoria blue B were identified as suitable photosensitizers.

Results

In this study it was observed that dyes such as Bromophenol blue sodium salt, New Fuchsin, Victoria blue B and Rosebengal sodium salt were suitable candidates for dental adhesive photopolymerization, leading to substantial degree of conversion to both the hydrophilic-rich phase and the hydrophobic-rich phase.

Conclusions

In addition to the ability of the photosensitizer to absorb light in the visible range and transition to an excited state as a result of the absorbed energy, other factors such as the efficiency of the photosensitizer/light curing unit (LCU) combination, stability/efficiency of the excited state of the photosensitizer and/or initiating reactive species play an important role in the photopolymerization of the dental adhesive.

Graphical Abstract

1. Introduction

A low viscosity dental adhesive is used to bond composite restorative materials to the tooth structure (dentin and enamel). The integrity of the bond formed at the composite/adhesive/tooth interface plays a vital role in the clinical lifetime of composite restorations [1]. While bonding dental adhesives to enamel has been largely successful, there are a multitude of biologic, mechanical and chemical stresses that challenge the integrity of the bond formed at the adhesive/dentin (a/d) interface. As an example, during the infiltration of dental adhesive resins through the wet demineralized dentin, they can separate into hydrophobic- and hydrophilic-rich phases [2, 3]. Adhesive phase separation threatens both the integrity and durability of the a/d interfacial bond [1, 4]. In addition to adhesive phase separation, suboptimal polymerization of the adhesive, partial adhesive infiltration of the demineralized dentin, water sorption, hydrolytic/enzymatic degradation of the adhesive, and diffusion of cariogenic bacteria undermine the a/d bond. The cumulative effect of these factors is a failed a/d bond and ultimately, clinical failure of the dental composite restoration [1].

The camphorquinone (CQ)/tertiary amine photoinitiator system is well-known and widely used for the free radical photopolymerization of dental adhesives and composites. Cook and colleagues proposed the specific mechanism for the CQ/tertiary amine system-the mechanism was based on tertiary amines, NN-3,5-tetramethylaniline and N,N-dimethylbenzylamine [5]. Hydrogen abstraction from the tertiary amine generates aminoalkyl and ketyl radicals. Aminoalkyl radicals participate in the initiation reaction while ketyl radicals are active in the termination [5, 6].

CQ and EDMAB are relatively hydrophobic and thus, preferentially accumulate in the hydrophobic-rich phase during adhesive phase separation [7, 8]. Investigations with the CQ/EDMAB photinitiator system and physically separated hydrophobic- and hydrophilic-rich phases, showed poor polymerization of the hydrophilic-rich phase [2, 3]. When an accelerator such as iodonium salt is added, the polymerization of the hydrophilic-rich phase improves but is still below an optimum level [7]. The poorly polymerized hydrophilic-rich phase is vulnerable to hydrolytic and enzyme-mediated degradation. Reduced polymerization leads to a loosely crosslinked hydrophilic-rich phase that will absorb water which will accelerate the diffusion of detrimental species into the a/d interface. These concomitant processes will compromise the integrity and stability of the a/d bond [1, 9].

The incorporation of hydrophilic co-initiator and photosensitizer has been explored as an approach for promoting the polymerization of the hydrophilic-rich phase [3, 10]. For example, Song et al. studied tris(trimethylsilyl)silane as a co-initiator for dental adhesive polymerization [11]. Kamoun et al. functionalized CQ with carboxylate group to improve the water solubility and compatibility of the photosensitizer [12]. Hydrogels containing carboxylated-CQ exhibited enhanced photo-reactivity as compared to non-functionalized CQ [12].

Dyes have been investigated as photosensitizers for polymerization reactions [13, 14]. Nan et al. used the dye-linked photoinitiator system (dyad) consisting of Erythrosin B (chromophore) chemically bonded to EDMAB (co-initiator) for the free radical polymerization of hexanediol diacrylate [15]. Xanthene dyes have been used as photosensitizers for the polymerization of acrylamide-bis(acrylamide) [16]. Xanthene dyes were also used with triethanolamine as the co-initiator to trigger photopolymerization of acrylamide in aqueous solution [17]. The lifetime of the singlet excited state of xanthene dyes was dependent on the type of solvent and pH [17].

Although dyes have been used in various photopolymerization reactions, there have been limited investigations using dyes as photosensitizers for dental adhesives. Moreover, previous studies did not focus on the photopolymerization of the partitioned phases in dental adhesives that experienced phase separation under conditions that simulate wet bonding. Prior investigations on the photopolymerization of the hydrophilic-rich phase focused on the photopolymerization of resin/water mixtures - not the photopolymerization of partitioned phases.

The objective of the current study was to investigate hydrophilic dyes as potential photosensitizers for the hydrophilic-rich phase of dental adhesive. Several hydrophilic dyes with maximum absorption wavelength in the visible range were selected. Properties of interest were degree of conversion (DC) of both hydrophobic- and hydrophilic-rich phases, molar extinction coefficient $(\xi )$ at the maximum emission wavelength (approximately 480 nm) of the halogen light curing unit (LCU), and relative normalized photon absorption efficiency (PAE) which was a measure of the efficiency of the combination of LCU and the photosensitizer. The goal was to identify dyes which would lead to substantial DC in the hydrophilic-rich phase without compromising the DC of the hydrophobic-rich phase.

2. Materials and Methods

The dental monomers used in this investigation were: 2-hydroxyethyl methacrylate (HEMA) and bisphenol A glycerolate dimethacrylate (BisGMA). The hydrophobic photosensitizer and co-initiator were: camphorquinone (CQ) and ethyl 4-(dimethyl)amino benzoate (EDMAB). The hydrophilic co-initiator and reaction accelerator were: 2-(dimethylamino)ethyl methacrylate (DMAEMA) and diphenyliodonium hexafluorophosphate (DPIHP).

The dyes were selected based on their hydrophilicity as compared to CQ, and whether their maximum absorption wavelength was within the visible range. The following hydrophilic dyes: Bromophenol blue sodium salt, Fluorescein sodium salt, Eosin Y disodium salt, Erythrosin B, New Fuchsin, Victoria blue B, Rosebengal sodium salt, and Methyleneblue chloride were investigated. Hydrophilic dyes, monomers, photosensitizer, co-initiators and reaction accelerator were purchased from Sigma Aldrich (St. Louis, MO, USA).

2.1. Hydrophilic and hydrophobic-rich phase preparation

The neat resin was prepared by mixing HEMA and BisGMA in 45:55 ratio by weight. The neat resins were selected based on 3-step etch and rinse dental adhesive system. The photoinitiator components were added to the neat resin in the ratio of 0.5 wt% CQ, 0.25 wt% hydrophilic dye, 0.25 wt% EDMAB, 0.5 wt% DMAEMA and 0.5 wt% DPIHP. The ratio in which photoinitiator (PI) components were added was based on the total weight of the neat resin. Approximately 33 wt% deuterium oxide (D₂O) was added to the mixture and the mixture was vortexed. The PI components were then added in the same ratio as discussed above but this time based on the weight of D₂O.

We have shown that for neat resin, HEMA/BisGMA with weight ratio of 45/55, the threshold limit to add D₂O without phase separation is 10 wt% [18]. Higher concentration of D₂O such as 33 wt% will create a model over-wet environment leading to phase separation [19].

The method used in the current investigation allowed the same ratio of PI components to be maintained based on the total weight of the mixture. The mixture was vortexed again to ensure that all the components dissolved. The mixture was centrifuged at 10,000 rpm for 20 minutes. Since the percentage of D₂O was beyond the miscibility limit for the resin, two separate phases (hydrophilic- and hydrophobic-rich) were obtained after centrifugation.

The separated hydrophilic-rich phase was collected from the top of the mixture using a pipette. The interface layer was removed and finally, the separated hydrophobic-rich phase was collected. Fig. 1 summarizes the specimen preparation technique. The mixtures were prepared in triplicate for each dye. Table 1 summarizes the composition of the mixture prior to the phase separation process. Fig. 2 shows the molecular structure of hydrophilic dyes used in this study.

Fig. 1.

Procedure for preparing dental adhesive hydrophobic- and hydrophilic-rich phases

Table 1.

Composition of each component by wt% within the dental adhesive mixture before phase separation

Photosensitizer in the model building set	HEMA	BisGMA	D2O	CQ	EDMAB	DMAEMA	Additional photosensitizer (Dyes)	DPIHP
Bromophenol blue sodium salt	29.31 ± 0.05	35.87 ± 0.06	32.81 ± 0.12	0.52 ± 0.01	0.26 ± 0.01	0.50 ± 0.01	0.25 ± 0.00	0.50 ± 0.00
Eosin Y disodium salt	29.34 ± 0.06	35.94 ± 0.01	32.73 ± 0.05	0.50 ± 0.01	0.25 ± 0.00	0.50 ± 0.01	0.25 ± 0.00	0.50 ± 0.01
Erythrosin B	29.38 ± 0.14	35.92 ± 0.17	32.7 ± 0.30	0.51 ± 0.01	0.26 ± 0.01	0.49 ± 0.00	0.25 ± 0.00	0.49 ± 0.01
Fluorescein sodium salt	29.31 ± 0.02	35.87 ± 0.02	32.84 ± 0.04	0.5 ± 0.01	0.25 ± 0.00	0.49 ± 0.01	0.25 ± 0.00	0.49 ± 0.00
Methylene blue chloride	29.42 ± 0.14	35.95 ± 0.17	32.64 ± 0.30	0.49 ± 0.01	0.25 ± 0.00	0.05 ± 0.00	0.25 ± 0.01	0.50 ± 0.01
New Fuchsin	29.24 ± 0.03	35.74 ± 0.04	33.03 ± 0.08	0.49 ± 0.01	0.25 ± 0.00	0.50 ± 0.00	0.25 ± 0.01	0.50 ± 0.00
Rosebengal sodium salt	29.38 ± 0.04	35.86 ± 0.05	32.78 ± 0.08	0.49 ± 0.01	0.25 ± 0.00	0.49 ± 0.00	0.25 ± 0.01	0.49 ± 0.01
Victoria blue B	29.27 ± 0.07	35.74 ± 0.08	33.00 ± 0.14	0.50 ± 0.00	0.25 ± 0.00	0.49 ± 0.00	0.25 ± 0.00	0.49 ± 0.01

Fig. 2.

Molecular structure of hydrophilic dyes

2.2. Photopolymerization kinetics study of hydrophilic and hydrophobic-rich phases

The polymerization kinetics was monitored using a Perkin-Elmer Spectrum 400 Fourier transform infrared spectrophotometer in the ATR sampling mode (PerkinElmer, Waltham, MA). The photopolymerization kinetics study was carried out separately for the hydrophilic- and hydrophobic-rich phases of each dye. The kinetics study was carried out using three samples per phase. The samples were exposed for 40s to a dental curing light (Spectrum® 800, Dentsply, Milford, DE) at 550 mW/cm² intensity. The dental curing light has maximum emission wavelength at 480 nm and the absorption peak of camphorquinone (CQ) is close to 480nm. This relationship between the emission wavelength and absorption peak allows optimal performance of CQ to trigger photopolymerization in the hydrophobic-rich phase since CQ is present in abundant quantity in this phase. A fixed volume of 30 μl of each phase, for each type of hydrophilic photosensitizer was placed on the ATR crystal. The sample was covered with a transparent coverslip and the edges of the coverslip were sealed to prevent evaporation of D₂O and to reduce oxygen diffusion into the system. The polymerization kinetics was monitored at a resolution of 4 cm⁻¹ for 2 hours in the case of the hydrophilic-rich phase and 1 hour for the hydrophobic-rich phase. Our prior investigations on dental hydrophilic-rich mimics showed that the polymerization kinetics must be monitored for at least 2 hours to observe the secondary gel effect [2]. Hence, based on our experience the polymerization kinetics of the hydrophilic-rich phase was monitored for extended time as compared to the hydrophobic-rich phase.

The degree of conversion (DC) was calculated based on the band ratio profiles of v(C=C) at 1637 cm⁻¹ to v(C=O) at 1716 cm⁻¹ (equation provided below). The rate of polymerization was obtained by differentiating the DC with respect to time using Microcal Origin (Version 6.0, Microcal Software, Northampton, MA). D₂O was used to decrease spectral interference from water, i.e. water exhibits spectral feature at the same wavelength as C=C, and hence would interfere with the monitoring of the conversion of C=C [2, 3].

$$DC=\left(1-\frac{{\text{ Absorbance }}_{1637{\text{cm}}^{-1}}^{\text{sample }}/{\text{ Absorbance }}_{1716{\text{cm}}^{-1}}^{\text{sample }}}{{\text{Absorbance}}_{1637{\text{cm}}^{-1}}^{\text{monomer}}/{\text{Absorbance}}_{1716{\text{cm}}^{\text{-1}}}^{\text{monomer}}}\right)$$ 2.1

2.3. Molar Extinction Coefficient

With the exception of Victoria blue B, a 9 μM solution of each dye in deionized (DI) water was prepared. In the case of Victoria blue B, a 27 μM solution in DI water was prepared - the higher concentration was necessary because the 9 μM solution resulted in very low absorption at 480 nm. Then 200 μl of each solution was added to wells of a 96 flat bottom well plate. The absorption of the solutions at 480 nm wavelength was measured using a Multi-Mode Microplate UV-Vis spectrophotometer (Bio Tek, Winooski, VT). The absorption for each dye was measured three times and the pathlength was corrected for 1 cm. The 480 nm wavelength was chosen for determining the molar extinction coefficient because it is approximately the same as the maximum emission wavelength of the halogen light curing unit (LCU) [20]. The molar extinction coefficient was determined using the equation shown below.

$$\mathit{\text{Absorbance}}=\xi \left[C_{mol/L}\right]X_{cm}$$ 2.2

Where $\xi$ is the molar extinction coefficient in L/mol.cm, [C] is the concentration of the dye in DI water in mol/L and X is the pathlength in cm.

2.4. Relative Normalized Photon Absorption Efficiency (PAE)

Photon absorption efficiency determines the overlap between the emission spectrum of the LCU and absorption spectrum of the dye. It is a measure of the number photons absorbed by the dye upon exposure to the LCU, and hence it represents the efficiency of the dye and LCU combination to generate excited species vital for generating initiating radicals [21–23]. Three solutions of each dye at the concentrations noted in 2.3 were prepared. The absorption spectra were measured using UV-vis spectrophotometer (Bio Tek, Winooski, VT). The number of photons per square centimeter with exposure to the halogen LCU (Dentsply Spectrum® 800) was determined using the spectral irradiance in mW/cm² in the equation below. The spectral irradiance was obtained from the published emission spectrum of the LCU [20, 21].

$$n_{ph\lambda }=\frac{w\lambda }{hc}$$ 2.3

Where w = spectral irradiance, λ = wavelength, h = Planck’s constant and c = speed of light

The photon absorption efficiency of each dye was obtained by plotting the product of n_ph at each wavelength and the absorption of each dye at the corresponding wavelength against wavelength, and then determining the area under the plot [21]. The photon absorption efficiency was normalized for concentration.

3. Results

3.1. Photopolymerization kinetics study of hydrophilic and hydrophobic-rich phase

Graphs representing the degree of conversion and rate of polymerization in the polymerization kinetics study of the hydrophilic-rich phase are presented in Fig. 3. With the exception of eosin Y disodium salt, fluorescein sodium salt and methylene blue chloride, all other dyes exhibited at least 47% degree of conversion for the hydrophilic-rich phase. Fluorescein sodium salt and methylene blue chloride did not show a single prominent peak and hence their rates were not reported in Fig. 3b. Methylene blue chloride showed the lowest peak rate for the hydrophilic-rich phase whereas eosin Y disodium salt showed the highest peak rate for this phase. The hydrophilic-rich phase containing Rosebengal sodium salt showed two peaks in a short interval - the rate reported here is for the peak with the greatest peak height. Table 2 summarizes the average degree of conversion and average peak rate of polymerization for the hydrophilic-rich phase.

Fig. 3.

Polymerization kinetics study of dental adhesive hydrophilic-rich phase showing (a) DC and (b) rate of polymerization

Table 2.

Degree of conversion and peak rate of polymerization for dental adhesive hydrophilic-rich phase in presence of dye as the second hydrophilic photosensitizer

Hydrophilic photosensitizer in the formulation before phase separation	Average DC	Average polymerization peak rate (s-1)
Bromophenol blue sodium salt	0.56 ± 0.04	0.021 ± 0.001
Eosin Y disodium salt	0.20 ± 0.07	0.037 ± 0.011
Erythrosin B	0.47 ± 0.08	0.016 ± 0.002
Fluorescein sodium salt	0.08 ± 0.03	0.036 ± 0.002
Methylene blue chloride	0.06 ± 0.01	0.010 ± 0.001
New Fuchsin	0.50 ± 0.07	0.020 ± 0.002
Rosebengal sodium salt	0.54 ± 0.05	0.025 ± 0.007
Victoria blue B	0.58 ± 0.02	0.023 ± 0.002

Graphs representing the degree of conversion and rate of polymerization for the polymerization kinetics study of the hydrophobic-rich phase are presented in Fig. 4. With the exception of Fluorescein sodium salt, all other dyes exhibited substantial degree of conversion for the hydrophobic-rich phase. The hydrophobic-rich phase with New Fuchsin, Methylene blue chloride and Victoria blue B showed an average degree of conversion of 90% or higher. The highest average degree of conversion was exhibited by hydrophobic-rich phase with New Fuchsin. The average peak rate of polymerization for Fluorescein sodium salt was the lowest and that for Bromophenol blue sodium salt was the highest.

Fig. 4.

Polymerization kinetics study of dental adhesive hydrophobic-rich phase showing (a) DC and (b) rate of polymerization

Table 3 summarizes the average degree of conversion and peak rate of polymerization of hydrophobic-rich phase. Both Rosebengal sodium salt and Eosin Y disodium salt showed two peaks for the rate of polymerization. For these two dyes, the rate maxima are the peak with the greatest peak height (Table 3). Since Fluorescein sodium salt did not exhibit a single prominent peak for rate, it was not reported in Fig. 4(b).

Table 3.

Degree of conversion and peak rate of polymerization for dental adhesive hydrophobic-rich phase in presence of dye as the second hydrophilic photosensitizer

Hydrophilic photosensitizer in the formulation before phase separation	Average DC	Average peak polymerization rate (s-1)
Bromophenol blue sodium salt	0.88 ± 0.02	0.242 ± 0.001
Eosin Y disodium salt	0.70 ± 0.05	0.046 ± 0.001
Erythrosin B	0.81 ± 0.04	0.091 ± 0.027
Fluorescein sodium salt	0.20 ± 0.05	0.028 ± 0.004
Methylene blue chloride	0.90 ± 0.01	0.175 ± 0.009
New Fuchsin	0.93 ± 0.02	0.199 ± 0.019
Rosebengal sodium salt	0.85 ± 0.02	0.058 ± 0.007
Victoria blue B	0.91 ± 0.01	0.195 ± 0.009

3.2. Molar Extinction Coefficient

Fig. 5(a) shows the representative absorption spectrum of the dyes in DI water at 9 μM concentration except for Victoria Blue B. The solution with Victoria Blue B had three times higher concentration due to its low overall absorption. Due to its peak absorption wavelength being significantly higher than 480 nm and its absorption being very low, molar extinction coefficient of Victoria blue B was low. Methylene blue chloride had the lowest molar extinction coefficient at 480 nm whereas New Fuchsin exhibited the highest. Fluorescein sodium salt had peak absorption wavelength at approximately 480 nm, and its molar extinction coefficient was also high despite having low overall absorption. Although the peak absorption wavelength of New Fuchsin was significantly higher than 480 nm, its overall absorption was high leading to an increased molar extinction coefficient. The peak wavelength of Methylene blue chloride was significantly higher than 480 nm leading to a low molar extinction coefficient. The peak absorption wavelengths of Rosebengal sodium salt and Bromophenol blue sodium salt were also significantly higher than 480 nm which could account for their low molar extinction coefficient at 480 nm. The peak absorption wavelength of Eosin Y disodium salt was slightly higher than 480 nm and its overall absorption was high which could account for its high molar extinction coefficient at 480 nm. Table 4 shows the average molar extinction coefficient at 480 nm for all the dyes along with their peak absorption wavelengths.

Fig. 5.

UV-vis spectroscopy results where the concentration of aqueous solution of all the dyes was 9 μM except for Victoria blue B. For the latter case the concentration was 27 μM. (a) Absorption spectrum of the dyes and (b) product of emission spectrum of LCU and absorption of each dye against wavelength.

Table 4.

Average molar extinction coefficient and peak wavelength of the hydrophilic dyes

Photosensitizer in the model building set	Peak wavelength (nm)	Average molar extinction coefficient at 480 nm (L/(mol cm))
Bromophenol blue sodium salt	590	14143 ± 510
Eosin Y disodium salt	515	51471 ± 1550
Erythrosin B	525	13750 ± 1674
Fluorescein sodium salt	480	46208 ± 1617
Methylene blue chloride	663	1594 ± 232
New Fuchsin	545	76883 ± 683
Rosebengal sodium salt	550	8001 ± 1792
Victoria blue B	615	5269 ± 134

3.3. Relative Normalized Photon Absorption Efficiency (PAE)

Since New Fuchsin showed the highest overall absorption followed by Eosin Y disodium salt, these dyes had high relative normalized photon absorption efficiency (PAE). Methylene blue chloride and Victoria blue B exhibited very low PAE due to their overall low absorption. Although this indicates poor efficiency for Victoria blue B as a photosensitizer with Spectrum 800 halogen LCU, its performance is also dictated by the efficiency and stability of the radicals generated by this dye to trigger the photopolymerization reaction.

Table 5 summarizes the average relative normalized PAE of all the dyes and Fig. 5(b) shows the representative product of photons per square centimeter per second and absorption of the dye against wavelength. The area under this graph was used to determine PAE.

Table 5.

Average relative normalized PAE of the hydrophilic dyes

Photosensitizer	Average normalized PAE
Bromophenol blue sodium salt	3.14 ± 0.08
Eosin Y disodium salt	6.30 ± 0.36
Erythrosin B	2.19 ± 0.07
Fluorescein sodium salt	4.42 ± 0.07
Methylene blue chloride	0.29 ± 0.04
New Fuchsin	10.65 ± 0.07
Rosebengal sodium salt	1.46 ± 0.03
Victoria blue B	0.85 ± 0.02

4. Discussions

Although molar extinction coefficient and PAE both play a vital role in the performance of a photosensitizer, its efficiency as well as stability when it is in an excited state also impact the performance in initiating the photopolymerization reaction. It should be noted that the polymerization of hydrophobic-rich phase is mostly triggered by the photosensitizer, CQ [7]. With the exception of Fluorescein sodium salt, the degree of conversion of the hydrophobic-rich phase was substantially high with all of the dyes. This result indicates that the dyes did not interfere with the activity of CQ.

A previous study showed that the average degree of conversion of hydrophobic-rich phase with CQ/EDMAB/DPIHP as the PI system was about 89% in the adhesive mixture with 33 wt% D₂O [7]. For CQ/EDMAB phototoinitiating system, the degree of conversion of the hydrophobic-rich phase was approximately 77% [7]. Although the wt% of EDMAB was greater, i.e. 0.5 wt% in the prior investigation, the average DC was close to the current study with dyes as the hydrophilic photosensitizer. This relationship was noted for all of the dyes except Fluorescein sodium salt. The average DC of the hydrophobic-rich phase with Victoria blue B, New Fuchsin and Methylene blue chloride was above 90%. This result indicates that these photosensitizers could facilitate the activity of CQ or participate in triggering the photopolymerization of this phase. With the exception of Fluorescein, the dyes did not impede the initiation process by CQ in the hydrophobic-rich phase.

Fluorescein sodium salt exhibited a very high molar extinction coefficient and moderate relative normalized PAE but poor DC for both hydrophobic- and hydrophilic-rich phases. The polymerization of the hydrophilic-rich phase is triggered mostly by the dyes since the concentration of CQ in this phase is low [7]. The poor DC of the hydrophilic-rich phase with Fluorescein sodium salt suggests that the excited state or the initiating radicals are not stable or effective at triggering the photopolymerization reaction in the hydrophilic-rich phase. Eosin Y disodium salt also led to a poor DC for the hydrophilic-rich phase, indicating that poor stability/effectiveness of the excited state/initiating radicals could account for its behavior since the molar extinction coefficient and relative normalized PAE were high. Methylene blue chloride possessed low molar extinction coefficient and relative normalized PAE. This indicated that its ability to absorb light at 480 nm and be promoted to the excited state to generate initiating radicals was poor. Therefore, poor generation of initiating radicals and/or poor stability/effectiveness of the initiating radicals could account for the low DC and polymerization rate observed for this dye.

With the exception of Eosin Y disodium salt, Fluorescein sodium salt and methylene blue chloride, all other dyes led to at least 47% or higher DC of the hydrophilic-rich phase. Results from a previous investigation that used CQ/EDMAB/DPIHP where each component was present in 0.5 wt% showed that the average DC was approximately 20% which was less than half of what was observed with most of the dyes [7]. For CQ/EDMAB photoinitiating components in dental adhesive mixture containing 33 wt% D₂O, the degree of conversion in the hydrophilic-rich phase was approximately 2.8% [7]. In some cases, such as the Bromophenol blue sodium salt, New Fuchsin, Victoria blue B and Rose Bengal sodium salt, the average DC exceeded 50% clearly indicating that incorporation of efficient hydrophilic photosensitizer can significantly improve the DC of the hydrophilic-rich phase. The higher DC may improve the mechanical properties as well as the degradation resistance of the hydrophilic-rich phase in phase-separated dental adhesives. Bromophenol blue sodium salt, Rosebengal sodium salt and Victoria blue B possessed low molar extinction coefficient and relative normalized PAE, indicating that the stability of the excited state or the stability/efficiency of the initiating reactive species could be high. Partitioning concentration of the dyes could also play a role in the photopolymerization of the hydrophilic-rich phase but since all the dyes were relatively hydrophilic they should be present in abundant quantity in this phase.

In our prior investigation using the properties discussed here, candidate hydrophilic photosensitizers were designed by computer-aided molecular design (CAMD) [24]. In that study, the newly designed candidate hydrophilic photosensitizers all contained iminium ions indicating that the functional group, C=NH ⁺₂ could play a vital role in generating effective initiating reactive species, and among the dyes studied here, New Fuchsin and Victoria blue B both possessed this functional group. The latter two yielded promising results in terms of DC of both phases.

This study indicates that with either Bromophenol blue sodium salt, Rosebengal sodium salt, Erythrosin B, New Fuchsin or Victoria blue B incorporated in the dental adhesive, the average DC of hydrophobic-rich phase is high and that of the hydrophilic-rich phase is improved. Therefore, these dyes are suitable photosensitizer candidates for dental adhesives that experience phase separation when photopolymerized in an over-wet environment. The toxicity and color are important issues that must be addressed in future studies.

The overall significance of the results is that appropriate hydrophilic dyes can improve the photopolymerization of the hydrophilic-rich phase of dental adhesive without compromising the performance of CQ in the hydrophobic-rich phase. Therefore, dyes are potential candidates for water-compatible dental adhesive photosensitizers. Further investigation of these dyes in dental adhesive formulations could advance the efforts of the dental research community to develop an efficient water-compatible photosensitizer.

Conclusion

The potential of hydrophilic dyes to serve as photosensitizers for the photopolymerization of dental adhesive phases was investigated. It was observed that the performance of the dye as the photosensitizer depends on a number of factors which include stability of the photosensitizer in its excited state, stability/efficiency of the initiating radicals, molar extinction coefficient and normalized photon absorption efficiency. In this study, it was observed that Fluorescein sodium salt, Eosin Y disodium salt and methylene blue chloride dyes were not ideal as photosensitizers for dental adhesive photopolymerization. Dyes such as Bromophenol blue sodium salt, New Fuchsin, Victoria blue B and Rosebengal sodium salt were more suitable candidates for dental adhesive photopolymerization, leading to substantial degree of conversion to the hydrophilic-rich phase without impairing the polymerization of the hydrophobic-rich phase.