Physical analysis of an acrylic resin modified by metal and ceramic nanoparticles

Abstract

Background. Nanoparticles (NPs) have gained significant attention in various fields due to their unique properties and potential applications. Polymethyl methacrylate (PMMA) is an acrylic resin widely used in dentistry and medicine. However, the effect of different types of NP fillers on the physical properties of PMMA-based resins has not been thoroughly explored in the literature.

Objectives. The present study aimed to evaluate the effects of 3 different types of NP fillers on the physical properties of an experimental PMMA-based resin as a function of the NP content and concentration.

Material and methods. Ten groups (n = 10) were designed. The specimens were composed of an acrylic resin, silicon dioxide (SiO₂), cerium dioxide (CeO₂), and titanium dioxide (TiO₂) at the following ratios (wt%): group 1 (G1) – control; group 2 (G2) – 0.5% SiO₂; group 3 (G3) – 1% SiO₂; group 4 (G4) – 3% SiO₂; group 5 (G5) – 0.5% CeO₂; group 6 (G6) – 1% CeO₂; group 7 (G7) – 3% CeO₂; group 8 (G8) – 0.5% TiO₂; group 9 (G9) – 1% TiO₂; and group 10 (G10) – 3% TiO₂. Transmission electron microscopy (TEM) was used to assess the quality of NP dispersion. Thermal stability was assessed with thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). The effects of the abovementioned NPs on the properties of the resin were evaluated using the Archimedes principle for density, the Vickers hardness (VH) test and the impact strength (IS) test. Data analysis employed the one- and two-way analysis of variance (ANOVA), followed by Duncan’s post hoc test at a significance level of 0.05.

Results. Transmission electron microscopy showed partial NP dispersion. All types of NPs enhanced the mechanical properties of the acrylic resin except for IS, which was similar to that of the control group. Among the types of NPs, irrespective of the weight percentage, CeO₂ showed higher thermal stability and higher IS for 0.5 wt% and 1 wt% as compared to other groups, as well as the highest values of density at 0.5 wt%, 1 wt% and 3 wt%. Titanium oxide at 1 wt% presented a higher VH as compared to other groups. The fracture pattern was the same for all groups.

Conclusions. Incorporating the tested NPs into the acrylic resin resulted in enhanced physical properties, primarily attributed to a lower NP content.

Keywords: nanoparticles, silicon dioxide, cerium dioxide, titanium dioxide, polymethyl methacrylate

Introduction

Acrylic resins have been used in dentistry and medicine since 1936, and have remained the primary material for manufacturing dental and craniofacial prostheses.¹ The material is typically fabricated with the use of polymethyl methacrylate (PMMA). The polymer has beneficial biological properties, as it is tasteless, insoluble and compatible with oral tissues. The favorable physical characteristics of PMMA include dimensional stability, resilience, resistance to compression, the ease of handling, and a low cost.² However, PMMA is not an ideal material in terms of mechanical properties required for long-term clinical use, such as flexural strength, hardness, density, and thermal stability.^{3, 4}

To improve the physical properties of different dental biomaterials, composites have been reinforced with inorganic nanofillers.^{5, 6, 7} Studies have shown that decreasing the size of the reinforcement particles to a nanoscale can completely alter the way they interact with the matrix, either due to the increased surface area or the possible interactions with the matrix in the molecular sphere, changing the chemical characteristcs of the matrix.⁸

Silicate-based minerals, such as silica (silicon dioxide (SiO₂), approx. 40 nm), have been used as inorganic fillers in methacrylate-based resins, and are associated with good mechanical and polishing properties. Numerous studies have demonstrated the correlation between the quantity of silica and the modulus of elasticity, wear resistance and polymerization shrinkage.^{9, 10, 11}

A previous study incorporated 3 different metal oxides in different proportions into PMMA, namely titania (titanium dioxide (TiO₂)), zirconium dioxide (ZrO₂) and aluminum oxide (Al₂O₃), and concluded that all experimental groups showed better resistance to impact, sorption and solubility as compared to the control group.³ Cerium-based metallic material, such as ceria (cerium dioxide (CeO₂)), is transparent to visible light, can act as a catalyst, and has been linked to better compression resistance properties and thermal expansion values of ceramics.¹² Ceria has demonstrated antibacterial properties against Pseudomonas aeruginosa after being treated with polyacrylic acid.¹³ However, there are no reports on the use of the material in the reinforcement of acrylic resins.

Titanium dioxide nanoparticles (NPs) are chemically inert, resistant to corrosion, non-toxic, and inexpensive. They have a high refractive index and antibacterial pro­perties under a variety of spectra.¹⁴ For these and other characteristics, TiO₂ NPs are incorporated into polymeric materials, such as PMMA.¹⁵ In comparison with its larger particle version, TiO₂ NPs primarily offer heat stability, clearly define the color of the polymer and provide improved reinforcing properties. As such, the particles do not lose their features during the polymerization process at high temperatures, contributing to the control of poly­merization shrinkage.³ The existing studies indicate that incorporating surface-treated TiO₂ into a PMMA matrix leads to a reduced shrinkage stress of the polymeric matrix, which is due to the formation of agglomerates that serve as concentration points for these stresses.¹⁶ According to a study by Acosta-Torres et al., with regard to the physical properties of conventional and nanopigmented PMMA, the matrices containing TiO₂ show less poro­sity.¹⁷ However, certain studies have highlighted the adverse effects associated with incorporating TiO₂ and SiO₂ into polymeric materials.^{18, 19}

In light of the current literature,^{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19} there is a notable research gap regarding the effects of 3 distinctive NPs – SiO₂, CeO₂ and TiO₂ – as potential fillers for experimental PMMA-based resins. While the use of NPs as additives has been explored in various materials, their interactions with PMMA-based resins remain surprisingly underrepresented in the current research. Therefore, the present study aimed to bridge this gap by conducting a comprehensive analysis of the influence of these NPs on the mechanical properties of a resin across a spectrum of concentrations. Our study encompassed a thorough examination of density, Vickers hardness (VH) and impact strength (IS) to assess the mechanical behavior of the resin. Furthermore, the study evaluated the thermal stability of the NP-integrated resin via thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC).

The purpose of the present study was to investigate the influence of different content and concentration of SiO₂, CeO₂ and TiO₂ on the physicochemical properties of an PMMA-based resin. The present hypothesis is that the incorporation of NPs does not alter the properties of PMMA.

Material and methods

To investigate the effects of different NPs on the pro­perties of a PMMA-based resin, 100 samples (N = 100) were prepared and divided into 10 distinct groups (n = 10 per group). A commercially available acrylic resin (Acron™ MC; GC America Inc., Alsip, USA) composed of powder and liquid was used. Three types of NPs, ranging in size from 15 nm to 50 nm, were added to the powder before the polymerization procedure by using a Hauschild SpeedMixer® DAC 150.1 (Hauschild & Co., Hamm, Germany) at 50/60 Hz and 3,500 rpm. The NPs comprised SiO₂ (15 nm), CeO₂ (25 nm) and TiO₂ (50 nm) in standardized sizes, as provided by the manufacturer (SkySpring Nanomaterials Inc., Houston, USA). The specimens in each group were composed of an acrylic resin, SiO₂, CeO₂, and TiO₂ at the following ratios (wt%): group 1 (G1) – control; group 2 (G2) – 0.5% SiO₂; group 3 (G3) – 1% SiO₂; group 4 (G4) – 3% SiO₂; group 5 (G5) – 0.5% CeO₂; group 6 (G6) – 1% CeO₂; group 7 (G7) – 3% CeO₂; group 8 (G8) – 0.5% TiO₂; group 9 (G9) – 1% TiO₂; and group 10 (G10) – 3% TiO₂.

The preliminary test using transmission electron microscopy (TEM) suggested that the NPs had better dispersion when added to the powder instead of the liquid. After microwave polymerization, the samples were cut with a vitreous fragment at 45°, and then an ultramicrotome (EM UC6; Leica Microsystems, Wetzlar, Germany) was used to obtain ultrathin (80-nanometer) sections. The slices were placed in a fine mesh (grid storage box; Agar Scientific Ltd., Stansted, UK) and observed under a TEM at 300 Kv (Tecnai G2 F30; FEI Company, Hillsboro, USA) to identify the NP content. The NPs were incorporated into the powder, and the flask was closed and placed inside the mixer at 2,000 rpm for 3 min, then rested on the bench for another 3 min. The liquid was added to the mixture and stirred at 500 rpm for 1 min to prevent the premature packing of the resin, which may occur due to the agitation and shear forces generated during this step. The liquid needs to be thoroughly and evenly dispersed within the powder to achieve a consistent resin.

The size of the samples depended on the specific characteristics of the respective tests. For instance, disk-shaped samples measuring 5 mm × 4 mm were employed in the VH test, while samples measuring 65 mm × 10 mm × 2 mm were utilized for the Charpy IS test. Similarly, the sample mass for TGA was standardized at 20 ±2 mg. By adjusting the sample sizes according to the unique demands of each test, we aimed to obtain reliable and comprehensive data that would accurately reflect the properties under investigation. However, the inclusion and polymerization processes were standardized for all samples.

Metal molds were used for the TEM and Charpy IS test samples, and composite resin molds were used for all other tested samples. The molds were invested individually with laboratory silicone (Titanium Zetalabor; Zhermack, Badia Polesine, Italy) and covered with type III dental stone (Scola-cast powder; Scola, Cheshire, UK) in a ceramic muffle oven (GC Europe, Leuven, Belgium). The experimental acrylic resin was prepared as previously described and immediately packed into the silicone index at its dough stage. The flasks were slowly and gradually pressed in a hydraulic press until estab­lishing 1.25 kg, and then polymerized in a micro­wave oven (Daewoo KQG-6L6B, 800 W maximum power; Daewoo Electronics, Wokingham, UK) for 10 min: phase 1 – 3 min at 40% power; phase 2 – 4 min at 0% power; and phase 3 – 3 min at 90% power. Then, the flasks were placed on the bench until completely cooled. The samples were deflasked and their surfaces were smoothed with an electric minigrinder (Dremel® 3000; Robert Bosch Tool Corporation, Mount Prospect, USA) at 5,000–33,000 rpm. For the VH test, one side of the sample was polished with 600–1,200 grit silicon carbide sandpaper (Jiangsu Jianda Grinding Technology Co. Ltd., Wuxi, China). The samples were stored dry at 24 ±1°C for 24 ±2 h before the tests.

A sample from each group was submitted for TGA. Thermal stability was determined by means of the thermogravimetric trace record (STARe software; Mettler-Toledo, Greinfensee, Switzerland) in an atmosphere of nitrogen (N), using small sample fragments. A heating rate of 10°C/min, with a temperature ranging from 35°C to 600°C, and a sample portion weighing 20 ±2 mg were used for this test. The balance arm was held in a horizontal reference position and changes in the sample mass caused the beam to bend, which was detected by the photoelectric cells. The TGA curve was obtained by recording variations in temperature and the sample mass continuously. Differential scanning calorimetry was conducted simultaneously with the TGA test, and was used to better understand the thermal behavior and homogeneity of the experimental resin samples. The same fragment from each group (20 ±2 mg) was analyzed to compare the glass transition temperature (T_g) of the polymer (control) and of the NP-modified polymer.

One hundred disk-shaped samples, divided into 10 groups (n = 10 per group), measuring 5 mm × 4 mm and with the top surfaces polished, were evaluated for VH in a microhardness tester (FM-700; Future-Tech Corp., Kawasaki, Japan). A 100-gram load was applied to the polished surface of the sample at the speed of 0.05 mm/s for 10 s, and the values were recorded in Vickers pyramid numbers (HV). Three indentations were made in each sample. Then, the Archimedes principle was applied to investigate density changes, with the mass (m) measured on an analytical precision scale and the volume (V) determined by the principle. The density of the sample (ρ) was calculated using the following formula (Equation 1):

where:

ρ – density [g/cm³];

m – mass [g]; and

V – volume [cm³].

One hundred samples, divided into 10 groups (n = 10 per group), measuring 65 mm × 10 mm × 2 mm were subjected to the IS test in the Charpy system (Otto Wolpert-Werke, Ludwigshafen am Rhein, Germany) with an impact load of 40 kpcm. The impact load recorderd at the moment of specimen failure was changed to the IS value [kgf/cm²] using the following formula (Equation 2):

where:

IS – impact strength [kgf/cm²];

I – impact load (kpcm);

L – sample length in the impact region [cm]; and

H – sample height in the impact region [cm].

A fractographic analysis was conducted for the 2 sample fragments obtained in the IS test, and a macroscopic analysis was done by the visual inspection of the fractured surfaces, using a claw-mounted magnifier (Western 9051L; Western Ophthalmics Corp., São Paulo, Brazil) at ×4 magnification. The fragments of each specimen were called fragment A (FA) and fragment B (FB). During visual inspection, when FA and FB could be repositioned at the fracture line, presenting a smooth surface, the fractures were classified as brittle. Those presenting plastic deformation, or exhibiting rough and jagged surfaces were recorded as intermediate (ductile-to-brittle transition) fractures. The fractographic examination verified that all samples could have FA and FB perfectly placed together, and not even the fracture line could be seen at ×4 magnification; therefore, all samples were classified as brittle. The control and the 1 wt% experimental samples were covered with a fine layer of gold, using a sputter coater (Emitech K550X; Quorum Technologies Ltd., Ashford, UK) and observed under a scanning electron microscope (SEM) (EVO MA10; Carl Zeiss, Oberkochen, Germany) to analyze and compare the fracture morphology and microstructure.

Statistical analysis

Data analysis employed the one- and two-way analysis of variance (ANOVA), followed by Duncan’s post hoc test at a significance level of 0.05.

Results

Transmission electron microscopy showed enhanced dispersion when NPs were incorporated into the powder component of the acrylic resin. However, the incorporation method did not completely prevent the formation of agglomerates (Figure 1). Furthermore, a strong bond between the NPs and the matrix was evident in cases of effective dispersion, while regions with more pronounced agglomeration displayed voids (Figure 1).

Table 1 presents the TGA and DSC values from the respective thermographs. The curve profiles in both ana­lyses exhibited pattern similarity, displaying only slight temperature shifts. The control group displayed a relatively higher mass loss following the initial event, and preliminary observations suggested that 1% TiO₂ ex­hibited a relatively higher initial degradation temperature (T_i). Moreover, all ceria groups presented higher T_i values than the control group. The final residue percentage increased proportionally to the amount of incorporated NPs in each group, except for the control group. In the 2^nd significant mass loss event, the control group displayed slightly higher temperatures, aligning with those observed in the ceria groups, which recorded relatively higher values as compared to the control.

The mean and standard deviation (M ±SD) values for VH, ρ and Charpy IS for all groups, including the control group, are reported in Figure 2, Figure 3, Figure 4. The experimental results demonstrated significant variations in the mechanical properties among the different groups. Notably, the discernible differences in VH, ρ and IS could be attributed to variations in group composition. The one-way ANOVA revealed a statistically significant p-value (p < 0.05), indicating the substantial influence of the NP content and concentration on the aforementioned mechanical properties, underscoring the importance of group composition in shaping the material characteristics.

The VH mean values for the control and 1% SiO₂ groups were significantly lower as compared to the 3% SiO₂, 3% CeO₂, 0.5% TiO₂, and 1% TiO₂ groups (Figure 2). Regarding density, the results show that the 3% CeO₂ group exhibited relatively higher scores, while the control group showed relatively lower values. The 0.5% SiO₂ and 0.5% TiO₂ groups had significantly lower means than the 3% TiO₂ group, though the results for the 3% SiO₂ and 1% TiO₂ groups did not differ statistically (Figure 3). For IS, the mean values in the 3% TiO₂ group were significantly lower than those in the control, 0.5% CeO₂, 1% CeO₂, and 0.5% TiO₂ groups. The 3% CeO₂ group scores were lower than those in the 0.5% SiO₂, 1% SiO₂, 3% SiO₂, 1% CeO₂, 0.5% TiO₂, and 1% TiO₂ groups, and significantly lower than those in the control, 0.5% CeO₂ and 1% CeO₂ groups (Figure 4).

A comprehensive multiple two-way ANOVA yielded significant outcomes, further confirming the substantial impact of the NP weight percentage and type on the observed results. The specific details and outcomes of this analysis are succinctly presented in Table 2. With regard to VH, only the 1% TiO₂ group showed higher values. There were no significant differences among the NPs for the tested weight percentages within the range of 0.5 wt% and 3 wt%. Regarding the NP type, TiO₂ showed no differences when various weight percentages were used. For density, CeO₂ showed higher values; for all NP types, the density values increased along with the weight percentage. Moreover, CeO₂ displayed higher IS than SiO₂ and TiO₂, which behaved similarly across all weight percentages. For SiO₂, the lowest IS value was obtained for 0.5 wt%, and all types of NPs showed lower values when 3 wt% was incorporated, with no difference between 0.5 wt% and 1 wt%.

The fractographic examination verified that all samples could have FA and FB perfectly placed together, and not even the fracture line could be seen at ×4 magnification. Therefore, the fracture patterns were similar (Figure 5).

Discussion

The hypothesis for this study was rejected, since the incorporation of NPs into the acrylic resin altered all the tested properties. The concept of adding reinforcement materials to enhance the mechanical properties of a dental polymer is not novel.²⁰ Nanoscale fillers have a significantly greater surface area per mass unit than their larger versions, which can entirely change the way they interact with the polymer matrix. The filler particles used in this study were smaller (15–50 nm) than the resin powder particles (121.2 μm); thus, it was expected that they would fill the interstices between the particles of the polymer, forming a homogeneous mixture without forcing the displacement of the polymer chain segments.^{3, 21}

The percentage of particles was kept low so that all of them could be involved in the matrix. Increasing the percentage to 3% was expected to improve mechanical properties, as, theoretically, there would be less space between the oxide NPs. However, it has often proved difficult to achieve a stable dispersion of NPs in polymer matrices, as NPs tend to agglomerate.^{3, 11, 22} The agglomerates can form stress concentration sites within the polymer matrix, leading to premature failure with the application of load. Synthesizing the nanocomposite and separating the aggregated NPs were the main challenges of this study.

To disperse the NPs in the hydrophobic polymer matrix, the modification of the NP surfaces was required; it involved silanization and dual asymmetric centrifugal mixing processing.²³ Transmission electron microscopy was employed to characterize the uniformity, or lack thereof, of the NP dispersion in the resin matrix. In­organic nanofillers were uniformly dispersed, more or less, when incorporated into the powder rather than the liquid. However, agglomeration could not be avoided, and the uniform dispersion throughout the polymer matrix was questionable (Figure 1). Nonetheless, this could be explained by the suspected imbalance between depletion effects and osmotic pressures in PMMA, adversely impacting the fracture behavior.

Thermogravimetric analysis is a thermal analysis technique in which the mass variation of the sample (loss or gain) is determined as a function of temperature and/or time while the sample is subjected to a controlled temperature program.²⁴ Such changes are evidenced by a decrease or increase in mass registered by the sensor. Our TGA analysis presented 2 distinct events. The 1^st event, at 175–330°C, could be related to the decomposition of second­ary groups, such as the free CH₃ groups attached to the main polymer chain, generating volatiles. The 2^nd and main mass loss event, at 328–455°C, can be attributed to the degradation of the main PMMA chain. As can be seen in Table 1, increases in the NP concentration and the NP content did not affect the initial degradation temperature (T_i). The glass transition temperature is the temperature at which the movement of the polymer chain segments starts, and at which the material changes from a ‘glassy’ state (frozen) to a ‘rubber’ state (more flexible), which can be observed in the DSC curve. Since PMMA is an amorphous polymer, it showed a T_g of approx. 105°C, as expected, which could be observed in the endo­thermic curve obtained in the heating run. Furthermore, we showed that the NPs may have interfered with the flowing process, which allows molecular chains to slide past each other, decreasing stiffness, and hence increased T_g.²⁵ The incorporation of NPs showed satisfactory thermal stabi­lity for use in dental PMMA, since all types, depending on the weight percentage, exhibited degradation at a temperature comparable to that for the control group and a T_g higher than in the control group.

The current study found that the incorporation of NPs increased microhardness, irrespective of the weight percentage, except for 1% SiO₂. The different densities of the NPs interfere directly with the light weight of the acrylic resin, which can be desirable. In this study, the incorporation of NPs, regardless of their type and weight percentage, significantly increased the density of the resin. A low density can be partly associated with the formation of pores or microgaps, resulting from processing, a pro­blem that affects acrylic resins for prostheses in general. Such high-porosity-level sites facilitate fluid transport into and out of the polymer, serving as places for the sequestration of molecules.²⁶

The NPs used in this study are insoluble in water and were meant to reduce the overall volume of the susceptible polymer matrix. Therefore, the higher density could be explained by a ‘packing’ effect, and not necessarily a lower porosity. In practice, impact fracture damage occurs when the patient accidentally drops the prosthesis. Studies have provided different proposals to minimize the susceptibility of the material to this flaw, such as increasing the thickness of the prosthesis in regions of increased susceptibility and modifying the matrix to improve the mechanical properties by adding polyfunctional agent crosslinks,²⁷ and the incorporation of a rubber phase,²⁸ fibers²⁹ and metal oxides.³ Impact strength can be defined as the energy required to fracture a material under the impact force. In this study, IS was not affected by the NPs as much as the other properties. Most groups presented lower IS values than the control, which could be related to an increase in density, thus reducing stiffness.

Regarding IS, the groups that received SiO₂ showed very similar mechanical behavior, which might have been caused by the clustering of the NPs heterogeneously inside the PMMA matrix, resulting in weaker spots to propagate fractures. In this case, the incorporation of NPs into the polymer with a mixer seems to be insufficient to ensure a fully uniform dispersion of NPs. Further studies evaluating different methods should be developed to incorporate NPs into PMMA for dental and medical applications.

No cracks or missing material were observed under macroscopic evaluation. The A and B fragments of all samples could be placed together perfectly, suggesting that the NPs limited the crack propagation and kept the samples from breaking into several pieces. Some macro­phase separations could be observed in the silica and titania groups under SEM, while the ceria morphology was similar to that of the control group. The SEM images showed a honeycomb-like configuration for silica due to intense porosity, and compact and wave-like surfaces for the control or in the presence of ceria. The titania groups had both of these characteristics. Nanoparticles play a relevant role in polymerization configuration, and thus, in the fracture propagation patterns. A previous study used SEM to observe agglomeration in PMMA specimens containing 0.91% silica, and stated that the particles appeared to be embedded in and semi-bonded to the PMMA matrix, indicating a relatively strong NP–matrix interaction.³⁰ The present study corroborates this interaction pattern, though it was more visible in the ceria groups.

Balos et al. added nanosilica to the liquid component of a commercial PMMA-based resin, and found that it increased the microhardness and fracture toughness values for the lowest content.³⁰ Cerium oxide has been used in dental materials to stabilize the polycrystalline tetragonal structure of zirconia.³¹ To our knowledge, there are no reports of its use for acrylic resin reinforcement. Interest­ingly, in this study, CeO₂ NPs showed better overall be­havior. In comparison with other groups containing NPs, the CeO₂ groups had a significantly higher degradation temperature, a higher mean IS, and similar VH and density values. Titanium oxide NPs are chemically inert, have a high refractive index and antibacterial properties under a variety of spectra, and are corrosion-resistant, non-toxic and inexpensive.¹⁴ These NPs have been previously incorpo­rated into polymeric materials, such as PMMA,¹⁵ and were reported to provide better reinforcing properties than their micrometric version.¹⁶ The impact of CeO₂ and TiO₂ on the resin color was readily apparent, potentially offering valuable insights. However, this effect should be given significant consideration in the evaluation and application of these NPs, warranting further in-depth investigation.

Limitations

Some limitations of the current study are acknow­ledged. Firstly, the dispersion method should be refined, perhaps with the use of ultrasonication. Also, surface roughness analysis could be very useful for corroborating the hypothesis of increased wear resistance in two- and three-body wear, as well as when considering biofilm formation and biological aspects.^{32, 33, 34}

Conclusions

Incorporating all 3 types of NPs into the PMMA matrix significantly enhanced thermal stability, surface hardness and density. Moreover, the impact resistance in the tested NP groups remained comparable to that of the control group, indicating that modifying the microwave-cured acrylic resin with specific amounts of NPs, particularly CeO₂, could serve as a viable solution to address the physico­chemical limitations of the material and to prevent un­wanted clinical failure.

Ethics approval and consent to participate

Not applicable.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Consent for publication

Not applicable.