Acid Resistance of CAD/CAM Resin Composites

Acid resistance of CAD/CAM resin composites. Erosion-related tooth surface loss is closely related to acid exposure, such as contact with acidic beverages or disease-related reflux. As a result, dental restorations in affected patients are also exposed to acids, which indicates that the performance and longevity of a dental restoration is impacted by the acid resistance of the individually employed restorative materials. However, unlike for ceramic materials, the acid resistance of CAD/CAM resin composites is not commonly evaluated by the manufacturers, and no standardised test methods have yet been established. Against this background, the present in vitro study aimed to examine the long-term resistance of CAD/CAM resin composites (Brilliant Crios, Cerasmart, Grandio blocs, Lava Ultimate, Shofu Block HC) against three acidic media (tonic water, acetic acid, hydrochloric acid) as well as demineralized water and to investigate potential damage mechanisms. Changes in surface roughness (Sa) were detected by confocal laser scanning microscopy (CLSM), and changes in surface hardness were measured using Vickers hardness (HV). The damage mechanisms were analysed by scanning electron microscopy (SEM) with energy dispersive X-ray spectroscopy (EDS) and micro X-ray computer tomography (µXCT). For each material, few changes in either Sa or HV were identified for at least one of the different media; for Cerasmart, the sharpest deterioration in surface properties was observed. SEM–EDS revealed leaching of barium, aluminium, and titanium from fillers in a 2 µm zone on the rough but not on the polished surface of the specimen. Within the limitations of the current study, it can be concluded that polished CAD/CAM resin composites can be recommended for clinical use in patients with erosive conditions.


Introduction
Aetiologically, tooth surface loss is the sum of cumulative, multifactorial events, which ultimately lead to an irreversible loss of superficial tooth structure. In addition to idiopathic and genetic factors, mechanical or chemical processes can also cause tooth surface loss [1]. Whereas attrition describes intrinsic mechanical wear, which is caused by direct functional or parafunctional tooth-to-antagonist contact [1], abrasion refers to extrinsic mechanical such as water or artificial saliva suggests that the material properties (strength, fracture behaviour, hardness) and the surface properties of CAD/CAM resin composites, which are decisive for microbiology, are affected by exposure to acids [26,27].
By using selected CAD/CAM resin composites, this study evaluated their resistance against different acidic solutions (tonic water, acetic acid, hydrochloric acid) and control medium (demineralized water) in terms of surface roughness (Sa) and Vickers hardness (HV). Based on current knowledge, the null hypothesis of our study was that exposure to acid solution does not significantly affect the near-surface properties, such as the hardness and roughness, of CAD/CAM resin composites.

Specimen Preparation
Five materials were investigated for the in vitro analysis of the acid resistance of CAD/CAM resin composites (Table 1). Table 1. CAD/CAM composites used in this study and composition of the inorganic fillers [12,17]. For the fabrication of the specimens, the corresponding CAD/CAM resin composite blocks were cut by using a precision saw (IsoMet4000, Buehler Ltd., Lake Bluff, IL, USA) under water cooling. Then, the specimens were subjected to a constant grinding and polishing regime by using a semiautomatic polishing device (Pedemin-2/DAV-5, Struers GmbH, Willich, Germany). Sandpaper discs up to P500 and grinding discs with a diamond suspension down to a particle size of 0.04 µm were used. After the polishing process, the slices had dimensions of 13 ± 1 × 14.5 ± 0.5 × 2.0 ± 0.2 mm 3 . This resulted in a ratio of sample surface to media volume of 4-5 mm 2 mL −1 . Finally, the specimens were cleaned with demineralized water, dried for 17 h at 40 ± 1 • C, and then stored in a desiccator. Four different media were used to stress the CAD/CAM test specimens (Table 2), which were selected in order to cover a wide range of frequently present acid-related erosive factors with a high erosive potential (see Section 1). For each specimen, 100 mL of the medium was initially pretempered in 250 mL screw-top vials at 40 ± 1 • C (the lower temperature was selected for imaging clinically relevant strains) in the drying cabinet. After stress loading for 232 h (t 1 ), the specimens were removed from the medium, cleaned with demineralized water, and dried with subsequent storage in a desiccator to prevent an undesirable influence on the results (especially mass, dimensions, microhardness) due to, e.g., impurities or pure swelling processes, as they occur in dental resin composites, and to ensure comparability of the test [28]. Constant stressing was ensured by monitoring of the pH value and the conductivity (σ) in each medium before and after insertion of the respective specimen. A data Biomedicines 2022, 10, 1383 4 of 13 logger (ALMEMO ® 2590 A, Ahlborn Mess-und Regelungstechnik GmbH, Holzkirchen, Germany), which was connected with a pH sensor (DULCOTEST PHER 112 SE, ProMinent, Heidelberg, Germany), a σ-sensor (FYA641LFP1, Ahlborn Mess-und Regelungstechnik GmbH, Holzkirchen, Germany), and a temperature sensor (R2E4, Ahlborn Mess-und Regelungstechnik GmbH, Holzkirchen, Germany) were used to analyse the media before and after loading. Table 2. Liquids to which the test specimens were exposed.

Roughness Parameters
The measurements to determine the surface roughness prior to (t 0 ) and after (t 1 ) loading were performed by confocal laser-scanning microscopy (CLSM) (VK-X1000/1050, Keyence, Osaka, Japan). For each specimen and loading cycle, two measurements were performed by using a 50× objective (CF IC EPI Plan 50×; N = 0.8; WD = 0.54 mm) on eight areas on the entire surface of each specimen to minimize the influence of local discontinuities on the surface analysis.
The recording and calculation of surface roughness was conducted with manufacturer software (VK Viewer 1.1.2.174 and MultiFileAnalyzer 2.1.3.89, both from Keyence, Osaka, Japan). A consistent correction was carried out for all surface scans. The form of the specimen surface was filtered out with a cut-off of 0.2 mm (F-filter). An S-filter of 0.5 µm was applied to analyse surface details based on an S-F-surface. In both approaches, a Gaussian filter was used, and an end effect correction was performed. To ensure unaltered surface characteristics of the SH samples through filtering, an L-filter was deliberately omitted. For the subsequent determination of the surface roughness, each scan of the S-F-surface was divided into four areas (each 100 µm × 100 µm). The arithmetical mean height (Sa) for each area was determined for each material group, media, and loading cycle [29]. A positive ∆Sa represents an increase, and a negative ∆Sa represents a decrease in surface roughness.

Determination of Vickers Hardness
Vickers hardness (HV) was determined prior to (t 0 ) and after (t 1 ) loading by using a microindentation tester (MHT-4 Anton Paar, Graz, Austria) with a pyramid-shaped diamond indenter. Each specimen was placed on the examination plate with the tested surface facing the indenter. On the surface of each specimen, three indentations with a testing force of 200 p = 1.961 N, a testing time of 12 s, and a testing speed of 35 p s −1 = 0.343 N s −1 were produced. Sufficient distance of the indentations from each other and from the edge of the specimen was verified [30]. To determine HV, two-point measurements via CLSM were applied to the indentation diagonals. HV was calculated according to Formula (1):

Statistical Analysis
Statistical analysis was performed (SPSS 25, SPSS Inc., IBM, Chicago, IL, USA), and normal distribution was determined by using the Kolmogorov-Smirnov and Shapiro-Wilk tests. Based on this evaluation, the data were analysed with paired t-test or Wilcoxon signed-rank tests to compare the material-specific properties in the various media. Surface roughness and HV were compared by using two-way ANOVA and multiple comparison analysis post-hoc test under Bonferroni correction. Statistical analysis was conducted at a significance level of α = 0.05.

Further Analysis to Investigate the Damage Mechanisms
Scanning electron microscopy (SEM) with energy dispersive X-ray spectroscopy (EDS) and micro X-ray computer tomography (µXCT) was used to qualitatively investigate changes in the microstructure of selected stressed materials that showed the strongest altered surface characteristics. As only a few samples showed changes, extensive investigations were carried out on an exemplary basis.
For the SEM-EDS measurements, two specimens of the selected material-one stressed in M H2O and the other stressed in M HCl -were sectioned and embedded in resin with the cross section facing upwards and were polished and sputtered with 10 nm Cr to prevent charging effects. The cross-section to the three surfaces that were previously in contact with the storage media was investigated. Measurements were performed with 5 kV (SEM images) and 15 kV (EDS mappings) on an Ultra 55 SEM (Carl Zeiss Microscopy GmbH, Jena, Germany) with an SE2 detector. EDS-Mappings were performed to investigate differences in elemental distribution (O, Al, Si, Ti, Cr, Ba, C) from the outer to the inner areas.
A directional X-ray tube "FXE 225.99" (200 kV, 150 µA, focal spot 0.6 µm, tungsten target) from YXLON International GmbH and a 2D-detector "1621 N" (2048 × 2048 pixels, CsI, pitch 200 µm) from PerkinElmer Inc. were used to analyse the cylindrical samples (Ø 2 mm) of selected material CE before and after the stress caused by M HCl . The edge length of the voxels was 2.8 3 µm 3 . The raw data were cut, oriented, and calibrated with ImageJ (National Institutes of Health, 1.51d, Bethesda, MD, USA) in accordance with Koenig (2020) [31].

Surface Roughness
The evaluation of the S-F-surface (filtered surface, see Section 2.2) for all specimens revealed the following gradation for the roughness parameter Sa before exposure to the storing media (t 0 ): SH > LA > GR > BR > CE (p ≤ 0.031); only BR did not significantly differ from GR and CE. The highest Sa value was recorded for SH with 0.036 µm, and the lowest was recorded for CE with 0.008 µm (Supplementary Table S1).
After exposure to the various media (t 1 ), the gradation of mean Sa was (t 1 ): SH > GR > CE > LA > BR (p ≤ 0.035), with no significant differences between BR and LA and a significant difference between CE and SH. The Sa value of SH was significantly higher than those of all other CAD/CAM resin composites (p < 0.001), with a maximum value of 0.038 µm after storage in M AcOH . With respect to the media, the roughness gradation after immersion (t 1 ) was as follows: M AcOH > M HCl > M H2O > M TW (p ≤ 0.003), while M H2O and M TW did not significantly differ from each other.
For the mean changes in Sa values ∆Sa (t 1 −t 0 ), a similar trend was observed ( Figure 1). The smallest changes were observed with M TW . After exposure to M AcOH (with the exception of LA) and M HCl , all CAD/CAM resin composites showed a significant change in Sa (p ≤ 0.044). The maximum increase in Sa (∆Sa = +0.010 µm) was observed with M AcOH for SH, followed by CE in M HCl , M AcOH , and M H2O . In addition, the maximum decrease in Sa was determined for SH (∆Sa = −0.025 µm) after storage in M H2O . for SH, followed by CE in MHCl, MAcOH, and MH2O. In addition, the maximum decrease in Sa was determined for SH (ΔSa = −0.025 µm) after storage in MH2O. In comparison with the other CAD/CAM resin composites, only the surface of SH was very heterogeneous, with more or less semispherical hills both before and after loading in the storing media (Supplementary Figure S1). The change here was not immediately apparent by examining the surface images. In contrast, CE showed a relatively homogeneous surface before immersion in MHCl. After stressing, clear changes appeared in the form of heterogeneously distributed spots of greater depth ( Figure 2).  In comparison with the other CAD/CAM resin composites, only the surface of SH was very heterogeneous, with more or less semispherical hills both before and after loading in the storing media (Supplementary Figure S1). The change here was not immediately apparent by examining the surface images. In contrast, CE showed a relatively homogeneous surface before immersion in M HCl . After stressing, clear changes appeared in the form of heterogeneously distributed spots of greater depth ( Figure 2). for SH, followed by CE in MHCl, MAcOH, and MH2O. In addition, the maximum decrease i Sa was determined for SH (ΔSa = −0.025 µm) after storage in MH2O. In comparison with the other CAD/CAM resin composites, only the surface of SH was very heterogeneous, with more or less semispherical hills both before and after load ing in the storing media (Supplementary Figure S1). The change here was not immediatel apparent by examining the surface images. In contrast, CE showed a relatively homoge neous surface before immersion in MHCl. After stressing, clear changes appeared in th form of heterogeneously distributed spots of greater depth ( Figure 2).

Vickers Hardness
The Vickers hardness of the CAD/CAM resin composites revealed the following gradation before exposure to storing media (t 0 ): GR > LA > BR > CE > SH (p < 0.001), while BR, CE, and SH did not significantly differ. The largest value of HV 161.5 was observed for GR and the smallest for SH with HV 66.8 (Supplementary Table S2).
The HV results after exposure to the different media were ranked as follows (t 1 ): GR > LA > CE > BR > SH (p ≤ 0.001); BR did not significantly differ from CE and SH. The HV values after immersion (t 1 ) ranged from 66.5 for SH after M TW to 161.6 for GR after M AcOH exposure. The following trend in mean HV was assessed after exposure to the media (t 1 ): M AcOH > M H2O > M TW > M HCl (p ≤ 0.007), with only the value after M HCl HV being significantly smaller than those after M AcOH and M H2O .
The mean changes in HV values ∆HV (t 1 −t 0 ) presented no significant changes after storage in M H2O and M TW (Figure 3). Through exposure to M AcOH , GR experienced the maximum increase in ∆HV of +12.9 (p ≤ 0.047). The highest changes occurred after exposure to M HCl : BR (∆HV = −2.4) and CE (∆HV = −5.2; maximum decrease), which was significant (p ≤ 0.044), whereas HV for SH increased by +6.3 (p < 0.001).

Vickers Hardness
The Vickers hardness of the CAD/CAM resin composites revealed the following gradation before exposure to storing media (t0): GR > LA > BR > CE > SH (p < 0.001), while BR, CE, and SH did not significantly differ. The largest value of HV 161.5 was observed for GR and the smallest for SH with HV 66.8 (Supplementary Table S2).
The HV results after exposure to the different media were ranked as follows (t1): GR > LA > CE > BR > SH (p ≤ 0.001); BR did not significantly differ from CE and SH. The HV values after immersion (t1) ranged from 66.5 for SH after MTW to 161.6 for GR after MAcOH exposure. The following trend in mean HV was assessed after exposure to the media (t1): MAcOH > MH2O > MTW > MHCl (p ≤ 0.007), with only the value after MHCl HV being significantly smaller than those after MAcOH and MH2O.

Damage Mechanisms
Only the CE material presented a significant reduction in HV in combination with increased Sa and a clear spatially resolved change in the surface after immersion in MHC ( Figure 2). Therefore, the microstructure of this material was selected for detailed investigations by SEM-EDS and µXCT after storage in water (MH2O) and hydrochloric acid (MHCl).
SEM images of CE after exposure to MH2O and MHCl presented the morphology of filler particles embedded in a polymeric matrix typical for CAD/CAM resin composites ( Figure  4). SEM-EDS mappings of the edges to side A and side B revealed no differences in elemental distribution and morphology between CE stressed in MH2O and MHCl. However, the cross sections at the edge of the unpolished side C showed clear differences between CE stressed in MH2O and MHCl in terms of morphology and elemental distribution ( Figure 4). While SEM images of different edges stressed in MH2O revealed no differences, CE stressed in MHCl had a zone of approximately 2 µm distance from the edge with particles that were less radiopaque than the more inward particles. The EDS mappings of the area confirmed the absence of titanium, aluminium, and barium in this layer, while the distribution of the other elements remained the same.

Damage Mechanisms
Only the CE material presented a significant reduction in HV in combination with increased Sa and a clear spatially resolved change in the surface after immersion in M HCl (Figure 2). Therefore, the microstructure of this material was selected for detailed investigations by SEM-EDS and µXCT after storage in water (M H2O ) and hydrochloric acid (M HCl ).
SEM images of CE after exposure to M H2O and M HCl presented the morphology of filler particles embedded in a polymeric matrix typical for CAD/CAM resin composites ( Figure 4). SEM-EDS mappings of the edges to side A and side B revealed no differences in elemental distribution and morphology between CE stressed in M H2O and M HCl . However, the cross sections at the edge of the unpolished side C showed clear differences between CE stressed in M H2O and M HCl in terms of morphology and elemental distribution (Figure 4). While SEM images of different edges stressed in M H2O revealed no differences, CE stressed in M HCl had a zone of approximately 2 µm distance from the edge with particles that were less radiopaque than the more inward particles. The EDS mappings of the area confirmed the absence of titanium, aluminium, and barium in this layer, while the distribution of the other elements remained the same.  The µXCT investigations of CE specimens before and after storage in hydrochloric acid have shown, taking into account the resolution of 2.8 µm, no changes in the grey value distribution and no material erosion at the surface ( Figure 5).

Discussion
The null hypothesis of the present study, which posits that exposure to acidic media does not significantly affect the near-surface properties such as the hardness and rough- The µXCT investigations of CE specimens before and after storage in hydrochloric acid have shown, taking into account the resolution of 2.8 µm, no changes in the grey value distribution and no material erosion at the surface ( Figure 5). The µXCT investigations of CE specimens before and after storage in hydrochloric acid have shown, taking into account the resolution of 2.8 µm, no changes in the grey value distribution and no material erosion at the surface ( Figure 5).

Discussion
The null hypothesis of the present study, which posits that exposure to acidic media does not significantly affect the near-surface properties such as the hardness and rough-

Discussion
The null hypothesis of the present study, which posits that exposure to acidic media does not significantly affect the near-surface properties such as the hardness and roughness of CAD/CAM resin composites, can be partly rejected. All five CAD/CAM resin composites examined in this study showed significant changes in either roughness (Sa) or Vickers hardness (HV) after exposure to at least one of the media: demineralized water (M H2O ), tonic water (M TW ), acetic acid (M AcOH ), and hydrochloric acid (M HCl ).
The main limitation of this in vitro study in comparison to clinical practice is the reduction in the number of simultaneously interacting types of stress. It is conceivable that in reality, the mechanical stress forms, e.g., abrasion, increase the progress of the damage caused by the acids.
Despite their statistical significance, the magnitude of the changes in roughness with max. ∆Sa = −0.025 µm (SH in M H2O ) and in Vickers hardness with max. 12.9 in ∆HV (GR in M AcOH ) were relatively small (Figure 3) in relation to the storage time (232 h). The change in Sa was approximately one-tenth of the roughness differences that, according to Jones et al. (2004), can be detected by the human tongue [32]. The change in HV (9%) was approximately three-quarters of the maximum percentage change observed by Sagsoz et al. (2019) after one week of ageing CAD/CAM ceramics in lactic or citric acid [33]. The intermaterial differences (max. 59%) far exceed the changes caused by acidic media. At the same time, even though the storage time was longer within the present study (232 h), the magnitudes of changes in roughness are similar to the maximum changes found by Munusamy et al. (2020); they observed a change in roughness Ra of 0.022 µm between LA aged for 168 h in lactic acid (Ra = 0.142 µm) and the control air (Ra = 0.120 µm) [25].
With regard to roughness, the yielded maximum absolute value in the present investigation of Sa = 0.038 µm (SH in M AcOH ) was due to a change of 0.010 µm (∆Sa) compared to Sa = 0.028 µm in the sample before immersion. This was lower than the surface roughness limit of Ra = 0.2 µm, which is crucial for bacterial adhesion [34][35][36][37]. Furthermore, a lower roughness no longer reduces bacterial retention on dental materials [35]. In addition, CAD/CAM materials with surface finishes for clinical purposes often present an even greater surface roughness [38], thus accounting for the precise labside polishing process performed in this study. The surface roughness values of high gloss polishing of CAD/CAM resin composites described in other publications with a labside approach [38,39] are comparable to the surface roughness Sa values determined in the present study. However, regulations are needed for a definite testing and evaluation concept for the acid resistance of dental CAD/CAM resin composites and resins.
Although the changes in Sa and HV were small, the differences induced through immersion in the media varied in number and intensity. When considering the effects of the different media on the CAD/CAM resin composites, it was noticeable that M TW had a lower effect on the roughness (Sa) than M H2O , despite the lower pH of M TW . This observation is in contrast to the results of Munusamy et al. (2020) [25], who noticed significantly increased roughness of CE, LA, and SH after 168 h immersion in citric acid (with a similar pH value to the tonic water used herein, which also contains citric acid) at 37 • C compared to demineralized water [25]. A possible explanation might be the prolonged time (232 h) and higher temperature (40 • C) applied in the current study. Munusamy et al. (2020) suggested that demineralized water might exert significant effects on roughness after longer conditioning times [25]. In the same way, Rosentritt et al. [39] were able to demonstrate water saturation after 90 to 180 days and effects in the resin by thermoanalytical studies. These effects can indicate ageing by water with increasing storage time [39]. Other factors might be explained by the lower ionic strength of demineralized water that enhances leaching, swelling, and causes the presence of additional ingredients in the medium M TW (compared to the pure citric acid in [25]), which may reduce the effect. In terms of Vickers hardness, no significant differences between M H2O and M TW were observed.
The maximum values of Sa and the maximum changes ∆Sa were detected for SH. The authors attribute this to the high heterogeneity of the SH surface even after a strict polishing protocol (Supplementary Figure S1). Compared to all other investigated CAD/CAM resin composites, filler particles of SH are predominantly spherical in nature [12]. This entails a minimal ratio between surface and volume. As a consequence, the bonding area between the filler and resin matrix is minimal compared to the other CAD/CAM resin composites. Presumably, the fillers of SH have a poorer bond to the resin matrix and can easily dissolve during abrasion. The heterogeneity of the SH surfaces is also reflected by the high standard deviations.
In addition to the already discussed heterogeneity of SH, the decreases in surface roughness Sa observed for SH (M H2O , M TW , M HCl ) and LA (M HCl ) in the respective media could also be explained by swelling. In the present study, this effect was limited due to the drying procedure before stressing the specimen: however, in general, it is well known for dental composites [28,40,41]. It is easy to imagine that when a relatively soft polymer matrix swells between rigid filler particles, the polymer matrix lifts due to the incorporation of solvent. Thus, the height differences would be levelled, and the surface roughness would be reduced. This effect was highest for M H2O (SH), which supports the interpretation. The increased Vickers hardness for SH in M HCl (∆HV = 6.3) and for GR in M AcOH (∆HV = 12.9) may also be explained by swelling in combination with postcuring [42]. Moraes et al. (2008) [42] explain the increase in Knoop hardness after water storage with increased monomer conversion or post-curing due to a crosslinking within the polymer matrix. The influences of large-scale local discontinuities in several CAD/CAM resin composites might also serve as a potential explanation for these unexpected observations [12].
For CE, the most pronounced surface changes were observed in M H2O , M AcOH , and M HCl . This was reflected by a strongly increased surface roughness parameter Sa (Figure 3). In contrast to the other CAD/CAM resin composites, this was already obvious with optical microscope images even without surface roughness analysis ( Figure 2). Furthermore, the most severe reduction in Vickers hardness HV was present for CE in M HCl . In general, degradation of CAD/CAM resin composites can be attributed to the direct attack towards the polymeric matrix, the leaching of filler particles, or the weakening of the bonding between fillers and matrix-addressing the silanisation [43]. Various examples exist for the hydrolysis of ester bonds that are present in the polymeric matrix of CAD/CAM resin composites [25,44,45]. Hence, it seems reasonable that at least some of the induced changes in CE were due to degradation of the polymer matrix. The filler particles of some CAD/CAM resin composites, such as CE, contain barium [17]. Söderholm et al. (2000) observed that barium-containing filler particles are susceptible to leaching [46].
Taking into account the resolution of 2.8 µm, the µXCT investigations showed that the grey value distribution in CE and the diameter of the specimens did not change after exposure to hydrochloric acid. Despite the effects on the surface roughness and Vickers hardness of CE, no deep damage was detected by µXCT. This is in accordance with the SEM-EDS results observed on the cross sections of CE in M H2O and M HCl (Figure 4). Only after stressing in M HCl did CE show a very narrow leached zone of approximately 2 µm (below the detection limit of µXCT). Therefore, only the surface, not the internal structure of CE, was changed by the acids. The elements leached were barium, titanium, and aluminium. As these are components of the filler particles in dental resin composites [12,47,48], it is reasonable to assume that the changes in surface roughness and Vickers hardness of CE after M HCl exposure are attributable to filler leaching. Moreover, filler leaching was only associated with the surface. The polymer matrix consists mainly of lightweight elements such as carbon or hydrogen. As EDS measurements are less sensitive to hydrogen, some changes in the volume of the polymer matrix might remain undetected.
A possible explanation for the fact that only the unpolished side seemed to be attacked by M HCl is that rough surfaces have high surface energy with scratches or holes. This is where dissolution processes of minerals and glasses are easily initiated, since less energy is needed to break the remaining bonds [49]. The increased surface area also accelerates the process. Buchwalter et al. (1982) reported that, in general, a higher surface roughness of simple silicate glasses increases their dissolution [50].
Taking our results and the limitations of the study into account, it can be assumed that the polishing process can increase the acid resistance for some CAD/CAM resin composites, such as CE. Although the present study design only investigated roughness as an effect of acid attack on CAD/CAM resin composites, there are indications of an increased susceptibility to acid exposure due to initially rougher surfaces. This should be investigated in further studies.

Conclusions
The resistance of five CAD/CAM resin composites against three relevant acidic media (tonic water, acetic acid, hydrochloric acid) and demineralized water were investigated by means of the surface roughness parameter Sa and the Vickers hardness. Within the limitations of this in vitro study, the null hypothesis that near-surface properties will not change significantly by acid solutions was partly rejected. The investigations revealed the following: (1) All of the investigated CAD/CAM resin composites were susceptible to at least one of the applied media. (2) The changes in surface roughness (max. ∆Sa = −0.025 µm) and Vickers hardness (max. ∆HV = 12.9) induced through the different media were small with regard to storage time, literature values, and clinically relevant thresholds [25,32,33,35]. (3) The greatest differences were observed for Cerasmart after storage in hydrochloric acid. Further investigations via SEM-EDS and µXCT revealed leached fillers with reduced quantities of barium, aluminium, and titanium that were present in a 2 µm surface region of the rough surfaces. The comparison between polished and unpolished surfaces suggested that acid resistance may increase with lower surface roughness.
In conclusion, the CAD/CAM resin composites showed sufficient resistance against different acidic media and demineralized water. For clinical usage, CAD/CAM resin composites can be recommended for use under erosive conditions. Nonetheless, regulations are needed for a definitive testing and evaluation concept for the acid resistance of CAD/CAM resin composites. Further investigations that focus on the possible influence of acids on the polymeric matrix and the effect of initial roughness on filler leaching behaviour remain interesting for future investigations.