Preparation methods, mechano-bactericidal performance and mechanisms of Ti-based nanostructure arrays: A review Authors: Wenfang Li, Shujiao Zhang, Rui Liu, Fangchong Xing, Xingcai Guo, Lidan Zhao Published: 2026 ABSTRACT AbstractMedical titanium (Ti) and its alloys have been widely employed in biomedical field, but Ti-based implants are prone to bacterial biofilm contamination, leading to implant failure. Nanostructure arrays, as a new strategy, exhibit effective bactericidal performance and biofilm inhibition only by using their own morphological characteristics, drawing extensive attention. This review mainly summarizes the progress on natural and biomimetic mechano-bactericidal nanostructure arrays, as well as the various preparation techniques for Ti-based bactericidal nanostructured surfaces, and highlights the necessity of developing controllable preparation methods. Furthermore, the critical discussion of the underlying mechano-bactericidal mechanisms is presented, identifying the need for additional experimental data to refine bacterial modeling and elucidate the complex interplay between nanostructure arrays and bacterial cells. Finally, the challenges and opportunities for Ti materials with bactericidal nanostructure arrays as implants or clinical devices are analyzed as well.Keywords: Bactericidal performance, Bactericidal mechanism, Nanostructure arrays, Titanium Introduction Titanium (Ti) and its alloys are widely used in modern medical fields, such as orthopedic and dental implants, due to its excellent biocompatibility, chemical inertness, and corrosion resistance [1–3]. Although Ti exhibits good tolerance in vitro environment, it is prone to complications in vivo, especially when bacterial biofilms form on its surface, which significantly affects the implants performance [2, 4]. In addition, it is a frequently cited historical estimate from the US National Institutes of Health that over 80% of microbial infections in the human body are associated with bacterial biofilms [5]. The enhanced antimicrobial resistance associated with such biofilms not only creates a severe clinical and economic burden but also drives the urgent need for novel, non-chemical antibacterial strategies, such as the development of physically bactericidal topographies. Following the discovery that the conical nanopillar structure on cicada wing surface and the nanoprotrusion structure on dragonfly wing surface can pierce bacterial cells through pure mechanical action [6–9], this kind of mechano-bactericidal surface with nanostructures has received more and more attention in recent years [10–12]. In addition, many natural mechano-bactericidal surfaces have been widely reported [6, 11, 13–17], such as gecko skin [15], damselfly wings [16], etc. In order to broaden the application of these natural surface nanostructures in different fields, a large number of bionic bactericidal nanostructures have been designed on surfaces of different materials, including Ti and its alloys [18–32], polymers [33–37], aluminum and its alloys [38], bSi [8, 39–41] and silicon wafers [42]. The main preparation methods of Ti-based surface nanostructures include: Reactive Ion Etching (RIE) [28, 29], Ion Irradiation (II) [43], Glancing Angle Sputter Deposition (GLAD) [27], Thermal Oxidation (TO) [44], Hydrothermal Process (HP) [18–26, 45] and Anodic Oxidation (AO) [46–54], etc. Among them, HP and AO are the most commonly used due to their advantages of simple operation, cost saving and environmental friendliness [55]. Although the latter can be used to prepare a variety of nanostructures on Ti-based surfaces, concerning nanowires [46–48], nanopores [49], and nanotubes [50, 51], etc., and some studies have reported that the nanostructures prepared by this method exhibit antibacterial performance, this performance is a coupling of multiple factors, involving topographical effects, surface roughness, crystallinity, fluoride residues, wettability, photocatalytic activity, or incorporated antibacterial elements, rather than being based on the pure physical geometry [52–54]. Nanostructures can lyse bacterial cells mainly through mechanical action [6, 8], and their surface bactericidal efficiency depends on themselves geometric characteristics [33, 41, 56, 57]. In order to comprehensively understand the mechano-bactericidal mechanism of nanostructures for further constructing a more efficient physical bactericidal surface and, it is necessary to clarify the mechanical response during the interaction between bacterial cells and different nanostructures. It should be mentioned that it is difficult to directly capture the mechanical response of bacterial cells to nanostructures in experimental studies, and it is impossible to control the geometric parameters of nanostructures, such as radius, height and spacing, with a single variable. Therefore, theoretical calculation is usually adopted to study the interaction between nanostructures and bacterial cells [56, 58–63]. Physical bactericidal surfaces with nanostructures provide a new idea for the anti-infectious treatment of medical Ti materials and have become attractive research interests in recent years. Therefore, it is necessary to overview the latest developments in this field. Over the past five years, several reviews have sorted out relevant topics from their respective perspectives, covering the physical mechanisms of mechano-bactericidal nanostructures across different materials [64, 65], the Ti-based mechano-bactericidal surfaces [66], the osseointegration performance of Ti-based nanotube arrays [50], the surface engineering and biocompatibility of Ti-based implants [67, 68], as well as the biomedical applications of Ti and its alloys [69, 70]. However, these reviews all focus on a specific aspect, and there is still a lack of a comprehensive review that systematically integrates the preparation methods, bactericidal performance and mechanisms, and clinical translation potential of Ti-based mechano-bactericidal nanostructured arrays. In this review, it highlights several key aspects (Fig. 1), including: (i) progress on natural/biomimetic mechano-bactericidal nanostructures; (ii) preparation technologies of Ti-based nanostructures and their effect on surface morphology and bactericidal performance; (iii) mechano-bactericidal mechanisms of nanostructure arrays; (iv) challenges and opportunities faced by Ti materials with bactericidal nanostructures as implants or clinical devices. Natural bactericidal surfaces as well as the morphology, preparation methods, and mechanisms of Ti-based bactericidal nanostructure arrays The literature search for this review was conducted in the Web of Science Core Collection. In order to avoid introducing a large amount of irrelevant literature, we adopted a stepwise focused retrieval strategy. A general primary search was first performed using broad keywords to obtain an initial candidate literature pool, followed by secondary retrieval for Sects. 3, 4, 5, and 7, in which topic-specific secondary keywords were respectively applied. The primary keywords and initial hits, the secondary keywords and secondary hits for each section, the inclusion/exclusion criteria, and the final included counts are summarized in Table 1. The final included counts were determined through manual screening based on the inclusion/exclusion criteria. It should be noted that Sects. 1 and 2 serve as background context for this review, which are summarized based on classical literatures and consensus knowledge in this field; Sect. 6 (Standardized evaluation) primarily references international standard documents. Therefore, these three sections were not subjected to systematic literature retrieval. Table 1Search keywords, inclusion/exclusion criteria, and literature screening results for some section of this reviewSectionInitial keywordsInitial hitsSecondary keywordsSecondary hitsinclusion/exclusion criteriaFinal hits3nanostructure or nanopillar or nanowire or nanotube or nanosheet or nanospike, bactericidal or antibacterial, surface or arrays3198natural or biomimetic or titanium, mechano-bactericidal or physical or topographical432Inclusion: Natural, biomimetic, or Ti-based nanostructures with bactericidal performance; explicit evidence of topography-driven bacterial killing. Exclusion: Chemically modified surfaces (e.g., Ag/Cu doping, drug loading); photocatalysis-dominated studies; no direct evidence of mechanical killing.1174titanium, mechano-bactericidal or physical or topographical210Inclusion: Fabrication of Ti-based nanostructures with mechano-bactericidal performance. Exclusion: chemically modified post-treatment (e.g., Ag loading).655Mechano-mechanism or membrane rupture or interaction or geometry287Inclusion: Experimental studies with direct morphological evidence of bacterial damage; computational studies on mechanical interactions; parametric studies correlating geometry with antibacterial efficacy. Exclusion: Studies without mechano-bactericidal focus; purely chemical/photocatalytic antibacterial studies.477titanium, mechano-bactericidal or physical or topographical, in vivo or in vitro or protein adsorption or clinical or osteogenic or cytotoxicity or biocompatibility120Inclusion: In vivo or vitro studies on Ti-based nanostructures with mechano-bactericidal performance; translational perspectives; biosafety evaluations. Exclusion: studies not addressing clinical relevance or biocompatibility, not mechano-bactericidal performance48 Search keywords, inclusion/exclusion criteria, and literature screening results for some section of this review Applications and challenges of medical Ti and its alloys Medical Ti and its alloy materials as a type of functional materials used in the biomedical field, are mainly utilized in the surgical implants, orthopedic instruments and other akin products [71]. According to the classification standards of material properties, it can be classified into the category of “metal materials” in “materials for implants” [71, 72]. Biomedical metal materials are the most suitable choice to replace the failed hard tissues due to their excellent mechanical properties. They mainly include three series: stainless steel, cobalt-based alloys, Ti and its alloys, accounting for approximately 40% of the entire biomaterials market share [71, 73]. Compared with other biomedical metal materials, Ti and its alloys are the most commonly used materials for bone tissue replacement and repair, because they have not only high strength and good toughness, but also good biocompatibility [72, 74, 75]. For instance, they are often adopted in the orthopedics field for hip and knee joint replacements, as well as shoulder and elbow joint implants [69], and also used in spinal correction and fixation devices, spinal fusion devices, and replacement of spinal intervertebral discs [76, 77]. In addition, the instruments that support fracture fixation and reconstruction are also usually made of Ti, such as bone plates, bone mesh, screws, bone rods, etc. [69, 78]. Besides the above applications, they are also widely utilized in the dental field [3, 79, 80] due to a repair effect similar to natural teeth [81]. Meanwhile, some alloys of TiMo and TiNi have been used to make orthodontic braces because of their memory properties and special elasticity [80]. Note that the cardiovascular diseases remain the leading global cause of death. Implanting a new Ti alloy stent with shape memory properties at the site of vascular stenosis provides a new strategy to solve this problem, which has received great attention and research [79, 82]. Figure 2A shows the distribution of medical Ti and its alloys used in the human body. Application of medical Ti and its alloys associated with their distribution in the human body (A) [69], Copyright © 2021 Springer Nature, and Images of Ti-based orthopedic and dental implants (B) It should be mentioned that, implants are susceptible to contamination by bacteria on the skin and mucous membranes during surgery [83], causing the formation of biofilms on the implant surface (Fig. 2B) [74]. As shown in Fig. 3A, once a mature biofilm forms, it is difficult to eradicate the bacterial infection, even with antibiotic treatment and repeated surgical irrigation and debridement, may leading to joint loosening, amputation or death [74, 84]. It is worth noting that the implant materials, such as Ti and its alloys, cannot resist bacterial adhesion, proliferation and biofilm formation by themselves. Hence, the development of surfaces with antimicrobial functionality is critical to the success of implantable medical device procedures. A scenario for the surface biofilm formation (A) [74], Copyright © 2018 Elsevier, and the advantages and disadvantages of chemical and physical bactericidal mechanisms (B) To prevent surface bacterial biofilm formation, two strategies are commonly employed to inhibit bacterial adhesion and growth (Fig. 3B). One of them is based on chemical mechanisms, such as applying the implant surface with antibiotics [85–87] or other bactericidal substances, including quaternary ammonium compounds [88] or antimicrobial peptides [89–91], heavy metals [92, 93] or oxidants [94], etc. While the main disadvantage for this method is that it cannot control the release concentration of bactericidal substances, which may shorten the effective bactericidal time and even lead to the emergence of inflammatory reactions, toxicity to animal cells, and the development of drug resistance in bacteria after long-term use [67, 86, 95]. In particular, the bacterial resistance to antibiotics often brings difficulties to treatment [96, 97]. It is predicted that the number of deaths caused by and related to bacterial antimicrobial resistance could reach up to 10 million by 2050 [98]. The other one is chemistry-independent, only dependent on the surface with micro/nanopatterned topography which can effectively inhibit bacterial adhesion [99, 100] or kill bacteria [13, 64, 101]. Compared with chemical method, surfaces with nanostructures can prevent microbial contamination and inhibit the formation of biofilms solely by utilizing their own morphological features, which is a safe and sustainable new approach [102]. Mechano-bactericidal nanostructures Natural/Biomimetic bactericidal nanostructures The natural surface with mechano-bactericidal performance is firstly reported by Ivanova et al. [6], and they found that the Pseudomonas aeruginosa (P. aeruginosa)adhered on cicada wing surface (Fig. 4Aa) can be killed within 3 min. This surface is composed of nanopillar arrays, and the corresponding height, top cap diameter, bottom diameter and spacing are 200 nm, 60 nm, 100 nm and 200 nm, respectively. Moreover, the phenomenon of nanopillars pinning and penetrating P. aeruginosa cells is clearly captured using a focused ion beam scanning electron microscope (FIB-SEM) (Fig. 4Ab), indicating that the cells move downward on cicada wing surface. In addition, in order to determine whether the chemical composition of cicada wing surface affects the bactericidal performance against P. aeruginosa, the gold film of 10 nm thick is deposited on the surface using magnetron sputtering. Although the surface chemical elements change after gold spraying, the morphology of nanopillars is not influenced, and there is no significant difference in the bactericidal effect of P. aeruginosa on the surface before and after gold spraying. The results suggest that the bactericidal performance of cicada wing is mainly attributed to the surface physical structure, rather than the surface chemical properties. This nanostructured surface can prevent bacteria from developing drug resistance and is a safe and sustainable new method with great potential for application in the field of antibacterial materials. Nanostructures on natural surfaces and their bactericidal performance. Photograph of a cicada and its wings (Aa), and the FIB-SEM image of P. aeruginosa cells adhered to the nanopillar surfaces (Ab) [6]. Copyright © 2012 John Wiley and Sons. Photograph of a dragonfly (Ba), and the SEM images of P. aeruginosa (Bb) and S. aureus (Bc) cells adhered to its surface nanostructures, respectively [8]. Photograph of a gecko (Ca), together with the microspine structures on its skin surface (Cb), and the SEM image of the Porphyromonas gingivalis on the ones (Cc) [103]. Copyright © 2015 Elsevier. Photographs (Da, Ea) of lotus and locust leaves and SEM images (Db, Eb) showing E. coli cells punctured by the nanostructures on the surface of lotus and locust leaves, respectively [104, 105], Copyright © 2020 & 2023 Elsevier Subsequently, the bactericidal performance of cicada wing surface against various bacteria is evaluated by Ivanova team [7], and they found that this surface is able to kill Gram-negative bacteria with different shapes, while Gram-positive bacteria remain highly active. In addition, they also studied the bactericidal performance of surfaces for dragonfly wings (Fig. 4Ba) and black silicon (bSi) with nanostructures, and their bactericidal behavior of these surfaces is compared to that previously obtained from cicada wing surfaces [8]. Dragonfly wings (Figs. 4Bb and 4Bc) and bSi surface not only show significant killing effects on Gram-negative P. aeruginosa, but also destroy the Gram-positive Staphylococcus aureus (S. aureus) cells and Bacillus subtilis (B. subtilis) spores. Both surface nanostructures have high bactericidal efficiency against S. aureus, which are 4.6 × 105 and 4.5 × 105 cells cm− 2 min− 1, respectively (Table 2). The nanocones on the bSi surface are synthesized by reactive ion etching, and its height is twice as high as the nanoprotrusions on the dragonfly surface (Table 2). Interestingly, the surfaces of dragonfly wing and bSi are observed to assemble into hierarchical nanoprotrusion clusters, and the corresponding size, shape, and spatial distribution are random. Note that this is the first time to report the physical bactericidal performance of surface hierarchical nanostructures. Table 2Comparison in characteristics and bactericidal performance between insect wings and bSi surfaces [8]SurfaceCicada wingDragonfly wingbSiSurfacecharacteristicsWater contact angle159°153°80°Chemical compositionLipids/waxesLipids/waxesMostly SiO2Nanostructure height200 nm240 nm500 nmBactericidalactivityEffectivenessGram-negativeGram-negativeGram-negative-Gram-positiveGram-positive-SporesSporesEfficiency (cells killed per cm− 2 min− 1) P. aeruginosa 2.0 × 1053.0 × 1054.3 × 105 S. aureus -4.6 × 1054.5 × 105 B. subtilis -1.4 × 1051.4 × 105Note: The bactericidal efficiency in the table is the mean value within 3 h Comparison in characteristics and bactericidal performance between insect wings and bSi surfaces [8] Surface characteristics Bactericidal activity Note: The bactericidal efficiency in the table is the mean value within 3 h In present, the structural characteristics of natural antimicrobial surfaces and their interactions with different bacteria/cells have received widespread attention. Watson et al. [103, 106] found that the skin of gecko (Fig. 4Ca) can efficiently kill oral pathogenic bacteria (e.g. Porphyromonas gingivalis, Fig. 4Cc), and exhibits a good biocompatibility to eukaryotic cells. It is composed of scales with micro-spines in length from hundreds nm to several µm, in spacing of sub-micron order and curvature radius of generally 10–20 nm (Fig. 4Cb). Nowlin et al. [107] studied the interaction between two types of Saccharomyces cerevisiae cells and surfaces of a dragonfly wing and two types of cicada wings, and found that the nanodendritic cluster with random height distribution on the surface of dragonfly wing shows a stronger killing effect on Saccharomyces cerevisiae cells than the nanopillar with uniform height on the surface of cicada wing. This study also proposes and demonstrates that the amount of yeast cells rupture induced by nanostructured surfaces depends on the adhesion strength of the ones. The similar results are also reported by Kelleher et al. [108], who investigated the nanostructure morphologies of three types of cicada wing surfaces and their bactericidal performance, and found that the bacterial activity is significantly correlated with the morphology characteristics of surface nanostructures. Dehghani et al. [109] further demonstrated that nanostructures on wing surface with higher roughness, skewness and kurtosis presents excellent bactericidal performance by using atomic force microscopy technology (AFM). Recently, Ivanova’s team [16] reported that damselfly wings can efficiently kill the S. aureus and P. aeruginosa cells in the early growth stage (1 h) and mature stage (24 h), and the corresponding bactericidal efficiency is 89.7% and 61.3%, 97.9 and 91.3%, respectively. This is because their surface is covered with densely arranged nanoprotrusions, having the average height of 433 ± 71 nm, tip diameter of 48 ± 11 nm and spacing of 116 ± 40 nm. In addition to animal wings, the surface of plant leaves with nanostructures also possesses a bactericidal performance. For instance, the bacteria attached on the surface of lotus leaves (Fig. 4Da) has been killed by the nanostructure on the surface via penetrating the bacteria’s body (Fig. 4Db) [104]. This surface is of a layered structure composed of micro-protrusions with 5–10 μm in diameter and nanowax tubes in length of 0.3–1.1 μm, and diameter of 0.1–0.2 μm. Chen et al. [105] reported that the bactericidal efficiency of Escherichia coli (E. coli) on the surface of Sophora acacia leaves (Fig. 4Ea) is as high as 99.8%. The surface is composed of randomly arranged nanosheets (Fig. 4Eb) with approximately 3.1 × 108 cm− 2 in density. Inspired by natural surface bactericidal nanostructures, a large number of bionic bactericidal nanostructures have been designed on the surfaces of different materials, such as bSi [39–41] and silicon wafers [42], polymers [33–37, 110, 111], Ti and its alloys [18–29, 44, 112]. In 2013, the Ivanova team [8] first studied the physical bactericidal performance of bSi surfaces with nanodendritic structure similar to the dragonfly wing surface. Soon afterwards, in 2017–2018, the team continues to design and prepare superhydrophobic and superhydrophilic bSi surfaces with different spatial geometric characteristics, and found that all types of nanopillar surfaces are effective in killing P. aeruginosa and S. aureus, with bactericidal efficiencies of 85% and 89% respectively. In addition, the adherent amount of both bacterium decreases with the decrease in the height, diameter and inter-pillar spacing of the nanopillars [39, 40]. They also demonstrated that the small variation in size for nanostructure can generate obvious influence on its physical sterilization performance [41]. Hasan et al. [42] prepared a “super surface” on a silicon wafer using ion etching, and it consists of nanopillars with a height of 4 μm and a diameter of 220 nm, with the randomly distributed spacing of pillars. And the bacterial experiments showed that both Gram-negative E. coli (Figs. 5Aa and 5Ab) and Gram-positive S. aureus cells (Figs. 5Ac and 5Ad) attached to the nanostructure surface are punctured and killed by the sharp tapered-nanopillars. Bactericidal performance of bionic surface nanostructures. (Aa, Ab) and (Ac, Ad) show the SEM images and fluorescence images of E. coli and S. aureus cells adhered to the nanopillars on silicon wafer surface, respectively [42]. (Ba), (Bb), and (Bc) present the SEM morphology images of nanocones with different heights on polycarbonate surface and the fluorescence image of E. coli cells adhered to its surface, respectively [33], Copyright © 2020 American Chemical Society The physical bactericidal performance of polymer surface nanopillars is first reported by Dickson et al. [34], who prepared three different bionic nanopillar arrays on the surface of polymethylmethacrylate film by utilizing nanoimprint technology, and found that the surface with the densest nanopillar can reduce the bacterial concentration of E. coli suspension by 50% within 24 h. Hazell et al. [35] constructed a polyethylene terephthalate nanocone array using the polystyrene nanosphere-mask colloid lithography process, and the height, tip width and bottom diameter of a single nanocone is regulated by changing the oxygen plasma etching time. The results show that the surface bactericidal efficiency of nanocones with a spacing of 200 nm, an aspect ratio over 3, and a tip width less than 20 nm is the highest, approximately 30%. At approximately the same time, Yamada et al. [110] developed an artificial bionic moth eye mask with antibacterial performance on polyethylene terephthalate film via nanoimprint technology, and it shows a high bactericidal efficiency in in vitro experiments. In 2020, our team [33, 111, 113] used anodized aluminum template-assisted hot embossing and wet etching to prepare a series of size-controllable nanostructures on the surface of polycarbonate and polymethylmethacrylate, and systematically studied the effect of geometric parameters (height, diameter and spacing) of the polycarbonate surface nanostructures on their bactericidal efficiency. The results show that for the bactericidal nanostructure, there is a critical height, approximately 200 nm (Figs. 5Ba–Bc); and the smaller the diameter of nanopillars’ top, the greater the pressure on the bacterial cells, which easily rupture the cells. In addition, there is also an optimal inter-pillar spacing. In 2022, Lohmann et al. [36] adopted UV imprint lithography technology to construct flexible nanopillar resin arrays with controllable size and various Young’s moduli ranging from 208 MPa to 4 GPa. They found that when the diameter and Young’s modulus of nanopillars are smaller than 100 nm and larger than 1.3 GPa, respectively, the bacterial mortality increases 3.5–5.6 times, while thicker or softer nanopillars do not reduce the bacterial cell activity. Ti-based bactericidal nanostructures In addition to constructing biomimetic bactericidal surfaces on non-metallic materials of silicon and polymers, the physical bactericidal nanostructures on the surface of Ti-based medical metal as orthopedic and dental implants have also been extensively studied (Fig. 6), to solve the implant failure caused by infections[18–24, 26, 29, 44, 112]. At present, the topological configuration of bactericidal nanostructures on the surface of Ti and its alloys mainly includes nanowires (Fig. 6A, C, E, H and I) [18–21, 23, 30, 31], nanopillars (Fig. 6D and F) [27–29, 44] and nanosheets (Fig. 6G and H) [26, 30]. Research progress of Ti-based bactericidal nanostructures. (A-C, E, I) nanowires [18–21, 31], Copyright © 2023 American Chemical Society. (D, F, J) nanopillars [28, 29, 43], Copyright © 2018 & 2024 Elsevier. (G) nanosheets [26], Copyright © 2019 Elsevier. (H) nanowire/sheet cluster hybrid structures [30], Copyright © 2021 American Chemical Society. (K) cactus-like nanospikes [45] Diu et al. [18] synthesized brush-like and network-like titanium dioxide (TiO2) nanowire arrays on the Ti surface (Fig. 6A) using hydrothermal method (240 ℃), in which the former is composed of uniform and dense spike-like structures with an average height of about 3 μm, while the latter is formed by entangled tips of longer nanowires, and the diameter of this network is in the range 10–15 μm. Meanwhile, the behavior of bacterial cells and mammalian cells adhered to both TiO2 nanostructures was also studied. The results show that both nanowire arrays can effectively kill P. aeruginosa, but have basically no effect on S. aureus. This selective bactericidal phenomenon is consistent with the surface of cicada wings. In addition, there is no significant difference in the bactericidal efficiency for the nanostructures before and after gold plating, but the surface chemical components have changed after gold spraying, indicating that TiO2 nanowires are physically bactericidal effect. Unlike the bacterial cells, these nanostructures have no effect on the viability of osteoblast-like cells MG-63. Tsimbouri et al. [20] constructed three types of nanowire arrays on the pure Ti surface by adjusting the hydrothermal reaction temperature (240 °C) and duration (2 h, 2.5 h and 3 h), and studied the interaction between human osteoblasts and P. aeruginosa cells and the nanowire arrays (Fig. 6C). It is found that the distribution and skeleton organization of osteoblasts on the nanowire surface prepared only for 2 h is the best, so this surface is selected for subsequent studies of cell differentiation and bacterial experiments. The results show that the bactericidal efficiency of the nanowire surface for 2 h is about 30% after 1 h of bacterial culture, and the bactericidal efficiency after 18 h of culture is about 58%. Cao et al. [21] studied the long-term bactericidal performance of nanostructures prepared on the Ti surface by hydrothermal method (240 ℃) against S. aureus (Fig. 6E). They found that although the number of bacteria adhering to the spear-shaped structure is small in a short period, a thick biofilm is formed on its surface after 6 days. The pocket structure can delay the formation of biofilm for 6 days and can kill bacterial cells through penetration or extrusion, with bactericidal efficiency of 47%. The micro/nano-scale pillar structure on the Ti-based surface also exhibits a good bactericidal performance. A bSi-surface-like anisotropic Ti nanopillar is built by Hasan et al. [28] employing reactive ion etching technology. The bactericidal efficiency of this surface against Gram-negative E. coli, P. aeruginosa, and Gram-positive S. aureus cells is approximately 95%, 98%, and 22%, respectively. And with the adhesion duration increasing from 4 h to 24 h, the bactericidal efficiency of S. aureus increases to 76%, which is similar to results reported by Tsimbouri et al. [20] (Fig. 6C), suggesting that prolonging adhesion duration can improve the surface bactericidal performance. In addition, they also found that the surface not only facilitates the attachment and proliferation of human bone marrow mesenchymal stem cells, but also enhances osteogenesis. Hence, they believe that this nanostructure surface may be very suitable for use in the next generation of orthopedic implants with multiple biological functions. The above preparation technology is also adopted by Ivanova et al. [29] to construct the dense and vertical micron-sized pillars on the Ti surface, and the bactericidal efficiency of this surface against P. aeruginosa and S. aureus is about 87.2% and 72.5%, respectively. In addition to bactericidal nanopillars, in 2020, a Ti-based nanosheet structure with bactericidal performance is reported by the Ivanova et al. [26] They systematically studied the relationship between the surface hydrothermal treatment duration and the corresponding bactericidal efficiency, and found that after hydrothermal treatment for 6 h, the surface with ~ 10 nm thick sharp nanosheets (Fig. 6G) exhibits the highest bactericidal efficiency against both Gram-negative P. aeruginosa and Gram-positive S. aureus, which are ~ 99% and ~ 90% respectively. In 2021, Zhao et al. [30] described a three-step method of acid etching-NaOH hydrothermal treatment (80 °C)-heat treatment to prepare a nanocluster structure with bactericidal performance on the Ti surface. The results show that the surface with a mixed nanowire/sheet cluster structure has the highest bactericidal efficiency for both E. coli and S. aureus, which are ~ 98.6% and ~ 69.8% respectively (Fig. 6H). In 2023, Morel et al. [31] constructed a nanowire array on the Ti surface through the KOH hydrothermal method under a relative low temperature of 75 °C (Fig. 6I), and the results suggested that the surface nanostructure morphology and the corresponding bactericidal performance are no different from those using high-temperature hydrothermal method (150 °C). Lately, Mulyukov et al. [43] utilized argon ion irradiation to construct spike-like nanostructures on Ti surface, and no alive E. coli cells were found on the homogeneous nanospike surface (Fig. 6J), indicating its bactericidal efficiency can reach 100%. In 2025, Huang et al. [45] prepared cactus-like nanostructures on the Ti surface through 3D printing-NaOH hydrothermal treatment (150 °C) (Fig. 6K), believing that surface roughness will significantly change the effectiveness of hydrothermal treatment. It should be mentioned that there are still several challenges for biomimetic bactericidal nanostructured surface even though it has been extensively studied: (1) The nanostructure surface will be covered by bacterial debris with the increase in service time, thus affecting its bactericidal performance [16]; (2) Nanostructures can only kill bacteria adhered to the surface, and when the environment is rich in bacteria, it is difficult to quickly kill a large number of bacteria by nanostructures alone, easily resulting in the biofilm formation. In sum, finding methods that can release the remaining dead bacteria on the implant surface [114, 115], non-contact bactericidal performance of implant surface [116, 117], and finding more effective physical bactericidal strategy [118] have become the focus in recent years. For instance, in 2024, Chen et al. [116] utilized the silver nanoparticles to coat the surface of hydrothermal method-prepared titanium dioxide nanoneedle, thereby making it have a strong bactericidal ability against E. coli and S. aureus, accompanied with the good biocompatibility. Preparation of Ti-based bactericidal nanostructures Reactive ion etching Reactive ion etching (RIE) technology belongs to plasma dry etching, which is a rapid etching method combining physical sputtering and chemical reaction [119]. Its principle is that when a high-frequency voltage (13.56 MHz) is applied between the two electrodes, the etching gas will generate a glow discharge to form a thick plasma layer in which the active free radicals react chemically with the material surface to produce volatile substances. The positive ions vertically impinge the sample surface at a high speed, thus achieving physical etching while increasing the chemical reaction rate (Fig. 7A) [120, 121]. Schematic diagram for preparation methods of Ti-based bactericidal nanostructures. A RIE device [122]. B II device [123]. C TO Device [122]. D two GLAD preparation processes [27], Copyright © 2014 IOP Publishing. E AO reaction device [124]. F HP device [66], Copyright © 2021 Elsevier Compared with a mask, the nanostructural characteristics prepared using RIE without a mask are more disordered [28]. In order to obtain an ordered nanopattern, a mask on the sample surface generally needs to prepare by using electron beam lithography and nanoimprint technology before conducting RIE [125, 126]. While, the mask preparation increases the preliminary work and requires additional equipment, complicating the process. Hasan et al. [28] and Ganjian et al. [127] reported the maskless RIE process-prepared the randomly distributed Ti nanostructures with a high aspect ratio. Although this nanostructured surface has been proven to be effective in killing Gram-negative and Gram-positive bacteria [8, 28, 29], only the top surface of substrate can be etched due to the unidirectional etching, meaning that it is difficult for RIE technology to produce pre-designed nanopatterns on medical implants with complex geometries. Ion irradiation Ion Irradiation (II) technology involves the interaction between ions (e.g. helium ions and argon ions) with a certain amount of energy and the atoms in the target material, and the ions can not only remain inside the material as foreign atoms, but also cause certain damage and defects in the material [128]. The schematic diagram of II device is shown in Fig. 7B. During the irradiation process, the helium (He), in the form of single atom, diffuses in the bulk material, and may combine with vacancies to form He bubbles [123]. As the number of the ones in the material increases, dislocation loops will be released, leading to the formation of voids on the material surface, and the He bubbles rupture may induce the generation of nanostructures. Nanocone structures can be constructed on the surfaces of various materials including tungsten, Ti, Si, and Fe, etc. using II technology [43, 123, 129–135]. The research on Ti materials primarily focuses on the preparation and formation mechanism of nanostructures after ion irradiation [123, 129, 135], and only Mulyukov et al. [43] analyzed the relevant bactericidal performance. Furthermore, the geometry of metal surface nanostructures can be controlled by changing the incident ion energy and surface temperature [129], which is a simple and low-cost physical method. It should be mentioned that like the RIE technology, the etching process of II technology is unidirectional, with only the top planar surface of the substrate being etched, making it difficult to generate pre-designed nanopatterns on material surfaces with complex geometries. Thermal oxidation Thermal oxidation (TO) is a simple and low-cost nanofabrication technique that can prepare nanostructured surfaces on Ti and its alloys. As shown in Fig. 7C, TO is typically conducted in a tubular furnace at high temperatures (400–900 ℃) in a carbon-containing atmosphere, and the growth of nanostructures depends on the oxidation control of the substrate [66, 122]. The parameters for regulating nanostructures include substrate type (pure Ti or Ti alloy), argon/acetone flow rates, holding temperature and time, as well as post-treatment [122]. The TO-synthesized TiO2 nanostructures surface exhibits super-hydrophobicity due to the presence of carbon, which consequently inhibits the adhesion of mammalian cells and reduces osseointegration performance [66]. In order to solve this problem, the post-processing method of annealing is usually employed to remove carbon, leaving only a TiO2 nanostructure on the surface that can enhance osseointegration [122]. Although the Ti surface treated by TO can exhibit bactericidal performance to some extent [44, 136], its effect is not as significant as that of the nanostructured surface prepared using the RIE. For instance, Sjöström et al. [44] prepared nanospikes with a diameter of approximately 20 nm on the Ti6Al4V surface by controlling the acetone vapor concentration in a tube furnace. This nanospike structure only reduced the viability of E. coli cells by 40%. In contrast, Hasan et al. [28] achieved a ~ 95% bactericidal efficiency against E. coli on a RIE-constructed nano-pillar structure Ti surface. Therefore, further research is needed on the use of TO technology for constructing nanostructured surfaces with high bactericidal efficiency. Glancing angle sputter deposition Glancing angle sputter deposition (GLAD) is a complex physical vapor deposition (PVD) method proposed by Robbie et al. in the mid-1990s [137]. Its principle refers to that the angular vapor deposition on a substrate is performed utilizing magnetron sputtering equipment, thereby obtaining ordered and periodically arranged nanostructures on the substrate surface [27]. The so-formed surface nanostructures are the result of the spatial interaction of two inputs, i.e., atomic self-shadowing effect (Fig. 7D) and the collisions of sputtered atoms in the plasma, and their morphologies can be regulated by adjusting the tilt angle of the substrate and the argon gas pressure [138]. If the deposited atoms arrive at the substrate surface with a highly inclined deposition angle θ (generally ≥ 80°), it will result in the deposition of a columnar thin film. When the substrate keeps stationary, columnar structures tilted towards the particle source direction can be observed, whereas on the rotating substrate, these tilted columnar structures can be carved into various structures such as vertical columns, helices, or serrations [137, 139]. Sengstock et al. [27] utilized the AJA ATC-2200 V sputtering deposition system to fabricate columnar Ti nanostructures and a dense film on the reference sample surface, employing a glancing-angle deposition source with a fixed inclination angle of β ≈ 88° and a confocal deposition source with a relative inclination angle α ≈ 27° to the substrate normal, respectively (Fig. 7D). This surface only exhibits 60% bactericidal efficiency against Gram-negative E. coli, and has a relatively minor effect on the viability of Gram-positive S. aureus cells. The García-Martín team [138, 140, 141] used a circular Ti target with a diameter of 5 cm to fabricate high-density Ti nanopillar structures on the surface of medical Ti6Al4V alloy of 12 mm in diameter, in which the alloy is positioned 22 cm away from the target and tilted 80° relative to its normal. Subsequently, the team proposed a novel magnetron sputtering method capable of simultaneously depositing Ti nanopillars on both sides of large-area, up to 15 cm², implants, utilizing a larger rectangular Ti target with a size of 20 × 7.5 cm². The obtained surfaces can reduce the adhesion amount of E. coli and S. aureus cells, thereby inhibiting the formation of biofilms. Although existing literatures have reported that nanostructured Ti surfaces with bactericidal and antibiofilm performance can be fabricated using the GLAD technique, these surfaces do not exhibit a very high bactericidal efficiency [66]. Moreover, the GLAD technique requires highly specialized equipment and precise operations, making it unsuitable for large-scale industrial production. Anodic oxidation Anodic oxidation (AO), also known as electrochemical anodization, is an electrochemical surface treatment process in which a metal substrate is immersed as an anode in an electrolyte, where a controlled electric current promotes the formation of an oxide layer on its surface [52, 124]. The Ti-based samples are placed in an electrochemical cell (Fig. 7E), and the TiO₂ nanotube arrays can be directly grown on a Ti substrate after applying a specific voltage. By adjusting the anodization parameters, such as voltage, electrolyte composition, and processing time, a variety of nanostructured morphologies, including nanotubes, nanopores, and even biomimetic nanopillar-like textures (e.g., spinules, daggers, spikes), can be prepared on their surface [54]. This one-step, scalable approach offers significant advantages for biomedical implant modification, as it enables the production of nanostructured surfaces with enhanced antibacterial and biocompatibility performance without the need for expensive equipment or complex multi-step procedures. There are some studies have reported that the nanostructures prepared by this method exhibit antibacterial performance. Rajeswari et al. [142] reported that the bactericidal performance of TiO₂ nanotubes can be improved by increasing their diameter from 40 nm to 80 nm. The similar results are also reported by Šístková et al. [52] i.e., the nanotubes with a diameter of 118 nm have better bactericidal effects against P. aeruginosa compared with the nanotubes of 73 nm in diameter, while the latter shows a better anti-adhesion against S. aureus. It should be mentioned that, although the above results indicate that the physical dimension of nanotubes can affect their antibacterial effect, their mechanical bactericidal performance is very limited only based on their own morphological characteristics. It is often necessary to combine antibacterial elements such as Ag, Zn, Cu, and Se to achieve efficient antibacterial performance [53, 143–145]. In addition, fluoride ions could be incorporated into the anodic nanostructures during the anodization process, and the presence of fluoride on anodized nanotextured surfaces is potential contribution to antibacterial activity [54]. The rutile TiO2 and surface hydrophilicity generated after heat treatment will also enhance the surface antibacterial effect [52]. A few studies have shown that the rutile TiO2 has photocatalytic activity under UV illumination, which not only exhibits potential bacterial effects, but also has great potential for application in killing RNA viruses and cancer cells [146, 147]. Hydrothermal process Hydrothermal process (HP), also known as hydrothermal synthesis or hydrothermal etching, generally refers to a heterogeneous reaction that occurs in an aqueous solvent or mineralizer under high-temperature and -pressure conditions. The definition of HP is further described by Byrappa and Yoshimura, in 2001, i.e., any heterogeneous chemical reaction occurring in a closed solvent system including aqueous or non-aqueous solution under temperatures above room temperature and pressures greater than 1 atm [148]. HP has many advantages of simplicity, low cost, and environmental friendliness, making it a commonly used technique for preparing nanostructures on Ti surfaces [66, 122]. The Ti-based samples are placed in a container filled with an alkaline aqueous solution (Fig. 7F) and after treated in high temperature, various nanostructures with different morphologies, such as nanowires, nanopillars, and nanosheets, can be formed on their surface [55, 66, 68]. The shape and size of Ti-based nanostructures can be regulated by adjusting the solution concentration, solute composition (NaOH or KOH), operating temperature, and reaction time [22, 24–26, 55]. Jaggessar et al. [22, 24] synthesized various TiO2 nanostructures on the Ti surface by changing the NaOH concentration, reaction time, and reaction temperature, and investigated the effects of hydrothermal reaction conditions on the geometric characteristics and mechanical properties of the nanostructures. The results indicate that large-size nanowire network arrays can be generated under a higher NaOH concentration, and their surfaces are of lower elastic modulus and hardness, while relatively small-size arrays can be formed at a shorter reaction time and lower temperature. Bright et al. [25] found that after reacting in NaOH for 4 h, a dense and relatively ordered surface nanospike structure can be generated, with a density of 75 spikes/µm², while after reacting for 5 h, the surface morphology is relatively sparse and disordered, with only 8 spikes/µm². Compared with GLAD and TO, the HP-prepared nanostructure Ti surface exhibits more significant bactericidal performance [23, 26, 27, 44]. Kapat et al. [23] constructed linear nanostructures on the surface of Ti6Al4V using HP, and their bactericidal efficiency against Gram-negative E. coli and Gram-positive S. aureus are approximately 93.3% and 85.6%, respectively. The same preparation method is also adopted by the team of Ivanova [26], and they synthesized flake-like nanostructures on the Ti surface, exhibiting high bactericidal efficiencies of up to approximately 99% against Gram-negative P. aeruginosa and approximately 90% against Gram-positive S. aureus, respectively. Compared with RIE technology, HP can synthesize nanostructures not only on 2D planes but also on different types of 3D surfaces, such as meshes, wires, and microbeads, making it an ideal method for orthopedic implants [122, 149]. In addition, a combination of HP and alkaline treatments to enhance bactericidal efficacy of Ti-based surfaces has also been reported [150]. In sum, HP is the preferred technology for constructing highly efficient bactericidal nanostructures on medical-grade Ti surfaces. Note that, in current research, its reaction temperature is generally above 120 °C (Table 3), indicating that the reaction occurs under high-temperature and -pressure conditions [22, 55], making it highly dependent on production equipment, and posing safety risks as evidenced by several recent reactor explosion accidents [151, 152]. Although Zhao et al. [30] and Morel et al. [31] prepared Ti-based nanowire/sheet clusters and one-dimensional nanostructures at 80 ℃ and 75 ℃, respectively, the low-temperature condition-prepared nanostructures exhibit poor bactericidal performance against Gram-positive bacteria. In addition, it is difficult to regulate the geometric dimensions of the nanostructures using HP method. Therefore, it is necessary to continue studying the HP technology for constructing Ti-based bactericidal nanostructures under relatively low temperature (< 100 °C) and normal pressure conditions, as well as developing controllable preparation methods for Ti-based bactericidal nanostructures. Table 3Summaries of bactericidal nanostructures on the Ti surfacesFabrication methodTypes of nanostructureBactericidal efficiency towards E. coli & P. aeruginosaBactericidal efficiency towards S. aureusReaction temperatureRef.HPNanowire arrays80%0%240 ℃2014 [18]GLADNanocolumnararrays40%0%-2014 [27]HPNanowire arrays47.1%19.8%121 ℃2015 [19]TONanospikes40%--2016 [44]HPNanowires58%-240 ℃2016 [20]RIENanopillars98%76%-2017 [28]HPPocket-typenanoarrays-47%240 ℃2018 [21]HPNanowire arrays-54%120–240 ℃2018 [22]HPWire-like nanostructures93.3%85.6%200 ℃2018 [23]RIEHierarchicaltopography87%76%-2019 [29]HPNanowire arrays< 50%46%120–240 ℃2019 [24]HPNanosheets99%90%150 ℃2020 [26]HPNanowire/sheet clusters98.6%69.8%80 ℃2021 [30]HPNanospikes96.5%61.5%150 ℃2022 [25]RIENanotextures50%--2022 [112]HPOne-dimensional nanostructure34.4%-75 ℃2023 [31]IINanospikes100%--2024 [43]AONanotube< 80%< 80%-2024 [52]HPNanopillars-63%150 ℃2025 [45]Note: The bactericidal efficiency against E. coli and/or P. aeruginosa listed in the table is all the maximum values reported in the literatures Summaries of bactericidal nanostructures on the Ti surfaces Nanocolumnar arrays Pocket-type nanoarrays Hierarchical topography Note: The bactericidal efficiency against E. coli and/or P. aeruginosa listed in the table is all the maximum values reported in the literatures Mechano-bactericidal mechanism of surface nanostructures In order to design more efficient physical bactericidal surfaces, it is necessary to further study the bactericidal mechanism of nanostructures [56, 59–61, 63, 65, 153–155]. Several literatures reported that the adhesive power between nanostructures and bacteria is the driving force [64, 155–157], propelling the deformation and rupture of bacterial cells, which causes the release of cytoplasmic components, thereby leading to the bacterial death [158]. There are two main rupture mechanisms for bacterial cells on the nanostructured surface (Fig. 8A and B): One is that the adhered bacterial wall/membrane is ruptured by tension on the nanostructured surface [61, 156, 159, 160]. The other is that the bacterial cells are punctured by sharp nanostructures [26, 59, 159, 161]. Bactericidal mechanisms of nanostructures according to experimental investigation. Schematic diagrams of rupture of bacterial cell wall/membrane (A) and puncturing of bacterial cells (B) on the nanostructures, respectively; (C) SEM images of state evolution for E. coli on black Si-based nanostructure arrays [159], Copyright © 2018 Royal Society of Chemistry. D FIB-SEM side profile images of P. aeruginosa and S. aureus cells adhered to Si-based nanopillar arrays with different heights [160], (E) SEM image of Ti-based nanosheet structures inserted into the interior of S. aureus cells [26], Copyright © 2018 Royal Society of Chemistry. F cross-sectional SEM image of S. aureus cells adhered to Si-based nanopillar arrays [161]. G SEM images of Gram-negative P. aeruginosa being penetrated or undergoing perturbation by nanopillars on the surface of cicada wings [59], Copyright © 2020 Elsevier. H High-magnification SEM images of E. coli and S. aureus cells adhered to Ti-based nanowire/sheet clusters [30], Copyright © 2021 American Chemical Society Experimental studies In experimental studies, the interaction between bacterial cells and nanostructures is usually based on SEM observation, and the bactericidal mechanism is inferred by the rupture state of bacterial cells attached to the nanostructures. Michalska et al. [159] suggested that longer and sharper nanopillars can directly pierce all types of bacterial cells, appearing non-selective to bacterial species, while shorter and blunter pillars require multi-faceted cell interactions to stretch and tear bacterial cell membranes (Fig. 8C). Ivanova et al. [160] proposed that mechanical lysis of bacterial cells may be affected by the extent of elasticity and aggregation of highly ordered silicon nanopillar arrays (Fig. 8D). Meanwhile, Wandiyanto et al. [26] believed that surface mechanical sterilization may be attributed to the direct physical contact between the sharp nanoedges of the TiO2 structure and the bacterial cell membrane, because the SEM image shows that the Ti-based nanosheet structure is directly inserted into the interior of S. aureus (shown by the red arrow in Fig. 8E), destroying the integrity of the cell. Modaresifar et al. [161] reported the killing mechanism of nanopillars directly penetrating the bacterial cell wall based on SEM cross-sectional images of S. aureus cells attached to Si-based nanopillar arrays (Fig. 8F). Subsequently, Velic et al. [59] observed that the bacterial envelope on the nanopillars of cicada wing surfaces was destroyed, while the envelope area suspended between the nanopillars seemingly remained intact, and the obvious penetration seemed to occur shortly after the bacterial envelope sank to the top of the pillars (Fig. 8G). The topological morphology of linear and blade-like nanostructures on three Ti-based surfaces was prepared by Zhao et al. [30], and they also observed the phenomenon of bacterial cells being penetrated and punctured by sharp nanotips or nanoedges (shown by the red arrow in Fig. 8H). Although the bactericidal mechanisms of nanostructure arrays have been reported by numerous experimental studies, the precise rupture mechanism of bacterial cells is still unclear. Theoretical calculation It is widely agreed that the bacterial cells adhered to the surface nanostructures are lysed through their mechanical action in essence [6, 8, 9]. Therefore, clarifying the mechanical response of bacterial cells on nanostructures is the key to understanding the mechano-bactericidal mechanism. Considering the difficulties in directly characterizing the mechanical response of bacterial cells to nanostructures by experimental method and the inability to independently regulating the geometric parameters of nanostructures, the theoretical calculation method is commonly employed to study the interactions between nanostructures and bacterial cells. Currently, there are some inconsistent results regarding the mechanical response of bacterial cells on nanostructured surfaces [56, 58–63, 154, 162], mainly including the critical action sites (weak points) of bacterial cells and the relationship between the geometric parameters of nanostructures and the forces and deformations experienced by bacterial cells. Ivanova et al. [61] developed a biophysical model of the interaction between bacterial cells and surface structure of cicada wing, and found that the cell membrane suspended in the area between two pillars is stretched (Fig. 9A). Specifically, the bacterial cell membrane in this model is divided into regions A and B, in which the former refers to the contact area with the nanopillars and the latter refers to the area suspended between the nanopillars. Due to the nanopillars adsorption, the surface area SA of region A increases, causing the stretching of elastic membrane in region B. When the stretching degree of region B exceeds the elastic limit of cells, it can lead to mechanical rupture of ones, indicating that the weak point of bacterial cells is located in the suspended area between the nanopillars. Bactericidal mechanisms of nanostructures according to biophysical and thermodynamic models. A Biophysical model of the interaction between nanopillars on the cicada wing surface and bacterial cells [61], Copyright © 2013 Elsevier. B Investigation of the bactericidal mechanism on the nanoarray surface by analyzing the total free energy change of bacterial cells [62], Copyright © 2016 Royal Society of Chemistry. C Effects of four types of nanopillar arrays on the deformation of bacterial cells [56]. D Schematic diagrams of bacillus cells adhering to a smooth surface and a nanopillar surface [63], Copyright © 2019 American Chemical Society In addition to the biophysical model, a quantitative thermodynamic model is proposed by Li et al. [62] to investigate the bactericidal mechanism on the surface of nanoarrays based on the total free energy change of bacterial cells (Fig. 9B). The results show that the bactericidal performance of cicada wing-like nanoarray surface is contributed to a substantial increase in the contact adhesion area of the cell envelope induced by this surface. They also found that appropriately increasing surface roughness, i.e. increasing the surface distribution density, radius, and height of nanopillars, can increase the stretching degree of the cell membrane, thereby improving the bactericidal performance. Wu et al. [56] utilized the biophysical model established by Ivanova et al. [61] to systematically investigate the effects of four types of nanopillar surfaces on the deformation degree of bacterial cells (Fig. 9C). They found that at an appropriate pillar density, a significant stretching occurs for the cell membrane covered the pillar tops and some of the vertical pillar walls. In addition, the heterogeneity in nanopillar height can further enhance the stretching degree of bacterial cells. In 2019, Liu et al. [63] given that the interfacial energy gradient between nanoarray surface and adhered bacteria serves as the driving force for bacterial cells to move into the interior of nanostructures. They also found that the smaller the nanopillars diameter and the larger the spacing, the greater the pressure exerted on the cells. Only enough height of nanopillars can ensure significant creep deformation of cells, ultimately leading to their rupture (Fig. 9D). In addition, this study also revealed that there is a significant relationship between pressure exerted by nanostructures and cell size, and the smaller the cell volumes, the greater the pressure on cells from nanostructures. The volume of bacterial cells is approximately 10− 18 m³, and the corresponding pressure exerted on cells can reach MPa order, whereas the volume of mammalian cells is approximately 10− 15 m³ (with a diameter of about 10 μm), and the corresponding pressure is near to zero. It should be mentioned that, the above biophysical and thermodynamic models cannot intuitively observe the distribution of cellular stress and deformation, whereas the finite element (FE) models can track the deformation process of cells on nanostructured surfaces and quantify their stress/strain [155]. Mirzaali et al. [58] used FE method (ABAQUS) to investigate the response of S. aureus cells on different nanoarrays and quantitatively characterize the effect of nanopillar geometric parameters on the von Mises strain (Fig. 10A) and sinking depth of bacterial cells. The results show that the critical action site of bacterial cells is located away from the contact region with the nanopillars, and the height of nanopillars do not alter the maximum equivalent strain of bacterial cell wall, indicating that the bactericidal behavior of the nanopillars is not affected by their height. As the width and spacing of nanopillars reduces and increases, respectively, the sinking depth and maximum strain of bacterial cells gradually increase, suggesting that nanopillar surfaces with smaller width and larger spacing exhibit better bactericidal performance. Yarlagadda et al. [60] established a 2D FE model for the interaction between Bacillus subtilis cells and nanoarrays using ABAQUS/Standard software, and studied the effect of the nanopillars geometric parameters (diameter, height, and spacing), as well as the surface/cell interaction force on the stress/strain of the cells. The results indicate that the local deformation stress generated within the bacterial cell wall (peptidoglycan) is sufficient to cause cell rupture and subsequent death. The strategy for enhancing surface bactericidal efficiency is to reduce the diameter of nanopillars and increase the interactive attractive force. Subsequently, they also established a 3D FE model of the interaction between Gram-negative bacteria and nanostructured surface, not only analyzing the mechanical mechanism of adhesion-driven envelope deformation, but also investigating the effect of different geometric parameters on cell deformation by varying the radius, spacing, and height of the nanopillars (Fig. 10B) [59]. They found that the nanoarray surface sterilization is mainly based on a critical action site at the pillar tips, which induces large in-plane strain and local rupture of cells. In addition, a reduction in the radius and spacing of the pillars on the nanoarray increases the deformation of bacterial cells. Alameda et al. [162] and Valiei et al. [154] adopted a 2D FE model and found that the maximum stress occurs at the site where the cell wall is significantly bent, namely, near the bacteria/nanocone separation edge, rather than at the nanocone tips (Fig. 10C). A 3D FE model of Gram-positive bacteria, including an elastic bacterial cell wall and a nonlinear viscoelastic cytoplasm, is established by Zheng et al. [163] to investigate the effect of contact friction and nanopillar geometric parameters on cellular stress/strain. The greatest cellular deformation and stress concentration near the cell-nanopillar contact area is observed, and the smaller nanopillar radii and larger spacings can induce the higher maximum stress and strain values on the cells. Bactericidal mechanisms of nanostructures according to FE calculations. A Mechanical responses of S. aureus cells to different nanoarrays [58] Copyright © 2018 IOP Publishing. B Strain distribution of bacterial cells adhered to nanopillars [59], Copyright © 2020 Elsevier. C 2D FE model of bacterial cell on nanopillars [154], Copyright © 2020 American Chemical Society. D 3D turgor pressure FE model of E. coli cells and the influence of nanopillar geometric parameters [155], Copyright © 2021 Elsevier. E Spatial and temporal evolution of stress and strain within the cell membrane of E. coli [164]. F Critical action site and stress/strain analysis of E. coli cells on hierarchical nanostructure arrays with different height differences [165], Copyright © 2023 American Chemical Society The FE models used in the above studies do not consider the presence of intracellular turgor pressure, which is difficult to accurately reflect the true mechanical responses of cells adhered to nanostructure arrays. Therefore, a 3D turgor pressure FE model of E. coli cells is developed by Cui et al. [155] to investigate the critical action site of bacterial cells as well as the influence of nanopillar geometric parameters. This model calculates the elastic and creep deformation of cells during cell-nanostructure interactions, and its reliability has been validated through AFM indentation experiment (Fig. 10D) [166]. Subsequently, the same model is adopted by Islam et al. [164] to further analyze the evolution of stress and strain within the E. coli cell membrane, observing that they are spatially linear but temporally nonlinear. They thought that the bactericidal mechanism of nanostructures is similar to that of a hole-punching machine (Fig. 10E). Currently, the aforementioned theoretical researches on the bactericidal mechanisms of nanostructures primarily focuses on flat nanostructure arrays [59, 61, 65, 154, 155, 162], that is, the height for all nanostructures is essentially uniform, which are mostly fabricated using nanoimprint lithography and plasma etching techniques [28, 33, 36, 167]. Note that surfaces with inconsistent nanostructure height, termed as hierarchical nanostructure arrays, also exhibit excellent bactericidal performance [30, 55, 56, 66, 157, 168]. Nevertheless, only a few studies have reported on the mechanical bactericidal mechanisms of hierarchical nanostructure arrays [56, 157]. A biophysical model is established by Wu et al. [56], and this model demonstrates that the cell membrane suspended between nanopillars with different heights is further stretched, but it cannot directly observe the sites of the maximum stress and strain of cells. Xie et al. [157], based on FE, found that the small tip width and uneven morphological characteristics of nanorods significantly enhanced the stress exerted by the nanostructures on the bacterial cell wall, leading to rapid bacterial inactivation. Zhao et al. [165] studied the mechanical response of bacterial cells on flat and hierarchical nanostructure arrays using a 3D turgor pressure FE model, clarifying that the key action site of bacterial cells is at the three-phase junction of cell-liquid phase-nanostructure. It is found that the cell rupture speed can be accelerated by reducing the diameter of nanopillars, the thickness/width of nanosheets, or appropriately increasing the spacing between nanostructures, as well as increasing the height difference of nanostructures (Fig. 10F), which provides a clear direction for the design of surface-efficient bactericidal nanostructures. It should be pointed out that: (i) The relationship between the mechanical properties of nanostructures and the stress and strain of cells has not been built in current studied. (ii) Theoretical calculations can be carried out using molecular dynamics simulations, which is beneficial to reveal the molecular-level mechanical bactericidal mechanism of nanostructures. (iii) The mechanical property parameters of various bacterial cells require further measurement and corroboration via experimental results, in order to establish more realistic bacterial cell models. In summary, further investigation is needed on the interaction between bacterial cells and different nanostructure arrays, as well as their corresponding mechanical responses. Note that, when clarifying the mechano-bactericidal mechanism, sometimes the effect of external forces is also a key consideration factor, especially those caused by the motion of gas-liquid interface (ALI) and capillary action. Valiei et al. [154] demonstrated that hydrophilic silicon and zinc oxide nanopillar surfaces do not exhibit bactericidal performance against P. aeruginosa for up to 24 h under fully submerged conditions, with over 98% of attached bacteria remaining viable. However, after the evaporation of bacterial droplets, mobile ALI is generated, and rapid and extensive cell death occurs within seconds (> 99.9% on NanoSi, > 90% on NanoZnO), but it has no effect on the flat controls. Computational modeling shows that the ALI motion-produced capillary normal force ranges from 150 to 250 nN per bacterium, which is sufficient to exceed the cell envelope’s ultimate tensile strength (estimated at ~ 1.1 nN/pillar for NanoSi and ~ 6.7 nN/pillar for NanoZnO). In addition, the critical role of surface tension as one of the external forces has also been confirmed by the significant reduction in killing effect after the addition of surfactants. Considering that many experimental schemes, including droplet drying, rinsing, and electron microscopy sample preparation, may inadvertently introduce ALI or capillary force, which may affect the observed bactericidal efficacy in published studies. But these early works indicate that the mechano-bactericidal performance depends on specific physical conditions. Although Valiei et al. [154] only studied Si and ZnO, the relevant results can be used to guide the construction of Ti-based nanostructure arrays, that is, Ti surfaces can be designed with different wettability and surface chemical properties, which may regulate the magnitude of capillary forces. Overall, these observations emphasize the necessity of strictly reporting experimental conditions, particularly immersion state and ALI exposure, in future research on mechano-bactericidal surfaces, as recommended by the minimum reporting framework presented in Sect. 6. Geometries effects The mechano-bactericidal performance of nanostructured arrays is not determined by the type of material or preparation technology itself, but is related to the geometric parameters of the nanostructures [33, 41, 56, 57]. Clarifying the influence of these parameters on bactericidal efficiency is crucial for designing high-performance bactericidal surfaces. The relevant results of experimental and computational studies are summarized in Table 4. Table 4Summary of experimental and computational investigations into the effects of geometric parameters on bactericidal performanceArticle typeShapeMaterials/ModelGeometries range/Relation between geometries and high bactericidal efficiency (nm)DiameterThicknessSpacingHeightAspect ratioExperimentnanopillar [34]Polymer70–215-100–380/Optimal=130–380210–300-nanopillar [186]Polymer10–210/smaller-360300–620-nanopillar [159]Silicon50–490/ smaller-130–600/larger150–400/Larger-nanopillar [187]Polymer270, 380-smaller214,803-nanopillar [33]Polymer25–60/smaller-100–300/Optimal = 170150–400/Critical=200-nanopillar [188]Black silicon80–460-300–1400/smaller1000–6000-nanopillar [170]Titanium35–7040–752400–395034.28–120/smallernanotube [169]Carbon50–100-50–1001000–32,000100–3000/largernanopillar [36]Polymer41–186/smaller-170–500145–285-nanosheet [189]Graphene-Thinner---nanosheet [26]Titanium-Optimal = 10---Nanowire/sheet clusters [30]Titanium23, 24, 69/ smaller8, 13/Thinner---Calculationnanopillar [60]FE20–80/smaller200–500/no significant effect200–500/no significant effect-nanopillar [62]thermodynamic model10–50/larger-91–316/smaller30–700/larger-nanopillar [63]thermodynamic mode30–90/smaller-170–250/largercertain height-nanopillar [155]FE20–200/smaller-100–300/Optimal = 200200-nanopillar [163]FE10–90/smaller-120–300/larger--nanopillar [190]FE10–120/smaller-25–160/larger--nanopillar [59]FE10–90/smaller-120–240/smaller--nanopillar/nanosheet [165]FE10–150/smaller8–75/smaller120–300/larger200Experiment+ Calculationnanopillar [191]Polymer/FE57–184/smaller-121–468/larger126–663-nanopillar [56]Polymer/Biophysical model--130–300/ Optimal = 170150–400-Note: Relation between geometries and high bactericidal efficiency refers to how to regulate the nanostructure geometries to obtain the high surface bactericidal efficiency, e.g., Diameter: smaller represents that the smaller the diameter, the higher the surface bactericidal efficiency Summary of experimental and computational investigations into the effects of geometric parameters on bactericidal performance 100–380/ Optimal= 130–380 10–210/ smaller 130–600/ larger 150–400/ Larger 25–60/ smaller 100–300/ Optimal = 170 150–400/ Critical= 200 300–1400/ smaller 34.28–120/ smaller 100–3000/ larger Nanowire/ sheet clusters [30] 8, 13/ Thinner 200–500/ no significant effect 200–500/ no significant effect 91–316/ smaller 30–700/ larger 100–300/ Optimal = 200 Polymer/ FE Polymer/ Biophysical model Note: Relation between geometries and high bactericidal efficiency refers to how to regulate the nanostructure geometries to obtain the high surface bactericidal efficiency, e.g., Diameter: smaller represents that the smaller the diameter, the higher the surface bactericidal efficiency Diameter is a key geometric parameter that affects bactericidal performance. Although it is difficult to control the independent parameter changes of the nanostructures prepared in the experiment, it is generally confirmed that reducing the tip diameter can significantly improve the bactericidal efficiency. For instance, polymer nanocolumn arrays (diameter 10–210 nm) exhibit higher bactericidal activity at smaller diameters [33]. The similar results can also be found in these studies of the silicon-based nanowires (diameter 50–490 nm) [159] and Ti-based nanowires/clusters (diameter 23–69 nm) [30]. Yet, the inconsistent results are reported by the thermodynamic model, namely, a larger diameter (10–50 nm) is more optimal [62]. Even so, most finite element simulations (Table 4) still support that the smaller the tip diameter, the more significant the von Mises stress, which can far exceed the rupture threshold of the outer membrane of Gram-negative bacteria. It should be noted that, the nanostructure should not be too sharp, as capillary forces can easily cause it to buckle, reducing its penetration ability [169]. The spacing between nanostructures is also a key geometric parameter that affects bactericidal performance, but there are inconsistent results on its impact on bactericidal efficiency (Table 4). The essential function of spacing is to regulate the density of contact sites between nanostructures and bacterial membranes, thereby determining the synergistic or competitive relationship between interfacial adhesion and local mechanical loads. Specifically, increasing the number of adhesive sites per unit area with small spacing enhances the interfacial bonding force [60]; but overly dense arrangement will shorten the suspended span between columns, limit the sinking depth of bacterial membranes in gaps, weaken local stress concentration, and instead be detrimental to rupture [155]. Although large spacing reduces the total number of adhesion sites and interfacial adhesion force, it increases the suspended span between columns. As a result, a single column can significantly increase local deformation of the cell membrane, resulting in high membrane strain [165]. In addition, the divergence of the influence of spacing in different models may stem from differences in model assumptions, such as the presence or absence of cell swelling pressure, material properties, adhesion strength, and consideration of creep. Both experimental and computational results indicate that the critical value of the height of the nanopillar is approximately 200 nm [155]. When it is less than this value, the higher the height, the better the bactericidal performance. When it exceeds this value, the height does not significantly improve the bactericidal efficiency [60]. This is because when the height is low (< 200 nm), the bacterial membrane has not fully sunk to touch the substrate, resulting in limited bactericidal effect [63]; as the height increases above the critical value, the membrane has gained sufficient sinking space, and the pillar tip can fully penetrate or stretch the cell membrane. At this point, further increasing the height will no longer significantly improve the bactericidal effect. It should be pointed out that an appropriate height difference can further enhance membrane stretching, thereby accelerating cell rupture [56, 165]. In addition, the aspect ratio can also affect the bactericidal efficiency of nanostructures. Rigid materials (Ti nanopillar) are mainly designed to resist bending and puncture, and the aspect ratio should not be too high [170]. Flexible materials (carbon nanotubes) are mainly designed for energy storage and rebound, and a high aspect ratio is more advantageous [169]. Standardized evaluation A further challenge is the lack of standardized protocols for mechano-bactericidal surfaces with nanostructure arrays. Although ISO 22,196 [171] and ISO 4768:2023 [172] are valuable for their intended purposes, they are not designed for mechano-bactericidal surfaces and do not reflect implant-relevant conditions. In order to improve the study comparability, it is necessary to propose a minimum reporting framework. This framework recommends a systematic reporting of: (i)Substrate characteristics, including Ti grade, fabrication parameters, nanostructure geometry, roughness, chemistry, wettability and crystallinity; Substrate characteristics, including Ti grade, fabrication parameters, nanostructure geometry, roughness, chemistry, wettability and crystallinity; (ii)Bacterial strain details, including species, strain number, and inoculum density; Bacterial strain details, including species, strain number, and inoculum density; (iii)Testing conditions, including incubation time, static vs. dynamic, and quantification methods such as CFU counting and Live/Dead staining; Testing conditions, including incubation time, static vs. dynamic, and quantification methods such as CFU counting and Live/Dead staining; (iv)Controls and statistical validation; Controls and statistical validation; (v)Crucially, whether the sample remains submerged or exposed to ALI. Crucially, whether the sample remains submerged or exposed to ALI. Although not a normative standard, these guidelines aim to improve transparency and reproducibility, and may provide information for future specialized standardization efforts. Clinical translation and challenges At present, the understanding of Ti-based mechano-bactericidal nanostructures is mainly based on laboratory-scale in vitro studies. However, after implantation into the human body, the physiological environment faced by the nanostructures is significantly different from the conditions in vitro, which may affect their bactericidal performance. Protein adsorption is the first step that occurs when biological materials are implanted into the body. It forms a complex protein layer within seconds, thereby altering the superficial biological properties and mediating subsequent biological responses [173–175]. This protein layer may reduce the sharp features of the nanostructure, which in turn reduces direct bacteria-nanostructure contact, and consequently affects the mechano-bactericidal efficacy. Contrary to this concern, recent studies have shown that the interstitial spaces of Ti-based nanostructures remain unfilled following adsorption of human serum proteins or fibronectin, with surfaces maintaining high bactericidal efficiency [176, 177]. In addition, Ivanova et al. [178] reported that macrophages exhibit wound healing behavior on the protein-modified Ti-based nanosheets infected with S. aureus. Visalakshan et al. [179] further demonstrated that nanostructured Ti surfaces can selectively adsorb more adhesion proteins and fewer complement proteins compared to planar control group, thereby reducing the production of pro-inflammatory cytokine and suppressing the formation of fibrous capsule in a subcutaneous mouse model. These findings indicate that protein adsorption does not necessarily reduce the bactericidal activity of the nanostructured array; in some cases, it may even promote beneficial immune regulation. However, the interaction between the protein layer composition, nanostructure geometry, and the bactericidal efficiency remains unclear and requires further systematic research. In addition to its bactericidal function, the clinical application of Ti-based nanostructured arrays as bone implants also requires meeting multiple functional requirements such as supporting bone integration, promoting tissue healing, and regulating immune response. At present, research on these properties is mainly conducted under in vitro conditions (Table 5). In terms of osteogenic differentiation, TiO₂ nanowires have been shown to promote the osteogenic differentiation of mesenchymal stem cells without the need for external biochemical factors [20, 28]. Similarly, both nanosheets and micro/nano hierarchical topographies can also induce osteogenic differentiation of human adipose stem cells without exogenous osteogenic factors, as confirmed by enhanced ALP activity and calcium deposition [180]. TiO₂ nanofibers can increase the metabolic activity of human osteoblasts to 71% [22], and hydrophilic surfaces are more conducive to osteoblast activity [24]. In terms of tissue healing, nanowire arrays support osteoblast adhesion and proliferation by guiding cytoskeleton reorganization and three-dimensional migration [18]. And hierarchical micro/nano structures support long-term expansion of stem cells and maintain their multi-directional differentiation potential [180]. In terms of immune regulation, nanospike structures can directly upregulate the expression of macrophage TLR2/4 and phagocytosis-related receptor through ligand-independent physical contact stimulation, activating the phagocytic activity of innate immune cells [181]. Hierarchical disordered spike-like nanostructures support macrophage adhesion and proliferation while maintaining good cell compatibility [182]. Following pre-incubation with human plasma, nanostructured surfaces still maintain significant bactericidal activity, with platelet adhesion and activation maintained at extremely low levels [183]. Overall, the above in vitro studies indicate that Ti-based bactericidal nanostructured arrays exhibit positive regulatory potential in key biological processes, but these conclusions still need to be validated through in vivo studies. Table 5Summary of representative in vitro and in vivo studies on Ti-based mechano-bactericidal nanostructured surfacesNanostructure/FabricationBactericidal performanceCellular/Immune responseRef.nanowire arrays/HP> 50% against motile bacteriasupporting the viability and proliferation of MG-63 human osteosarcoma cells2014 [18]TiO2 nanowires/HP~ 30% and 58% bactericidal efficacy after 1 h and 18 h, respectivelypromoting osteogenesis2016 [20]randomly nano-structures/RIE95% of E. coli, 98% of P. aeruginosa, 92% of M. smegmatis, and 22% of S. aureuscytocompatible, supporting attachment and proliferation of human mesenchymal stem cells (hMSCs)2017 [28]nanowires/HP54% and 33% against S. aureus after 3 h and 18 h, respectivelybiocompatible and beneficial to osteoblast cell growth2018 [22]nanowires/HPreducing the colony-forming units of S. aureus by 46% over 18 hsupporting osteoblast activity without piercing or damaging human cells2019 [24]hierarchical micro/nanopillars/RIE; nanosheets/HPreducing the attachment of bacteria and mechanically rupturing bacteriasupporting human adipose-derived mesenchymal stem cells (hASC) adhesion, growth, and proliferation while preserving their viability and trilineage differentiation potential, exhibiting intrinsic osteoinductive properties, successfully inducing osteogenic differentiation2020 [180]nanospikes/HPsignificantly improving the macrophages’ ability to phagocytose fluorescent microbeadspolarizing macrophages toward an M1-like pro-inflammatory phenotype, significantly activating macrophage phagocytosis2022 [181]hierarchical disordered spike-like nanostructures/HPP. aeruginosa viability never exceeds 4% over the 21 days. S. aureus viability drops to 6.8% by day 3 and recovers to 40% by day 21.After 48 h of incubation, the macrophages display normal morphology, healthy proliferation, and a high viability of 94.0 ± 4.2%, without signs of cytotoxicity or membrane damage.2022 [182]sharp nanostructures passivated with native human plasma/HPThe viability of S. aureus and P. aeruginosa is reduced to approximately 35% and 5%, respectively.superior blood biocompatibility profile and minimal platelet adhesion. Macrophages remain predominantly in a resting (M0) state. Bone-marrow-derived mesenchymal stem cells (BMSCs) maintain steady viability and show a slight tendency toward chondrogenic differentiation.2024 [183] Summary of representative in vitro and in vivo studies on Ti-based mechano-bactericidal nanostructured surfaces Although in vivo studies remain limited, the feasibility and potential of Ti-based mechano-bactericidal nanostructures have been preliminarily demonstrated (see Table 6). Kapat et al. [23] implanted nanowire-structured porous Ti6Al4V into rabbit femoral defects, achieving a bone volume fraction of 44.0% after 12 weeks, significantly superior to untreated group (36.9%) and empty defect group (19.6%), with subcutaneous implantation showing good vascular penetration and no observable inflammation or infection. In addition, after 84 days of implantation of nanospikes Ti6Al4V into mice, the thickness of its subcutaneous fibrous capsule significantly decreases [179]. Ziegelmeyer et al. [184] directly compared micropillars and nanosheets in a rat skull defect model: in vitro, Nano Ti promotes M2 polarization of RAW 264.7 macrophages, whereas Micro Ti has a macrophage survival rate of less than 20% by day 4; in vivo, the bone volume on the Nano Ti surfaces achieves a 3.2-fold increase within 12 weeks compared to 2 weeks, while 4/10 of rats with Micro Ti shows no bone formation response. These results indicate that in vitro bactericidal activity cannot be directly extrapolated to in vivo osseointegration outcomes, and that macrophage compatibility is a critical determinant of in vivo success. Nevertheless, current in vivo research is still mainly focused on short-term (≤ 12 weeks) and small animal models. In the future, priority should be given to long-term evaluations of large animals to systematically assess their safety, durability, and antimicrobial performance. Table 6Summary of representative in vivo studies on Ti-based mechano-bactericidal nanostructured surfacesNanostructure/FabricationBactericidal performance in vitroCellular/Immune response in vitroanimal model/ implant location/ Bacterial Challenge/ follow-up periodBacterial burden/ inflammatory response/ bone–implant contact/ histology/ adverse effectsRef.flower-like, rod-like, wire-like, & saw-like nanostructures/HPWithin 4 h, the wire-like and saw-like surfaces kill ~ 93.3% and ~ 82.4% of E. coli, and ~ 85.6% and ~ 79.8% of S. aureus, respectively.highly cytocompatible, promoting human adipose-derived mesenchymal stem cells (hASC) adhesion and proliferationNew Zealand white rabbits/ Subcutaneous implantation and bilateral femoral defects/ No in vivo bacterial challenge/ 4 and 12 weeksNo inflammation or infection. Osteochondral defect healing with restoration of subchondral bone and partial hyaline cartilage formation. a dense collagen matrix deposition and calcification2018 [23]nanospikes/HP94.7% of P. aeruginosa cells and 61% of S. aureus cellsenhancing monocyte adhesion while attenuating inflammatory activation, resulting in lower production of pro-inflammatory cytokines and higher production of anti-inflammatory factors. Macrophages display an elongated morphology, with the significant downregulation of pro-inflammatory genes.8-week-old female Balb/ C mice/ Subcutaneous implantation (bilateral flank pockets)/ No in vivo bacterial challenge/ 12 weekssignificant reduction in fibrous capsule thickness, indicating an attenuated inflammatory foreign body response and improved overall biocompatibility2023 [179]sharp nanoedges array/HP80% against S. aureus after 24 hmultiple functions including procoagulant activity, anti-inflammatory immunomodulation, osteogenic induction, and pro-angiogenic effects, synergistically driving endothelial cell migration and capillary network formationAdult male Wistar Han rats/ Calvarial (skull) bone defect model (8 mm critical-size defect)/ No in vivo bacterial challenge/ 12 weekssignificantly enhancing osseointegration, with markedly higher bone volume fraction, bone mineral density, and bone-implant contact, a fibrous tissue-free interface, and newly formed mature bone covering the entire surface2025 [184] Summary of representative in vivo studies on Ti-based mechano-bactericidal nanostructured surfaces In summary, the clinical translation of Ti-based bactericidal nanostructured arrays faces several key challenges. First, geometric controllability and reproducibility are crucial for industrial-scale manufacturing. And it is still not easy to scale up the production from laboratory samples to complex three-dimensional implant geometries. Second, their long-term bactericidal performance in the physiological environment of the body requires verification. Nanostructures must withstand mechanical loads, corrosion, protein contamination, and biofilm maturation over months to years without losing sharpness or undergoing delamination. Note that, the impact of the inflammatory environment on nanostructure stability warrants particular attention. Bordbar-Khiabani and Gasik [185] reported that under simulated inflammatory conditions, the corrosion current density of patterned Ti alloys increases significantly, and the adsorption of BSA in the pitting area changes the diffusion behavior of corrosion ions, suggesting that the inflammatory environment may accelerate material degradation through modification of surface chemical states, thereby compromising the long-term service performance of nanostructures. Third, safety concerns regarding cytotoxicity and chronic inflammation require comprehensive evaluation; although in vitro studies generally indicate favorable cell compatibility, the balance between bactericidal efficiency of nanostructures and host cell viability warrants further investigation. Fourth, the existing standardized testing schemes are not designed for pristine topographical surfaces and fail to replicate physiological conditions. Future research should prioritize long-term in vivo experiments in clinically relevant animal models and establish unified evaluation criteria to comprehensively verify the safety and effectiveness of Ti-based bactericidal nanostructured arrays. Conclusions and outlooks Inspired by natural surfaces, physical bactericidal nanostructure arrays provide a novel approach for inhibiting biofilm formation on implant surfaces. Ti and its alloys are commonly used in the production of surgical implants, orthopedic devices, etc. Therefore, this review summarizes the development of natural/biomimetic bactericidal nanostructure arrays, mainly concerning preparation methods, surface morphologies, bactericidal performance, and bactericidal mechanisms of Ti-based physical bactericidal nanostructure arrays. Based on the current research results, the following five facets can be given priority attention in the future: (i) The controllable preparation technology of Ti-based nanostructure arrays and the HP method under ordinary pressure. (ii) The high-efficiency and long-term bactericidal performance of nanostructure arrays. (iii) Bactericidal mechanism of nanostructure arrays. (iv) The adsorption modes and mechanisms of proteins on nanostructure arrays. (v) Safety of nanostructure arrays in vivo. Acknowledgements The authors acknowledge financial support from the Natural Science Foundation of Shandong Province and Youth Innovation Team of Higher Education of Shandong Province. Abbreviations Reactive ion etching Ion irradiation Glancing angle sputter deposition Thermal oxidation Hydrothermal process Anodic oxidation black silicon atomic force microscopy technology Pseudomonas aeruginosa Staphylococcus aureus Bacillus subtilis Escherichia coli Gas-liquid interface Finite element model Author contributions WFL, SJZ and LDZ contributed to the design of this work. RL, FCX and XCG contributed to data collection. WFL, RL, FCX and XCG contributed to manuscript preparation and editing. SJZ and LDZ revised critically for important intellectual content. WFL, SJZ, and LDZ contributed to funding acquisition and conceptualization. All authors read and approved the final manuscript. Funding This work was supported by the National Natural Science Foundation of China (32000977), the Natural Science Foundation of Shandong Province (ZR2024QH602/ ZR2025MS482/ZR2025QC2225Z), the Youth Innovation Team of Higher Education of Shandong Province (2022KJ262/2025KJH027), and Weifang Science and Technology Development Plan Project (2026GX048). Data availability No datasets were generated or analysed during the current study. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Shujiao Zhang, Email: SJZhang_sdvcst@163.com. Lidan Zhao, Email: ldzhao@sdsmu.edu.cn. References Associated Data Data Availability Statement No datasets were generated or analysed during the current study.