Synthesis and characterization of porous structures of rutile TiO2 /Na0.8Ti4O8/Na2Ti6O13 for biomedical applications

Graphical abstract


Specifications
The most commonly used methods to promote porosity are: slurry foaming, salt leaching, phase separation and lyophilization. Hydrothermal treatment of TiO 2 with NaOH is the method commonly used to obtain sodium titanates. Resource availability: N/A

Overview
In the tissue-engineering field, porous scaffolds must be biocompatible and mechanically stable. Different synthesis methods have been investigated to obtain porous materials for biomedical applications. Among the types of materials used are polymers, such as hydrogels and thermoplastics [1,2], and some types of metals, such as stainless steel [3] and gold [4]. These materials may have limitations due to its composition and synthesis method, including the lack of sufficient biocompatibility with the host, low mechanical stability [5] and, in the case of some metals, insufficient chemical stability in vivo due to the release of toxic ions [6].
Many researchers have studied different synthesis methods such as slurry foaming [7], salt leaching [8], phase separation and lyophilization [9] to promote porosity, but removing undesirable substances generally involve the use of toxic solvents [10,11], which causes cell death when they are not completely eliminated during the process. In this context, ceramic structures such as TiO 2 and sodium titanate have good biocompatibility and chemical stability [12][13][14][15].
Regarding to the sodium titanate preparation routes, hydrothermal treatment of TiO 2 with NaOH is the commonly used method, however, this method requires a synthesis time of up to 24 h [16,17], and its result is a precipitation that must be separated by centrifugation and washed with abundant water and ethanol to finally obtain titanate powder [17]. In general, there are still some technological problems involved in the preparation of biocompatible porous structures, requiring several high-cost synthesis methods. This paper proposes the following method: By applying molding, pressing and sintering techniques in mixtures of TiO 2 /NaHCO 3 powder materials, it was possible to obtain porous rutile TiO 2 /Na 0.8 Ti 4 O 8 / Na 2 Ti 6 O 13 composites ( Fig. 1) of the required size and shape, with minimal variations and adequate mechanical properties. Unlike traditional methods, this method does not require the use of toxic solvents to remove residues, because the porous structure formed is mainly composed of crystalline phases of Na 2 Ti 6 O 13 /TiO 2 reported as biocompatible [14,18] with adequate chemical stability [12][13][14][15]. Another advantage of this method was the porosity and roughness control by varying the amount of NaHCO 3 and the sintering time [19]. At 230 C the Na 2 CO 3 formation (Eq. (1)) responds to a reaction of NaHCO 3 at 230 C [20].
Then the solid-state reaction between Na 2 CO 3 and TiO 2 (anatase or rutile phase) takes place at temperatures around of 800 C [21][22][23]. Na 2 CO 3 can also be transformed (Eq. (2)) into Na 2 O [24]: The release of gases during the heating treatment can promote voids in the structure. Depending on the molar concentration of Na 2 O and TiO 2 , both can react in the solid state to form different types of sodium titanates according to the binary Na 2 O-TiO 2 phase diagram [25]. For a high molar content of TiO 2 (>85 mol%), the main phases formed according to the phase diagram are TiO 2 rutile and Na 2 Ti 6 O 13 [22]. Based on the literature, the formation of Na 2 Ti 6 O 13 occurs via the decomposition of Na 2 Ti 3 O 7 at temperatures from 400 C to 1100 C [15].
In this paper, porous structures of rutile TiO 2 /Na 0.8 Ti 4 O 8 /Na 2 Ti 6 O 13 and Na 0.8 Ti 4 O 8 /Na 2 Ti 6 O 13 have been synthesized by molding, pressing and sintering different mixtures of TiO 2 /NaHCO 3 powder materials. We study the effects of the sintering time and the weight ratio of TiO 2 /NaHCO 3 on the porosity, and structure proportion of rutile TiO 2 wt%/Na 0.8 Ti 4 O 8 wt%/Na 2 Ti 6 O 13 wt% and Na 0.8 Ti 4 O 8 wt %/Na 2 Ti 6 O 13 wt% crystalline phases. Likewise, the biological properties of the materials studied are evaluated in-vitro through cytotoxicity tests with the CHO cell line.

Experimental
Sol-gel synthesis 6 g of TiO 2 were prepared by the sol-gel method. In a typical synthesis, 21 mL of titanium(IV) n-butoxide, 99% (TBT-ACROS) were mixed with 76 mL of absolute ethanol (Aldrich 99.8%) for 10 min under magnetic stirring. Then, 2 mL of deionized distilled water were added at the rate of 1 mL min À1 to promote the gelation of the components. The solution was heated at 100 C for 10 days. Subsequently, the TiO 2 powder was placed inside a muffle furnace (see Fig. 2), heated at the rate of 3.5 C min À1 , sintered at 900 C for 3 h, and cooled down to 25 C at the rate of 5 C min À1 .

Mixing and pressing
The TiO 2 and NaHCO 3 powders were ball-milled using a Spex Mixer/Mill for 10 min and 2 min respectively. Subsequently, both materials (TiO 2 with 15 wt% NaHCO 3 ; TiO 2 with 30 wt% NaHCO 3 ) were mixed manually. Then, the mixture was poured into a steel mold (13 mm in diameter and approximately 3 mm in height) and pressed uniaxially (35 Kg cm À2 ) for 5 min using a hydraulic press to form green compacts, as can be seen in Fig. 3.

Sintering
After drying at 45 C for 2 h, the green compact discs were sintered in a tube furnace (Carbolite Gero HTRH) at 1200 C for 4 h and 8 h in an inert atmosphere (Ar) with a flow rate of 1 L min À1 (Fig. 4). The heating rate was 3.5 C min À1 and then cooled to 25 C at a rate of 5 C min À1 . For all tests, the discs of rutile TiO 2 /Na 0.8 Ti 4 O 8 /Na 2 Ti 6 O 13 were polished with 800 and 1200 SiC paper for 2 min each side, and ultrasonically rinsed in acetone for 20 min to remove impurities.

Surface characterization
The characterization of the prepared materials was carried out to validate the formation of rutile TiO 2 /Na 0.8 Ti 4 O 8 /Na 2 Ti 6 O 13 and Na 0.8 Ti 4 O 8 /Na 2 Ti 6 O 13 porous structures. XRD patterns were obtained using a Rigaku DMAX 2100 diffractometer; CuK α radiation (1.54056 Å and 1.54439 Å) was used as an X-ray source in the 2u range of 20-80 , with a step size of 0.02 .

Identification and quantification of crystalline phases using X'Pert HighScore Plus
Moving forward, the samples prepared under different conditions will be referred to as T1: TiO 2 with 15 wt.% of NaHCO 3 at 1200 C for 4 h; T2: TiO 2 with 15 wt.% of NaHCO 3 at 1200 C for 8 h, T3: TiO 2 with 30% NaHCO 3 at 1200 C for 4 h; and T4: TiO 2 with 30 wt.% NaHCO 3 at 1200 C for 8 h.
For the XRD patterns of the materials obtained at conditions T1, T2, T3 and T4, the following methodology was used to identify and quantify three crystalline phases in the materials using the PANalytical X'Pert HighScore Plus software. For the T1 sample, the following analysis procedure was used: Treatment→Search peaks:  Select the compounds that most fit the diffractogram It is also possible to refine the diffractogram by using the Rietveld method to identify the major crystalline phase proportion (wt%) (Fig. 5): "Analysis" → "Rietveld" → "Star Rietveld Refinement"

Determination of crystallographic orientation
For all samples, it was possible to identify the preferential orientation of the more intense peaks of the crystalline phases. Here are the steps taken: Right click over the refined diffractogram → Label Peaks→Select: n most intense Peaks (for this diffractogram, it is calibrated up to 9). In Labels select: Phase Title and the boxes.

Morphological analysis
The morphology of the materials was characterized using a Phillips ESEM XL30 scanning electron microscope (SEM), operating at 20 kV. The associated energy-dispersive spectrometer (EDS) provided quantitative information about surface elemental composition. The images were obtained at magnifications of 65X and 5000 Â . The pore sizes and surface porosities of the materials were quantified by digital image analysis using ImageJ 1

Cytotoxicity test
Cytotoxicity in-vitro tests of 24 h and 48 h were performed using the cytotoxicity Kit LDH [26], which quantifies death and cell lysis based on the measurement of lactate dehydrogenase activity (LDH), released from the cytosol of damaged cells. The biological procedures were carried out in a presterilized laminar flow hood. The working solutions and Kit LDH controls were prepared and applied according to protocol. The LDH release was measured at 24 and 48 h using an Elisa microplate reader (iMark Bio-Rad), with a working length of 492 nm and a reference length of 655 nm.
By organizing the data and graphing it in OriginLab1 (Fig. 6)

Effects of sintering time and NaHCO 3 percentage on the crystalline phases
At T1, the results showed three crystalline phases, in different proportions: 14.8 wt% of rutile TiO 2 , 7.7 wt% of Na 2 Ti 6 O 13 and 77.5 wt% of Na 0.8 Ti 4 O 8 , as can be seen in Table 1. The majority phase was sodium titanate or titanium bronze Na 0.8 Ti 4 O 8 . According to previous research [27], this phase is formed from the reduction of Na 2 Ti 3 O 7 in inert gas atmosphere (hydrogen) at 950 C for two days. As previously mentioned, the formation of Na 2 Ti 6 O 13 also requires the reaction of Na 2 Ti 3 O 7 at temperatures around 400 C to 1100 C. However, the Na 2 Ti 3 O 7 phase was not observed in any of the samples analyzed, which suggests that this phase (under reducing atmosphere), can be transformed into Na 0.8 Ti 4 O 8 or Na 2 Ti 6 O 13. When comparing the T1 with T2 results, it is observed that in T2, higher concentration of the rutile TiO 2 and Na 2 Ti 6 O 13 phases were obtained, while the percentage of Na 0.8 Ti 4 O 8 was lower. The main difference between T1 and T2 is the increase of the sintering time from 4 h to 8 h. As the sintering time increases, it is possible that more Na 2 Ti 6 O 13 is converted into rutile TiO 2 as it occurs at T2 condition  [15]. The presence of rutile TiO 2 and Na 2 Ti 6 O 13 in the T1 and T2 samples corresponds to Na 2 O-TiO 2 phase diagram [25], due to the high mole amounts of TiO 2 (> 85 mol% TiO 2 ).
While at T3, the proportion of phases obtained was 27.6 wt% of Na 2 Ti 6 O 13 and 72.4 wt% of Na 0.8 Ti 4 O 8 . No rutile TiO 2 peaks were observed in the T3 and T4 samples, probably due to the high content (wt%) of NaHCO 3 . When comparing this result with the one obtained under condition T4, it was found that the proportion of Na 2 Ti 6 O 13 was higher while the proportion of the Na 0.8 Ti 4 O 8 phase was lower. It is possible that a sintering time of 8 h promotes higher formation of the Na 2 Ti 6 O 13 phase. The main differences between T1-T2 and T3-T4 is the increase on the percentage (wt%) of NaHCO 3 .

SEM images and EDS pattern
The scanning electron micrographs showed particle sizes with different morphologies (Fig. 7T1 (a)): small spherical particles of 1 mm in diameter and semi-cubic particles of about 3 mm in size. The Table 1 Rietveld refinement values from XRD patterns of T1, T2, T3 and T4 samples.

Sample
Phase fraction (wt%)  maximum pore size was 170.5 mm (Fig. 7T1(b)). The grain sized with polygonal structures [28] most probably due to the highest sintering time, is shown in Fig. 7T2(a). Similar results have been reported by other authors [15,22] calcining Na 2 Ti 3 O 7 from 800 C for several hours. Fig. 7T2(b) exhibited a maximum pore size of 180 mm.
When comparing, the T3(b)-T4(b) samples, showed higher surface porosity than the T1(b)-T2(b) samples. It is possible to observe that the porosity increased as the amount (30 wt%) of NaHCO 3 and the sintering time (8 h) increased. In other words, the higher the porogenic percentage and the sintering time, the greater the porosity.
Likewise, it is observed that the grain growth continued with the increase of the sintering time, as can be seen in Fig. 7T2(a) and T4(a). Grain boundaries are evident in the T4(a) sample, demonstrating the good crystallinity and sintering property.
The corresponding elemental composition of the materials is shown in Fig. 7(c), and the elemental ratios as obtained from de EDS detector are presented in Table 2. EDS analysis revealed oxygen, sodium and titanium to be present on the surface.
Depending on the tissue application, appropriate pore sizes can vary from 5 to 500 mm in diameter.
For bone tissue engineering applications and designing of macro devices to treat cellular immunodeficiency diseases, the optimal pore diameters vary from 40 to 450 mm [3,29]. In this context, the maximum pore size of the samples obtained at the different treatment conditions were calculated between 50 and 496 mm, which is a widely accepted range for use in biomedical applications.

Cytotoxicity evaluation
The success in the preparation and characterization of any material for use in the biomedical field depends on the positive response that cells, and tissues present upon being in contact with such material once implanted.
The LDH release results at 24 h and 48 h from the CHO cell line showed in Fig. 8. LDH release from all materials in the presence of the cells was low compared to high toxicity control of 1% Triton-X 100 The materials (T3 and T4) with more NaHCO 3 (30 wt%) constituted by the Na 2 Ti 6 O 13 and Na 0.8 Ti 4 O 8 phases caused less LDH release at 24 and 48 h than the materials (T1 and T2) constituted by rutile TiO 2 , Na 2 Ti 6 O 13 and Na 0.8 Ti 4 O 8 phases prepared with less NaHCO 3 . It is possible that the presence of the rutile phase in the T1 and T2 samples increased the LDH release compared to the materials constituted by Na 2 Ti 6 O 13 and Na 0.8 Ti 4 O 8 phases. In previous investigations, sodium titanates were prepared via hydrothermal process and sintered at temperatures between 900-1050 C for 1 h [14]. These materials presented low LDH release in the cytotoxicity assay. The in-vitro biocompatibility of porous barium titanate manufactured by a direct foaming technique was investigated, using an LDH concentration assay on osteoblast cells cultured for 72 h [29]. In all cases, the results showed that the materials were significantly less toxic to cells, than those cultured with 1% Triton X-100 for 45 min (high toxicity control).
It was demonstrated that the materials evaluated (TiO 2 /Na 0.8 Ti 4 O 8 /Na 2 Ti 6 O 13 and Na 0.8 Ti 4 O 8 / Na 2 Ti 6 O 13 ) in the present investigation induce low LDH release (less than 13%).

Conclusion
Porous structures composed of the rutile TiO 2 /Na 0.8 Ti 4 O 8 /Na 2 Ti 6 O 13 and Na 0.8 Ti 4 O 8 /Na 2 Ti 6 O 13 crystalline phases were obtained by mixing, compacting and sintering TiO 2 with NaHCO 3 at different times in an inert Ar atmosphere. It can be concluded that varying these treatment conditions influences the response variables, such as porosity and structure proportion. Moreover, these properties can be controlled to produce a potential suitable material for use in the biomaterials field.
According to the results of the cytotoxicity test, the ceramic porous structures under study induce low LDH release in the presence of the CHO cell line in vitro, regardless of the treatment condition and the rutile TiO 2 /Na 2 Ti 6 O 13 /Na 0.8 Ti 4 O 8 phase fraction, the porous ceramic materials evaluated could be used for biomedical applications. Finally, this method of synthesis could be used for future applications in the biomedical field, replacing conventional manufacturing techniques because of its porous structure and specific crystalline phases that do not affect the in-vitro cell viability.