Effect of side-group methylation on the performance of methacrylamides and methacrylates for dentin hybridization

Abstract

The stability of the bond between polymeric adhesives to mineralized substrates is crucial in many biomedical applications. The objective of this study was to determine the effect of methyl substitution at the α- and β-carbons on the kinetics of polymerization, monomer hydrolytic stability, and long-term bond strength to dentin for methacrylamide- and methacrylate-based crosslinked networks for dental adhesive applications.

Methods.

Secondary methacrylamides (α-CH₃ substituted = 1-methyl HEMAM, β-CH₃ substituted = 2-methyl HEMAM, and unsubstituted = HEMAM) and OH-terminated methacrylates (α- and β-CH₃ mixture = 1-methyl HEMA and 2-methyl HEMA, and unsubstituted = HEMA) were copolymerized with urethane dimethacrylate. The kinetics of photopolymerization were followed in real-time using near-IR spectroscopy. Monomer hydrolysis kinetics were followed by NMR spectroscopy in water at pH 1 over 30 days. Solvated adhesives (40 vol% ethanol) were used to bond composite to dentin and microtensile bond strength (μTBS) measured after 24 h and 6 months storage in water at 37 °C.

Results.

The rate of polymerization increased in the following order: OH-terminated methacrylates ≥ methacrylamides > NH₂-terminated methacrylates, with minimal effect of the substitution. Final conversion ranged between 79% for 1-methyl AEMA and 94% for HEMA. 1-methyl-HEMAM showed the highest and most stable μTBS, while HEMA showed a 37% reduction after six months All groups showed measurable degradation after up to 4 days in pH 1, with the methacrylamides showing less degradation than the methacrylates. Additionally, transesterification products were observed in the methacrylamide groups.

Significance.

Amide monomers were significantly more stable to hydrolysis than the analogous methacrylates. The addition of a α- or β-CH₃ groups increased the rate of hydrolysis, with the magnitude of the effect tracking with the expected base-catalyzed hydrolysis of esters or amides, but opposite in influence. The α-CH₃ substituted secondary methacrylamide, 1-methyl HEMAM, showed the most stable adhesive interface. A side reaction was observed with transesterification of the monomers studied under ambient conditions, which was not expected under the relatively mild conditions used here, which warrants further investigation.

1. Introduction

Dental adhesive interfaces with reduced susceptibility to degradation could lead to dental restorations with extended clinical lifetimes. Degradation is the result of two factors: (1) collagen degradation by endogenous proteases [1,2], and (2) polymer hydrolysis. The hydrolysis of dental adhesives – specifically the ester functionality within the polymer – is catalyzed by acid (low pH) as well as bacterial/salivary esterases [3–6]. A host of strategies have been suggested to achieve a more stable dental adhesive interface to promote longer clinical lifetimes, including the use of compounds shown to reduce the activity of the metalloproteinases and cysteine cathepsins responsible for the proteolysis of collagen fibrils [7–11]. One common example is chlorhexidine digluconate, but this compound shows cytotoxicity, high water solubility, low substantivity and only short-term effects [12–16]. Another strategy relies on the dentin biomodification by flavonoid-type polyphenolics (such as proanthocyanidins, quercetin, and curcumin) or 1-Ethyl-3-(3-dimethylamino-propyl) carbodiimide (EDC), which may function as collagen cross-linking agents and inhibit the activity of the endopeptidases [13,17–21]. The undesirable effects of staining dentin along with the uncertain longevity of the benefits make the clinical feasibility of this approach questionable.

Another recent strategy to improve dental adhesive performance is the use of hydrolysis-resistant compounds. After more than 60 years using purely methacrylate-based compositions, researchers have concentrated their efforts on the inclusion of more hydrolytically stable compounds as co-monomers for dental adhesives. Methacrylamide-methacrylate blends have emerged as materials with significantly improved properties [22–25]. The replacement of the oxygen atom in the ester group of hydroxyethyl-methacrylate (HEMA) by an NH (amide group) resulted in formulations with markedly higher bond stability, which-was mainly attributed to the replacement of ester bonds with amide bonds, making the polymers more resistant to hydrolysis [26]. Methacrylamide-based polymer performance is highly dependent on the chemical structure localized about the amide functional group. For example, α-substituted secondary methacrylamides (40% weight blend in methacrylates) lead to materials with more stable mechanical adhesion after 6 months than unsubstituted secondary methacrylamides [22]. However, the degree to which the side-group substituents affect polymerization kinetics and subsequent polymer stability (hydrolysis kinetics) is not well understood for methacrylamides, especially in comparison with methacrylates. Further, the effects that α- and β-carbon substituents have on hydrolysis in materials remain unexplored.

Based on homogenous chemical model systems, the maximum expected effect of the substitution on the hydrolysis rates of amides is ten times slower (i.e., ten times longer lifetime) compared to esters [27], and a maximum factor of ~2.5 slower hydrolysis is expected for sterically imposing side-chains [28] – see Scheme 1 in the supplemental materials. This sterically derived hydrolytic stability factor is approximately the same for both ester and amide hydrolysis. Thus, additional hydrolytic stability could be imparted to the material both at the ester and the amide groups by the synthetic inclusion of chemically inert methyl groups.

One important point to consider is that the same factors that make amide bonds more stable than acrylate bonds also affect polymerization rate (Scheme 2 in supplemental materials). Neat methacrylamides tend to have slower polymerization kinetics compared to methacrylates [29] because the vinyl radical that is generated during the polymerization is better resonance stabilized by the amide nitrogen in a non-propagating form (Scheme 2) than an analogous ester oxygen. The difference in resonance forms makes the vinyl amide radical more stable and subsequently, less reactive than the less stable and more reactive methacrylate radical. In addition, (meth)acrylamides are more prone to water sorption, and therefore, are expected to lead to a reduction in bulk mechanical properties after exposure to the oral environment [30]. To overcome these limitations, methacrylamide-methacrylate co-polymerizations can lead to significant gains in dentin bond strength stability, while only marginally affecting polymerization rate [22].

It is currently unclear how methylation as mentioned above affects the polymerization kinetics and the final material properties. The aim of this study was to explore the effect of α- and β-CH₃ side-group substituents in methacrylamide-methacrylate-based dental adhesives on the polymerization kinetics, monomer hydrolysis kinetics, and microtensile bond strength to dentin. Our hypotheses were: (1) the side-group substitutions will affect the degradation profile of the monomers, and (2) the side-group substitutions will not affect the degree of conversion.

2. Materials and methods

2.1. Tested Co-monomers

The commercially-available monomers used in this study were purchased from Sigma Aldrich (Milwaukee, WI, USA) at 95% or greater purity, and used as received: 2-hydroxyethyl methacrylate – HEMA, 2-hydroxyethyl methacrylamide – HEMAM, 2-hydroxy-2-methylethyl methacrylamide - 2-methyl HEMAM. Hydroxypropyl methacrylate was obtained as a mixture of isomers consisting of the α-substituted 2-hydroxy-1-methylethyl methacrylate - 1-methyl HEMA, and β-substituted 2-hydroxy-2-methylethyl methacrylamide – 2-methyl HEMA. The composition of the mixture was determined by ¹H NMR spectroscopy to be 72% 2-methyl HEMA and 28% 1-methyl HEMA, consistent with the distributer’s analysis. The hydroxypropyl methacrylate isomeric mixture was used as received due to facile isomerization equilibrium (discussed later). 2-hydroxy-1-methylethyl methacrylamide – 1-methyl HEMAM was synthesized de novo (see supplementary information).

The NH₂-terminated methacrylate used in this study (2-aminoethyl methacrylate (AEMA) was obtained as a hydrochloride salt. The α-substituted, 2-amino-1-methylethyl methacrylate – 1-methyl AEMA, and the β-substituted, 2-amino-2-methylethyl methacrylate - 2-methyl AEMA, amino-terminated monomers were synthesized as hydrochloride salts following procedures adapted from a previous report [31]. Detailed information can be found in the supplementary information. In order to prevent the confounding variable of the presence of a salt in the polymer formulations, the hydrochloride salts were neutralized prior to formulation of the tested materials (see supplementary information) (see Fig. 1).

Fig. 1 –

Monomers copolymerized with UDMA as alternative options for dental adhesives.

2.2. Formulations and photocuring conditions

The co-monomers were mixed at 40 wt% with UDMA (urethane dimethacrylate, purchased from ESSTECH, Essington, PA, USA). The mixtures were made polymerizable by the addition of 0.2 wt% DMPA (2,2-dimethoxy-2-phenyl acetophenone) and 0.4 wt% DPI-PF₆ (diphenyliodonium hexafluorophosphate). BHT (butylated hydroxytoluene) was incorporated at 0.1 wt% into the formulations as an inhibitor. All photocuring procedures were carried out with a mercury arc lamp (Acticure 4000, 320–500 nm filtered) at 630 mW/cm² measured directly on the surface of the samples using a power meter (PowerMax 5200, Molectron Detector Inc., Portland, OR, USA).

2.3. Kinetics of polymerization

Polymerization kinetics were followed with real time by near-IR spectroscopy. Discs of 6 mm diameter and 0.8 mm thickness were sandwiched between glass slides and photoactivated for 300 s with the tip of the light guide placed 4 cm away and perpendicular to the glass surface, delivering 630 mW/cm² at the sample surface (n = 3). Spectra were collected for 330 s, with 2 scans per spectrum at 4 cm⁻¹ resolution. The light was kept on for the duration of the experiment to provide isothermal conditions, and avoid overestimation of conversion due to potential IR pathlength reduction (had the light been turned off during the experiment, causing shrinkage of the specimen). The followed peaks were 6165 and 6135 cm⁻¹ for methacrylates and methacrylamides, respectively. The maximum rate of polymerization (RP_MAX) was calculated as the first derivative of the degree of conversion vs. time curve, and the final degree of conversion (Final DC) was based on the change in area of the vinyl overtone peaks. The degree of conversion at the maximum rate of polymerization (DC at RP_MAX) was used as a proxy for the onset of vitrification.

Since the β-substituted secondary methacrylamide 2-methyl AEMA was not soluble in the organic matrix at room temperature, the mixture was heated on a hot plate to 50 °C and the kinetics tested immediately at the same conditions described above. For an appropriate comparison, the methacrylamide and methacrylate controls – HEMAM and HEMA, were also tested at 50°C, as controls. Formulations that did not cure or cured very slowly were not subjected to dentin microtensile bond strength or monomer hydrolysis kinetics.

2.4. Dentin microtensile bond strength

Sound human dentin from extracted third molars was used as the substrate for microtensile bond strength (μTBS) (project approved by Oregon Health & Science University – IRB #00012056). Ethanol was added at 40 vol% to the selected monomer compositions. Briefly, enamel was removed to expose a flat surface of mid-coronal dentin. A smear layer was created on this surface using 600 grit sandpaper for 30 s followed by etching with 35% phosphoric acid (3M ESPE, St. Paul, MN, USA) for 15 s and rinsing for 10 s. After blotting the surface dry, two consecutive coats of the experimental adhesives were applied and solvent evaporated using a gentle air spray for 10 s. The second coat of the adhesive was photoactivated for 60 s with the light guide 4 cm away from the surface delivering 630 mW/cm². The restorative procedures consisted of a composite block (Filtek Supreme, A2 – 3M ESPE, St. Paul, MN, USA) built in 2 increments of 2 mm thickness, photoactivated for 30 s each at 1100 mW/cm² (Elipar DeepCure-S, 3M ESPE, St. Paul, MN, USA). Adper Single Bond (composed of dimethacrylates and HEMA, 3M ESPE, St. Paul, MN, USA) was tested as commercial control and used as described previously except for the photoactivation of the second coat, carried out for 20 s using Elipar (Mono-wave LED, 3M-ESPE) at 1100 mW/cm² placed directly over the dentin surface (n = 6). All experimental adhesives, including the experimental control, were photoactivated with light parameters that were optimized for the type and concentration of initiator (DMPA/DPI = PF6, λ_max = 365 nm), as determined in a preliminary study. The commercial control was included as an external benchmark, and photoactivated according to manufacturer’s instructions using a commercially available light source. After 24 h, the teeth were cut on a slow speed diamond saw to obtain 1 mm² transversal section area sticks, which were stored for an additional 24 h or 6 months in distilled water at 37 °C. At the end of the storage time, sticks were fixed to a custom-made metal jig (Odeme Equipment, Luzerna, SC, Brazil) using super glue (Zapit, Dental Ventures of America Inc, Corona, CA, US), attached to a universal testing machine (Criterion MTS, Eden Prairie, MN, USA), and tested until failure at 0.5 mm/min.

2.5. Monomer hydrolysis kinetics

An aqueous solution (H₂O, pH = 1) was prepared using HPLC grade water and adjusted using 1.0 M HCl. A 50 mM solution of each monomer (n = 3) was prepared using 1.0 mL of the acidic aqueous solution. A capillary tube was filled with a 50 mM solution of tetramethylammonium bromide dissolved in D₂O and flame sealed. The capillary tube was placed at the bottom of the NMR tube to allow the locking of the magnet on the instrument onto the deuterium of the inner-tube with the ammonium acting as an internal standard. ¹H NMR spectra were obtained using a water suppression by excitation sculpting experiment [32]. After the initial reading, the NMR tubes were flame sealed and incubated at 37°C. At 4, 9, and 17 days (4, 7, 12, and 19 days for HEMA), the samples were removed from incubation to obtain water suppressed ¹H NMR spectra. To determine the amount of monomer degradation, spectra were first aligned using the ITSD singlet peak and then an integration region unique to vinyl protons of the monomer, methacrylic acid, and transesterification product (if applicable) were compared to calculate the percentage of intact monomer. To determine the rate constant and half-lives for the hydrolysis of each monomer, data were fit to a pseudo-first order exponential decay function (Eq. (1)) where A is the percentage of intact monomer, A₀ is the initial monomer purity (>99%), and t is time in days.

$$A=A_0{e}^{-kt}$$ (1)

2.6. Statistical analysis

Data were statistically analyzed with one-way ANOVA and Tukey’s test (α = 0.05), after normality and homoscedasticity tests. Student’s T-test was carried out to analyze the effect of the storage time on the μTBS (α = 0.05).

3. Results

3.1. Polymerization kinetics

Kinetics of polymerization results for the groups tested at room temperature are shown in Fig. 2 and Table 1. The OH-terminated methacrylate, HEMA α-,β-CH₃ mixture showed the highest RP_MAX, 20.2%·s⁻¹, and the NH₂-terminated 1-methyl and 2-methyl AEMA the lowest at 5.3%·s⁻¹. The methacrylamides, HEMAM and 1-methyl HEMAM, presented intermediate RP_MAX values: 13.0 and 13.3%.s⁻¹, respectively. The DC at RP_MAX ranged between 35.1% and 16.7%. In general, OH-terminated methacrylates showed the highest values (35.1% and 31.6% for HEMA and HEMA α-,β-CH₃ mixture, respectively), followed by the methacrylamides (22.2% and 24.3% for HEMAM and 1-methyl HEMAM, respectively). The NH₂-terminated 1-methyl and 2-methyl AEMA presented the lowest values: 15.7% and 16.7%, respectively. Final DC values ranged between 94.0% and 79.0%, with HEMA and 2-methyl AEMA presenting the highest and the lowest values, respectively. All other groups were statistically similar to each other. The non-substituted NH₂-terminated AEMA did not polymerize.

Fig. 2 –

Degree of conversion (%) as a function of time (s), and polymerization rate (%·s⁻¹) as a function of the degree of conversion (%) for all tested copolymers containing UDMA as base monomer. The kinetics of polymerization was followed at room temperaure in real time by near-IR spectroscopy during photopolymerization for 300 s at 630 mW/cm².

Table 1 –

Average (standard deviation) for maximum rate of polymerization (RP_MAX – %·s⁻¹), degree of conversion at the maximum rate of polymerization (DC at (RP_MAX – %) and final degree of conversion (Final DC – %) for all tested copolymers. Values followed by different letters indicate significant differences among the tested groups (p < 0.05).

Groups	RPMax (%·s−1)	DC at RPMax (%)	Final DC (%)
HEMA	14.2 (0.16) B	35.1 (3.46) A	94.0 (0.77) A
HEMA α-β mixture	20.2 (0.53) A	31.6 (1.72) A	91.5 (2.95) AB
HEMAM	13.0 (0.71) B	22.2 (2.52) B	89.8 (1.10) AB
1-methyl HEMAM	13.3 (0.48) B	24.2 (0.84) B	86.6 (3.38) B
2-methyl HEMAM	Not soluble at room temperature
AEMA	NA	NA	NA
1-methyl AEMA	5.6 (0.21) C	15.7 (1.83) C	79.0 (1.51) C
2-methyl AEMA	5.3 (0.49) C	16.7 (1.11) C	86.2 (1.32) B
p	<0.0001	<0.0001	<0.0001

Kinetics of polymerization results at 50 °C for HEMA, HEMAM and 2-methyl HEMAM are presented in Fig. 3 and Table 2. While there was no statistical difference among the tested groups in terms of RP_MAX (values ranged between 16.1 and 18.1%·s⁻¹), the methacrylate control HEMA showed the highest values, 41.7% and 92.0%, for DC at RP_MAX and final DC, respectively, and the non-substituted HEMAM and the β-substituted 2-methyl HEMAM methacrylamides presented the lowest values of DC at RP_MAX, 29.6% and 31.3%, respectively, and final DC, 86.3% and 86.6%, respectively.

Fig. 3 –

Degree of conversion (%) as a function of the time (s) and polymerization rate (%·s⁻¹) as a function of the degree of conversion (%) for the OH-terminated methacrylate control HEMA, and the non-substituted HEMAM and the β-substituded 2-methyl HEMAM methacrylamides heated at 50 °C. The kinetics of polymerization was followed in real time by near-IR spectroscopy during the photopolymerization for 300 s at 630 mW/cm².

Table 2 –

Maximum rate of polymerization (RP_MAX – %·s⁻¹), degree of conversion at the maximum rate of polymerization (DC at (RP_MAX – %) and final degree of conversion (Final DC – %) for the methacrylate and methacrylamide controls, and the β-substituted methacrylamide 2-methyl HEMAM. Values followed by different letters indicate significant difference among the tested groups (p < 0.05).

Groups	RPMax (%·s−1)	DC at RPMax (%)	Final DC (%)
HEMA	17.4 (2.97) A	41.6 (2.41) A	92.0 (0.39) A
HEMAM	16.1 (1.53) A	29.6 (3.21) B	86.2 (1.34) B
2-methyl HEMAM	18.1 (1.19) A	31.3 (2.06) B	86.6 (1.32) B
p	0.5316	0.0026	0.0012

3.2. Dentin microtensile bond strength

Dentin μTBS for the selected groups after 24 h and 6 months storage time are shown in Fig. 4. At 24 h, the commercial control Single Bond and the methacrylamides, HEMAM and the α-substituted 1-methyl HEMAM, showed the highest values (53.4 ± 9.8, 40.4 ± 5.9, and 45.5 ± 6.4 MPa, respectively), whereas HEMA α-,β-CH₃ mixture presented the lowest bond strength (23.2 ± 6.8 MPa). After 6 months, Single Bond and 1-methyl HEMAM presented the highest results (43.3 ± 5.3 and 42.9 ± 7.2 MPa, respectively). The methacrylates HEMA and HEMA α-,β-CH₃ mixture presented the lowest values: 21.9 ± 5.2 MPa and 12.7 ± 3.2 MPa, respectively. The reduction in μTBS over time was statistically significant only for the HEMA α-,β-CH₃ mixture, but all groups showed a numeric reduction ranging between 37.5% and 5.7% for HEMA and 1-methyl HEMAM, respectively.

Fig. 4 –

Dentin microtensile bond strength after 24 h and 6 months aging for the selected groups. Different uppercase letters indicate significant difference among the groups at the same storage condition (p < 0.05), and different lowercase letters indicate significant difference between the storage conditions within the composition (p < 0.05).

3.3. Monomer hydrolysis kinetics

All tested monomers showed a measurable amount of degradation after 4 days or less exposure to acidic aqueous conditions (H₂O, pH = 1) (Fig. 5). The unsubstituted methacrylamide, HEMAM, experienced the least degradation with 89.1 ± 0.01% of the monomer remaining intact after 17 days of incubation. The substituted methacrylamides, 2-methyl HEMAM and 1-methyl HEMAM, both exhibited more degradation than HEMAM, with 83.5 ± 0.5% and 65.4 ± 1.3% intact monomer remaining. The two methacrylate monomers exhibited similar degradation amounts, HEMA with 25.2 ± 0.1% intact monomer after 19 days incubation and the α,β-CH₃ HEMA mixture with 31.9 ± 0.4% intact monomer after 17 days incubation. The half-life results from fitting to an exponential decay model and a transformed linear regression showed that HEMAM had the greatest half life at 104 days, nearly 10× longer than the α,β-CH₃ HEMA mixture (Table 3). Additionally, transesterification products were observed in the methacrylamide groups. NMR spectra and further analysis can be found in the supplementary information.

Fig. 5 –

Degradation of monomer over time in acid aqueous conditions (H₂O, pH = 1) at 37 °C. NMR spectroscopy was used to determine the amount of remaining monomer compared to degradation products (and transesterification products where applicable). Error bars are too small to be seen. Data were fit to a pseudo-first order exponential decay curve to determine monomer half-life. Half-lives for monomers in acidic aqueous conditions at 37 °C. Data were fit to an exponential decay model (Fig. 5) as well as linear regression using a natural logarithmic transformation (Fig. S14).

Table 3.

Half-lives for monomers in acidic aqueous conditions at 37°C. Data were fit to an exponential decay model (Fig. 5) as well as linear regression using a natural logarithmic transformation.

	Linear regression half-life (days)	R2	Non-linear regression half-life (days)	R2
HEMAm	101 ± 6	0.967	104 ± 12	0.968
2,1-HPMAm	68.8 ± 1.5	0.996	68.8 ± 3.5	0.996
1,2-HPMAm	27.7 ± 0.6	0.996	27.9 ± 1.0	0.997
HEMA	9.52 ± 0.09	0.999	9.89 ± 0.22	0.999
HPMA	9.52 ± 0.08	0.999	9.57 ± 0.20	0.999

4. Discussion

4.1. Dynamic isomeric transesterification equilibrium

The degradation of the adhesive interface is believed to be one of the most important causes of the reduced clinical lifetime of adhesive dental restorations. In addition to the collagen degradation, the hydrolysis of the polymeric constituents play a crucial role on the adhesive interface instability [3–6], leading to the development of more degradation resistant monomers. The original experiment was designed to systematically evaluate the effect of the carbon substitutions on monomer reactivity and stability, in which alpha and beta-substituted HEMA monomers would be tested individually. However, the systematic evaluation of the effect of the methyl substitutions for the HEMA derivatives was not possible due to the facile susceptibility to α and β-CH₃ isomerization. In fact, even the commercially available HEMA methylated derivative is only available as a mixture of isomers. This isomerization is very likely occurring via a low-energy transesterification mechanism [33]. For example, the β-substituted 1-methyl HEMA is shown as an example in Fig. 6 and is particularly susceptible to this isomerization. The terminal hydroxyl group participates in a transesterification resulting in a dimethacrylate and propane-1,2-diol (Scheme 3 in the supplemental information). Another transesterification occurs resulting in two monomethacrylates. In order to return to the original monomethacrylate, the secondary alcohol of the diol would need to participate in the transesterification while the more nucleophilic primary alcohol results in 2-methyl HEMA. This difference in nucleophilicity explains why an eventual equilibrium of a 3:1 ratio of 2-methyl HEMA to 1-methyl HEMA is reached in commercial HEMA α-,β-CH₃. This transesterification will not result in an isomerization in unsubstituted monomers, like HEMA, though the dynamic behavior is likely still occurring. This mechanism explains the observation that the common impurities in commercial HEMA are dimethacrylate and ethylene glycol [34]. A detailed exploration of this phenomenon is beyond the scope of this paper.

Fig. 6.

Amino-terminated methacrylates (AEMA, 1-methyl AEMA, and 2-methyl AEMA) were obtained or synthesized as hydrochloride salts and did not form isomers like their hydroxyl-terminated counterparts. However, when the hydrochloride salts were free-based before being incorporated into the formulations, they rapidly formed amides from the primary amine, acting as a nucleophile, resulting in a mixture of methacrylates and methacrylamides (and likely hybrid methacrylate-methacrylamide) monomers. Additional details are discussed in the supplementary information.

4.2. Polymerization kinetics

The reactivity rates as observed from the polymerization kinetics experiments ranked as follows: NH₂-terminated methacrylates < OH-terminated methacrylamides < OH-terminated methacrylates.

The NH₂-terminated methacrylates (AEMA, 1-methyl AEMA and 2-methyl AEMA) resulted in a copolymerization characterized by markedly low values of RP_MAX and DC at RP_MAX (Fig. 2 and Table 1). One possible explanation is potential phase-separation, as the final polymer presented a nacre-like structure. While one may expect the slow kinetics for these monomers would result in low final DC values, the final DCs were actually comparable or only slightly lower as compared to the other monomers. This was likely due to the long period of photoactivation which compensated for the slow curing kinetics. However, the low reactivity and phase-separation make the NH₂-terminated monomers unsuitable for dental material compositions and, therefore were not subjected to further tests.

For the remaining compounds, in general, OH-terminated methacrylates showed higher values of RP_MAX and DC at RP_MAX than the secondary methacrylamides. This was expected since the tested methacrylates show lower molecular weight and viscosity than the methacrylamides, which likely increased the molecular mobility [35]. In addition, due to the known strong resonance stabilization of the carbonyl with the lone pair of electrons from the nitrogen, amide functionalities show significantly lower reactivity than methacrylates [36].

The side-group substitution played no significant role in the monomer reactivity, especially for the methacrylamides. It had previously been shown that the incorporation of ethyl or methyl side-group substituents on the α carbon of secondary methacrylamides resulted in marginally increased polymerization reactivity and, subsequently, increased bonding performance [22]. An increase in polymerization kinetics of a methacrylamide monomer with side-chain substitution can likely be attributed to amide twisting (a twisting of the amide C(O)–NH bond), which changes the geometry, reducing the contribution from the non-propagating resonance form (Scheme 2) due to de-conjugation of the nitrogen lone pair with the carbonyl π-system [37]. In this study, the lack of an observable effect is likely due to the small steric profile of the methyl substitutions.

The OH-terminated methacrylate HEMA (control) presented the highest values of DC at RP_MAX and final DC when polymerized at 50°C, with HEMAM and 2-methyl HEMAM being similar to each other. This was expected due to the differences in molecular weight, viscosity and reactivity among the compounds, as discussed above. The similarity of RP_MAX among HEMA and the secondary methacrylamides was also observed for the polymerization kinetics evaluated at room temperature in this study and previously reported [22]. As mentioned above, HEMA has low molecular weight and viscosity (130 g/mol and 0.007 Pa.s), which increases the overall mobility within the comonomer system. This allows for a rapid increase in the rates of propagation and termination at the beginning of the polymerization reaction, until the formation of high molecular weight species severely hamper diffusion [35]. Despite methacrylates having higher reactivity than methacrylamides, the observed RP_MAX values were similar. This observation reinforces that the methacrylamide-methacrylate blend ratio used in this study provides good properties without a significant loss of polymerization reactivity.

Another interesting finding is that the polymerization kinetics was less affected than expected when the polymerization was carried out at 50 °C, compared with room temperature. The increase in RP_MAX was similar for HEMA and HEMAM (22% and 24%, respectively, and the DC at RP_MAX increased by 19% for HEMA and 33% for HEMAM. The increase in and final DC was negligible. The effect on DC at RP_MAX observed for HEMAM was expected based both on the increased mobility and on the decrease in activation energy at higher temperatures [35], mainly because of its greater hydrogen bonding potential and viscosity at room temperature [38–40]. The absence of significant effects on final DC is likley due to the fact that the materials were polymerized at a relatively high intensity, for a relatively long time (300 s at 630 mW/cm²).

4.3. Dentin microtensile bond strength

The similarity of the μTBS (after 24 h) for commercial control Single Bond and the experimental secondary methacrylamides (HEMAM and 1-methyl HEMAM formulations) suggests that the methacrylamides may be useful candidates as co-monomers for improved dental adhesive formulations.

The low μTBS of the HEMA α-,β-CH₃ mix stands in contrast to the excellent polymerization kinetics observed. It is possible that the high polymerization rates and side-group substitutions at the α- and β-carbons might have resulted in a poorly packed polymer network with compromised mechanical properties [41]. The β-substituted 2-methyl HEMAM showed a statistically equivalent μTBS to the other secondary methacrylamides (HEMAM and 1-methyl HEMAM), though lower than Single Bond. A previous study has shown that the performance of the (meth)acrylamide copolymers is highly dependent on the chemical structure localized about the amide, as well as the blending monomer [22].

As mentioned previously, 2-methyl HEMAM is a powder and, even after the addition of solvent for the adhesive formulation, still led to a product with noticeably higher viscosity compared with the remaining monomers. Even though viscosity was not measured in this study, it can be speculated that this affected the quality of the hybridization of the collagen substrate in the 2-methyl HEMAM group. At 6 months, while Single Bond, HEMAM and 1-methyl HEMAM maintained the highest bond strengths, 2-methyl HEMAM showed intermediate results, and the experimental methacrylates HEMA and HEMA α-, β-CH₃ mix the lowest bonds, which once again highlights the degradation resistance of the methacrylamides. Finally, the reduction in μTBS over time ranged between 37.5% for the methacrylate control, HEMA, and 5.7% for the α-substituted secondary methacrylamide, 1-methyl HEMAM, which can be explained by the increased resistance to hydrolysis of the methacrylamides compared to their methacrylate counterparts.

4.4. Monomer hydrolysis kinetics

The hydrolysis results confirmed the expected increased resistance to hydrolysis, as all methacrylate monomers showed significantly more degradation in acidic aqueous conditions than the methacrylamide monomers. Notably, the hydrolysis rates of the amides vs esters were in agreement with simple chemical models which predict a factor of ten difference ([28], Scheme 1 in supplemental information). The unsubstituted methacrylate, HEMA, and mixture of isomers, HEMA α-,β-CH₃, performed similarly to each other and poorly in comparison to the methacrylamides in the degradation experiment. Both HEMA and HEMA α-,β-CH₃ had a half-life of 9.52 days (linear regression), while the worst performing methacrylamide, 1-methyl HEMAM, had a half-life of 27.7 days. The HEMA α-,β-CH₃ mixture of isomers was mostly composed of the β-substituted 2-methyl HEMA (3:1), but this appeared to have no benefit or detriment to the stability to acid-catalyzed hydrolysis compared to HEMA.

Interestingly, the addition of a α- or β-CH₃ groups had a detrimental effect (i.e., increased hydrolysis rate). The α-,β-CH₃ methacrylamides (1-methyl and 2-methyl HEMAM) showed higher degradation rates (half-lives of 68.8 and 27.7 days, respectively) compared to unsubstituted HEMAM (half-life of 101 days). The α-CH₃ substituted 1-methyl HEMAM was hydrolyzed about 3.5 times faster than HEMAM, with the the β-CH₃ substituted 2-methyl HEMAM being hydrolyzed about 1.5 times faster than HEMAM. This trend is the opposite of what was expected from a simple chemical steric model for base-assisted hydrolysis (Scheme 1). Notably, the magnitude of the effect is approximately the same. These observations suggest that for the acid catalyzed hydrolysis reaction, increased steric influence accelerates the leaving group (X-R in Scheme 1 of the supplemental materials). This would be the case, for example, if protonation of the heteroatom (O for ester, NH for amide) were the rate-limiting step instead of the H₂O nucleophilic attack of the carbonyl, which also tracks with the basicity of the heteroatom increasing in the order (e.g., HEMAM < 1-methyl HEMAM < 2-methyl HEMAM).

During the hydrolysis experiment, transesterification products were observed for the methacrylamide monomers. Acid and base catalyzed transesterification of monomers like HEMA have been reported under concentrated conditions, such as hydrogels [42]; however, this was an unexpected chemical equilibrium at the dilute aqueous conditions of the experiment. 1-methyl HEMAM exhibited the most transesterification product after 17 days (Fig. S12), making up 9.0%. While this number may seem small, only 34.6% of the monomer hydrolyzed at this time point meaning 26.0% of the 2-aminopropan-1-ol degradation product re-esterified into the methacrylate transesterification product, 2-methyl AEMA.

The re-esterification likely occurred due to the low pH of the aqueous environment. At pH = 1, the amino group is more than 9 units below its pKa and the population of water molecules is essentially all hydronium ions, leaving the alcohol as the best available nucleophile for the transesterification reaction. This concept is consistent in the other methacrylamide groups, HEMAM (Fig. S11) and 2-methyl HEMAM (Fig. S13). In the case of HEMAM, there is very little monomer hydrolysis, only 10.9%, but 36.3% of the less sterically hindered degradation product, aminoethanol, re-esterifies to form AEMA. Of the 16.51% of hydrolyzed 2-methyl AEMA, only 22.8% re-esterifies into the transesterification product, 1-methyl AEMA (see Figs. 6 and 7).

Fig. 7.

The secondary alcohol of the 1-aminopropan-2-ol is less nucleophilic than the primary alcohol of 2-aminopropan-1-ol and aminoethanol, resulting in less re-esterification of the degradation products. The fact that weak nucleophiles such as secondary alcohols are able to participate in this transesterification suggests that the activated carbonyls of methacrylates and methacrylamides are very prone to transesterification. In the case of HEMA, much like in the discussion of amino-terminated monomers, this transesterification can go unnoticed as ethylene glycol can only re-esterify into HEMA. This would mean that the HEMA molecule is breaking and reforming, which makes the degradation of HEMA appear artificially slow compared to the methacrylamide groups where the amine is protonated and unable to participate in re-esterification of the original monomer. More importantly, this evidence of transesterification would suggest that more care should be used when analyzing degradation products of dental materials as the degradation product mixtures have potential to be more complex than simply primary degradation products.

5. Conclusions

The blend of methacrylamides (40 wt%) in methacrylates produced good bond strengths and excellent hydrolytic stability, while retaining acceptable polymerization kinetics. α-and β-CH₃ derivatives had a non-measurable effect on polymerization kinetics, suggesting that the methyl moiety is not sufficiently sterically hindering to affect the reactivity as previously seen with larger ethyl groups. NH₂-terminated monomers had unacceptable polymerization rates, making these non-starters for dental materials. Even the fast polymerization kinetics of HEMA α-β-CH₃ mixture along with the side-group substitutions compromised bond strength.

Amide monomers were approximately ten times more stable to hydrolysis than the analogous methacrylates. The addition of a α- or β-CH₃ groups increased the rate of hydrolysis. The magnitude of the effect was approximately the same as a model system (base-catalyzed hydrolysis of esters or amides) but opposite in influence.

Finally, a side reaction was observed with transesterification of the monomers studied under ambient conditions. Transesterifications of this nature typically occur under much harsher reaction conditions, and it is not clear why the transesterification of the diol moiety is so facile for methacrylates compared to simple esters.