Introduction

Due to its favourable mechanical [1,2,3] aesthetic [4] properties, dental ceramics made from yttria-partly stabilized zirconia (Y-PSZ) are widely used in dental applications [5, 6].

However, the final mechanical and optical properties (e.g. translucency) are mainly determined by the (micro) structural properties like phase composition and grain size [7]. The phase composition is influenced by the yttria content and other process conditions, including mechanical pressing [8], (hydro-) thermal treatment conditions [9, 10], and grain size of the raw material [11]. Y-PSZ utilized in dentistry may consist of several phases: The tetragonal phases T and T′′ (space group P42/nmc) are principally responsible for exceptional strength, whilst the cubic phase C (space group Fm\(\overline{3}\)m; mineral name: fianite) is responsible for high translucency [12,13,14].

The monoclinic phase M (P21/c; mineral name: baddeleyite) is the only thermodynamically stable phase (without stabilization with external elements) under standard laboratory conditions. Y-PSZ may undergo a very slow tetragonal to monoclinic (T → M) phase transformation [15,16,17,18] when subjected to thermal [9, 19, 20] or mechanical [6, 21, 22] stress under certain conditions, which is fostered by moisture [9, 21]. This results in a significant alteration of the mechanical properties of Y-PSZ, which can be beneficial [23,24,25,26] or detrimental. Due to the optical biaxiality of the monoclinic phase, the optical properties would degrade severely [9, 10, 19, 20, 27].

Additionally, the formation of a rhombohedral phase R (R\(\overline{3}\)) [28,29,30,31] was discussed, which results in a larger calculated shear and volume change compared to the monoclinic phase [8].

The exceptionally inert properties of Y-PSZ impede etching with HF to ensure appropriate preparation for bonding between the luting cement and the restoration [32, 33]. Therefore surface enhancing methods like sandblasting (interior) are in use to increase the microretention [32, 34, 35]. Small distance (≤ 1 mm [36]), high pressures (> 3 bar [36, 37]), larger particle sizes [38], or the type of blasting material (SiC has a higher impact than corundum/Al2O3 [37]) can potentially cause phase transformations with a monoclinic phase fraction as a result [36, 37, 39].

For the necessary chairside adjustments of the internal, proximal, and occlusal fit, the use of burs is required [40,41,42]. To prevent crack formation, microbial growth on the surface, wear of the antagonist, and to enhance the mechanical properties of the brittle ceramic, surface roughness (arithmetical mean height of the surface, Sa) must be reduced through an intraoral polishing regime, either dry or wet [5, 8, 42, 43].

After making any minor adjustments to the occlusion with a bur (either wet or dry), the surface requires polishing. The choice between wet or dry grinding for the adjustment will depend on the specific situation. If the need for adjustments arises after placing the restoration, it is necessary to use water cooling to avoid heat damage to the dental pulp and soft tissues in the patient’s mouth. The use of burs in general is postulated to cause a preferred orientation shift of the crystallites, but had no effect on the phase content [8, 44, 45].

Polishing regimes involve the use of two or three consecutive polishing tools, starting with a coarse one and ending with a fine one, in order to gradually reduce surface roughness.

With a speed of 0.1 mm/s and 10 swiping moves, exemplary surface roughness (Ra; arithmetical mean height) at 15,000 rpm (Tencor D-300, KLA-Tencor Co., Milpitas, CA, USA).

Exemplary surface roughness (Ra; arithmetical mean height) with 15,000 rpm (Tencor D-300, KLA-Tencor Co., Milpitas, CA, USA) and 10 swiping moves with 0.1 mm/s speed: glaze firing (0.14–0.19 µm), grinding (0.33–0.36 µm), coarse polisher (0.26–0.28 µm), medium polisher (0.13–0.19 µm), and fine polisher (0.03–0.08 µm) [40]. Recently proposed magnetic field-assisted batch polishing (MABP) regimes could achieve surface roughness of 0.0035 µm (Sa; arithmetical mean height) using diamond polishing slurry mixed with carried fluid [46].

Any kind of surface enhancement affects the surface roughness and may induce phase transformations into the monoclinic (M) or rhombohedral (R) phases [42, 47]. Additionally, the rhombohedral phase (R) might also retransform into the tetragonal phase (R → T [30]) and further into the monoclinic (M) phase over time.

For glazing, previous works reported no or a very small monoclinic phase formation [48, 49]. Grinding induced depended on the machining parameters, either a reflection broadening, which was attributed to a smaller crystallite size [8], or a rhombohedral phase formation [50, 51]. Additionally, the formation [23, 50], lacking [8, 28, 52], or reducing [52, 53] of a monoclinic phase fraction was reported. Grinding also induced a preferred orientation between 001T and 110T shift at the reflections around 35°, 60°, and 74°, reported by several works [8, 50].

All phase transformations can induce local internal stresses due to local volume expansion and shear [8, 9]. Several theories exist regarding the outcomes of such internal stress. When the internal stress is opposite to the prevailing stress, it can result in an increase in strength/hardness [23,24,25,26]. In reverse, when they are the same, it can result in a decrease in strength/hardness. In contrast to laboratory experiments that identify strengths based on standardized test specimens, such as beams and plates, chewing creates more complex multiaxial stress conditions in reality. These conditions can have a negative impact on the mechanical behaviour of the monolith due to pre-stress in the structure. If residual stress exceeds the strength, microcracks occur primarily at the grain boundaries, resulting in microcrack toughening and decreased stiffness (elastic modulus) in that region [26]. Furthermore, the mechanical properties of the monolith depend on the phase-specific mechanical properties, particularly on the surface, in addition to residual stress. The flexural strength and fracture toughness of the monoclinic phase are lower than those of the tetragonal or cubic phases [54]. These processes at room temperature are referred to as low-temperature degradation.

To recommend one or more polishing regimes, this study investigates the following questions:

  1. 1.

    Can the phase composition of the surface be influenced by wet and dry grinding occlusion adjustment and various polishing methods, and if so, are there differences between the regimes and are there possible explanations for their different behaviour?

  2. 2.

    Are there significant differences between the polishing regimes in the reduction of surface roughness? The null hypothesis is that surface roughness is not affected by any polishing regime.

Experimental approach

Materials

The manufacturing and analysis of flattened dental crowns are summarized in Fig. 1.

Figure 1
figure 1

Manufacturing and analysing of the dental crowns.

Full size image

For the analysis, top-flattened pre-molar dental crowns from the upper jaw are produced using a computer-aided design and manufacturing process (CAD/CAM) from blanks specified in Table 1. The specimens underwent sintering and glaze firing in accordance with the manufacturing recommendation (VITA Zahnfabrik H. Rauter GmbH & Co. KG, Bad Säckingen, Germany). Please refer to [48] for a detailed description of the standardized procedure.

Table 1 Overview of the Y-PSZs from VITA Zahnfabrik used for producing dental crowns
Full size table

The surfaces of the dental crowns were examined following various stages of preparations (Table 2).

Table 2 Different steps of processing of sintered dental crowns
Full size table

The occlusion adjustment of the dental crowns was carried out using a Meisinger diamond bur 850G (Hager & Meisinger GmbH, Neuss, Germany) for wet or dry grinding (Appendix, Table 4). Each diamond bur was utilized for 25 s at about 200,000 rpm. For every dental crown, a new diamond bur was employed (Appendix, Table 4).

Polishing was conducted for two minutes following the operational guidelines. The NTI polishing regime (NTI-Kahla GmbH, Kahla, Germany) includes three polishers, whereas the Komet (Komet Dental, Gebr. Brasseler GmbH & Co. KG, Lemgo, Germany) and EVE (EVE Ernst Vetter GmbH, Keltern, Germany) polishing regimes include only two polishers, so we used the NTI polisher for 40 s and the others for 60 s each (Appendix, Table 4).

The general approach for analysing dental polishing tools from the different manufacturers is summarized in Fig. 2.

Figure 2
figure 2

Analysing of the dental technical tools (Appendix, Table 4 and Fig. 11) of different manufacturers.

Full size image

Methods

Confocal laser scanning microscopy (CLSM)

A confocal laser scanning microscope (CLSM), Keyence VK-X1000/1050 (Keyence Corporation, Osaka, Japan), with a “Nikon CF IC EPI Plan 50X” objective (NA: 0.5, NIKON, Tokyo, Japan) and “VK Viewer 1.1.2.174” software (Keyence Corporation, Osaka, Japan) was used to take magnified images of the flattened top of the polishers. The “MultiFileAnalyser 2.1.3.89” software (Keyence Corporation, Osaka, Japan) was employed for analysis. The maximum particle size in the moulds was determined using 50× and 100× light and laser images.

For the roughness evaluation, a confocal laser scanning measurement was taken with a red laser (λ = 661 nm) at a resolution of 1024 × 768 pixels. The calculation of the arithmetical mean height Sa was carried out on 10 sections of the surfaces as specified in DIN EN ISO 25178-2. The procedure employed suitable filtering (S-filter: 1 µm; F-filter: 0.25 mm; L-filter: 0.08 mm, filter type: double Gaussian, with end-effect correction) to obtain the desired results.

All measurements for arithmetical mean height were averaged as mean, and the standard deviation was calculated. A normal distribution test was carried out using the Shapiro–Wilk test, and subsequently, an analysis of variance (ANOVA) was performed along with a multiple comparison post-hoc test that employed the Bonferroni correction. IBM SPSS Statistics 29.0.0.0 software was used for all computations with the level of significance set at α = 0.05.

Field emission scanning electron microscopy (FE-SEM)

Electron micrographs of the flattened top of the dental crowns were captured using a Zeiss Ultra 55 field emission scanning electron microscope (FE-SEM) (Carl Zeiss AG, Oberkochen, Germany) at 20-kV accelerating voltage. Before the analysis, the samples were embedded in resin, which was left to harden overnight at room temperature (RT). The embedded samples were grinded, polished, and vapour-deposited with gold.

X-ray diffraction (XRD)

For XRD analysis of the dental crowns, we utilized the “Bruker D8 series 1 Advance” (Bruker AXS Advanced X-ray Solutions GmbH, Karlsruhe, Germany) diffractometer in reflective Bragg–Brentano geometry. The instrument was equipped with a Cu-sealed tube operating at 40 kV/20 mA, Göbel mirror optics, a 0.2-mm divergence slit, a fixed knife edge to suppress air scattering, and a VANTEC-1 detector in scanning mode. The sintered ceramic crowns had a flat surface and were adjusted and aligned in a sample holder designed for large samples. The measurement time ranged between approximately 1and 3 s/step with a step size of 0.008° from 25° to 105° 2Θ. During the measurement, the samples were rotated. The integration was carried out with the software DIFFRAC.EVA (Version 3.1; Bruker AXS Advanced X-ray Solutions GmbH, Karlsruhe, Germany) (Fig. 3).

Figure 3
figure 3

Observed diffraction patterns of the unprocessed references with different yttria contents. Since the reflections at around 30° are much more intense than the other reflections, we rescaled the diffraction pattern at 33° (27% of the scattering intensity) and 70° (4% of the scattering intensity).

Full size image

To identify the abrasive and polishing particles present in the polishing tools, the tools were cut into small fragments and mounted on a height-adjustable stage to provide the flat surface necessary for X-ray diffraction. A Bruker D8 Series 1 Advance diffractometer was employed in a reflective Bragg–Brentano configuration. The system was outfitted with a Cu-sealed tube running at 40 kV/20 mA, Goebel mirror optics, a 0.2-mm divergence slit, a fixed knife edge to suppress air scattering, and an energy discriminating LYNXEYE first Gen detector in scanning mode. The LYNXEYE detector’s cut off energy was adjusted to remove most of the Fe-fluorescence that could be present in the sample to enhance the peak to background ratio. The measuring procedure involved 67 s/step, with each step being 0.01° taken between 3° and 60° 2Θ.

Rietveld refinement

For Rietveld refinement, the software TOPAS 4.2 (Bruker AXS Advanced X-ray Solutions GmbH, Karlsruhe, Germany) was used. Structural data were obtained from the literature reference [8]. For refinement, a structural model which includes the monoclinic phase M, the two tetragonal phases T (yttria-lean [55]) and T′′ (“pseudocubic” [24]; yttria-rich [55]), the cubic phase C, and the rhombohedral phase R [28] was used. In addition, to simulate the formation of a small grain-sized layer on the surface through grinding or polishing regimes, the tetragonal main phase was divided into two phases with different grain sizes, specifically a larger and smaller grain size (3Y: T, 4Y: T and T′′, 5Y: T′′).

The grain size starting values were extracted from FE-SEM images. A surface roughness correction was applied, if necessary, according to Pitschke et al. [56]. Texture effects were simulated using a preferred orientation approach according to March–Dollase [57, 58].

MicroX-ray computer tomography (µXCT)

MicroX-ray computed tomography (industrial tomograph) was used to image the 3D microstructure of small individual parts (< 2 µm) of each dental tool. The X-ray tube, manufactured by YXLON International GmbH (Hamburg, Germany) (focal spot diameter 0.6 µm, tungsten target), was operated at 140 kV/120 µA without filter. A resolution of 2.7 µm was achieved with the manipulator settings used and the 1621 × N 2D detector (2048 × 2048 pitches, CsI, pitch size 200 µm2) fabricated by PerkinElmer (Waltham, MA, USA).

Results

Phase composition and surface roughness of the dental crowns

Tables 5, 6, and 7 (Appendix) display the phase composition of all dental crowns, whilst Tables 8 and 9 (Appendix) present the mean surface roughness (Sa) and pair comparisons. We will only cover significant effects that correspond to the thesis formulated.

Prior to polishing (unprocessed: 3Y/4Y/5Y_1, glaze firing 3Y/4Y/5Y_2, wet bur: 3Y/4Y/5Y_3, dry bur: 3Y/4Y/5Y_7).

The unprocessed crowns (3Y/4Y/5Y_1) consisted of visible grains approximately 300–400 nm in diameter (Fig. 4).

Figure 4
figure 4

FE-SEM images show an unprocessed dental crown (3Y_1; left) with visible grain boundaries and a polished dental crown (3Y_4; NTI polisher; right) with a layer covering the previously visible grain boundaries, which are now invisible.

Full size image

As demonstrated in prior studies [48], glaze firing led to decreased intensities due to the amorphous content and did not result in any additional phase transformations (Appendix; Tables 5, 6, and 7).

The burs induced a reflection broadening (shown in Fig. 5) regardless of whether water was used, suggesting smaller crystallites on the surface of the dental crown and a shift in preferred orientation already described in a previous study [8]. The use of a wet or dry bur did not affect the phase composition.

Figure 5
figure 5

Comparison of the observed diffraction patterns: the unprocessed and the dry-grinded dental crowns (left) and the wet- and dry-grinded dental crowns (right).

Full size image

Polishing (wet grinded: 3Y/4Y/5Y_3-5 and dry grinded: 3Y/4Y/5Y_8-10).

Processing using a polishing regime caused the previously visible grains to be transformed into very small crystallites. As shown in Fig. 4 (right), the grain boundaries are no longer visible on the polished surface.

Mechanical processing resulted in a small monoclinic phase fraction for the 3Y/4Y_3-10 specimen. The polishing regimes NTI (wet: 3Y/4Y/5Y_4; dry: 3Y/4Y/5Y_8, Fig. 6a) and Komet (wet: 3Y/4Y/5Y_5; dry: 3Y/4Y/5Y_9; Fig. 6b) induced a signal next to the tetragonal and cubic reflections at 30°, which we assign to the rhombohedral phase (Fig. 6d; Tables 5, 6, and 7 in Appendix) [8, 28,29,30,31, 47, 51, 59]. The EVE polishing regime (3Y/4Y/5Y_6; 3Y/4Y/5Y_10) did not induce a rhombohedral phase transformation (Fig. 6c).

Figure 6
figure 6

Comparison of observed reflections at approximately 30° between the grinded (bur) and polished dental crowns: a NTI polishing regime (5Y_8; rhombohedral phase formation), b Komet polishing regime (5Y_9; rhombohedral phase formation), c EVE polishing regime (5Y_10; no phase transformation, and d exemplary comparison of the diffraction patterns before (5Y_7) and after the Komet polishing regime (5Y_9) with subsequent phase mapping obtained by Rietveld refinement.

Full size image

Polishing with any dental polishing regime (NTI, Komet, or EVE) resulted in a significant reduction of surface roughness (Sa) for all yttria concentrations (Fig. 7; Tables 5, 6, and 7 in Appendix). However, when used wet (3Y/4Y/5Y_6) or on a 3Y-PSZ surface (3Y_10; Table 5 in Appendix), the polishing regime from EVE reduced surface roughness (Sa) significantly less than the polishing regimes from NTI and Komet.

Figure 7
figure 7

Median of arithmetical mean heights (Sa) of the dental crown surface roughness after the processing steps 3: bur (wet), polishing regimes: 4: NTI, 5: Komet, 6: EVE, 7: bur (dry), polishing regimes: 8: NTI, 9: Komet, 10: EVE (see also Fig. 2, Table 2). The marked treatments have a significantly different surface roughness (Sa) than the preceding bur (Appendix, Tables 8 and 9).

Full size image

Material composition and texture of the polishers

Phase composition by XRD

All polishers consist of different types of inorganic crystalline phases dispersed in an organic matrix.

All three NTI polishers contained diamond (C) and corundum (Al2O3). The coarse NTI green polisher also contained eskolaite (Cr2O3), anatase (TiO2), several modifications of silicium carbide (SiC; e.g. moissanite), and some weak amorphous signals (probably organic matter). The medium NTI blue polisher contained rutile (TiO2) and amorphous signals, whilst the fine yellow polisher contained anatase (TiO2).

Diamond (C) and anatase (TiO2) were identified in the Komet polishers. Furthermore, the coarse Komet blue polisher contained several modifications of silicium carbide (SiC), such as moissanite, and some amorphous signals.

In both EVE polishers, diamond (C) and some amorphous signals (probably organics) were found. Furthermore, rutile (TiO2) was detectable in the coarse EVE green polisher, and hematite (Fe2O3) in the fine EVE red polisher, as well (Table 3).

Table 3 Overview of the main phase composition incl. loss on ignition (LOI: weight loss in %) in the fillers based on XRD analysis
Full size table

Polisher texture

The medium (blue; Appendix, Fig. 12) and coarse NTI green and Komet blue polishers contain larger abrasive particles in their easily reversible deformable elastomer matrix, which are significantly smaller for the coarse polisher EVE green. Additionally, the coarse and medium polishers of Komet blue and NTI green/blue have a network structure with low densities, resulting in a less dense texture compared to the coarse polisher EVE green (Fig. 8).

Figure 8
figure 8

Comparison of the surface texture with CLSM focus variation (top images) and µXCT (bottom images) cross-sectional images of NTI green, Komet blue, and EVE green coarse polishers. The maximal particle sizes (distance between two points) on cross section based on CLSM were NTI green: 69 µm, Komet blue: 70 µm, EVE green: 45 µm.

Full size image

Bright areas in cross sections of µXCT indicate a high density, which can be attributed to abrasive particles. Dark areas, in contrast, indicate regions with low atom density, characteristic of either pores or the organic matrix. The CLSM images suggest that the larger, dark areas correspond to the presence of pores. The content of abrasive particles is evaluated on the basis of different images, the phase composition in combination with the loss on ignition (LOI, Table 3).

The NTI green and Komet blue coarse polishers feature larger abrasive particles (Fig. 8) made of diamond and SiC (Table 3) when compared to fine polishers (Fig. 9) with anatase probably as a colouring agent. It should be noted that the abrasive particles used by EVE green are exclusively diamond and are reduced in size (Fig. 8).

Figure 9
figure 9

Comparison between the texture of the fine polishers NTI yellow, Komet white, and EVE red using cross-sectional images taken by CLSM focus variation (top images) and µXCT (bottom images). Equivalent maximal diameter of the abrasives based on µXCT: NTI yellow: 20 µm, Komet white: 90 µm, EVE red: 40 µm. Max. distance on cross section based on CLSM: NTI green: 69 µm, Komet white: 6 µm, EVE green: 8 µm.

Full size image

The fine polisher EVE red displayed only very small (8 µm) single (low content, see Fig. 9 and Table 3, LOI: 94/98%) visible abrasive particles (Table 3, diamond as an abrasive with hematite probably as a colouring agent; some larger inhomogeneities with high density are also present). The particle size of Komet (6 µm) white is similar to EVE red (8 µm), but the content is maybe higher (Table 3, LOI: 89/90%) and the composition is different (Table 3, diamond and rutile—black areas—as abrasives). The largest particles (36 µm) and the highest content (Table 3, LOI: 41/75%) were found in NTI yellow (Table 3, diamond and corundum as abrasives). In contrast to the NTI yellow fine polisher, the Komet white and EVE red fine polishers do not contain a visible pore system (Fig. 9).

Discussion

All polishers consist of abrasive particles that are statistically distributed in a porous, elastic organic matrix (Table 3). However, there are significant differences in the grain size and material composition of the pores and abrasives. This results in different behaviour in terms of phase transformation and surface roughness reduction.

Surface roughness

The null hypothesis can be rejected as all polishers significantly reduced the surface roughness. However, the EVE polishing regime yielded a significantly slower surface roughness reduction than the NTI and Komet polishing regimes when used wet and/or on a 3Y-PSZ surface.

The EVE polishing regime featured small diamond abrasives without the larger SiC or corundum abrasives. The abrasive content is lower, and the individual small abrasive particles are distributed in their easily reversible, deformable elastomer matrix, so that less energy is likely to be transferred to the surface than with the other methods. This would result in a more ductile-based material removal mechanism (D-MRM; chip formation through deformation [60, 61]) than fracture-based material removal mechanism (F-MRM; propagation and intersection of cracks [60]). Since the D-MRM removes material more slowly on brittle materials with high E-modulus, the EVE polishing regime would reduce surface roughness more slowly than the other regimes on brittle materials like Y-PSZ [60,61,62].

Additionally, wet polishing causes a lubrication effect, which reduces the material removal rate and changes the surface roughness less. This approach is probably negated by the large corundum and/or SiC abrasive particles of the coarse NTI green and Komet blue polishers [62, 63].

Therefore, the EVE polishing regime featured a slower substance removal and might reduce the surface roughness less than the other regimes when used wet or on a 3Y-PSZ surface (Fig. 10; Appendix, Tables 8, 9).

Figure 10
figure 10

Exemplary images of the surfaces (217 µm × 289 µm) of the dental crowns made from 3Y-PSZ after exposure to dry grinding and different polishing regimes (3Y_7-10).

Full size image

Phase composition

New reflections were observed following polishing with the NTI (3Y/4Y/5Y_4 and 3Y/4Y/5Y_8) and Komet (3Y/4Y/5Y_5 and 3Y/4Y/5Y_9) polishing regimes, which can be assigned to the rhombohedral phase [8, 28,29,30, 47, 51, 59]. Our own calculations demonstrate that the rhombohedral phase could result in a 15% increase in volume (monoclinic: 9%) and a shear of 120° (monoclinic: 99°) compared to the tetragonal and cubic phases [8].

The rhombohedral phase can be indicated best by its most distinct reflections 1 -1 -1R, 1 1 1R, and -1 1 1R which form a peak (“shoulder”) below the tetragonal and cubic (1 1 1C and 1 0 1T) reflections around 30° [28, 29, 51, 59]). Other relevant reflections include 1 -3 -2R and 1 -2 -3R (around 35°), 1 2 -4R and 4 1 0R (around 50°), 1 2 5R, 5 2 0R, 4 1 3R, and 2 4 1R (around 60°), and 2 4 4R (around 73°).

The rhombohedral “shoulder” has also been depicted in diffraction patterns by several other works as a result of polishing regimes or ageing processes, but not indicated as a rhombohedral phase [40, 42, 64].

Mechanical processing in general, such as milling, sandblasting, or removal of material by grinding or polishing, may cause phase transformations that lead to changes in the mechanical and optical properties of Y-PSZ (rhombohedral phase: [8]; monoclinic phase: [9, 10, 23, 24]).

Since phase transformations are enabled by mechanical stress or pressure and temperature, this means that these boundary conditions for phase transformation into the rhombohedral phase were not reached for the EVE polishing regime.

An explanatory approach includes the smaller abrasive particles (Fig. 8) of the coarse polisher EVE green compared to NTI green and Komet blue. These particles are not firmly embedded in the elastomer matrix and can dislodge or avoid acting on the elastomer matrix more easily. Consequently, these smaller abrasive particles exert weaker forces on the ceramic surface than larger ones. Other works [8, 24], where individual polishers were studied, support this theory.

Furthermore, the abrasive particles differed in their composition (Table 3) and particle size (Figs. 8 and 9). Whilst the NTI green and Komet blue coarse polishers included additional silicon carbide abrasives, the EVE green coarse polishers relied solely on diamond abrasive particles (diamond was used in all polishers). Beside the abrasives, the coarse polishers NTI green and Komet blue contained anatase (TiO2), whereas EVE green contained rutile (TiO2).

Another approach includes the denser elastomer matrix of the polisher EVE green compared to the NTI green/blue and Komet blue coarse and medium polishers. A polisher texture that is less dense and more open increases the specific surface area of the abrasives on the tool surface and thus the rotating impact force on the dental crown surface. This higher mechanical stress might contribute to reach the phase boundary of the rhombohedral phase.

In summary, the coarse polisher EVE green (compared to its NTI green and Komet blue counterparts) is likely to transfer less energy to the dental crown surface due to its overall finer structure. This would result in not exceeding the rhombohedral phase boundary and a slower material removal (higher surface roughness, Figs. 7, 10) when used wet and/or on a 3Y-PSZ (highest modulus and flexural strength) dental crown.

The number of polishers (NTI: 3, Komet/EVE: 2) and the chemical composition of the abrasive particles (Table 3) embedded in the elastomer matrix had only a small influence on the phase composition of the Y-PSZ.

Grain size and crystallite size

In addition to the symmetry and orientation of the individual phases in Y-PSZ, the absorption and scattering of light are also affected by the size of the grains and crystallites. The grain size (FE-SEM images; Fig. 4) and the crystallite size (Rietveld refinement/XRD; Fig. 5) were reduced by the burs and the polishers. The FE-SEM images suggest that the layer is of low thickness; thus, the effect on optical (e.g. decreased translucency) and mechanical (lower critical stress intensity) properties may also be minimal [65]. It is also possible that a multilayered structure with deformed grains, as described by Munoz-Tabares et al. [49] in a TEM study, is present.

Clinical implications

Due to its high hardness, the adjustment, and especially the adequate polishing of Y-PSZ, is a challenge for the dentist. When polishing fixed dental prostheses wet in the patient’s mouth, it is important to achieve an adequate reduction in surface roughness as quickly as possible to keep the preparation time as short as possible. Since the NTI and Komet polishing regimes reduced roughness significantly faster than the EVE polishing regime and showed no lubricating effect (“surface roughness” section; Figs. 7, 10; Appendix, Tables 8, 9), they were well suited for this task. However, the rhombohedral phase formation induced by the NTI and Komet polishing regimes could affect the long-term behaviour and possibly induce pre-stress effects.

The results suggest that the mechanical and optical properties at the top surface of Y-PSZ are affected not only by the roughness but also by the change of phase composition during the chairside polishing of the surface. Since the results of our XRD studies give only surface-near information, further investigations on the influence on the final bulk properties (e.g. translucency, bulk mechanics) of dental restoration are necessary.

Limitations

One of the study’s limitations is the restricted selection of dental technical tools and Y-PSZs that we employed, along with the inadequately defined information depth below the surface which could only be assessed with typical Bragg–Brentano geometry XRD setups like the one used in our investigation. The resolution of the image depends on various parameters, including the wavelength, intensity of the primary X-ray beam, absorption in the material under analysis, and, most importantly, the diffraction angle. The path length of the incident and diffracted X-rays, as well as the intensity of the individual reflection, also play a significant role in determining the resolution. Typically, the resolution ranges from a few micrometres to several tens of micrometres.

Conclusions

In this study, the effect of different polishing regimes on the phase composition and surface roughness of 3Y-/4Y-/5Y-PSZ was investigated. Therefore, explanatory approaches based on characteristic parameters of dental tools were established. Considering the limitations of this study, the following conclusions can be formulated:

  1. 1.

    Processing with a bur induced a small monoclinic phase fraction and a reflection broadening indicating smaller crystallites at the surface. Based on the current knowledge, reduced translucency and low, lower critical stress intensity may be possible outcomes.

  2. 2.

    The polishing regimes of NTI and Komet induced a rhombohedral phase transformation and a significant reduction in surface roughness (Sa). Smoothing of surface roughness and imperfections is expected to result in improved strength. The rhombohedral phase formation could be either beneficial (pre-stressing) or detrimental to strength. Since the volume expansion and shear are greater than that of the monoclinic phase, the effects might be stronger.

  3. 3.

    The EVE polishing regime did not cause a rhombohedral phase formation but resulted in a significantly slower substance removal (lower surface roughness reduction) than the other polishing regimes when used with water and/or on a 3Y-PSZ surface.

Reasons might be the smaller abrasive particle sizes, the higher density of the easily reversible deformable elastomer matrix, and the absence of SiC or corundum of the EVE polishing regime as compared to the NTI and Komet polishing regimes. Further research is needed to investigate the extent of rhombohedral phase within the material. In addition, the possible mechanical and optical changes (e.g. translucency) and the long-term behaviour of the dental crowns should be investigated experimentally.