Nanostructuring Ti-Alloys by HPT: Phase Transformation, Mechanical and Corrosion Properties, and Bioactivation

Abstract

Due to their unique properties, titanium (Ti) and Ti-alloys are particularly suitable for biomedical devices. Ti has a high specific strength and low Young’s modulus (reducing stress shielding), high corrosion resistance, and superior biocompatibility. However, Ti’s moderately low Young’s modulus (100–110 GPa) is still considerably higher than to bones (5–30 GPa). The β-Ti phase, whose elastic modulus is closer to the bone, can be kept by increasing the contents of non-toxic β stabilizing elements. Besides stress shielding and corrosion resistance, adjusted bioactivity (bone-bonding ability) is another primary prerequisite for implants that can be improved by ultrafine-grained (UFG) microstructures and surface modification (anodization and acid+alkaline treatment). UFG by HPT also enhances wear resistance and mechanical properties. Representative alloys (Ti–6Al–7Nb (TAN), Ti–13–Nb–13Zr (TNZ), and Ti–35Nb–7Zr–5Ta (TNZT)) and cp-Ti, were presented in this overview. Samples started with different phases and morphologies. Deformation by HPT induced phase transformation in the alloys, which depended on the amounts of α or β stabilizers, the strain rate, applied loads, and starting phases and α morphologies. Grain sizes were reduced to about 120 nm. Mechanical properties depended mainly on the number of grain boundaries and their nature and different phases, sizes, and strengths. Young’s modulus diminished when the β was increased. Polished surfaces and cp-Ti presented similar corrosion resistance, improved by surface treatments, which reached maximum protection in anodized samples processed by HPT. After bioactivity tests, different growth rates for various processing conditions and alloys were observed, the highest for the TNZ alloy, and improved after HPT processing.

(a) The average grain sizes (upper) and the corresponding hardness (lower) of the alloys analyzed here as processed by HPT according to applied pressures, number of turns (2t, 3t, and 5t), and initial conditions. (b) Hall-Petch plot resultant from experimental results of Figure (a). Full lines represent the experimental behavior, and dash-dot ones the Hall-Pecth relationship.

1. Introduction

Soon skeletal injuries and disorders due to aging will affect 25% of the steadily growing populations in the world.¹⁾ However, only limited technological and clinical breakthroughs in implant materials development have been achieved in the past decades.²⁾

Indeed, developing new biomedical materials has been one of the last decade’s most explored defiance in materials science. Materials conceived to be employed in the human body must meet many well-known but frequently contradictious prerequisites,³⁾ such as high strength and fatigue life, low modulus, and high wear resistance.

Essentially, implant devices’ effectiveness is guaranteed by their human body’s inertness because of properties like corrosion resistance, lack of toxicity, biocompatibility, etc. Nevertheless, inertia incompletely fulfills ideal conditions of interaction for the body, reacting to the device’s insertion and driving it to its rejection.⁴⁾ Nearly all implant failures ensue in the early stages after implantation or in the first year’s load support, increasing the necessity for revisions.⁵⁾ Such revisions occur primarily because of extreme inflammatory state, low osteointegration, and bacterial infection, leading to the implant’s complete failure, and causing substantial distress for patients, often submitted to highly aggressive and costly surgeries.^4,6)

Ideally, a prosthesis’s primary attributes must be (i) biocompatible and free of toxic elements (such as V, Ni, Cu, Co, etc.). These elements are present in most utilized biomedicine alloys, may be freed by corrosion and/or friction over long-term implantation, and can be thoroughly dangerous and possibly lead to severe troubles such as Alzheimer’s, allergic reactions, and cancer.^7–9) (ii) elevated bioactivity, where the implanted material yields fast, driven, controlled bone rehabilitation, protein adsorption, and cell attachment. High bioactivity induces an adequate interface between bone tissue and prosthesis, leading to good mechanical properties between them^10,11) (iii) high resistance to corrosion and wear, and excellent mechanical properties, including Young’s modulus (E) that should be close to the bone’s one, preventing the stress shielding effect, which reduces bone density.^12,13)

For many decades, scientists from materials science and implantology have been striving for advancements in the design of prosthesis devices that may comply with the above attributes, thus leading such devices to be useful for implant application in replacing body components. As one may observe above, bone is one of the prime targets because of its unique properties and the prospect of having its functional restoration employing bulk biocompatible implants jointly with suitable surgical methods. Such context leads Ti and its alloys to be the currently more appropriate and pursued materials for various developed implants for orthopedics and dentistry employment. In recent decades, considerable experimental and clinical investigations have demonstrated promising biocompatibility and good corrosion resistance of Ti and Ti-based alloys, two of the crucial attributes of a metallic implant.

Among the above prerequisites for ideal medical bone implants, adjusting bio-activity and corrosion rates is challenging. The natural passive oxide layer on the surface of Ti-based materials greatly helps the necessary corrosion properties¹⁴⁾—the chemical inertness of this layer, predominantly formed of TiO₂, protects the metallic surface. However, Ti inertness is ineffective in combating infections or controlling any activity linked to bone cell interactions, which is valuable for bone healing implants.¹⁵⁾ Therefore, the improvement of these alloys’ surfaces by several modifications and coating methods is now available, improving bone formation or even decreasing the chance of infections.^{16–18,25–27)}

Titanium-based implants exhibit superior mechanical capability considering load capacity (maximum load, bending, and fatigue strength)^16,17,28) and stiffness. This last may favor controlling bone cell phenotypic necessity.^29,30)

Ti and Ti-alloys’ mechanical properties are nearer to the bone tissues’ compared to other metallic biomaterials such as stainless steels and cobalt-chromium alloys, satisfying a critical requirement for the successfulness of an implant. Their Young’s moduli values and hardness are comparable to the bone ones, diminishing the stress shielding effect, reducing the possibility of bone resorption, and loosening the implant from the bone.¹³⁾

Though pure titanium and the alloy Ti6Al4V are materials largely commercially employed for titanium-based implants,³¹⁾ there are critical considerations concerning their elastic moduli, which are still much higher compared to the bones (10–30 GPa).³²⁾

Pure titanium is the option whenever corrosion resistance is the most profound concern, such as in dentistry. When mechanical properties are more pertinent, the Ti6Al4V (α + β) alloy is chosen. Also, heat treatment or thermomechanical processing can adjust its chemical stability, microstructure, and mechanical behavior.^7,33–35) Nonetheless, there are growing considerations concerning the cytotoxicity effects of Al and V ions that can be released into the body because of corrosion or friction of implanted Ti6Al4V alloy,⁷⁾ as mentioned above, resulting in adverse bone tissue reactions.^7–9)

Consequently, the latest objective is designing single-phase atoxic and free of allergic elements β-titanium alloys with outstanding workability and mechanical properties. Furthermore, β-titanium alloys display intrinsic lower elastic moduli than the ones of α-Ti and α + β Ti6Al4V alloy.^7,36) The most recent titanium alloys designed for orthopaedical uses, including Ti₁₂Mo₆Zr₂Fe (TMZF),³⁷⁾ Ti₁₅Sn₄Nb₂Ta_0.2Pd,³⁸⁾ Ti₁₃Nb₁₃Zr (TNZ),³⁹⁾ and Ti₂₉Nb₁₃Ta_4.6Zr (TNTZ)⁴⁰⁾ alloys, has achieved lowest modulus and better biocompatibility. More thorough and updated overviews of the mechanical behavior, the influence of thermomechanical processing on the microstructure, and alloying of Ti-based alloys can be encountered elsewhere.^3,7)

Further improvement of the mechanical properties of Ti-based alloys may be acquired by microstructural refinement by severe plastic deformation (SPD) processes, including equal channel angular pressing (ECAP) and high-pressure torsion (HPT). Such methods can lead grain sizes to the ultrafine (UFG) or even nano-size range.^{18–20,41–51)} Therefore, SPD may enhance the wear resistance and bioactivity by causing apatite precipitation and amplified biological response compared to coarse-grained alloys due to resulting Hall–Petch strengthening, augmented surface energy, and increased number of nano-grooves, which may lead to a significant advancement in the cell adhesion/proliferation.⁵²⁾ Also, the metallic materials’ nanostructuring can alter static and dynamic mechanical behavior.^18,53–57) Indeed, mechanical stability, further improved by the surface nano/micro-roughness,⁵⁸⁾ is crucial for producing reliable implants, decreasing implant failure incidence, and, most importantly, revision surgeries. Moreover, with the increased strength due to the Hall-Petch effect, implants can be produced with smaller sizes, providing the same functionality but using lesser invasive surgeries.⁵⁹⁾

A further characteristic of processing titanium and its alloys via HPT is the dynamic phase transformation leading to the precipitation of metastable phases.^{19,20,60–63)} The literature mainly reports on the ω-phase and the martensite α′′ phase, which may affect Ti alloy’s tenacity and elastic modulus.^19,60,64,65)

In investigating Ti-alloys for implant application in the last decade, the authors and colleagues conducted studies on their processing by HPT and its consequent influence on the microstructure evolution, phase transformation, and some properties. Also, some investigations were performed on surface treatments to improve their bioactivity. This overview shortly examines and describes the primary discoveries acquired in such above studies. However, the readers must be aware that the widespread laboratory HPT equipment involves, among a few limitations, small-sized samples unsuitable for medical implant manufacture.^46,66) Processing techniques that are more suitable for that, among others, include ECAP^18,27,49) and ECAP-Conform,^67,68) which can be used to produce materials in relatively large volumes efficiently. However, as with other laboratory equipment, conventional HPT, taking due care in the analysis as in the papers discussed in this overview, continues to be very valuable and broadly used in simulating larger-scale types of equipment.

2. Mechanically Induced Phase Transformation in Ti-Alloys by HPT

HPT was conducted using cylindrical anvils fabricated from tool steel, which were surface nitrided after machining. The samples were in the form of discs with 7 or 10 mm diameter and 1 mm thickness. The applied pressure was in the range of 1 to 6 GPa, the rotation speed was 3 rpm, and the total number of turns was up to 5. Initial samples hardness is low; hence, there was no contamination from the anvils.

In this session, some results achieved regarding mechanically induced phase transformation by applying HPT to three different representative materials are reviewed, and some new results are included to improve the discussion of previous results.

Processing titanium and its alloys via HPT can lead to mechanically induced phase transformation, where metastable phases may precipitate.^{19,20,60–63)} When samples are deformed by HPT, there are indications that grains larger than the submicron can induce the β to ω transformation under a relatively low pressure of 3 GPa⁶⁰⁾ in high alloyed β-Ti–Nb–Ta–Zr–O alloy or transformation of α phase to ω phase by applying loads as high 6 GPa or higher no matter the number of revolutions or rotation speed in the Ti–6Al–7Nb alloy⁶⁹⁾ or commercially pure Ti,^61,70) or yet, also in cp-Ti, the amount of ω phase is dependent on the rotation speed and the number of rotations⁷¹⁾ applying 5 GPa, the fraction of the ω-phase decreases when the rotation speed is increased. There is a substantial augment of such a phase with the number of rotations. An additional critical aspect in the role of α to ω transformation is the effect of Ti phase stabilizers in Ti, and oxygen⁷²⁾ may play a role by hampering the α to ω-phase transformation.⁷³⁾ Hence, the more the oxygen content, the more stabilized the α phase. Consequently, even after processing cp-Ti having a reasonably high oxygen content at 6.0 GPa, authors did not observe phase transformation.⁷⁴⁾

We studied the effect of the number of turns by processing the Ti–6Al–7Nb alloy by HPT using the high load of 5 GPa, 3 rpm, and 1, 3, and 5 turns.⁵⁰⁾ As shown in Fig. 1, there was no phase transformation from the initial as received α phase to the ω one at any number of turns. In this case, one must consider that the alloy was prepared by arc-melting in an Ar-protective atmosphere after several cycles of vacuum and argon cleaning. Therefore, the oxygen content may be neglected. However, as aluminum is the most important α-stabilizer and as the amount of β-stabilizer (Nb⁷⁵⁾) is small, we believe that the effect of Al is similar to the one of oxygen as observed in Ref. 74) and, therefore, α is kept. Such an observation may be reinforced by the results obtained in Ref. 69), where the amount of the transformed ω was minimal, even using 6 GPa, 20 revolutions, and 1 rpm. Therefore, one would say that also the high strain rate used in our work may have hampered α from being transformed in the ω phase during HPT due to plastic localization and adiabatic shearing that produces local heating.⁷¹⁾

Fig. 1

X-ray diffraction patterns of the Ti–6Al–7Nb alloy processed by HPT using a different number of turns and a constant load of 5 GPa, and 3 rpm. (some similar results can be found in Ref. 50)).

After processing by HPT, the (α + β) Ti–6Al–7Nb alloy produced only the α phase, which has Young’s modulus of 114 GPa, still high when compared to moduli for human bones (10–30 GPa).³²⁾ Therefore, we studied the biomedical β-rich TNZ alloy, which has the advantage of increasing the amount of Nb, a β stabilizer,⁷⁵⁾ and removing Al from composition, which has raised concerns about its cytotoxicity effects when its ions are dissolved into the body as a consequence of corrosion, hence, resulting in aggressive reactions with the bone tissue.⁸⁾ For such a study, we produced different starting microstructures for HPT processing. We analyzed its influence on dynamic phase transformation that may induce an increase in the β-Ti phase and provide a better Young’s modulus than the previous Ti–6Al–7Nb alloy. In summa, the different heat treatments (HT)⁷⁶⁾ were performed as follows: (1) - Slow cooling (SC) from the β-field (750°C/1 h - furnace cooling; (2) Fast cooling (FC) from the β field (750°C/1 h - water quenched), and (3) Quenching (HT 2) and aging (QA) (750°C/1 h - water quenched + 500°C/5 h + air cooled). HPT was performed using pressures of 2, 3, and 4.5 GPa, at 3 rpm and three turns. In a previous work,⁴⁸⁾ we published results for the HT type (1) using loads of 1, 4.5, and 6 GPa.

Figure 2 presents the resulting microstructures (left side) and structures (right side) after the different heat treatments applied to the TNZ alloy. The slow cooling HT resulted in the α phase with lamellae morphology in the retained β phase (∼35%), the fast cooling HT in the acicular martensite (α′) within the retained β phase (∼10%), and the quenching and aging HT in the martensite transformed into α and β (∼36%), where one may observe a distribution of fine globular α along pre-existing martensite plates.

Fig. 2

Scanning electron microscopy images (left side) and X-Ray diffraction patterns (right side) showing microstructures and structures of the Ti–13Nb–13Zr alloy, respectively, after different heat treatments (mentioned in Figure). Similar slow and fast cooling results can be found in Refs. 20), 48).

Figure 3 presents the mapping of phases using Automated Crystal Orientation Mapping (ACOM) in transmission electron microscopy (TEM) in nanoprobe mode after HPT processing of the alloy TNZ. Pressures, initial conditions, and phases’ fractions are annotated in Figure. The yellow color represents the α and α′ phases in the respective heat treatment, the blue the β phase, and the red the ω one. It is noted that samples processed by HPT using the slow cooling condition and loads up to 3 GPa produced only α and β phases. When the pressure is increased to 4.5 Gpa, besides grain refinement (please, refer to the grain boundary lines in the images), the microstructure is broken and redistributed. There is a sudden augmentation of β, and the ω begins to emerge. Considering data published before,⁴⁸⁾ there is a significant increase of β from ∼35% to about 75% for the undeformed and deformed at 1 GPa conditions, respectively. Hence, one may affirm that the β amount decreases with the pressure and becomes almost constant between 2 and 3 GPa, after which it increases again at the highest pressures of 4.5 and 6 GPa;⁴⁸⁾ the phases are better distributed along the sample volume, the quantity of β is kept at a stable value from 4.5 to 6 GPa, and the ω is increased. Observing the quenched and aged condition, one may assume that phase transformation to ω increases from pressures of 2 GPa. Also, as for the slow cooling state, the amount of β is almost stable between 2 and 3 GPa, then decreases with the load. After fast-cooling, there is an increase of β up to the pressure of 3 GPa, then a decrease. There is no precipitation of ω in this condition.

Fig. 3

Phase Mapping using Automated Crystal Orientation Mapping (ACOM) in transmission electron microscopy (TEM) in nanoprobe mode for the Ti–13Nb–13Zr alloy after being processed by HPT for the different heat treatments: Slow cooling on the left side, fast cooling in the center, and quenched and aged on the right side. Samples were deformed using three turns and a rotation speed of 3 rpm. HPT pressures and phases’ fractions are annotated in Figure. (some similar results can be found in Ref. 19)).

In another work,¹⁹⁾ we processed the alloy Ti–35Nb–7Zr–5Ta (TNZT) by HPT using a pressure of 4.5 GPa, three turns, and a rotation speed of 3 rpm. The alloy was previously heat-treated and fast-cooled from the β field. This route was chosen given the ω phase non-precipitation under these conditions for the TNZ alloy. Despite being able to increase the strength, the ω phase is embrittling^77–79) and can increase Young’s modulus.^79,80) Thus, in the case of implants, it is an undesirable phase.

Furthermore, the pressure of 4.5 GPa was found to be more efficient for grain refining, thus providing the abovementioned properties required for an implant material. The heat treatment resulted in an equiaxed grains microstructure mainly composed of β-phase.¹⁹⁾ However, the XRD pattern shown in Fig. 4(a) also reveals a small amount of α phase. After HPT processing, Fig. 4(a) shows that practically all the α transformed into β, confirmed by the Phase Mapping in Fig. 4(b) that shows a minimal amount (∼7.4%) of α phase (magenta) still present in an equiaxed refined β matrix (white), where black lines represent grain boundaries.

Fig. 4

Results for the Ti–35Nb–7Zr–5Ta (TNZT) alloy. (a) XRD patterns comparing both the undeformed and HPT deformed and (b) ACOM-TEM showing a minimal amount (∼7.4%) of α phase in the magenta present in an equiaxed refined β matrix in white (black lines represent grain boundaries). The alloy was previously heat-treated and fast-cooled from the β field, then deformed by HPT using a pressure of 4.5 GPa, three turns, and a rotation speed of 3 rpm. (some similar results can be found in Ref. 19)).

Considering the above-reported alloys, it is evident that HPT has induced solute redistribution,¹⁹⁾ which may lead to different phase transformations. As one may observe, the precipitation of the ω phase occurs only in the presence of the α phase in the case of the TNZ alloy, where the amount of β stabilizer (Nb) is still low. Analyzing Fig. 3, where there is ω precipitation, it is noteworthy that the ω phase is precipitating in β regions. Considering the amount of β stabilizer in the alloy and a pseudo-binary isomorphous phase diagram published elsewhere,⁸¹⁾ one may conjecture that those precipitation regions may have an amount of β stabilizer that may lead such areas to be inside the ω + β field. Therefore, it is believed that the transformation sequence is from α to β, then β is destabilized and transforms into ω. When α′ phase is present, ω is destabilized and does not precipitate. Observing the results for the high alloyed TNZT alloy, where the amount of β stabilizers (Nb and Ta) was increased by about three times compared to the 13Nb in the other alloy, and considering the same phase diagram, even with the minor presence of α one may infer that solute redistribution moved the alloy to the β-phase field. Hence ω is destabilized and does not precipitate.

3. Mechanical Properties

As it is well known, HPT processing can lead grain sizes to the ultrafine (UFG) or even nano-size range, and hardening occurs.^{18–20,41,46–48,50,51)} Figure 5(a) presents average grain sizes and the corresponding hardness of the alloys analyzed here as processed by HPT according to applied pressures, number of turns (2t, 3t, and 5t), and initial conditions. The average grain size was measured by ACOM-TEM in the case of TNZ and TNZT alloys (Figs. 3 and 4) and by TEM⁵⁰⁾ in the case of the Ti–6Al–7Nb (TAN) one. We considered all the phases present to stand for such an average. Average hardness measurements were taken from half of the discs’ radius circumference, where TEM samples were observed.

Fig. 5

(a) The average grain sizes (upper) and the corresponding hardness (lower) of the alloys analyzed here as processed by HPT according to applied pressures, number of turns (2t, 3t, and 5t), and initial conditions. (b) Hall-Petch plot resultant from experimental results of Fig. 5(a). Full lines represent the experimental behavior, and dash-dot ones the Hall-Pecth relationship. (some similar results can be found in Refs. 19), 50)).

As one may keep from Fig. 5(a), grain sizes decrease with the applied pressure for the same number of turns (corresponding to the shear strain). If one pays attention to the grain size behavior for the TAN alloy, as we are not working in conditions of Ultra-Severe Plastic Deformation,⁸²⁾ its grain sizes reach a steady state after three turns. Due to the decreasing grain sizes, there is a hardness increase. However, this is an expected, superficial and well-known result, and it could be interesting to analyze the hardness evolution behavior under the light not only of grain sizes but also the presence of second phases.

Bearing that the TAN and TNZT alloys are a single phase (Fig. 1 and Ref. 50)), their hardness shall depend only on the grain size and forming phase, which is not the case for multiphase TNZ alloy in the diverse initial conditions, influencing the hardness behavior. See in Fig. 5, for instance, very different grain sizes resulting in almost the same hardness in the case of the TNZ alloy, or the smaller grain size of the TNZT alloy resulting in hardness similar to the ones for larger grain sizes of the other alloys, or, yet, smaller hardness of the TAN alloy compared to similar grain sizes of the TNZ alloy. That is why the effect of second phases by dynamic aging arising from SPD (HPT in our case), as discussed above, has drawn considerable awareness because, besides their precipitation kinetics, precipitates’ morphology differs significantly from the ones of conventional aging, as observed here, providing new opportunities in the development of advanced alloys that can be hardened with aging.^83,84)

Hall⁸⁵⁾ suggested a hardness dependency on grain size directly obeys the Hall-Petch relation,^86,87) as follows:   

\begin{equation} H = H_{0} + \frac{K_{H}}{\sqrt{D}} \end{equation} (1)

where H₀ and K_H are materials-dependent constants whose physical meanings are still far from being well explained. Therefore, it is suggested that eq. (1) can be restated stresses, as in eq. (2):   

\begin{equation} \sigma_{Y} = \sigma_{i} + \frac{K_{Y}}{\sqrt{D}} \end{equation} (2)

where σ_Y is the yield stress, D is the average grain size, σ_i is the friction stress, denoting the global crystal lattice resistance to dislocation movement, and K_Y is the Petch parameter, or the unpinning constant or “locking parameter”, measuring the relative hardening contribution of grain boundaries. Assuming that the average strength of a metallic material is roughly 3.295 times the HV,⁸⁸⁾ one may readily convert hardness values to stress (in MPa). Hall-Petch plots corresponding to data in Fig. 5(a) are presented in Fig. 5(b). Values of σ_i and K_Y for alloys and conditions are annotated in the Figure.

As expected for the single-phase TAN alloy, the experimental behavior follows a linear relation indicating total agreement with the Hall-Petch relationship, i.e., a hardness dependency on grain boundaries (K_Y) and forming phase (σ_i). However, it is not the case for the TNZ alloy, where the hardness experimental behavior is not linearly related to the D^−0.5 evidencing other strengthening mechanisms than the ones for the TAN alloy. Analyzing Figs. 5(a) and 5(b), it is worthy of note that for the same pressures, the strength for the diverse starting conditions of the TNZ alloy is very similar, regardless of grain sizes. Observing the curves’ behaviors (full lines representing the experimental behavior) of the TNZ alloy in Fig. 5(b) and correlating them with Fig. 3, it is also noteworthy that the sudden slope increase for FC and QA conditions at 4.5 GPa is directly related to the copious α precipitation and α and ω precipitations for FC and QA, respectively. Also, the concomitant grain refinement for the QA condition accounting for the fine ω size. In contrast, the decreasing slope from 3 GPa for the SC condition is correlated to the α decrease and the considerable grain refinement, also considering the fine ω phase. As the strength of the single-phase α-TAN alloy is the smallest, one may think of the lesser solid solution effect of Al and Nb than Nb and Zr, not only regarding elements but also their amounts.

Analyzing the values of K_Y in the Hall-Petch relationship (dash-dot lines), which expresses the efficiency of grain/phases’ boundaries, one may say that grain boundaries dominate the behavior of the TNZ-FC condition and the TAN alloy. In contrast, the strength of the grain interior (σ_i), represented by the strength of different phases present, contributes more to the other starting conditions of the TNZ alloy. These influences can be better visualized by comparing the product K_YD^−0.5 and σ_i for the different situations. By calculating such a product, one would get similar values of about 1100 MPa on average for all cases, which is much higher than σ_i for the FC-TNZ condition and TAN alloy (43 MPa and 179 MPa, respectively) or σ_i ∼ 9% of the total yield stress (σ_Y), while other σ_i values (∼500 MPa) represent almost 50% of the average product or one-third of σ_Y.

Therefore, despite having a vital role in mechanical strength, as observed for pure metals and single-phase materials, that may be improved by about 3–6 times^89,90) due to severe grain refinement via SPD, the grain size and fine precipitation and their strength, will finally, indeed, increase the number of boundaries. Hence, as observed above, boundaries (K_Y) dominate the strength behavior of all alloys at any condition, even in the ones where σ_i had an augmented influence. As with any SPD-processed materials, the HPT-processed ones of this overview may have various grain boundary nanoscale features (e.g., dislocation hardening, precipitation hardening) that may further strengthen,^89,91) resulting in higher values and linearity behavior that do not follow those predicted by the standard Hall–Petch relationship,^89,92) as also observed here. Grain boundary segregation also supplies extra strengthening.^93–96) Such segregation strengthening is physically explained and linked to dislocation generation at grain boundaries.⁹⁴⁾

Finally and very importantly, using HPT, no matter the alloy or route, an outstanding strength was acquired by grain refinement and/or induced phase transformation. However, as stated above, ω precipitation is not desirable for implant alloys. Therefore, the fast cooling condition was chosen as the best, avoiding omega precipitation and the load of 4.5 GPa to refine grains, increasing grain refinement and, consequently, strength, hence, increasing surface energy and the number of nano-grooves, which may direct to a vital expansion in cell adhesion/proliferation. Considering all the above aspects, Fig. 6 presents Young’s modulus of undeformed and after HPT processing for the different alloys. TNZ and TNZT samples achieved remarkable elastic modulus, slightly dropping after HPT processing. Such a drop was inferred to the increase of β phase in both alloys. Measurements gave an average value of ∼63 GPa and ∼60 GPa for the TNZ alloy and ∼45 GPa and ∼44 GPa for the TNZT alloy for the undeformed and deformed conditions, respectively. The α phase was kept in the Ti–6Al–7Nb alloy, and even with the high microstructural refinement, as expected, Young’s modulus was kept at ∼114 GPa, still high compared to moduli for human bones (10–30 GPa),³²⁾ while the values acquired by the TNZT alloy are very close to them.

Fig. 6

Young’s modulus for undeformed and after HPT processing samples of the Ti–6Al–7Nb, Ti–13Nb–13Zr, and Ti–35Nb–7Zr–5Ta alloys. (some similar results can be found in Refs. 19), 50)).

4. Surface Treatment, Corrosion Resistance, and Bioactivity

4.1 Surface treatments

As stated in the introduction, biomaterials research is developing from bioinert and biologically passive implants that stimulate tissue regeneration. Hence, the surface physicochemical properties of bone implants ought to be best able to guide and promote bone tissue healing biologically. Thus, surface modifications have been carried out by anodization^{19–21,23,25)} and chemical methods^{18,19,22,25–27)} to induce specific responses in osteoblastic cells after implantation.

Figure 7 presents SEM images after surface-treated samples by anodization (following procedures in Refs. 19), 20)) in Figs. 7(a)–(d), chemically treated with HCl 37% at 60°C and NaOH 10 mol·L⁻¹ at 60°C for 24 h in Figs. 7(e)–(h) (see procedures in Ref. 25)), chemically treated with concentrated phosphoric acid at 80°C for 30 min, then alkaline in a 10 mol L⁻¹ NaOH solution at 60°C for 24 h in Figs. 7(i)–(l) (see methods in Refs. 18), 19), 27)).

Fig. 7

SEM images after surface-treated samples. (a)–(d) by anodization (following procedures in Refs. 19), 20)), (e)–(h) chemically treated with HCl 37% at 60°C and NaOH 10 mol·L⁻¹ at 60°C for 24 h (see methods in Ref. 25)). (i)–(l) chemically treated with concentrated phosphoric acid at 80°C for 30 min and alkaline in a 10 mol L⁻¹ NaOH solution at 60°C for 24 h (see methods in Refs. 18), 19), 27)). Undeformed samples are shown in Figures (a), (e), (i) for the TNZ alloy and (c), (g), (k) for the TNZT alloy. Figures (b), (f), (j) and (d), (h), (l) are for the deformed TNZ and TNZT alloys, respectively. Insets in Figures (a)–(d) of anodized samples show nanostructures’ cross-section views. Insets in Figures (e)–(g), (i), (k)–(l) are higher magnifications for comparison purposes with Figures (h) and (j). (similar results for anodized and H₃PO₄-treated samples can be found in Refs. 19), 20)).

Indeed, Fig. 7 shows the successful production of diversified nanostructures on samples’ surfaces for both compositions and deformed states. For the first time,^19,20) we showed the presence of two different self-organized nanostructures after the anodization of the undeformed Ti13Nb13Zr alloy (Fig. 7(a)), where nanotubes grow in the α′-phase and nanopores in the β-phase (regarding phase, please, refer to Figs. 2 and 3), no matter phases’ spatial distribution. The HPT processing (Fig. 7(b)) induced phases’ refinement and spatial redistribution, and nanotubes have the same diameter as for the non-deformed condition are not well organized and deformed in their longitudinal direction, indicating a growth following the spatial distortion of the α′ phase. Regarding the TNZT alloy, almost entirely β (Fig. 4), only nanopores are observed on the samples’ surfaces for undeformed (Fig. 7(c)) and deformed (Fig. 7(d)) conditions. Indeed, such results indicate that only nanopores are formed in the β phase following the conditions we used.

Nonetheless, the nanostructures of the deformed sample are much less ordered than the ones of the deformed TNZ alloy (Fig. 7(b)). Outer diameters of nanopores and nanotubes are similar (∼87 nm) and, even with similar lengths (∼2 µm), indicate favorable nanostructures’ growth in the TNZ alloy and more in its undeformed situation. As reported before,^{19,20,23–25)} nanotubes and nanopores formed are amorphous and have low strength after anodization. Thus, samples were heat-treated at 550°C for 2 h resulting in the highest strength mix of crystalline phases from different oxides, notably anatase and rutile from TiO₂. However, such a heat treatment did not change any dimension but relieved the stress of deformed microstructures after HPT processing.

The micrographs of Figs. 7(e)–(h) and 7(i)–(l) present the results of HCl and H₃PO₄ chemical surface treatments, respectively. Figures 7(e), (i) and 7(g), (k) concern undeformed conditions of the TNZ and TNZT alloys, respectively, and Figs. 7(f), (j), and 7(h), (l) regard their deformed conditions, respectively. One may observe that a nanotopographic sponge-like morphology^18,19,27) was formed on the samples’ surface, not mattering the acid plus alkali treatment. Nevertheless, noteworthy are sponges’ tentacles, having somewhat dissimilar sizes (please, refer to related Figures and insets for comparison). Such sizes depend on the deformation condition; they have similar sizes for HCl-treated and H₃PO₄-treated samples. The tentacles’ sizes decrease from undeformed and deformed TNZ alloy to undeformed and deformed TNZT one. Also, there are more microcracks on the surfaces of HCl-treated specimens than the H₃PO₄-treated ones.

The significance of micron, submicron, and primarily nano features development on surfaces are significant in accelerating biological responses. Surface treatments, by anodization and chemical ones, have shown to be practical and straightforward ways of modifying surfaces producing a uniform surface nanoporosity and sponge or coral-like nanotopographic on substrates. As observed in Refs. 18), 24), the feature mentioned above sizes showed to be adequate to improve wettability, hence increasing surface free energy and adhesion interfacial free energy, leading to an improved effect on cell membranes, cell proliferation, and adhesion, mainly in the ultrafine-grained alloy.¹⁸⁾

4.2 Corrosion resistance

Figure 8 presents the potentiodynamic polarization curves comparing corrosion behaviors of the diverse surface-treated surfaces of the alloys TNZ (Fig. 8(a)) and TNZT (Fig. 8(b)) in the deformed and non-deformed conditions and cp-Ti. Such curves depict diverse corrosion potential values for the different conditions, higher for anodized and thermally treated samples (nanotubes or nanopores), meaning that anodization significantly enhances the material’s surface stability. Comparing the potentials of polished specimens (cp-Ti, and the alloys TNZ and TNZT), they are electrochemically very similar. These differences may be ascribed to surface preparation (grinding and polishing). Despite some dissimilarities, roughly, potentials increase in the sequence from polished surfaces, H₃PO₄-treated, and HCL-treated surfaces to anodized surfaces, indicating that oxides formed by the different chemical treatments and anodization ennoble surfaces.

Fig. 8

Potentiodynamic polarization curves recorded in Simulated Body Fluid (SBF) at 37°C comparing corrosion behaviors of the diverse surface-treated surfaces of the alloys TNZ (a) and TNZT (b) in the deformed and non-deformed conditions and cp-Ti. (some similar results can be found in Refs. 20), 23), 24)).

All samples exhibited a passive behavior after an active first region near the corrosion potential. Current densities achieved a steady and broad-ranging plateau with passivation currents of less than 0.5 µA cm⁻² for the anodized samples and close to a narrowed range between ca. 2 and 5 µA cm⁻² for the other conditions. The observed low values for the current density plateaus may be ascribed to the controlled mass transport through the oxide layer.⁹⁷⁾ In all cases, the different protective coatings appear to be already protecting surfaces at their very early formation stages,^97,98) marked by the lack of an active region in which the current density is more elevated than the passivation plateau current density, observed in the case of stainless steels subjected to an acidic media. Hence, the corrosion current density (j_corr) was evaluated as the current density at the passivation plateau (j_pass), as in Refs. 98), 99). This way, a deep analysis of corrosion potential levels and j_pass values in Fig. 8 indicates a much better corrosion resistance for anodized samples than any other condition, even better for the anodized HPT-processed specimens, as in Ref. 20).

4.3 Bioactivity

Osseointegration demands the growth of a bone-like apatite layer on the implant surface, starting after implantation.¹⁰⁰⁾ Usually, osseointegration is estimated in relationship with its bioactivity, conducted by immersing the material in SBF solution¹⁰⁰⁾ to confirm its effectiveness in apatite generation. Hence, the samples’ bone-bonding ability of specimens in the diverse surface conditions was assessed by soaking them for 1, 7, and 14 days in SBF. A quantitative apatite formation assessment was performed by weighing samples before and after immersion in SBF (for a qualitative SEM analysis and other related techniques, please refer to Refs. 19), 20)). Figure 9 presents values of the average of three weighing measurements, where error bars regard the standard deviation and weighing balance uncertainties (0.2 mg). It is worthy of note that polished surfaces are not bioactivated even after 14 days, denoting the importance of surface activation treatments even for HPT-deformed samples.

Fig. 9

Mass gain of apatite after immersing samples in SBF as a function of the soaking time, comparing treated and untreated surfaces of deformed and undeformed specimens. (some similar results can be found in Refs. 19), 20)). (a) Ti–13Nb–13Zr and (b) Ti–35Nb–7Zr–5Ta.

It is also evident from Fig. 9 that even showing some bioactivity, anodized samples’ surfaces of undeformed and HPT-processed and the ones for HCl-treated deformed TNZT alloy started to be activated and delivered a minimal apatite quantity after 14 days. On the whole, for the TNZT alloy, one may infer that grain refinement is essential, but the three types of surface treatments could not activate the samples before 14 days.

Regarding the apatite formation for the TNZ alloy, there was precipitation since the first day of soaking only for anodized HPT-deformed samples. Also, Fig. 9 evidences again that the smallest the grain size, the highest the mass gain, also confirms the surface state’s effect by the redistribution and refinement of α′ and β phases led by HPT. The H₃PO₄ chemical surface treatment of the undeformed TNZ alloy was not as helpful for surface bioactivation as it was for the HCl-treated surface. However, superior activation is marked for the deformed samples, more for HCl treatment. Finally, comparing anodized and chemically treated surfaces, one may say that anodization and HCl treatment are foremost for the bioactivation of deformed and undeformed surfaces of both alloys than the H₃PO₄ one.

Deep explanations for the above behaviors are given in Refs. 19), 20). However, in summa, hydrophilicity was shown to be required because polished surfaces were hydrophobic, leading to the non-formation of apatite on their surfaces. Nevertheless, all other conditions were hydrophilic, which was insufficient to explain their behaviors. Interestingly, hydrophilic surfaces helped symbiosis between cells and biological fluids.^18,24) Additionally, anodized surfaces of cp-Ti showed to be much more bioactive than the anodized alloys in the deformed or not condition.^20,23) These facts mean that not only chemical reactions and chemistry impact apatite precipitation but also surface morphology and dissimilar charges emerging on the surfaces from various oxides naturally grown or produced by surface treatments. Morphologically, a surface composed of nanopores+nanotubes is not as good, and nanopores are even worse for apatite precipitation. Surface charges make precipitation worse in both anodized and chemically treated surfaces. In the case of chemically treated surfaces, it was also observed that the tentacles’ sizes, as in the case of cells,¹⁸⁾ seem to need a critical dimension to permit the bonding of biological fluids, favoring apatite formation. Grain refinement was vital for protein adsorption and cell fixing¹⁸⁾ and surface bioactivation, given the best results achieved by anodized and HCl-treated surfaces of HPT-deformed samples.

5. Conclusions

In this overview, besides cp-Ti for comparison, we have chosen three representative alloys used for implant application to describe the effect of HPT processing on grain refinement, phase transformation, mechanical and corrosion properties, and bioactivation. Samples were in the range of 1 to 6 GPa, the rotation speed was 3 rpm, and the total number of turns was up to 5. The main conclusions can be drawn as follows:

Dynamic phase transformation by HPT processing depends on the amounts of α or β stabilizers, the strain rate, applied loads, and starting phases and α morphologies. High quantities of α stabilizers (oxygen or aluminum) and α′ avoid phase transformation to ω. In the presence of α′, β is increased with decreasing pressure. An α phase with lamellae morphology is induced to ω transformation at higher pressures, while a fine globular α is induced to ω from low loads; however, at higher loads β is also transformed into α. When the amount of β is high enough, α transforms in β, ω is destabilized and does not precipitate. The transformation sequence is from α to β, then β is destabilized and transforms into ω. When α′ phase is present, ω is destabilized and does not precipitate.

As expected, HPT induced a very high grain refinement either by high pressures or an increased number of turns. However, the strength will depend not only on the number of grain boundaries but also on their nature, different phases and sizes, and strength. Young’s modulus depends, as expected, only on the phase. After processing, when β was increased, the modulus diminished.

There was a successful production of diversified nanostructures on samples’ surfaces for both compositions and deformed states by either anodization or chemical methods. Anodization produced nanotubes of oxides on surfaces where α was the phase, and nanopores (tubular morphology with thicker walls when β was the phase. Acid+alakaline treatment produced a sponge-like morphology, where the tentacle sizes of HCl-treated samples had similar sizes and similar to H₃PO₄-treated on TNZ deformed alloy. Tentacle sizes depend on sample composition and contaminants after their etching, mainly Zr.

Corrosion properties were very good for any condition, better for HPT-processed samples, and even better for anodized deformed samples. Other conditions (non-treated and chemically treated) had similar behavior as for cp-Ti. Corrosion potentials (surface stability) increased from polished surfaces to anodized ones. The current densities were very low for all the cases, indicating a controlled mass transport through the formed oxide layer (natural or after the various surface treatments).

Regarding bioactivity, polished surfaces were hydrophobic, resulting in their non-bioactivity. However, despite explaining a good symbiosis between cells and biological fluids, the hydrophilicity of treated surfaces (all hydrophilic) does not justify their bioactivity. The cp-Ti hydrophilic anodized surfaces showed to be much more bioactive than hydrophilic anodized alloys in the deformed or not condition, meaning that not only chemical reactions and chemistry impact apatite precipitation. Morphologically, compared to cp-Ti, a surface composed of nanopores+nanotubes or only nanopores are harmful for apatite precipitation. Surface charges make precipitation worse in both anodized and chemically treated surfaces. Tentacles’ sizes of chemically treated surfaces need a critical dimension to enable biological fluids’ bonding and favor apatite formation. Grain refinement was vital for protein adsorption, cell fixing, and surface bioactivation, given the best results achieved by anodized and HCl-treated surfaces of HPT-deformed samples.

Acknowledgments

A.M.J.J acknowledges the São Paulo Research Foundation - FAPESP (Brazil) under the grant FAPESP #2021/06546-1. R.Z.V. gratefully acknowledges the financial support partly from the Russian Science Foundation in the framework of Project No. 20-63-47027 and in part from the Mega-grant State Program (agreement 075-15-2022-1114 dated by June 30, 2022).

REFERENCES

• 1)   B.A.  Baker: J. Sport Rehabil. 26 (2017) 180–188. doi:10.1123/jsr.2016-0075
• 2)   M.P.  Soares dos Santos and  R.M.C.  Bernardo: Bioelectron. Med. 8 (2022) 15. doi:10.1186/s42234-022-00097-9
• 3)   M.  Geetha,  A.K.  Singh,  R.  Asokamani and  A.K.  Gogia: Prog. Mater. Sci. 54 (2009) 397–425. doi:10.1016/j.pmatsci.2008.06.004
• 4)   M.  Sloan,  A.  Premkumar and  N.P.  Sheth: J. Bone Joint Surg. Am. 100 (2018) 1455–1460. doi:10.2106/JBJS.17.01617
• 5)   K.  Pirruccio,  M.  Sloan and  N.P.  Sheth: J. Orthop. 16 (2019) 280–283. doi:10.1016/j.jor.2019.03.011
• 6)   S.G.  Memtsoudis,  M.C.  Besculides,  L.  Gaber,  S.  Liu and  A.  González Della Valle: Int. Orthop. 33 (2009) 1739–1745. doi:10.1007/s00264-008-0659-z
• 7)   M.  Long and  H.  Rack: Biomaterials 19 (1998) 1621–1639. doi:10.1016/S0142-9612(97)00146-4
• 8)   P.G.  Laing,  A.B.  Ferguson and  E.S.  Hodge: J. Biomed. Mater. Res. 1 (1967) 135–149. doi:10.1002/jbm.820010113
• 9)   M.  Niinomi: Metall. Mater. Trans. A 33 (2002) 477–486. doi:10.1007/s11661-002-0109-2
• 10)   T.  Albrektsson and  A.  Wennerberg: Int. J. Prosthodont. 17 (2004) 536–543 http://www.ncbi.nlm.nih.gov/pubmed/15543910.
• 11)   R.  Junker,  A.  Dimakis,  M.  Thoneick and  J.A.  Jansen: Clin. Oral Implants Res. 20 (2009) 185–206. doi:10.1111/j.1600-0501.2009.01777.x
• 12)   C.M.  Agrawal: JOM 50 (1998) 31–35. doi:10.1007/s11837-998-0064-5
• 13)  I. Sabirov, N.A. Enikeev, M.Y. Murashkin and R.Z. Valiev: Bulk Nanostructured Materials with Multifunctional Properties, (Springer International Publishing, Cham, 2015). doi:10.1007/978-3-319-19599-5.
• 14)   J.M.  Hernández-López,  A.  Conde,  J.  de Damborenea and  M.A.  Arenas: Corros. Sci. 94 (2015) 61–69. doi:10.1016/j.corsci.2015.01.041
• 15)   R.  Hazan,  R.  Brener and  U.  Oron: Biomaterials 14 (1993) 570–574. doi:10.1016/0142-9612(93)90172-X
• 16)   R.  Bosco,  E.R.U.  Edreira,  J.G.C.  Wolke,  S.C.G.  Leeuwenburgh,  J.J.J.P.  van den Beucken and  J.A.  Jansen: Surf. Coat. Technol. 233 (2013) 91–98. doi:10.1016/j.surfcoat.2013.02.039
• 17)   R.  Bosco,  J.  Van Den Beucken,  S.  Leeuwenburgh and  J.  Jansen: Coatings 2 (2012) 95–119. doi:10.3390/coatings2030095
• 18)   D.  Pedreira de Oliveira,  T.V.  Toniato,  R.  Ricci,  F.R.  Marciano,  E.  Prokofiev,  R.Z.  Valiev,  A.O.  Lobo and  A.M.  Jorge, Jr.: Int. J. Nanomedicine 14 (2019) 1725–1736. doi:10.2147/IJN.S197099
• 19)   D.A.G.  Pérez,  A.M.  Jorge, Jr.,  V.  Roche,  J.-C.  Lepretre,  C.R.M.  Afonso,  D.N.  Travessa,  G.H.  Asato,  C.  Bolfarini and  W.J.  Botta: J. Alloy. Compd. 812 (2020) 152116. doi:10.1016/j.jallcom.2019.152116
• 20)   D.A.G.  Pérez,  A.M.  Jorge, Jr.,  G.H.  Asato,  J.-C.  Lepretre,  V.  Roche,  C.  Bolfarini and  W.J.  Botta: J. Alloy. Compd. 796 (2019) 93–102. doi:10.1016/j.jallcom.2019.04.167
• 21)   J.E.  Berger,  A.M.  Jorge, Jr.,  G.H.  Asato and  V.  Roche: J. Alloy. Compd. 865 (2021) 158837. doi:10.1016/j.jallcom.2021.158837
• 22)   V.P.  Ricci,  C.R.M.  Afonso,  R.F.M.  dos Santos,  A.M.  Jorge, Jr. and  V.  Roche: Ceram. Int. 48 (2022) 27575–27589. doi:10.1016/j.ceramint.2022.06.052
• 23)   F.  Hilario,  V.  Roche,  R.P.  Nogueira and  A.M.  Jorge, Jr.: Electrochim. Acta 245 (2017) 337–349. doi:10.1016/j.electacta.2017.05.160
• 24)   R.P.  Nogueira,  J.  Deuzimar Uchoa,  F.  Hilario,  G. de F.  Santana-Melo,  L.M.R.  de Vasconcellos,  F.R.  Marciano,  V.  Roche,  A.M.  Jorge, Jr. and  A.O.  Lobo: Int. J. Nanomedicine 16 (2021) 667–682. doi:10.2147/IJN.S285805
• 25)   C.A.  Escobar Claros,  L.  Contri Campanelli,  A.M.  Jorge, Jr.,  J.-C.  Leprêtre,  C.  Bolfarini and  V.  Roche: Corros. Sci. 188 (2021) 109544. doi:10.1016/j.corsci.2021.109544
• 26)   G.V.B.  da Rocha,  J.H.  Lopes,  V.  Roche,  A.M.  Jorge,  R.  Riva,  A.C.  de Oliveira and  D.N.  Travessa: Appl. Surf. Sci. 603 (2022) 154460. doi:10.1016/j.apsusc.2022.154460
• 27)   D.P.  Oliveira,  E.  Prokofiev,  L.F.R.  Sanches,  V.  Polyakova,  R.Z.  Valiev,  W.J.  Botta,  A.M.  Jorge, Jr. and  C.  Bolfarini: J. Alloy. Compd. 643 (2015) S241–S245. doi:10.1016/j.jallcom.2014.11.115
• 28)   C.  Leinenbach and  D.  Eifler: Biomaterials 27 (2006) 1200–1208. doi:10.1016/j.biomaterials.2005.08.012
• 29)   A.J.  Engler,  S.  Sen,  H.L.  Sweeney and  D.E.  Discher: Cell 126 (2006) 677–689. doi:10.1016/j.cell.2006.06.044
• 30)   G.C.  Reilly and  A.J.  Engler: J. Biomech. 43 (2010) 55–62. doi:10.1016/j.jbiomech.2009.09.009
• 31)  D. Williams: Biomaterials Science: An Introduction to Materials in Medicine, ed. by B.D. Ratner et al., (Academic Press, 2004) (864pp., $95/£49.99), ISBN: 0-12-582463-7, (2005). doi:10.1016/j.biomaterials.2005.01.034.
• 32)  S.-W.H.E. Wintermantel: Medizintechnik: Life Science Engineering, (Springer Science & Business Media, Springer Berlin Heidelberg, Berlin, Heidelberg, 2009). doi:10.1007/978-3-540-93936-8.
• 33)   Y.G.  Ko,  W.S.  Jung,  D.H.  Shin and  C.S.  Lee: Scr. Mater. 48 (2003) 197–202. doi:10.1016/S1359-6462(02)00356-1
• 34)   S.  Semiatin,  V.  Seetharaman and  I.  Weiss: Mater. Sci. Eng. A 263 (1999) 257–271. doi:10.1016/S0921-5093(98)01156-3
• 35)   T.  Seshacharyulu,  S.C.  Medeiros,  W.G.  Frazier and  Y.V.R.K.  Prasad: Mater. Sci. Eng. A 284 (2000) 184–194. doi:10.1016/S0921-5093(00)00741-3
• 36)   M.  Niinomi: Int. J. Fatigue 29 (2007) 992–1000. doi:10.1016/j.ijfatigue.2006.09.021
• 37)  K. Wang: The Characterization of Ti–12Mo–6Zr–2Fe A New Biocompatible Titanium Alloy Developed for Surgical Implant, Beta Titan. Alloy. 1990’s, (1993) pp. 49–60. https://cir.nii.ac.jp/crid/1570854175107697152.bib?lang=en (accessed January 19, 2023).
• 38)   Y.  Okazaki,  Y.  Ito,  A.  Ito and  T.  Tateishi: Mater. Trans. JIM 34 (1993) 1217–1222. doi:10.2320/matertrans1989.34.1217
• 39)   J.A.  Davidson,  A.K.  Mishra,  P.  Kovacs and  R.A.  Poggie: Biomed. Mater. Eng. 4 (1994) 231–243. doi:10.3233/BME-1994-4310
• 40)   M.  Niinomi: Biomaterials 24 (2003) 2673–2683. doi:10.1016/S0142-9612(03)00069-3
• 41)   R.K.  Islamgaliev,  V.U.  Kazyhanov,  L.O.  Shestakova,  A.V.  Sharafutdinov and  R.Z.  Valiev: Mater. Sci. Eng. A 493 (2008) 190–194. doi:10.1016/j.msea.2007.08.084
• 42)   R.Z.  Valiev,  Y.  Estrin,  Z.  Horita,  T.G.  Langdon,  M.J.  Zechetbauer and  Y.T.  Zhu: JOM 58 (2006) 33–39. doi:10.1007/s11837-006-0213-7
• 43)   V.  Segal: Materials (Basel) 11 (2018) 1175. doi:10.3390/ma11071175
• 44)   R.  Valiev,  R.  Islamgaliev and  I.  Alexandrov: Prog. Mater. Sci. 45 (2000) 103–189. doi:10.1016/S0079-6425(99)00007-9
• 45)   R.  Valiev: Nat. Mater. 3 (2004) 511–516. doi:10.1038/nmat1180
• 46)   A.P.  Zhilyaev and  T.G.  Langdon: Prog. Mater. Sci. 53 (2008) 893–979. doi:10.1016/j.pmatsci.2008.03.002
• 47)   V.V.  Stolyarov,  L.O.  Shestakova,  Y.T.  Zhu and  R.Z.  Valiev: Nanostruct. Mater. 12 (1999) 923–926. doi:10.1016/S0965-9773(99)00268-8
• 48)   D.A.G.  Pérez,  A.M.  Jorge, Jr.,  C.S.  Kiminami,  C.  Bolfarini and  W.J.  Botta: Mater. Res. 20 (2017) 404–410. doi:10.1590/1980-5373-mr-2017-0037
• 49)   J.  Gallego,  T.S.  Pinheiro,  R.Z.  Valiev,  V.  Polyakova,  C.  Bolfarini,  C.S.  Kiminami,  A.M.  Jorge, Jr. and  W.J.  Botta: Mater. Res. 15 (2012) 786–791. doi:10.1590/S1516-14392012005000100
• 50)   T.S.  Pinheiro,  J.  Gallego,  C.  Bolfarini,  C.S.  Kiminami,  A.M.  Jorge, Jr. and  W.J.  Botta Filho: Mater. Res. 15 (2012) 792–795. doi:10.1590/S1516-14392012005000102
• 51)   A.  Zhilyaev,  G.  Nurislamova,  B.-K.  Kim,  M.  Baró,  J.  Szpunar and  T.  Langdon: Acta Mater. 51 (2003) 753–765. doi:10.1016/S1359-6454(02)00466-4
• 52)   T.N.  Kim,  A.  Balakrishnan,  B.C.  Lee,  W.S.  Kim,  K.  Smetana,  J.K.  Park and  B.B.  Panigrahi: Biomed. Mater. 2 (2007) S117–S120. doi:10.1088/1748-6041/2/3/S06
• 53)  R.Z. Valiev, A.P. Zhilyaev and T.G. Langdon: Bulk Nanostructured Materials: Fundamentals and Applications, (Wiley, New York, 2014). doi:10.1002/9781118742679.
• 54)   I.P.  Semenova,  G.K.  Salimgareeva,  V.V.  Latysh,  S.A.  Kunavin and  R.Z.  Valiev: Met. Sci. Heat Treat. 51 (2009) 87–91. doi:10.1007/s11041-009-9123-y
• 55)   Y.  Estrin and  A.  Vinogradov: Acta Mater. 61 (2013) 782–817. doi:10.1016/j.actamat.2012.10.038
• 56)   M.  Zehetbauer,  R.  Grössinger,  H.  Krenn,  M.  Krystian,  R.  Pippan,  P.  Rogl,  T.  Waitz and  R.  Würschum: Adv. Eng. Mater. 12 (2010) 692–700. doi:10.1002/adem.201000119
• 57)   A.  Azushima,  R.  Kopp,  A.  Korhonen,  D.Y.  Yang,  F.  Micari,  G.D.  Lahoti,  P.  Groche,  J.  Yanagimoto,  N.  Tsuji,  A.  Rosochowski and  A.  Yanagida: CIRP Ann. 57 (2008) 716–735. doi:10.1016/j.cirp.2008.09.005
• 58)   H.-S.  Ryu,  C.  Namgung,  J.-H.  Lee and  Y.-J.  Lim: J. Adv. Prosthodont. 6 (2014) 547. doi:10.4047/jap.2014.6.6.547
• 59)   I.P.  Semenova,  G.V.  Klevtsov,  N.A.  Klevtsova,  G.S.  Dyakonov,  A.A.  Matchin and  R.Z.  Valiev: Adv. Eng. Mater. 18 (2016) 1216–1224. doi:10.1002/adem.201500542
• 60)   Y.B.  Wang,  Y.H.  Zhao,  Q.  Lian,  X.Z.  Liao,  R.Z.  Valiev,  S.P.  Ringer,  Y.T.  Zhu and  E.J.  Lavernia: Scr. Mater. 63 (2010) 613–616. doi:10.1016/j.scriptamat.2010.05.045
• 61)   K.  Edalati,  E.  Matsubara and  Z.  Horita: Metall. Mater. Trans. A 40 (2009) 2079–2086. doi:10.1007/s11661-009-9890-5
• 62)   R.B.  Figueiredo,  F.L.  Sicupira,  L.R.C.  Malheiros,  M.  Kawasaki,  D.B.  Santos and  T.G.  Langdon: Mater. Sci. Eng. A 625 (2015) 114–118. doi:10.1016/j.msea.2014.11.091
• 63)   A.R.  Kilmametov,  Y.  Ivanisenko,  A.A.  Mazilkin,  B.B.  Straumal,  A.S.  Gornakova,  O.B.  Fabrichnaya,  M.J.  Kriegel,  D.  Rafaja and  H.  Hahn: Acta Mater. 144 (2018) 337–351. doi:10.1016/j.actamat.2017.10.051
• 64)   H.J.  Rack and  J.I.  Qazi: Mater. Sci. Eng. C 26 (2006) 1269–1277. doi:10.1016/j.msec.2005.08.032
• 65)   Q.  Guo,  Y.  Zhan,  H.  Mo and  G.  Zhang: Mater. Des. 31 (2010) 4842–4846. doi:10.1016/j.matdes.2010.05.047
• 66)   K.  Edalati and  Z.  Horita: J. Mater. Sci. 45 (2010) 4578–4582. doi:10.1007/s10853-010-4381-z
• 67)   R.Z.  Valiev and  T.G.  Langdon: Prog. Mater. Sci. 51 (2006) 881–981. doi:10.1016/j.pmatsci.2006.02.003
• 68)   G.I.  Raab,  R.Z.  Valiev,  T.C.  Lowe and  Y.T.  Zhu: Mater. Sci. Eng. A 382 (2004) 30–34. doi:10.1016/j.msea.2004.04.021
• 69)   M.  Ashida,  P.  Chen,  H.  Doi,  Y.  Tsutsumi,  T.  Hanawa and  Z.  Horita: Procedia Eng. 81 (2014) 1523–1528. doi:10.1016/j.proeng.2014.10.184
• 70)   R.B.  Figueiredo and  T.G.  Langdon: Mater. Res. 19 (2016) 1144–1148. doi:10.1590/1980-5373-MR-2016-0403
• 71)   H.  Shahmir and  T.G.  Langdon: Mater. Sci. Eng. A 667 (2016) 293–299. doi:10.1016/j.msea.2016.05.001
• 72)  I. Chattoraj: Stress Corrosion Cracking, (Woodhead Publishing, Elsevier, 2011) pp. 381–408. doi:10.1533/9780857093769.3.381.
• 73)   R.G.  Hennig,  D.R.  Trinkle,  J.  Bouchet,  S.G.  Srinivasan,  R.C.  Albers and  J.W.  Wilkins: Nat. Mater. 4 (2005) 129–133. doi:10.1038/nmat1292
• 74)   C.T.  Wang,  A.G.  Fox and  T.G.  Langdon: J. Mater. Sci. 49 (2014) 6558–6564. doi:10.1007/s10853-014-8248-6
• 75)  K.K. Sankaran and R.S. Mishra: Metallurgy and Design of Alloys with Hierarchical Microstructures, (Elsevier, 2017) pp. 177–288. doi:10.1016/B978-0-12-812068-2.00005-9.
• 76)   P.  Majumdar,  S.B.  Singh and  M.  Chakraborty: J. Mech. Behav. Biomed. Mater. 4 (2011) 1132–1144. doi:10.1016/j.jmbbm.2011.03.023
• 77)   A.  Korneva,  B.  Straumal,  A.  Kilmametov,  A.  Gornakova,  A.  Wierzbicka-Miernik,  L.  Lityńska-Dobrzyńska,  R.  Chulist,  Ł.  Gondek,  G.  Cios and  P.  Zięba: Materials (Basel) 14 (2021) 2262. doi:10.3390/ma14092262
• 78)   W.  Wang,  X.  Zhang,  W.  Mei and  J.  Sun: Mater. Des. 186 (2020) 108282. doi:10.1016/j.matdes.2019.108282
• 79)   A.X.Y.  Guo,  C.  Geng,  Z.  Lin and  S.C.  Cao: J. Mater. Sci. Eng. 11 (2022) 620. doi:10.37421/jme.2022.11.620
• 80)   G.T.  Aleixo,  C.R.M.  Afonso,  A.A.  Coelho and  R.  Caram: Solid State Phenom. 138 (2008) 393–398. doi:10.4028/www.scientific.net/SSP.138.393
• 81)   R.  Kolli and  A.  Devaraj: Metals (Basel) 8 (2018) 506. doi:10.3390/met8070506
• 82)   K.  Edalati: Mater. Trans. 60 (2019) 1221–1229. doi:10.2320/matertrans.MF201914
• 83)   R.Z.  Valiev,  M.Yu.  Murashkin and  I.  Sabirov: Scr. Mater. 76 (2014) 13–16. doi:10.1016/j.scriptamat.2013.12.002
• 84)   K.  Ma,  Y.  Zheng,  S.  Dasari,  D.  Zhang,  H.L.  Fraser and  R.  Banerjee: MRS Bull. 46 (2021) 250–257. doi:10.1557/s43577-021-00066-8
• 85)   E.O.  Hall: Nature 173 (1954) 948–949. doi:10.1038/173948b0
• 86)   E.O.  Hall: Proc. Phys. Soc. B 64 (1951) 747–753. doi:10.1088/0370-1301/64/9/303
• 87)  N.J. Petch: The cleavage strength of polycrystals, I. Iron Steel Inst. (1953).
• 88)   E.J.  Pavlina and  C.J.  Van Tyne: J. Mater. Eng. Perform. 17 (2008) 888–893. doi:10.1007/s11665-008-9225-5
• 89)   R.Z.  Valiev and  Y.T.  Zhu: Trans. Mater. Res. Soc. Jpn. 40 (2015) 309–318. doi:10.14723/tmrsj.40.309
• 90)   N.  Balasubramanian and  T.G.  Langdon: Metall. Mater. Trans. A 47 (2016) 5827–5838. doi:10.1007/s11661-016-3499-2
• 91)   R.Z.  Valiev,  Y.  Estrin,  Z.  Horita,  T.G.  Langdon,  M.J.  Zehetbauer and  Y.T.  Zhu: Mater. Res. Lett. 4 (2016) 1–21. doi:10.1080/21663831.2015.1060543
• 92)   R.Z.  Valiev,  N.A.  Enikeev and  T.G.  Langdon: Kovov. Mater. 49 (2011) 1–9. doi:10.4149/km_2011_1_1
• 93)   P.V.  Liddicoat,  X.-Z.  Liao,  Y.  Zhao,  Y.T.  Zhu,  M.Yu.  Murashkin,  E.J.  Lavernia,  R.Z.  Valiev and  S.P.  Ringer: Nat. Commun. 1 (2010) 63. doi:10.1038/ncomms1062
• 94)   R.Z.  Valiev,  N.A.  Enikeev,  M.Yu.  Murashkin,  V.U.  Kazykhanov and  X.  Sauvage: Scr. Mater. 63 (2010) 949–952. doi:10.1016/j.scriptamat.2010.07.014
• 95)   M.M.  Abramova,  N.A.  Enikeev,  R.Z.  Valiev,  A.  Etienne,  B.  Radiguet,  Y.  Ivanisenko and  X.  Sauvage: Mater. Lett. 136 (2014) 349–352. doi:10.1016/j.matlet.2014.07.188
• 96)   S.  Scheriau,  Z.  Zhang,  S.  Kleber and  R.  Pippan: Mater. Sci. Eng. A 528 (2011) 2776–2786. doi:10.1016/j.msea.2010.12.023
• 97)   V.  Roche,  Y.  Champion,  I.A.  Bataev and  A.M.  Jorge, Jr.: Corros. Sci. 207 (2022) 110607. doi:10.1016/j.corsci.2022.110607
• 98)   A.  Carnot,  I.  Frateur,  P.  Marcus and  B.  Tribollet: J. Appl. Electrochem. 32 (2002) 865–869. doi:10.1023/A:1020510506504
• 99)   S.  Chakri,  P.  David,  I.  Frateur,  A.  Galtayries,  P.  Marcus,  E.  Sutter,  B.  Tribollet,  V.  Vivier and  S.  Zanna: Matériaux & Tech. 103 (2015) 209. doi:10.1051/mattech/2015016
• 100)   X.  Liu,  P.K.  Chu and  C.  Ding: Mater. Sci. Eng. R Rep. 47 (2004) 49–121. doi:10.1016/j.mser.2004.11.001