Novel laser textured surface designs for improved zirconia implants performance

Journal Pre-proof Novel laser textured surface designs for improved zirconia implants performance D. Faria, S. Madeira, M. Buciumeanu, F.S. Silva, O. Carvalho PII:

S0928-4931(18)33617-8

DOI:

https://doi.org/10.1016/j.msec.2019.110390

Reference:

MSC 110390

To appear in:

Materials Science & Engineering C

Received Date: 28 November 2018 Revised Date:

27 October 2019

Accepted Date: 31 October 2019

Please cite this article as: D. Faria, S. Madeira, M. Buciumeanu, F.S. Silva, O. Carvalho, Novel laser textured surface designs for improved zirconia implants performance, Materials Science & Engineering C (2019), doi: https://doi.org/10.1016/j.msec.2019.110390. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier B.V.

Novel laser textured surface designs for improved zirconia implants performance D. Fariaa; S. Madeiraa*; M. Buciumeanub; F. S. Silvaa; O. Carvalhoa

a

Center for Micro-Electro Mechanical Systems (CMEMS-UMinho), University of

Minho, Campus de Azurém, 4800-058 Guimarães, Portugal b

Department of Mechanical Engineering, Faculty of Engineering, “Dunărea de Jos”

University of Galaţi, Domnească 47, 800008 Galati, Romania

*Corresponding author: Madeira, S. email: [email protected]; Center for Micro-Electro Mechanical Systems (CMEMS-UMinho), University of Minho, Campus de Azurém, 4800-058 Guimarães, Portugal

Tel.: +351253510732 Fax.: +351253516007

Abstract The development of new surface designs to enhance the integration process between surgically placed implants and biological tissues remains a challenge for the scientific community. In this way and trying to overcome this issue, in this work, laser technology was explored to produce novel textures on the surface of green zirconia compacts produced by cold pressing technique. Two strategies regarding line design (8 and 16 lines design) and different laser parameters (laser power and number of laser passages) were explored to assess their influence on geometry and depth of created textures. The produced textures were evaluated with Scanning Electron Microscopy (SEM) and it was observed that well-defined textured surfaces with regular geometric features (cavities or pillars) were obtained by laser combining different strategies lines design and parameters. The potential of proposed textures was also evaluated regarding surface wettability, friction performance (static and dynamic coefficient of friction evolution) against bone, aging resistance and flexural strength. Results demonstrated that all the produced textures display a super hydrophilic or hydrophilic behavior. Regarding the friction behavior, it was experimentally observed a high initial static coefficient of friction (COF) for all produced textures. Concerning the aging resistance, all the textured surfaces revealed a low monoclinic content, less than 25% after 5 h of hydrothermal aging. The flexural strength results showed that the mechanical resistance of zirconia was not significantly compromised with the laser action. Based on the obtained results, it is possible to prove that the processing route used for manufacturing the new and different surface designs (cold pressing technique followed by laser texturing) showed to be particularly effective for the production of zirconia implants with customized surface designs according to the properties required in a specific application. These new surface designs besides to enhance the surface wettability and 2

also to improve the fixation at the initial moment of the implantation, do not significantly compromise the resistance to aging and the mechanical performance of zirconia. Hence, a positive impact on the long-term performance of the zirconia implants may be expected with the proposed novel laser textured surface designs.

Keywords: Zirconia; Laser surface texturing; Surface wettability; Coefficient of friction; Aging resistance; Flexural strength. 1. Introduction Zirconia (ZrO2) is a ceramic biomaterial that has been increasingly used as a metal substitute for biomedical applications including implants (such as dental implants [1], hip and knee prostheses [2,3]) and prosthesis structures (e.g. dental crowns [1,3]). In particular, it has been reported that the addition of 2–3 mol% of yttria (Y2O3) to the zirconia chemical composition, fully stabilise the tetragonal phase, resulting in small 100% metastable tetragonal grains [4–6]. High flexural strength (up to 900-1200 MPa) [7], high fracture toughness (7–10 MPa m1/2) [8], wear resistance, optical and biological properties, biocompatibility, and low bacterial affinity are the remarkable properties reported for 3% yttria-stabilized tetragonal zirconia polycrystal (3Y-TZP) in literature [4,7,9–11]. Furthermore, zirconiabased material is considered as biologically inert. This means that when implanted it is not able to start an adverse reaction in host tissues after implantation to surrounding tissue. However, some studies have reported that this characteristic compromises osseointegration since bone cannot naturally grow on zirconia surface [12,13]. In this sense, some strategies have been studied regarding surface modification trying to tailor surface properties through roughness improvements and patterns production [14,15].

3

Among existent surface treatments (such as machining, polishing, ultraviolet light treatment, sol-gel processing, lithography and coatings [16–18]) sandblasting followed by an acid-etching is known as the gold standard surface treatment present in the most implants available on market due to its homogeneous micro-roughness [19,20]. However, there are some risks associated to this surface treatment which might be detrimental to the clinical survival of zirconia-based implants such as the contamination of implant surface with residual ceramic particles; the spreading of bacteria colonization and the alteration on the structure and properties of the underlying material [21]. On the other hand, no different and localized geometric features and structures can be produced by this technique to promote customized areas according clinical and mechanical requirements. In this context, the laser technology has emerged as a promising processing technique to create complex microstructures at micro-nano scale (regarding surface texture design) due to its versatility and ability to remove material quickly, with low waste and without surface contamination [22,23] since there is no contact between the biomaterial and laser (energy transfer from the laser to the zirconia occurs through the irradiation) [14,16,24,25]. Besides this, laser technology offers other important advantages some of which comprise high speed operation, local treatment, high precision and low cost [26,27]. Additionally, some studies have reported that this technique changes the way as biological fluids interact with the biomaterial and also enhances zirconia surface wettability improving, therefore, the bone-implant interface [5,16,28]. A variety of laser sources such as Nd:YAG (Neodymium-doped yttrium aluminum garnet), CO2 (Carbon dioxide) and excimer lasers have been used for texturing ceramics and each one of them have its own wavelength of absorption. 4

According to Samant et al. [29] the Nd:YAG laser is the most widely used for texturing ceramics surfaces owing to its high energy density and small focused spot. Furthermore, it has also been reported that pulsed lasers like Nd:YAG lasers are more suitable for texturing ceramics due to the ability to control the processing parameters as compared to continuous mode [24]. It has also been showed that the zirconia surface roughness values have significantly increased after being textured with Nd:YAG laser as well as the surface wettability [21,28]. Despite the advantages of laser technology comparing to conventional surface treatments, some undesirable side-effects like poor thermal shock resistance that conduct to initiation and propagation of micro-cracks; spatter and heat-affected zones have been reported when laser is used for texturing hard and brittle materials, like sintered zirconia [5,23,30]. Additionally, it has been reported that these micro-damages can accelerate the aging of zirconia (destabilize the zirconia metastable tetragonal phase and transforming it to monoclinic) [5,31]. To the authors’ best knowledge, literature is very scarce in studies involving the texturing by laser of green zirconia compacts. Notwithstanding, there is a study conducted by Liu et al. [23] that successfully used a pulsed laser to produce micro-scale grooves on green zirconia compacts. In their study, the influence of laser parameters on the heat-affected zone around the machined grooves and on micromorphology of laserirradiated surfaces were studied and better surface quality was obtained with frequency below 40 Hz, power below 6 W, and scanning velocity above 200 mm/s. Nevertheless, no further studies using this approach are currently available in the literature for other surface topographies. In this sense, the present work goes further and presents novel surface topography designs with different geometrical definitions produced on green zirconia compacts by means of laser technology. A detailed study focused on the 5

influence of laser parameters and strategies is presented aiming to study their influence on geometry and depth of created textures. The potential of produced textures is also evaluated regarding surface wettability, friction performance (static and dynamic coefficient of friction evolution) against bone, aging resistance and flexural strength. In the end of the paper, the global performance of the new designed textures is graphically displayed to make it easier to assess the most adequate set of laser parameters and strategies for the development of zirconia implants with customized surface designs, according to the properties required in a specific application.

2. Experimental details 2.1. Starting material In this study, a 3 mol% yttria-stabilized zirconia spray-dried powder (TZ-3YSBE, Tosoh Corporation, Japan), with high purity (99%) and theoretical density of 6.05 g/cm3 was used. This commercial powder is formed by spherical agglomerates with an average size of 60 µm and contains small crystallites that are about 36 nm in diameter. Fig. 1 a presents the Scanning Electron Microscopy (SEM) image of TZ-3YB-E powder in agglomerate form, while in Fig. 1 b is shown the cumulative weight distribution curve as a function of powder size (according to manufacturer). The chemical composition of TZ-3YB-E powder is listed in Table 1.

6

Fig. 1. TZ-3YB-E powder: (a) SEM image and (b) cumulative weight distribution curve as a function of size (according to manufacturer) [32]. Table 1. Chemical composition of TZ-3YB-E powder (according to manufacturer) [32]. Elements Y 2O 3 ZrO2 + HfO2 + Y2O wt% 5.15±0.20 >99.9 *Calculated value ----- 100 - (SiO2 + Fe2O3 + Na2O).

Al2O3 0.25 ±0.10

SiO2 ≦ 0.02

Fe2O3 ≦ 0.01

Na2O ≦ 0.04

2.2. Production of green zirconia compacts The green zirconia compacts were produced by cold pressing technique, as schematically represented in Fig. 2. The compaction of the TZ-3YB-E powder was performed on a steel cylindrical mold with an internal diameter of 18 mm and 30 mm of height. First, the TZ-3YB-E powder was introduced into the mold (Fig. 2a) and then, it was applied a pressure of 200 MPa for 30 s (Fig. 2b). After that, the pressure was released evenly and the mean dimensions of the obtained green zirconia compacts (Fig. 2c) were 18 mm of diameter and 2 mm of thickness.

7

Fig. 2. Schematic representation of the cold pressing technique: (a) introducing the TZ-3YB-E powder into the steel mold; (b) overall arrangement by applying pressure; (c) green zirconia compact.

2.3 Laser surface texturing of green zirconia compacts After production of green zirconia compacts, the surfaces texturing was performed by using a Nd:YAG laser (OEM Plus, SISMA, Italy), with an output power of 6W, a spot size of 3 µm and a pulse width of approximately 35 ns. The laser beam was focused on the surface of green zirconia compacts by a focusing unit containing a fused quartz lens with a nominal focal length of 160 mm, a fundamental wavelength of 1.064 µm and a maximum pulse energy of 0.3 mJ/pulse. It is important to highlight that the surfaces texturing was carried out in normal air under atmospheric pressure and assisted with a jet of air braided to remove debris produced during laser processing. Firstly, the design of crosslines was defined in a computer-aided design system and then it was engraved on the surface of green zirconia compacts according to the specific laser parameters. The schematic representation of the laser texturing is illustrated in Fig. 3. 8

Fig. 3. Schematic representation of the laser texturing process.

In this study, two different strategies were performed to produce the surface textures: Strategy 1: Z8PmLn – zirconia surface texturing using a laser pattern with 8 lines (see Fig. 4a), where m corresponds to the laser power and n corresponds to the number of laser passages. Strategy 2: Z16PmLn – zirconia surface texturing using a laser pattern with 16 lines (see Fig. 4b), where m corresponds to the laser power and n corresponds to the number of laser passages.

9

Fig. 4. Schematic representation of lines design on (a) strategy 1 and (b) strategy 2.

In both strategies, different laser parameters such as laser power and number of laser passages were tested to produce the textures. The remaining parameters such as scan speed, fill spacing, wobble amplitude and wobble frequency were maintained constant, as listed in Table 2. The time required for texturing each sample, with a surface area of 162.9 mm2, varied from 592 to 4980 s, according to the laser parameters (laser power and number of laser passages) and strategies (Z8 and Z16).

10

Table 2. Summary of the laser texturing parameters tested.

Experiment Z8P0.3L1 Z8P0.9L1 Z8P1.5L1 Z8P1.5L2 Z8P1.5L4 Z8P1.5L8 Z16P0.3L1 Z16P0.9L1 Z16P1.5L1 Z16P1.5L2 Z16P1.5L4 Z16P1.5L8

Strategy

Laser power [W] 0.3 0.9

1 1.5 0.3 0.9 2 1.5

Number of laser passages

Scan speed [mm/s]

Fill spacing [mm]

Wobble amplitude [mm]

Wobble frequency [Hz]

200

8

8

550

200

8

8

550

1 2 4 8 1 2 4 8

For comparison purposes, zirconia samples as sintered (hereafter referred to as AS), and zirconia samples subjected to the conventional sandblasting and acid-etching treatment (henceforward referred to as SB-AE) were produced. Two steps were performed to produce the SB-AE samples with a moderate roughness topography (between 2 and 4 µm) [20,33]: Step 1. The zirconia surfaces were sandblasted with 250 µm spherical alumina particles for 30 s. The sandblasting was carried out at a constant pressure of 6 bars at a distance of 10 mm from the blasting nozzle and with an impact angle of 90°. After that, the samples were ultrasonically cleaned in isopropyl alcohol during 5 min. Step 2. The sandblasted zirconia surfaces were immersed in hydrofluoridic acid (48%) during 30 min. and subsequently cleaned in an ultrasonic bath immersed in isopropyl alcohol for 5 min. It should be noted that the conventional sandblasting and acid-etching treatment was performed in sintered zirconia samples.

11

The surface roughness of the AS and SB-AE samples was measured using a contact profilometer (Surftest SJ 201, Mitutoyo, Tokyo, Japan) composed by a sharp diamond stylus with 2 µm of diameter. Six measurements at randomly selected areas on each samples group were performed with a speed of 0.25 mm/s during 7 mm of length to obtain roughness mean values and standard deviations. The measured surface roughness parameter was the average roughness, Ra, (average obtained between peaks and valleys distance). 2.4. Sintering of green zirconia compacts After laser texturing, the green zirconia compacts were sintered using a high temperature furnace (Zirkonofen 700, Zirkonzahn, Italy) with a sintering temperature of 1500˚C, a heating and cooling rate of 8 ˚C/min and 2 h of holding time. The temperature was chosen based on sintering temperatures (T = 1350 – 1550 °C) required by supplier (Tosoh Corporation, Japan) [32] for conventional sintering processes. The same conditions of sintering were applied to the AS and SB-AE samples. After sintering, zirconia samples with 14.4 mm in diameter and 1.8 mm thick were obtained. All samples were ultrasonically cleaned in isopropyl alcohol for 10 min. and then in distilled water for 10 min. to remove any loose debris or surface contamination. 2.5. Optical and scanning electron microscopy To analyze powder morphology, evaluate the surface of laser-generated textures as well as the microstructure of the AS and SB-AE samples, Scanning Electron Microscopy (SEM - Nova NanoSEM 200, FEI, Netherlands) equipped with an Energy Dispersive Spectrometer (EDS) system was used. SEM was also used to observe the fracture zone of the samples after ball-on-three-balls tests. The morphology and chemical composition of the zirconia surfaces and the bone plates were evaluated by 12

means of SEM/EDS, after the friction tests. An optical microscope (Leica DM 2500 equipment from Leica Microsystems (Wetzlar, Germany)) connected to a computer for image processing, using Leica Application Suite software was used to acquire the cross section view of textures. Afterwards, Image J software was used to measure the depth of the textures based on the acquired images. 2.6. Surface wettability In order to determine the wettability properties of the AS, SB-AE and laser textured samples, contact angle measurements for each group were performed. The measurements were performed by the sessile drop method using the optical goniometer OCA 15 plus (Dataphysics, Germany). For the experiments, droplets of ultrapure deionized water (at 18.2 Ohm) with a volume of 5 µL and dosing rate of 2.5 µl/s were dispensed from a micrometric syringe, brought into contact with the surface and allowed to stabilize for 15 s before the reading is taken. Fig. 5 presents representative pictures of the shapes of a droplet, before contact with the surface (Fig. 5 a), at the moment of contact with the surface and lifting of the micrometric syringe (Fig. 5 b) and 15 s after contact with the surface (Fig. 5 c). The behavior of the droplets was video recorded with a frame rate of 52 frames per second at a resolution of 768 x 576 pixels. The experiments were conducted at room temperature (20°C) and a minimum of three contact angle readings were taken on each sample always at the same location (in the centre of the sample). It is worth noting that prior to each measurement (one measurement per day), all the samples were ultrasonically cleaned in isopropyl alcohol for 1 min. and after each measurement were stored in distilled water until to next cleaning in order to maintain surface properties [34]. Before the contact angle measurements, the samples stored in liquid were dried in vacuum, for about 30 min.

13

Fig. 5. Representative pictures of the shapes of a droplet: (a) before contact with the surface (b) at the moment of contact with the surface and lifting of the micrometric syringe and (c) 15 s after contact with the surface. The arrows symbolize the micrometric syringe movement during the measurements.

2.7. Details of the friction tests The friction tests were employed to evaluate the friction performance (static and dynamic coefficient of friction evolution) and the structural integrity of the produced textures in an environment mimicking the insertion of an implant. To this end, the laser textured samples were tested against real bone. In the tests, the pins were the studied zirconia samples, while the counter body plates were the rectangular bone plates (4x16x20 mm) cut from a fresh femoral part of a young bovine. The bone plates were mounted in an acrylic electrochemical cell attached to the tribometer. The tests were carried out on a reciprocating pin-on-plate tribometer (Bruker-UMT-2, USA). An illustration of the friction tests performed in this work is shown in Fig. 6. The friction tests were performed at 100 N loading (nominal), a frequency of 1 Hz and a 5 mm total stroke length for 17 s. A Phosphate Buffered Solution (PBS) fluid with the standard composition found in the literature [35] was used as lubricant in this study once it is a simulated body fluid commonly used in biomedical research for implant materials applications [36–40]. 14

Fig. 6. Schematic representation of the friction test: (a) Initial and final static friction test and (b) dynamic friction test.

The friction tests were performed in three stages as follows: (i) measurement of the initial static coefficient of friction (COF) by a single displacement in one direction (Fig. 6 a – solid arrow); (ii) determination of the dynamic coefficient of friction during 17 s of reciprocating sliding (Fig. 6 b); (iii) measurement of the final static coefficient of friction by a single displacement in the opposite direction of the first reading (Fig. 6 a – dashed arrow). At least three samples of each laser experiment were tested using a new bone plate for every new test. For comparison purposes, the friction performance of AS and SB-AE samples was also assessed.

15

2.8. Low-temperature autoclave aging Samples of each group (AS, SB-AE and laser textured samples) were subjected to a hydrothermal attack in order to evaluate the stability of the tetragonal zirconia phase. It is important to highlight that prior to hydrothermal degradation, the chemical composition of each sample was assessed with X-ray diffraction using a Bruker AXS D8 Discover diffractometer with a Cu-Kα radiation (λ=1.54060Å). Diffraction data were collected from 27º to 32º (2θ), with a step size of 0.04º and counting time of 1 s/step. The accelerated aging tests were performed at 134ºC in water steam, under 2 bars pressure during 5 h. After this time, the samples were again characterized by X-ray diffraction in order to evaluate the monoclinic phase content. Monoclinic surface phase transformation was identified by using the software for identification peaks diffraction (X’pert High Score Plus, Netherlands). Afterwards, Toraya equation [41,42] (equation 1) was used to quantify the percentage of monoclinic phase present on all tested samples.

= Where,

111 = 28.2º and

(1) 111 = 31.5º represent the intensities of the peaks

along the 111 and the 111 diffraction planes of the monoclinic phase, respectively and

101 = 30.2º represents the intensity of the peak along the (101) diffraction

plane of the tetragonal phase. The volumetric fraction,

, of monoclinic phase was then calculated by equation 2:

=

.

.

(2)

16

2.9. Details of the ball-on-three-balls (B3B) tests The flexural strength of all tested conditions (AS, SB-AE samples and laser textured samples) was measured using the ball-on-three-balls test. Tests were performed at room temperature (~23 °C) in a servohydraulic machine (Instron 8874 MA, USA), equipped with a 25 kN capacity load cell at a loading rate of 1 mm/min. The experimental procedure was made in a custom-made stainless steel apparatus. In the test, the intended side of the sample was positioned in the samples holder on top of the three supporting steel balls equidistant from its center. The opposite surface of the sample was centrally loaded by a fourth ball coupled to the pin crosshead of the testing machine. The test started from a small preload and then the load increased until sample fracture. After that, the fracture load was recorded and the maximum tensile stress σ

, which occurs in the centre of the sample, on the opposite side of the

loading ball was calculated according to the equation 3 [43,44]:

=

+

"# "

⁄$

"& ⁄$

"( ⁄$

"' ⁄$ &

)1 +

$ *$

+

,-

(3)

where F is the maximum force at fracture (N), t the sample thickness (mm), . the sample radius (mm), . the support radius (Ra=5 mm), and the parameters c0 to c6 refers to fitting factors for the geometrical correction term (c0= -17,346 c1= 20,774 c2= 622.62 c3= 76.879 c4= 50.383 c5= 33.736 c6= 0.0613) [44]. It is worth noting that the latter parameters values are dependent of Poisson’s ratio of the studied material (ʋ= 0.3) [45].

17

2.10. Statistical analysis The one-way ANOVA with post hoc Bonferroni multiple comparison test was used (GraphPad Prism v, GraphPad Software, La Jolla, California, USA) to assess the statistically significant differences in the depth of created textures; water contact angle; coefficient of friction; monoclinic content and flexural strength between the different laser experiments (laser parameters and strategies). The analysis was also performed in comparison to the SB-AE samples (representing the conventional treatment). The level of significance was set at Pvalues < 0.05. In all the diagrams, the respective significances were marked with solid lines in the case of strategy 1 (Z8) and dash dot lines in the case of strategy 2 (Z16).

18

3. Results and Discussion 3.1. Morphology and microstructure characterization As previously mentioned, two different laser strategies (strategy 1 – Z8 and strategy 2 – Z16) as well as two different laser parameters (laser power and number of laser passages) were tested to produce the surface textures. SEM micrographs (120x and 500x magnifications) of the laser-generated textures obtained according to the set of the experiments listed in Table 2 are shown in Fig. 7 c-n. First, it is worth noting from Fig. 7 c-n the high quality on the laser texturing once well-designed (without distortion), reproducible and equally spaced textures were produced for all tested laser experiments. This adequate geometrical accuracy is due to the fact that the laser pulses are applied in a sequential way, i.e. returning to each texture a number of times rather than applying all the pulses in a burst. In fact, it has been reported that in the burst mode the material in and around the texture remains hot for a longer time and the molten material can flow to cover the texture [46,47]. This suggests that local heating of the zirconia samples plays an important role in the laser texturing process. Additionally, no micro-cracks; spatter and heat-affected zones are observed on the laser textured surfaces. This fact demonstrates that texturing zirconia surfaces in their green (unsintered) state by laser is an effective way for producing textures with different and complex geometries without the introduction of harmful defects to zirconia mechanical resistance [3,48]. Apart from this, in general, it can be verified from Fig. 7 c-n that textures with different geometrical definitions were achieved using different laser strategies. In the case of Z8 (Fig. 7 c-e and Fig. 7 i-k), cavities were produced through the removal of material by laser, while pillars were obtained from Z16 (Fig. 7 f-h and Fig. 7 l-n).

19

Fig. 7. SEM micrographs of the zirconia surface: (a) and (a1) AS; (b) and (b1) SB-AE. Lasergenerated textures obtained with strategy Z8 or Z16, as a function of: (c-h) and (c1-h1) laser power (number of laser passages remains constant – L1), and (i-n) and (i1-n1) number of laser passages (laser power remains constant – P1.5).

20

Regarding the influence of laser parameters on the produced textures, it can be observed from Fig. 7 c-h that when increasing the laser power from P0.3 to P1.5 there is an increase on the depth of machined volume (in particular for P1.5), for both laser strategies, once higher energy is irradiated in same area (considering the same number of laser passages – L1). Additionally, as shown in Fig. 7 c-h, the increase of machined volume (depth) is more pronounced on Z16 than Z8. This observation is in line with the results presented in Fig. 8 a. As it can be observed, in the case of Z8 there is a slight increase on depth of cavities from P0.3 to P1.5 (from 11.9 to 22.4 µm). A similar behavior (increase on depth of pillars) can be also seen in the case of Z16 from P0.3 to P0.9 (from 14.4 to 20.1 µm) but followed by a more pronounced increase on depth of pillars for P1.5 (67.9 µm) (p < 0.05). Similar trend was also observed in Fig. 7 i-n concerning the effect of laser passages on the textures depth, i.e. increasing the number of laser passages from L1 to L8, while maintaining constant the laser power (P1.5) led to high marking depths, as a result of the higher irradiated energy on those areas. The opposite effect was achieved for a smaller number of laser passages (L1). Furthermore, it can be also observed from Fig. 7 i-n that the textures produced with Z16 have much higher depth than the ones produced with Z8. This observation is in accordance with the results presented in Fig. 8 b. It can be observed, in the case of Z8, an increase on depth of cavities from L1 to L8 (from 22.4 to 39.8 µm) while in the case of Z16 there is a significant increase on depth of pillars from L1 to L8 (from 67.9 to 251.8 µm) (p < 0.05). From these results, it can be concluded that the number of laser passages has more influence on depth of created textures as compared to the laser power. This means that the depth of textures is strongly dependent of the amount of energy used to produce them. 21

Fig. 8. Depth of the produced textures for both laser strategies (Z8 and Z16), as a function of (a) laser power (number of laser passages remains constant – L1), and (b) number of laser passages (laser power remains constant – P1.5). The asterisk (*) indicates statistically significant differences (p < 0.05).

The microstructure of AS and SB-AE samples was also recorded for comparison purposes. It can be observed from Fig. 7 a that the microstructure of AS samples is dense and almost totally homogeneous, with an average roughness (Ra) of 0.14 ± 0.03 µm, whereas the SB-AE samples (Fig. 7 b) present a moderate roughness throughout the surface, Ra of 2.01 ± 0.07 µm, usually found in most implants available on the market [20,33]. Additionally, by comparing the surface of SB-AE samples with the surface of laser textured samples, it is possible to conclude that SB-AE samples have a homogenous texture throughout the surface while the laser textured samples display well-defined textured surfaces with regular geometric features, according to the properties required in a specific application. In fact, there is evidence in the literature that the cell adhesion and proliferation on biomedical surfaces are commonly achieved by well-defined geometric features like pits [49], grooves [50], pillars [51], cavities and ridges [52], ranging from micro- to macroscale [53]. Therefore, to induce a controlled response in the bone-implant contact region, the implant contact surface needs to be 22

well controlled [53]. Most of the common techniques used in implants surfaces modification (e.g. sandblasting and chemical etching) generate randomly distributed surface features which can lead to random bone cell orientations and consequently contribute to scar tissue formation [53,54]. Instead of those random surface features, production of regular and geometrically defined features by laser will create highly regulated and reproducible surfaces, which will aid to have more precise control over the human body response [53]. 3.2. Surface wettability The surface wettability has been considered as a key factor on the biological response of implantable biomedical materials. In this work, the surface wettability was assessed by water contact angle (WCA). Fig. 9 displays the average WCA values for both laser strategies (Z8 and Z16) as a function of (a) laser power and (b) number of laser passages. For comparison purposes, the WCA values of AS and SB-AE samples were also measured.

Fig. 9. Average water contact angle for both laser strategies (Z8 and Z16), as a function of (a) laser power (number of laser passages remains constant – L1), and (b) number of laser passages (laser power remains constant – P1.5). The asterisk (*) indicates statistically significant differences (p < 0.05). 23

Analyzing the obtained results, a WCA value of 21.3 ± 0.9° was found for AS samples which means that zirconia material have a great ability to spread water right after contact (hydrophilic character) as expected and reported in the literature [22]. Additionally, it can be seen from Fig. 9 that improvements on wettability can be reached by performing surface treatments on zirconia. Regarding standard treatment (SB-AE samples), a WCA value of 8 ± 0.9° (superhydrophilic behavior) was obtained. However, a significant improvement on wettability was achieved for the textured samples as a consequence of the changes (mainly physical) induced by laser in the zirconia surface. The WCA values of textured samples ranged from 0º to 13.2º (superhydrophilic and hydrophilic behavior). Concerning the influence of the surface topography, obtained by different laser parameters, on the wettability several aspects can be pointed. First, it may be clearly observed from Fig. 8 a and Fig. 9 a that when increasing the laser power from P0.3 to P1.5, maintaining the number of laser passages constant (L1), an increase on the depth of created textures was obtained resulting in significant improvement on wettability through WCA reduction, for both laser strategies. Additionally, it can be seen from Fig. 9 a that Z8P0.9L1 and Z8P1.5L1 experiments have further wettability as compared to the SB-AE samples (p < 0.05), demonstrating therefore that laser texturing is a promising technique to modify zirconia surfaces as compared to the conventional treatment available on the market. In fact, some studies have reported improvements on wettability by performing laser treatments on ceramics surfaces [28,55,56]. On the other hand, considering the best conditions of both strategies (Z8P1.5L1 and Z16P1.5L1), the second approach was to investigate the influence of the surface topography, obtained by the number of laser passages, on the wettability. It can be seen from Fig. 8 b and Fig. 9 b that when increasing the number of laser passages (from L1 24

to L8) and the laser power is maintained constant (P1.5), a more pronounced effect on depth of created textures was achieved resulting in further improvements on wettability through WCA reduction. In the case of Z16, it can be observed a sharp increase on wettability from L1 to L2 followed by a small increase for the remaining laser passages (L4 and L8) (p < 0.05), whereas on Z8 there is an overall gradual increase on wettability from L1 to L8 (p < 0.05). Additionally, it should be noted that Z8P1.5L4; Z8P1.5L8 and Z16P1.5L8 experiments possess a great wettability since between 0 – 0.3 s the droplet is totally absorbed by the surface immediately after contacting it. Therefore, the most important conclusion in this section is that the increase of laser power leads to an increase on depth of created textures that in turns leads to an improvement on the wettability of zirconia, especially on deeper textures resulting from high number of laser passages. 3.3. Static and dynamic coefficient of friction It is well known that the coefficient of friction has a major influence on stability and, as a consequence, on the lifetime of the implants [11,22,57]. According to the literature, during the lifetime of the implant, a low COF is preferable to avoid bone loss and space closure [11,22]. However, during implantation (insertion of the implant), it seems that a high COF may lead to a greater initial stability and, as a consequence, to a long-lasting stability of the implant [58–61]. From a tribological point of view, the static friction represents the force necessary to start the movement. This force will always be higher than the one to maintain this motion (dynamic COF) because in the beginning of the implant insertion, the implant is positioned and, firstly, at the implant surface level, there is an opposition of the bone to implant material movement (rough samples or

25

samples with surface steps) [11,22]. Thus, a higher COF during the insertion of the implant can indicate a better fixation at the initial moment of implantation. In the present study, to perform the friction tests and consequently analyse the coefficient of friction of laser textured samples against bovine bone, a 100 N load was selected, once, it is the load typically found in the literature for assessing the friction performance (static and dynamic coefficient of friction evolution) and the structural integrity of the zirconia implants against bone [11,22]. Fig. 10 shows the average static (initial and final) and dynamic COF values for both laser strategies (Z8 and Z16), as a function of (a) laser power and (b) number of laser passages. For comparison purposes, the static and dynamic COF values of AS and SB-AE samples were also assessed. It can be seen from Fig. 10 that the initial static COF of the AS samples was around 0.61, while in the case of SB-AE samples the obtained static COF was higher around 0.71. Indeed, in the present study, it was expected an increase of the initial static COF value in the case of SB-AE samples (conventional treatment) due to the presence of peaks and valleys throughout the surface (Ra = 2.01 ± 0.07 µm). Thus, it was required a large tangential force to initiate relative motion of SB-AE samples on the bone as compared to the one for AS samples. On the other hand, as the surface of the samples textured by laser is not homogenous (complex localized textures with cavities and pillars), the friction force required to initiate relative motion was even higher than in the case of the SB-AE samples, proved by the high COF values in the case of the textured samples. Concerning the influence of laser parameters and strategies on the (initial and final) static and dynamic COF evolution, several aspects may be pointed.

26

Fig. 10. Average static and dynamic coefficient of friction (COF) for both laser strategies (Z8 and Z16), as a function of (a) laser power (number of laser passages remains constant - L1), and (b) number of laser passages (laser power remains constant - 1.5 W): (top row - initial static COF; middle row - dynamic COF and bottom row - final static COF). The asterisk (*) indicates statistically significant differences (p < 0.05).

27

First, it can be clearly observed from Fig. 10 a (top row) that when increasing the laser power (from P0.3 to P1.5), while maintaining the number of laser passages constant (L1), there is an increasing tendency in the initial static COF values for both laser strategies (from 0.81 to 1.01 in the case of Z8 and from 0.54 to 0.89 in the case of Z16) (p < 0.05). The highest initial static COF values were obtained for Z8P1.5L1 and Z16P0.9L1. In this sense, when changing from conventional treatment (SB-AE samples) to the new proposed alternatives (laser textured samples), the initial static COF value was increased from 0.71 to 1.01 in the case of Z8 and from 0.71 to 0.89 in the case of Z16 (p < 0.05). As expected, the dynamic COF values (Fig. 10 a - middle row) when increasing the laser power (from P0.3 to P1.5), while maintaining the number of laser passages constant (L1) were always smaller as compared to the ones of the initial static COF (Fig. 10 a – top row), for both laser strategies. This means that the required friction force to maintain the relative motion is slightly lower as compared to the one to initiate the relative motion. Additionally, no significant differences were observed in the dynamic COF values, among all the studied cases (around 0.50). This is a normal tendency, but in this particular case when the textured samples are sliding against bone, the bone adheres to the surface (cavities or pillars are filled with bone) and there may be regions with contact bone-to-bone. A slightly lower value was obtained in the case of SB-AE samples (0.41). The obtained dynamic COF values are in agreement with the values reported into the technical literature when zirconia was tested in wet conditions (from 0.40 to 0.60) [10,22]. The final static COF values when increasing the laser power (from P0.3 to P1.5), while maintaining the number of laser passages constant (L1) are presented in Fig. 10 a 28

(bottom row). It can be seen that the final static COF values are higher than the ones of the dynamic COF (Fig. 10 a - middle row), among all the studied cases, and slightly lower than the ones of the initial static COF (Fig. 10 a – top row). The higher final static COF values as compared to the ones of the dynamic COF, among all the studied cases, may be explained based on the fact that it was necessary a higher energy to start the motion (to move bone against zirconia samples), after the dynamic tests. This means that in all the cases there was a good adhesion of the zirconia samples to the bone. The lower final static COF values as compared to the ones of the initial static COF can be explained based on the fact that in the case of the final static COF, the motion is started mostly between bone (cavities or pillars are filled with bone) and bone, while in the case of the initial static COF the motion is initiated between the peaks of cavities or pillars and bone. Additionally, it can be observed from Fig. 10 a (bottom row) a gradual increase in the final static COF values, for both laser strategies (from 0.65 to 0.82 in the case of Z8 and from 0.51 to 0.66 in the case of Z16) (p < 0.05). It can be also seen from Fig. 10 a (bottom row) an increase in the final static COF value from the conventional treatment to the new proposed alternatives (33% in the case of Z8 (p < 0.05) and 17% in the case of Z16). Thus, based on all these observations it can be concluded that combining a high laser power (P1.5) with the strategy 1 (Z8) is a good choice for increasing the COF value, and as a consequence improving the adhesion of the bone to the zirconia textured surfaces. Taking into account such observations, the second approach was to investigate the influence of the number of laser passages (L1, L2, L4 and L8) on the (initial and final) static and dynamic COF evolution, while maintaining constant the laser power (P1.5). It can be seen from Fig. 10 b (top row) that when increasing the number of laser passages (from L1 to L8), there is an increase in the initial static COF values (from 1.01 to 1.09), 29

for both laser strategies (Z8 and Z16). In the case of Z16, by increasing the number of laser passages from L1 to L2 was possible to increase more the initial static COF from 0.83 to 1.07 (p < 0.05). The same initial static COF value (1.07) was obtained for Z16P1.5L8. It can be also seen from Fig. 10 b (top row) that a significant increase in the initial static COF value was obtained from the conventional treatment to the new proposed alternatives (46% in the case of Z8 and 48% in the case of Z16) (p < 0.05). Regarding the variation of dynamic COF values with increasing the number of laser passages (Fig. 10 b – middle row), it can be seen that there is a slight increase from 0.51 (L1, L2) to 0.59 (L4, L8), for both laser strategies. Concerning the influence of the number of laser passages on final static COF values (Fig 10 b – bottom row), it can be verified that also in this case the values are higher than the ones of the dynamic COF, among all the studied cases, and slightly lower than the ones of the initial static COF. It can be also seen that in the case of Z8, by increasing the number of laser passages from L1 to L2 it was possible to increase more the final static COF value from 0.82 to 0.91. The same final static COF value (0.91) was also obtained in the case of Z16P1.5L8. It should be noted that was obtained an increase on the final static COF values for all new proposed alternatives, except for Z16P0.3L1. The adhesive component of friction seems to be more active for Z8P1.5L2 and Z16P1.5L8, as in these cases was obtained the highest final static COF value (0.91). Thus, when changing from conventional treatment to the new proposed alternatives, the final static COF increased 50%. Fig. 11 shows the SEM micrographs of textured samples after the friction tests against bone, for both laser strategies (Z8 and Z16). For comparison purposes, SEM micrographs of AS and SB-AE samples were also recorded.

30

Each SEM micrograph has three zones, the first two zones are the micrographs of the zirconia samples with two different magnifications (500 µm and 100 µm), while in the third zone is presented the micrograph of the bone surface (500 µm magnification). The sliding directions are indicated by the red arrows on the micrographs. From Fig. 11, it may be observed that in all the tested conditions there is the presence of adhesive mechanism on zirconia samples (bone adhered to zirconia surface) and abrasive mechanism on the bone surfaces (grooves aligned with the sliding direction). The adhesion of the bone to the zirconia surface can be clearly observed on the SEM micrographs presented in Fig. 11, but it can be also proved by the EDS analysis performed in all studied samples. Fig. 12 shows a representative EDS spectrum.

31

Fig. 11. SEM micrographs of zirconia surfaces after the friction tests against bone: AS (a); SB-AE (b); laser textures for both laser strategies (Z8 and Z16) as a function of laser power (c-e and i-k) and number of laser passages (f-h and l-n). The bones surface used against the zirconia samples are also presented (a1- n1).

32

Fig. 12. EDS spectrum of Z8P0.3L1 experiment.

Zr (Zirconium), Y (Yttrium), Ca (calcium) and P (phosphorus) were the main elements detected on the zirconia samples. Zr and Y are coming from the zirconia samples, while Ca and P from the bone. The bone transfer was observed in all studied cases, except in the case of the AS samples (Fig. 11 a). The laser textured surfaces presented a higher bone transfer (Fig. 11 c-n) as compared to the SB-AE surfaces (localized bone adherence – Fig. 11 b). Dantas et al. [62], Moura et al. [22] and Marques et al. [11] also confirmed a high amount of bone adhered to zirconia surfaces after performing sliding tests. Scarano et al. [63] analyzed the bone response to zirconia implants inserted in male rabbits and results showed great amount of newly formed bone and osteoblastic presence on the zirconia surfaces. Besides the chemical affinity of the bone to the zirconia surface [64], in the case of the textured samples it was also created a mechanical connection due to the presence of local specific features. Additionally, it can be also seen from Fig. 11 c-n that these laser textures are filled by bone, but not totally covered once it is possible to see the ridges of cavities or pillars. 33

Moreover, it should be noted that all textured surfaces maintained their integrity after friction tests, except in the case of Z8P1.5L8. Indeed, in this laser experiment, it can be observed some regions where the created pillars did not resist to friction forces which occurred during the friction test (Fig. 11 k). From Fig. 11 c-h, it can be seen that when increasing the laser power (from P0.3 to P1.5) the grooves on the bone surface present a lot of thin grooves in the case of strategy 1 (Z8) for P1.5 (Fig. 11 e). In the case of strategy 2 (Z16), the grooves are larger and no significant changes are observed by the increase of laser power. Nevertheless, it seems that a better response can be obtained combining a high laser power (1.5 W) with the strategy 1 (Z8). This observation is in line with the results presented in Fig. 9 a (bottom row), since the highest final static COF (1) was obtained for Z8P1.5L1. On the other hand, the grooves on the bone surface with the increase of the number of laser passages (from L1 to L8) are more pronounced. In fact, it can be observed from Fig. 11 i-n that the depth of the grooves on bone surface is increasing with the number of laser passages. This behavior seems to be more active in the case of the strategy 2 (Z16). It can be also observed from Fig. 11 n that in the case of Z16P1.5L8 there is the most severe abrasive mechanism of the bone, evidenced by the great amount of bone transferred to the laser textured surface. Also in this laser experiment the final static COF has the highest value (see Fig. 9 b). From this section, it can be concluded that all the cavities or pillars created by laser have a great influence on the COF values. In this sense, by combining the suitable laser parameters and strategies, it can be possible to improve the primary stability and, as a consequence, to increase the lifetime of the zirconia implants.

34

3.4. Analysis of monoclinic content after low-temperature autoclave aging Some studies have reported that laser treatment can induce tetragonal to monoclinic phase transformation of zirconia, however, until to now this matter remains unclear [65–67]. In this sense, aiming to evaluate the zirconia phase after laser texturing, X-ray diffraction analysis were performed and the monoclinic content was assessed. Additionally, the monoclinic content was also evaluated after 5 h of lowtemperature autoclave aging once this aging time represents the range of the lifetime expected for implants (10 – 20 years in vivo). Fig. 13 presents the average of monoclinic content for both laser strategies (Z8 and Z16), as a function of laser power and number of laser passages (a) before and (b) after 5 h of low-temperature autoclave aging. For comparison purposes, the monoclinic content of AS and SB-AE samples was also assessed.

35

Fig. 13. Average monoclinic content for both laser strategies (Z8 and Z16), as a function of laser power (number of laser passages remains constant - L1) and the number of laser passages (laser power remains constant - 1.5 W) (a) before and (b) after 5 h of low-temperature autoclave aging. The asterisk (*) indicates statistically significant differences (p < 0.05).

From Fig. 13 a, it can be observed that for 0 h of hydrothermal aging, monoclinic phase was found for all tested conditions. The monoclinic content of the AS samples was around 0.92 ± 0.13%, whereas in the case of SB-AE samples the monoclinic content was significantly higher around 7.79 ± 0.35%. Inokoshi et. al [65] also reported that sandblasting treatment on zirconia surfaces result in higher monoclinic content as compared to the AS samples. Concerning the influence of laser texturing on tetragonal to monoclinic phase transformation of zirconia some aspects can be pointed. First, it can be observed from 36

Fig. 13 a (top row) that when increasing the laser power from P0.3 to P1.5, while maintaining the number of laser passages constant (L1), there is a decreasing tendency in the monoclinic content for both laser strategies, from 2.79 ± 0.66 % to 1.19 ± 0.18 % in the case of Z8 and from 3.64 ± 0.23 % to 2.13 ± 0.45 % in the case of Z16 (p < 0.05). The lowest monoclinic content was obtained for P1.5 in both laser strategies. Therefore, it is possible to claim that when changing from conventional treatment (SB-AE samples) to the new proposed alternatives (laser textured samples), the monoclinic content was significantly decreased from 7.79 ± 0.35 % to 1.19 ± 0.18 % in the case of Z8 and from 7.79 ± 0.35 % to 2.13 ± 0.45 % in the case of Z16 (p < 0.05). Based on such observations, the second approach was to verify the influence of the number of laser passages (L1, L2, L4 and L8) on the monoclinic content. It can be seen from Fig. 13 a (bottom row) that when increasing the number of laser passages, maintaining the laser power constant (P1.5), there is an increasing tendency in the monoclinic content for both laser strategies (from 1.19 ± 0.18 % to 2.05 ± 0.28 % in the case of Z8 (p < 0.05) and from 2.13 ± 0.45 % to 3.03 ± 0.21 % in the case of Z16. The highest monoclinic content was obtained for Z16P1.5L8. Therefore, it is possible to state that the content of monoclinic phase increases with number of laser passages and this increase is in turn strongly dependent on the time of laser texturing. From these results, it can be concluded that although the laser textured samples have a higher monoclinic content as compared to the AS samples, they present greater aging resistance when compared to the SB-AE samples. The same tendencies were found after 5 h of hydrothermal aging, however, the monoclinic content was substantially higher, as it can be seen in Fig. 13 b. Yet, despite this increase, the obtained results are in accordance with ISO 13356:2008, which states

37

that after 5 h of accelerated aging the acceptable monoclinic content must not overcome 25% [68]. 3.5. Flexural strength results Considering that the flexural strength is a mechanical property determining to investigate the influence of laser parameters and strategies on the mechanical performance of laser textured zirconia surfaces, a biaxial bending strength test of brittle materials designated ball-on-three-balls test was performed. According to the ISO 13356:2008 (Implants for surgery – ceramic materials based on yttria-stabilizes tetragonal zirconia (Y-TZP)), the flexural strength of zirconia should be greater than 500 MPa [68]. Fig. 14 shows the average flexural strength values for both laser strategies (Z8 and Z16), as a function of (a) laser power and (b) number of laser passages. For comparison purposes, the flexural strength of AS and SB-AE samples was also assessed.

Fig. 14. Average flexural strength values for both laser strategies (Z8 and Z16), as a function of (a) laser power (number of laser passages remains constant – L1), and (b) number of laser passages (laser power remains constant – P1.5). The asterisk (*) indicates statistically

significant differences (p < 0.05).

38

Analyzing the results presented in Fig. 14, flexural strength values of 795 ± 87 MPa and 858 ± 110 MPa were obtained for AS and SB-AE samples, respectively, which are in agreement with the standard defined by the ISO 13356:2008. Within the laser textured samples, the flexural strength values ranged from 331 ± 37 MPa to 692 ± 41 MPa, which means that there are significant differences between them. By comparing flexural strength values of laser textured samples with AS and SB-AE samples, a decrease was verified, i.e. the mechanical resistance of zirconia was significantly affected with laser texturing. Despite this reduction on flexural strength values, it is important to highlight, that only Z8P1.5L4; Z8P1.5L8; Z16P1.5L4; Z16P1.5L8 experiments are bellow of the standard defined by the ISO 13356:2008. Regarding the influence of laser parameters on flexural strength values several aspects can be pointed. First, it can be seen from Fig. 14 a that when increasing the laser power (from P0.3 to P1.5) and the number of laser passages remains constant (L1) there is a slight decrease on flexural strength values, for both laser strategies. Apart from this, no significant differences were found on the flexural strength values between both laser strategies, with exception of Z8P1.5L1 and Z16P1.5L1 experiments. A significant decrease on flexural strength (p < 0.05) can be also observed in Fig. 14 b when the number of laser passages is increased from L1 to L8 and the laser power remains constant (P1.5), as a result of the higher laser ablation on those areas. Despite this substantial reduction, flexural strength values above 500 MPa were found for Z8P1.5L2 (521 ± 20 MPa) and Z16P1.5L2 (559 ± 27 MPa) experiments. Thereby, from these results, it is possible to conclude that although there is a decrease on flexural strength values for laser textured samples, most of them have flexural strength values above of the standard defined by the ISO 13356:2008. This

39

means that the produced textures from these laser experiments are potential candidates to apply in areas of implant where there is a mechanical request. Fig. 15 presents the fracture zone of the different tested conditions (AS, SB-AE and laser textured samples). It can be observed from Fig. 15 that, in all conditions, failure started from the bottom part of the samples, which corresponds to the tensile surface in the flexural test. The fracture surfaces exhibited typical fracture patterns of ceramic materials, namely the presence of hackle lines i.e. pronounced lines that form without a specific cause, frequently owing to variations in material's microstructure [69]. Additionally, it can be also observed from Fig. 15 that the fracture surfaces revealed the absence of detectable porosity within the samples.

40

Fig. 15. SEM micrographs of fracture zone of the samples: (a) AS; (b) SB-AE; laser-generated textures obtained with strategy Z8 or Z16, as a function of: (c-h) laser power (number of laser passages remains constant – L1), and (i-n) number of laser passages (laser power remains constant – P1.5).

41

In general, all the results of this work demonstrated that the cold pressing technique followed by laser texturing is a promising approach to develop novel textures designs on the zirconia surface. These new surface designs besides to enhance the surface wettability and also to improve fixation at the initial moment of the implantation, do not significantly compromise the resistance to aging and the mechanical performance of zirconia. Therefore, two 3D decision plots were designed (see Fig. 16) to easily correlate the depth of produced textures with (Fig. 16 a) water contact angle and initial static COF and (Fig. 16 b) monoclinic content after 5 h of aging and flexural strength. Based on the analysis of these two plots, it is possible to select a specific range of properties required in a specific application and, in turns, to produce zirconia implants with customized surface designs.

42

Fig. 16. 3D plots correlating the depth of produced textures (µm) with: (a) water contact angle (º) and initial static COF and (b) monoclinic content after 5 h of aging (%) and flexural strength (MPa).

43

Conclusions The following detailed conclusions can be drawn from this work: •

Laser technology showed to be an effective and versatile method to produce different and complex textured zirconia surfaces under green state, without compromise zirconia mechanical performance.

•

The laser parameters and strategies showed to have a great influence on geometrical definition and depth of created textures.

•

Regarding the laser strategy, cavities were obtained from strategy 1 (Z8), while pillars were created from strategy 2 (Z16).

•

The number of laser passages was found to have more pronounced effect on depth of created textures as compared to laser power. The deeper textures were obtained by strategy 2.

•

All produced laser textures revealed super hydrophilic or hydrophilic behavior. However, the surface topography obtained by different laser parameters and strategies showed to have a pronounced influence on wettability.

•

Higher marked depths were achieved using high laser powers which results on improved wettability, being this tendency more pronounced on strategy 1.

•

Further improvements on wettability were achieved when deeper textures were created by increasing the number of laser passages. A sharp increase on wettability from L1 to L2 followed by a small increase for the remaining laser passages (L4 and L8) was observed on strategy 2, while an overall gradual increase on wettability from L1 to L8 was verified on strategy 1.

44

•

The textured zirconia surfaces showed to have high initial static COF, indicating that these new designs can lead to a better fixation at the initial moment of implantation. Furthermore, in general, for both laser strategies, an increase on initial static COF was obtained when deeper textures were obtained by increasing the laser power and number of laser passages.

•

For both laser strategies, a less amount of monoclinic phase was obtained when increasing the laser power, while a higher amount of monoclinic phase was found by the increase of number of laser passages. Furthermore, the textures produced by strategy 1 showed to have less monoclinic phase than the ones created by strategy 2. After 5 h of hydrothermal aging, the monoclinic content was less than 25%.

•

A slight decrease on flexural strength was observed in both laser strategies when increasing the laser power. The same tendency, although more pronounced, was observed by the increase of number of laser passages. The majority of textured samples presented flexural strength values above 500 MPa and no significant differences on the flexural strength values between both laser strategies were found.

•

Novel and different zirconia surface designs with enhanced wettability and improved fixation at the initial moment of the implantation were produced, validating this new approach for the development of zirconia implants with customized surface designs.

45

Acknowledgements This work is supported by Fundação para a Ciência e Tecnologia (FCT) through the grants SFRH/BD/146324/2019 and SFRH/BD/112280/2015 and by the projects UID/EEA/04436/2019, NORTE-01-0145-FEDER-000018-HAMaBICo and FunImp, POCI-01-0145-FEDER-030498.

References [1]

Z. Ozkurt, E. Kazazoglu, Zirconia Dental Implants: A Literature Review, J. Oral Implantol. XXXVII (2011) 367–376. doi:10.1563/AAID-JOI-D-09-00079.

[2]

A. Aherwar, A. K Singh, A. Patnaik, Current and future biocompatibility aspects of biomaterials for hip prosthesis, AIMS Bioeng. 3 (2015) 23–43. doi:10.3934/bioeng.2016.1.23.

[3]

S. Mohanty, A.P. Rameshbabu, S. Mandal, B. Su, S. Dhara, Critical issues in near net shape forming via green machining of ceramics: A case study of alumina dental crown, J. Asian Ceram. Soc. 1 (2013) 274–281. doi:10.1016/j.jascer.2013.06.005.

[4]

Y.-D. Cho, J.-C. Shin, H.-L. Kim, M. Gerelmaa, H.-I. Yoon, H.-M. Ryoo, D.-J. Kim, J.-S. Han, Comparison of the Osteogenic Potential of Titanium and Modified Zirconia-Based Bioceramics, Int. J. Mol. Sci. 15 (2014) 4442–4452. doi:10.3390/ijms15034442.

[5]

B. Henriques, D. Fabris, J.C.M. Souza, F.S. Silva, O. Carvalho, M.C. Fredel, J. 46

Mesquita-Guimarães, Bond strength enhancement of zirconia-porcelain interfaces via Nd:YAG laser surface structuring, J. Mech. Behav. Biomed. Mater. 81 (2018) 161–167. doi:10.1016/j.jmbbm.2018.02.031. [6]

D. Faria, J.M. Pires, A.R. Boccaccini, O. Carvalho, F.S. Silva, J. MesquitaGuimarães, Development of novel zirconia implant’s materials gradated design with improved bioactive surface, J. Mech. Behav. Biomed. Mater. 94 (2019) 110–125. doi:10.1016/j.jmbbm.2019.02.022.

[7]

N. Cionca, D. Hashim, A. Mombelli, Zirconia dental implants: where are we now, and where are we heading?, Periodontol. 2000. 73 (2017) 241–258. doi:10.1111/prd.12180.

[8]

C. Piconi, G. Maccauro, Zirconia as a ceramic biomaterial, Biomaterials. 20 (1999) 1–25. doi:10.1016/S0142-9612(98)00010-6.

[9]

M. Buciumeanu, D. Faria, J. Mesquita-Guimarães, F.S. Silva, Tribological characterization of bioactive zirconia composite layers on zirconia structures, Ceram. Int. 44 (2018) 18663–18671. doi:10.1016/j.ceramint.2018.07.094.

[10] M. Buciumeanu, J.R.C. Queiroz, A.E. Martinelli, F.S. Silva, B. Henriques, The effect of surface treatment on the friction and wear behavior of dental Y-TZP ceramic against human enamel, Tribol. Int. 116 (2017) 192–198. doi:10.1016/j.triboint.2017.07.016. [11] A. Marques, G. Miranda, D. Faria, P. Pinto, F. Silva, Ó. Carvalho, Novel design of low modulus high strength zirconia scaffolds for biomedical applications, J. Mech. Behav. Biomed. Mater. 97 (2019) 375–384. doi:10.1016/j.jmbbm.2019.05.005. 47

[12] M. Dehestani, L. Ilver, E. Adolfsson, Enhancing the bioactivity of zirconia and zirconia composites by surface modification, J. Biomed. Mater. Res. - Part B. 100B (2012) 832–840. doi:10.1002/jbm.b.32647. [13] C. Caravaca, L. Shi, S. Balvay, P. Rivory, E. Laurenceau, Y. Chevolot, D. Hartmann, L. Gremillard, J. Chevalier, Direct silanization of zirconia for increased biointegration, Acta Biomater. 46 (2016) 323–335. doi:10.1016/j.actbio.2016.09.034. [14] J. Lawrence, L. Hao, H.R. Chew, On the correlation between Nd:YAG laserinduced wettability characteristics modification and osteoblast cell bioactivity on a titanium alloy, Surf. Coat. Technol. 200 (2006) 5581–5589. doi:10.1016/j.surfcoat.2005.07.107. [15] L. Hao, J. Lawrence, K.S. Chian, Osteoblast cell adhesion on a laser modified zirconia based bioceramic, J. Mater. Sci. Mater. Med. 16 (2005) 719–726. doi:10.1007/s10856-005-2608-3. [16] G. Soon, B. Pingguan-Murphy, K.W. Lai, S.A. Akbar, Review of zirconia-based bioceramic: Surface modification and cellular response, Ceram. Int. 42 (2016) 12543–12555. doi:10.1016/j.ceramint.2016.05.077. [17] S. Ferraris, A. Bobbio, M. Miola, S. Spriano, Micro- and nano-textured, hydrophilic and bioactive titanium dental implants, Surf. Coat. Technol. 276 (2015) 374–383. doi:10.1016/j.surfcoat.2015.06.042. [18] M. Liu, J. Zhou, Y. Yang, M. Zheng, J. Yang, J. Tan, Surface modification of zirconia with polydopamine to enhance fibroblast response and decrease bacterial activity in vitro: A potential technique for soft tissue engineering applications, 48

Colloids Surfaces B Biointerfaces. 136 (2015) 74–83. doi:10.1016/j.colsurfb.2015.06.047. [19] S.-H. Park, K.-S. Park, S.-A. Cho, Comparison of removal torques of SLActive® implant and blasted, laser-treated titanium implant in rabbit tibia bone healed with concentrated growth factor application, J. Adv. Prosthodont. 8 (2016) 110. doi:10.4047/jap.2016.8.2.110. [20] Straumann, Straumann SLA, Sci. Evid, (2011) 1–36. [21] D. Liu, J.P. Matinlinna, J.K.-H. Tsoi, E.H.N. Pow, T. Miyazaki, Y. Shibata, C.W. Kan, A new modified laser pretreatment for porcelain zirconia bonding, Dent. Mater. 29 (2013) 559–565. doi:10.1016/j.dental.2013.03.002. [22] C.G. Moura, R. Pereira, M. Buciumeanu, O. Carvalho, F. Bartolomeu, R. Nascimento, F.S. Silva, Effect of laser surface texturing on primary stability and surface properties of zirconia implants, Ceram. Int. 43 (2017) 15227–15236. doi:10.1016/j.ceramint.2017.08.058. [23] Y. Liu, L. Liu, J. Deng, R. Meng, X. Zou, F. Wu, Fabrication of micro-scale textured grooves on green ZrO2 ceramics by pulsed laser ablation, Ceram. Int. 43 (2017) 6519–6531. doi:10.1016/j.ceramint.2017.02.074. [24] D. Faria, C.S. Abreu, M. Buciumeanu, N. Dourado, O. Carvalho, F.S. Silva, G. Miranda, Ti6Al4V laser surface preparation and functionalization using hydroxyapatite for biomedical applications, J. Biomed. Mater. Res. - Part B. 106B (2018) 1534–1545. doi:10.1002/jbm.b.33964. [25] Z. Rui, W. Yi, A. Xing, M. Bo, W. Teng, L. Zhanqiang, Z. Dong, Fabrication of 49

Micro/Nano-textured Titanium Alloy Implant Surface and Its Influence on Hydroxyapatite Coatings, J. Wuhan Univ. Technol. Sci. Ed. 31 (2016) 440–445. doi:10.1007/s11595-016-1389-5. [26] B.S. Yilbas, Laser treatment of zirconia surface for improved surface hydrophobicity, J. Alloys Compd. 625 (2015) 208–215. doi:10.1016/j.jallcom.2014.11.069. [27] B.S. Yilbas, Laser texturing of zirconia surface with presence of TiC and B4C: Surface hydrophobicity, metallurgical, and mechanical characteristics, Ceram. Int. 40 (2014) 16159–16167. doi:10.1016/j.ceramint.2014.07.047. [28] L. Hao, J. Lawrence, Effects of Nd: YAG laser treatment on the wettability characteristics of a zirconia-based bioceramic, Opt. Lasers Eng. 44 (2006) 803– 814. doi:10.1016/j.optlaseng.2005.08.001. [29] A.N. Samant, N.B. Dahotre, Laser machining of structural ceramics-A review, J. Eur. Ceram. Soc. 29 (2009) 969–993. doi:10.1016/j.jeurceramsoc.2008.11.010. [30] B. Guimarães, D. Figueiredo, C.M. Fernandes, F.S. Silva, G. Miranda, O. Carvalho, Laser machining of WC-Co green compacts for cutting tools manufacturing, Int. J. Refract. Met. Hard Mater. 81 (2019) 316–324. doi:10.1016/j.ijrmhm.2019.03.018. [31] J. Li, L. Ji, Y. Hu, Y. Bao, Precise micromachining of yttria-tetragonal zirconia polycrystal ceramic using 532 nm nanosecond laser, Ceram. Int. 42 (2016) 4377– 4385. doi:10.1016/j.ceramint.2015.11.118. [32] Tosoh Corporation, Tosoh zirconia powder: Specification and typical properties 50

Grades: TZ-3Y-E, 3YS-E, 3YB-E, 3YSB-E, 3YSB-C, (2004) 13. [33] P. Ming, S. Shao, J. Qiu, J. Yang, Y. Yu, J. xi Chen, W. qing Zhu, C. Tang, Superiority of calcium-containing nanowires modified titanium surface compared with SLA titanium surface in biological behavior of osteoblasts: A pilot study, Appl. Surf. Sci. 416 (2017) 790–797. doi:10.1016/j.apsusc.2017.04.152. [34] A. Noro, M. Kaneko, I. Murata, M. Yoshinari, Influence of surface topography and surface physicochemistry on wettability of zirconia (tetragonal zirconia polycrystal), J. Biomed. Mater. Res. - Part B Appl. Biomater. 101B (2013) 355– 363. doi:10.1002/jbm.b.32846. [35] V.G. Pina, A. Dalmau, F. Devesa, V. Amigó, A.I. Muñoz, Tribocorrosion behavior of beta titanium biomedical alloys in phosphate buffer saline solution, J. Mech. Behav. Biomed. Mater. 46 (2015) 59–68. doi:10.1016/j.jmbbm.2015.02.016. [36] M.J. Runa, M.T. Mathew, L.A. Rocha, Tribocorrosion response of the Ti6Al4V alloys commonly used in femoral stems, Tribol. Int. 68 (2013) 85–93. doi:10.1016/j.triboint.2013.09.022. [37] T.M. Manhabosco, S.M. Tamborim, C.B. dos Santos, I.L. Müller, Tribological, electrochemical and tribo-electrochemical characterization of bare and nitrided Ti6Al4V in simulated body fluid solution, Corros. Sci. 53 (2011) 1786–1793. doi:10.1016/j.corsci.2011.01.057. [38] S. Hiromoto, S. Mischler, The influence of proteins on the fretting-corrosion behaviour of a Ti6Al4V alloy, Wear. 261 (2006) 1002–1011. doi:10.1016/j.wear.2006.03.032. 51

[39] S. Karimi, T. Nickchi, A. Alfantazi, Effects of bovine serum albumin on the corrosion behaviour of AISI 316L, Co-28Cr-6Mo, and Ti-6Al-4V alloys in phosphate buffered saline solutions, Corros. Sci. 53 (2011) 3262–3272. doi:10.1016/j.corsci.2011.06.009. [40] V. Swaminathan, H. Zeng, D. Lawrynowicz, Z. Zhang, J.L. Gilbert, Electrochemical investigation of chromium nanocarbide coated Ti-6Al-4V and Co-Cr-Mo alloy substrates, Electrochim. Acta. 59 (2012) 387–397. doi:10.1016/j.electacta.2011.10.085. [41] H. Toraya, M. Yoshimura, S. Sômiya, Calibration curve for quantitative analysis of the monoclinic-tetragonal ZrO2 system by X-ray diffraction, Commun. Am. Ceram. Soc. (1984) 119–121. doi:10.1111/j.1151-2916.1984.tb19715.x. [42] D. Bucevac, T. Kosmac, A. Kocjan, The influence of yttrium-segregationdependent phase partitioning and residual stresses on the aging and fracture behaviour of 3Y-TZP ceramics, Acta Biomater. 62 (2017) 306–316. doi:10.1016/j.actbio.2017.08.014. [43] S.L. Stares, M.C. Fredel, P. Greil, N. Travitzky, Paper-derived hydroxyapatite, Ceram. Int. 39 (2013) 7179–7183. doi:10.1016/j.ceramint.2013.02.062. [44] A. Borger, P. Supancic, R. Danzer, The ball on three balls test for strength testing of brittle discs: Part II: analysis of possible errors in the strength determination, J. Eur. Ceram. Soc. 24 (2004) 2917–2928. doi:10.1016/j.jeurceramsoc.2003.10.035. [45] W. Pabst, G. Tichá, E. Gregorová, Effective elastic properties of aluminazirconia composite ceramics - Part 3. Calculation of elastic moduli of polycrystalline alumina and zirconia from monocrystal data, Ceram. Silikaty. 48 52

(2004) 41–48. [46] J. Byskov-nielsen, J. V. Boll, A.H. Holm, R. Højsholt, P. Balling, Ultra-highstrength micro-mechanical interlocking by injection molding into laser-structured surfaces, Int. J. Adhes. Adhes. 30 (2010) 485–488. doi:10.1016/j.ijadhadh.2010.03.008. [47] B. Henriques, M. Sampaio, M. Buciumeanu, J.C.M. Souza, J.R. Gomes, F. Silva, O. Carvalho, Laser surface structuring of Ti6Al4V substrates for adhesion enhancement in Ti6Al4V-PEEK joints, Mater. Sci. Eng. C. 79 (2017) 177–184. doi:10.1016/j.msec.2017.04.157. [48] B. Su, S. Dhara, L. Wang, Green ceramic machining: A top-down approach for the rapid fabrication of complex-shaped ceramics, J. Eur. Ceram. Soc. 28 (2008) 2109–2115. doi:10.1016/j.jeurceramsoc.2008.02.023. [49] S. Çelen, C. Efeoǧlu, H. Özden, Pulsed Laser-Induced Micro-Pits: As Bone Stabilizers, Phys. Procedia. 12 (2011) 245–251. doi:10.1016/j.phpro.2011.03.130. [50] L. Tiainen, P. Abreu, M. Buciumeanu, F. Silva, M. Gasik, R. Serna, O. Carvalho, Novel laser surface texturing for improved primary stability of titanium implants, J. Mech. Behav. Biomed. Mater. 98 (2019) 26–39. doi:10.1016/j.jmbbm.2019.04.052. [51] M. Nematollahi, D.W. Hamilton, N.J. Jaeger, D.M. Brunette, Hexagonal micron scale pillars influence epithelial cell adhesion, morphology, proliferation, migration, and cytoskeletal arrangement, J. Biomed. Mater. Res. - Part A. 91A (2009) 149–157. doi:10.1002/jbm.a.32202. 53

[52] D.M. Kalaskar, F. Alshomer, Micro- and Nanotopographical Cues Guiding Biomaterial Host Response, in: Situ Tissue Regen., Elsevier Inc., 2016: pp. 137– 163. doi:10.1016/B978-0-12-802225-2.00008-8. [53] S. Mukherjee, S. Dhara, P. Saha, Enhancing the biocompatibility of Ti6Al4V implants by laser surface microtexturing: an in vitro study, Int. J. Adv. Manuf. Technol. 76 (2015) 5–15. doi:10.1007/s00170-013-5277-2. [54] J.L. Ricci, H. Alexander, Laser Microtexturing of Implant Surfaces for Enhanced Tissue Integration, Key Eng. Mater. 198–199 (2001) 179–202. doi:10.4028/www.scientific.net/kem.198-199.179. [55] L. Hao, J. Lawrence, CO2 laser induced microstructure features in magnesia partially stablised zirconia bioceramic and effects thereof on the wettability characteristics, Mater. Sci. Eng. A. 364 (2004) 171–181. doi:10.1016/j.msea.2003.08.011. [56] L. Hao, D.R. Ma, J. Lawrence, X. Zhu, Enhancing osteoblast functions on a magnesia partially stabilised zirconia bioceramic by means of laser irradiation, Mater. Sci. Eng. C. 25 (2005) 496–502. doi:10.1016/j.msec.2005.03.003. [57] B. Ivkovic, M. Djukdjanovi, D. Stamenkovic, The influence of the contact surface roughness on the static friction, Tribol. Ind. 22 (2000) 41–44. [58] M. Jasty, C. Bragdon, D. Burke, D. O’Connor, J. Lowenstein, W.H. Harris, In vivo skeletal responses to porous-surfaced implants subjected to small induced motions, J. Bone Jt. Surg. 79A (1997) 707–714. doi:10.2106/00004623199705000-00010.

54

[59] J.E. Biemond, R. Aquarius, N. Verdonschot, P. Buma, Frictional and bone ingrowth properties of engineered surface topographies produced by electron beam technology, Arch. Orthop. Trauma Surg. 131 (2011) 711–718. doi:10.1007/s00402-010-1218-9. [60] M. Dammak, A. Shirazi-Adl, M. Schwartz, L. Gustavson, Friction Properties at the Bone – Metal Interface : Comparison of Four Different Porous Metal Surfaces, J. Biomed. Mater. Res. 35 (1997) 329–336. doi:10.1002/(SICI)10974636(19970605)35:3<329::AID-JBM7>3.0.CO;2-J. [61] S. Arango, A. Peláez-Vargas, C. García, Coating and Surface Treatments on Orthodontic Metallic Materials, Coatings. 3 (2013) 1–15. doi:10.3390/coatings3010001. [62] T.A. Dantas, S. Roedel, J. Mesquita-Guimarães, P. Pinto, J.C.M. Souza, M.C. Fredel, F.S. Silva, B. Henriques, Sliding behavior of zirconia porous implant surfaces against bone, J. Biomed. Mater. Res. Part B Appl. Biomater. (2018) 1–9. doi:10.1002/jbm.b.34204. [63] A. Scarano, F. Di Carlo, M. Quaranta, A. Piattelli, Bone Response to zirconia ceramic implants; an experimental study in rabbits, J. Oral Implantol. 29 (2009) 8–12. doi:10.4028/www.scientific.net/kem.192-195.403. [64] U. Hempel, T. Hefti, M. Kalbacova, C. Wolf-Brandstetter, P. Dieter, F. Schlottig, Response of osteoblast-like SAOS-2 cells to zirconia ceramics with different surface topographies, Clin. Oral Implants Res. 21 (2010) 174–181. doi:10.1111/j.1600-0501.2009.01797.x. [65] M. Inokoshi, K. Vanmeensel, F. Zhang, J. De Munck, G. Eliades, S. Minakuchi, 55

I. Naert, B. Van Meerbeek, J. Vleugels, Aging resistance of surface-treated dental zirconia, Dent. Mater. 31 (2015) 182–194. doi:10.1016/j.dental.2014.11.018. [66] A. Usumez, N. Hamdemirci, B.Y. Koroglu, I. Simsek, O. Parlar, T. Sari, Bond strength of resin cement to zirconia ceramic with different surface treatments, Lasers Med. Sci. 28 (2013) 259–266. doi:10.1007/s10103-012-1136-x. [67] M. Noda, Y. Okuda, J. Tsuruki, Y. Minesaki, Y. Takenouchi, S. Ban, Surface damages of zirconia by Nd:YAG dental laser irradiation, Dent. Mater. J. 29 (2010) 536–541. doi:10.4012/dmj.2009-127. [68] I. Standard, Implants for surgery —Ceramic materials based on yttria-stabilized tetragonal zirconia (Y-TZP), 2008. [69] J.B. Quinn, G.D. Quinn, J.R. Kelly, S.S. Scherrer, Fractographic analyses of three ceramic whole crown restoration failures, Denatl Mater. J. 21 (2005) 920– 929. doi:10.1016/j.dental.2005.01.006.

56

Highlights • Laser texturing was proven effective for producing novel surface designs on ZrO2. • Laser parameters and strategies had influenced the geometry and depth of textures. • ZrO2 surface textures with improved wettability and primary stability are proposed. • ZrO2 aging resistance and mechanical behavior were not greatly affected by the laser. • This approach showed to be a promising way to produce customized surface implants.

Declaration of interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: