Cartilage-bone inspired the construction of soft-hard composite material with excellent interfacial binding performance and low friction for artificial joints

Inspired by the cartilage-bone structure in natural joints, soft-hard integrated materials have received extensive attention, which are the most promising candidates for artificial joints due to their combination of excellent load-bearing properties and lubricating properties. The latest progress showed that the combination of hydrogel and titanium alloy can realize a bionic natural joint lubrication system on the surface of titanium alloy. However, obtaining a tough interface between the hydrogel (soft and wet) and the titanium substrate (hard and dry) is still a great challenge. Here, we designed a “soft (hydrogel)-hard (Ti6Al4V)” integrated material with outstanding combination, which simulates the structure and function of cartilage-bone in the natural joint. The load-bearing properties, binding performance, and tribological behaviors for different forms of the soft-hard integrated materials were investigated. The results showed that the hydrogel layer and Ti6Al4V substrate possess ultra-high interfacial toughness (3,900 J/m2). In addition, the combination of the hydrogel layer and Ti6Al4V substrate provided a good lubrication system to endow the “soft (hydrogel)-hard (Ti6Al4V)” integrated material with high load-bearing and excellent tribological properties. Therefore, this study provided an effective strategy for prolonging the service life of Ti6Al4V in the biomedical field.


Introduction
Articular cartilage is a soft tissue with significant functions covering the ends of the joint bones. It provides an excellent lubricating and bearing joint surface, which can transmit 7-9 times the weight of the human body and maintain an extremely low friction coefficient (0.001-0.04) [1,2]. Cartilage would be injured due to various factors such as the increase in age, congenital diseases, abrasion, or trauma.
However, articular cartilage is difficult to repair itself after damage due to its avascular and limited self-healing capacity, which may lead to the occurrence of osteoarthritis and seriously affect the quality of life of patients [3,4]. Therefore, it is necessary to explore effective methods to repair or replace the damaged articular cartilage, and then reconstruct the function of the joint.
Currently, joint replacement was considered an effective method for treating osteoarthritis. The commonly used artificial joint materials are mainly metal materials, ceramic materials, and ultra-high molecular weight polyethylene [5][6][7]. Among them, titanium alloy has been widespread concerned in artificial joint materials because of its excellent mechanical performances, high biocompatibility, fatigue resistance, and corrosion resistance [8,9]. Therefore, titanium can form long-term and effective interaction with soft tissues and bone to replace damaged joints [10][11][12]. Although titanium and its alloys have been used clinically in the middle of the last century, there are still some shortcomings in their surface characteristics, which affect their service life in many applications. For instance, titanium alloys as artificial joints can produce wear debris, release metal ions, and cause bone dissolution around the prosthesis, resulting in the failure of joint materials after long-term friction [13]. The existence of articular cartilage can keep the joints with excellent boundary lubrication under high load and lower speed bending. Thus, a bionic lubricating cartilage layer should be designed on the surface of hard artificial joint materials by surface modification technology to make it similar to natural human joints in structure and function, which is the most attractive method to enhance the bio-tribology of titanium alloy surfaces. For example, hydrophilic polymer coatings have been developed for surface modification of titanium and its alloys [14][15][16].
Hydrogels were considered to be the most promising substitute for artificial articular cartilage by their good biocompatibility, adjustable mechanical properties, excellent biotribological properties, and the similarity of the microstructure of natural cartilage [17,18]. Recently, inspired by the structure and functions of the cartilage, hydrogels have received widespread attention as a coating for hard artificial joint materials, which can dramatically moderate the friction effect of hard materials due to soft and moist characteristics [19,20]. Among them, polyvinyl alcohol (PVA) hydrogels have received extensive attention as a substitute for articular cartilage because of their excellent biocompatibility, and biotribological properties [17,21,22]. However, the traditional PVA hydrogels are weak and brittle, which is difficult to meet the requirements of articular cartilage. Therefore, various methods (double or triple network structure, nano/micro materials reinforcement, annealing, etc.) have been developed to enhance the mechanical properties of the hydrogels [17,23,24]. Despite the excellent mechanical properties, outstanding lubricating performance, and biological activity are also the criteria for hydrogels as the cartilage substitute. Therefore, we should be chosen biocompatible materials as additives to promote the mechanical properties, bioactivity, and biotribological properties of PVA hydrogel. As far as we know, nano-hydroxyapatite (HA) is the main inorganic phase of bones, which can achieve chemical bonding with tissues, then stimulate or induce bone hyperplasia, and promote the repair of defective tissues. It showed that nano-hydroxyapatite (HA) with excellent biocompatibility and biological activity [25,26]. Moreover, HA as an inorganic nanofiller in the hydrogel plays a role in the enhancement of mechanical properties (nanoparticle reinforcement) and tribological properties (HA can interact with polymer chains to enhance the network structure, which can decrease the deformation and contact area of the hydrogel during the friction tests). The studies showed that the load-bearing capacity, biological activity, and tribological performance of hydrogels were improved by the addition of HA [17,27,28]. Polyacrylic acid (PAA) is a hydrophilic polymer with a largr number of carboxyl groups, which can hydrogen bonding with other polymer chains to form the interpenetrating network, showing higher mechanical properties [7,17]. Also, studies have shown that PAA can improve the surface roughness and biocompatibility of the hydrogel [29]. At present, PAA has been widely used in tissue engineering [30,31]. In summary, our hydrogel layer takes PVA, HA, and PAA as the main materials.
In addition, a substantial obstacle to coating a soft, wet hydrogel on a hard substrate is how to obtain a strong bonding between them. Most of the methods connecting hydrogel with the substrate are through physical bonding, which cannot achieve strong bonding at the interface. Recent reports illustrated that mechanical interlocking and chemical bonding have been used to obtain a tough soft-hard interface [32][33][34][35]. Kurokawa et al. [35] utilized the double-network principle to anchor the dual-network (DN) hydrogel in a porous solid substrate, which built a strong interface between a DN gel and a porous solid. To www.Springer.com/journal/40544 | Friction obtain both high mechanical properties and excellent tribological performance of materials for artificial joints, Zhou et al. [20] used laser processing and surface sulfonation to prepare porous laser and sulfonation poly(ether-ether-ketone) (LSPEEK), and then cast PVA hydrogel on the surface of LSPEEK to simulate natural cartilage-subchondral bone in structure and function successfully. Yuk et al. [36] realized the covalent immobilization of the longchain polymer network of a hard hydrogel on a non-porous solid surface through silanization of the substrate surface. The above studies have improved the binding force between the hydrogel and the hard substrate to a certain extent, but it is still a challenge to meet both the high mechanical properties of the hydrogel and the tough interface for long-term stability. Based on mussels can adhere to almost all inorganic and organic surfaces and maintain strong adhesion even in humid environments, polydopamine (PDA) was widely used to increase the adhesion of materials due to the presence of catechol [19,34,37], we hypothesized that the introduction of PDA coating on titanium alloy surface can effectively improve the binding performance of the soft-hard interface by chemical adhesion. Moreover, PDA has attracted more and more attention as an additive in hydrogels that can promote cell adhesion and provide the excellent biocompatibility to hydrogels [38][39][40]. Therefore, the introduction of PDA can not only enhance the binding performance but also improve the biocompatibility of the soft-hard integrated material.
Here, inspired by the structure of natural joints and the multifunctional adhesion characteristics of mussels, we proposed a "soft surface-hard substrate" approach to design soft-hard integrated materials for artificial joints, which can obtain both high mechanical properties and excellent tribological properties. We initially processed the Ti6Al4V alloy by laser processing and PDA surface modification. Then, PVA/HA/PAA/PDA hydrogel was cast on the Ti6Al4V by freezing-thawing technology, achieving a tough soft-hard integrated material. Mechanical interlocking was formed between the hydrogel and the Ti6Al4V substrate, and the PDA on the Ti6Al4V substrate can interact with hydrogel by chemical adhesion, effectively enhancing the binding between the hydrogel and the substrate. In addition, the tribological properties of soft-hard materials were rarely studied. However, the tribological properties are of great significance to the use of hydrogel-Ti6Al4V integrated materials as artificial joint implants. In this work, the sliding, twisting, and swinging tribological behaviors of the hydrogel-Ti6Al4V integrated material are studied by simulating the motion form of the human body, and the results showed that the material has excellent tribological properties. Nano-Hydroxyapatite (n-HA, 20 nm) was obtained from Nanjing Emperor Nano Material Co., Ltd.

Pretreatment of titanium alloy (Ti6Al4V)
All Ti6Al4V samples were polished with sandpaper, and then laser treatment was performed on the polished Ti6Al4V surface to obtain porous LTi6Al4V. The samples were ultrasonically cleaned with deionized water and absolute ethanol 3 times to remove wear debris and then dried at 50 °C for 24 h. Next, PDA was used as a coating to modify the surface of the LTi6Al4V. 0.528 g CuSO 4 ·5H 2 O, and 1.5 g DA were dissolved in 60 mL deionized water to form solution A. Solution B is a mixture of 240 mL deionized water, 120 mL ethanol, 3 mL NH 3 ·H 2 O, and 0.93 g 30% H 2 O 2 solution. Solution B was poured into solution A and stirred evenly to obtain CuSO 4 /H 2 O 2 treatment solution for soaking the LTi6Al4V to form the PDA coating. Finally, the LTi6Al4V was taken out after 2 h and ultrasonically cleaned with ethanol for 15 min to remove excess polydopamine, obtaining a PDAmodified LTi6Al4V substrate (PDA-LTi6Al4V).

Preparation of Ti6Al4V-hydrogel soft-hard integrated material
The preparation of PDA was prepared according to the previous report [41]. First, 1 wt% PDA was uniformly dispersed in the deionized water by ultrasound for 5 min. Next, 16 wt% PVA, 2 wt% HA, and 4 wt% PAA were added to the PDA solution and stirred at 95 °C for 4-5 h in a water bath to form PVA/HA/PAA/PDA sol. Finally, the sol was poured on the PDA-LTi6Al4V substrate, and subjected to freezing-thawing and annealing treatments (-20 °C to freeze for 6 h, room temperature to thaw for 2 h, repeated 9 times; annealing at 120 °C for 1 h) to obtain Ti6Al4V-Hydrogel soft-hard integrated material.

Component and morphology characterizations
The surface elementary composition of LTi6Al4V, PDA-LTi6Al4V, and Ti6Al4V substrate after tearing off the hydrogel were characterized by the X-ray photoelectron spectroscopy (XPS, ESCALAB 250Xi, USA). The microscopic morphology of Ti6Al4V-hydrogel was detected by scanning electron field emission scanning electron microscope (FESEM), Hitachi SU8220, Japan), and its element distribution was analyzed by energy dispersive spectroscopy (EDS, XflashFQ5060, Germany).

Mechanical properties measurement
The compression properties of the Ti6Al4V-Hydrogel were tested by WDW-2 microcomputer-controlled electronic universal testing machine. The samples were cubic (25 mm × 25 mm × 5 mm of Ti6Al4V substrate, and the thickness is 1 mm of hydrogel layer), and the compression test was carried out at a rate of 5 mm/min and a strain of 50%. The compressive strength is obtained by dividing maximum force (F) by the cross-sectional area (S), and the compressive modulus was defined as the slope of the stress-strain curve in the 0-10% strain region. The creep performance was determined by the UMT-Ⅱ multifunctional friction and wear tester. The Ti6Al4V-hydrogel sample was fixed in the lower clamp and the upper sample is CoCrMo ball (diameter: 28 mm) to load the Ti6Al4Vhydrogel sample. During the experiment, the creep load was set as: 5, 10, and 20 N, the creep time was 1 h, and the medium is deionized water.

Binding strength and interfacial toughness
The binding strength between the Ti6Al4V substrate and hydrogel was tested by the adhesion tester (F510-20T, Elcometer, UK). Fix the Ti6Al4V-hydrogel on the fixture, stick the forging die for the tester on the upper hydrogel and the lower Ti6Al4V substrate of Ti6Al4V-hydrogel, and then pull up the forging die in the direction perpendicular to the Ti6Al4Vhydrogel binding surface until the hydrogel was pulled off the Ti6Al4V substrate by adhesion tester. The interfacial toughness of Ti6Al4V-hydrogel was detected by the standard 90-degree peeling test. The Ti6Al4V-hydrogel materials were prepared with 50 mm in length, and 25 mm in width. In addition, half of the hydrogel layer was bonded to the substrate, while another half was separated from the substrate. The peeling speed of the 90-degree peeling test was 2 mm/min. In the peeling process, the peel force will enter a stable state (plateau force), that is, the force needed to peel off the hydrogel from the substrate. The interfacial toughness Γ was determined by dividing the plateau force F by the width of the hydrogel sheet W.

Friction test
The sliding friction was carried on the UMT-Ⅱ multifunctional friction and wear tester. The friction pair was CoGrMo ball (diameter: 28 mm), the sliding The Prosim 2 station deep flexion knee simulator was used to determine the twisting friction properties of the Ti6Al4V-hydrogel materials. The friction pair was CoGrMo ball (diameter: 28 mm), the loads are 400, 600, and 800 N, the twisting angle are 5°, 10°, 15°, and 20°, the frequency is 1 Hz, the lubricating medium is deionized water, and the twisting cycle is 2,000.
Similarly, the swing friction of the Ti6Al4V-hydrogel composite was tested by the PROSIM knee joint wear simulator. The friction pair was CoGrMo ball (diameter: 28 mm), the loads are 100, 150, and 200 N, the swing angle are 5°, 10°, 15°, and 20°, the frequency is 1 Hz, the lubricating medium is deionized water, and the swing cycle is 2,000.

Statistical analysis
In this work, three parallel samples were measured and the final data are reported as means ± SD (standard deviation). Moreover, the calculated results were analyzed by the Tukey test to determine the statistical significance (p < 0.05).

Formation of the Ti6Al4V-hydrogel and basic characterizations
The formation mechanism of Ti6Al4V-hydrogel softhard integrated material is shown in Fig. 3. Inspired by the structure and function of articular cartilagesubchondral bone, biomimetic soft-hard integrated material was designed with soft hydrogel and hard Ti6Al4V for artificial joint material. The key point in the formation of Ti6Al4V-hydrogel soft-hard integrated material is the bonding between Ti6Al4V and hydrogel, so it is essential to enhance the bonding between Ti6Al4V and hydrogel by surface modification. The surface modification of Ti6Al4V is mainly composed of two parts. Firstly, the porous structure was formed on the surface of Ti6Al4V by laser printing, and the hydroxylated surface was obtained on the Ti6Al4V substrate at the same time. Secondly, inspired by the adhesion of mussels, PDA was synthesized in situ on the LTi6Al4V surface. PDA can adhere closely to the surface of almost all materials, and the catechol groups in PDA and the -OH on the surface of the Ti6Al4V substrate could be dehydrated to form bidentate coordination bonds, which are firmly attached to the surface of LTi6Al4V to obtain PDA-LTi6Al4V.
In addition, it also can improve the mechanical properties of hydrogels as a nano-reinforcing phase and hydrogen bonding with other polymer chains to enhance the network structure. PVA/HA/PAA/PDA was used as a hydrogel layer to cover the surface of PDA-LTi6Al4V to form the soft-hard integrated structure. PVA/HA/PAA/PDA sol was poured on the PDA-LTi6Al4V surface, and then frozen-thawed to form a gel which was attributed to the formation of the hydrogen bonds and microcrystalline structure. Due to the existence of the microporous structure on the Ti6Al4V substrate surface, the hydrogel can be embedded in the micropores to produce a tight pinning effect with the substrate. Furthermore, the PDA on the Ti6Al4V surface can adhere to the hydrogel due to the multifunctional adhesion, and a large number of hydrogen bonds could be formed between the PDA (Ti6Al4V surface) and the polymer chains so that the hydrogel tightly combined with the Ti6Al4V substrate. Similarly, the PDA in the hydrogel can enhance the adhesion strength of the hydrogel on the Ti6Al4V substrate. Overall, the soft-hard integrated materials Ti6Al4V-hydrogel were successfully prepared by laser fitting, PDA adhesion, and freezing-thawing technology.
In previous work, the basic properties of PVA/ HA/PAA/PDA hydrogels were characterized, and the results showed that the porous three-dimensional network structure of PVA/HA/PAA/PDA hydrogels is comparable to the articular cartilage. Moreover, the PVA, PAA, and PDA in the PVA/HA/PAA/PDA hydrogel were well cross-linked [41]. Therefore, the surface of the titanium alloy substrate was characterized by SEM, XPS, and EDS. As shown in Fig. 4(c), a uniformly dense microporous matrix was displayed on the surface of Ti6Al4V substrate by laser printing, which can enlarge the contact area between the hydrogel and the Ti6Al4V substrate. In addition, PDA particles were uniformly and densely distributed on the surface of the substrate, which was conducive to the combination between Ti6Al4V and the hydrogel. Figures 4(a) and 4(b) showed the XPS spectra of the Ti6Al4V substrate to analyze the chemical composition www.Springer.com/journal/40544 | Friction in different treatment processes. The Ti 2p narrow scan pattern showed that the characteristic peak of Ti disappeared, while the characteristic peak of TiO 2 increased in LTi6Al4V, indicating that the surface of Ti6Al4V was oxidized to generate TiO 2 (hydroxylated surface). The characteristic peak of TiO 2 significantly weakened, and the characteristic peak of C-C enhanced on the surface of PDA-LTi6Al4V ( Fig. 4(b)), which indirectly proved that the PDA modified layer adhered to the surface of Ti6Al4V. Besides, the EDS spectrums indicated that PDA-LTi6Al4V surface has an increase in N content and a decrease in O content compared with LTi6Al4V (Fig. 4(d)). Thus, it is proved that the PDA was attached to the surface of Ti6Al4V.

Mechanical properties
The mechanical properties of the hydrogel layer are of great significance to the application of soft-hard integrated materials. The results showed that the PDA can significantly enhance the mechanical properties of Ti6Al4V-hydrogel material (Fig. 5). As shown in Fig. 5(a), the typical compression stress-strain curves (50% strain of hydrogel layer) indicated that the hydrogel layer of Ti6Al4V-hydrogel (PVA/HA/PAA/ PDA) soft-hard integrated materials has excellent mechanical properties (compressive strength ~2.63 MPa, compressive modulus ~1.56 MPa). Moreover, it is important for artificial cartilage material to endure the successive multiple loading-unloading for a short time. 5 cyclic compression loading-unloading test was performed for the Ti6Al4V-Hydrogel material at a compressive strain of 50% of the hydrogel layer. The hysteresis loops were almost coincided except for the first cycle (Figs. 5(c) and 5(d)). Taking the Ti6Al4V-hydrogel (PVA/HA/PAA/PDA) as an example, the maximum stress of the hydrogel maintained at 2.28 MPa after 5 cyclic compression tests (approximately 86.7% of the first cycle), showing that the hydrogel can recover immediately to withstand repeated loads without being damaged. At the same time, the area of the stress-strain hysteresis loop of the PVA/HA/PAA/ PDA hydrogel layer is larger than that of the PVA/HA/PAA hydrogel layer, indicating that the three-dimensional network became denser due to the

Evaluation of binding performance between hydrogel layer and Ti6Al4V
The excellent interfacial binding performance is a significant prerequisite to ensure the long-term service life and stability of the artificial joints. We found that the hydrogel was easily separated from the Ti6Al4V when the hydrogel was directly cast on the Ti6AL4V. On the contrary, the hydrogel and the Ti6Al4V exhibited excellent binding performance when the hydrogel was cast on the Ti6Al4V after laser treatment and PDA modification. It can be seen that the excellent combination between the hydrogel and the Ti6Al4V owing to the formation of the mechanical interlocking by the gel embedded into the microporous structure on the Ti6Al4V surface (Figs. 6(a) and 6(b)). The effect of PDA was measured by the binding strength and interfacial toughness.
The uniformly dense microporous matrix was distributed on the surface of the substrate that the hydrogel flowed into the laser micropores to form a mechanical interlocking effect, which can effectively enhance the binding strength between the substrate and the hydrogel [20]. Therefore, we have performed different times of laser processing on the Ti6Al4V substrate. As shown in Fig. 6(c), the interface binding strength of the soft-hard integrated material increased with the increase of the laser treatment times, but the increment of the binding strength of thrice laser treatment is small by comparing to twice laser treatment (there is no significant difference), twice laser treatment was finally selected as the optimal parameters. Similarly, the effect of PDA modification on the binding performance was evaluated by the determination of the binding strength. Figure 6(d) showed that the interface binding strength with or without PDA modification are ~ 52.7 MPa and ~ 41.2 MPa, respectively. Thus it can be seen that the chemical adhesion of the catechol element in the PDA greatly improved the binding strength.
In addition, the interfacial toughness between the substrate and the hydrogel was measured by a www.Springer.com/journal/40544 | Friction standard 90° peel test. Figure 6(e) showed the force versus displacement curve of the hydrogel peeling from Ti6Al4V substrate. The combination of the Ti6Al4V substrate without PDA modification and the hydrogel belongs to mechanical interlocking (physical combination), the hydrogel can easily peel from the interface without significant deformation of the hydrogel, so that the stress required for the peeling of the hydrogel was small, and the corresponding interface toughness was only ~480 J/m 2 . On the contrary, the combination of the PDA-modified Ti6Al4V substrate and the hydrogel belongs to the interaction of mechanical interlocking and chemical adhesion. The peeling of the hydrogel goes through  | https://mc03.manuscriptcentral.com/friction three stages. In the first stage, the hydrogel around the front of the interface crack was highly deformed as the peeling stress increased, and no separation between the hydrogel and the substrate. In the second stage, the drop in the peeling stress indicated that the interface cracks have begun to expand and the boundary edge of the hydrogel and the Ti6Al4V substrate has begun to separate. In the third stage, the peeling force reached a steady-state, and the crack spread along with the interface until it is completely peeled off [36,42]. The measured interfacial toughness was ~3,900 J/m 2 (Fig. 6(f)), which is much higher than that of the tendon-bone, cartilage-bone interface [43,44], and most hydrogel-substrate interfaces [16,19,35,36,[45][46][47][48].
To further reveal the binding mechanism between the hydrogel and the PDA-LTi6Al4V substrate, the substrate was characterized by SEM, EDS, XPS after the hydrogel was peeled off. Figure 6(g) showed the characteristic peaks of Ca and P elements in the XPS spectrum of the substrate, indicating that there was residual nHA on the surface of the substrate after the hydrogel was peeled off. The EDS spectrums showed that there was hardly any Ti element in the middle area of the microporous array on the Ti6Al4V surface after the hydrogel was peeled, while the C, O, and N elements significantly increased (Figs. 6(d) and 6(h)). In summary, it is proved that the red area in Fig. 6(h) was covered with the residual hydrogel, which indicated the excellent binding between the hydrogel the Ti6Al4V substrate. On the other hand, the morphology of the microporous area was characterized after the hydrogel layer was peeled. It can be seen from Fig. 6(c) that there were micropores with an inner diameter of about 100 μm on the surface of the Ti6Al4V. However, there was residual hydrogel in the laser micropores on the Ti6Al4V surface after the hydrogel was peeled off, indicating that the hydrogel was embedded in the laser micropores, and the laser micropores play a role in mechanical interlocking (Fig. 6(i)). In addition, Fig. 6(j) showed that the residual hydrogel in the laser micropores adheres to the inner wall of the micropores in a filamentous shape, similar to the "foot silk" of mussels. It can be inferred that the PDA will adhere to the inner wall of the micropore when the Ti6Al4V substrate was modified with PDA. The adhesion effect of PDA makes the hydrogel tightly adhere to the substrate and the inner wall of the micropores, further enhancing the binding between the hydrogel and the substrate.

Tribological properties
In the various applications of soft-hard integrated materials, its practical application as an artificial joint requires excellent tribological properties. Therefore, the tribological properties of Ti6Al4V-hydrogel softhard integrated materials were studied through sliding friction, torsional friction, and swing friction experiments by simulating the motion form of the human body.
The sliding tribological properties of Ti6Al4Vhydrogel (PVA/HA/PAA/PDA) are better than Ti6Al4Vhydrogel (PVA/HA/PAA), which indicated that the addition of PDA can improve the tribological properties of the Ti6Al4V-hydrogel materials ( Fig. 7(a)). On the one hand, the addition of PDA improves the compactness of the hydrogel network structure. The compressive strength of Ti6Al4V-hydrogel (PVA/HA/ PAA/PDA) is significantly better than Ti6Al4Vhydrogel (PVA/HA/PAA) (Fig. 7). Therefore, the deformation of the PVA/HA/PAA/PDA hydrogel layer was smaller relatively under the same load, so that the frictional resistance generated by hydrogel deformation during sliding friction was reduced, which is manifested as a lower friction coefficient (Fig. 7(b)). Moreover, the introduction of PDA also can improve the surface roughness of the hydrogel, lowering the coefficient of friction.
We further studied the effects of load and sliding speed on tribological properties. Figure 7(c) showed the average friction coefficient of Ti6Al4V-hydrogel (PVA/HA/PAA/PDA) under the load of 20, 30, and 40 N, respectively. It can be seen that as the load increased, the friction coefficient increased. When the load is 20 and 40 N, the friction coefficient is 0.065 and 0.117, respectively. The increase of the load caused greater deformation of the hydrogel, which lead to the frictional resistance generated by the deformation of the hydrogel increased during the sliding process, so the friction coefficient became larger. Also, the greater load caused the lubricant to diffuse from the surface of the hydrogel, resulting in a higher friction coefficient. In different sliding speeds (5, 20, and 35 mm/s), the www.Springer.com/journal/40544 | Friction friction coefficient first increased and then decreased, and its changing law conformed to the adsorptionrepulsion model [49] (Fig. 7(d)). The sliding friction between CoCrMo and hydrogel was mainly affected by two factors: on the one hand, it originates from the elastic deformation force generated by the stretching of the polymer chain of the hydrogel. On the other hand, it comes from the lubrication of the hydration layer between the hydrogel and the ball, and its resistance was manifested by the viscous force of the hydration layer. The friction between the hydrogel and CoCrMo manifested as elastic friction at a lower sliding speed; On the contrary, it is mainly manifested as fluid lubrication when the sliding speed is higher.
The tribological dynamic characteristics of the torsional friction interface were exhibited by the friction torque-angular displacement (T-θ) curve [50]. As shown in Figs. 8(a)-8(c), the torque value gradually increased as the cycles of torsional friction increased, and the T-θ curve changed from a regular parallelogram to a narrower parallelogram-like shape, indicating that the motion state between the hydrogel and the CoCrMo interface has changed from a completely slip state to a stick-slip mixed state. The hydrogel layer of the Ti6Al4V-hydrogel materials is a viscoelastic material with high moist content. At the initial stage, the interface between the hydrogel and the CoCrMo ball was in a completely lubricated state, showing the liquid-phase loading and the lower sliding resistance. As the cycles increases, the fluid was extruded from the interface and the fluid load support decreased, resulting in the lubrication state of the friction interface change from full slip state into stick-slip state [51,52]. Therefore, both the torque and the equivalent friction coefficient showed a rapid increase at the beginning of torsion, and it slowly increased until stable with the increase of the period.
In addition, the torque and the coefficient of friction were affected by the load. As the load increased, the friction torque increased (Figs. 8(a)-8(d)). As shown in Fig. 8(d), the torque at a load of 800 N (0.313 N/m) is much greater than that at a load of 400N (0.216 N/m). Conversely, the equivalent friction coefficient decreased as the load increased. the equivalent friction coefficient (0.0463) at a load of 800 N is much smaller than when the load is 400 N (0.0978).
The torsional friction performance of Ti6Al4Vhydrogel materials was also affected by the torsional angle. Figures 9(a)    The results showed that the friction torque and equivalent friction coefficient increased gradually with torsional angle, which results from the tangential force of the friction interface between the hydrogel and the CoCrMo ball increased with the increase of torsional angle. Moreover, with the increase of the torsional angle, the fluid flow velocity of hydrogel increase will lead to the decrease of lubrication, leading increase of friction torque and equivalent friction coefficient.
Similarly, the swing friction coefficient was affected by the load and swing angle (Fig. 10). As the load increased, the swing friction coefficient increased ( Fig. 10(a)). On the one hand, the increase of load will increase the deformation of the hydrogel, and the contact area between the CoCrMo and the hydrogel will increase, resulting in the increase of resistance in the swing process. On the other hand, the greater load caused the lubricant to diffuse from the surface of the hydrogel, resulting in a higher friction coefficient. However, the friction coefficient can still be maintained at 0.096 under the high load of 200 N. Similarly, the swing friction coefficient increased as the swing angle increased (Fig. 10(c)). The fluid loss at the surface of the hydrogel accelerated with the increase of the swing angle, which weakened the liquid phase lubrication and fluid load support of the hydrogel layer, causing the increase of the friction force during the swing process. In addition, the swing angle has a greater influence on the coefficient of friction. When the swing angle is 20°, the friction coefficient (0.365) is 3 times larger than the swing angle of 5° (0.0836). Figures 10(b) and 10(d) showed the reaction forces of Ti6Al4V-hydrogel materials in a swing cycle under different loads and swing angles, respectively. As the load increased and the swing angle increased, the reaction forces increased, which is the same as the results of the swing friction coefficient. The change curve of the frictional force in the swing direction in a cycle is similar to a sinusoidal curve. At the beginning of a cycle (0 s), the reaction force is 0 due to there no relative motion between the CoCrMo ball and hydrogel  | https://mc03.manuscriptcentral.com/friction layer in the frictional direction. Once the swing friction test begins, the reaction force increased rapidly in the negative direction until the swing angle reached the maximum (0.25 s). Subsequently, the swing friction gradually returned to the position of 0°, and the reaction force (negative) gradually decreased. When the CoCrMo ball returned to 0° (0.5 s), the reaction force in the frictional direction decreased to 0. The process mechanism of 0.5-1 s is the same as 0-0.5 s, and the reaction force direction is opposite.
Generally speaking, the friction force in the friction process is mainly derived from the shear force. According to the results of tribological properties, the maximum frictional force (shear force) of sliding friction, torsional friction and swinging friction is tens or tens of Newtons. However, according to the results of the interface toughness, the binding force between the hydrogel layer and the Ti6Al4V substrate is about 100 N, which is much larger than the frictional shear force. Therefore, the excellent binding performance between the hydrogel and the Ti6Al4V substrate can prevent the separation of the Ti6Al4V-hydrogel material during friction process. Overall, The Ti6Al4Vhydrogel materials possess ultra-high interfacial toughness and excellent tribological properties, which can be used as a candidate for artificial joints.

Conclusions
Inspired by the structure and function of natural articular cartilage, this work developed a soft-hard integrated material with an excellent tribological performance by laser printing, polydopamine (PDA) modification, and freezing-thawing, which simulated the natural cartilage-subchondral bone. The hydrogel lubricating layer was fixed on the surface of the titanium alloy by mechanical interlocking and chemical adhesion. The interface between the hydrogel and the Ti6Al4V substrate exhibited excellent binding performance that the normal bonding strength and interface toughness are as high as 52.7 MPa and 3,900 J/m 2 , respectively, which are better than the tendon-bone interface, cartilage-bone interface, and most hydrogel-substrate interfaces. In addition, the results showed that the PDA can significantly enhance the interface binding performance to obtain a tough soft-hard (hydrogel-Ti6Al4V) interface. The function of PDA to design a strong soft-hard interface makes various research directions and applications possible in the future. Moreover, the Ti6Al4V-hydrogel exhibited excellent tribological performance, which was mainly attributed to the excellent lubricating performance of the hydrogel and the high mechanical properties of the Ti6Al4V. This simple method was expected to be used in the development of medical wear-resistant materials, and has broad prospects in the field of artificial joint implants.