Effect of the photoinitiator system on the polymerization of secondary methacrylamides of systematically varied structure for dental adhesive applications

Abstract

Objective.

The aim of this study was to investigate the influence of the photoinitiator system on the polymerization kinetics of methacrylamide-based monomers as alternatives to methacrylates in adhesives dental-based materials.

Methods.

In total, 16 groups were tested. Monofunctional monomers (2-hydroxyethyl methacrylate) – HEMA; (2-hydroxy-1-ethyl methacrylate) −2EMATE, (2-hydroxyethyl methacrylamide) – HEMAM; and (N-(1-hydroxybutan-2-yl) methacrylamide) −2EM; were combined with bifunctional monomers containing the same polymerizing moieties as the monofunctional counterparts (HEMA-BDI; 2EMATE-BDI; HEMAM-BDI; and 2EM-BDI) at 50/50M ratios. BHT was used as inhibitor (0.1 wt%) and the photoinitiators used were: CQ+EDMAB (0.2/0.8), BAPO (0.2), IVOCERIN (0.2), and DMPA (0.2), in wt%. The polymerization kinetics were monitored using Near-IR spectroscopy (~6165cm⁻¹) in real-time while the specimens were photoactivated with a mercury arc lamp (Acticure 2; 320–500nm, 300mW/cm²) for 5min, and maximum rate of polymerization (Rp_max, in %.s⁻–¹), degree of conversion at Rp_max (DC@Rp_max, in %), and the final degree of conversion (Final DC, in %) were calculated (n = 3). Initial viscosity was measured with an oscillating rheometer (n = 3). Data were analyzed using Two-way ANOVA for the polymerization kinetics and one-way ANOVA for the viscosity. Multiple comparisons were made using the Tukey’s test (α= 0.05).

Results.

There was statistically significant interaction between monomer and photoinitiator (p < 0.001). For the methacrylates groups, the highest Rp_max was observed for HEMA + DMPA and 2EMATE + BAPO. For methacrylamides groups, the highest Rp_max were observed for HEMAM and 2EM, both with DMPA. Final DC was higher for the methacrylate groups, in comparison with methacrylamide groups, independent of the photoinitiators. However, for the methacrylamide groups, the association with BAPO led to the lowest values of DC. In terms of DC@Rp_max, methacrylate-based systems showed significantly higher values than methacrylamide formulations. DMPA and Ivocerin led to higher values than CQ/EDMAB and BAPO in methacrylamide-based compounds. BAPO systems showed de lowest values for both HEMA and HEMAM formulations. For the viscosity (Pa.s), only 2EM had higher values (1.60±0.15) in comparison with all monomers. In conclusion, polymerization kinetics was affected by the photoinitiators for both monomers. Viscosity was significantly increased with the use of secondary methacrylamide.

1. Introduction

The composition of the vast majority of dental adhesive systems available on the market includes methacrylates, mainly because their chemical and mechanical properties have shown satisfactory results in this application [1]. The mono-methacrylate 2-Hydroxyethyl (HEMA) was incorporated in the composition in the late 1960’s due to its “water-chasing” capabilities, which in turn allowed for wet bonding to the dentin substrate using total etch techniques [1]. Until the introduction of more hydrophilic adhesive compositions, the bonding could only be reliably performed within enamel margins, so the introduction of bonding to dentin and the novel description of the hybrid layer were revolutionary [2]. However, since that time, in spite of active research in the area [3,4], very few products with monomers different from HEMA have been introduced in the market. The clinical studies that followed this products have demonstrated acceptable performance [5,6], in spite of lower in vitro results compared to other commercial products [7]. There is interest in pursuing alternatives to HEMA, since this monomer imparts significant water sorption and solubility to the adhesive, and contains ester bonds that make it prone to hydrolysis in the oral environment [4,8,9]. In addition, studies have shown that the degradation of methacrylates is accelerated at low pH, a common condition in some adhesive compositions and also when caries-forming biofilm is present mainly in gingival margins at proximal surfaces [10–13]. As a result, polymer properties can be compromised, resulting in premature bonding degradation, secondary decay or restoration de-bonding [14,15]. The degradation of the adhesive can also pose toxicity to pulp cells through the dentinal tubules or even to adjacent soft tissues in regions of the restoration close to gingival margins [16].

Since the maintenance of the integrity of the bonded interface is crucial to the longevity of the restoration, novel monomer alternatives such as acrylamides and methacrylamides have been proposed [4,17]. Methacrylamides and acrylamides have an amide group instead of an ester group, which makes them less prone to degradation in the oral environment. This is due to the fact that the electronegativity of the nitrogen is greater than the oxygen present in the methacrylate, which stabilizes the bond via resonance with the lone pair [18]. The energy required to break the bond between the carbon in the carbonyl and the nitrogen in the amide is much greater than the energy required to break the bond with the oxygen in the methacrylate [19]. Therefore, it can be envisioned that these monomers will be less prone to the hydrolysis via acid-base reactions in aqueous environment, and even through the great advantage from the bond preservation standpoint. Also it means that the vinyl bond attached to the methacrylamide is more stable, which leads to concerns over the reactivity of acrylamides and methacrylamides [17]. Previous study has demonstrated that steric constraints such as the ones observed in tertiary methacrylamides significantly reduce monomer reactivity [20].

One approach to circumvent the lower reactivity of methacrylamides is to use initiator systems with greater quantum yields than what is observed for the conventional camphorquinone/amine pair [21]. The efficiency of this Norrish-type photoinitiator system can be improved by the addition of electron acceptors such as diphenyl iodonium salts [22]. Studies with methacrylate monomers have shown significantly improved rates of polymerization and conversion with the addition of iodonium salt concentrations of around 1 mol% [22]. In addition, the use of cleavage-type photoinitiators such as phenylbis(2,4,6-trimethylbenzoyl) phosphine oxide (BAPO) and 2,2-dimethoxy-2-phenylacetophenone (DMPA) can also improve reactivity, since they rely on a single electron transfer event to generate twice the amount of radicals compared to CQ/amine [23]. In fact, their quantum yields of conversion are two to four-fold that of the CQ/amine pair [23]. More recently, photoinitiators based on germanium chemistry have also been described [24]. These are also cleavage-type initiators, and have been shown to be suitable for the polymerization ofmethacrylates and methacrylamides [25] . Bis-(4-methoxybenzoyl)diethyl germane (brand name Ivocerin, Ivoclar-Vivadent, Liechtenstein) and BAPO can be activated by visible light, at a slightly shorter wavelengths compared to CQ (405 and 470 nm, respectively), which also improves light penetration through the bulk of the material [26].

The present study had two distinct aims: firstly was to develop alternative monomers to BisGMA and HEMA, based on secondary methacrylamides, for future potential use as adhesive monomers. The monomers were prepared with systematically varied structures to allow for the assessment of the effect of steric hindrance on the kinetics of polymerization. Methacrylate analogs with similar steric constraints were also synthesized to allow for direct comparisons with the novel methacrylamides. Secondly was to evaluate the polymerization kinetics of these monomers when in combination with different types of initiators, including the conventional CQ/amine and three other examples of cleavage-type initiators. The null hypotheses of this study were: (1) the methacrylamide-based monomers would have the same polymerization kinetics behavior as the methacrylate-based monomers, and (2) the polymerization kinetics would not be influenced by the type of initiator used.

2. Material and methods

2.1. Monomer synthesis, material composition, and specimen preparation

Unless otherwise noted, all commercially-available materials were obtained from Sigma-Aldrich (Milwaukee, WI, USA). The synthesis procedures for the materials produced de novo are described in the supplemental materials. Four different monofunctional monomers were used: two commercially available (2-hydroxyethyl methacrylate (HEMA); and N-hydroxyethyl methacrylamide (HEMAM)) and two synthesized de novo (α-substituted methacrylate 2-hydroxy-1-ethyl methacrylate (2EMATE); and 2-hydroxy-1-ethyl methacrylamide (2EM)). The OH-bearing monofunctional monomers were then reacted with a difunctional isocyanate (1,3-bis (1-isocyanato-1-methylethylbenzene) – BDI) at 2:1mol ratio, resulting in difunctional urethanes. The structures of all monomers used in this study are shown in Fig. 1.

Fig. 1 –

Molecular structure of all monomers used in this study. The monofunctional monomers were HEMA (2-hydroxyethyl methacrylate), HEMAM (2-hydroxyethyl methacrylamide), 2EMATE (2-hydroxy-1-ethyl methacrylate) and 2EM (2-hydroxy-1-ethyl methacrylamide) – the latter two bearing one ethyl substituent on the alpha-carbon. The di-functional monomers were produced by reacting 2 mols of each of the monofunctional monomers with 1mol of BDI (1,3-bis (1-isocyanato-1-methylethylbenzene)). The monomethacrylate monomers were used in combination with dimethacrylate monomers of analogous polymerizable structure, at a mass ratio of 50–50.

Each monofunctional monomer was combined at 50:50 mass ratio with the corresponding difunctional urethane, resulting in four different monomer systems – HEMA+HEMA-BDI, 2EMATE+2EMATE-BDI, HEMAM+HEMAM-BDI and 2EM+2EM-BDI. For simplicity’s sake, throughout the manuscript, the groups containing a pair of monofunctional monomer plus its analogous difunctional monomer with the BDI core will be referred to by the name of the monofunctional monomer. For example, HEMA+HEMA-BDI will be referred to as HEMA. These were made photopolymerizable by the addition of one of the following four photoinitiator systems: .2/0.8wt% dl-camphoroquinone/ethyl 4-dimethylaminobenzoate (CQ/EDMAB - λ_max = 470nm), 0.2 wt% phenylbis(2,4,6-trimethylbenzoyl) phosphine oxide (BAPO- λ_max = 380nm); 0.2wt% 2,2-dimethoxyphenoxy acetophenone (DMPA- λ_max = 365 nm); or 0.2wt% bis-(4-methoxybenzoyl)diethyl germane (IVOCERIN™, donated by Ivoclar-Vivadent, Schaan, Liechstein- λ_max = 405 nm). 0.1 wt% 2,6-di-tert-butyl-4-methylphenol (BHT) was used as inhibitor. Materials were formulated under filtered yellow lights. Molecular structures for photoinitiator systems are shown in Fig. 2.

Fig. 2 –

Photoinitiator systems used in this study: CQ+EDMAB (dL-camphorquinone + Ethyl 4-dimethylaminobenzoate), BAPO (phenylbis(2,4,6-trimethylbenzoyl) phosphine oxide), DMPA (2,2-dimethoxyphenoxy acetophenone) and Ivocerin ™ (bis-(4-methoxybenzoyl)diethyl germane, kindly donated by Ivoclar-Vivadent).

2.2. Viscosity

The initial viscosity in Pa.s n= 3 wa 3 measured with an oscillating rheometer DH-R1, TA Instruments, New Castle, DE, USA. Approximately 20mg of each monomer mixture were placed between 20-mm diameter plates separated by a 300 μm gap, and tested at room temperature with strain rate ranging from 10 to 1000 Hz flow sweep mode.

2.3. Photopolymerization reaction kinetics and degree of conversion

Specimens were made using a rubber mold 10mm in diameter and 0.8mm thick, laminated between two glass slides. Polymerization kinetics (n = 3) was monitored using near-infrared (NIR) spectroscopy in real time. The vinyl overtones for the methacrylate and methacrylamide (6165 and 6135cm⁻¹, respectively) were followed with 2 scans per spectrum at 4cm⁻¹ resolution, resulting in 2Hz data acquisition rate. The test was recorded continuously during 300 s of irradiation with a mercury arc lamp filtered to 320–500nm (Acticure 4000, EXFO, Mississauga, ON, Canada). The tip of the light guide was positioned 1.5cm away from the surface of the specimen, so to deliver approximately 300mW/cm². Selected groups (HEMA and HEMAM-containing materials) were also evaluated using a mono-wave LED light source (Elipar, 3M-ESPE, St. Paul, MN, USA), centered at 470nm and delivering an irradiance of 840mW/cm² to the surface of the specimen for 60 s. Polymer conversion was calculated based on the area of the overtone peak before and during polymerization [27], and the rate of polymerization was calculated as the first derivative of the conversion vs. time curve.

2.4. Characterization of the light curing units

The emission spectra and irradiance of all light curing units (LCU) used in this study were obtained using a UV–vis detector (MARC Resin Calibrator, Blue Light Analytics Inc., Halifax, NS, Canada). The tip of the LCU was positioned parallel to the surface of the top detector, separated by 1.5cm to mimic the conditions of the polymerization kinetics on the IR chamber. Spectra were recorded after light stabilization, 5 s after power up. The irradiance for each individual wavelength band was calculated in mW/cm².

2.5. Characterization of the absorption spectra of each photosensitizer

Each initiator (0.008 g of EDMAB and 0.002 g of the others) was dissolved in 1ml of ethanol. These concentrations were selected to match the mass ratio of initiators in the materials as tested. Therefore, the results are reported in arbitrary units, rather than in molar absorptivity units. The solutions were placed in UV-transparent cuvettes and UV–vis spectra (300–540 nm) were obtained using a UV–vis spectrophotometer (Evolution 201, Thermo Fisher Scientific, Waltham, MA), with 1nm sampling interval.

2.6. Statistical analysis

The data was tested for normality (Anderson-Darling) and homoscedasticity (Bartlett/Levene). Degree of conversion, maximum rate of polymerization and conversion at rate maximum data were analyzed with two-way ANOVA (monomer type and photoinitiator system). The data for initial monomer viscosity failed the normality and homocedasticity tests, so one-way ANOVA was performed on ranks (Kruskall-Wallis). Multiple comparisons were done using Tukey’s test. The significance level of 95% was used for all tests.

3. Results

3.1. Viscosity

Results for viscosity are presented in Table 1. Both methacrylate-based formulations were statistically similar to each other (HEMA= 0.067 and 2 EMATE = 0.05 Pa.s), and showed statistically lower viscosity than the methacrylamide systems (HEMAM= 3.567 and 2 EM =40.0 Pa.s). The methacrylamide 2 EM monomer system presented the highest viscosity of all materials (p = 0.001; F = 874.7).

Table 1 –

Mean±standard deviation for viscosity (Pa s) for the monomer systems tested in this study (HEMA/HEMA-BDI, HEMAM/HEMAM-BDI, 2EMATE/2EMATE-BDI, 2EM/2EM-BDI). Values followed by the same superscript are statistically similar (α = 5%), according to one-way ANOVA on ranks (F = 874.7 and p < 0.001).

Monomer system	Viscosity (Pa s)
HEMA	0.067±0.006 c
2EMATE	0.050±0.007 c
HEMAM	3.567±0.303 b
2EM	40.0±12.578 a

3.2. Light source and photoinitiator spectra

The emission spectra of the two light curing units used in this study are shown in Fig. 3. The Acticure presents three distinct emission peaks, centered at 365, 405 and 435 nm, with approximately 157, 53 and 90 mW/cm² absolute irradiance, respectively. These values were calculated by integrating the area under those peaks. The Elipar is a mono-wave LED, with emission spectra centered at 450nm and irradiance of 840mW/cm². Irradiance for each individual peak is shown in Table 2.

Fig. 3 –

Left – Emission spectra for the mercury arc lamp (Acticure) and the commercial LED (Elipar) light sources, obtained with the MARC resin calibrator (values expressed in mW/cm²/nm). Right – Absorption spectra for each photosensitizer utilized in this study (CQ: camphorquinone; DMPA: 2,2-dimethoxyphenhoxy acetophenone; BAPO: phenylbis(2,4,6-trimethylbenzoyl) phosphine oxide and Ivocerin ™: bis-(4-methoxybenzoyl)diethyl germane.

Table 2 –

Top — Maximum emission wavelength (nm), range of emission (nm) and irradiance at range (mW/cm²) for the mercury arc lamp (Acticure) and the commercial LED (Elipar) light sources. Bottom — Maximum absorption wavelength (nm), range of absorption (nm) and potential useful irradiance from the Acticure (calculated as the sum of irradiances at the wavelengths overlapping the absorption ranges of the initiators).

Light source characterization
Light source	Maximum emission wavelength (nm)	Range of emission (nm)	Irradiance at range (mW/cm2)
Acticure	365	345–380	157	300
	405	380–410	53
	435	430–460	90
Elipar	450	430–480	840
Photosensitizer characterization
Initiator	Lambda max (nm)	Range of absorption (nm)		Useful irradiance from Acticure (mW/cm2)
CQ	471	400–500		143
EDMAB	320	300–450		300
DMPA	365	300–390		157
BAPO	380	300–435		236*
Ivocerin ™	405	300–340/340–460		300

^*The Acticure provides 26 mW/cm² up to 435 nm.

The absorption spectra of the four different photosensitizer molecules are shown in Fig. 3. Absorption maxima and absorption ranges for each photoinitiator system is shown in Table 2.

3.3. Photopolymerization reaction kinetics and degree of conversion

Results for degree of conversion and polymerization kinetics obtained with the Acticure are presented in Table 3 and Figs. 4a/b and 5. In terms of degree of conversion (final DC), two-way ANOVA showed that both factors (monomer and photoinitiator) as well as the interaction between them were statistically significant (p < 0.001). The methacrylamide formulations presented lower conversion than the methacrylate counterparts for all photoinitiator systems tested. Within each polymerizing functionality (methacrylate or methacrylamide), the presence of the ethyl substituent on the alpha carbon led to statistically lower conversion (except for BAPO and Ivocerin with the methacrylamides). The type of initiator system did not affect the conversion outcomes for the alpha-substituted methacrylate. For all the other monomers, BAPO consistently led to the lowest values of conversion.

Table 3 –

Mean±standard deviation for degree of conversion (Final DC, %), maximum rate of polymerization (RP_MAX, %s⁻¹) and degree of conversion at rate maximum (DC@RP_MAX, %) for the monomer systems tested in this study (HEMA/HEMA-BDI, HEMAM/HEMAM-BDI, 2EMATE/2EMATE-BDI, 2EM/2 EM-BDI), as a function of the photoinitiator system (CQ/EDMAB, DMPA, BAPO and Ivocerin™). Values followed by the same uppercase letter on the same column or lowercase letter on the same row are statistically similar, according to two-way ANOVA (α = 5%).

Photoinitiator systems	Final DC (%)				RPMAX (% s−1)				DC @ RPMAX (%)
	Methacrylates		Methacrylamides		Methacrylates		Methacrylamides		Methacrylates		Methacrylamides
	HEMA	2EMATE	HEMAM	2EM	HEMA	2 EMATE	HEMAM	2EM	HEMA	2 EMATE	HEMAM	2EM
CQ/EDMAB	86.2±2.2 ABb	94.6± 3.3 Aa	63.8±1.6 Ac	56.1±1.96 Ad	2.1±0.3 Da	1.2±0.2 Dab	0.6±0.1 Cb	1.8±0.2 Cab	42.1±2.4 Aa	48.4±5.3 Aa	1.0±0.1 Cc	8.9±1.5 Bb
DMPA	89.5±2.5 Ab	93.0±0.6 Aa	65.2±3.0 Ac	59.8±4.4 Ad	14.0±0.6 Aa	4.3±0.2 Cb	3.6±0.2 Ab	3.5±0.6 Bb	41.9±2.9 Aa	43.9±1.1 Aa	8.1±1.4 Ac	11.8±0.8 Ab
BAPO	81.1±1.7 Bb	88.3±0.2 Aa	41.7±2.3 Cc	43.8±5.4 Bc	10.6±1.6 Cb	15.1±0.6 Aa	2.2±0.1 Bc	0.9±0.5 Cc	25.5±6.8 Ba	30.2±3.4 Ca	3.9±0.4 Bb	4.0±3.5 Bb
Ivocerin ™	81.4±1.7 ABb	90.9±1.8 Aa	52.2±4.3 Bc	58.7±3.2 Ac	12.4±0.3 Ba	8.9±0.4 Bb	3.6±0.2 Ad	6.7±1.2 Ac	38.0±1.7 Aa	38.7±0.6 Ba	7.4±0.9 Ac	13.0±2.3 Ab
p (monomer)	0.001				0.001				0.001
p (photoinit.)	0.001				0.001				0.001
p (interaction)	0.001				0.001				0.001

Fig. 4 –

(a) Polymerization kinetic profiles (degree of conversion as a function of time) for the methacrylate- (HEMA and 2 EMATE) and methacrylamide-terminated (HEMAM and 2EM) monomers used in this study. Photoinitiator systems used in this study: CQ+EDMAB (dL-camphorquinone + Ethyl 4-dimethylaminobenzoate), BAPO (phenylbis(2,4,6-trimethylbenzoyl) phosphine oxide), DMPA (2,2-dimethoxyphenoxy acetophenone) and Ivocerin ™ (bis-(4-methoxybenzoyl)diethyl germane, kindly donated by Ivoclar-Vivadent). Photopolymerization kinetics was followed in near-IR in real time, while the materials were exposed to 300 mW/cm² (320–500nm) for 5min. (b) Polymerization kinetic profiles (degree of conversion as a function of time) for one methacrylate- (HEMA) and one methacrylamide-terminated (HEMAM) combined with CQ+EDMAB (dL-camphorquinone + Ethyl 4-dimethylaminobenzoate) as the photoinitiator system. Photopolymerization kinetics was followed in near-IR in real time, while the materials were exposed to 300mW/cm² (320–500nm) for 5min with a mercury arc lamp (Acticure) or to 840mW/cm² (440–480nm) for 60 s with a mono-wave LED light source (Elipar, 3M-ESPE).

Fig. 5 –

Polymerization kinetic profiles (rate of polymerization as a function of conversion) for the methacrylate- (HEMA and 2EMATE) and methacrylamide-terminated (HEMAM and 2EM) monomers used in this study. Photoinitiator systems used in this study: CQ+EDMAB (dL-camphorquinone + Ethyl 4-dimethylaminobenzoate), BAPO (phenylbis(2,4,6-trimethylbenzoyl) phosphine oxide), DMPA (2,2-dimethoxyphenoxy acetophenone) and Ivocerin ™ (bis-(4-methoxybenzoyl)diethyl germane, kindly donated by Ivoclar-Vivadent). Photopolymerization kinetics was followed in near-IR in real time, while the materials were exposed to 300mW/cm² (320–500nm) for 5min.

In terms of maximum rate of polymerization (RP_MAX), two-way ANOVA showed that both factors (monomer and photoinitiator) as well as the interaction between them were statistically significant (p < 0.001). In general, methacrylamide monomers presented lower rates of polymerization compared to the analogous methacrylate monomers. In general, CQ/amine systems led to the lowest rates of polymerization, but the behaviour of the other three initiators varied widely with the monomer system. For HEMA, the rate of polymerization increased in the order CQ< BAPO< Ivocerin <DMPA. For 2EMATE, the ranking was: CQ<DMPA< Ivocerin < BAPO. For HEMAM, the ranking was: CQ< BAPO<DMPA= Ivocerin. For 2EM, the ranking was CQ= BAPO<DMPA< Ivocerin. In general, the degree of conversion at Rp_max, used in this study as an indication of the conversion at the onset of vitrification [28] was statistically lower for the methacrylamides compared to the methacrylate counterparts (Table 3). For BAPO, Rp_max was observed earlier in conversion compared to all other adhesives.

For the two materials photoactivated with the commercial LED light source, the final conversion values were 73.3±0.7 and 17.7±3.3 % for HEMA- and HEMAM-based materials, respectively, and the maximum rate of reaction values were 2.2±0.1 and 1.0±0.1%s⁻¹ for HEMA- and HEMAM-based materials, respectively (Fig. 4b).

4. Discussion

The use of methacrylamide monomers as replacements for methacrylate counterparts in dental material compositions was evaluated in this study. The rationale is the absence of hydrolytically unstable ester bonds in methacrylamides, which has shown potential for decreased degradation in aqueous environments and in the presence of enzymes [19,29]. Methacrylamides have also been demonstrated to be less reactive when polymerized via radical-mediated methods [17], and therefore, the objective of this study was to evaluate the effect of different photoactivated radical initiator systems on the polymerization kinetics of methacrylamides. The study design included monomers with alkyl substituents on the alpha carbon, which has been shown to increase reactivity [30] and decrease the likelihood for hydrolytic degradation [31]. In general, based on the results obtained, the null hypotheses of the study were rejected, as methacrylates and methacrylamides presented distinct behaviors, which were influenced by the initiator used.

In general, both methacrylamide monomers polymerized at a slower rate and to a lesser extent compared to the methacrylate counterparts. The presence of the lone pair on the nitrogen provides resonance stabilization to the methacrylamide bond, and while this is advantageous from the standpoint of potentially protecting it from hydrolytic degradation, it also causes the vinyl bond to be more stable and less susceptible to propagation in the presence of free radicals [20]. In addition, the methacrylamides produced higher viscosity mixtures, which affects the initial kinetics of polymerization [28]. In higher viscosity systems, diffusional limitations to termination, and subsequently, to propagation, take place earlier in conversion, as evidenced by the much lower conversion at Rp_max, used in this study as a proxy for the onset of vitrification [28]. This helps explain both the lower rate and lower limiting conversion achieved by methacrylamides in this study.

The effect of the presence of alkyl substituents was less pronounced, but statistical differences in final conversion were found within the same polymerizable functionality (methacrylate or methacrylamide) for a given initiator system. In general, conversion was lower for all alpha-substituted monomers, with only two exceptions (methacrylamides polymerized with BAPO and Ivocerin), when the conversion values were statistically similar for HEMAM and 2EM, within the same initiator. The decreased conversion is explained by steric hindrance around the vinyl functionality [30]. That same stabilization is hypothesized to increase stability of the amide and ester bonds when the resulting polymeric network is presented with hydrolytic challenges, which will be the subject of a separate study. In summary, the two mechanisms that increase the hydrolytic stability (lone pair on the nitrogen on the amide bond and the alpha substitution on either monomer type) also lead to decreased monomer reactivity (lower polymerization rates). The effects of the substituent on the rate of polymerization were more complex, and dependent on the initiator system. For CQ/EDMAB, all rates of polymerization were very low, ranging from 0.6 to 2.1% s⁻¹. HEMAM presented the lowest result, but it was only statistically lower than HEMA, showing that even at low rates, it was possible to notice the effect of the stabilization of the vinyl by the resonance provided by the lone pair on the nitrogen of the amide bond. At those lower rate levels, the presence of the alkyl substitution did not affect the results. All other initiators led to much higher rates of polymerization, and in general, the methacrylamides presented lower rates than the methacrylates, as already mentioned. For the methacrylates, DMPA and Ivocerin led to higher rates in the non-substituted monomer (HEMA), while BAPO led to lower rates for HEMA. For the methacrylamides, there was either no difference between HEMAM (non-substituted) and 2EM (bearing the alkyl side chain), as was the case for DMPA and BAPO, or the rate was higher for 2EM, as was the case for Ivocerin.

The differences among initiator systems were due to two well-explored factors: the quantum yield of conversion of each molecule and the irradiance at the wavelength of maximum absorption (λ_max) [23,28]. In terms of quantum yield of conversion, CQ has been reported to present the lowest initiator efficiency by each packet of photons absorbed. This is in part due to the fact that CQ is a Norrish-type initiator, whose radical-generating events are dependent both on the bleaching of CQ and on the successful abstraction of a proton from a co-initiator [23]. The band of absorption for CQ is broad, ranging from 400 to 500nm. At the λ_max, centered at 470 nm, the absorption efficiency is maximized, and outside of 470 nm, excitation of the initiator is still possible, albeit less efficient. The light source used in this study was a mercury arc lamp, with three distinct peaks of irradiance, centered at 365, 405 and 435 nm, each with an absolute irradiance of approximately 157, 53 and 90mW/cm², respectively. Of those peaks, CQ overlaps with the ones centered at 405 and 435 nm, and therefore, received useful irradiance totaling approximately 143mW/cm². However, the irradiance at 470nm was very low. In other words, on top of being the least efficient initiator, the mercury arc lamp did not efficiently overlap CQ’s λ_max, both of which explain the low rate of polymerization. The degree of conversion was still greater than 50% for all monomers, due to the extended exposure time (5min), resulting in 50 J/cm² radiant exposure. Since the low rates of polymerization obtained did not allow for the assessment of the potential differences among the monomers, HEMA and HEMAM were selected to be tested using a commercial dental mono-wave LED light source, with emission spectrum centered at 450 nm and effective irradiance of 840 mW/cm². For HEMA, the rate of polymerization increased as expected, and the final conversion was lower to what was obtained with the mercury arc lamp, also as expected due to the shorter exposure time and consequently lower radiant exposure. Interestingly and unexpectedly, for HEMAM, both the rate and the conversion significantly decreased with the use of the mono-wave LED compared to what was obtained with the mercury arc. This was unexpected because of the direct and exponential relationship between rate of polymerization and irradiance [28], and especially because the wavelengths delivered by the LED light more closely match the absorption spectrum of the photosensitizer CQ [32]. One possible explanation for the higher conversion of HEMAM materials when polymerized with the light source extending to the UV is the fact that the amine co-initiator also has absorption in that range (300–450nm). Previous reports have demonstrated that when exposed to light at those wavelengths, organic bases (electron donors in this case) can potentially aid in direct radical generation [32]. Therefore, the presence of the amine in Acticure polymerizations helped improve the conversion, while for the mono-wave LED light source, centered at 470nm, the same was not observed. Another possible explanation relates to the high initial viscosity of HEMAM. This initially hampered the diffusion of polymerizing species, and as a consequence, delayed the onset of diffusion-controlled termination and propagation, keeping the overall rate of polymerization low. For the lower intensity combined with longer exposure (Acticure polymerizations), the final conversion was not affected because the final radiant exposure was much higher. However, for the shorter exposure at higher irradiance (Elipar polymerization), it is likely that most of the radicals generated decayed before any meaningful chain growth could take place, keeping both the rate and the final conversion low.

For all other initiators, the effective irradiance (within the absorption range of each molecule) ranged from 150mW/cm², and that contributed to higher rates of polymerization, as expected [28]. In addition, the increase in comparison with CQ is due to the fact that BAPO, DMPA and Ivocerin are cleavage type initiators, with higher quantum yields of conversion [21]. For those initiators, each photon absorbed has the potential to generate 2 radicals per molecule of initiator, independent of the interaction with a co-initiator. In contrast, for camphorquinone/amine pairs, each photon absorbed can potentially take one CQ molecule to a triplex state, which then has a very short time to interact with the amine to generate one radical [23]. Therefore, the process is expected to be much more efficient for cleavage-type initiators, as long as they are exposed to wavelengths at or close to their lambda max [23]. This was indeed observed for almost all monomer/initiator pairs, with the exception of the 2EMATE combined with DMPA and 2EM combined with BAPO - therefore, both exceptions involve an alpha-substituted monomer. These aspects are likely explained by a complex interplay between the size of the formed radical for each initiator and the steric hindrance around the vinyl radical for the acrylate/acrylamide monomer, as well as possibly due to other electronic effects [30]. Within the limitations of this study, it is not possible to confidently determine which factor was prevalent in explaining these exceptions. However, for the non-substituted monomers, the rate of polymerization was consistently higher for the cleavage-type initiators compared to CQ/amine, as already mentioned.

5. Conclusion

In conclusion, this exploratory study demonstrated that the polymerization of methacrylamides is less efficient and progresses to a lesser extent compared to methacrylate analogues, regardless of the presence of alkyl substitutions on the alpha carbon. Cleavage-type initiators led to faster rates of polymerization compared to CQ/amine, with two exceptions, but that did not translate into higher conversion for any of the monomers tested. Conversion values with CQ/amine, a clinically-relevant initiator system, were above 55% for all materials. Future studies will evaluate monomer and resulting polymer stability to determine whether methacrylamides are a suitable alternative for dental materials applications.