Biofunctionalization of Porous Ti Substrates Coated with Ag Nanoparticles for Potential Antibacterial Behavior

: Ti prosthesis have shown better biological compatibility, mechanical performance, and resistance to corrosion in cases of bone replacements. Nevertheless, fully dense Ti in connection with bone-host tissues show stress-shielding phenomenon that, together with the development of frequent undesirable microbial infections, may lead to implant failures. To overcome these issues, the present study aimed at the development of a novel combination of a chemically functionalized porous Ti substrate with a potentially therapeutic AgNPs coating. Fully dense and porous Ti substrates (30 and 60 vol.%, 100–200 and 355–500 µ m, as spacer particles) were studied. Ti surface was treated with acid or basic medium followed by silanization and deposition of AgNPs by “submerged” and “in situ” methods. In general, for similar porosity, mechanical resistance decreased as pore size increased. Acidic reagent and submerged methodology were the best combination for fully dense Ti substrates. Hence, they were also employed for porous Ti substrates. Depending on the porosity of the substrates, variations can be observed both in the size and degree of agglomeration of the deposited AgNPs, entailing differences in the antibacterial behavior of the samples.


Introduction
The use of prosthetic solutions helps to overcome the limitations of people who have suffered trauma, accidents, or illnesses and are designed to meet functional or aesthetic requirements of the missing or damaged body part [1]. In this sense, its variety and complexity have increased dramatically in the last decades because of a rise in life expectancy and quality of life [2]. This social demand has turned this area into a growing business giving biomaterials for prosthetic applications a critical role in the search for new devices in biomedicine. In this sense, the utilization of materials like metals for healing was an early practice stretching back thousands of years. However, the use of metal implants was first introduced in 1895 by Lane as a metal plate for bone fracture [3]. Unfortunately, at this point all prostheses faced critical problems related to early corrosion or insufficient strength capabilities that were solved in part due to the introduction of stainless steel in 1920s, which had far superior corrosion resistance [4]. Significantly, for biomedical purposes, it is mandatory to evaluate the performance of the type of metal in terms of strength, resistance to fracture, biocompatibility, or stiffness depending on the specific implant or prosthesis application.
That said, ceramics, polymers, composites, or even titanium (Ti)-based alloys have been widely used for hard and soft tissues replacements. In particular Ti prostheses, Interestingly, AgNPs clinical treatment avoids the phenomenon called antibiotic resistance that happens when the human body develops the ability to defeat the drugs designed to kill them; therefore, alternative clinical treatments involving this therapeutically moieties are critical and desirable [50,51], such as the combination including antibacterial protection by phosphate or hydroxyapatite coatings implemented with AgNPs [52,53]. Another example is the use of porous or nanotubular matrices containing this type of nanoparticles [54,55]. Hence, the global rise of antibiotic resistance and the lack of new antibiotics open a new research era including metal-based antimicrobial therapies which is moving toward clinical studies. Although the benefits of the combination of antibacterial coatings of AgNPs onto Ti substrates have been previously reported in the literature [56], this study is based on a novel combination of a chemically functionalized porous Ti substrate with a potentially therapeutic coverture to implement mechanical and pharmacological characteristics. In this sense, the main aim for this investigation was the fabrication, by the space-holder technique, and the characterization of superficially modified porous Ti samples. These substrates presented a potential biomechanical and biofunctional balance, which guaranteed the requirements of the cortical bone tissue, while the proposed functionalization of the surface treatment was expected to improve the antibacterial behavior, particularly with the deposition of AgNPs.

Materials and Methods
The summary workflow of this investigation comparing fully dense, as a reference, and porous Ti samples and the AgNPs synthesis is shown in Figure 1. Furthermore, two different approaches to the biofunctionalization treatments were carried out, either acid or basic hydroxylation, and afterwards, AgNPs were both independently deposited on fully dense and porous Ti discs, to select the best choice between an "in situ" or "submerged" methodology. Substrate manufacturing details, surface modification treatments, microstructural characterization, and antibacterial behavior of the modified discs are described along the following sections.

Fabrication and Characterization of the Ti Substrates
Commercial pure titanium powder (c.p. Ti) grade IV [57] with a mean particle size of d [50] = 23.3 µm [58], provided by SEJONG Materials Co. Ltd. (Seoul, Korea), was used in this study to obtain two types of substrates: fully dense and porous substrates. Fully dense substrates were fabricated by conventional powder metallurgy technique, using an Instron 5505 universal testing machine (Instron, UK) for pressing at 1300 MPa and then, sintering in a molybdenum chamber furnace (Termolab-Fornos Eléctricos, Lda., Águeda, Portugal) at 1300 • C for 2 h, in a high-vacuum condition (10 −5 mbar). On the other hand, porous substrates were obtained by the space-holder technique, using as spacer particles 30 and 60 vol.% of ammonium bicarbonate (NH 4 HCO 3 ), (BA), with a purity of 99%, supplied by Cymit Química S.L. (Barcelona, Spain), and two ranges in size (100-200 µm and 355-500 µm). The spacer particles and c.p. Ti powder were mixed, and once the blend was homogeneous (Turbula ® T2C Shaker-Mixer, Basel, Switzerland for 40 min), it was pressed at 800 MPa with the same equipment. Then, the BA was removed in an oven at 10 −2 mbar: firstly, at 60 • C for 12 h and secondly, at 110 • C for 12 h more. Finally, the porous green discs were also sintered in a molybdenum chamber furnace at 1250 • C for 2 h, in a high-vacuum condition (10 −5 mbar). The manufactured discs presented an approximate average diameter of 11.9 and 2 mm of thickness.
The porosity of all obtained substrates was studied by Archimedes' method and Image analysis (IA). The equivalent pore diameter, pore shape factor, and total and interconnected porosity (D eq , F f , P T , and P i , respectively) [59] were evaluated by these methods. The shape factor was defined as F f = 4πA/(PE) 2 , where A is the area of the pore and PE is its experimental perimeter. Furthermore, the total area occupied by the micropores (less than 50 µm), macropores (associated with the spacer particles used), and the flat area of remaining Ti were evaluated using the image analysis software and with at least five pictures of 5× for each type of substrate. Finally, the mechanical behavior of the porous substrates (dynamic Young's modulus, E d , yield strength, and σ y ), was estimated from the experimental porosity results (at least three measurements for each processing condition) and using fit equations reported in the literature [60].

Synthesis and Characterization of the AgNPs
AgNPs were obtained using the methodology proposed by Lee and Melsel [61], which consisted of dissolving AgNO 3 (2 mM) in deionized water. Under a continuous stirring at 115 rpm and 0 • C, NaBH 4 (1 vol.%) was added, and the mixture was left reacting for 30 min [62,63]. A yellowish suspension of AgNPs was obtained. Chemicals and solvents were purchased from Aldrich Chemicals Co (Madrid, Spain) and were used without further purification.
UV-Vis absorbance spectroscopy was used to corroborate the presence of AgNPs. Measurements were carried out in an Agilent Cary 5000 spectrophotometer at 298 K from 300 to 600 nm with a wavelength accuracy of ±0.3 nm and a spectral bandwidth of 0.5 nm. In order to check the quality (size distribution and composition) of the AgNPs, they were characterized by dynamic light scattering (DLS) using a NanoSizer Nano ZS ® (Malvern Panalytical, UK) and a scanning electron microscopy (SEM) using a JEOL, JSM 6490-LV (Tokio, Japan). Compositional analyses were performed by energy dispersive spectroscopy (EDS) coupled to the scanning electron microscope (EDS-SEM). In addition, transmission electron microscope techniques (TEM) were performed on a FEI Tecnai G2 F20 (FEI, Eindhoven, The Netherlands). For TEM studies, a drop of the solution was deposited on an ultrathin carbon film on a nickel grid.

Modification and Characterization of the Surface of the Ti Substrates
In this work, concerning the Ti surface modifications, two alternative hydroxylation treatments were evaluated: a basic treatment, consisting of a conventional chemical etching with a NaOH solution [64,65], and acid treatment by the exposing the Ti surface to concentrated oxidant sulfuric acid, ("piranha solution", H 2 O 2 /H 2 SO 4 ) [66,67]. To perform the attack treatment, either with the basic medium or with acid medium, the Ti samples were soaked in 5 M NaOH at 60 • C for 24 h or with the "piranha solution" at 75 • C (acid solution: 30% H 2 O 2 and 70% H 2 SO 4 ) for 1.5 h, respectively. Next, after the basic or acid treatments, all samples were subjected to excessive washing with milli-Q water and were completely dried with a nitrogen stream. Then, the hydroxyl groups in Ti surface attacked and displaced the alkoxy groups on the silane to form a covalent O-Si-bond via a silanization reaction, immersing them in a solution of (3-aminopropyl)triethoxysilane (APTES) at 1% (v/v) for 15 min, under constant stirring, followed by a heat treatment in a conventional drying oven (BINDER ® , BINDER GmbH, Tuttlingen, Germany) with convective air at 115 • C for 1.5 h. Additionally, three washes were carried out to rinse out with water and, finally, to dry it with a stream of nitrogen.
Once the best hydroxylation treatment (basic or acid treatment) was chosen in terms of the best homogenization and roughness of the surface followed by a posterior silanization, this study also evaluated as a second implementation two different routes of AgNPs deposition on the hydroxylated Ti surfaces. On the one hand, the synthesis of the AgNPs was performed in the presence of the Ti substrates, and it was called "in situ" synthesis. On the other hand, the suspension of AgNPs was previously obtained, and then, the Ti substrates were submerged in such solution, calling this method "submerged".
The modified surface of the Ti discs was then characterized by different techniques; microstructural features (morphology, roughness, and AgNPs distribution) were studied using a scanning electron microscopy (SEM, JSM 6490-LV by Jeol, Japan), while the chemical composition was analyzed by energy dispersive spectroscopy (EDS) couple to the SEM microscope (EDS-SEM).
Moreover, the antibacterial behavior of AgNPs-coated samples was evaluated following the Kirby-Bauer disc-diffusion method [68] against Staphylococcus aureus (ATCC 25923) as bacterial model. Petri dishes containing Mueller-Hinton medium were covered with a suspension of bacteria (104 CFU/mL) using a sterile hyssop. Immediately afterward, the modified surface of the Ti discs was deposited on top of the seeded medium. Samples were incubated at 37 • C for 24 h in aerobic conditions. Inhibition halos were measured using ImageJ software. For statistical analysis, at least three measurements in three samples were taken for n = 9 of the different Ti discs.

Results and Discussion
The microstructural characterization (porosity parameters) of the uncoated titanium substrates using the Archimedes' method and image analysis, as well as the mechanical properties estimated from the these experimentally values [60,69], are summarized in Table 1. Table 1. Experimental porosity parameters and estimation of the mechanical properties. The pore sizes and contents obtained corroborate the effectiveness of the use of the space-holder technique. The pores size depends both on the size of the spacer applied and on the potential coalescence phenomena. This fact is particularly critical on substrates with higher pore content and/or smaller sizes. From the data shown in Table 1, it can be seen that the pores obtained by the conventional power metallurgy (PM) route are more rounded (pore shape factor values closer to 1) than those obtained with the spacers technique. On the other hand, as expected, there is an inverse relationship between the pore content and the mechanical properties (stiffness and yield strength) of titanium substrates.

Microstructural Characterization Estimated Mechanical Behavior
In general, it has been considered that there are two types of pores: one of them due to the spacer used (content and size) and the other inherent to the sintering conditions. In this sense, pore diameters greater than 50 µm have been identified as macropores and less than 50 µm as micropores. The percentage of area occupied by each one of them, over the total titanium discs surface (2.835 mm 2 ), are shown in Table 2. In general, a clear relation can be established between the area occupied by macropores and the probability to occur biggest defect (macropore that controls cracking). In this context, as detailed in this discussion later, the role played by pores and the titanium flat zone in bacterial behavior could also be evaluated. To achieve an efficient deposition through chemical stabilization of AgNPs on the substrate, the Ti surface was first treated with basic or acidic solutions (NaOH or piranha, respectively) and later modified with the chemical linker APTES via a silanization reaction to form strong covalent bonds and to introduce the necessary amino terminal moieties. The successful interaction between free primary amino groups (-NH 2 ) from APTES and AgNPs has been widely demonstrated in the literature [70,71]. Figure 2 shows the images of the chemically etched surfaces, as well as the chemical composition spectra obtained by EDS, of porous Ti substrates. Interestingly, it is observed that the surfaces modified with the NaOH treatment presented a much greater degradation compared to the piranha treatment, since the first one induces a more pronounced material loss as well as a critical increase of numerous irregularities on the material surface. This eroded morphology could provoke a potential steric obstruction impeding an efficient cell adhesion and therefore a partial antibacterial effect. At the same time, it is worth mentioning that a flatter surface promotes the presence of silver nanoparticles and consequently guarantees a greater bactericide effect; therefore, the balance of these two behaviors should be taken into consideration.
On the other hand, the synthesis of AgNPs employed a strong reducing agent as NaBH 4 and low temperatures. Figure 3 collects the main results related to the synthesis and characterization of AgNPs, emphasizing the spherical morphology, size distribution, and the spectrophotometry study of the particle suspension stability. The presence of AgNPs in suspension was confirmed by the appearance of a wide band centered at 387 nm in the UV-Vis, while the mean particle size estimated by TEM was 16 nm. treatment, since the first one induces a more pronounced material loss as well as a critic increase of numerous irregularities on the material surface. This eroded morpholog could provoke a potential steric obstruction impeding an efficient cell adhesion and ther fore a partial antibacterial effect. At the same time, it is worth mentioning that a flatt surface promotes the presence of silver nanoparticles and consequently guarantees greater bactericide effect; therefore, the balance of these two behaviors should be take into consideration. On the other hand, the synthesis of AgNPs employed a strong reducing agent NaBH4 and low temperatures. Figure 3 collects the main results related to the synthes and characterization of AgNPs, emphasizing the spherical morphology, size distributio and the spectrophotometry study of the particle suspension stability. The presence AgNPs in suspension was confirmed by the appearance of a wide band centered at 3 nm in the UV-Vis, while the mean particle size estimated by TEM was 16 nm. However, two different deposition methods were attempted: "submerged", in whic the synthesis reaction was conducted prior to the substrate submersion in the nanopartic suspension; and "in situ", in which the substrate is submerged in the reaction mixtu before adding the AgNO3, inducing the nanoparticle synthesis directly on top of the su strate surface. These different deposition mechanisms lead to small differences in th AgNPs characteristics. Thus, the particles obtained by NaOH (119.5 nm) are slightly bi ger than the ones prepared by the piranha treatment (116.2 nm), as displayed in Table  both measured by DLS. The observation of the zeta potential (ZP) shows key differenc for the colloidal system obtained from the "submerged" compared to the "in situ" AgNP methodology. It is worthy to mention that, in general, the most stable nanoparticle collo dal solutions are in the range of less than −30 mV or values greater than +30 mV. In o case, the observed zeta potential values for both acid or basic treatments indicate that th colloidal suspension of Ag reached values of −29.6 mV and −36.4 mV at 25 °C, respectivel evidencing a colloidal AgNPs system with stable regions.  However, two different deposition methods were attempted: "submerged", in which the synthesis reaction was conducted prior to the substrate submersion in the nanoparticle suspension; and "in situ", in which the substrate is submerged in the reaction mixture before adding the AgNO 3 , inducing the nanoparticle synthesis directly on top of the substrate surface. These different deposition mechanisms lead to small differences in the AgNPs characteristics. Thus, the particles obtained by NaOH (119.5 nm) are slightly bigger than the ones prepared by the piranha treatment (116.2 nm), as displayed in Table 3, both measured by DLS. The observation of the zeta potential (ZP) shows key differences for the colloidal system obtained from the "submerged" compared to the "in situ" AgNPs methodology. It is worthy to mention that, in general, the most stable nanoparticle colloidal solutions are in the range of less than −30 mV or values greater than +30 mV. In our case, the observed zeta potential values for both acid or basic treatments indicate that the colloidal suspension of Ag reached values of −29.6 mV and −36.4 mV at 25 • C, respectively, evidencing a colloidal AgNPs system with stable regions.  Figure 4 shows the EDS spectra corresponding to the surfaces of the fully dense Ti substrates pretreated first with the piranha reagent and after silanized and coated with AgNPs, using the two different deposition routes previously described (immersion and "in situ"). Significantly, the EDS results indicate the presence of titanium and silver primarily. A higher concentration of Ag deposited on the surface of the titanium discs immersed in the nanoparticle suspension is observed compared to the one obtained with the "in situ" method. Unfortunately, the method of disc immersion in the suspension causes agglomerates of AgNPs. The opposite case was observed during the elaboration of colloidal systems with the "in situ" method, where reduction in the size of AgNPs was noted. Pris and Krzysztof et al. [72] reported the impact of a vigorous stirring during the suspension (degree and time) not only on the color but also to prevent agglomeration of the AgNPs. In this sense, authors recommend implementing vigorous shaking for times less than 30 min. In their study, they observed a direct relationship between the stirring time vs. the agglomeration of the AgNPs and the dark color of the suspension. Furthermore, the presence of agglomerates of AgNPs can be detected in the study shown in Figure 5, independent of the functionalization treatment carried out on the surface of the fully dense titanium substrate (NaOH or piranha). Considering the best semiquantitative results previously discussed corresponding to the surfaces of the fully dense Ti substrates, from this moment on, this work only takes into account the combination of etching with the piranha reagent and the "immersion" methodology for AgNPs deposition to modify the surface of porous titanium substrates as the best procedure.

Characterization of Surfaces of Ti Substrates with AgNPs
Metals 2021, 11, x FOR PEER REVIEW 9 suspension (degree and time) not only on the color but also to prevent agglomerati the AgNPs. In this sense, authors recommend implementing vigorous shaking for less than 30 min. In their study, they observed a direct relationship between the st time vs. the agglomeration of the AgNPs and the dark color of the suspension. Fu more, the presence of agglomerates of AgNPs can be detected in the study shown in ure 5, independent of the functionalization treatment carried out on the surface of the dense titanium substrate (NaOH or piranha). Considering the best semiquantitativ sults previously discussed corresponding to the surfaces of the fully dense Ti subst from this moment on, this work only takes into account the combination of etching the piranha reagent and the "immersion" methodology for AgNPs deposition to m the surface of porous titanium substrates as the best procedure.   . EDS spectra of the Ti surfaces coated with AgNPs using two routes of deposition (a) submerged reduction and (b) "in situ" reduction. Figure 5. SEM micrographs and EDS-SEM of fully dense Ti substrates, treated with NaOH (left), or piranha (right), and after coating with AgNPs via a "submerged" reduction in both cases. Figure 5. SEM micrographs and EDS-SEM of fully dense Ti substrates, treated with NaOH (left), or piranha (right), and after coating with AgNPs via a "submerged" reduction in both cases.
Once the surface modification methodology was optimized on fully dense Ti substrate, the surface functionalization was studied in porous substrates. Figure 6 shows images obtained by SEM of the modified surface (AgNPs deposition by the immersion method, after etching with piranha treatment) of the porous titanium substrates obtained with different content and spacer size range. In general, it can be observed that AgNPs are mainly deposited on the flat area between the pores but in lower degree on the inner surface of them. This would entail a higher AgNP content in the samples with 30 vol% of pores, since they have larger flat area where AgNP can attach to than the substrates with 60 vol%, as displayed in Table 2. In addition, depending on the porosity of the substrates, differences can be observed both in the size and degree of agglomeration of the deposited AgNPs, as well as in the homogeneity of their distribution. In this sense, smaller AgNPs are obtained for the 60 vol.% and the smallest spacer size range (100-200 µm), while the presence of nanoparticle agglomerates is greater for the substrates obtained with larger spacer sizes. The reason for this behavior is not so clear, but it could be related to the differences in the dip deposition process, depending on the flat area of titanium (surface wettability). As discussed later, the conjunction of these effects has a direct influence on the antibacterial behavior of the samples. Thus, although the substrates with 30 vol% of pores show greater absolute antibacterial inhibition due to a larger flat surface where AgNP can attach to, samples with 60 vol% present greater relative antibacterial inhibition due to a higher NP agglomeration. spacer sizes. The reason for this behavior is not so clear, but it could be related to the differences in the dip deposition process, depending on the flat area of titanium (surface wettability). As discussed later, the conjunction of these effects has a direct influence on the antibacterial behavior of the samples. Thus, although the substrates with 30 vol% of pores show greater absolute antibacterial inhibition due to a larger flat surface where AgNP can attach to, samples with 60 vol% present greater relative antibacterial inhibition due to a higher NP agglomeration.

Characterization of the Antibacterial Behavior of Ti Substrates
For the reasons stated above, the study of antibacterial behavior was carried out on discs superficially modified (hydroxylated) with the piranha reagent followed by silanization and then submerged in the suspension of silver nanoparticles. To demonstrate the effectiveness of AgNPs coating, a qualitative experiment of bacterial growth inhibition was conducted using Staphylococcus aureus as a Gram+ bacterial model [73]. Initially, tests were performed on fully dense discs comparing the antimicrobial effect on only silanized samples and samples with a silanization and AgNP-coating treatment. Once the antibacterial character of these substrates was confirmed, the same experiments were carried out on porous samples, considering both pore samples (30 and 60 vol.%) as well as both pore size ranges (100-200 and 355-500 µm). Additionally, the samples were studied and compared with only silanization treatment (as blank sample) and the ones with silanization followed by AgNPs coating. As expected, none of the samples with just the silanization treatment presented any inhibition halo. In general, in porous substrates, independent of pore density and pore size distribution, all samples displayed antibacterial behavior (Figure 7a), although they were lower than the one displayed by the fully dense substrate. In the particular case of porous samples, those ones containing a 30 vol.% of porosity presented slightly higher bacterial inhibition than the ones with a 60 vol.% of porosity in absolute terms, probably due to a higher concentration of AgNPs related to a larger flat Ti surface where they can get attached. To prove this, a study of the surface of the different sample was conducted: a quantitative evaluation including flat areas, macropore areas and micropore areas were included, considering only micropores as those with a diameter smaller than 50 µm. Hence, the flat area proportion of the samples (calculated from a total titanium discs surface of 2.835 mm 2 , considering an ideal fully dense substrate without micropores) (Table in Figure 7b), where the AgNPs can attach easily, was higher for the substrates with 30 vol.% of porosity than the flat area of those samples with higher porosity. Therefore, in total, the substrates of lower porosity presented higher amounts of Ag. However, if a relation between the inhibition area and the flat surface of each sample is carried out, the results revealed that the relative inhibition was more pronounced for the substrates with 60 vol.% of porosity, probably due a higher content of AgNPs, both inside the pores and agglomerated in the flat area (as above revealed by SEM) that compensate the lower amount of Ag corresponding to the flat area.
sented slightly higher bacterial inhibition than the ones with a 60 vol.% of poros absolute terms, probably due to a higher concentration of AgNPs related to a large Ti surface where they can get attached. To prove this, a study of the surface of the diff sample was conducted: a quantitative evaluation including flat areas, macropore and micropore areas were included, considering only micropores as those with a diam smaller than 50 μm. Hence, the flat area proportion of the samples (calculated from a titanium discs surface of 2.835 mm 2 , considering an ideal fully dense substrate wi micropores) (Table in Figure 7b), where the AgNPs can attach easily, was higher fo substrates with 30 vol.% of porosity than the flat area of those samples with higher p ity. Therefore, in total, the substrates of lower porosity presented higher amounts o However, if a relation between the inhibition area and the flat surface of each sam carried out, the results revealed that the relative inhibition was more pronounced fo substrates with 60 vol.% of porosity, probably due a higher content of AgNPs, both i the pores and agglomerated in the flat area (as above revealed by SEM) that compe the lower amount of Ag corresponding to the flat area. These results are consistent with the total inhibition area observed in a sample 40% porosity recently reported by our group [56]. This inhibition area is located bet the inhibition area obtained for samples with 30 and 60 vol.% porosity. In compar previously reported results by Jingchao et al. [51], concerning the evaluation of inhib halos of fully dense Ti coated with AgNPs showed negligible zones of inhibition du low concentration of silver ions in the surrounding environment. These results are consistent with the total inhibition area observed in a sample with 40% porosity recently reported by our group [56]. This inhibition area is located between the inhibition area obtained for samples with 30 and 60 vol.% porosity. In comparison, previously reported results by Jingchao et al. [51], concerning the evaluation of inhibition halos of fully dense Ti coated with AgNPs showed negligible zones of inhibition due to a low concentration of silver ions in the surrounding environment.
Finally, no statistical analyses concerning the antibacterial studies have been developed as limitations of the study; a more accurate way to evaluate this effect for future studies should be carried out involving CFU data.

Conclusions
This study is based on a novel combination of a chemically functionalized porous Ti substrate with a potentially therapeutic coverture AgNPs. The absolute antibacterial activity of fully dense samples is higher than those showed by the porous materials. However, the high difference on the stiffness between the implant and the cortical bone tissue compromises the clinical success (resorption of the bone surrounding the implant). The superficially modified porous implants presented in this work showed an improved biomechanical and biofunctional balance, achieving at the same time, a reduction of the stress shielding, allowing the bone-in-growth and better antibacterial behavior. In this context, the main conclusions are as follows: (1) It culd be generally observed that mechanical resistance decreases by increasing the size of the pores, for similar pore content.
(2) To achieve an efficient deposition through chemical stabilization of AgNPs on the substrate, the Ti surface was first treated with basic or acidic solutions (NaOH or piranha, respectively) and later modified with the chemical linker APTES via a silanization reaction.
(3) Interestingly, it was observed that the surfaces modified with the NaOH treatment presented a much greater degradation compared to the piranha treatment. (4) The synthesis of AgNPs employed a strong reducing agent as NaBH 4 and low temperatures. Two different deposition methods were attempted: "submerged" and "in situ". Considering the best semiquantitative results previously discussed corresponding to the surfaces of the fully dense Ti substrates, the best combination of etching was piranha reagent and the "immersion" methodology for AgNPs deposition. (5) For porous Ti substrates, differences could be observed both in the size and degree of agglomeration of the deposited AgNPs. In particular, those using a 30 vol.% as a spacer presented slightly higher bacterial inhibition than the ones with a 60 vol.% as a spacer in absolute terms, probably due to a higher concentration of AgNPs related to a larger flat Ti surface where they can get attached. However, if a relation between the inhibition area and the flat surface of each sample is carried out, the results revealed that the relative inhibition was more pronounced for the substrates with 60 vol. % as spacer, probably due a higher content of AgNPs both inside the pores and agglomerated in the flat area that compensate the lower amount of Ag corresponding to the flat area.